Gelfand, Michele J. - Moaddel, Mansoor-Values, Political Action, and Change in The Middle East and The Arab Spring-Oxford University Press (2017)

i
Values, Political Action, and Change in

the Middle East and the Arab Spring
ii
iii
Values, Political Action, and

Change in the Middle East
and the Arab Spring
Edited by
Mansoor Moaddel
Michele J. Gelfand
1
iv
1
Oxford University Press is a department of the University of Oxford. It furthers
the University’s objective of excellence in research, scholarship, and education
by publishing worldwide. Oxford is a registered trade mark of Oxford University
Press in the UK and certain other countries.
Published in the United States of America by Oxford University Press

198 Madison Avenue, New York, NY 10016, United States of America.
© Oxford University Press 2017
All rights reserved. No part of this publication may be reproduced, stored in

a retrieval system, or transmitted, in any form or by any means, without the
prior permission in writing of Oxford University Press, or as expressly permitted
by law, by license, or under terms agreed with the appropriate reproduction
rights organization. Inquiries concerning reproduction outside the scope of the
above should be sent to the Rights Department, Oxford University Press, at the
address above.
You must not circulate this work in any other form

and you must impose this same condition on any acquirer.
Library of Congress Cataloging-in-Publication Data

Names: Moaddel, Mansoor, editor. | Gelfand, Michele J., editor.
Title: Values, political action, and change in the Middle East and the Arab
Spring /edited by Mansoor Moaddel, Michele J. Gelfand.
Description: New York, NY : Oxford University Press, 2017. | Includes index.
Identifiers: LCCN 2016027913 | ISBN 9780190269098 (hardback : alk. paper)
Subjects: LCSH: Social values—Arab countries. | Social change—Arab
countries. | Arab Spring, 2010–| Protest movements—Arab countries. |
Arab countries—Politics and government—21st century. | Arab
countries—Social conditions—21st century.
Classification: LCC HN766.Z9 S6783 2017 | DDC 306.0917/4927—dc23
LC record available at https://lccn.loc.gov/2016027913
9 8 7 6 5 4 3 2 1
Printed by Sheridan Books, Inc., United States of America

v
The support of the 2010 Cairo workshop provided by the late Dr. Terrence Lyons
made this book possible. In his fond memory, we dedicate this book.
vi
vii
CONTENTS
Acknowledgments ix
Contributors xi
Introduction xiii
Mansoor Moaddel and Michele J. Gelfand
SECTION I | Changes in Values and the Arab Spring

CHAPTER 1 hanging Values in the Islamic World and the
C
West: Social Tolerance and the Arab Spring 3
Ronald F. Inglehart
CHAPTER 2 outh Perceptions and Values During the Arab

Y
Spring: Cross-national Variation and Trends 25
Mansoor Moaddel and Julie de Jong
SECTION II | Perspectives on Change: Development

and Modernization
CHAPTER 3 ational Identity Versus National Pride in the
N
Modalities of Liberal Territorial Nationalism
and Islamic Nationalism in Muslim-Majority
Countries 61
Mansoor Moaddel
CHAPTER 4 Modernization, World System, and Clash of

Civilization Perspectives in Lay Views of the
Development–Morality Nexus in the United States
and the Middle East 101
Arland Thornton, Kathryn M. Yount,
Linda Young-DeMarco, and Mansoor Moaddel
viii
CHAPTER 5 ay Accounts of “Modern” and “Traditional” Family in

L
Greater Cairo: A Test of Developmental Models
of Family Life 133
Kathryn M. Yount, Arland Thornton, Sohair Mehanna,
and Shilpa N. Patel
SECTION III | Social-Scientific

Perspectives on Collective
Action, Political Engagement, and Voting
Behavior
CHAPTER 6 he Roots of Political Activism in Six Muslim-Majority
T
Nations 171
Nancy J. Davis, Robert V. Robinson, and Tom VanHeuvelen
CHAPTER 7 he Arab Spring and Egyptian Revolution

T
Makers: Predictors of Participation 205
Mansoor Moaddel
CHAPTER 8 hange and Continuity in Arab Attitudes toward

C
Political Islam: The Impact of Political Transitions
in Tunisia and Egypt from 2011 to 2013 249
Mark Tessler
CHAPTER 9 utocratic Recidivism: Computational Models

A
of Why Revolutions Fail 271
Andrzej Nowak, Michele J. Gelfand, Wojciech Borkowski,
and Arie Kruglanski
SECTION IV | BasicMethodological Issues in the Study

of Values
CHAPTER 10 est Practices: Lessons from a Middle East Survey
B
Research Program 295
Julie de Jong and Linda Young-DeMarco
CHAPTER 11 n Analysis of Subjective Culture in the Middle

A
East: Lessons Learned from a Qualitative Research
Program 325
Janetta Lun, Michele J. Gelfand, C. Bayan Bruss,
Lily Assaad, Zeynep Aycan, Munqith M. Dagher, and
Abdel-Hamid Abdel-Latif
Index 351
viii | CONTENTS
ix
ACKNOWLEDGMENTS
W e thank the US Air Force Office of Scientific Research for funding a

workshop in Cairo, Egypt, in May 2010, where many authors of this
volume convened to discuss theory and research on the causes and con-
sequences of human values and value change in the Middle East and North
Africa. Special thanks goes out to Dr. Ivy Estabrooke, Gary Kollmorgen, and
Ken Boff for their support; and Abby Gross, Courtney McCarroll, and Kripa
Guruprasad of Oxford University Press for their remarkable work in bringing
this book to print.
Mansoor Moaddel also thanks the National Science Foundation, the Office
of Naval Research, Africa Command of the US Military, MITRE, the Society for
the Scientific Study of Religion, Göttingen University and Max Planck Institute
(Germany), the University of Michigan, and Eastern Michigan University for
their financial contribution in support of data collection, and the Office of Naval
Research (N00014-14-1-0579, N00014-09-1-0985) for its support of this book.
Michele J. Gelfand also thanks the US Army Research Laboratory and the
US Army Research Office (grant W911NF-08-1-0144, US Federal Bureau of
Investigation contract J-FBI-10-009 [Subaward 2014054568]), and the Office
of Naval Research (grant 019183-001) for their support of this book.
x
xi
CONTRIBUTORS
Abdel-Hamid Abdel-Latif Nancy J. Davis

The Egyptian Research and Department of Sociology
Training Center DePauw University
Cairo, Egypt Greencastle, IN
Lily Assaad Julie de Jong
Department of Psychology Survey Research Center, Institute for
University of Maryland Social Research
College Park, MD University of Michigan
Zeynep Aycan Ann Arbor, MI
Departments of Psychology and Michele J. Gelfand
Management Department of Psychology
Koc University University of Maryland
Istanbul, Turkey College Park, MD
Wojciech Borkowski Ronald F. Inglehart
Institute for Social Studies Department of Political Science
University of Warsaw University of Michigan
Warsaw, Poland Ann Arbor, MI
C. Bayan Bruss Arie Kruglanski
Accenture Department of Psychology
Washington, DC University of Maryland
Munqith M. Dagher College Park, MD
The Independent Institute for Janetta Lun
Administration and Civil Society National Institutes of Health
Studies Bethesda, MD
Baghdad, Iraq
xii
Sohair Mehanna Mark Tessler

Social Research Center Department of Political Science
The American University in Cairo University of Michigan
Cairo, Egypt Ann Arbor, MI
Mansoor Moaddel Arland Thornton
Department of Sociology Department of Sociology
University of Maryland Population Studies Center, Survey
College Park, MD Research Center
Andrzej Nowak University of Michigan
Department of Psychology Ann Arbor, MI
University of Social Sciences and Tom VanHeuvelen
Humanities Department of Sociology
Warsaw, Poland University of Illinois
and Urbana, IL
Department of Psychology Linda Young-DeMarco
Florida Atlantic University Institute for Social Research
Boca Raton, FL University of Michigan
Shilpa N. Patel Ann Arbor, MI
Department of Behavioral Sciences Kathryn M. Yount
and Health Education Hubert Department of Global Health
School of Public Health, Emory and Department of Sociology
University Emory University
Atlanta, GA Atlanta, GA
Robert V. Robinson
Department of Sociology
Indiana University
Bloomington, IN
xii | CONTRIBUTORS
xiii
INTRODUCTION
Mansoor Moaddel and Michele J. Gelfand
T he anti-authoritarian movements known as the Arab Spring signify as one

of the most remarkable series of events in the Middle East and North Africa
in recent decades. They have also been distinctive in the manner in which they
emerged spontaneously and the rapidity with which they managed to unseat
entrenched dictators. It was known that the current rulers of Middle Eastern and
North African countries were unpopular and that significant sections of the subject
populations were dissatisfied. Nonetheless, no one knew that the self-immolation
of one Tunisian street vendor in December 2010, who protested the confiscation
of his wares and the humiliation that he felt was inflicted on him by the municipal
agents, would create such a powerful uproar that forced President Zein al-Abedin
Ben Ali to flee the country. Nor was it fathomable that the Tunisian action would
inspire and galvanize discontented people in other countries, leading to the over-
throw of the ruling regimes in Egypt, Libya, and Yemen, and to a civil war in Syria.
Equally unexpected have been the consequences of these movements.
Although they all started as predominantly peaceful and pragmatic collective
actions for democracy and economic prosperity (Moaddel, 2013), they have thus
far produced diverse and, in most cases, undesirable outcomes. Tunisia is the
only country that appears to be making a steady transition to a stable democratic
regime. In Egypt, on the other hand, the worsening economic conditions coupled
with the authoritarian and exclusivist presidency of Muslim Brother Mohammed
Morsi (June 30, 2012–July 3, 2013) produced mass demonstrations throughout the
country demanding his resignation. Emboldened by popular support, the military
removed Morsi from office. However, General Abdel Fattah el-Sisi, Egyptian army
chief turned elected president in 2014, has thus far failed to curb the mounting
political violence launched by Muslim terrorists in that country. And it is not quite
clear whether he will be able to renew and reconstitute the tradition of praetorian
rule practiced by the former regime. Yemen’s prospect for political stability and
xiv
transparent government also gave way to sectarian strife between the Shia Houthi
rebels in the north (backed by Iran’s Islamic regime) and the forces connected to
former president Ali Abdullah Saleh on the one hand, and the Sunni groups who
were loyal to the government of Abd Rabbuh Mansur Hadi, supported militarily
by a Saudi-led coalition of Persian Gulf Arab states on the other. Similarly, Libya’s
transition to a stable, functioning government has been hindered by the presence
of armed militias that divided the country into semi-autonomous regions. The
most tragic outcome is the situation in Syria, where the initial peaceful protests
against the authoritarian regime of Bashar al-Assad turned deadly and the country
was thrown into a vicious cycle of violence between the regime forces, supported
by Russia and the Islamic Republic of Iran, and Sunni extremist groups, leading
to the loss of more than 400,000 lives and millions of refugees. Further compli-
cating the political situation in the region is the formation of the Islamic State of
Iraq and the Levant (ISIL; or Daesh, which is the Arabic acronym for the group),
one of the most horrific terror groups in recent memory.
These developments, naturally, have turned the euphoria of the Arab Spring
into despair and prompted many to conclude that life under the former repres-
sive regimes in Egypt, Iraq, Libya, and Yemen was more bearable than the cur-
rent worsening economic conditions and the chaos of recurrent political violence.
Nevertheless, considering a broader historical perspective, the pattern of inter-
necine conflict and the scale of human tragedy experienced in the Middle East
and North Africa since the beginning of the Arab Spring in 2011 are not totally
unusual. Major political transformations in human society during different his-
torical periods have often transpired after a period of intense contention for power
among warring individuals and groups. The most recent historical experience of
sectarian violence and religious bigotry in the region was the 1860 civil war in
Lebanon and the massacre of Christians in Damascus, Syria, that same year, dur-
ing which tens of thousands perished. And like the current situation in Iraq and
Syria, the 1860 ethnic strife was the culmination of decades of tensions among
different groups, which in that case involved the Muslim Druze and the Maronite
Christians, following the breakup of their three-century-long alliance (1514–1840)
within the ambit of Ottoman rule. It also occurred against the backdrop of the
Ottomans’ decline, the rise of European powers, and the divide-and-rule policy
of Egyptian Ibrahim Pasha in Syria during the 1830s (Agoston & Masters, 2009;
Makdisi, 2000; Phares, 1995; Salibi, 1977, 1988; Spagnolo, 1977). There is, how-
ever, little comfort in the knowledge that the current human tragedy and mass
suffering are not unique in the contemporary history of the Middle East, North
Africa, and beyond.
The unpredictable manner in which the movements of the Arab Spring
emerged as well as the recurrence of political violence in the Middle East and
North Africa are certainly posing serious challenges for the extant social–scientific
theories to explain the causes and processes of change, conflict, and peace in the
region. Key questions addressed in this book include the following: Who are the main
actors in these processes? What forces constrain, limit, or empower these actors?
What are the issues being contended? To what extent do recent events reflect
xiv | INTRODUCTION
xv
changes in values among the Middle Eastern and North African public? Is the pro-
cess of change identical across different age groups? Are the youth more support-
ive uniformly of change than the rest of the population? In what areas of human
concerns does the younger generation differ from the older? To what extent are
changes in values connected to changes in identities—be they religious, national,
or ethnic? Which factors explain people’s participation in collective actions for
change? Which types of individuals or groups are more likely to participate in
such actions? Why would the same individuals who risked their lives to overthrow
their despotic government later embrace new autocratic regimes with great fer-
vor? How can we predict different scenarios following the collapse of autocracy
and understand diverse outcomes in countries such as Egypt, Syria, and Tunisia?
More generally, what are the key variables that contribute to changes in value
orientation? Do people’s conceptions of change and development conform to
a traditional–modern duality? How do people perceive the connection between
developmental change and morality? What are the major elements that contribute
to changes in people’s orientation toward the relationship between religion and
politics, and particularly the orientation toward political Islam? What are the best
methodological tools for understanding changes in values?
Part of the difficulties faced by researchers interested in understanding and
explaining the trajectory of change lie in the absence of plausible macro theoreti-
cal frameworks. The macro sociological theories used by social scientists living
during the end of World War II and the breakdown of the former Soviet Union
to explain events transpiring in the Middle East and North Africa are no longer
adequate or useful to understand the Arab Spring and its aftermath. The emer-
gence of major rebellions and revolutionary changes was thought to be caused
by tensions between the traditional order and modernity in the modernization
perspective, class conflicts and/or class–state relations in various Marxian or neo-
Marxian frameworks, and the economic dynamics of the hierarchically organized
zones of the world capitalist system. Parallel to this theoretical simplicity was the
international-relations-made-simple image provided during that historical episode
by the division of the world into the capitalist camp dominated by Western democ-
racies and the socialist camp ruled by the Soviet totalitarian system. However, the
breakdown of the former Soviet Union and the traditional consensus in social the-
ory almost simultaneously not only revealed the complexity of world politics, but
also the futility of formulating an all-encompassing discourse that would purport-
edly explain historically significant changes in different societies in terms of a set
of social dynamics common to all cases. Before theorizing about the phenomenon
of change, it is thus necessary for social scientists, first, to collect empirical data,
and then try to establish patterns of stability and change in the empirical reality.
The dearth of empirical evidence is thus the most serious difficulty that research-
ers in the area of the Middle East and North Africa have faced.
For sure, since the turn of the 21st century, there has been an impressive expan-
sion of empirical—qualitative, historical, and quantitative—research on various
aspects of Middle Eastern and North African societies. A substantial part of this
expansion is driven by the interests of Western governments and policymakers
Introduction | xv
xvi
in figuring out the causes and processes of political violence and terrorism. Of
considerable significance is the use of survey research techniques as an important
tool for data collection and concomitant improvements in data quality. The sur-
veys carried out in the region have covered major areas of human concerns—from
family to religion, to culture, to morality, to politics, to economy, and to gender
relations. This expanding scientific empirical infrastructure has made it possible
to bring the attitudes, value orientations, and long-term concerns of the people of
the Middle East and North Africa into mainstream public opinion research and
the social–scientific study of values, mass belief systems, and political engage-
ments. Complementing this explosion of survey research is the use of other
diverse methodologies, including qualitative methods, experiments, and even
computer simulations.
The chapters in this edited volume attempt to contribute to a more effective
social–scientific understanding of the changes transpiring in some of the coun-
tries in the Middle East and North Africa. Change is the central theme of this
edited volume; the authors advance new frameworks and data analysis to help
understand the dynamics of changes occurring in the region. To this end, they
have gone beyond traditional explanations of change such as those drawn on mod-
ernization and world system theories or the deductive text-based perspective of
the Islamicists. Rather, these authors formulate analytical models that consider
carefully people’s actual life situations as the sites where changes in values and
collective political action transpire. Some of these authors assess the predictive
power of modernization, the world system, and clashes of civilization theories.
Furthermore, they draw on the available empirical data and identify specific mech-
anisms that explain change across different Middle Eastern societies. They also
use theories of collective action in explaining cross-national variation in political
engagement in the Middle East, while at the same time consider variables specific
to particular Middle Eastern cultures. Other authors assess and test theses from
the Islamicist perspective to explain changes in the public’s attitudes toward politi-
cal Islam. And still others use state-of-the-art computational modeling techniques
to forecast different scenarios following the collapse of autocratic regimes.
Collectively, the chapters cover many aspects of the changes transpiring in
the region. These chapters are organized into four sections. Section I, “Changes
in Values and the Arab Spring,” provides an overview and analysis of cross-
national variation, variation by age groups, and trend in values. Section II focuses
on “Perspectives on Change: Development and Modernization.” Because the
problem of development and modernization has been addressed by intellectual
leaders, political activists, and policy makers in the region, the chapters in this
section assess (a) the type of modalities these individuals have employed as the
most appropriate framework for change and modernization of their societies,
and (b) how the public at large understood the meanings of such concepts as
development, modern, and traditional as well as the relationship between devel-
opment and morality. Section III on “Social-Scientific Perspectives on Collective
Action, Political Engagement, and Voting Behavior” consist of four chapters that
address the causes or predictors of different forms of collective action and political
xvi | INTRODUCTION
xvii
engagement, both at macro (country) and micro (individual) levels, focusing a vari-
ety of sociological and social-psychological factors. The final chapter in this sec-
tion coins the term autocratic recidivism and implements computer simulations
to explain the meso-institutional processes that explain the failure of many revolu-
tionary movement to produce a democratic form of government. Finally, Section IV
consists of two chapters that address the methodological issues and problems in
carrying out comparative cross-national empirical research in the Middle East and
North Africa.
Section I in the volume begins with Ronald Inglehart’s chapter, “Changing
Values in the Islamic World and the West: Social Tolerance and the Arab
Spring.” In this chapter, Inglehart advances a revision of the classic modern-
ization theory and illustrates how traditional approaches that related values
change to processes such as commercialization, industrialization, and other
measures of economic development are inadequate to explain the changes hap-
pening today. He argues these processes may lead to a shift in values from
materialism to postmaterialist self-expressive values when they contribute to a
change in people’s perception of security. He distinguishes between existential
security, a social condition in which people take their security for granted, and
existential insecurity, where security is the major concern. Even the creation
of existential security, he argues further, may not produce a change in values
automatically, because the older people who experienced existential insecurity
in the past may still cling to materialistic values. Thus, according to Inglehart,
change occurs among those experiencing a relatively high level of existential
security during their impressionable years. Applying this framework to the case
of the Middle East and North Africa, Inglehart notes significant differences
between the younger and older generations in terms of self-expressive values.
He concludes that the Middle Eastern and North African public is in the early
stage of a major values change.
Mansoor Moaddel and Julie de Jong’s chapter, “Youth Perceptions and Values
during the Arab Spring: Cross-national Variation and Trends in Values,” analyzes
data from cross-national and longitudinal surveys to assess cross-national variation
and trends in values among the younger and older age groups in several Middle
Eastern and North African countries. They conceptualize and measure social indi-
vidualism, gender equality, secular politics, liberal values, and identity, and eval-
uate variations in these values. Findings from their comparative cross-national
analysis of data from Egypt, Iraq, Lebanon, Pakistan, Saudi Arabia, Tunisia, and
Turkey from 2011 to 2013, as well as longitudinal surveys from Egypt, Iraq, Saudi
Arabia, and Turkey, show (a) the Arab Spring represented a dramatic change in
values toward nationalism and a decline of political Islam in the region; (b) these
changes were event driven and not simply a reflection of economic development;
(c) the changes in value orientations of youths across these countries does not war-
rant the conclusion that youths were the harbingers of the Arab Spring, although
younger people tended to be more liberal than older people across the seven coun-
tries; and (d) national context is important in explaining intergenerational dif-
ferences in values (for example, the difference in value orientation between the
Introduction | xvii
xviii
younger and older age groups is wider among Pakistani and more narrow among
Egyptian respondents compared with the other countries, and Turkey displayed
the most inconsistent relationships between age groups and value orientations of
the seven countries).
The next three chapters in Section II offer different modalities and perspec-
tives on change. In Chapter 3, “National Identity Versus National Pride in the
Modalities of Liberal Territorial Nationalism and Islamic Nationalism in Muslim-
Majority Countries,” Moaddel illustrates varying relations of identity and the feel-
ing of national pride with people’s value orientations across Muslim-majority
countries. While advocating this perspective, he argues that extant theories of
nationalism cannot account for historical and cross-national variations of nation-
alist movements. These theories, he argues, “are formulated in terms that are too
general to account for this diversity. They refer to such processes as industrial-
ization, the expansion of mass education, print capitalism, and uneven capitalist
development in different theories within the modernist tradition; to substantialist
accounts of the rise of a nation in various versions of the primordialism perspective;
or to cycles of nationalism in varied perennial interpretations.” As an alternative,
Moaddel proposes the concept of modalities to manage and classify diverse forms
of nationalism into homogeneous types. Modalities are distinguished by identity
and are associated with different feelings of collective solidarity. He also argues
that nationalism—far from reflecting the conditions of modernity, the primordial
attachments arising from group life, or perennial features of a historical cycle—is
produced by intellectual leaders as they try to resolve issues related to the nature
of their political community, its boundaries, and the basis of identity. Because
issues are resolved in oppositional relations to the ideological targets these intel-
lectual leaders encountered, variation in these targets produces different modali-
ties of nationalism. Moaddel then explains the production of the modalities of
territorial nationalism, pan-Arab nationalism, and religious nationalism in 20th-
century Middle East and North Africa in oppositional relations to the ideology of
monarchical absolutism, ulama obstructionism, and foreign occupation; colonial
partitioning of the Arab territories into disparate states after World War I; and the
secularist discourses of an interventionist state, respectively. Finally, he focuses on
liberal territorial nationalism and Islamic nationalism, and reasons that because
liberals are more secular, favorable toward gender equality, and less hostile toward
outgroups and Islamic nationalist just the opposite, these modalities exist in peo-
ple’s perception insofar as identity and national pride are linked to these values
in a consistent manner. Using the data from cross-national surveys carried out in
the region from 2000 to 2008, he tests these linkages by showing a shift in the
basis of identity from religion to territorial nation is connected to a significant
increase in favorable attitudes toward gender equality, secular orientation, secular
politics, Western culture, and, except in two cases, democracy, but is not linked
consistently to attitudes toward outsiders. National pride, on the other hand, tends
to reinforce the values supported by Islamic nationalism.
In Chapter 4, Arland Thornton, Mansoor Moaddel, Kathryn Yount, and
Linda Young-DeMarco analyze the association between changes in morality and
xviii | INTRODUCTION
xix
economic development in the perceptions of Middle Eastern and American peo-

ple. This analysis draws on Thornton’s larger project on developmental idealism
in which (a) development is considered to be desirable; (b) development is associ-
ated with certain familial, gender, political, religious, and cultural attributes; and
(c) these features are considered the causes and consequences of development.
Development in developmental idealism, for example, promotes and is promoted
by the rise of the small family, gender equality, and secular politics; the decline
of organized religion; and the rise of individualistic values. In his project, Arland
assessed the role of developmental perspective in projecting a Western-centric
evolutionary model of historical change and in shaping public policies.
In this chapter, Arland and colleagues address the relationship between
development and morality by assessing whether a more developed society is
perceived to be associated with a higher standard of morality. This assessment
is done in terms of how the ordinary public connects development to moral-
ity. The authors consider the modernization, world system, and clashes of
civilization theories; draw hypotheses from these theories on the development–
morality nexus; and evaluate the degree of affinities of the public’s perceptions
with these varied theories. Their data analysis shows that most Americans per-
ceive the association between development and morality to be positive, sug-
gesting views that are consistent with modernization theory. The data from
three Middle Eastern countries—Egypt, Lebanon, and Saudi Arabia—however,
reveal extensive heterogeneity of viewpoints: some consistent with moderniza-
tion theory, others with world system theory, the third with clashes of moral-
ity (as a revised version of the clashes of civilization theory), and the fourth
with a combination of the three perspectives. These findings are important in
today’s globalized world, because perceptions of development–morality rela-
tionships prevalent in diverse cultures shape national priorities and interna-
tional relations.
The fourth chapter in this section, Chapter 5, by Kathryn Yount, Arland
Thornton, Sohair Mehanna, and Shipa Patel, focuses on the family and evalu-
ates Egyptians’ understanding and use of developmental models to conceptualize
this important social institution. The authors explore the perception of Egyptians
with regard to the distinction between traditional family and modern family, and
whether they follow the schema of modernization scholars and practitioners in
defining traditional families as undesirable or backward, and in carrying attributes
such as extended households, arranged marriage, young age at marriage, uncon-
trolled and high fertility, and gender inequality on the one hand, and modern
families as desirable or civilized, and displaying other attributes such as nuclear
households, love marriages, older age at marriage, controlled and low fertility, and
gender equality, on the other. Yount and colleagues use a sample of 84 informants
who were asked to list all the attributes they associate with traditional families and
modern families. The findings from analyzing the data offer insights into how
Egyptians have received, resisted, and adapted this model to resonate with their
local history, society, and culture.
Introduction | xix
xx
Section III includes four chapters that provide rigorous analyses of the criti-
cal domain of political action. Chapter 6, by Nancy Davis, Robert Robinson,
and Tom VanHuevelen, analyzes the roots of political activism in six Muslim-
majority nations: Algeria, Bangladesh, Egypt, Indonesia, Jordan, and Pakistan.
Drawing on theories of collective action, the authors assess the linkages of
macro and micro sociological variables in shaping political engagement in
these six countries. Their findings indicate the most politically engaged citi-
zens are male, young or old (vs. middle-aged), have no/few children, are highly
educated (but often with a low income), are embedded in social networks, have
a strong interest in politics, and live in countries with greater foreign invest-
ment and with low or high (vs. middle) levels of repression. Contrary to popular
perception, their analysis shows that poverty and inequality do not increase
political engagement. By assessing how differential recruitments to political
action are linked to such factors as relative deprivation, social dislocation, social
networks, socioeconomic status, regime repressiveness, and economic depen-
dency, Davis and colleagues not only contribute to a fuller understanding of
politics in the social context of Muslim-majority countries, but also propound
sociological theories of collective action.
In Chapter 7, “The Arab Spring and Egyptian Revolution Makers: Predictors
of Participation,” Moaddel also focuses on political engagement by analyzing
self-report participation by Egyptians with regard to the political movement that
toppled President Hosni Mubarak. He juxtaposes two clusters of theories, rep-
resenting two diverse traditions in sociological theories of collective action and
revolution. One consists of political conflict, resource mobilization, and orga-
nizational and political opportunity theories; the other includes mass society,
structural–functional, and relative deprivation theories. He assesses their explana-
tory power in predicting participation in the Egyptian revolution of 2011, analyzing
data from a nationally representative sample of 3,143 Egyptian adults. The study
identifies three sets of variables linked to participation: (a) mobilizing ideas and
attitudes, efficacy, and moral flexibility; (b) media of communicative power (the
Internet, cell phones, newspapers); and (c) demographics, including being male,
residing in an urban area, and living impressionable years during the presidency
of Mubarak. That socioeconomic status has an inverted-U relationship with par-
ticipation suggests that the revolution was made by members of the middle class.
The data, however, provide support for some of the hypotheses drawn from both
clusters of theories. The analysis thus suggests rethinking the predictors of par-
ticipation, which entails departing from the conception that presumes the partici-
pants as monolithic individuals rather than manifold and heterogeneous, taking
a new look at the relationship between morality and participation, and refocusing
on the monolithic state as the unifying element in the revolutionary process.
In Chapter 8, “Change and Continuity in Arab Attitudes toward Political
Islam: The Impact of Political Transitions in Tunisia and Egypt from 2011 to
2013,” Mark Tessler discusses the impact of political transition in Tunisia and
Egypt between 2011 and 2013. He takes issue with the Islamicists’ precept that
relates public support for political Islam to Islamic theory of government, which
xx | INTRODUCTION
xxi
purportedly does not recognize the separation of religion and politics. He argues
that Islamic political parties in Egypt and Tunisia have, in the past, drawn sup-
port from “strategic” voters who do not favor an Islamic platform. Analyzing data
from Arab barometer surveys, Tessler shows that recent experience with Islamic
governance has increased and intensified opposition to political Islam among the
general public in the two countries. He concludes that the rise or decline in sup-
port for political Islam is a function of the existing political conditions and not the
alleged Islamic teachings on politics.
In Chapter 9, “Autocratic Recidivism: Computational Models of Why
Revolutions Fail” Andrzej Nowak, Michele J. Gelfand, Wojciech Borkowski, and
Arie Kruglanski introduce an innovative construct—autocratic recidivism—and
develop a theory of why groups return to autocracy after participating in concerted
efforts to overthrow autocratic regimes. The authors observe that, throughout the
past century, a similar scenario repeats itself in different parts of the world: auto-
cratic regimes are first brought down by a popular uprising, followed by a short
period of political openness, when diverse contenders for power compete freely
for the control of the government. In many cases, these attempts at establishing
democracy fail and the government is replaced either by the formation of a similar
autocratic government or by a brand new and even more repressive regime—
a phenomenon they term autocratic recidivism. In their chapter, Nowak and col-
leagues explain such pendulum shifts. Adopting a dynamic systems perspective
(Nowak & Vallacher, 1998), they argue that the capacity to synchronize actions is
critical to efficient functioning of society, and they examine the notion that meso-
level institutions provide the conditions that allow societies to achieve and main-
tain synchronization during a transition to democracy. The lack or weakness of
meso-level institutions weakens synchronization, which may in turn result in fail-
ure to meet basic needs, frustrate attempts to establish democracy, and facilitate a
return to autocracy. Nowak and colleagues implement computer simulations that
illustrate some initial support of their theory. Their models also suggest reasons
why some nations will return to another form of autocracy versus total chaos and
civil war, as in the case of Egypt and Syria, respectively.
The two chapters in the final section of this volume address two diverse meth-
odological orientations to acquire a better understanding of Middle Eastern and
North African countries. Chapter 10, by Julie de Jong and Linda Young-DeMarco,
discusses the problems involved in carrying out comparative cross-national sur-
veys in the Middle East and North Africa, including questionnaire construction,
interviewer training, and sampling procedures in comparative survey research,
particularly in the context of the contemporary Middle East. The authors note
that, during the past several decades, the practice of administering multicoun-
try surveys to collect data to acquire comparative data on a wide array of topics
has become ubiquitous. Researchers generally believe that if the research design
process is developed with attention to standardization, the resulting data are
appropriate for use in both inter-and intracountry comparisons. Yet, despite a
researcher’s best attempts to achieve standardization, the reliability and validity
of comparisons may remain questionable as a result of differences in cultural
Introduction | xxi
xxii
sensitivities, infrastructure, census data availability, and survey methods training.

Although each region of the world poses its own unique set of challenges, for
many researchers the Middle East has proved to be a particularly difficult place in
which to achieve cross-national comparative research objectives. In this chapter,
the authors provide an overview of protocols that are the most critical in designing
and administering a multicountry survey for data collection in the Middle Eastern
context, while at the same time remaining cognizant of the requirement of as
much standardization as possible across countries. When discussing strategies
to ameliorate many of the potential challenges, they draw on issues we face in an
ambitious cross-national comparative surveys across six Middle Eastern countries.
Last, but not least, is Chapter 11 by Janetta Lun and colleagues titled “An
Analysis of Subjective Culture in the Middle East: Lessons Learned from a
Qualitative Research Program.” The authors argue that to understand values and
value change, survey methods need to be complemented with qualitative methods
that enable researchers to understand how values are understood subjectively by
citizens in the region. The authors describe their own work involving 400 in-
depth interviews in Afghanistan, Egypt, Iraq, Jordan, Lebanon, Pakistan, and
Turkey that explored values such as honor, wasta, fatalism, and modesty, among
others. They describe the strategies, processes, and stages of analyzing large-scale,
qualitative cross-cultural research, and they pinpoint issues and potential solu-
tions during each stage of the research process (e.g., translations, development
of coding schemes, coding reliability). They end with a summary of practical and
theoretical considerations for conducting cross-cultural qualitative research in the
Middle East. Both chapters provide critical advice based on extensive research on
how to conduct high-quality research in the region.
All in all, in the chapters that follow, the authors offer different analytical
frameworks to assess the causes and processes of changes in values and politi-
cal engagement in Middle Eastern and North African countries, focusing on the
effects of economic development on existential security, expressive values, and
social tolerance; cross-national variation and trend in values; the factors shaping
political engagement, changes in voting behavior, and orientation toward political
Islam, and authoritarian recidivism and the failure of democracy; modalities of
nationalism, development and morality in people’s perceptions, and perceptions
of the modern and traditional family; and methodological issues in comparative
cross-national values surveys and qualitative research.
Our hope is that this collective effort not only contributes to the development of
the social sciences in the Middle East and North Africa, but also suggests practical
political actions and public policies that serve social tolerance and harmony, peace,
and economic prosperity for the people of the region.
References
Agoston, G., & Masters, B. A. (2009). Encyclopedia of the Ottoman Empire. New York:
Infobase Publishing.
xxii | INTRODUCTION
xxiii
Makdisi, U. (2000). The culture of sectarianism: Community, history, and violence in

nineteenth-century Ottoman Lebanon. Berkeley: University of California Press.
Moaddel, M. (2013). A report: The birthplace of the Arab Spring: Values and perceptions of
Tunisians [On-line]. Available: www.mevs.org.
Nowak, A., & Vallacher, R. R. (1998). Dynamical social psychology (Vol. 647). Guilford
Press.
Phares, W. (1995). Lebanese Christian nationalism: The rise and fall of an ethnic resis-
tance. London: Lynne Rienner.
Salibi, K. (1977). The Modern History of Lebanon. New York: Caravan Books.
———. (1988). A House of Many Mansions: The History of Lebanon Reconsidered.
Berkeley: University of California Press.
Spagnolo, J. P. (1977). France and Ottoman Lebanon. London: Ithaca Press.
Introduction | xxiii
xxiv
1
SECTION I Changes in Values and

the Arab Spring
2
3
CHAPTER 1 Changing Values in the Islamic World

and the West
Social Tolerance and the Arab Spring
Ronald F. Inglehart
About 45 years ago, I suggested “a transformation may be taking place in the

political culture of advanced industrial societies. This transformation seems to be
altering the basic value priorities of given generations as a result of changing con-
ditions influencing their basic socialization” (Inglehart, 1971, p. 991). This chapter
traces the evolution of values in Western countries since 1970, and examines to
what extent similar value changes are transforming other countries today, with
special attention to Muslim-majority countries. As Inglehart and Welzel (2005)
demonstrated, the rise of postmaterialist values is part of a broader set of cultural
changes that tend to bring democratization. But, although the recent Arab Spring
seemed to be installing democratic regimes across the Arab world initially, it met
an early frost. In the 2014 Freedom House ratings, not a single Muslim-majority
country was coded as “Free”—meaning, the country has sufficiently high levels of
political rights and civil liberties to be considered a democracy—although a clear
majority of non-Muslim countries were ranked as “Free.”
The initial evidence of intergenerational value change came from surveys car-
ried out in six West European societies. They revealed large differences between
the value priorities of older and younger generations. Among the older cohorts,
materialist values, emphasizing economic and physical security, were overwhelm-
ingly predominant; however, as one moved from older to younger birth cohorts,
postmaterialist values, emphasizing autonomy and self-expression, became increas-
ingly widespread. The differences were striking. Among those age 65 years or
older, materialists were fully 12 times as numerous as postmaterialists; among
those born after World War II (who were younger than 25 years in 1970), postma-
terialists were slightly more numerous than materialists.
If, as I argued, these age differences reflected intergenerational value change
(and not simply a tendency for people to become more materialist as they aged),
then we should expect to find a gradual shift from materialist to postmaterialist
values as younger birth cohorts replaced older ones in the adult population. The
implications were far-reaching, for these values were linked closely with a number
of important orientations, ranging from emphasis on political participation and
4
freedom of expression, to support for environmental protection, gender equality,

and democratic political institutions.
The value change thesis was controversial from the start. Various critics argued
that the large age differences found in 1970 reflected life cycle effects rather than
intergenerational change (e.g., Boeltken & Jagodzinski, 1985). Younger people had
an inherent tendency to prefer postmaterialist values such as participation and
free speech; but, as they matured, they would come to have the same solid mate-
rialist preferences as their elders, so we would find no change in the values of the
society as a whole.
The value change hypothesis, in contrast, holds that young people are not nec-
essarily more postmaterialist than their elders. This happens only if they have
grown up under substantially more secure living conditions than their elders. We
would not expect to find intergenerational value differences in stagnant societies.
And, if future generations no longer grew up under more secure conditions than
their elders, we would no longer find intergenerational value differences even
in western Europe. On the other hand, these intergenerational differences exist
because the degree of security or insecurity that one experiences during one’s
preadult years has a lasting impact. Consequently, as postwar birth cohorts replace
older ones in the adult population, we witness a shift from materialist to postma-
terialist values in these societies.
Intergenerational value change, by its very nature, moves slowly, but its long-
term impact can be profound. More than four decades have passed since the
hypothesized shift from materialist to postmaterialist values was published. Have
the predicted changes actually taken place?
The shift toward postmaterialist values, I argue, is driven by changing exis-
tential conditions—above all, the change from growing up with the feeling that
survival is precarious, to growing up with the feeling that survival can be taken
for granted. Throughout most of history, survival has been uncertain for the vast
majority of the population. But, the remarkable economic growth that occurred
during the era following World War II, together with the rise of the welfare state,
brought fundamentally new conditions in advanced industrial societies. The post-
war birth cohorts spent their formative years under levels of prosperity that were
unprecedented in human history, and the postwar welfare states reinforced the
feeling that survival was secure. Finally, although the older birth cohorts had expe-
rienced the life-threatening events of World War I and World War II, the postwar
birth cohorts grew up in a new era. Since 1945, the world has experienced the
longest period in recorded history without war between major powers. There were
huge differences between the formative experiences of the postwar birth cohorts
and all the older cohorts, producing major differences in their value priorities.
These differences started to become obvious when a postwar birth cohort emerged
into political relevance two decades after World War II, contributing to the era of
student protest during the late 1960s and 1970s. A widespread slogan among the
protesters was “Don’t trust anyone over 30!”
As we will see, a massive body of evidence demonstrates that an intergenera-
tional shift from materialist to postmaterialist priorities has been occurring. But, it
is only one aspect of a broader cultural shift from survival values to self-expression
4 | Changes in Values and the Arab Spring

5
values, which is bringing new political issues to center stage and encouraging the
spread of democracy.
This theory of intergenerational value change is based on two key hypotheses
(Inglehart, 1977):
1.
A scarcity hypothesis: Virtually everyone aspires to freedom and autonomy,
but people tend to place the highest value on their most pressing needs.
Material sustenance and physical security are linked immediately with sur-
vival, and when they are scarce, people give top priority to these materialistic
goals. However, under prosperous conditions, people become more likely to
emphasize postmaterialist goals such as belonging, esteem, and aesthetic
and intellectual satisfaction.
2.
A socialization hypothesis: The relationship between material conditions
and value priorities is not one of immediate adjustment. To a large extent,
one’s basic values reflect the conditions that prevailed during one’s preadult
years, and these values change mainly through intergenerational popula-
tion replacement.
The scarcity hypothesis is similar to the principle of diminishing marginal utility.
It reflects the basic distinction between the material needs for physical survival
and safety, and nonmaterial needs such as those for self-expression and aesthetic
satisfaction.
During the past several decades, advanced industrial societies have diverged
strikingly from the prevailing historical pattern; most of their population has
not grown up under conditions of hunger and economic insecurity. This has
led to a gradual shift in which needs for belonging, esteem, and intellectual
and self-expression have become more prominent. The scarcity hypothesis
implies that prolonged periods of high prosperity tends to encourage the
spread of postmaterialist values—and that enduring economic decline has the
opposite effect.
But, there is no one-to-one relationship between socioeconomic development
and the prevalence of postmaterialist values, for these values reflect one’s subjec-
tive sense of security, not simply one’s objective economic level. One’s sense of
security is shaped by a society’s social welfare institutions as well as its income
level, and is also influenced by the general sense of security prevailing in one’s
society. Furthermore, people’s basic value priorities do not change overnight;
the scarcity hypothesis must be interpreted in connection with the socialization
hypothesis.
One of the most pervasive concepts in social science is that one’s basic
personality structure crystallizes by the time one reaches adulthood. A large
body of evidence indicates that people’s basic values are largely fixed when they
reach adulthood, and change relatively little thereafter (Inglehart, 1977, 1997;
Rokeach, 1968, 1973). If so, we would expect to find substantial differences
between the values of the young and the old in societies that have experienced
a rising sense of security. People are most likely to adopt those values consis-
tent with what they experienced first-hand during their formative years. This
implies that intergenerational value change occurs if younger generations grow
Changing Values In The Islamic World And The west | 5

6
up under different conditions from those that shaped earlier generations, so

that the values of the entire society change gradually through intergenerational
replacement.
These two hypotheses generate several predictions concerning value change.
First, although the scarcity hypothesis implies that prosperity is conducive to the
spread of postmaterialist values, the socialization hypothesis implies that funda-
mental value change takes place gradually. To a large extent, it occurs as younger
generations replace older ones in the adult population. After an extended period
of rising economic and physical security, one would expect to find substantial dif-
ferences between the value priorities of older and younger groups, because they
would have been shaped by different experiences during their formative years.
However, a sizeable time lag would occur between economic changes and their
political effects. Fifteen or 20 years after an era of prosperity began, the birth
cohorts that spent their formative years in prosperity would begin to enter the
electorate.
Per-capita income and educational levels are among the best readily available
indicators of the conditions leading to the shift from materialist to postmaterialist
goals, but the theoretically crucial factor is not per-capita income itself, but one’s
sense of existential security—which means the impact of economic and physical
security is mediated by the given society’s social security system.
To test the value change hypothesis, we asked people which goals they con-
sidered most important, choosing between things such as economic growth, a
fight against rising prices, maintenance of order, and the fight against crime
(which tap materialist priorities); and freedom of speech, a society in which
people have more say in important government decisions and more say on the
job, and a society in which ideas count (which tap postmaterialist priorities).
Representative national surveys in six West European countries in 1970 revealed
huge differences between the values of young and old in all these societies. As
Figure 1.1 indicates, among those age 65 years and older, people with material-
ist value priorities outnumbered those with postmaterialist value priorities by
more than 12:1. But, as one moves from older to younger cohorts, the balance
shifts gradually toward a diminishing proportion of materialists and a growing
proportion of people with postmaterialist values. Among the youngest cohort
(those from 18–25 years in 1970), postmaterialists outnumber materialists. If we
assume the value priorities of given birth cohorts are stable, this implies that,
during the 1930s, when the two oldest cohorts were in their 20s and 30s, mate-
rialists must have outnumbered postmaterialists by at least 10:1 among the adult
population of these countries. In that era, the Marxist model of politics, domi-
nated by class conflict and economic issues, provided a reasonably good first
approximation of reality. However, the cross-sectional evidence in Figure 1.1 also
implies that, as the four oldest birth cohorts died off during the four decades
following 1970, we should observe a major shift in the motivations of these soci-
eties, with postmaterialists becoming as numerous as materialists, bringing a
corresponding shift away from economic issues toward increasing emphasis on
quality of life and expressive issues.

7
50%
Materialist
45%
Postmaterialist
Percentage of Age Group in Given Value Type

40%
35%
30%
25%
20%
15%
10%
5%
0%
Ages: 65+ 55−64 45−54 35−44 25−34 15−24
(Postwar
generation)
Figure 1.1 Value type by age group, among the populations
of England, France, West Germany, Italy, Belgium, and the
Netherlands in 1970. Source: European Community Survey
of February 1970. Based on original four-item materialist/
postmaterialist values battery. Reprinted from Inglehart,
1990: 76.
Cohort Effects Versus Life Cycle Effects

But are these value differences stable? Do these age differences reflect enduring
birth cohort effects or transient life cycle effects? With data from just one time
point, one cannot be sure—and the two interpretations have very different impli-
cations. The life cycle reading implies that the young will become increasingly
materialist as they age, so that by the time they are 65 years old, they will have
become just as materialist as the 65 year olds in 1970, which means that society
as a whole will not change at all. The cohort effects interpretation implies that the
younger cohorts will remain relatively postmaterialist over time, and that as they
replace the older, more materialist cohorts, the prevailing values of the society will
change profoundly.
Cohort analysis provides the only conclusive way to answer this question and
it requires (a) survey data covering a long time period; (b) surveys carried out at
numerous time points, enabling one to distinguish period effects from life cycle
and cohort effects; and (c) large numbers of respondents in each survey, because
when one breaks a given national sample down into six or seven birth cohorts, the
sampling error margin rises to the point where noise begins to drown out the signal.

8
Figure 1.2 shows the results of a cohort analysis that follows given birth cohorts
for almost 40 years, using data from Euro-barometer surveys that included
the materialist/postmaterialist battery in almost every year from 1970 to 1997;
supplemented with data from the fourth and fifth waves of the World Values
Surveys (WVSs), carried out in 1999 and in 2007 to 2009.1 This figure pools the
data from Britain, France, West Germany, Italy, Belgium, and the Netherlands
to provide large samples and relatively stable estimates of each cohort’s position
at a given time, which is calculated by subtracting the percentage of materialists
from the percentage of postmaterialists. Thus, at the zero point on the y-axis, the
two groups are equally numerous. The proportion of postmaterialists increases
as one moves up; the proportion of materialists increases as one moves down
in Figure 1.2.
If the age differences shown in Figure 1.1 reflected a life cycle effect, then each
of the cohort lines would move downward toward the materialist pole, with each
cohort becoming more materialist as one moves across Figure 1.2, from 1970 to
2009. If the age differences reflect stable birth cohort effects, the pattern would be
horizontal, with each birth cohort remaining about as postmaterialist at the end of
the time series as it was at the start.
But, we also need to take period effects into account. Our theory implies that
negative short-term effects such as a major recession will tend to push all cohorts
downward in response to current conditions; but, with recovery, they will return to
their former level, so that in the long run they will remain about as postmaterial-
ist as they were at the start. Over short periods, a period effect that pushed all the
cohorts downward could give the misleading impression that the age differences
reflected life cycle effects. But, in the long run, positive and negative fluctuations
tend to cancel each other out.
Because we have data from numerous time points, we can see that period
effects clearly are present. As Inglehart and Welzel (2005) demonstrated, they
reflect current economic conditions, particularly inflation levels. During periods
of economic difficulty, each birth cohort moves downward, becoming more mate-
rialist; with recovery, each birth cohort moves upward again, becoming more post-
materialist, but the differences between given birth cohorts are relatively stable.
Consequently, these period effects have no lasting impact. The younger cohorts
remain relatively postmaterialist despite short-term fluctuations and, over four
decades, we find no overall tendency for the members of given birth cohorts to
become more materialist as they age. Indeed, most cohorts are slightly more post-
materialist at the end of this time series than they were at the start.
During this four-decade span, the three oldest birth cohorts have left the
sample. First, the 1896-to-1905 cohort, then the 1906-to-1915 cohort, and finally
the 1916-to-1925 cohort disappeared, as the number of surviving members in
the cohort became too small to provide reliable estimates. These cohorts were
replaced by three new ones, born during 1956 to 1965, 1966 to 1975, and, most
recently, 1976 to 1985. The cohort analysis presented in Figure 1.2 shows no evi-
dence whatsoever of life cycle effects. Time series evidence covering four decades
makes it clear that the age-related differences found in 1970 reflect lasting cohort

9
25 25
1976−85
20 20
1986−95
15 15
10 10
5 1946−55 Total Sample 5
1966−75 0
0
1956−65
−5 −5
−10 −10
−15 −15
1936−45 −20
−20
−25 1926−35 −25
−30 −30
−35 −35
1916−25
Postmaterialists less materialists (%)

−40 −40
−45 1906−15 −45
−50 −50
1886−1905 −55
−55
−60 −60
1970 1973 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008
Year
Figure 1.2 Cohort analysis. Percentage of postmaterialists less the percentage of materialists in six West European countries—
Britain, France, West Germany, Italy, Belgium, and the Netherlands—1971 to 2009. (Data from 1970 through 1997 are from
Euro-Barometer surveys; data from 1999, 2006 and 2008–2009 are from European Values Study/World Values Survey.)
10
differences. This implies that as the younger, less materialist cohorts, replace the
older ones in the adult population, these societies should shift from materialist
toward postmaterialist values.
This is precisely what happened. During the past four decades, we find a sub-
stantial net shift toward postmaterialist values among the six populations first
surveyed in 1970 (and in the United States and other Western countries for which
we have time series data). The heavy shaded line in Figure 1.2 shows the net shift
toward postmaterialist values among the adult population as a whole at vari-
ous time points from 1970 to 2009. During the early 1970s, materialists heavily
outnumbered postmaterialists in all these countries. In the six West European
countries, materialists were four times as numerous as postmaterialists (and 21
times as numerous as postmaterialists among the oldest cohort). Similarly, in the
United States, materialists were three times as numerous as postmaterialists.
During the ensuing years, a major shift occurred. By 2000, postmaterialists were
slightly more numerous than materialists in western Europe and twice as numer-
ous as materialists in the United States. The predicted shift toward postmaterialist
values took place.
Strikingly similar findings have been reported by researchers in other
disciplines—from anthropology to biology. Thus, Gelfand et al. (2011) distin-
guished between cultures that are “tight” versus “loose,” arguing that these
qualities are shaped by the ecological and human-made threats the societies
encountered historically. These threats increase the need for strong norms and
severe punishment of deviant behavior. Tight societies have autocratic govern-
ments that suppress dissent, control crime sternly, and are relatively religious.
Testing these predictions against survey data from 33 countries, Gelfand et al.
(2011) found that nations that encountered severe ecological and historical threats
had relatively strong norms and low tolerance of deviant behavior. Severe existen-
tial pressures tend to produce “tight” cultures whereas fading pressures lower the
need for rigid norms, producing “loose” cultures that are less restricted by religion
and more tolerant of deviance.
Similarly, Thornhill and colleagues (Thornhill, Fincher, & Aran, 2009;
Thornhill, Fincher, Murray, 2010) found that historic vulnerability to infectious
disease is linked with collectivist attitudes, xenophobia, and low support for gen-
der equality—all of which hinder the emergence of democracy. Fincher, Thornhill,
Murray, and Schaller (2008), rated 98 societies on a collectivist–individualist scale,
and found that a high threat of disease goes with collectivist attitudes, controlling
for wealth and urbanization. As the threat from disease diminishes, individualism
and tolerance increase. These findings have very similar implications to those of
Inglehart et al.
Western levels of economic security have not continued to rise during the
past two decades. Economic growth has been relatively stagnant and increas-
ing income inequality has created a situation in which there has been little or
no growth in real income for most of the population. The negative impact of
economic stagnation and income inequality on economic security has been rein-
forced by cutbacks in the welfare state and high levels of unemployment, par-
ticularly among youth. The shift toward postmaterialist values has tapered off in
10 | Values, Political Action, and Change in the Middle East

11
these Western countries. Thus, in the most recent surveys, the youngest cohorts
are no more postmaterialist than the other postwar cohorts. The one surviving
prewar cohort remains distinctive, with materialists still outnumbering postma-
terialists, but intergenerational population replacement no longer brings a shift
toward postmaterialist values.
In contrast with the striking intergenerational value differences in 1970 that
are shown in Figure 1.1, Western countries no longer show sizeable differences
between the values of younger and older birth cohorts. A major value transition
occurred. In 1970, materialists vastly outnumbered postmaterialists in all Western
countries. By 2000, postmaterialists were slightly more numerous than material-
ists, but because postmaterialists tend to be concentrated among the more secure,
better educated, and more articulate strata of society, they set the tone; their values
have become politically correct. The culture of high-income Western countries has
been transformed, but intergenerational value change is no longer a major factor.
Except for the very old, young and old have pretty similar values.
But, the logic of the postmaterialist shift has significant implications for many
countries. Although the rates vary widely from country to country, the world as a
whole has been experiencing unprecedented economic growth since1980, with
India and China recently experiencing annual growth rates well more than 6%.
These countries are still in the phase of rising materialism that characterizes early
industrialization. But, if they continue on their current trajectories, they will even-
tually reach a stage when younger generations will have grown up under condi-
tions in which they take survival for granted. Many other countries—from Mexico
to Singapore—are approaching or have already attained this level.
In the world as a whole, the ratio between materialists and postmaterialists
varies tremendously according to a society’s level of economic development. Low-
income countries and strife-torn countries show an overwhelming preponderance
of materialists, whereas prosperous and secure ones show a preponderance of
postmaterialists. Materialists outnumber postmaterialists in Pakistan 55:1 and, in
Russia, 28:1; but, in the United States, postmaterialists outnumber materialists 2:1
and, in Sweden, postmaterialists prevail 5:1. There is no guarantee that the rapid
economic growth the world as a whole experienced from 1980 to 2008 will con-
tinue, but in those countries that do attain high levels of existential security, we
would expect intergenerational value change to take place.
Postmaterialist Values: Part of a Broader Cultural Change

The shift toward postmaterialist values is only one aspect of a still broader pro-
cess of cultural change reshaping the political outlook, religious orientations, gen-
der roles, and sexual mores of advanced industrial society (Inglehart 1990, 1997;
Inglehart & Welzel 2005). The emerging orientations place less emphasis on tra-
ditional cultural norms, especially those that limit individual self-expression.
To identify the main dimensions of global cultural variation, Inglehart and
Baker (2000)2 carried out a factor analysis of each society’s mean level on scores of
variables, tapping into a wide range of values. The two most significant dimensions

12
that emerged reflected (a) a polarization between traditional and secular–rational

values and (b) a polarization between survival and self-expression values.
Traditional values place strong emphasis on religion and respect for authority,
have relatively low levels of tolerance for abortion and divorce, and have relatively
high levels of national pride. Secular–rational values have the opposite characteris-
tics. Agrarian societies tend to emphasize traditional values; industrializing soci-
eties tend to emphasize secular–rational values.
The second major dimension of cross-cultural variation is linked with the
transition from industrial society to postindustrial societies, which brings a polar-
ization between survival and self-expression values. As Table 1.1 demonstrates, the
polarization between materialist and postmaterialist values is a sensitive indica-
tor of this dimension; the conditions that give rise to postmaterialist values are
also conducive to self-expression values. But, self-expression values encompass a
number of issues that go well beyond the items tapped by postmaterialist values.
For example, self-expression values reflect mass polarization over such issues as
whether “When jobs are scarce, men have more right to a job than women” or
whether “Men make better political leaders than women.” This emphasis on gen-
der equality is part of a broader syndrome of tolerance of outgroups, including
foreigners, gays, and lesbians. Self-expression values give high priority to environ-
mental protection, tolerance of diversity, and rising demands for participation in
decision making in economic and political life, and support for individual auton-
omy in general.
The shift from survival values to self-expression values also includes a shift in
child-rearing values, from emphasis on hard work toward emphasis on imagina-
tion and tolerance as important values to teach a child. Societies that rank high on
self-expression values also tend to rank high on interpersonal trust and have rela-
tively high levels of subjective well-being. This produces an environment of trust
and tolerance in which people place a relatively high value on individual freedom
and self-expression, and have activist political orientations—the attributes that the
political culture literature defines as crucial to democracy.
A major component of rise of self-expression values is a shift away from defer-
ence to all forms of external authority. Submission to authority has high costs; the
individual’s personal goals must be subordinated to those of external authorities.
Under conditions of insecurity, people are generally willing to do so. Under threat
of invasion, internal disorder, or economic collapse, people seek eagerly strong
authority figures that can protect them from danger.
Conversely, conditions of prosperity and security are conducive to tolerance
of diversity in general and democracy in particular. This helps explain a long-
established finding that rich societies are much likelier to be democratic than
poor ones. Under conditions of insecurity, people may be willing to submit to
authoritarian rule, but with increasing levels of existential security, they become
less willing to do so.
The rise of self-expression values brings an intergenerational change in a
wide variety of basic social norms—from cultural norms linked to survival of
the species, to norms linked to the pursuit of individual well-being. For example,
younger birth cohorts are markedly more tolerant of homosexuality than their

13
Table 1.1 Orientations linked with survival versus self-expression values
Item Correlation
Survival values emphasize the following:

Materialist/postmaterialist values .87
Men make better political leaders than women .86
Respondent is not highly satisfied with life .84
A woman has to have children to be fulfilled .83
Respondent rejects foreigners, homosexuals, and people with .81
AIDS as neighbors
Respondent has not and would not sign a petition .80
Respondent is not very happy .79
Respondent favors more emphasis on the development of technology .78
Homosexuality is never justifiable .78
Respondent has not recycled something to protect the environment .76
Respondent has not attended a meeting or signed a petition to .75
protect the environment
A good income and safe job are more important than a feeling of .74
accomplishment and working with people you like
Respondent does not rate own health as very good .73
A child needs a home with both a father and a mother to grow up .73
happily
When jobs are scarce, a man has more right to a job than a women .69
A university education is more important for a boy than for a girl .67
Government should ensure that everyone is provided for .69
Hard work is one of the most important things to teach a child .65
Imagination is not one of the most important things to teach a child .62
Tolerance is not one of the most important things to teach a child .62
Leisure is not very important in life .61
Scientific discoveries will help, rather than harm, humanity .60
Friends are not very important in life .56
You have to be very careful about trusting people .56
Respondent has not and would not join a boycott .56
Respondent is relatively favorable to state ownership of business .54
and industry
Self-expression values take the opposite position on all the above.
The original polarities vary; the statements made in the table show how each item relates to
this values index.
elders. And younger cohorts are substantially more favorable to gender equality
and are more permissive in their attitudes toward abortion, divorce, extramari-
tal affairs, and euthanasia. Economic accumulation for the sake of economic
security was the central goal of industrial society. Ironically, its attainment set in
motion a process of gradual cultural change that has made these goals less cen-
tral and is now bringing a rejection of the hierarchical institutions that helped
attain them.

14
Intergenerational Shift from Survival Values

toward Self-Expression Values
Throughout advanced industrial societies, the younger age cohorts emphasize self-
expression values much more heavily than their elders, in a pattern similar to that
found earlier with postmaterialist values. As we have seen, given birth cohorts did
not become more materialistic as they aged. This holds true for the shift from sur-
vival to self-expression values as well, although we do not yet have a massive time
series database comparable with what is available with materialist/postmaterialist
values. But, we do have evidence from six waves of the World Values Surveys (WVSs)
carried out from 1981 to 2013. From the start of this time series, younger birth cohorts
placed more emphasis on self-expression values than older cohorts, and they did not
move away from self-expression values toward survival values as they aged from 1980
to 2000. Throughout this period, younger birth cohorts continued to place more
emphasis on self-expression values than older ones. And although each of the birth
cohorts aged by several decades during this period, none of them placed less empha-
sis on self-expression in 2012 than in 1981—as would have happened if these age
differences simply reflected life cycle effects.
The evidence suggests that major cultural changes are occurring through an
intergenerational value shift linked with the fact that the younger birth cohorts
have grown up under greater levels of existential security than those that shaped
the formative years of the older cohorts. As Inglehart and Welzel (2005) dem-
onstrated, emphasis on self-expression values is linked strongly to the extent to
which a given society actually approaches gender equality in political and social
life. Even more strikingly, the shift toward self-expression values is conducive to
the spread and flourishing of democratic institutions.
How Different Is the Islamic World?

The Arab Spring that erupted in December 2010 inspired widespread hope that it
would bring a wave of democratization similar to the one that swept East Asia and
eastern Europe during the late 1980s and early 1990s. Most of the new democra-
cies that were established in East Asia and eastern Europe are still flourishing,
but the democratizing surge of the Arab Spring has largely floundered. In a world
where most other countries have democratic governments, the 2014 Freedom
House report does not code a single Muslim-majority country as “Free.” Among 18
Arab-speaking countries, only four—Morocco, Tunisia, Libya, and Lebanon—are
even listed as “Partly Free.” The situation is slightly better among the 19 non-Arab
Muslim- majority countries, where nine— Albania, Bangladesh, Burkina Faso,
Indonesia, Kyrgyzstan, Kuwait, Malaysia, Mali, and Turkey—are “Partly Free.”
Throughout the rest of the world, in contrast, 57% of all countries are considered
“Free,” and an additional 30% are categorized as “Partly Free.”
These findings might be interpreted as giving new credibility to Samuel
Huntington’s (1993, 1997) hotly contested thesis that the Islamic world and
the West are divided by a clash of civilizations. Why did the recent wave of

15
democratization fail in most Arab countries? Huntington’s response would be

that the Islamic world lacks the core political values that gave birth to representa-
tive democracy in Western civilization: separation of religious and secular author-
ity, rule of law and social pluralism, parliamentary institutions of representative
government, and protection of individual rights and civil liberties as the buffer
between citizens and the power of the state. This claim is not entirely implausible.
According to the latest Freedom House rankings, about two thirds of the world’s
countries are now electoral democracies. But, among the 47 countries with an
Islamic majority, only one fourth are electoral democracies, and none of the core
Arabic-speaking societies falls into this category.
But, this does not prove Huntington was correct, since it tells us nothing about
the underlying beliefs of the Islamic populations. The results of the latest wave
of the WVS, conducted in 2010 to 2013 provide an extensive body of relevant evi-
dence. Asking questions that explore values and beliefs in more than 100 coun-
tries, the WVS is an investigation of sociocultural and political change that covers
more than 90% of the world’s population.
A comparison of the data yielded by these surveys in Islamic and non-Islamic
societies around the globe confirms the first claim in Huntington’s thesis: Culture
does matter; indeed, it matters a lot. Historical religious traditions have left an
enduring imprint on contemporary values. But, Huntington was mistaken in
assuming that the core clash between the West and Islam is over political values.
Today, the populations of almost every country in the world—whether Muslim,
Christian, Hindu, Buddhist, or Confucian—see democracy as the best form of
government.
The 2010 to 2014 WVS asked representative national samples of the public of
each country whether they thought various ways of governing their country would
be very good, fairly good, fairly bad, or very bad. The options included “Having
a strong leader who does not have to bother with parliament and elections”;
“Having experts, not government, make decisions according to what they think
is best for the country”; “Having the army rule”; and “Having a democratic politi-
cal system.” In virtually every country in the world, democratic political systems
were endorsed by a larger share of the population than any other alternative. This
includes the populations of the 17 Muslim-majority countries. As Figure 1.3 dem-
onstrates, overwhelming majorities of the populations of both the Arab-speaking
Islamic countries and the other Islamic countries rated a democratic system as
either “very good” or “good.” Support for these options among Islamic popula-
tions was about as high as among the populations of Western countries, and was
significantly greater than among the populations of post-Communist countries or
societies with a Confucian historical heritage.
At this point in history, democracy has an overwhelmingly positive image
throughout the world. In country after country, a clear majority of the popula-
tion describes “having a democratic political system” as either “good” or “very
good.” This represents a dramatic change from the 1930s and 1940s, when fascist
regimes won overwhelming mass approval in many societies; and from the 1920s
until the 1980s, Communist regimes had widespread support. However, in the
past few decades, democracy has become virtually the only political model with

16
100%
95%
90%
85%
80%
75%
Western Arab- Other Confucian Other Post-
Islamic Islamic Communist
Figure 1.3 Support for a democratic system in six cultural zones. Values are
a percentage of those who view a democratic system as “good” or “very good.”
(2010–2014 WVS.)
global appeal, regardless of cultural heritage. A solid majority of the public in both
Western and Muslim countries gives democracy high marks as the most efficient
form of government, with 68% disagreeing that “democracies are indecisive” and
“democracies aren’t good at maintaining order” (all other cultural regions and
countries, except East Asia and Japan, being far more critical). And equal num-
bers of respondents on both sides of the civilizational divide (61%) firmly reject
authoritarian governance, expressing disapproval of “strong leaders who need not
bother with parliament and elections.”
As Inglehart and Norris (2003) and Norris and Inglehart (2004) point out,
the real fault line between the West and Islam—which Huntington overlooked
completely—concerns gender equality and sexual liberalization. The values sepa-
rating the two cultures have more to do with eros than demos. As younger gen-
erations in the West have gradually become more liberal on these issues, Islamic
societies have remained the most traditional ones in the world.
Commenting on the disenfranchisement of women throughout the Middle
East, the United Nations Development Program has argued that “no society can
achieve the desired state of well-being and human development, or compete in a
globalizing world, if half its people remain marginalized and disempowered.” But
this “sexual clash of civilizations” taps into far deeper issues than how Muslim
countries treat women. A society’s commitment to gender equality and sexual
liberalization is a powerful indicator of how strongly that society supports social
tolerance—a core prerequisite for effective democracy.
In attitudes toward gender equality and sexual liberalization, the cultural gap
between Islam and the West widens into a chasm. On the matter of equal rights

17
and opportunities for women—measured by such questions as whether men

make better political leaders than women or whether university education is more
important for boys than for girls—Western and Muslim countries score 82% and
55%, respectively. Islamic societies are also much less tolerant of homosexuality,
abortion, and divorce. The comparative cross-national surveys in seven Middle
Eastern countries cited elsewhere in this book include several questions on moral-
ity. Respondents were asked to rate between 1 point and 10 points the morality of
violence against other people, stealing other’s property, telling lies to protect one’s
interests, and premarital sex. In many of these surveys, premarital sex was con-
sidered more immoral than prevarication, theft, or violence against others. This
is one more piece of evidence that the contrast between the values of Western and
Muslim-majority societies involves norms concerning sexual behavior rather than
democracy. Virtually all Muslim-majority populations express favorable attitudes
toward democracy, but they are much less tolerant than Western populations con-
cerning gender equality, premarital sex, and homosexuality.
Along with postmaterialist values, these issues are part of the broader syn-
drome of tolerance, trust, political activism, and emphasis on individual auton-
omy that constitutes self-expression values, as Table 1.1 indicated. The extent to
which a society emphasizes these self-expression values has a surprisingly strong
bearing on the emergence and survival of democratic institutions. Among the
countries included in the WVS, support for gender equality—a key indicator of
social tolerance—is linked closely with a society’s level of democracy.
The world as a whole shows a clear intergenerational shift toward growing sup-
port for gender equality, but the trend is very uneven. Among the 35 high-income
non-Muslim majority societies included in the WVSs, all 35—without a single
exception—show strong intergenerational differences, with the younger birth
cohorts being more favorable to gender equality than the older cohorts, and in
most countries the differences are very large. This holds true not only of Western
high-income countries but also of Japan, Hong Kong, Singapore, Taiwan, and
South Korea.
Muslim-majority societies show a mixed pattern. In 13 countries—Albania,
Lebanon, Morocco, Tunisia, Iran, Turkey, Palestine, Bangladesh, Jordan, Malaysia,
Burkina Faso, Egypt, and Iraq—we find statistically significant tendencies for
younger respondents to be more favorable to gender equality than their older com-
patriots. This tendency ranges from rather strong in Albania to relatively weak in
Iraq (though 34 of the 35 non-Islamic high-income countries show stronger inter-
generational differences than Albania). And 14 other Muslim-majority countries—
Libya, Pakistan, Mali, Yemen, Kazakhstan, Kosovo, Saudi Arabia, Uzbekistan,
Algeria, Kyrgyzstan, Qatar, Bahrain, Indonesia, and Azerbaijan (plus half-Muslim
Nigeria)—show little or no evidence of an intergenerational shift toward growing
support for gender equality.
What accounts for this bifurcation of Muslim-majority countries on movement
toward gender equality? One point is clear: None of the four ex-Soviet Muslim
majority countries shows evidence of an intergenerational shift toward gender
equality, although why this is true is not clear, particularly as these four countries
are markedly more secular than most Muslim-majority societies. For the world

18
as a whole, high-income countries show much stronger evidence of an inter-

generational shift toward growing support for gender equality than low-income
countries, but within the Islamic world this tendency is less evident. The popu-
lations of rich countries such as Qatar and Bahrain show no intergenerational
differences whatsoever. The Muslim-majority countries that show the largest
intergenerational differences tend to be located on the Mediterranean and have
relatively strong exchanges with Europe, but there are several exceptions. The rea-
sons underlying the bifurcation of the Muslim world into two groups, only one
of which shows evidence of an emerging intergenerational shift toward growing
support for gender equality demands further research.
This question has important political implications because, as Figure 1.4 indi-
cates, attitudes toward gender equality have a surprisingly strong linkage with
democracy (r = .82). In every stable democracy, a majority of the public disagrees
with the statement: Men make better political leaders than women. None of the
societies in which less than 30% of the public rejects this statement has a dem-
ocratic government. In China, one of the world’s most autocratic countries, a
majority of the public agrees that men make better political leaders than women,
despite a party line that has long emphasized gender equality. In practice, Chinese
women occupy few positions of real power and face widespread discrimination
in the workplace. India is a borderline case. The country is a long-standing par-
liamentary democracy with an independent judiciary and civilian control of the
armed forces, but it is also marred by a weak rule of law, arbitrary arrests, and
210
United States Australia Sweden
200 New Norway
190 Spain Zealand
Japan
180 Finland
Germany
Sum of Freedom House ratings, 1981−1998
170 Venezuela
160 Dominican
150 Republic Uruguay
140 Argentina
Brazil
130 India
Colombia
Philippines
120 South Korea El Salvador Peru
Hungary
110 Poland Mexico Chile
100 Bangladesh Taiwan Slovenia
90 Turkey Czech
Lithuania
80 Latvia Slovakia South
Pakistan Africa
70 Estonia
Bulgaria Croatia
Jordan
60 Ukraine
Nigeria Romania Russia
50 Egypt
Armenia Moldova Bosnia Macedonia
40 Belarus
Georgia Yugoslavia
30 Iran Azerbaijan
20
10 China
0
5 15 25 35 45 55 65 75 85
Respondents who disagree with the statement
"Men make better political leaders than women" (%)
Figure 1.4 Gender equality and democracy.

19
extrajudicial killings. The status of Indian women reflects that duality. Women’s
rights are guaranteed in the Constitution, and Indira Gandhi led the nation for
15 years. But, domestic violence and forced prostitution remain widespread in
India and, as Figure 1.4 indicates, about half of the public believes that men make
better political leaders than women.
As Figure 1.5 indicates, Muslim-majority countries rank lower than any other
major cultural group in support for gender equality. According to evidence from
the 2010 to 2014 WVS, fully 81% of the population of Western countries disagrees
with the statement that men make better political leaders than women. In both the
Confucian-influenced countries and the “Other” category, clear majorities of 58%
and 59%, respectively, reject that claim. In the post-Communist countries, this
figure is 47%. But, belief in gender equality drops sharply among Islamic popula-
tions; the claim of male superiority is rejected by only 21% in the Arab-Islamic
countries and by 26% in the “Other” Islamic countries.
Responses to the statement “When jobs are scarce, men have more right to a
job than women” produces similar results. In Western countries, 70% of the pub-
lic disagree with that statement. In Post-Communist countries, the figure is 55%;
in the “Other” category, it is 50%; and in Confucian societies, it is 30%. However,
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
Western Other Confucian Post- Other Arab-
Communist Islamic Islamic
Figure 1.5 Support for gender equality in six cultural zones. This graph indicates
the percentage of respondents who disagreed with the question: Do men make
better political leaders than women? Arab-Islamic: Morocco, Tunisia, Libya, Egypt,
Jordan, Palestine, Qatar, Yemen, Lebanon, and Iraq; Confucian: China, Japan,
Taiwan, and South Korea; Other: Mexico, Colombia, Argentina, Chile, Peru,
Trinidad, Thailand, Philippines, Ghana, and Zimbabwe; Other Islamic: Turkey,
Pakistan, Uzbekistan, Kazakhstan, Kyrgyzstan, Malaysia, and Azerbaijan; Post-
Communist: Russia, Ukraine, Belarus, Poland, Romania, Slovenia, Estonia,
Armenia, and Georgia; Western: United States, Sweden, Germany, Cyprus, Australia,
and New Zealand. (2010–2014 WVS.)

20
it falls to 21% and 20% in Arab-Islamic countries and Other Islamic countries,
respectively.
Attitudes toward homosexuality provide an excellent litmus test of a society’s
commitment to equality. Tolerance of well-liked groups is never a problem; the
crucial test of social tolerance is how a society treats it most unpopular groups
(Gibson, 1992). Today, relatively few people express overt hostility toward other
classes, races, or religions, but rejection of homosexuals remains widespread.
The WVS asked representative national samples of populations of throughout the
world about which groups they would be willing to have as neighbors. Globally,
homosexuals proved to be the least-liked group. In response to another WVS
question, measuring whether homosexuality is ever justifiable on a 10-point scale,
about half the world’s population, selected 1 point—meaning, “never.” Like gender
equality, this attitude is linked strongly to a country’s level of democracy. Among
authoritarian and quasi-democratic states, rejection of homosexuality is deeply
entrenched. In the latest available survey, 99% of the public in both Egypt and
Bangladesh said homosexuality is “never” acceptable; the corresponding figures
were 94% in Iran, 92% in China, and 71% in India. The populations of stable
democracies were much more tolerant. Only 32% of the US public said that
homosexuality was never acceptable, and the figures were 26% in Canada, 25% in
Britain, 19% in Germany, and 10% in the Netherlands.
Islamic societies are neither uniquely nor monolithically low on tolerance
toward sexual orientation and gender equality. Many of the Soviet successor states
rank about as low as most Islamic societies. On the whole, however, the popula-
tions of Muslim-majority countries not only lag behind the West but behind all
the other categories of societies, as Figure 1.6 indicates. Even more strikingly, the
gap between the West and Islam is even wider among the younger age groups
than it is among older respondents. Although the younger birth cohorts in many
Muslim-majority societies are becoming more tolerant, more liberal, and less reli-
gious than their elders, the intergenerational differences in most Western societ-
ies are considerably larger. Consequently, cultural change is moving much faster
in Western societies than in the Muslim world, producing a growing cultural gap.
Any claim of a “clash of civilizations” based on fundamentally different political
goals held by Western and Islamic societies is a gross oversimplification. Support
for the goal of democracy is widespread among Islamic populations, even among
those living in the most authoritarian regimes. But, Huntington was correct in
claiming that cultural differences have taken on a new importance, forming the
fault lines for future conflict. Although the populations of virtually every country
in the world now endorse the goal of democracy, there is no global acceptance of
the self-expression values (such as social tolerance, gender equality, and trust) and
the postmaterialist emphasis on participation and freedom of speech that are cru-
cial to democracy. Today, these divergent values constitute the real clash between
Islamic societies and the West.
But, economic development is conducive to changing attitudes in virtually any
society. Modernization brings systematic, predictable changes in gender roles.
Industrialization brings women into the paid workforce and reduces fertility rates
dramatically. Women become literate and begin to participate in representative

21
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
Western Confucian Other Post- Other Arab-
Communist Islamic Islamic
Figure 1.6 Tolerance of homosexuality in six cultural zones. This graph indicates
the percentage of respondents who expressed any degree of tolerance to the
question: Is homosexuality ever justifiable? Arab-Islamic: Morocco, Tunisia, Libya,
Egypt, Jordan, Palestine, Qatar, Yemen, Lebanon, and Iraq; Confucian: China,
Japan, Taiwan, and South Korea; Other: Mexico, Colombia, Argentina, Chile, Peru,
Trinidad, Thailand, Philippines, Ghana, and Zimbabwe; Other Islamic: Turkey,
Pakistan, Uzbekistan, Kazakhstan, Kyrgyzstan, Malaysia, and Azerbaijan; Post-
Communist: Russia, Ukraine, Belarus, Poland, Romania, Slovenia, Estonia,
Armenia, and Georgia; Western: United States, Sweden, Germany, Cyprus,
Australia, and New Zealand. (2010–2014 WVS.)
government, although still have far less power than men. Then the postindustrial
phase brings a shift toward greater gender equality as women move into higher
status economic roles in management and gain political influence within elected
and appointed bodies. Thus, the populations of relatively industrialized Islamic
societies such as Turkey hold views on gender equality and sexual liberalization
that are very similar to those held in other new democracies.
Cultural attitudes—and deep-rooted support for democracy—is linked closely
with modernization. Women did not attain the right to vote in most historically
Protestant societies until about 1920 and, in Roman Catholic Europe, not until
after World War II. In 1945, only 3% of the members of parliaments around the
world were women. In 1965, the figure had risen to 8%, in 1985 to 12%, in 2002
to 15%, and in 2014 to 22%.
The forces of modernization are starting to transform Islamic societies. The
2007 to 2013 wave of the WVS shows evidence that a process of intergenerational
value change similar to the one that transformed the political cultures of Western
societies during the decades after 1945, is now at work in Muslim-majority coun-
tries. Interestingly, the process tends to be more advanced in the countries that

22
50
45 Materialist
40 Postmaterialist
35
30
25
20
15
10
5
0
36
46
56
66
76
86
96
9
9
−1
−1
−1
−1
−1
−1
−1
27
37
47
57
67
77
87
19
19
19
19
19
19
19
Years
Figure 1.7 Materialist versus postmaterialist values by birth cohort in nine
Muslim-majority countries. (World Values Survey and European Values
Study, 2007–2013.)
played leading roles in the recent Arab Spring uprisings than in other Muslim
societies.
Figure 1.7 shows the shifting balance between materialist and postmaterial-
ist values in nine Muslim-majority countries: Morocco, Tunisia, Libya, Palestine,
Jordan, Turkey, Albania, and Indonesia. Among the oldest birth cohort, material-
ists outnumber postmaterialists by a ratio of more than 10:1; among the youngest
birth cohort, the ratio is slightly more than 2:1, which is roughly the ratio found
among the youngest prewar cohort in western Europe in 1970. The ratio is clearly
shifting, although it has not yet produced a cohort in which postmaterialists
outnumber materialists, as was true of the first postwar West European cohort
in 1970.
This pattern does not hold true for all Islamic countries for which we have
data. In 14 other Muslim-majority countries (Kazakhstan, Uzbekistan, Azerbaijan,
Kyrgyzstan, Lebanon, Iran, Saudi Arabia, Qatar, Bahrain, Yemen, Mali, Pakistan,
Bangladesh, and Malaysia) the age-related differences are relatively weak (median
r = –.05). The countries that show evidence of relatively strong intergenerational
value change do not have the highest levels of per-capita gross domestic prod-
uct (which are found in the Gulf states), but they do have significantly greater
life expectancies, lower infant mortality rates, and lower fertility rates than the
other Islamic countries. Thus, the median life expectancy in the countries shown
in Figure 1.7 is 74.8 years; the median life expectancy in the other countries is
68.9 years. Survival tends to be relatively secure in the Muslim-majority coun-
tries that show relatively high rates of intergenerational value change. Moreover,
apart from Indonesia, these countries are clustered on or near the Mediterranean
and have relatively flows of population to and from western Europe. These coun-
tries also show relatively strong evidence of intergenerational change in attitudes
toward gender equality.

23
The bifurcation of Muslim-majority countries into two groups—one of which

shows clear signs of an intergenerational shift toward growing support for gender
equality and postmaterialist values whereas the other does not—suggests that the
cultural changes linked with modernization have only begun to impinge on the
Muslim world and they are doing so unevenly. The fact that these changes have
been transforming the populations of literally all non-Muslim high-income coun-
tries suggests, but does not prove, that in the long run they will also affect increas-
ing numbers of Muslim-majority societies.
Simply holding elections and having a parliament will not establish effective
democracy in the Muslim world. And it seems unlikely that many Arab-speaking
countries will, in the near future, establish an enduring wave of democratization
like the one that swept eastern Europe during the final days of the Cold War. But,
there are signs that a process of intergenerational change is beginning to trans-
form the culture of these countries. Culture has a lasting impact on how societies
evolve, but culture is not destiny.
Notes
1. The samples are weighted to reflect each country’s population. Because the
2006 World Values Survey did not include Belgium, we used data from the 1999
Belgian survey in the pooled analysis. This tends to reduce the amount of change
observed from 1999 to 2006, but the distortion is minimal because Belgium con-
tains only 4% of the population of the six countries.
2. For details on how these factor analyses were carried out at both the individual
and societal levels, see Inglehart and Baker (2000).
References
Boeltken, F., & Jagodzinski, W. (1985). Postmaterialism in the European commu-
nity, 1970–1980: Insecure value orientations in an environment of insecurity.
Comparative Political Studies, 17(4), 453–484.
Fincher, C. L., Thornhill, R., Murray, D. R., & Schaller, M. (2008). Pathogen preva-
lence predicts human cross-cultural variability in individualism/collectivism.
Proceedings of the Royal Society B, 275, 1279–1285.
Freedom House. (2014). Freedom in the world [On- line]. Available: http://www.
freedomhouse.org/report/freedom-world/freedom-world-2014.
Gelfand, M. J., et al. (2011). Differences between tight and loose cultures: A
33-nation study.
Gibson, J. L. (1992). Alternative measures of political tolerance: Must tolerance be
“least-liked?” American Journal of Political Science, 36, 560–577.
Huntington, S. P. (1993). Huntington, The clash of civilizations. Foreign affairs, 72(3),
22–49.
Huntington, S. P. (1997). The clash of civilizations and the remaking of world order.
Penguin Books India.

24
Inglehart, R. (1971). The silent revolution in Europe: Intergenerational change in

post-industrial societies.” American Political Science Review, 65(4), 991–1017.
Inglehart, R. (1977). The silent revolution: Changing values and political styles among
western publics. Princeton, NJ: Princeton University Press.
Inglehart, R. (1990). Culture shift in advanced industrial society. Princeton, NJ: Princeton
University Press.
Inglehart, R. (1997). Modernization and postmodernization: Cultural, economic and
political change in 43 societies. Princeton, NJ: Princeton University Press.
Inglehart, R., & Baker, W. (2000). Modernization, cultural change and the persis-
tence of traditional values. American Sociological Review, 65, 19–51.
Inglehart, R., & Norris, P. (2003). Islam and the West: A “clash of civilizations?”
Foreign Policy, March/April, 62–70.
Inglehart, R., & Norris, P. (2004). Rising tide: Gender equality in global perspective.
Cambridge: Cambridge University Press, 2003.
Inglehart, R., & Welzel, C. (2005). Modernization, cultural change and democracy.
New York: Cambridge University Press.
Norris, P., & Inglehart, R. (2004). Sacred and secular: Religion and politics worldwide.
Rokeach, M. (1968). Beliefs, attitudes and values. San Francisco: Jossey-Bass.
Rokeach, M. (1973). The nature of human values. New York: Free Press.
Thornhill, R., Fincher, C., & Aran, D. (2009). Parasites, democratization, and the
liberalization of values across contemporary countries. Biological Reviews, 84(1),
113–131.
Thornhill, R., Fincher, C., & Murray, D. R. (2010). Zoonotic and non-zoonotic diseases
in relation to human personality and societal values. Evolutionary Psychology, 8,
151–155.

25
CHAPTER 2 Youth Perceptions and Values During

the Arab Spring
Cross-national Variation and Trends
Mansoor Moaddel and Julie de Jong
It has been claimed that youths played a major role in the protest movements
known as the Arab Spring that toppled the ruling regimes in Egypt, Tunisia, Libya,
and Yemen (Campante & Chor, 2012; Chaaban, 2009; Hoffman & Jamal, 2012;
Hvistendahl, 2011; LaGraffe, 2012). Without the constraints of family obligations
and work, youths may have a greater level of political engagement than the older
population. In addition, there may be intergenerational differences in value ori-
entation, with the younger generation subscribing to a specific set of values,
norms, and lifestyles generally known as youth culture. The differences in the lev-
els of personal constraints and formative life experiences may thus explain why
youths might have different value orientations and why they chose to participate
in the Arab Spring more often than the rest of the population. Nonetheless, scant
empirical evidence has been presented to substantiate these intergenerational
differences.1
In this chapter, we construct attitudinal measures of social individualism, gen-
der equality, secular politics, and basis of identity. We assess (a) cross-national vari-
ation in these measures in Egypt, Iraq, Lebanon, Pakistan, Saudi Arabia, Tunisia,
and Turkey; and (b) trends in some of these measures in a smaller set of countries
(Egypt, Iraq, Saudi Arabia, and Turkey) for which longitudinal data are available.
Finally, we evaluate whether variations in values are a function of age. Are younger
cohorts more individualistic and more supportive of gender equality and secular
politics—that is, more liberal—than older cohorts? Are these cohorts different
in terms of identity? Do they identify predominantly with their religion, nation,
or ethnicity? More specifically, do younger cohorts consider themselves primarily
Muslim or Christian, as members of the territorial nation (e.g., Egyptians, Iraqis,
Saudis, or Turkish), or as Arabs, Kurdish, or Turkish? To what extent are trends
in values over time different across age groups? Finally, based on the analysis
presented, we determine (a) whether it is possible to characterize the Arab Spring
as the ushering in of a new cultural episode in the contemporary history of the
Middle East and North Africa (MENA), and, if so, (b) the role played by youths in
driving this trend. We answer these questions by focusing on measures of social
26
individualism, gender equality, secular politics, and basis of identity using cross-
national survey data.
The Significance of Values

Social scientists have widely recognized the role of values in shaping human
behavior, perception, and emotion (Boudon, 2001; Durkheim, 1893; Harrison &
Huntington 2000; Inglehart & Welzel, 2005; Parsons, 1951; Rokeach, 1971, 1973;
Schwartz, 2006, 2012; Weber, 1958 [1905]). Despite this recognition, it is not
quite clear on which set of values researchers should focus to understand more
fully stability and change within the historical context of a society or a region that
displays a distinct cultural tradition. Moreover, one of the problems associated
with the study of the causes and consequences of human values is that the things
that people value in life and the reasoning underpinning their values vary cross-
nationally and in different historical periods. For example, people cherish and
value the scriptures because they are religious, they value their national flag and
treat it as a sacred object because they are raised as patriots, and they value educa-
tion for women because they believe in gender equality. They may, however, aban-
don these values as a result of becoming secular, immigrating to another country,
and joining a religious fundamentalist movement, respectively.
In attempting to explain variation in values, social scientists have considered
changes in aspects of social conditions such as division of labor, economic develop-
ment, demographic change, and state formation. In the seminal works of Parsons
and others (Parsons, 1949, 1951, 1966; Sahlins & Service, 1960; Service, 1971), this
variation is linked to social differentiation produced as a result of development in
what they consider different subsystems of society. Inglehart and Welzel (2005)
consider the rise of self-expressive values, as opposed to materialistic values, to be
associated with a sense of security provided as a result of an increase in the level
of economic prosperity. Schwartz (2006, 2012) develops an a priori proposition
concerning the existence of three basic dimensions in all cultures. He then identi-
fies seven cultural value orientations (affective autonomy, egalitarianism, embed-
dedness, harmony, hierarchy, intellectual autonomy, and mastery) as different
resolutions of the critical issues that these dimensions reflect: autonomy versus
embeddedness, egalitarianism versus hierarchy, and harmony versus mastery.
These value orientations are then related to gross domestic product per capita,
democratization, and household size.
These approaches are useful in categorizing societies in terms of value empha-
ses, tapping into the world’s cultural diversity, and explaining cross-national varia-
tion in values. Additional explanatory models, however, are necessary to describe
more fully how changes in values transpire and to determine which factors
explain historical discontinuity in the process of value change. Objective factors
such as level of economic development, gross domestic product per capita, and
demographic changes cannot provide an adequate explanation of, for example,
the remarkable differences in value orientations among the dominant political
and cultural movements that emerged in MENA during the first quarter of the

27
20th century and those that transpired during the last quarter. During the first
quarter, these movements were shaped by territorial nationalism, had a secular
orientation, and were predisposed favorably toward gender equality. During the
last quarter of the century, however, movements were informed by religious fun-
damentalism and the notions of gender segregation and male supremacy.
The set of values selected for analysis in this chapter—social individualism,
gender equality, secular politics, and identity—cannot be construed simply as a
derivative of the type of economic development or political transformation expe-
rienced by people in MENA. For sure, compared with other aspects of human
concern, such as economic prosperity, political stability, security, and health, these
values may appear less significant. After all, as Marx has said, humans must be in
a position to live—that is, have food, shelter, and habitation—to be able to make
history (Geras, 1983). Thus, where people are struggling to make ends meet under
the conditions of food scarcity, inadequate housing, and political instability, they
may care little about individual rights, equality of political voices, and the basis of
their identity.
Nonetheless, these values are important in a historical sense; they constitute
a significant component of the sociopolitical and cultural issues debated by intel-
lectual leaders and political activists in contemporary MENA, as these intellec-
tuals and activists attempt to construct the parameters of the social order they
deem desirable. These debates revolve on the utility and significance of individual
autonomy vis-à-vis obedience to parental authority or patriarchy, gender equality
vis-à-vis gender hierarchies and male supremacy, secular politics vis-à-vis Islamic
government, the nation as the primary basis of identity vis-à-vis the primacy of
religion, religious equality vis-à-vis religious intolerance, and the West as the
exemplar of progress and civilization vis-à-vis the West as the epitome of cultural
decadence (Moaddel 2005).
Positions on these issues decided the ideological orientations of the indig-
enous intellectual leaders, indicating whether they were Islamic modern-
ists, territorial nationalists, ethnic nationalists (pan-Arab, Turkish, or Kurdish
nationalists), or Islamic fundamentalists. For example, 19th-century Islamic
modernists were theologically individualistic, favored laissez-faire capitalism,
supported the separation of religion and politics, had a favorable orientation
toward Western culture, and advanced a feminist interpretation of the Quran to
defend women’s rights. Twentieth-century Islamic fundamentalists, in contrast,
were theologically communitarian, defended the unity of religion and politics
in an Islamic government, considered Western culture as decadent, and pro-
moted gender inequality and segregation rigorously. In terms of identity, the
followers of territorial nationalism consider the territorial nation as the basis of
identity, whereas ethnic nationalists favor varied forms of ethnic identity such as
pan-Arab nationalism, Kurdish nationalism, or Turkish nationalism, with eth-
nicity specified as Arab, Kurd, or Turkish, respectively. For those who identify
with Islamic fundamentalism-cum-nationalism, religion is the basis of identity
(Moaddel, 2005; see also the chapter on “National Identity vs. Pride in Liberal
Territorial Nationalism and Islamic Nationalism in the Middle East and North
Africa,” Chapter 3 in this volume).
Youth Activity During The Arab spring | 27

28
For sure, intellectual leaders and activists in the MENA region also addressed
such material issues as economic underdevelopment, authoritarian political
institutions and lack of transparency in politics, foreign interventions, economic
inequality and poverty, and inadequate health. Their expressions on how to resolve
these issues, however, did not emanate from the exigencies of these problems.
Nor can it be said that their discourses were affected fully or even primarily by the
value emphases characterizing the cultural traditions of their society. Rather, what
they actually said about these issues and their solutions to economic and political
problems were shaped, as it were, by the type of ideological targets they encoun-
tered. And, these targets varied under different historical conditions: monarchical
absolutism and ulama obstructionism during the late 19th and early 20th century,
Western interventions during the first half of the 20th century, the secular author-
itarian and interventionist state during the second half of the 20th century, and,
among Iranian and Saudi populations, the monolith and obstructionist religious
institutions imposed from above under current conditions. These targets have
been perceived as the obstacles to the formulation of an independent and prosper-
ous society. Thus, the dominant sociopolitical and cultural values proposed and
defended by these leaders were produced or invoked in oppositional relation to the
perceived obstacles: the targets (Moaddel 2005).
Social Individualism, Gender Equality, Secular Politics,

Liberalism, and Basis of Identity
Considering this background, the aim of this chapter is not a full assessment
of the value orientations of the ordinary public in all aspects of life. Rather, by
focusing on social individualism, gender equality, secular politics, and basis of
identity, this chapter contributes to the evaluation of the proximity of the popula-
tion to liberal values and the type of modality of politics more likely to emerge in
the seven countries. This understanding is also useful for assessing whether the
trend in values indicates a cultural turn in the region. Finally, the difference in
value orientations between the younger and older generations provides a better
comprehension of the youth’s role in the process of change.
In the next section, we first develop measures of social individualism, gender
equality, secular politics, and the basis of identity. Because the first three con-
structs are components of liberalism, we combine measures of these components
to create a liberalism index. Then, we assess the orientations of the public toward
liberal values and how these orientations vary cross-nationally and by age group.
Social Individualism
Liberal democracy rests on the recognition of the individual as the ultimate legiti-
mizing authority in a self-governing political community. Liberal social order is
predicated on individuals’ autonomy in making choices freely that affect their lives
in the family, labor market, and politics. Understanding and recognizing individu-
alism as a given of human nature is the great virtue of liberalism. This recognition

29
of individual autonomy is a formal and principal aspect of legal order in Western

democracies. In many MENA societies, on the other hand, the institution of patri-
archy and the values supporting patrimonialism are still dominant and play a sig-
nificant role in dictating individual behavior.
The difference in the institutional development of individualism between
Western and Islamic cultural traditions has been the subject of considerable
debate among historians and Islamicists. It has been argued that, in contrast to
Western culture, Islamic cultural tradition exhibits conceptual inadequacy in the
area of individual rights, and this inadequacy explains the failure of liberal democ-
racy in Muslim-majority countries. According to Lewis (1993), Western democracy
is rooted in Roman law of the legal person, defined as a corporate entity with
certain rights and obligations. This principle provided the legal scaffolding for
the emergence of a form of council or assembly through which qualified indi-
viduals participate in the formation, conduct, and, on occasion, replacement of
the government. Over the centuries, such bodies in the West gained increasing
legislative power. Although Christianity “was forced to recognize the authority of
Roman law” (Gibb, 1947, p. 85), in Islam, in contrast, there is no recognition of
corporate persons—hence, no legislative function. And without legislative func-
tion, there is no need for legislative institutions and any principle of representa-
tion (Lewis, 1993).
The presence of cultural categories favoring patriarchy is claimed to have
undermined societal recognition of individual autonomy in Muslim- majority
countries further. It has been said that concepts such as za’im (leader) and za’ama
(leadership), as well as the associated norms, dissolve individuality in the institu-
tion of patriarchy. Although the term za’im in Arabic refers to charismatic political
leader, during the modern period it has been used by leaders of the military who
seized power in Arab countries such as Algeria, Egypt, and Iraq to give credence
to authoritarian rule (Lewis, 1988’ Sharabi, 1963; Vatikiotis, 1973). For Lapidus
(1992), on the other hand, modern authoritarianism is rooted in the second golden
age of historical Islam during which power—far from being an expression of the
total society—became the prerogative of certain individuals or groups. Because
“the exercise of power is organized through networks of clients and retainers”
(Lapidus, 1992, p. 17), there would be little room for free individual expression.
In this chapter, we measure the concept of social individualism and assess
cross-national variation in respondents’ attitudes toward individualistic values. As
is seen in the following pages, societies in the MENA region that are less authori-
tarian and more secular tend to show more widespread support for individualistic
values than those that are more authoritarian and less secular.
Gender Equality
The social status of women has been one of the most hotly contested issues
among intellectual leaders of different persuasions during the modern period. An
early proponent of women’s rights, John Millar believed that the test of civiliza-
tions is provided by the status of women. “The condition of women,” he said, “is
one of the most remarkable circumstances in the manners of nations. Among

30
the rude people, the women are generally degraded; among civilized people they
are exalted” (Millar, 1781, p. 309). In Western societies, recognition of the signifi-
cance of gender equality for building a civilized order shaped the women’s move-
ments of the 19th and 20th centuries, culminating in the formulation of laws
that banned discrimination against women and established equal treatment of
men and women before the law. In practice, however, women still face system-
atic biases and their reproductive rights are hotly debated. In the contemporary
Middle East, on the other hand, gender inequality in diverse domains of social life
such as politics, the labor market, education, and the family is tolerated, practiced,
and even sanctioned by the state.
In different historical episodes during the contemporary period, orientations
toward women in MENA countries have been a function of the nature of ideo-
logical discourses dominant in these countries. Encountering Western modernity
during the 19th century, maltreatment of women in historical Islam became the
subject of considerable debate and ideological warfare among diverse intellectual
leaders, political activists, and the ruling elite. The institution of male supremacy
and practices such as female infanticide, gender segregation, early marriage, and
polygamy were among the most visible targets of polemics and criticisms of Islam
and the traditional cultures of Muslim communities not only by Westerners and
followers of the Enlightenment, but by Christian evangelicals as well (Ahmad,
1967). Intellectual debates among the proponents and opponents of greater free-
dom for women shaped the intellectual context within which Islamic feminism
emerged during the late 19th and early 20th centuries, with the most important
spokespersons of this feminist movement in Islam being Sayyid Mumtaz Ali
(1860–1935) in India (Minault, 1990), and Qasim Amin (1865–1908) in Egypt.
The rise of territorial nationalism and the secular cultural shift during the first
quarter of the 20th century contributed to and were reinforced by the rise of wom-
en’s movements for equality in different MENA countries (Charrad, 2001; Nashat
& Tucker, 1999; Reid, 1975; Safran, 1961; Shaarāwī, 1987; Vatikiotis, 1980). These
movements also provided a context favorable to the rise of state feminism (known
also as feminims from higher authorities)—a series of policies designed to promote
and enforce Western-style dress among women, to expand female education,
and to encourage the participation of women in public affairs. State feminism
was, in fact, the hallmark of developmental policies implemented by the modern
national states in MENA, most notably in Egypt, Iran, and Turkey. In opposition
to state feminism, the harbingers of Islamic fundamentalism launched their cul-
tural assault on modernity, criticizing the state for allegedly promoting nudity and
“decadent” Western culture, rejecting the idea of gender equality and defending
gender segregation and male supremacy in the family, politics, and the labor mar-
ket (Motahhari, 1969 [1347]; Taraki, 1996).
Since the Iranian Revolution of 1979 and the rise of political Islam in MENA
countries, religious fundamentalists have had considerable impact in shaping atti-
tudes toward gender equality in these countries. However, as this chapter shows,
these attitudes vary cross-nationally and across age groups. Nations that are less
authoritarian and more secular display more egalitarian attitudes toward women
than those that are more authoritarian and less secular.

31
Secularism and Orientation toward the Secular State

The separation of religion and politics was not a serious issue in the nationalist
movements for the construction of the modern state that emerged during late 19th-
and early 20th-century MENA countries. Although the public in these countries
was religious, and Muslim theologians and activists played a leading role, these
movements were secular in the sense that the separation of religion and politics
was taken for granted. The justificatory rationale underpinning these movements
was formulated neither in reference to religion nor in opposition to it. Rather, this
rationale was in terms of the necessity of liberating the native land from foreign
domination in countries such as Algeria and Egypt, which were under French and
British colonial rule, respectively, or in opposition to monarchical absolutism and
the obstructionism of the official ulama, which were perceived as obstacles to the
formation of a responsive government, economic development, and prosperity, as
in Iran, Lebanon, Syria, and Turkey.
In addition, the favorable attitudes of these movements toward secular poli-
tics and the admission of political modernity by a substantial section of Muslim
scholars were rooted in historical changes in Islamic political theory. Accordingly,
prominent Muslim theologians-cum-political theorists in different periods gradu-
ally relaxed some of the alleged principles concerning the unity between religion
and politics in the institution of the caliphate in favor of recognizing the sultan’s
discretionary power in various interpretations of these principles, first by Abu
al-Hasan Ali Ibn al-Mawardi (972–1058), then by Imam Muhammad al-Ghazali
(1058–1111), followed by Taqi al-Din Ahmad Ibn Taymiyya (1263–1328), and finally
by Abd al-Rahman Ibn Khaldun (1333–1406). These changes in the conception of
legitimate political authority in historical Islam amounted progressively to a de
facto admission of the reality of secular politics in which the legitimate power was
differentiated between the realm of politics under the control of the sultan and the
realm of religious affairs under the supervision of the ulama.
Therefore, the idea that the state is a secular institution with a legitimacy that
is not derived or originating from religion (i.e., the notion that considered the
sultan as the shadow of God on Earth was the outcome of his military power,
not the cause of it) is not exclusively a Western invention, although the forerun-
ners of the Enlightenment were among the first to articulate the utility of secular
politics for the construction of a better society, defined as democratic, prosperous,
and egalitarian. Although it is also true that the ideology of the Enlightenment
had influenced the political ideas of a substantial number of indigenous intellec-
tual leaders in MENA, it is a mistake to think that these intellectual leaders were
passive receptors of Western ideologies. On the contrary, through their distinctly
religious reasoning, Muslim intellectual leaders, such as Egyptian Muhammad
Abduh (1949–1905), Syrian Abdul Rahman al-Kawakibi (1849–1903), the ayatol-
lahs Sayyid Abdullah Bihbahani (1840–1910) and Sayyid Muhammad Tabataba’ie
(1842–1920), who led the Iranian Constitutional Revolution in 1906, and other
prominent Shia theologians during the early 20th century, realized that the reign-
ing absolutist monarch was as despotic and corrupt as he was incompetent in
defending the Islamic community against foreign interventions. Therefore, from

32
the standpoint of religious reasoning, they opted to substitute monarchical abso-

lutism with a constitutional system (Adamiyat, 1976 [1355]; Moaddel, 2005). The
social context of the late 19th and early 20th centuries also enhanced the popular-
ity of secular politics. In many countries of the MENA region, this movement was
directed against European colonial rule, which was both foreign and non-Muslim.
The political struggle against foreigners pacified the conflict between religion and
nationalism and made it possible for “the concept of ‘la Patrie,’ ” as Hourani (1983,
p. 194) aptly commented, to conquer “without struggle.”
Secularism in 20th-century MENA, however, failed to produce a distinctly sec-
ular state that was neutral ideologically, or a secular order in which religion was
kept within the individual’s private life. Secularism turned into what Thornton
(2005) describes as developmental idealism—a set of interrelated ideas that
considered changes in culture as a condition for economic development. The
key aspect of this cultural change entailed restricting the sphere of activities
of religious institutions in society. The secularist policies implemented by the
20th-century authoritarian state, however, created a context that was favorable
to the rise of Islamic fundamentalism. On the one hand, state authoritarianism
effectively dismantled or disorganized substantially the collectivities within civil
society. As a result, oppositional politics were channeled through the medium
of religion. On the other hand, the perception of state policies as an attack on
religion and religious prerogatives of the ulama resulted in the politicization of
religion. At the same time, being associated with authoritarian rulers, secularism
not only lost much of its former appeal among intellectual leaders and the greater
public, but also formed a clear target in opposition to which religious fundamen-
talist discourse was produced during the 20th century (Moaddel 2005). If the
secular state failed to allow mass participation in politics, and if secular reason-
ing was used as a tool to attack religion, then Muslim activists during the latter
part of the 20th century opted to reject secularism in favor of the unity of religion
and politics, and to abandon rational reasoning in favor of the literal reading of
the scriptures.
The rise of religion as the dominant framework in shaping oppositional dis-
course provided a favorable opportunity for the leaders of Islamic fundamental-
ism to self-proclaim as the sole interpreting authority of the faith, and to manage
successfully to insert in popular political discourse the historically tested-and-
found-wanting idea of the unity between religion and politics. In Afghanistan,
Iran, and Saudi Arabia, Muslim political activists who seized power were able
to impose on society a despotic Islamic government. For many of these activ-
ists, the idea of secular politics reflected the Western cultural invasion of the
Muslim world. As the harbinger of Shi’i fundamentalism, Ayatollah Khomeini
(1981, p. 38) stated, “this slogan of the separation of religion and politics and
the demand that Islamic scholars not intervene in social and political affairs
have been formulated and propagated by the Imperialists; it is only the irre-
ligious who repeats them.” Likewise, as an exemplar of Sunni fundamental-
ism who espoused the ideas of Islamic government and Islamic nationalism,
Egyptian Sayyid Qutb rejects other forms of nationalism for being un-Islamic
and tyrannical:

33
The homeland (watan) a Muslim should cherish and defend is not a mere piece
of land; the collective identity he is known by is not that of a regime. … Any land
that combats the Faith, hampers Muslims from practicing their religion, or does
not apply the Shari’a, becomes ipso facto part of the Abode of War (Dar al-Harb).
It should be combated even if one’s own kith and kin, national group, capital and
commerce are to be found there. … A Muslim’s homeland is any land governed
by the laws of Islam. Islam is the only identity worthy of man. … Any other
group identity … is a jahili identity of the type humanity has known during its
periods of spiritual decadence (cited in Sivan, 1985, p. 31).
In this chapter, we develop measures of secular politics and assess the extent
to which the public is oriented toward the separation of religion and politics or
toward support of the Islamic government. We also consider whether the younger
generation is more strongly in favor of secular politics than the older generation.
Basis of Identity
Territory, ethnicity, and religion are diverse cultural categories over which intel-
lectual leaders and political activists debate to define the identity of their political
community in contemporary MENA. During the early 20th century, this issue
was resolved in terms of the territorial nation, giving rise to and reinforced by
territorial nationalist movements. These movements created modern nation-
alist regimes first in Egypt, Iran, Turkey, and then in other countries in the
region from the early 1920s onward. The period between the two world wars
witnessed the rise of pan-Arab nationalism by Iraqi and Syrian intellectual lead-
ers. This ideology, however, shaped the political discourse and behavior of the
military regimes that seized power through coups in Egypt, Iraq, Libya, and
Syria between 1952 (Egypt) and 1969 (Libya). Parallel to the rise of Islamic oppo-
sitional political discourse during the last quarter of the 20th century, religion
became a more important basis of identity than territorial nation or ethnicity.
In the perspective of the Islamic fundamentalist-cum-nationalist regimes that
seized power, the political community over which they ruled was defined decid-
edly in terms of religion.
In the following section, we evaluate whether people consider territorial nation,
ethnicity, or religion to be the basis of their identity and whether this definition
varies cross-nationally, historically, and by age group.
Cross-national Variation in Values: Egypt, Iraq, Lebanon,

Pakistan, Saudi Arabia, Tunisia, and Turkey
Complete cross-national comparable survey data are available for Egypt, Iraq,
Lebanon, Pakistan, Saudi Arabia, Tunisia, and Turkey for 2011 to 2013.2 We focus on
these seven countries to evaluate similarities and differences in value orientations
between nationally representative samples of respondents from adult populations
(18 years and older). We compare and contrast these respondents across country
and age groups (18–24 years, 25–34 years, 35–44 years, 45–54 years, 55 years and

34
older) on measures of social individualism, gender equality, secular politics, basis

of identity, and liberalism.
Measures of Social Individualism

Individualism, or the recognition of the right of choice for the individual, is a rele-
vant value in certain social domains. It exists in a variety of social practices, includ-
ing egalitarianism in politics, laissez-faire capitalism, a childrearing philosophy
that encourages the child to be independent and imaginative, individual choice
in spouse selection, and the freedom to choose one’s style of dress. Although the
equality of all voices in politics, gender equality, equality of opportunity in the
labor market, freedom to choose one’s religious faith, and individual autonomy
in lifestyle indicate various aspects of individualism, in this chapter, we narrow
the range of the construct to include only those aspects that are outside politics,
gender relations, the labor market, and religious framework. We include prefer-
ences for love as the basis for marriage, women’s right to dress as they please, and
characteristics that enhance independence as favorable qualities for children to
have. These constructs are measured as follows.
Basis for marriage: The recognition of the individual’s right to select one’s
mate, dubbed “the Romeo and Juliet revolution” (Deutsch, 1981; Huntington
1996), is a cornerstone of individualism. To measure this variable,
respondents were asked: Which of the following is the more important
basis for marriage: (1) parental approval or (2) love?
Woman dresses as she wishes: Respondents where asked: Do you (4) strongly
agree, (3) agree, (2) disagree, or (1) strongly disagree that it is up to a woman
to dress as she wishes?
Child qualities: Respondents were asked to select five from a list of 10 qualities they
considered favorable for children to have. Based on their responses, we constructed
a child quality index as the sum of four dummy variables: (a) independence
(1 = those who selected independence, 0 = otherwise), (b) imagination (1 = those
who selected imagination, 0 = otherwise), (c) nonreligious faith (0 = those who
selected religious faith, 1 = otherwise), and (d) nonobedience (0 = those who
selected obedience, 1 = otherwise), resulting in a variable with a possible range
of 0 to 4 points.
We then constructed a social individualism index by averaging these three

measures.3 A greater value on this measure indicates a stronger orientation
toward social individualism.
The measures of social individualism varied considerably across the seven
countries and age groups. Figure 2.1 shows variation in preference for love as the
basis for marriage. Lebanon was the most individualistic, with 69% of respon-
dents endorsing love, whereas Pakistan was the least individualistic, with just 7%
favoring love as the basis for marriage. The other countries fell in between; 54%
of Turkish, 46% of Saudis, 31% of Iraqis, 29% of Egyptians, and 26% of Tunisians
considered love as a more important basis for marriage than parental approval.
Considering variation by age groups, a greater percentage of the respondents in

35
90%
80%
70% Egypt (29%; −.19**)
60% Iraq (31%; −.20**)
Respondents
50% Lebanon (69%; −.21**)

40% Pakistan (7%; −.16**)
30% Saudi Arabia (46%; −.33**)
20% Tunisia (26%; −.11**)
Turkey (54%; −.20**)
10%
0%
18−24 25−34 35−44 45−54 55+
Age (yr)
Figure 2.1 Percentage of respondents who reported love as the most important basis
for marriage.
(% for country; r with age group; *p<.01, *p<.05)
the younger age groups considered love as the basis for marriage than in the older
age groups, and age correlated negatively with preference for love. Saudi respon-
dents displayed a much greater age difference than respondents from other coun-
tries (r = –.33), whereas Tunisia showed the most homogeneity across age groups,
with the lowest correlation between age and love marriage endorsement (r = –.11).
Correlation coefficients for other countries ranged from –.21 to –.19.
Respondents from the seven countries also expressed varying levels of support
for women’s freedom to dress as they wish, with the greatest endorsement coming
from Tunisia (56%), Turkey (53%), Lebanon (50%), and Saudi Arabia (47%), and
least from Iraq (27%), Pakistan (22%), and Egypt (14%). As Figure 2.2 shows, the
difference between the youngest and oldest age groups also varies by country, but
only Pakistani participants’ responses varied significantly by age (r = –.26).
Furthermore, countries and age groups vary on the child quality index, as
shown in Figure 2.3. Lebanese, with a mean index of 2.5, are most individualistic,
70%
60%
Egypt (14%; −.03)
50% Iraq (27%; −.05**)
Respondents
40% Lebanon (50%; −.07**)

Pakistan (22%; −.26**)
30%
Saudi Arabia (47%; −.03)
20% Tunisia (56%; −.03)
10% Turkey (53%; −.07**)
0%
18−24 25−34 35−44 45−54 55+
Age (yr)
Figure 2.2 Percent of respondents who indicated they strongly agree/agree with the
statement: It is up to a woman to dress as she wishes.

36
2.80
2.60 Egypt (1.71; .00)
Child quality index 2.40
Iraq (1.79; −.01)
2.20
2.00 Lebanon (2.50; −.10**)
1.80 Pakistan (1.71; −.35**)
1.60
Saudi Arabia (2.11; −.06*)
1.40
1.20 Turkey (2.25; −.19**)
1.00
18−24 25−34 35−44 45−54 55+
Age (yr)
Figure 2.3 Child quality index.
whereas Egyptians and Pakistanis are least individualistic, each with a mean
index of 1.71. In between are Turkey (mean, 2.25) and Saudi Arabia (mean, 2.11),
with moderate scores. An analysis of variance showed that variance in the child
quality index across countries was significant (F = 554.359, p < .001). Post hoc
analyses using the Scheffé’s test for significance indicated statistically significant
differences among all countries with the exception of Pakistan and Egypt, which
were nearly identical. With the exception of Egypt and Iraq, where attitudes are
relatively stable across all age groups, participants’ attitudes concerning desirable
child qualities and the age groups was related negatively, with younger age groups
reporting more individualistic attitudes than older age groups. Pakistan displayed
the largest age-driven difference (r = –.35), followed by Turkey (r = –.19), Lebanon
(r = –.10), and Saudi Arabia (r = –.06).
Last, combining these three indicators to construct a social individual-
ism index, we examine the values of this index by country and age group in
Figure 2.4.4 The analysis of variance of the social individualism index was signif-
icant among the seven countries (F = 980.340, p < .001), and post hoc analyses
3.00
2.80
Egypt (1.76; −.15**)
Social individualism index
2.60
2.40 Iraq (1.89; −.17**)
2.20 Lebanon (2.68; −.20**)
2.00 Pakistan (1.61; −.38**)
1.80 Saudi Arabia (2.31; −.25**)
1.60 Tunisia (2.23; −.10**)
1.40 Turkey (2.46; −.24**)
1.20
1.00
18−24 25−34 35−44 45−54 55+
Age (yr)
Figure 2.4 Index of social individualism.

37
using the Scheffé’s test for significance indicated statistically significant differ-
ences among all countries. With an index value of 2.68, Lebanon is the most
individualistic country, followed by Turkey, 2.46; Saudi Arabia, 2.31; Tunisia,
2.23; Iraq, 1.89; Egypt, 1.76; and Pakistan, 1.61. Although it was expected that
respondents from Lebanon, Turkey, and Tunisia, which are relatively more dem-
ocratic and open countries, would be more supportive of social individualism
than the less democratic countries, the Saudis’ comparatively high support is
paradoxical. The Saudis are quite conservative in terms of attitudes toward gen-
der equality and support for religious fundamentalism (Moaddel & Karabenick,
2013), yet their score on our social individualism index is much greater than that
of Egyptians, Iraqis, or Pakistanis, and is similar to Tunisians, who are relatively
more liberal and secular.
One plausible explanation for this paradox may be related to varying levels of
existential security that may be a consequence of the differences in the level of
economic conditions among these countries. In a prosperous economy with secure
employment and income stability, people are in a more favorable subjective condi-
tion to think and reflect about marriage and love. They are also better situated to
make independent financial decisions. As a result, in economically well-off coun-
tries such as Lebanon, Saudi Arabia, Tunisia, and Turkey, people may tend to favor
individualistic values. Egypt, Iraq, and Pakistan, on the other hand, face relatively
poor economic conditions, which tend to generate a feeling of insecurity. Such con-
ditions may prompt individuals to fall back on tradition and seek refuge in the pro-
tective hierarchical shield of family relations (Inglehart, Moaddel, & Tessler, 2006;
Inglehart & Welzel, 2005). Furthermore, a greater level of economic prosperity
allows people to travel abroad more often and learn about alternative lifestyles.
These processes may, in turn, weaken one’s attachment to conservative values.
There is also considerable variation across the age groups among the seven
countries, as shown in Figure 2.4. Without exception, younger age groups are more
individualistic than older. All correlation coefficients between the social individu-
alism index and age groups in each country are statistically significant (p < .001)
and negative—meaning, the younger age groups are more individualistic than
the older age groups. The strength of this correlation is highest among Pakistanis
(r = –.38), followed by Saudis (r = –.25), Turkish (r = –.24), Lebanese (r = –.20),
Iraqis (r = –.17), Egyptians (r = –.15), and Tunisians (r = –.10). A comparison of the Z
scores calculated through the Fisher transformation confirms that the correlation
between age and social individualism in Pakistan is statistically different—in this
case, stronger—than in the other countries (p < .01). We speculate that intergen-
erational differences in social individualism contribute significantly to the clash of
values in Pakistan and Saudi Arabia, because the youths in these two conservative
societies appear to differ considerably with the older age groups on issues related
to individual choice in child qualities, style of dress, and marriage.
Measures of Attitudes toward Gender Equality

To assess variations in attitudes toward gender equality, we first measured the
construct in different spheres of social life, then created a gender equality index

38
and assessed its variation across the five age groups in the seven countries. We
considered five survey questions asked in Likert-scale format: Do you (1) strongly
agree, (2) agree, (3) disagree, or (4) strongly disagree that (a) a wife must always
obey her husband (obedience); (b) men make better political leaders than women
(political leadership); (c) when jobs are scarce, men should have more rights to a job
than women (job market); (d) university education is more important for boys than
it is for girls (university education), and (e) it is acceptable for a man to have more
than one wife (polygamy). These variables were averaged to create a single gender
equality index, ranging from 1 to 4 points, with a greater value indicating stronger
support for gender equality.
Table 2.1 below shows cross-national variation in attitudes toward gender rela-
tions in these five areas. Considering wife obedience first, 6% of Egyptians, 11%
of Iraqis, 39% of Lebanese, 8% of Pakistanis, 21% of Saudis, 22% of Tunisians,
and 30% of Turkish citizens either disagreed or strongly disagreed with the state-
ment: A wife must always obey her husband. On this measure, the majority of the
respondents from the seven countries—even in Lebanon, which is considered the
most liberal country in the region—supported male supremacy in the family. On
political leadership, however, respondents across the seven countries were less
uniform. The majority of respondents considered men as better political lead-
ers and gave priority to men over women in a tight job market, but Lebanese,
Tunisians, and Turks favored men less strongly than respondents from the other
countries. The majority of the respondents, however, disagreed with the idea of
gender inequality in university education. Only in Saudi Arabia (where only 44%
Table 2.1 Measures of Gender Equality
Variable Egypt Iraq Lebanon Pakistan Saudi Tunisia Turkey

Arabia
A wife must 6% 11% 39% 8% 21% 22% 30%

always obey her
husband
Men make better 17% 25% 44% 30% 21% 45% 46%
political leaders
Men should have 14% 22% 34% 16% 22% 26% 45%
more of a right
to a job
A university edu- 65% 71% 74% 50% 44% 80% 71%
cation is more
important for
boys
It is OK for men 70% 53% 71% 75% 50% 81% 93%
to have more
than one wife
Sample size 3,018 2,714 2,757 3,398 1,453 2,871 2,743
note: Results measured as a percentage of respondents who reported “disagree” or “strongly

disagree.”

39
of participants strongly disagreed or disagreed that “university education is more

important for boys than it is for girls”) and in Pakistan (where 50% of participants
disagreed with gender inequality in the educational realm) did participants tend
to endorse education inequality by gender. The majority of the respondents did
not support polygamy, with levels of disapproval ranging from 50% among Saudis
to 93% among Turkish respondents. These rates are unsurprising, for of the five
different domains of social life where male supremacy is exercised, the practice of
polygamy is the weakest in virtually all seven countries.
We argue that inegalitarian attitudes toward women are linked to gender dis-
crimination in society. For example, in Saudi Arabia, although 58% of university
students are women, they make up only 5% of the labor force (Cordesman, 2003,
pp. 175–176; Moaddel, 2006). The marked differential participation in the labor mar-
ket by gender corresponds to a strong bias against women among the Saudi public;
78% Saudi respondents give more rights to men over women in a tight job market.
Figure 2.5 shows variation in the gender equality index by country and age
group. An analysis of variance showed a significant difference in the gender equal-
ity index across the seven countries (F = 563.882, p < .001). Post hoc analyses
using Scheffé’s test indicated that Saudis and Egyptians, with a mean index value
of 2.05 and 2.07, respectively, showed the least egalitarian orientation. Pakistanis
and Iraqis did not differ significantly and (after Saudis and Egyptians) were the
least egalitarian, with mean index values of 2.20 and 2.22, respectively. Tunisians
and Lebanese, with index levels of 2.50 and 2.58, respectively, both reported more
egalitarian attitudes, although Turks showed the most egalitarian orientation
toward women, with a mean index of 2.73.
There is also variation by age across the seven countries. Considering the dif-
ference between the youngest and oldest age groups, Pakistani respondents again
displayed the greatest disparity, showing the highest correlation coefficient among
gender equality index and age groups (r = –0.34) of all seven countries. Other coef-
ficients ranged from r = –.13 (Iraq) to r = –.05 (Egypt). These coefficients, however,
are all statistically significant, indicating that younger age groups had more egali-
tarian attitudes toward women than older age groups (Figure 2.5).
2.90
Egypt (2.07; −.05**)
2.70
Iraq (2.22; −.13**)
Gender equality index
2.50
Lebanon (2.58; −.08**)
2.30
Pakistan (2.20; −.34**)
2.10
Saudi Arabia (2.05; −.11**)
1.90
Tunisia (2.50; −.07**)
1.70
Turkey (2.73; −.09**)
1.50
18−24 25−34 35−44 45−54 55+
Age (yr)
Figure 2.5 Index of gender equality.

40
Variation in Attitudes toward Secularism

To measure orientations toward secular politics, we analyzed responses to several
survey questions, including the following, that asked respondents their attitudes
toward the separation of religion and politics, Islamic government, and the shari’a:
1. Do you (1) strongly agree, (2) agree, (3) disagree, or (4) strongly disagree that
[Study Country] would be a better place if religion and politics were sepa-
rated? (separation of religion and politics)
2. Would it be (1) very good, (2) fairly good, (3) fairly bad, or (4) very bad for
[Study Country] to have an Islamic government where religious authori-
ties have absolute power [ for Muslim respondents]/to have a Christian
government where religious authorities have absolute power [ for Christian
respondents]? (religious government)
3. Is it (1) very important, (2) important, (3) somewhat important, (4) least impor-
tant, or (5) not at all important for a good government to implement only the
shari’a law [ for Muslim respondents]/to implement only the laws inspired by
Christian values [ for Christian respondents]? (shari’a/Christian values)5
Responses to the first question were recoded so that greater values indicated
stronger support for the separation of religion and politics. We then constructed a
secular politics index by averaging responses to these questions, ranging between
1 point and 4 points, with a greater value indicating stronger support for secular
politics. Because the first two questions were not allowed to be included in the
Saudi survey, a secular politics index could not be determined for Saudi Arabia.
According to Table 2.2, with the exception of Pakistan, the majority of respon-
dents favored the separation of religion and politics. As with other issues, however,
these countries vary in terms of their public support for secular politics, ranging
between 51% of Egyptian and 80% of Lebanese respondents strongly agreeing or
agreeing their country would be a better place if religion and politics were separated.6
Egypt was the only country in which a clear majority, 59%, considered it very
good or fairly good to have an Islamic government. This fact may explain the
Table 2.2 Measures of Secular Politics

Variable Egypt Iraq Lebanon Pakistan Saudi Tunisia Turkey
Arabia
Separation of 51% 69% 80% 9% — 72% 76%

religion and
politics (agree/
strongly agree)
Religious govern- 59% 49% 31% 33% — 37% 34%
ment (very/
fairly good)
Implement 56% 48% 24% 74% 68% 27% 20%
sharia (very/
important)
Sample size 2,843 2,484 2,705 3,336 1,594 2,744 2,415

41
success of the Muslim Brothers and other Islamic groups in the 2011 elections.
This percentage dropped to 49% for Iraqis, and was between 30% and 40% for
respondents in the other countries. Except for Egypt and Iraq, only a minority of
the respondents was favorable to Islamic government. The relatively low support
for Islamic government among Pakistani respondents, despite having a conserva-
tive orientation on other issues, may be a result of its resonance with the Taliban-
style system of rule.7
Favorability toward the implementation of the shari’a laws was greater, with a
majority of Egyptians (56%), Pakistanis (74%), and Saudis (68%) reporting that
the implementation of shari’a law is very important or important. Conversely, only
a minority of Lebanese (24%), Tunisian (27%), and Turkish (20%) citizens shared
this sentiment. This value for Iraqis was 48%, indicating they were somewhat
divided over this issue.
Figure 2.6 shows cross-national variation in the secular politics index (as an
average of the three measures) and by age group for each country. An analysis of
variance showed a strong group effect (F = 1,164.936, p < .001). Post hoc analyses
using Scheffé’s test of significance produced three groups—Turkish, Lebanese,
and Tunisian respondents—with mean secular politics indices of 3.08, 3.02, and
2.96, respectively, which displayed the strongest support for secular politics among
the seven countries. Iraqis and Egyptians, with indices of 2.58 and 2.37, reported
significantly lower levels of endorsement of secular politics, and Pakistanis (with
an index value of 1.96) reported the lowest level of support for secular politics.
Of all the countries sampled, Pakistani respondents had the greatest dif-
ference in orientation toward secular politics between the younger and the
older age categories (r = –.32), again confirmed by post hoc analysis. The sig-
nificant strength of the correlation with age is also observed in the indices
of social individualism and gender equality among Pakistanis. Lebanese and
Tunisian respondents also had a noticeable difference between the younger
and older cohorts—as reflected by age–secularism correlations of –.13 and –.09,
respectively—with younger people reporting more secularism. The correlations
between age cohort and the secular politics index were not significant among
3.30
3.10 Egypt (2.37; .03)
2.90
Secular politics index
Iraq (2.58; −.02)

2.70
2.50 Lebanon (3.02; −.13**)
2.30 Pakistan (1.96; −.32**)
2.10
Tunisia (2.96; −.09**)
1.90
Turkey (3.08; .06**)
1.70
1.50
18−24 25−34 35−44 45−54 55+
Age (yr)
Figure 2.6 Index of secular politics.

42
Egyptians and Iraqis. Although we could not calculate an index for Saudis, the
relationship between age cohort and preference for shari’a law was also signifi-
cantly negative (r = –.09), with youths reporting more secular attitudes (not
shown). In Turkey, on the other hand, this correlation coefficient was reversed
(r = .06), which means that the older respondents, who came of age when secu-
larism was popular, appear to support secular politics more strongly than the
younger generation, which may be influenced by the ruling Islamic-oriented
Justice and Development Party.
Liberalism Index
Because the indices of social individualism, gender equality, and secular politics
correlated positively, we created a liberalism index by averaging these indices.8
Greater values on this measure indicate stronger liberal orientations and vary
between 1 point and 4 points. Post hoc analysis of variance (F = 1,498.536, p < .001)
confirms significant differences in the liberalism index by country, with Scheffé’s
test showing that Pakistan is the least liberal country, with a mean index value of
1.93. Egypt is the second to least liberal country (2.05), followed by Saudi Arabia
(2.18) and Iraq (2.21), which are not significantly different, and then by Tunisia
(2.55). Turkey and Lebanon are the most liberal countries, with index values of
2.73 and 2.75, respectively, and do not differ significantly.
Across the seven countries, the youngest age group is consistently more liberal
than the oldest, as shown by the values of the correlation coefficient reported in
Figure 2.7, with the strongest correlation coefficient among Pakistanis (r = –.44)
and the least among Egyptians (r = –.11). As we would expect from the pattern seen
thus far, the size of the correlation between liberalism and age among Pakistanis
differs significantly from that in the other countries, as does the size of the cor-
relation among Saudis (r = –.25), whereas the correlations in the other countries
were more similar. However, the correlations between liberalism and age group
were negative and significant across all seven countries, indicating that younger
participants have stronger liberal orientations.
3.10
2.90 Egypt (2.05; −.11)
2.70 Iraq (2.21; −.16**)
Liberalism index
2.50 Lebanon (2.75; −.18**)

2.30 Pakistan (1.93; −.44**)
2.10
Saudi Arabia (2.18; −.25**)
1.90
Tunisia (2.55; −.12**)
1.70
Turkey (2.73; −.16**)
1.50
1.30
18−24 25−34 35−44 45−54 55+
Age (yr)
Figure 2.7 Liberalism index.

43
Identity Measurement
The basis of identity was measured by asking respondents whether they defined
themselves in terms of their nation, ethnicity, or religion. Respondents were
asked: Which of the following best describes you: (1) I am an Egyptian, Iraqi,
Lebanese … , above all; (2) I am a Muslim (Christian [ for Christian respondents]),
above all; and (3) I am an Arab, Kurd, Berber … , above all. They were also asked
about the different ways in which they related to the world, by selecting among
the following five statements: (a) I see myself as a citizen of the world, (b) I see
myself as a citizen of [Country], (c) I see myself as a citizen of the Islamic umma
(worldwide Christian community [ for Christian respondents]), or (d) I see myself
as a citizen of the Arab, Berber, or Kurdish community.
Because findings from the seven country surveys showed that a great majority
of the respondents either identified with their religion or with their nation, we
report only the contrast between the respondents’ identification with nation and
religion. With the exception of Pakistan and Tunisia, the percentage of respon-
dents who considered nation as the basis of their identity is greater than those
who considered religion. For Pakistanis and Tunisians, this order was reversed,
with 70% of Pakistanis and 59% of Tunisians defining themselves as Muslim
above all.
Figure 2.8 displays two interesting features. One is that the majority of Saudis,
whose country is a bastion of conservative Islam, identified with the nation rather
than with their religion. The second is that a clear majority of Tunisian respon-
dents, despite having relatively liberal orientations on many other issues, saw
themselves first and foremost as Muslims rather than Tunisians. It is not clear
why Tunisian national identity is weaker than it is in most other countries. The
second survey question on identity, which taps into the contrast between one’s
recognition of national community and supranational religious community, tells
a similar story; compared with other nations, Tunisians are much less likely to
recognize their nation as the basis of their identity. The disparity between the
Tunisian respondents who see themselves as citizens of Tunisia versus citizens of
the Islamic umma (or worldwide Christian or Jewish communities for Christian
or Jewish respondents, respectively) is 46% versus 43%, which is much smaller
80% 70%
57% 60% 59%
60% 52%
47% 48% 46% 44%
36% 39%
40% 31%
29% 28%
20%
0%
Egypt Iraq Lebanon Pakistan Saudi Arabia Tunisia Turkey
Country
I am [National Identity] above all else. I am Muslim/Christian above all else.
Figure 2.8 National versus religious identity.

44
1.00
Secular national identity index

.90 Egypt (.59; .02)
.80 Iraq (.75, −.05*)

Lebanon (.76; −.10**)
.70
Pakistan (.53; −.21**)
.60
Saudi Arabia (.66; −.03)
.50 Tunisia (.49; .03)
.40 Turkey (.69; .06**)
.30
18−24 25−34 35−44 45−54 55+
Age (yr)
Figure 2.9 Index of secular national identity.
than it is for other countries, where the disparity is much greater: 61% versus 36%
for Egyptians, 70% versus 15% for Iraqis, 59% versus 19% for Lebanese, 65% ver-
sus 24% for Saudis, and 62% versus 23% for Turkish (not shown).
We created a secular national identity index first by coding those who identified
with religion or the religious community as 0 and other participants as 1 for both
variables. Then, we averaged the two variables so that a greater value indicates a
stronger identification with secular nationalism. Figure 2.9 shows variations in
the secular nationalism index across the seven countries and age groups.
According to Figure 2.9, the mean secular national identity index varies cross-
nationally and by age group. The mean value of this index is highest among
Lebanese and Iraqis (.76 and .75, respectively), followed by Turks (.69), Saudis
(.66), Egyptians (.59), and, last, Pakistanis and Tunisians (.53 and .49, respec-
tively). Post hoc analysis of variance (F = 235.045) shows that index values among
countries are significantly different, with the exception of the Lebanese and
Iraqis at one end, and the Pakistanis and Tunisians at the other. The relationship
between age group and secular national identity index is also negative among
some of these countries, as shown by the correlation coefficients between the two
measures. Again, the relationship is strongest among Pakistanis (r = –.21) and dif-
fers significantly from that in the other countries. The relationship was also rela-
tively strong and negative in Lebanon (r = –.10) and Iraq (r = –.05), and positive in
Turkey (r =.06). Age and secular national identity are not correlated significantly
in Egypt, Saudi Arabia, and Tunisia.
Trends in Values: Egypt, Iraq, Saudi Arabia, and Turkey

The foregoing analysis shows that the younger generation voiced greater support
for the values of social individualism, gender equality, and liberalism than the older
members of the population. Some other measures, such as the secular politics index
and the secular national identity index, showed inconsistent relationships with age

45
groups across the seven countries. The younger generation was more secular only
in Lebanon, Pakistan, and Tunisia, and they identified more often with nation,
rather than religion, in Lebanon and Pakistan only. In contrast, on these two indices,
in Turkey the younger generation tended to be less secular and identified less often
with nation. It is not clear, however, whether these variations by age group reflect
simply age effects (where people’s value orientations change as they grow older) or
cohort effects (representing trends in values, where the younger generation has a
distinctly new experience in life compared with the older generation). Of the seven
countries, Pakistan stood alone in displaying dramatic gaps in value orientations
between the youngest and oldest age groups. Therefore, we speculate that these dif-
ferences reflect cohort effects rather than age effects.
To assess changes in values and whether these changes reflect a trend in other
countries, it is necessary to have nationally representative sample data on the spec-
ified indicators (in this case, social individualism, gender equality, secular politics,
and basis of identity) for at least three points in time. We had such data available
for Egypt, Iraq, and, to a limited extent, Turkey, but data from Saudi Arabia were
available for only two points in time. Despite these limitations, we present the
available data to make a reasonable estimate of the trend in values among differ-
ent age groups during the past decade.
Social Individualism
The only measure of social individualism for which data were available for several
countries is the belief about the basis for marriage. Data on this measure were
collected in Iraq in 2004, 2006, and 2011; Saudi Arabia in 2003 and 2011, and
Turkey in 2013 and 2015. As depicted in Table 2.3, support for love as the basis
for marriage increased among Iraqis from 27% in 2004 and 28% in 2006 to 31%
in 2011. Among Saudis, it remained stable between 2003 and 2011. And, among
Turkish respondents, it dropped from 54% in 2013 to 46% in 2015—a significant
decline. It is, however, important to note that the data on only two points in time
is not adequate to establish a trend in value.
The correlation between age groups and preference for love as the basis for
marriage is negative across these surveys, indicating that younger age groups are
more individualistic than older age groups. The magnitude of this relationship
increased among Iraqis from –.08 in 2004, to –.12 in 2006, and to –.20 in 2011;
among Saudis from –.08 in 2003 to –.33 in 2011; and among Turkish respondents
from –.20 in 2013 to –.25 in 2015. Post hoc analyses of differences in correlation
coefficients across time using the Fisher transformation shows that the change in
the strength of correlation was statistically significant between 2006 and 2011 in
Iraq, and between 2003 and 2011 in Saudi Arabia. This increase in the size of the
correlation thus shows a widening gap between the younger and older age groups
among the surveys (Table 2.3).
Gender Equality
Longitudinal data on the measures of gender equality are available for Egypt,
Iraq, Saudi Arabia, and Turkey. These measures, however, are available across the

46
Table 2.3 Trends in Social Individualism: Love is the Basis for Marriage
Country Year Age Total N r (with age

groups)
18–24 25–34 35–44 45–54 55+
Iraq 2004 32% 25% 30% 21% 19% 27% 2,258 –.08a
2006 35% 31% 26% 22% 19% 28% 2,615 –.12a
2011 44% 34% 26% 21% 17% 31% 2,935 –.20a
Change: 2004–2011 39% 35% –13% 0% –9% 15%
Saudi Arabia 2003 53% 54% 43% 39% 51% 49% 999 –.08c
2011 65% 53% 39% 20% 20% 47% 1,552 –.33a
Change: 2003–2011 23% –1% –11% –49% –61% –4%
Turkey 2013 68% 62% 53% 47% 40% 54% 2,945 –.20a
2015 61% 56% 43% 41% 25% 46% 915 –.25a
Change: 2013–2015 –10% –10% –19% –13% –38% –15%
a
<.001.
b
<.01.
c
<.05.
47
countries and waves of surveys. For effective comparison, we use the only measure
of gender equality that is common across the four countries and for more than
one wave. This measure refers to respondents’ assessment of women’s political
leadership compared with men: Do you strongly agree, agree, disagree, or strongly
disagree that men make better political leaders than women?
Trends in responses vary by country and age group (Table 2.4). On the aggre-
gate level, the attitudes of Saudis remained stable, whereas among Egyptian,
Iraqi, and Turkish respondents, they grew more egalitarian. The mean response
to whether men make better political leaders rose in Egypt from 1.63 in 2001 to
1.84 in 2014, an increase of about 13% (1 indicates “strongly agree” and 4 indicates
“strongly disagree”); in Iraq from 1.42 to 1.95 between 2004 and 2011, an increase
of 37%; and in Turkey, where more egalitarian attitudes prevail, from 2.27 to 2.63
between 1996 and 2015, an increase of 16%. However, among Saudis there was
essentially no shift in attitudes.
Trends in gender equality by age varied in Egypt and Saudi Arabia as well.
Among Egyptians, there was no change among those age 18 to 24 years, but a
significant increase (between 12% and 20%) across other age groups. Overall, this
trend resulted in a leveling off in the distribution of attitudes toward gender equal-
ity by age group. Among Iraqi and Turkish citizens, the rate of change across age
groups remained about the same (if we disregard those between 25 years and
34 years among Iraqis, and those 55 years and older among Turkish citizens).
In Saudi Arabia, although there was an increase in egalitarian attitudes among
younger age groups, those individuals 45 years and older grew more conservative
between the two surveys.
We may conclude that, among Saudis, older age groups have developed stron-
ger inegalitarian attitudes toward women whereas the younger age groups dis-
played an opposite trend. In Iraq and Turkey, all age groups developed consistently
more egalitarian attitudes. Finally, in Egypt, the youngest age group has remained
unchanged whereas the older age groups have become more egalitarian in their
attitudes toward women.
Secular Politics
Only one indicator of secular politics was available for multiple years for Egypt,
Iraq, and Turkey. This indicator measured attitudes toward the separation of reli-
gion and politics: Do you strongly agree, agree, disagree, or strongly disagree that
[Study Country] would be a better place if religion and politics were separate?
We used these data to assess trends in attitudes toward secular politics and to
make a cross-national comparison of trends. For Saudi Arabia, data on attitudes
toward the sharia—Do you consider it (1) very important, (2) important, (3) not
very import, or (4) not at all important for a good government to implement only
the laws of the sharia?—is used as a proxy measure for attitudes toward secular
politics (Table 2.5).
Egypt and Iraq have experienced a significant increase in favorable attitudes
toward secular politics. Saudi Arabia also displays a similar change, assuming a
trend in attitudes toward the shari’a reflects a trend toward secular politics. That is,

48
Table 2.4 Trends in Gender Equality: Men Make Better Political Leaders
Variable Age Total N r (with age group)
18–24 25–34 35–44 45–54 55+
Egypt
2001 1.79 1.67 1.53 1.56 1.53 1.63 2,882 –.110a
2011, summer 1.66 1.66 1.64 1.65 1.61 1.65 3,079 –.015
2011, fall 1.52 1.54 1.64 1.57 1.53 1.56 1,663 .007
2014 1.80 1.89 1.85 1.87 1.72 1.84 1,340 –.021
Change: 2001–2014 1% 13% 21% 20% 12% 13%
Iraq
2004 1.46 1.43 1.41 1.41 1.40 1.42 2,244 –.022
2006 1.43 1.45 1.40 1.51 1.35 1.43 2,622 –.012
2011 1.99 2.02 1.92 1.86 1.92 1.95 2,919 –.046b
Change: 2004–2011 36% 41% 36% 32% 37% 37%
Saudi Arabia
2003 1.84 1.83 1.70 1.68 2.21 1.79 930 –.011
2011 1.89 1.81 1.82 1.57 1.72 1.81 1,613 –.085b
Change: 2003–2011 3% –1% 7% –7% –22% 1%
Turkey
1996 2.33 2.33 2.30 2.22 2.03 2.27 1,782 –.090a
2001 2.36 2.36 2.30 2.27 2.20 2.31 3,293 –.053b
2007 2.38 2.39 2.30 2.26 2.18 2.33 1,294 –.072b
2013 2.49 2.54 2.47 2.44 2.43 2.48 2,866 –.038c
2015 2.67 2.67 2.61 2.56 2.60 2.63 914 –.035
Change: 1996–2015 15% 15% 13% 15% 28% 16%
note. Higher values indicate a greater endorsement of egalitarian attitudes.

a
<.001.
b
<.01.
c
<.05.
49
Table 2.5 Trends in Secular Politics
Variable Age Total N r (with age group)
18–24 25–34 35–44 45–54 55+
Egypt: [Egypt] would be better if religion and politics were separate.

2007 2.42 2.32 2.49 2.43 2.45 2.41 735 .019
2011, summer 2.58 2.62 2.68 2.66 2.70 2.65 2,939 .034
2011, fall 2.61 2.55 2.58 2.76 2.94 2.67 1,535 .111a
2014 3.12 3.18 3.16 3.29 3.28 3.20 1,279 .062c
Change: 2007–2014 29% 37% 27% 35% 34% 33%
Iraq: [Iraq] would be better if religion and politics were separate.
2004 2.55 2.58 2.67 2.70 2.78 2.63 2,021 .104a
2006 2.89 2.85 2.78 2.84 3.04 2.86 2,268 .064b
2011 2.93 2.91 2.94 2.97 2.82 2.92 2,790 –.011
Change: 2004–2011 15% 13% 10% 10% 1% 11%
Turkey: [Turkey] would be better if religion and politics were separate.
2013 3.06 3.14 3.01 3.07 3.11 3.08 2,764 .003
2015 2.71 2.59 2.77 2.92 3.04 2.79 890 .124a
Change: 2013–2015 –11% –18% –8% –5% –2% –9%
Saudi Arabia: A good government implements only the shari’a.
2003* 1.38 1.32 1.32 1.31 1.23 1.33 970 –.047
2011* 1.84 1.83 1.71 1.76 1.62 1.82 1,594 –.085b
Change: 200–-2011 33% 39% 30% 34% 32% 37%
note: Higher values indicate greater endorsement of secularism. In computing means, all variables that used response categories 1
to 5 (i.e., a Likert scale with a middle category) were rescaled into variables ranging from 1 to 4 using the formula 1 × 0.75 + 0.25 = 1
and 5 × 0.75 + 0.25 = 4. Surveys in which variables were rescaled using this formula are noted with an asterisk.
a
<.001.
b
<.01.
c
<.05.
50
there has been a significant decline in favorable attitudes toward the implementa-
tion of the sharia as a characteristic of the government. In Turkey, on the other
hand, there has been a decline in favorable attitudes toward separation of religion
and government between 2013 and 2015.
In Egypt and Saudi Arabia, attitudes change toward secular politics did not dis-
play a consistent pattern across age groups. In Iraq, younger age groups displayed
slightly greater percentages of change than older age groups, and in Turkey the
younger age groups grew more conservative.
National Identity
Consistent with changes in attitudes toward social individualism, gender
equality (except among Saudis), and secular politics (except among citizens of
Turkey), our evidence indicates a change in conception of identity from religion
to nation. In our data, identity was measured by responses to the question of
whether respondents primarily define themselves as (a) Egyptian, Iraqi, Saudi, or
Turkish; (b) Muslim [Christian for Christian respondents]; or (c) Arab or Kurd.
Table 2.6 reports these responses for Egypt, Iraq, Saudi Arabia, and Turkey in
different years.
According to this table, across all the countries in the past decades, a much
greater percentage of the respondents identified with their nation than with their
religion in the later surveys than they did in the earlier surveys, indicating a sig-
nificant shift away from religion and toward nationality as the basis of identity.
In Egypt, only a small percentage of respondents in the 2001 and 2007 surveys
defined itself as Egyptian (10% and 12%, respectively), whereas the majority con-
sidered itself primarily as Muslim. This percentage increased dramatically to 52%
in summer 2011, 56% in fall 2011, and 61% in 2014. This increase may suggest
a rise of national identity within the context of the collective struggle against
the authoritarian, nonresponsive, and nontransparent state under President
Mubarak. The other countries for which data were available also experienced a
rise in national identity during the same period. Among Iraqis, those considering
themselves Iraqi above all rose from between 24% and 28% in 2004 to 2006,
to 57% and 46% in 2011 and 2013, respectively. Among Saudis, those defining
themselves as Saudi jumped from 17% in 2003 to 48% in 2011, and in Turkey,
identification with the nation rose from 34% in 2001 to 44% in 2013, before a
slight decrease to 40% in 2015.
Table 2.6 also reports the increase in the percentage of respondents adhering
to national identity between the latest and earliest surveys for the five countries.
Among Egyptians, there was a very significant increase across all age groups. The
magnitude of this increase was much greater among those 18 to 26 years old: 611%
(but less than 600% across all other age groups). Among Iraqis and Saudis, this
increase did not show a consistent pattern across all age groups. In Turkey, on
the other hand, adherence to national identity declined among the youngest age
group, but increased significantly across other age groups.

51
Table 2.6 Trends in National Identity
Country Date Age Total N r (with age group)
18–24 25–34 35–44 45–54 55+
Egypt 2001 9% 10% 11% 11% 8% 10% 3,000 .004

2007 8% 8% 14% 20% 18% 12% 989 .133a
2011, summer 49% 50% 51% 56% 53% 52% 3,136 .037c
2011, fall 59% 53% 52% 60% 59% 56% 1,675 .020
2014 64% 61% 63% 61% 53% 61% 1,382 –.052
Change: 2001–2014 611% 510% 473% 455% 563% 510%
Iraq 2004 24% 19% 27% 26% 24% 24% 2,307 .032
2006 28% 29% 29% 32% 24% 28% 2,646 –.008
2011 56% 54% 62% 60% 52% 57% 2,949 .014
2013 47% 46% 40% 50% 51% 46% 932 –.071c
Change: 2004–2013 99% 142% 50% 89% 114% 92%
Saudi Arabia 2003 14% 16% 18% 18% 15% 17% 1,024 .026
2011 50% 50% 46% 47% 49% 48% 1,494 –.023
Change: 2003–2011 251% 207% 160% 159% 232% 182%
Turkey 2001 34% 34% 36% 33% 32% 34% 3,188 –.011
2013 39% 40% 44% 47% 48% 44% 2,980 .065a
2015 30% 36% 44% 43% 49% 40% 930 .133a
Change: 2001–2015 –12% 6% 22% 30% 53% 18%
note: Results measured as a percentage of reporting, “Above all, I am [nationality].”

a
<.001.
b
<.01.
c
<.05.
52
Discussion
The analysis of comparative cross-national survey data from the seven countries in
2011 to 2013 showed, first, that considerable cross-national variation exists in peo-
ple’s orientations toward social individualism, gender equality, and secular poli-
tics. Combining these measures to create a liberalism index, the distribution of
the aggregate data across the seven countries indicated a major division between
those country respondents who are more secular and less authoritarian, including
Lebanese, Tunisians, and Turks; and those who are less secular and more authori-
tarian, including Egyptians, Iraqis, Pakistanis, and Saudis. Second, although
identity preferences varied across nations, except for Pakistan and Tunisia, only a
sizable minority defined itself in religious rather than national terms. The analysis
of the variation in value orientation across age groups showed that younger people
(a) are more individualistic; (b) display more favorable attitudes toward gender
equality; (c) are more secular in Lebanon, Pakistan, and Tunisia, but are less so
in Turkey; and (d) more often identify with nation than religion in Iraq, Lebanon,
and Pakistan, but less so in Turkey.
Our analysis of the longitudinal data from Egypt, Iraq, Saudi Arabia, and
Turkey also showed variations in trends in values across these countries. As
one of the countries that experienced the overthrow of its ruler as a result of
the Arab Spring movement, Egypt has displayed dramatic changes in values.
The most significant of such changes was in the definition of identity among
Egyptians. Although during the pre-Arab Spring period, Egyptians defined
themselves primarily as Muslim above all, after the overthrow of President
Hosni Mubarak, their conception of identity turned nationalistic; more than
60% of the respondents identified as Egyptian rather than Muslim in the after-
math of the Arab Spring. The second most important change in attitude was
toward secular politics, with a minority in 2007, but a clear majority in 2014,
expressing support for the separation of religion and politics. Finally, Egyptian
attitudes toward gender equality turned more egalitarian after the Arab Spring,
although a majority still holds inegalitarian attitudes. Except for the change
toward national identity, the attitudes among youths were not more dramatic
than among older Egyptians.
Likewise, among Iraqis, the most dramatic changes in attitudes were also
toward national identity; only a minority defined itself as Muslim whereas a major-
ity considered itself Iraqi above all—or, among Kurdish respondents, as Kurdish
above all. Iraqis’ attitudes also became more favorable of individualistic values
and gender equality, although a majority among them, as among Egyptians, holds
inegalitarian attitudes toward women. These changes in attitudes among Iraqis
were stronger among youth only vis-à-vis individualistic values and, to a lesser
extent, secular politics.
Among Saudis, the most dramatic change in attitudes was toward national
identity; only a minority of Saudis defined itself as Muslim above all. Among
Saudi youth, there is evidence of increased individualism and gender egalitarian-
ism over time.

53
Among Turkish citizens, the most important trend was toward national iden-
tity, followed by support for gender equality. There was, however, a reverse trend
away from supporting individualistic values or secular politics. Overall, the avail-
able data showed an inconsistent pattern of change among Turkish respondents,
and the role that youths played was even more inconsistent. As in the cases of
secular politics and national identity, the younger generations turned toward more
conservative values.
All in all, three general claims may be advanced. First, if the Arab Spring was
to signify the commencement of a new cultural episode in Arab-majority coun-
tries, this cultural episode is characterized by (a) the decline of political Islam and
(b) the rise of nationalism. Second, changes in values were event driven rather
than driven by economic development. In Turkey, where there has been substan-
tial economic growth in past decades, this growth has yet to be translated into a
significant change in values. Conversely, among countries where there has been
little economic development in recent decades, such as in Egypt, there has been a
change in values concurrent with the events of the Arab Spring. Finally, the role
of youths in spearheading changes in values has been exaggerated; our empirical
data demonstrate that youths are not unique among the general population when
assessing changes in values and attitudes over time.
Notes
1. In Chapter 8, Moaddel found that people who lived their formative years (i.e.,
18–42 years old) under President Hosni Mubarak are reported to have had a greater
participation in the revolutionary movements against Mubarak than those who
were older.
2. See Chapter 10 in this volume, in which de Jong and Young-DeMarco discuss
in detail the methodological aspects of the seven-country comparative survey project.
See also Moaddel (2013) for further details on the demographic composition of the
samples in each country.
3. Before doing so, the ranges of the measures of the basis of marriage and child
quality index are transformed to values ranging between 1 and 4. This was done by
recoding the dummy variable on the basis of marriage from 1 to 4, where love is coded as
4 and parental approval is 1, and by condensing the child quality index to range from
1 to 4 using the following formula: [(Child quality + 1) × (3/4)] + 1/4. The question
about child qualities was excluded from the Tunisian questionnaire, and the social
individualism index we created includes only variables on the basis for marriage and
women’s dress for Tunisian respondents.
4. Because Tunisia had no data on child quality measures, which prevented the
construction of a child quality index, its overall measure of social individualism is not
comparable with the other six countries.
5. To maintain consistency, the range of the last variable was converted from
a scale ranging from 1 to 5 to a scale ranging from 1 to 4 by multiplying this variable
by 0.75 then adding 0.25 (i.e., 1 × 0.75 + 0.25 = 1 and 5 × 0.75 + 0.25 = 4).

54
6. Given that, in our data, 81% of Iraqi Sunni Arabs agree that Iraq would be a
better place if religion and politics were separated (not shown), then the presence of
ISIL (or Daesh, which is the Arabic acronym for the group), which rests on religious
absolutism of the caliphate, is puzzling. We contend that the terror group’s success in
the country was made possible by the political space created as a result of a remark-
able conjunction of disparate regional and international forces. Although posing a
varying level of danger to the political stability of all the regimes in the region, ISIL’s
rise is benefited by conflicting priorities of some of the key regional and international
actors: (a) the sectarianism of Iraqi Shia Nouri al-Maliki that alienated a substantial
section of Iraqi Sunni Arabs who might have been led to believe that ISIL was the
lesser of the two evils, their strong support for secular politics notwithstanding; (b)
the preoccupation of the Saudi-led Persian Gulf Arab states trying to overthrow the
Syrian regime of Bashar al-Assad, the defeat of the Houthi rebels in Yemen, and
Iran’s conflicting regional ambitions, which pushed the ISIL threat to the sideline;
(c) Turkish President Recep Tayyip Erdoğan’s strategic calculations that considered
Kurdish separatists and the Assad regime as more serious enemies than the ISIL,
which prompted him to allow his country to serve as a transit for foreign fighters
pouring into Syria; and, finally, (d) the lack of desire or interest on the part of the
Obama Administration and other Western countries to get involved in yet another
conflict in the region. As a result, the ISIL terror group was not only able to solidify
its presence in Iraq and Syria, but also it managed to expand its influence in Egypt,
Libya, and Yemen. However, it appears that the only group that has undermined
ISIL is ISIL itself. The gruesome beheading of Western hostages in front of a video
camera, the massacre of religious minorities and the Shia, the raping and enslave-
ment of Yazidi women, and participation in terrorist activities abroad all turned world
public opinion against ISIL and prompted the formation of a broad-based consensus
to defeat the terror group militarily.
7. Consistent with this view is a 2014 Pew Research Center poll in Pakistan that
found little support for extremist groups in Pakistan, with only 8% giving the Taliban
a positive rating and just 12% saying they had a favorable opinion of al Qaeda (Pew
Research Center, 2014).
8. Using data pooled from all seven countries, these three indices make one fac-
tor with an Eigen value of 1.76 and with a Cronbach’s alpha of 0.64.
References
Adamiyat, F. (1976 [1355]). Idi’olozhi-ye nahzat-i mashrutiyat-i Iran [The ideology of the
constitutional movement in Iran]. Tehran: Payam Publications.
Ahmad, A. (1967). Islamic modernism in India and Pakistan: 1857–1964. London: Oxford
University Press.
Boudon, R. (2001). The origins of values: Essays in the sociology and philosophy of beliefs.
New York: Transaction Publishers.
Campante, F. R., & Chor, D. (2012). Why was the Arab world poised for revolution?
Schooling, economic opportunities, and the Arab Spring. Journal of Economic
Perspectives, 27(2), 167–187.

55
Chaaban, J. (2009). Youth and development in the Arab countries: The need for a
different approach. Middle Eastern Studies, 45(1), 33–55.
Charrad, M. (2001). States and women’s rights: The making of postcolonial Tunisia,
Algeria, and Morocco. Berkeley: University of California Press.
Cordesman, A. H. (2003). Saudi Arabia enters the twenty-first century. Westport,
CT: Praeger.
Deutsch, W. K. (1981). On nationalism, world religion and the nature of the West.
In P. Torsvik (Ed.). Mobilization, center-periphery structures and national building
(pp. 51–93). Bergen, Norway: Universitestsvorlaget.
Durkheim, E. (1893). The division of labor in society. Glencoe, IL: Free Press.
Geras, N. (1983). Marx’s theory of human nature: Refutation of a legend. London: Verso.
Gibb, H. A. R. (1947). Modern Trends in Islam. Chicago: University of Chicago Press.
Harrison, L. E., & Huntington, S. P. (Eds.). (2000). Cultures and matters: How values
shape human progress. New York: Basic Books.
Hoffman, M., & Jamal, A. (2012). The youth and the Arab Spring: Cohort differences
and similarities. Middle East Law and Governance, 4(1), 168–188.
Hourani, A. (1983). Arabic thought in the liberal age (1798–1939). Cambridge: Cambridge
University Press.
Huntington, S. P. (1996). The West unique, not universal. Foreign Affairs,
75(6), 28–46.
Hvistendahl, M. (2011). Young and restless can be a volatile mix. Science, 333, 552–554.
Inglehart, R., Moaddel, M., & Tessler, M. (2006). Xenophobia and in-group solidarity
in Iraq: A natural experiment on the impact of insecurity. Perspective on Politics,
4(3), 495–505.
Inglehart, R., & Welzel, C. (2005). Modernization, cultural change and democracy.
Cambridge: Cambridge University Press.
Khomeini, R. (1981). Islam and revolution: Writings and declarations of Imam Khomeini
(H. Algar, Trans.). Berkeley, CA: Mizan Press.
LaGraffe, D. (2012). The youth bulge in Egypt: An intersection of demographics, secu-
rity, and the Arab Spring. Journal of Strategic Security, 5(2), 65–979.
Lapidus, I. M. (1992). The Golden Age: The political concepts of Islam. Annals of the
American Academy of Political and Social Science, 524(November), 13–25.
Lewis, B. D. (1988). The Political Language of Islam. Chicago: University of
Chicago Press.
———. (1993). Islam and liberal democracy. Atlantic Monthly, 271, 2 (February), 89–98.
Millar, J. (1781). The origin of the distinction of ranks; or, an inquiry into the circumstances
which give rise to influence and authority in the different members of society (3rd ed.).
London: J. Murray.
Minault, G. (1990). Sayyid Mumtaz Ali and ‘Huquq un-Niswan’: An advocate of wom-
en’s rights in Islam in the late nineteenth century. Modern Asian Studies, 24(1),
147–172.
Moaddel, M. (2005). Islamic modernism, nationalism, and fundamentalism: Episode and
discourse. Chicago: University of Chicago Press.
———. (2006). The Saudi public speaks: Religion, gender, and politics. International
Journal of Middle East Studies, 38, 79–108.

56
———. (2013). The birthplace of the Arab Spring: Values and perceptions of Tunisians and
a comparative assessment of Egyptian, Iraqi, Lebanese, Pakistani, Saudi, Tunisian,
and Turkish publics [On- line]. Available: http://mevs.org/files/tmp/Tunisia_
FinalReport.pdf. Retrieved September 18, 2015.
Moaddel, M., & Karabenick, S. (2015). Cross-national analysis of religious fundamental-
ism in seven countries. Paper presented at the annual meeting of the American
Sociological Association, Chicago, IL, August 21–25.
Motahhari, M. (1969 [1347]). Mas’aleh-ye hijab [The problem of veiling]. Tehran: Islamic
Society of Physicians.
Nashat, G., & Tucker, J. E. (1999). Women in the Middle East and North Africa: Restoring
women to history. Bloomington: Indiana University Press.
Parsons, T. (1949). The structure of social action: A study in social theory with special refer-
ence to a group of recent European writers. Glencoe, IL: Free Press.
———. (1951). The social system. Toronto: Collier-Macmillan.
———. (1966). Societies: Evolutionary and comparative perspective. Englewood Cliffs,
NJ: Prentice-Hall.
Pew Research Center. (2014). “A Less Gloomy Mood in Pakistan.” Available: http://
www.pewglobal.org/2014/08/27/a-less-gloomy-mood-in-pakistan/. Retrieved October
20, 2015.
Reid, D. M. (1975). The Odyssey of Farah Antun: A Syrian Christian’s quest for secularism.
Chicago: Bibliotheca Islamica.
Rokeach, M. (1971). Long-range experimental modification of values, attitudes, and
behavior. American Psychologist, 25(5), 453–459.
———. (1973). The nature of human values. New York: Free Press.
Safran, N. (1961). Egypt in search of political community. Cambridge, MA: Harvard
University Press.
Sahlins, M. D., & Service, E. R. (1960). Evolution and culture. Ann Arbor: University
of Michigan Press.
Schwartz, S. H. (2006). A theory of cultural value orientations: Explication and appli-
cations. Comparative Sociology, 5, 136–182.
———. (2012). An overview of the Schwartz theory of basic values. Online Readings in
Psychology and Culture, 2(1). http://dx.doi.org/10.9707/2307-0919.1116.
Service, E. R. (1971). Cultural evolutionism: Theory in practice. New York: Holt, Rinehart
& Winston.
Shaarāwī, H. (1987). Harem years: The memoirs of an Egyptian feminist (1879–1924),
translated and edited by M. Badran. Feminist Press at the City University of
New York.
Sharabi, H. (1963). Power and leadership in the Arab world. Orbis, 7(3), 583–595.
Sivan, E. (1985). Radical Islam: Medieval theology and modem politics. New Haven,
CT: Yale University Press.
Taraki, L. (1996). Jordanian Islamists and the agenda for women: Between discourse
and practice.” Middle Eastern Studies, 32(1), 140–158.

57
Thornton, A. (2005). Reading history sideways: The fallacy and enduring impact of the
developmental paradigm on family life. Chicago: University of Chicago Press.
Vatikiotis, P. J. (1973). Tradition and Political Leadership: The Example of Algeria.
In I. W. Zartman (Ed.). Man, State, and Society in the Contemporary Maghrib
(pp. 309–29). New York: Praeger.
———. (1980). The history of Egypt. Baltimore, MD: The Johns Hopkins
University Press.
Weber, M. (1958 [1905]). The Protestant ethic and the spirit of capitalism. New York:
Scribner’s.

58
59
SECTION II Perspectives on Change

Development
and Modernization
60
61
CHAPTER 3 National Identity Versus National

Pride in the Modalities of Liberal
Territorial Nationalism and Islamic
Nationalism in Muslim-Majority
Countries
Mansoor Moaddel
Anthony D. Smith may be accused of making a sweeping historical generalization

in describing nationalism at the beginning of his remarkable book Nationalism
and Modernism as “a single red line” that “traverses the history of the modern
world from the fall of the Bastille to the fall of the Berlin Wall” (Smith, 1998, p. 1).
For sure, his generalization is defensible insofar as the term applies to a variety of
movements that (a) define as a nation a population of individuals based on certain
signified characteristics shared by these individuals, (b) bestow on this popula-
tion the right to organize a political regime, and (c) advance an exclusive claim to
represent them. Nonetheless, Smith’s generalization overlooks the fact that the
nationalist movements that emerged on the world stage in different places during
the modern period have been too diverse and heterogeneous to be represented
adequately by a single line with only one color. Even in a single region of the
world and within the confines of a much narrower historical period, such as 20th-
century Middle East, nationalism covers such dissimilar movements as territorial
nationalism, pan-Arab nationalism, and religious nationalism-cum fundamental-
ism. Although all are instances of nationalism, these movements differ vastly in
their sociopolitical and cultural outlooks, and the type of regimes they shaped.
Lumping them together under the umbrella of nationalism may lead one to gloss
over serious differences among these and other varieties of nationalism that may
coexist even in the same society, giving rise to different historical outcomes.
Some scholars prefer the application of the term nationalism to territorial
nationalism only and other cases such as pan-Arab nationalism, religious nation-
alism, or pan-Africanism, which cut across national boundaries and are rooted in
race, ethnicity, or religion rather than territorial nation, as alternatives to a nation-
alist project (e.g., Aminzade, 2013; Gershoni & Jankowski, 1995). Other schol-
ars, in contrast, have widely used the terms pan-Arab nationalism and religious
nationalism. Religious nationalism, in particular, has been used to capture the
political specificity of the Islamic fundamentalist movements that have emerged
62
in the Middle East in recent decades (e.g., Friedland, 2001, 2002; Fox, 2004;
Juergensmeyer, 1996). Although the analysis presented in this chapter is expected
to demonstrate the merits of the latter position, my intention is to use the concept
of modality as one way to manage more effectively the variety of the nationalist
movements that may exist contemporaneously in the same society, and to demon-
strate its fruitfulness for a better understanding of this diversity.
The concept of modality categorizes heterogeneous cases of nationalism into
relatively homogeneous types. This classification thus reduces and simplifies his-
torical complexity and contributes to a better comprehension of the variety of the
nationalist movements. Modalities represent different historical configurations of
the principles of collective sovereignty. I suggest that modalities are distinguished
from one another in terms of the conception of the basis of identity of belong-
ing. They are also associated with different feelings of group solidarity or emo-
tional expression of belonging to a national community. I propose that these two
features of modalities predict other variable aspects of collective sovereignty, the
demarcation of social boundaries that separate in-groups from outgroups, the pro-
jection of political map and national territories, and status valuations of different
members of the political community, including women. Modalities thus serve as
a transmission belt connecting the general concept of nationalism to particular
historical cases.
Modalities are constrained by social structures and exigencies of historical
development. However, they cannot be derived from these structures or exigen-
cies. Rather, modalities are produced as intellectual leaders address the problem
of collective sovereignty and attempt to resolve sociopolitical issues facing their
communities. Variation in modalities is thus a function of the different ways in
which these issues are resolved. Modalities are distinguished from one another
in terms of their conception of the basis of identity and the manner in which the
feeling of national solidarity is mobilized to promote certain cultural and political
values among the subject population. Modalities thus represent competing under-
standings of the nature of collective sovereignty.
To demonstrate the fruitfulness of these theoretical propositions, I first pres-
ent territorial nationalism, pan-Arab nationalism, and religious nationalism as
instances of modalities of collective sovereignty. These modalities are illumi-
nated when they are contrasted with one another. Then, I argue that these modal-
ities are the outcomes of the efforts of intellectual leaders to resolve historically
significant sociopolitical issues related to the problem of collective sovereignty.
Next, I focus on liberal territorial nationalism and religious nationalism as two
currently competing modalities in Muslim-majority countries, and I argue that
changes in identity and national pride are associated with changes in attitudes
toward significant sociopolitical and cultural issues. More specifically, I propose
that the difference between people who considered religion as the primary basis
of their identity and those who consider nation as such is associated with signifi-
cant differences in their value orientations toward gender equality, secularism
and secular politics, democracy, Western culture, and outsiders, whereas national
pride has just the opposite relationships with all these variables. The population
may thus be configured into those leaning toward a liberal modality of collective
62 | Perspectives on Change

63
sovereignty and those inclined to religious nationalism. To assess these proposi-

tions, I use data from 12 surveys carried out in Algeria, Bangladesh, Indonesia,
Iran 2000 and 2005, Iraq 2004 and 2006, Jordan, Lebanon, Morocco, Saudi
Arabia, and Turkey. Finally, I discuss the implications of the findings for the
study of nationalism.
Theoretical Development
The social–scientific literature on the causes and consequences of nationalism is
vast and engaging. Nonetheless, there has been little theorizing about how to han-
dle the diversity of nationalism. Many of the existing explanations are formulated
in terms that are too general to account for this diversity. They refer to historical
changes such as industrialization, the expansion of mass education, print capital-
ism, and uneven capitalist development in different theories within the modernist
tradition; to substantialist accounts of the rise of nation in various versions of the
primordialism perspective; and to cycles of nationalism in varied perennial inter-
pretations. It thus appears that whatever the forms of nationalism—territorial,
linguistic-cum-ethnic, or religious—they are all somehow related to the condi-
tions of modernity, the primordial attachments underpinning the formation of
nations, or the perennial features of historical cycles. Even then, there is little
mechanismic explanation that connects these aspects of social conditions to the
genesis of nationalism.
Building their theories on the premise that the modern era is, ipso facto, an
age of nationalism (Smith, 1998, pp. 35–36), proponents of the modernist para-
digm link the origins of nationalism to (a) the social dislocations caused by eco-
nomic transformation of ethnically divided empires in which language serves
as a medium for turning the ensued social conflict into a nationalist movement
(Gellner, 1964, 1983); (b) uneven development that creates center–periphery divi-
sion in the world economy (Nairn, 1977) or internal colonialism in the center
(Hechter, 1977), prompting the elite in the periphery or the subordinated groups
in the center to engage in nationalist activities; (c) the expansion of networks of
discursive literacy by the agencies of the military state and capitalist development,
which serve as a medium for the rise of national communities (Mann, 1993);
(d) the state’s bureaucratic expansion that creates a chasm between the state and
society, shaping political conflict, and the rise of movements for national self-
determination (Breuilly, 1994); (e) the pulverization of traditional society and
regimentation of the colony caused by imperialism, which lead to the decline of
traditional industry on the one hand and to the rise of mass literacy and the emer-
gence of new marginal people who embrace Western ideals of self-determination
on the other (Kedourie, 1971); and (f) the decline of the sacred language and king-
ship, the rise of the Reformation, the standardization of the vernacular as a tool
for administrative centralization, and the rise of print capitalism that generates
unified fields of communication, gives a new fixity to language, and creates lan-
guages of power—all contributing to the rise of nation as “an imagined commu-
nity” (Anderson, 1983, pp. 4, 44–45).
National Identity Versus National Pride In The Modalities | 63

64
The primordial and perennial perspectives decouple nationalism and moder-

nity. Their alternative accounts, however, fly even more widely over the historical
horizon and thus fail to capture variation in nationalism. Primordialism seeks the
roots of nationalism in either people’s genetic relatedness or cultural givens. In
the former, kinship, ethnicity, and nation are forms of cooperation that expand
genetic relatedness, enhance inclusive fitness, and thus improve an organism’s
overall reproductive success (Hamilton, 1964). Nation is made possible by the cul-
tural inventions of unilinear descent and lineage exogamy that extended the “pri-
mordial model of social organization to much larger societies running into tens
of thousands of people” (Smith, 1998, p. 147; van de Berghe, 1978, pp. 403–404).
Given that altruism is a function of genetic relatedness (Burnstein et al., 1994;
Korchmaros & Kenny, 2001), with nation being an extension of kin, this theory
readily explains self-sacrifice vis-à-vis threats from other nations. The cultural ver-
sion treats nationalism as an outgrowth of a primordial attachment, stemming
from the assumed givens “of social existence: immediate contiguity and kin con-
nection mainly, [which] are seen to have an ineffable, and at times, overpowering
coerciveness in and of themselves” (Geertz, 1973, pp. 259–260).
Likewise, perennialists do not appear to be concerned with historical variation
of nationalism. They differ with primordialists in rejecting the naturalness and
immutability of ethnicity and nationalism. They consider ethnicity and nation-
alism as perennial phenomena that emerge, transform, and vanish throughout
history. An earliest exemplar of this approach is by Ibn Khaldun (1967 [1377])
who relates the rise and fall of political dynasties in historical Islam to changes
in the conditions of ethnic solidarity (i.e., asabiyya). Nonetheless, the perennialist
time perspective, like that of the primordialists, is too vast to allow for a system-
atic understanding of variation in nationalism in a much more narrow histori-
cal context. They do, however, acknowledge that nationalism can take varied and
heterogeneous forms: “organic and political; rational and irrational; contractual
and inherited; stable and changeable; conflictual, competitive and cooperative”
(Fishman & Gertner, 1985, p. 494; see also Armstrong, 1982, p. 4; Connor, 1994,
p. 202; Fishman, 1972, p. 497; Smith, 1998, p. 159).
Modalities of Collective Sovereignty

A proposition shared by all sides of the debate is the social structure–nationalism
nexus. Although the operational definitions of social structure vary among theo-
rists, all attempt to connect the genesis of nationalism to conditions of modernity,
to the primordialism of ethnic structures, or to the nature of ethnic solidarity. This
proposition is problematic, however, for if the emergence of nationalism is unpre-
dictable, if there is a general admission regarding the difficulty of defining the
concept of nation, and if there is considerable historical variation in the national-
ist movements (Anderson, 1983; Fishman, 1972, p. 494; Hobsbawm, 1977, p. 3;
Kohn, 1967; Özkirimli, 2000, pp. 57–60; Seton-Watson, 1977, p. 5; Smith, 1998,
p. 4), then it would be hard to connect nationalism to conditions of modernity, eth-
nicity, or some presumed preexisting primordial quality of nations. Furthermore,

65
given that the same social structure may support diverse forms of nationalism,
that different societies sustain nationalist movements that are similar, and that
people shift loyalty from one modality of collective sovereignty to another in a
relatively short timeframe, it may be necessary to depart from the premise that
nationalism is caused by social structure in a determinate way. This is true, how-
ever, notwithstanding the significance of social structure in selecting and limiting
the rise and institutionalization of different modalities.
Given the malleability of culture, as the critics of primordialism have noted
(Brass, 1991; Eller & Coughlan, 1993; Eriksen, 1993), and the difficulty of connecting
the production of nationalist ideas to the objective conditions of modernity, then
one may be justified in searching for the emergence of nationalism in processes
that signify a category of people as a nation and produce the ideas supporting
the right of these people to form a political community. Theorists of nationalism
have long recognized the fluidity of the phenomenon of nation, realizing that
nation is a socially constructed entity, that “nationalism … invents nations where
they do not exist” (Gellner, 1964, p. 168), that nation is an “invented tradition”
(Hobsbawm & Ranger, 1983, p. 1), and that nation is “an imagined political com-
munity” (Anderson, 1983, p. 6).1 The invention, imagination, or as Smith (1983,
p. 3) has argued, reconstruction of nations out of the cultural values, norms, and
identities that were inherited from premodern eras are, however, euphemisms for
the processes in which the production of the ideas of collective sovereignty is the
starting point. The producers of these ideas thus stand at the initiating phase in
the genesis of nationalism. This phase begins when intellectual leaders start to
reflect on the problem of collective sovereignty or “problematized” the parameters
of the existing modality currently in power.
The production of the modalities of collective sovereignty is a dynamic pro-
cess; it transpires within the context of discussions, debates, and conflict among
diverse intellectual leaders on issues related to the identity of their community, the
status of different individuals within that community (including women) and its
boundaries, and how to establish a sovereign political regime. Issues are resolved
in oppositional relation to the ideology of the ruling regime. The latter thus consti-
tutes the target of criticism by intellectual leaders, while at the same time serves as
a key factor shaping the production of oppositional nationalist discourses. These
leaders are, in fact, seriously engaged in “the war of positions,” to use Gramsci’s
(2007, p. 168) terminology, against the ideologues of the ruling regime. On a
more concrete level, one such a target is the presence of foreign threat or foreign
domination, in opposition to which nationalist ideas are produced. The form and
intensity of these responses are moderated by the form of this threat or type of
this domination—whether it is directly cultural, political, economic, or the actual
occupation of the country, or whether this domination is perceived to be indirect,
when a domestic regime is allied with a foreign power. An indigenous regime
may also face nationalist oppositions when that regime is monolithic ideologi-
cally. That is, an overly secularist regime begets a religious oppositional response,
and a religious authoritarian regime tends to contribute to the rise of liberal ter-
ritorial nationalism. This oppositional context thus determines whether collective
identity is defined in terms of nation, language, ethnicity, or religion. Variation in

66
domination thus produces variation in nationalist oppositional discourses, gener-

ating different modalities of collective sovereignty.
The genesis of nationalism is also an affective process as its shapes the feel-
ings of collective solidarity and pride. This is true because nationalism is an activ-
ist ideology. It demands from the faithful investments in time, money, and even
one’s life. Thus, the affective dimension of nationalism—the sense of belonging
to a national community; the feeling of pride in its people, territory, history, and
achievements; and the love of the country and hate of the outsider—may thus
contribute to the core emotion that motivates its champions. As the revolutionary
icon of 20th-century Latin America Che Guevara put it, “at the risk of seeming
ridiculous, let me say that the true revolutionary is guided by a great feeling of
love. It is impossible to think of a genuine revolutionary lacking this quality” (Che
Guevara 1967, p. 169).
The affective side of nationalism, however, does not form a coherent set of
emotions. Rather, national pride is said to be manifested in conflicting or even
contradictory emotional expressions. As it has been suggested, “love of country
can be generous, compassionate, and intelligent, but it can also be exclusive,
deaf, and blind” (Viroli, 1995, p. 6; see also Aminzade, 2013). Nationalism and
patriotism are thus proposed to capture these contradictory feelings; although
the former taps into the feeling and perception of the superiority of one’s
nation and xenophobia, patriotism is believed to have no such connection and
relates only to love of country, its cultural heritage, and historical achievements.
De Figueiredo and Elkins (2003) elaborate further and measure the concept
of national pride along the nationalism and patriotism dimensions. Assessing
empirically the relations of these dimensions with attitudes toward immigrants,
they conclude that “while nationalists have a strong predilection for hostility
toward immigrants, patriots show no more prejudice than does the average
citizen” (p. 171).
In this chapter, I use national pride as having varying relations with different
modalities of collective sovereignty. I propose that national pride is linked more
strongly to the modalities that are overly concerned with foreign enemies and
conspiracies by outsiders than liberal nationalist ideologies that tend to be more
self-referential and rest on the critique of civil society, patriarchal values, and pat-
rimonial domination. The liberal component of nationalism often uncovers faults
and deficiencies in the existing social order and institutional practices as factors
hampering the formation of egalitarian social relationships. Such criticisms tend
to lower national pride.
I use these theoretical propositions to explain the emergence of such modali-
ties of collective sovereignty as territorial nationalism, pan-Arab nationalism, and
religious nationalism in Muslim-majority countries. I propose that these modali-
ties are produced in oppositional relation to different ideological targets: liberal
territorial nationalism to the ideology of monarchical absolutism, ulama (Muslim
theologians) obstructionism, and foreign occupation; pan-Arab nationalism to
the European domination of Arab societies and the creation of disparate states;
and religious nationalism to secularist policies of the nationalist authoritarian
interventionist state. Then, I consider the two modalities of liberal territorial

67
nationalism and religious nationalism, and assess how variation in identity and
national pride predicts sociopolitical and cultural values.
Finally, I test a modernity thesis that relates the support for liberal nationalist
values to socioeconomic status.
A Two-Pronged Methodological Strategy

I use a two-pronged methodological strategy to assess these propositions. The
first presents a comparative historical analysis of the modalities of (liberal) ter-
ritorial nationalism, pan- Arab nationalism, and religious nationalism. This
analysis outlines the differences in the historical contexts that prompted indig-
enous intellectual leaders to produce these diverse nationalist discourses. The
second strategy focuses on the modalities of Islamic nationalism and territorial
nationalism. I assess the linkages of national identity, national pride, and socio-
economic status with attitudes toward gender equality, secularism and secular
politics, democracy, Western culture, and outsiders, using data from a dozen
national values surveys carried out in Muslim-majority countries between 2000
and 2008.
Territorial Nationalism, Pan-Arab Nationalism,

and Religious Nationalism
Among the variety of the political regimes that emerged during 20th-century
Middle East and North Africa, three different types stand out: (a) the territorial
nationalist regimes inaugurated most notably in Egypt, Iran, and Turkey almost
simultaneously between 1919 and 1925, which identified with the territorial nation;
(b) the pan-Arab nationalist regimes that seized power through military coups in
Egypt, Iraq, Libya, and Syria between 1952 (Egypt) and 1969 (Libya), which rested
on the claim that Arabs constituted one nation; and (c) the religious nationalist
regimes, exemplified by the Islamic Republic of Iran, the Taliban in Afghanistan,
the National Islamic Front in the Sudan, the Justice and Development Party in
Turkey, the Ennahda in Tunisia, and the Muslim Brothers in Egypt—all stressed
primarily Islam as the identifying mark of their community.
These three types represented the institutionalization of the modalities of terri-
torial nationalism, pan-Arab nationalism, and religious nationalism, respectively.
These modalities were produced by indigenous intellectual leaders many years
before the inauguration of these regimes and reflected different ways in which
these leaders resolved issues related to the basis of identity of their political com-
munity, form of government, the relationship between religion and politics, the
status of women, and the nature of the outside world—the West, in particular.
Understanding how the modalities of nationalism are produced provides a better
insight into the process of political transformation. At the same time, given that
the production of modalities precedes political formation in time, understanding
the development of modalities may provide clues concerning the kind of political
regime most likely to emerge in the future.

68
Territorial Nationalism
A set of historical factors preceded the rise of territorial nationalism in such
diverse places as Algeria, Egypt, Iran, Syria, and Turkey. Included among these
factors were changes in the structure of domestic production, the development of
capitalism, the integration of the domestic economy into the world capitalist mar-
ket, the decline of traditional social institutions (including organized religion), the
rise of new social classes (merchants, landowners, and, to a limited extent, indus-
trialists), the introduction of the press, and the expansion of modern education
(Abrahamian, 1982; Crecelius, 1972; Floor, 1976; Heyworth-Dunne, 1968; Issawi,
1966; Lapidus, 1988; Ma’oz, 1968; Marsot, 1984; Robinson, 1993; Ruedy, 1992;
Vatikiotis, 1973). These changes, however, only set the stage for the rise of territo-
rial nationalism. The proximate condition that shaped this nationalist discourse
was a distinctive alignment of political and cultural forces. Broadly speaking, this
alignment provided two major ideological targets in opposition to which (liberal)
territorial nationalist discourse was produced. One was the political discourses of
the ruling regime; the other was the traditionalist religious-cum-political discourse
of the conservative ulama. This alignment certainly varied in Algeria, Egypt, Iran,
and Turkey. Algeria was under French colonial rule (1848–1962) and Egypt was
under the British (1882–1922). These colonial powers justified their rule on the
grounds that Algerians and Egyptians were not quite ready to self-govern (Cromer,
1908; Marsot, 1968; Milner, 1892; Ruedy, 1992). The colonial discourses of the
French and British thus constituted the target of nationalist agitations. Although
the conservative ulama resisted reforms, including the introduction of modern
education and Western liberal–nationalist ideas and lifestyle, which entailed a
shift in the basis of loyalty from religion to nation, the struggle against the domi-
neering colonial power, which was both foreign and non-Muslim, had pacified
the conflict between religion and nationalism and thus made it possible for “the
concept of ‘la Patrie,’ ” in Hourani’s apt remark, to conquer “without struggle”
(Hourani, 1983, p. 194).
Iran and Turkey, on the other hand, were under monarchical absolutism, the
power of which had become too arbitrary in running the affairs of the nation and
too weak to defend the country’s interests vis-à-vis European intervention during
the 19th century (Amanat, 1997; Kinross, 1977; Moaddel, 2005). Aligned with the
ruling monarch were the conservative ulama, who had blocked the reformist mea-
sures initiated within or outside the ruling regime during the second half of the
19th century. Opposition to the ulama provoked secular discourse whereas opposi-
tion to monarchical absolutism created awareness concerning the wickedness of
political despotism on the one hand, and the utility of a constitutional representa-
tive system for the country’s technological progress and economic prosperity on
the other. The Allies’ occupation of Turkey and the ensuing Turkish war of inde-
pendence (1919–1923) accentuated the significance of the nationalist element in
the modality of territorial nationalism among Turkish intellectual leaders. Among
Iranians, on the other hand, the nationalist element was much weaker because
the country was not under colonial domination (Affary, 1996; Ahmed, 1960;
Christelow, 1985, 1987; Deeb, 1997; Hourani, 1983; Marsot, 1968, 1977; Moaddel,
2005; Reid, 1996; Ruedy, 1992; Safran, 1961; Wendell, 1972; Zürcher, 2004).

69
In territorial nationalism, political sovereignty belongs to the people who

inhabit a given territory. Connections to a territory, rather than religious affilia-
tions or ethnicity, define membership in the political community. The nation is the
source of legitimacy, religious and political functions are differentiated, and indi-
vidual identity rests on the territorial nation, like being an Algerian, Egyptian, or
Iranian. An example that underscores the significance of the land in the discourse
of territorial nationalism is the manner in which Mustafa Kamil (1874–1908), a
prominent harbinger of Egyptian nationalism, glorified Egypt in a language suf-
fused with a strong feeling of affection. For him, it was not language or religion,
but the feeling of belonging to the land of Egypt that was the most potent unifying
force among Egyptians:
Nationalism is a sentiment before which all nations and all communities bow
because it is the feeling of the worth and dignity of man, of the bounty of God
and His care, of the meaning of existence itself. …
Nationalism is the food which the body and soul of Egypt need before any
other food. … It is the mainspring of all miracles and the principle of all prog-
ress … it is the blood in the veins of nations and the life of all living things. …
Nationalism is the noblest tie for men and the solid foundation upon which
great and mighty kingdoms are built. … Life is merely transitory and it has no
honor without nationalism and without work for the welfare of the fatherland
and its children. … Fatherland, O fatherland: To you my love and my heart. To
you my life and my existence. To you my blood and my soul. To you my mind
and my speech. … You, you, O Egypt are life itself, and there is no life but in you.
(cited in Safran, 1961, p. 87)
The exaltation of the country and panegyric description of its physical landscape
and purity of climate also featured prominently in the polemics of the Iranian
nationalist discourse (Adamiyat, 1955 [1334], 1970; Kashani-Sabet, 1999).
Alternatives to Territorial Nationalism

Liberal territorial nationalism went down in 1939 to 1953 before the onslaught of
two major ideological movements. One consisted of secular authoritarian ideolo-
gies of the Left and the Right, and the other was religious nationalism—both were
united only by their common hostility to parliamentary politics and democratic
elections. In Iran, liberal territorial nationalism was opposed by monarchy-cen-
tered nationalism and fundamentalism, and in Arab-majority countries by pan-
Arab nationalism and Sunni fundamentalism, spearheaded by the Society of the
Muslim Brothers.
Multiple historical factors contributed to the decline of liberal territorial
nationalism. A common feature of this decline across the region appears to
relate to the failure of the indigenous dominant classes (landowners and mer-
chants) to maintain hegemony in national politics on the one hand, and to the
radicalization of the members of the middle classes on the other. The national
context was also important not only in shaping how the members of these
social classes operated under varying (internationally induced) economic crisis,

70
but also in highlighting the factors that were specific to each country as well. In
Algeria, for example, the failure of the French to make peace with the moderate
leaders of the liberation movement contributed to the decline of liberalism and
the radicalization of the anticolonial struggle, leading to the Algerian war of
independence in 1954 to 1962 (Ruedy, 1992). In Egypt and Syria, the inability of
the politicians from landowning-merchant or industrialist families to promote
economic development increased the level of mass dissatisfaction with the gov-
ernment. At the same time, the identification of the national government and
the parliament with the particular interests of these classes eroded the uni-
versalistic appeal of parliamentary politics and undermined its legitimacy in
society. Furthermore, the members of the upper classes attaching low esteem to
the military professions created a structural opportunity for people of humble
origins to use the military as a channel of upward mobility. As a result, they
gradually managed to control this powerful institution and seize power through
coups. Pan-Arab nationalism and socialism provided the military a powerful
ideological tool and policy frame to destroy the basis of the power of their
upper-class rival by nationalizing their properties. At the same time, social-
ist policies were intended to expand and consolidate popular support for the
ruling military regime (Gershoni & Jankowski, 1995; Khoury, 1987; Moaddel,
2005). Finally, in Iran, the royalists, backed by a United States–British alliance,
and with a tacit approval of the organized Shia establishment overthrew the
premiership of Mohammad Mosaddiq in 1953 and abruptly ended the national–
liberal episode. During the postcoup period, Western government steadily sup-
porting the dictatorial monarch delegitimized pro-Western liberal politics, and
contributed to the decline of liberalism and the rise of anti-Western cultural
movements (Moaddel, 1993).
Pan-Arab nationalism and Sunni fundamentalism emerged somewhat simul-
taneously in the same social context in places like Egypt and Syria. That is, the
two ideological discourses were preceded by major changes in social conditions
during the first half of the 20th century. The population increased rapidly, rural-to-
urban migration accelerated, traditional social classes continued to decline, a new
educated middle class connected to the expanding bureaucratic structure of the
state and to an emerging industrial sector grew swiftly, a working class emerged,
and landowners and merchants continued to expand their political power and
interests in the society. Parallel with this process was also enrichment and “bour-
geoisification” of the leaders of the nationalist movements. The installment of
Faisal as the king of Iraq in 1921, following the overthrow of his kingship in Syria
a year earlier, was also a major factor in the development of pan-Arab national-
ism in Iraq (Ansari, 1986; Dawisha, 2003; Dawn, 1973; Deeb, 1979; Gershoni &
Jankowski, 1995; Issawi, 1954, 1966; Khoury, 1983, 1987; Marsot, 1977; Quraishi,
1967).
It would be hard, however, to contrast the structural changes of the 19th cen-
tury with those of the early 20th to explain why the first set of changes produced
territorial nationalism whereas the second set resulted in pan-Arab nationalism
and religious nationalism. For sure, the bourgeoisification of the nationalist
leaders and their exclusivist policies might have undermined the legitimacy of

71
territorial nationalism and parliamentary politics in Egypt, Iraq, and Syria, hence
contributing to the rise of alternative oppositional ideologies, including pan-Arab
nationalism. However, the production of pan-Arab nationalist discourse occurred
during the period between the two world wars, years before the decline of territo-
rial nationalism in the 1940s through the 1950s.
Here, again, the ideological target the indigenous intellectual leaders encoun-
tered appears to have been the key factor in the production of these discourses.
If liberal territorial nationalism was produced in oppositional relation to monar-
chical absolutism, ulama obstructionism, and foreign occupation during the late
19th and early 20th centuries, pan-Arab nationalism and religious nationalism
were responses to ideological targets that were quite different. The intellectual
leaders who arrived at the pan-Arab nationalist discourse to resolve the issue of
legitimate political power had encountered European colonial domination of the
Arab world after the disintegration of the Ottoman Empire. The Islamic funda-
mentalists, on the other hand, first arose in opposition to the secularist policies of
the nationalist state and then became radicalized as the state’s exclusivist policies
and interventions in the economy expanded considerably during the second half
of the 20th century.
Pan-Arab Nationalism
Although the rise of Arab consciousness was rooted in the 19th century, pan-
Arab nationalism as a full-blown political ideology was an outcome of the reflections
of Arab intellectual leaders between the two world wars on the sociopolitical prob-
lems facing their societies. It was formulated in reaction to the post-World War I
colonial partitioning of the Arab lands into disparate states and the imposition
of the French mandate on Lebanon (1920–1943) and Syria (1920–1945), and the
British mandate on Iraq (1920–1932) and Transjordan (1922–1946). Arab commu-
nities were diverse, and even the word Arab was reserved only for the Bedouins
during the premodern period. Yet, European domination appears to have had a
uniformitarian effect on the indigenous intellectual leaders’ perception of Arabs;
diverse Arabs were singularized as one people, hence the necessity to establish an
all-inclusive Arab state. The Arab identity thus became the key cultural differen-
tia specifica that distinguished a perceived subjugated people from the domineer-
ing Europeans (Chartouni-Dubarry, 1993; Cleveland, 1971; Dawisha, 2003; Dawn,
1973, 1988; Haim, 1962; Hourani, 1983; Khalidi, Anderson, Muslih, & Simon,
1991; Moaddel, 2005; Zeine, 1973).
In the pan-Arab–nationalist modality, nation is defined in terms of language-
cum-ethnicity. Attachments to Arab identity, rather than to a territory or a par-
ticular religious affiliation, define the criterion for membership in the political
community. As harbingers of pan-Arab nationalism, Sati’ al-Husri (1880–1968)
and Darwish al-Miqdadi (1879–1968) best articulated the notion that the Arab con-
stituted one single eternal nation. For them, only the followers of imperialism and
regionalism would deny the existence of the Arab nation. According to al-Husri,
There is left no room to doubt that the division of the Arab provinces into several
states took place because of the bargaining and ambitions of the foreign states,
and not according to the views and interests of the people of the countries. So,

72
too, were the borders of these states determined by the wishes and agreements of
the foreign powers, and not according to the natural demands of the situation or
the requirements of indigenous interests. … Is it possible for us to consider, for
example, the people of Syria as forming a true nation, different from the people
of Iraq and Lebanon? Never, gentlemen. All that I have explained indicates clearly
that the differences we now see between the people of these states are tempo-
rary and superficial. … We must always assert that the Syrians, Iraqis, Lebanese,
Jordanians, Hejazis, and Yemenis all belong to one nation, the Arab nation. (cited
in Cleveland, 1971, p. 130)
For pan-Arab ideologues, Arabs are one singular indivisible people. They believed
in the naturalness of the Arab nation, where its different provinces formed parts
of an organic whole, having historical permanency. In textbooks designed for use
in the schools of Palestine, Syria, and Iraq, the Arab homeland is portrayed as
a natural geographic unit, consisting of the “Arab Island”—a living body where
“the head” was the Fertile Crescent; “the heart,” central Arabia; and “the extremi-
ties,” the Arabian coastlands from the Gulf of Aqaba to “the Gulf of Basra.” This
“Island” was “the cradle of the Arabs and their fortress” (Dawn, 1988, p. 69).
Geography, however, is of secondary importance. The people, not territory, are
the decisive element, because they are the creators of the homeland. All lands
inhabited by Arabs are Arab lands. For Muhammad Izzat Darwaza (1888–1984),
“the lands of Syria, Iraq, and Palestine were always Arab because they were filled
with Arabs” (Dawn, 1988, p. 70). For Miqdadi also, the Arab homeland is the ter-
ritory inhabited by Arabs, which expanded as Arabs expanded into Asia, Africa,
and Europe. The Arabs inhabiting the homeland was accomplished in remote
antiquity. Arabic was the most advanced among the ancient Semitic languages.
The ancient pre-Islamic Arabs, however, were only a small part of the glorious
Arab history (Dawn, 1988, p. 70).
The Europeans, the Greeks, Turks, and Iranians are portrayed as enemies of
Arabs. The textbooks reinforced the belief that borrowing from outsiders, par-
ticularly Persians and Westerners, had been nearly fatal to the ancient Arabs. The
Persians are depicted as being filled with hatred of Arabs, and having a fanatical
desire for revenge for the loss of their sovereignty and glory (Dawn, 1988, p. 75).
Islamic Nationalism
The rise of Islamic nationalism, as a political dimension of Islamic funda-
mentalism, was in part a consequence of the breakdown of the religious–liberal
alliance that led the nationalist movements, most notably in Egypt and Iran,
during the early 20th century. For a group of the ulama and religious activists,
who played a leading role in the emergent nationalist movements for the con-
struction of the modern secular state, a constitutional government was far supe-
rior to the existing monarchical absolutism or foreign colonial rule. As one of
the leaders of the Iranian Constitutional Revolution of 1906, theologian Sayyid
Muhammad Tabataba’ie registered his support for a constitutional political sys-
tem on the ground that it “will bring security and prosperity to the country” (cited
in Adamiyat, 1976 [1355], p. 193). Similarly, for Islamic reformer Mufti of Egypt
Muhammad Abduh and his followers, although British rule was unacceptable in

73
principle, it had created the only viable opportunity for the gradual education of
their fellow Egyptians and would also be easier to get rid of than the khedivial
autocracy (Ahmed, 1960, p. 52).
After the overthrow of traditional monarchical absolutism in Iran and Turkey,
and nationalist revolution in Egypt during the first quarter of the 20th century,
the intellectual climate turned overly secularist and antireligion. The ruling cul-
tural and political elite did not confine their activities to the realm of politics.
Subscribing to a Europe-centered secularist project, they narrowed down the
cultural and social spheres of religious institutions. They rewrote history to fit
their nationalist conception of the past and to overlook the Islamic period, glori-
fied pre-Islamic kingship and ancient history, reformed the educational institu-
tions to undermine the influence of religion, imposed feminism from above, and
attacked religion and religious rituals in terms of Western standards (Moaddel &
Karabenick, 2013; Vatikiotis, 1980, pp. 306–307).
The cultural onslaughts on Islam, however, begot a strong reaction from both
the ulama and Muslim activists. The discourse of Islamic nationalism was thus
produced in opposition to the secularist ideas and policies of the nationalist state.
The formation of the Society of the Muslim Brothers in Egypt is 1928, which
expanded to other Arab countries in subsequent decades, represented an orga-
nizational embodiment of the antisecularist movement and exemplified 20th-
century Sunni fundamentalism-cum-religious nationalism. Hasan al-Banna, the
leader and founder of the Muslim Brothers, often expressed his disapproval of
the secular trend in the country as well as the secular policies implemented by
the national government. He blamed Egypt’s problems squarely on the country’s
secular nationalist parties such as the Wafd and Liberal Constitutionalists as well
as the state’s educational policies (Mitchell, 1969).
At the outset, the Muslim Brothers were a moderate group and expressed little
interest in revolutionary activities in their attempts to realize their ideal Islamic
order. In fact, they enjoyed a cordial relationship with the reigning monarch.
Furthermore, they even tried to pursue their religious objectives through partici-
pation in parliamentary politics. However, as the government launched repressive
measures against the Brothers, first by blocking their participation in the elec-
tions, then by rigging the elections after they were allowed to participate, dissolv-
ing the Society of the Muslim Brothers in 1948, and finally assassinating al-Banna
a year later, the Muslim Brothers began to question the legitimacy of the parlia-
mentary system. As one prominent member of the Muslim Brothers claimed,
the upper class monopolized the government, and the people were compelled to
choose the parliament from among their oppressors: the landlord commanded
the votes of his tenants; the lord of finance, those of his debtors. All the elections
since 1923 were spurious (Mitchell, 1969, p. 219).
With the active support of the Muslim Brothers, the 1952 military coup ended
the era of constitutional monarchy in Egypt. It did not, however, produce a friend-
lier environment for the Muslim Brothers. The totalitarianism of the postcoup pan-
Arab nationalist regime, as a consequence, was associated with the rise of religious
extremism in the country. The new extremism reflected most notably in the politi-
cal discourse of Sayyid Qutb (1906–1966), a secular intellectual-turned-Islamic

74
activist. Qutb did not simply criticize the secularism of the state, he questioned
the very Islamic nature of the existing order. For him, Egypt was a throwback to
the pre-Islamic conditions in Arabia, where people lived under a state of igno-
rance (jahiliyya). By implication, it was thus incumbent on the faithful to rebel
against it. An even more extremist and militant version of religious nationalism
was formulated by Muhammad Abd al-Salam Faraj (1954–1982). In his view, the
current rulers of Muslim countries were all apostates and should be overthrown
to establish a truly Islamic state (Akhavi, 1992, pp. 94–95). Faraj was the head
of the Cairo branch of the Tanzim al-jihad (Jihad Organization) that assassinated
President Anwar al-Sadat. Building on Sayyid Qutb’s argument, Faraj proclaimed
that “the establishment of an Islamic State is an obligation for the Muslims. …
The laws by which the Muslims are ruled today are the laws of Unbelief; they are
actually codes of law that were made by infidels who then subjected the Muslims
to these [codes]” (cited in Sageman, 2004, p. 15). These ideas formed the foun-
dations of such radical organizations as Jamaat Islamiyyat, Islamic Jihad, and al
Qaeda, justifying the use of terror for the realization of their ostensibly Islamic
objectives.
Similarly, Iranian Shi’i fundamentalism- cum- Islamic nationalism first
emerged as an antisecular religious extremist group, Fedaiyan-i Islam (Devotees
of Islam), in 1946. The group believed that Iranian society had strayed from the
right path and needed to be purified. It called for a strict enforcement of the
shari’a, which included prohibitions of alcohol, tobacco, opium, films, gambling,
and foreign clothing; the amputation of hands of thieves; the veiling of women;
and an elimination from school curricula the teaching of all non-Islamic subjects
(Abrahamian, 1982, p. 259). It also launched a campaign of terror against the
politicians and intellectuals the group considered responsible for corrupting the
society. One of its most heinous acts was the assassination of Ahmad Kasravi, a
prominent historian and noted social critic, in 1946. Despite the suppression of
the Fedaiyan and the execution of its leader, Navvab Safavi (1924–1955), in 1955,
the group survived and its members became staunch supporters of Ayatollah
Khomeini. It was reorganized by Sadiqh Khalkhali, the infamous ruthless judge
who, after the 1979 Revolution, executed thousands of the supporters of the for-
mer regime and political dissidents.
The modality of religious nationalism considers Islam as the basis of legiti-
macy and the source of legislation, negates secular politics and instead advances
the idea of the unity of religion and politics in an Islamic government, promotes
religious centrism and intolerance of other faiths, endorses male supremacy
and restricts women’s involvements in the public sphere, and rejects Western
culture as decadent. The view of Egyptian Sayyid Qutb on the Islamic concep-
tion of political sovereignty displays a clear contrast with a conception in territo-
rial nationalism and pan-Arab nationalism. Sayyid Qutb questioned pan-Arab
nationalism and other forms of nationalism first for being tyrannical, reasoning
that “there was no sense in liberating the land from a Byzantine or a Persian
tyrant in order to put it in the hands of an Arab tyrant. Any tyrant is a tyrant”
(cited in Sivan, 1985, p. 30). In a later text, he flatly rejected pan-Arabism as a
concept incompatible with Islam:

75
The homeland (watan) a Muslim should cherish and defend is not a mere piece
of land; the collective identity he is known by is not that of a regime. … Neither is
the banner he should glory in and die for that of a nation (qawn). … His jihad is
solely geared to protect the religion of Allah and His Shari’a and to save the Abode
of Islam and no other territory. … Any land that combats the Faith, hampers
Muslims from practicing their religion, or does not apply the Shari’a, becomes
ipso facto part of the Abode of War (Dar al-Harb). It should be combated even if
one’s own kith and kin, national group, capital and commerce are to be found
there. … A Muslim’s homeland is any land governed by the laws of Islam. Islam
is the only identity worthy of man. … Any other group identity … is a jahili iden-
tity of the type humanity has known during its periods of spiritual decadence.
(cited in Sivan, 1985, p. 31)
Likewise, harbinger of Shi’i fundamentalism Ayatollah Ruhollah Khomeini
proclaimed that:
Present Islam to the people in its true form, so that our youth do not picture
the akhunds as sitting in some corner in Najaf or Qum, studying the questions
of menstruation and parturition instead of concerning themselves with politics,
and draw the conclusion that religion must be separate from politics. This slogan
of the separation of religion and politics and the demand that Islamic scholars
not intervene in social and political affairs have been formulated and propa-
gated by the Imperialists; it is only the irreligious who repeats them. (Khomeini,
1981, p. 38)
To summarize, indigenous intellectual leaders formulated diverse modalities of
collective sovereignty in oppositional relations to the type of ideological targets
they encountered. As these targets changed—the old targets declined, trans-
formed, or withered away—under different historical episodes, a new the type of
modality was produced in opposition to new targets: (liberal) territorial nation-
alism in opposition to monarchical absolutism, foreign occupation, and ulama
obstructionism during the late-19th and early-20th century; pan-Arab nationalism
to the colonial partitioning of the Arab territories in the period between the two
world wars; and religious nationalism to the rise of the interventionist secular
ideological state during the second half of the 20th century. Furthermore, a sub-
stantial part of the intellectual attention span was fixated on signifying a particular
cultural feature as the identifying mark, and hence the most important organiz-
ing principle, of collective representation. The justificatory accounts of their view
on other issues have been mainly derivative of this core principle. That is to say,
historically, under the conditions that the establishment of collective sovereignty
was predicated on freeing the territorial community from monarchical absolutism
and ulama obstructionism, those who were more committed to this struggle were
also drawn toward secularism, the liberal value of gender equality, and apprecia-
tion of Western culture. On the other hand, those who rested their conception of
collective representation on the idea of Arab unity—hence the necessity of the
liberation of all Arabs from European domination and their local agents, includ-
ing members of the upper classes—were drawn toward anti-Western militancy
and became strong advocates of state socialism. Finally, those who believed in

76
Islam as the distinguishing mark of their collectivity and tied the restoration of
the Islamic identity to the overthrow of the Western-backed secular governments
were also directed toward Islamic government, religious fundamentalism, gender
inequality, and the institution of male domination.
Islamic Nationalism versus Liberal Nationalism in

Muslim-Majority Countries
The available evidence shows that the changes that transpired in the national-
ist perspectives of Iranians, Iraqis, and Saudis in recent decades also appeared
to have been affected by the working of factors that are broadly similar to
those that shaped the nationalist views of the intellectual leaders in Algeria,
Egypt, and Iran during the early 20th century. Two propositions are relevant
here: (a) religious authoritarianism and ulama obstructionism tend to provoke
liberal and secular responses, and (b) foreign occupation contributes to the
shaping of national territorial identity. Cultural trends in Iran after the 1979
revolution and changes in values among Iranians and Saudis appear to support
the first proposition, and the rise of territorial nationalist identity among Iraqis
following the US invasion in 2003 exemplifies the second. This is true because,
first, although the Iranian Revolution and the subsequent formation of reli-
gious absolutism under Ayatollah Khomeini was a major setback for the follow-
ers of liberal values and secular politics, decades of clerical rule did not create a
religious order in the country. Scholars of Iran’s intellectual history have shown
that the postrevolutionary period exhibits a dramatic decline in support for the
Shia fundamentalism that once dominated the discourse of the 1977 through
1979 revolution on the one hand, and the rise of secular and religious reform-
ist oppositional discourses on the other (Boroujerdi, 1996; Jahanbegloo, 2004;
Kamrava, 2008; Mahdavi, 2011; Rajaee, 2007; Vahdat, 2002, 2003). Findings
from values surveys have also shown that the Iranian public appeared to be less
religious than the populations from many other Muslim-majority countries, and
the trend in values among Iranians has been toward social individualism, gen-
der equality, democracy, and national identity (Moaddel, 2009). A similar trend
appears emerging among the Saudi public during the past decade (Moaddel &
de Jong, 2013). Second, a series of values surveys carried out in Iraq between
2004 and 2011 shows an increase in support for national identity and secular
politics. The percentage of Iraqis who defined themselves as “Iraqis above all”
(as opposed to Muslims, Arabs, or Kurds) increased consistently from 23% in
2004 to 63% in 2008, and then dropped slightly to 57% in 2011, and those who
agreed that Iraq would be a better place if religion and politics were separated
increased from 50% in 2004 to about 70% in 2011 (Moaddel, de Jong, & Dagher,
2011). Thus, in the same way that territorial nationalism in early-20th-century
Algeria and Egypt was a result of French and British occupation, respectively
(Moaddel, 2005), the dramatic increase in the percentage of Iraqis who defined
themselves in territorial nationalist terms, rather than religious, in 2004 to
2011 may reflect a response to American occupation of the country.

77
In the second strategy, I test further the proposition concerning the signifi-
cance of the identity of belonging and national solidarity in predicting values
using micro survey data. I focus on liberal territorial nationalism and religious
nationalism- cum-
fundamentalism— two currently competing “ideologies of
order” in Muslim-majority countries (Friedland, 2002; Juergensmeyer, 1993, p.
29), and assess how changes in identity and national pride predict sociopolitical
and cultural values. Although the modalities of liberal territorial nationalism and
religious nationalism in the perspectives of intellectual leaders are defined explic-
itly and their boundaries specified clearly, I propose that, among the members
of the ordinary public, the support for these modalities are a matter of tendency;
people who identify with the territorial nation tend to be more strongly in favor
of liberal values whereas those who identify with Islam have stronger conserva-
tive tendencies. Therefore, the change in the basis of identity from religion to
territorial nation predicts changes in value orientations toward gender equality,
secularism, secular politics, democracy, Western culture, and outsiders. National
pride, on the other hand, is linked negatively to gender equality, secularism, secu-
lar politics, democracy, Western culture, and outsiders. As was discussed earlier,
because religious nationalism tends to be predominantly hostile toward outsid-
ers and rigorously promoting intrafaith solidarity (Moaddel & Karabenick, 2013),
and that liberalism being self-referential and critical of the social institutions that
block the realization of individual liberty, national pride reflects religious nation-
alist or Islamic fundamentalist modality in the current context of Muslim-major-
ity countries.
Hypotheses of the Study
National Identity Versus National Pride
I propose that a shift in the basis of identity from religion to nation and the
feeling of national pride has varied relationships with orientations toward liberal
values. That is:
Hypothesis 1: The shift from religious to national identity entails
(a) A more favorable attitude toward gender equality
(b) A more favorable attitude toward Western culture
(c) A more favorable attitude toward democracy or secular politics
(d) A stronger secular orientation
(e) A weaker hostility toward outsiders
Hypothesis 2: The feeling of national pride, in contrast, is linked to

(a) A less favorable attitude toward gender equality
(b) A less favorable attitude toward Western culture
(c) A less favorable attitude toward democracy and secular politics
(d) A weaker secular orientation
(e) A stronger hostility toward outsiders

78
Modernity and Liberal Nationalism

According to the modernization theory, the higher one’s socioeconomic status,
the stronger is one’s support for liberal values:
Hypothesis 3: A higher socioeconomic status is linked to
(a) A more favorable attitude toward gender equality
(b) A more favorable attitude toward Western culture
(c) A more favorable attitude toward democracy or secular politics
(d) A stronger secular orientation
(e) A weaker hostility toward outsiders
Measurement and Models
Dependent Variables
Gender equality: Three indicators measure this construct. Respondents were
asked the following questions: (a) How important is it that a woman wears the
veil in public places (1 = very important, 6 = not at all important [unfavorable to
veil])? (b) Do you strongly agree, agree, disagree, or strongly disagree that men
make better political leaders than women (1 = strongly agree, 4 = strongly disagree
[unfavorable to men better leaders])? (c) Do you strongly agree, agree, disagree, or
strongly disagree that a wife must always obey her husband (1 = strongly agree,
4 = strongly disagree [unfavorable to wife obedience])?
Higher values on these measures indicate stronger unfavorable attitudes
toward the veil, men as better political leaders, and toward wife obedience—
hence, a stronger support for gender equality and a weaker advocacy of the values
promoted by religious fundamentalism on the proper role for women in the fam-
ily and society.
Secular orientation: Four indicators measure secular orientation of the respon-
dents. All these measures are recoded so that higher values indicate stronger secu-
lar orientation (weaker religiosity). One is on the importance of religion in one’s
life: (a) How important is religion in your life (1 = very important, 4 = not very
important [religion unimportant])? The other is on mosque attendance: (b) How
often do you participate in mosque services (1 = more than once a week, 7 = do
not go to mosques [church for Christian respondents in Lebanon] [mosque/church
attendance])? The third focuses on trust in the religious institutions: (c) How much
trust do you have in mosques ([church for Christian respondents in Lebanon]
1 = a great deal [trust in mosque/church], 6 = none at all). Finally, the fourth mea-
sures people’s attitudes toward the shari’a as the guiding principle of a good gov-
ernment: (d) How important is it for a good government to implement only the
laws of the shari’a ([laws inspired by Christian values for Christian respondents in
Lebanon] 1 = very important, 4 = not very important [shari’a])?
Western culture: Religious fundamentalism holds that the West is culturally
decadent and Islamic culture must be protected against invasion by Western cul-
ture. One question measures attitudes toward Western culture: In your view, how

79
important is Western cultural invasion (1 = very important, 6 = not important at

all)? Lower values on this measure indicate stronger attitudes against Western
culture.
Hostility toward outsiders: To assess the extent to which the two modalities
differ in orientations toward outsiders, a question that taps into hostility toward
immigrants is included in the model: How about people from other countries
coming here to work. Which one of the following do you think the government
should do (1 = let anyone comes who wants to, 4 = prohibit people coming here
from other countries [hostility toward immigrants])? Higher values indicate a
stronger hostility toward immigrant workers.
Form of government: The question concerning the most desirable form of gov-
ernment has been the subject of considerable debate among intellectual leaders
and the public at large in Muslim-majority countries in the contemporary period.
Here, I consider two constructs: democracy and secular politics. An important caveat
should be noted here. Except for extremist groups that reject democracy categori-
cally in favor of religious absolutism, the followers of religious fundamentalism
(e.g., Muslim Brothers or antireformist and conservative supporters of the ruling
regime in Iran) often express support for “Islamic” democracy. Therefore, given
that democracy means a different thing to different people, we consider secular
politics to be a better distinguishing mark between liberalism and fundamental-
ism. When data are available, attitudes toward secular politics is used instead of
attitudes toward democracy.
To measure attitude toward democracy, respondents were asked: Do you
strongly agree, agree, disagree, or strongly disagree that democracy may have
problems but it is a better system of government? And to assess orientation
toward secular politics, they were asked: Do you strongly agree, agree, disagree,
or strongly disagree that [Study Country] would be a better place if religion and
politics were separated? Both variables were recoded such that a higher value indi-
cates a stronger support for democracy or secular politics.
Independent Variables
National versus religious identity: This is the key independent variable constructed
in terms of the responses to the question on identity: Which of the following best
describes you: (a) above all, I am an Iranian, Egyptian, Iraqi, Saudi … ; (b) above
all, I am a Muslim (Christians [ for Christian respondents]); (c) above all, I am an
Azeri-Turk, a Kurd, or other, specify____?
National identity is measured as a dummy variable (0 = Muslim, above all;
1 = else, above all). It measures whether the respondents’ primary allegiance is to
the nation/ethnicity or to religion.
National pride is measured in terms of the strength of one’s pride to be a citizen
of the country: How proud are you to be a citizen of [Study Country]: 1 = very proud,
2 = quite proud, 3 = not very proud, or 4 = not at all proud? This variable is recoded
so that a higher value is associated with a stronger feeling of national pride.
Socioeconomic status: Three variable indicators measure this construct: edu-
cation, income, and self- report social-
class position. Education: Education is

80
measured by asking respondents: What is the highest educational level that you

have attained?: 1 = no formal education, 2 = incomplete primary school, 3 = com-
plete primary school, 4 = incomplete secondary school (technical/vocational type),
5 = complete secondary school (technical/vocational type), 6 = incomplete second-
ary (university–preparatory type), 7 = complete secondary (university–preparatory
type), 8 = some university-level education without degree, and 9 = university-level
education with degree. Social class: Self-reported class identification is measured
by asking respondents to describe their class background: People sometimes
describe themselves as belonging to the working class, the middle class, or the
upper or lower class. Would you describe yourself as belonging to the (1) upper
class, (2) upper middle class, (3) lower middle class, (4) working class, or (5) lower
class? The self-reported measure of social class is recoded so that higher values
indicate higher social classes. Income: This variable is measured in terms of the
respondents’ placement of their household income on a scale of incomes: On this
card is a scale of incomes on which 1 indicates the lowest income decile and 10 the
highest income decile in your country. We would like to know in what group your
household is. Please specify the appropriate number, counting all wages, salaries,
pensions, and other incomes that come in.
Gender: A dummy variable on gender is constructed (1 = male and 0 = female).
Age: The age of the respondents is included in the analysis as a continuous
variable.
Structural Equations Models and the Data

Data on these variables are not available for all countries included in the analysis.
The most complete data is from the 2005 Iran survey. For other countries, data
on only a subset of the variables are available. As a result, three different mod-
els are developed and are estimated (Figure 3.1, Figure 3.2, and Figure 3.3). The
model 1 in Figure 3.1 estimates the linkages of national identity, national pride,
and socioeconomic status with gender equality, Western culture, hostility toward
immigrants, secular politics, and secular orientation, controlling for gender and
age among Iranians, using the data from the 2005 survey. Modified versions of this
model, model 2 and model 3, are presented in Figure 3.2 and Figure 3.3, respec-
tively. Figure 3.2 uses the survey data from eight countries, and Figure 3.3 from
Iraq in 2004, Iraq in 2006, and Lebanon. Several differences among the three
models should be noted. Model 1 in Figure 3.1 and model 3 in Figure 3.3 consider
secular politics as a more appropriate indicator of form of government than democ-
racy. Attitude toward secular politics indicates a more clear difference between the
people who favor liberal nationalism and those supporting religious nationalism
than attitude toward democracy, which has different practical meanings for the fol-
lowers of liberalism and religious nationalism. Therefore, an indicator of attitude
toward secular politics is used instead of attitude toward democracy in these two
models. This question, however, was not asked in the surveys carried out in the
countries included in model 2 in Figure 3.2, and as a result, attitude toward democ-
racy is used in this model. Finally, questions on attitude toward Western culture
and hostility toward immigrants were not asked in the Iraq and Lebanon surveys.

81
0, 0, 0,
0, 0, 0, e1 e2 e3
e13 e14 e15
Unfavorable to Unfavorable to Unfavorable to
Men better leader veil wife obedience
Education Self-report Income 1
class status
Gender equality 0
Socioeconomic 1
0,
1
status 0, Mosque attendance e4
e11 0,
1
Secular 1 Religion important e5
National pride orientation 0 0,
1
1
Conf in mosque e6
0, 0,
1
e12 Sharia only e7
Age 0 0,
Secular 1
politics e8
0
National identity e9
1
Western culture
Male
0,
Hostility toward 1
e10
immigrants
Figure 3.1 Structural relations of the exogenous variables with gender equality,

secular orientation, secular politics, Western culture, and hostility toward immigrants
based on the 2005 Iran survey data.
0,0 0, 0,
e12 e13 e14
1 1 1
Unfavorable to Unfavorable Unfavorable to
Self-report men better leader to veil wife obedience
Education social class Income
1 0, 1
0,
1
0,
1
0, e10 e1 e2
Gender equality 0
1
Socioeconomic 0,
status e3
e6 0,
0,
Religion 1
e4
1 important
Secular
National pride orientation 0, 0,
Trust 1
e5
in mosque
National identity 0,
1
Democracy e7
Male
0,
1
Anti-immigrant e8
Age
Figure 3.2 Structural relations of the exogenous variables with gender equality,

secular orientation, democracy, and hostility toward immigrants in Algeria,
Bangladesh, Indonesia, Iran 2000, Jordan, Morocco, Saudi Arabia, and Turkey.
82
0, 0,
0,1 0, 0, 0,
e12 e13 e14 e9 e1 e2

1 1 1 1 1 1
Self-report Unfavorable to Unfavorable Unfavorable to
Education social class Income
men better leader to veil wife obedience
1
Gender equality
Socioeconomic 0 1
status
0 e3
0 0
National pride Religion 1
e4
e6 important
National identity
Secular 0
orientation Trust 1
e5
in mosque
0
Male
0
1
Secular politics e7
Age
Figure 3.3 Structural equations of the relations of the exogenous variables with

gender equality, secular orientation, and secular politics for Iraq 2004, Iraq 2006, and
Lebanon.
As shown in the three figures, the models have four exogenous variables and
between five (model 1 in Figure 3.1) and three (model 3 in Figure 3.3) dependent
variables. Although there may be reciprocal causations between the dependent
variables—for example, people who favor gender equality may have stronger
secular orientation, support secular politics more strongly, or be more favorable
toward Western culture and vice versa—the intention of these models is to assess
the linkages between the exogenous and the dependent variables and how well
these models fit the data.
Table 3.1 provides information on the sample size and dates of each of the sur-
veys carried out in the Muslim-majority countries. All the surveys used multi-
stage probability sampling procedures, broken down into urban and rural areas
in proportion to their size, with roughly equal male and female respondents. The
interviews, which required approximately 1 hour on average to complete, were
conducted face-to-face in respondents’ residences. For more information on these
surveys, see www.mevs.org.
Table 3.2 shows the mean and standard deviation of the variables of the study.
As this table shows, data on all the dependent variables are not available across the
12 surveys. Although the 2005 Iran survey includes data on all the variables, data
on only a subset of these variables are available for Algeria, Bangladesh, Indonesia,
Iran 2000, Jordan, Morocco, Saudi Arabia, and Turkey. Data on a slightly different
subset are also available for Iraq 2004, Iraq 2006, and Lebanon. Altogether these
data are adequate to evaluate rigorously the different ways in which national iden-
tity and national pride are linked to orientations toward liberal values.

83
Table 3.1 Sample Size and Date of Survey from Each

of the Countries
Country Sample size Survey dates
Algeria 1,282 March–May 2002

Bangladesh 1,500 August–September 2002
Indonesia 1,004 March–September 2001
Iran 2,532 January–February 2000
Iran 2,537 June–August 2005
Iraq 2,325 November–December 2004
Iraq 2,701 March–April 2006
Jordan 1,223 September 2001
Lebanon 3,039 April–September 2008
Morocco 1,013 February 2002
Saudi Arabia 1,502 April–May 2003
Turkey 3,401 December 2001
source: www.mevs.org.
Analysis and Findings

Estimates of Structural Equation Models
Table 3.3 reports the results of the estimates of the three structural equation mod-
els described in Figure 3.1, Figure 3.2, and Figure 3.3. Model 1 (Figure 3.1) uses data
from the 2005 Iran survey; Model 2 (Figure 3.2), from eight surveys carried out in
Algeria, Bangladesh, Indonesia, Iran 2000, Jordan, Morocco, Saudi Arabia, and
Turkey; and Model 3 (Figure 3.3) from three surveys conducted in Iraq 2004, Iraq
2006, and Lebanon. Table 3.3 reports only the standardized structural coefficients
(βs) and the level of significance. It also reports the measures of goodness-of-fit
and the chi-square value for the three models. As shown, these three models fit
the data (Model 1: χ2 = 721.625, df = 84, α < .001, CFI = .93, RMSEA = .05; Model
2: χ2 = 2,430, df = 408, p <.000, CFI = .90, RMSEA = .04; Model 3: χ2 = 1,118, df =
120, p <.000, CFI = .914, RMSEA = .03).
National Identity Versus National Pride

As Table 3.3 shows, the estimates of the structural coefficients of the linkages
of national identity and national pride with the dependent variables in model 1,
which includes information on all variables of interests, are significant and in the
expected direction. That is, national identity is linked positively to attitudes toward
gender equality (β = .173), secular orientation (β = .319), secular politics (β = .177),
and Western culture (β = .124). The only exception is that national identity has
no significant connection with hostility toward immigrants. National pride, in
contrast, is linked negatively to gender equality (β = –.224), secular orientation
(β = –.388) secular politics (β = –.227), and Western culture (β = –.329), but posi-
tively to hostility toward immigrants (β = .074).

84
Table 3.2 Mean and (Standard Deviation) for the Variables Used in the Analysis
Variable in Country
the data set
Iran 2005 Algeria Bangladesh Indonesia Iran 2000 Jordan Morocco Saudi Arabiaa Turkey Iraq 2004 Iraq 2006 Lebanon
National identity/national pride

National identity .50 (.50) 0.33 (.47) 0.76 (.43) 0.58 (.49) 0.38 (.48) 0.29 (.45) 0.37 (.48) 0.25 (.44) 0.36 (.48) 0.37 (.48) 0.36 (.48) 0.83 (.38)
(1 = else,
0 = Muslim)
National pride 3.55 (.68) 3.68 (.60) 3.69 (.53) 3.41(.64) 3.85 (.55) 3.67 (.45) 3.84(.44) 3.67 (.61) 3.53 (.78) 3.70(.62) 3.79 (.52) 3.28 (.77)
(4 = very proud)
Gender equality
Woman dresses 3.71 (1.61)
as she wants
(6 = very
important)
Wearing a veil 2.27 (1.29) 1.87 (1.09) 1.89 (1.02) 2.09 (1.08) 1.49 (.87) 1.56 (1.04) 1.84 (1.13) 1.24 (.647) 4.22 (1.18) 1.34 (.83) 1.27 (.75) 2.67 (1.32)
(5 = not at all
important)
Wife obedience 2.76 (1.26) 1.92 (1.03) 1.92 (.90) 2.11 (1.02) 2.65 (1.34) 2.10 (1.26) 1.71 (.96) 1.73 (.942) 2.44 (1.39) 1.61 (1.0) 1.57 (.95) 3.19 (1.41)
(5 = strongly
disagree)
Men better 2.07 (.86) 1.96 (1.02) 2.16 (.89) 2.32 (.78) 2.14 (.97) 1.46 (.81) 1.98 (1.04) 1.79 (.977) 2.31 (.96) 1.42 (.76) 1.43 (.74) 2.33 (.94)
political leaders
(4 = strongly
disagree)
85
Secular orientation
Religion impor- 3.73 (.59) 3.91 (.323) 3.86 (.40) 3.98 (.15) 3.75 (.57) 3.95 (.25) 3.92 (.31) 3.88 (.40) 3.65 (.74) 3.93 (.30) 3.95 (.25) 3.19 (.94)
tant (4 = very
important)
Mosque/church 3.97 (2.18)
(7 = more than
once a week)
Trust in mosque 3.40 (.72) 3.48 (.79) 3.88 (.36) 3.71 (.54) 3.46 (.80) 3.54 (.68) 3.74 (.61) 3.81 (.48) 2.84 (1.04) 3.18 (.93) 3.27 (.84) 2.70 (1.03)
(4 = a great deal)
Shari’a law (6 = 4.46 (1.31)
very important)
Secular politics
Religion and 2.44 (.95) 2.63 (1.06) 2.86 (1.14) 3.17 (1.03)
politics separate
(4 = strongly
agree)
Western culture
Western cultural 1.61 (.895) 1.34 (.80) 1.22 (.65)
invasion (5 = not
serious)
Attitudes toward immigration

Anti-immigrant 3.0 (.80) 2.25 (.83) 2.51 (.78) 2.68 (.70) 2.81 (.84) 2.84 (.79) 1.95 (.94) 2.38 (.818) 2.68 (.95)
(4 = prohibit
people)
Democracy
Democracy 3.33 (.78) 3.6 7(.52) 2.78 (.77) 2.89 (1.01) 3.26 (.71) 3.70 (.59) 2.93 (.964) 3.28 (.73)
(4 = strongly
agree)
(continued)
86
Table 3.2 Continued
Variable in Country
the data set
Iran 2005 Algeria Bangladesh Indonesia Iran 2000 Jordan Morocco Saudi Arabiaa Turkey Iraq 2004 Iraq 2006 Lebanon
Demographics
Education 4.54 (2.54) 5.08 (2.32) 3.83 (2.47) 5.29 (2.36) 4.75 (2.49) 3.96 (2.42) 1.79 (2.43) 4.73 (1.00) 3.63 (2.31) 3.42 (2.72) 3.31 (2.64) 6.00 (2.32)
Social class 3.12 (.99) 2.58 (.98) 2.82 (1.09) 2.97 (.84) 2.96 (1.02) 2.83 (.92) 2.33 (1.00) 3.57 (.83) 2.79 (1.01) 2.47 (.93) 2.38 (.85) 2.73 (1.12)
Income 4.87 (1.94) 3.12 (1.79) 4.68 (1.83) 5.83 (1.74) 5.03 (1.70) 3.48 (1.99) 4.37 (1.65) 6.71 (3.22) 3.19 (1.60) 4.20 (1.88) 4.19 (1.70) 5.09 (2.02)
Age 32.8 (12.8) 35.6 (13.5) 33.7 (10.9) 44.6 (13.7) 34.2 (15.4) 36.1 (14.7) 33.5 (12.6) 32.0 (10.9) 37.0 (13.6) 37.0 (13.8) 37.1 (13.8) 32.8 (13.0)
Gender .50 (.50) 0.51 (.50) 0.55 (.50) 0.50 (.50) 0.54 (.50) 0.49 (.50) 0.49 (.50) 0.50 (.50) 0.50 (.50) 0.48 (.50) .48 (.50) .56 (.50)
a
For respondents from Saudi Arabia, valid response codes for education range from 1 through 6, and for income range from 1 through 25.
87
Table 3.3 Standardized Structural Coefficients Showing the Effects of National Identity and National Pride
Variable Model 1 Model 2 Model 3
Iran Algeria Bangladesh Indonesia Iran Jordan Morocco Saudi Turkey Iraq Iraq 2006 Lebanon
2005 2000 Arabia 2004
National identity
Gender equality .173a 0.259a 0.189a 0.323a 0.238a 0.106b 0.189a 0.099c 0.31a 0.240a 0.230a 0.258a
Secular orientation .319a 0.405a 0.134a 0.274d 0.330a 0.210a 0.290a 0.141c 0.356a 0.250a 0.126a 0.393a
Democracy 0.059d 0.159a 0.081c 0.045ᵈ –0.014 0.011 0.125a 0.066a
Secular politics .177a 0.102a 0.165a 0.204a
Western culture .124a 0.104a 0.060b
Anti-immigrant –.010 –0.063c 0.059c –0.044 –0.052c –0.051d –0.089b 0.028 –0.061b
National pride
Gender equality –.224a –0.018 –0.185a –0.044 –0.182a –0.048 –0.077c –0.347a –0.168a –0.435a –0.345a –0.057b
Secular orientation –0.388a –0.336a –0.412a –0.189d –0.223a –0.334a –0.287a –0.237b –0.434a –0.475a –0.398a –0.203a
Democracy 0.024 0.060c 0.127a 0.000 –0.022 0.042 0.03 –0.025
Secular politics –.227a –0.123a –0.088a –0.074a
Western culture –.329a –0.147a –0.197a
Anti-immigrant .074a –0.037 –0.083b –0.049 0.104a –0.006 –0.074c –0.078c 0.088a
Socioeconomic status
Gender equality .610a 0.238a 0.481a 0.185a 0.262a 0.298a 0.586a -0.076d 0.474a 0.191a 0.115a 0.265a
Secular orientation 0.329a 0.165c 0.243a 0.065 0.106a 0.369a 0.208a -0.045 0.254a 0.446a 0.296a 0.167a
Democracy 0.024 0.107a 0.139a 0.057d 0.077c 0.158a 0.012 0.080a
Secular politics .062b 0.127a 0.086a 0.058b
Western culture .004 –0.013 0.040
Anti-immigrant –.010 –0.171a –0.176a 0.058 –0.052c –0.055d 0.027 0.008 –0.063b
(continued)
88
Table 3.3 Continued
Variable Model 1 Model 2 Model 3
Iran Algeria Bangladesh Indonesia Iran Jordan Morocco Saudi Turkey Iraq Iraq 2006 Lebanon
2005 2000 Arabia 2004
Male
Gender equality –.239a –0.391a –0.201a –0.367a –0.389a –0.477a –0.300a –0.412a –0.355a –0.048c –0.082a –0.258a
Secular orientation 0.035 –0.008 –0.009 0.089 –0.028 0.247a 0.114c 0.127c 0.008 0.051d 0.062d 0.082a
Democracy –0.037 –0.024 –0.003 0.013 0.131a 0.021 0.026 –0.019
Secular politics –.001 0.029 0.059b 0.024
Western culture –.030c 0.010 0.016
Anti-immigrants .026 –0.035 0.047ᵈ –0.068c –0.024 0.149a 0.018 0.025 –0.009
Age
Gender equality .074a 0.091c –0.236a 0.089 –0.183a 0.025 –0.119b –0.039 0.013 0.016 –0.016 –0.011
Secular orientation 0.004 –0.014 –0.047 0.037 –0.032 0.036 –0.035 0.002 –0.027 0.069c 0.043 –0.119a
Democracy 0.007 0.009 0.095c 0.111a –0.020 0.060 0.024 0.088c
Secular politics –.001 0.094a 0.020 –0.055b
Western culture –.038 0.044c 0.004
Anti-immigrant .026 –0.075c –0.017 0.030 –0.021 0.005 –0.036 –0.018 –0.019
Measurement: Socioeconomic status
Education .854 1 1 1 1 1 1 1 1 1 1 1
Social class .364a 0.303c 0.459c 0.42c 0.348c 0.311c 0.441c 0.191c 0.255c 0.321a 0.345a 0.258a
Income .300a 0.180c 0.330c 0.148c 0.203c 0.265c 0.454c 0.215c 0.475c 0.266a 0.260a 0.350a
89
Measurement: Gender equality
Veil Important .374a 0.430 0.608 0.358 0.384 0.275 0.458 0.290 0.408 0.852 0.813 0.683
Wife Obedience .557a 0.620a 0.177a 0.613a 0.548a 0.619a 0.61a 0.595a 0.64a 0.467a 0.595a 0.645a
Men better leader .740a 0.579a 0.053 0.529a 0.407a 0.472a 0.375a 0.413a 0.413a 0.510a 0.355a 0.544a
Measurement: Secular orientation
Religion important .609a 0.391 0.608 0.15 0.688 0.253 0.423 0.311 0.728 0.375 0.271 0.658
Trust in Mosque .622a 0.356a 0.679a 0.658c 0.634a 0.455a 0.555a 0.484a 0.448a 0.535a 0.536a 0.619a
Mosque attending .555a
Shari’a .503a
note: Model 1: χ2 = 721.625, df = 84, α < .001, CFI = .93, RMSEA = .05; model 2: χ2 = 2,430, df = 408, p <.000, CFI = .90, RMSEA = .04; model 3: χ2 = 1,118, df = 120, p
<.000, CFI = .91, RMSEA = .03.
a
p < .001.
b
p < .01.
c
p < .05.
d
p < .10.
90
The data thus show that people who identified with religion and have a stron-
ger feeling of national pride are more favorable toward the values of religious
nationalism, whereas those who are identified with nation and have a weaker
national pride are oriented toward liberal values.
Model 2 estimates the structural coefficients of the relationships between
the independent variables and a subset of the variables for which data from
the eight countries are available, including gender equality, secular orientation,
democracy, and hostility toward immigrant workers. In this model, also, all
the estimates of the structural coefficients linking national identity to gender
equality and secular orientation are significant and in the expected direction.
The values of the structural coefficient of national identity on gender equality
range between β = 0.099 for Saudi Arabia and β = 0.329 for Indonesia. The
link between this variable and secular orientation is also positive across the
eight countries; its values range from β = 0.134 for Bangladesh and β = 0.405
for Algeria. Furthermore, the link between national identity and democracy
is positive and significant for six of the eight countries: Algeria (β = 0.059),
Bangladesh (β = 0.159), Indonesia (β = 0.081), Iran 2000 (β = 0.045), Saudi
Arabia (β = 0.125), and Turkey (β = 0.066). Finally, national identity is linked
significantly negatively to hostility toward immigrants in five of the eight
cases: Algeria (β = –0.063), Iran 2000 (β = –0.52), Jordan (β = –0.051), Morocco
(β = –0.089), and Turkey (β = –0.061). In Bangladesh (β = 0.059), it is linked
positively to hostility toward immigrants; in the other two countries this link-
age is not significant.
National pride, on the other hand, is linked significantly negatively to gen-
der equality among five of the eight countries: Bangladesh (β = –0.185), Iran
2000 (β = –0.182), Morocco (β = –0.077), Saudi Arabia (β = –0.347), and Turkey
(β = –0.168). It is, however, linked negatively to secular orientations across the
eight countries, with values ranging from β = –0.434 for Turkey to β = –0.189 for
Indonesia—all of which are significant. Moreover, national pride has a significant
effect on democracy only in Bangladesh (β = 0.060) and Indonesia (β = 0.127).
Finally, concerning the linkage between national pride and hostilities toward
immigrants, the estimates are inconsistent. Only five of the estimates are signifi-
cant. Two are linked positively (Iran [β = 0.104] and Turkey [β = 0.088]) and three
are linked negatively (Bangladesh [β = –0.083], Morocco [β = –0.074], and Saudi
Arabia [β = –0.078]).
Finally, Model 3 uses data from Iraq 2004, Iraq 2006, and Lebanon, which
include information on a different subset of the dependent variables: gender
equality, secular orientation, secular politics, and Western culture. Attitude toward
secular politics is used instead of democracy in this model. Data on hostility
toward immigrant labor were not available in these surveys, and attitudes toward
Western culture were available for only the two Iraq surveys. Given this difference
between Iraq and Lebanon, two versions of Model 3 are estimated. The first ver-
sion excludes the variable Western culture to estimate the structural parameters
for the three countries simultaneously. The second version includes this vari-
able and estimates the structural parameters for Iraq 2004 and Iraq 2006. The
results show there is no significant difference in the estimates of the structural

91
coefficients between the two versions of Model 3 (not shown). Therefore, esti-
mates from version one of Model 3, which includes Iraq 2004, Iraq 2006, and
Lebanon, and estimates of the linkages of the independent variables with Western
culture from version two are reported in Table 3.3.
According to the estimates of Model 3, the relationship of national identity
with gender equality, secular orientation, secular politics, and Western culture (for
Iraq surveys only) are positive and significant across Iraq 2004, Iraq 2006, and
Lebanon. On the other hand, national pride has consistently opposite relation-
ships with these variables and is linked negatively to gender equality, secular ori-
entation, secular politics, and Western culture (for Iraq surveys) across the three
national values surveys.
Findings from the 12 surveys thus show that the people who consider nation
as the primary basis of their identity, compared with those who consider religion
as such, have more favorable attitudes toward gender equality, secular orientation,
secular politics (in the four cases for which data are available), democracy (except
for Jordan and Morocco), and Western culture (in the three cases for which data
are available), and significantly weaker hostility toward immigrant labor in five of
the nine cases for which data are available (in three cases it was not significant
and in one case it was linked positively). National pride, on the other hand, has
consistently negative relationships with gender equality (in 9 of 12 cases), secular
orientation, secular politics, and Western culture. It is linked positively to democ-
racy in only two cases and has inconsistent relationships with hostility toward
immigrants across the eight countries. All in all, the estimates appear to support
the view that the difference in the basis of identity between religion and nation
is associated with a significant difference in orientation toward liberal values (the
liberal modality of political sovereignty) and that national pride is a driver of con-
servative values (the Islamic fundamentalist modality).
Socioeconomic Status
Socioeconomic status is linked positively to gender equality and secular ori-
entation across all the cases, except in Saudi Arabia, where socioeconomic status
is linked negatively to gender equality and has no relationship with secular ori-
entation. The variable is linked positively to democracy in six of the eight cases:
Bangladesh (β = 0.107), Indonesia (β = 0.139), Iran 2000 (β = 0.057), Jordan (β
= 0.077), Morocco (β = 0.158), and Turkey (β = 0.080), and with secular poli-
tics across the four surveys in Model 3. The relationship between socioeconomic
status and hostility toward immigrants is either negative in five cases—Algeria
(β = –0.171), Bangladesh (β = –.176), Iran 2000 (β = –0.052), Jordan (β = –0.055),
and Turkey (β = –.063) or not. Finally, socioeconomic status has no relation-
ship with attitudes toward Western culture. Such a generally positive connection
between socioeconomic status and liberal values is consistent with the modernist
proposition that relates the rise of liberal democracy to economic prosperity.
Gender and Age
Men are less favorable to gender equality than women across all 12 surveys. In
half the cases—Jordan (β = 0.247), Morocco (β = 0.114), Saudi Arabia (β = 0.127),

92
Iraq 2004 (β = 0.051), Iraq 2006 (β = 0.062), and Lebanon (β = 0.082)—men
have a stronger secular orientation than women. On democracy and secular pol-
itics, there are no gender differences, except in Jordan (β = 131) and Iraq 2006
(β = 0.059), where men are more favorable to these values. On Western cul-
ture, only in Iran 2005 were men less favorable to Western culture than women
(β = –0.030), and there is no significant gender difference in other countries.
Finally, there is no relationship between gender and attitude toward immigrant
labor, except in Bangladesh (β = 0.047) and Jordan (β = 0.149), where men are more
hostile than women, and Indonesia (β = –0.068), where the opposite is the case.
Age does not display any discernible pattern of relationship with the depen-
dent variables. In Iran 2005 (β = 0.074) and Algeria (β = 0.091), older people have
more egalitarian attitudes toward women. This relationship is just the opposite
in Bangladesh (β = –0.236), Iran 2000 (β = –0.183), and Morocco (β = –0.119),
and is not significant in other countries. Furthermore, age has a positive effective
on secular orientation in Iraq 2004 (β = 0.069) and a negative effect in Lebanon
(β = –0.119). It is linked positively to democracy only in Indonesia (β = 0.095),
Iran 2000 (β = 0.111), and Turkey (β = 0.088). It is also linked positively to secular
politics in Iraq 2004 (β = 0.094), but negatively in Lebanon (β = –0.055). Finally,
age is linked significantly negatively to hostility toward immigrants in Algeria
only (β = –0.075).
Discussion
Nationalism and nationalist movements appear in a variety of forms and some-
times contemporaneously in the same society. This variety cannot be explained
fully in terms of the parameters of the modernity, primordial, or perennial per-
spectives. I argued that nationalisms are the outcomes of the efforts of indigenous
intellectual leaders to resolve issues concerning the nature of their political com-
munities. The way issues are resolved is the function of the ideological targets
these intellectual leaders encountered. As these targets change or fade away and
new targets emerge, issues are resolved in a different manner—hence, the rise of
diverse forms of nationalism.
I used the concept of modality in order to handle this diversity more effectively.
Modalities represent the configurations of the principles of collective sovereignty
into homogeneous constellations. They capture competing types of nationalism
that emerge in a given country, region, or broader cultural tradition. Modalities thus
indicate distinctive resolutions of issues on the identity of the political community,
the line demarcating in-group and outgroup members, and the ideal form of politi-
cal regime. I proposed that identity is the key feature of these modalities. I also
proposed that modalities are also associated with different types of feeling about
one’s political community; people who are liberal nationalists, for example, feel
differently about their political community than those who are Islamic nationalists.
Using these theoretical propositions, I explained the production of such modal-
ities as liberal territorial nationalism, pan-Arab nationalism, and Islamic nation-
alism in terms of the varying ideological targets the intellectual leaders of these

93
movements encountered in their sociopolitical environment. Liberal territorial

nationalism was produced in oppositional relations to the ideology of monarchical
absolutism, ulama obstructionism, and foreign domination; pan-Arab national-
ism to the post-World War I colonial partitioning of Arab territories and the forma-
tion of disparate Arab states; and Islamic nationalism to the secular ideologies and
policies of the interventionist state in the 20th century.
To demonstrate further the fruitfulness of the concept of modality for a better
understanding of the diversity of nationalist movements, I considered the modali-
ties of territorial nationalism and Islamic nationalism, and assessed the extent to
which the two types of modalities are configured in perceptions of ordinary indi-
viduals. Using data from 12 values surveys carried out in Muslim-majority coun-
tries and considering that these modalities are distinguished in terms of (a) basis of
identity and (b) nationalist affectivity, I reasoned that changes in the basis of identity
and national pride are linked to changes in attitudes toward significant sociopolitical
issues, including gender equality, secular politics, Western culture, and outsiders.
The analysis of the data from the 12 surveys showed that the change in people’s
basis of identity from religion to nation was linked to changes in favorable atti-
tudes toward gender equality, secular orientation, secular politics, and Western
culture across all the countries for which data were available and toward democ-
racy in all the surveys, except in Jordan and Morocco. National pride, on the other
hand, was linked negatively to attitudes toward gender equality, secular orienta-
tions, secular politics, and Western culture, except that it was not linked signifi-
cantly to gender equality in Algeria. It had no significant link with attitudes toward
democracy, except in Bangladesh and Indonesia, where it was linked positively. In
terms of attitudes toward outsiders, findings were inconsistent. People who iden-
tified with nation were less hostile to immigrant labors than people who identified
with religion in five of the nine cases for which information on this variable was
available, more hostile in one case only, and no significant difference was noted
for the remaining three cases. The relationship between this variable and national
pride was even more inconsistent across nine surveys: three cases were shown to
be positive; three, negative; and the rest had no significant relationship.
Nonetheless, that the two key variables of this study—national identity and
national pride—had consistently significant links with almost all the dependent
variables in predictable directions across the 12 surveys is remarkable. The analy-
sis thus provides significant support for using the concept of modalities to capture
different clusters of homogeneous elements that rest on varying understandings
of identity and national affectivity. As shown by findings from the 12 surveys, indi-
viduals who identified with territorial nation and have lower national pride tend
to cluster toward liberal values, whereas those identified with religion lean toward
religious fundamentalism.
Finally, the analysis provided support for a version of the modernist perspec-
tive on nationalism. It showed that liberal values are tied significantly to higher
socioeconomic status across virtually all the surveys. This connection between
socioeconomic status and liberalism lends credence to a view that the modernist
interpretations of historical cases of nationalism may apply only to the liberal–
territorial modality of nationalism.

94
The contrasting linkages of national identity and national pride with the two
modalities of collective sovereignty, where national identity reinforces liberal
values whereas national pride emphasizes Islamic nationalism, are drawn on
analyzing cross-sectional survey data. However, if these linkages are interpreted
as snapshots of the dynamics of encounters between those who are oriented
toward liberal values and those oriented toward religious nationalism, the fore-
going analysis thus provides a framework for speculating why a section of the
Islamic population in Muslim-majority countries has responded violently, almost
always, to (perceived) anti-Islamic behaviors by some individuals or groups in the
West, whereas it appears that these same people have often remained dormant
in responding to violent extremism in their midst or to brutal acts of repressions
committed by the ruling regimes in these countries. Insofar as liberal national-
ists and Islamic nationalists are competing for the intellectual control of their
society, uproar against the West under the excuse of the latter’s violence against
Islam is functional for religious nationalism, because it contributes to the mobi-
lization of national pride. And this mobilization in turn functions to strengthen
the fundamentalist-cum-religious nationalist values and weaken liberalism. As a
prime example of this process, the Islamic clerical rulers in Tehran—to monopo-
lize power and reproduce this monopoly—have used anti-Western and, more so,
anti-American rhetoric and policies effectively as the cornerstone of their rule in
the country. The seizure of the US embassy in Tehran, the continuation of Iran–
Iraq war for 6 years after the Iraqis forces were pushed back to the prewar borders,
the call for the execution of Salman Rushdie, and the brutal repression of the
green movement, which were all conducted under the pretext of war against the
United States, were, in reality, conducted to weaken the liberal national alterna-
tive to their rule and strengthen Shia fundamentalism. As this analysis shows,
it appears that the same dynamic is at work in other Muslim-majority countries,
where the mobilization of national pride tends to reinforce the idea of gender
inequality, religious rule, and the rejection of secular politics and the West—all
consistent with the Islamic fundamentalist modality of nationalism.
The theoretical propositions suggested in this work may also be applicable to
understanding the dynamics of nationalism and national politics in other cases
as well. For example, Russian nationalism may not be construed as a unitary and
monolithic phenomenon. Rather, conflicting modalities of collective sovereignty
may be at work in the country today, as the ruling elite uses hostilities toward
gay and lesbian communities and the West—the United States in particular—
as a means to provoke national pride, and hence solidify its rule in the country
and undermine the influence of the liberal opposition. In western Europe, on the
other hand, the formation of the European Union and the rise of European iden-
tity may moderate the traditional oppositional dichotomy between the modalities
of nationalism and patriotism.
In these and other cases, however, to comprehend the range of variation of
modalities of nationalism entails understanding the number and nature of the
sociopolitical and cultural issues being contested in the society. Included among
these issues are (a) the significance of secular versus religious discourses in shap-
ing the norms of the sociopolitical community, (b) people wishes versus religious

95
laws as the governing principle of this community, (c) individual liberty as the foun-
dation of the nation versus the subordination of this liberty to the idea of maintain-
ing national unity against a (perceived) enemy of the nation, (d) territorial nation
versus religion as the basis of identity, (e) individual equality versus patrimonial
domination, (f) gender equality versus gender hierarchy and male domination,
(g) religious monopoly and intolerance versus religious diversity and tolerance,
and (h) fulfillments of economic needs versus protection of cultural values in
formulating national priorities. Identifying the significance of these and similar
issues in sociopolitical and cultural debates and understanding the process of the
resolution of these issues provide clues for understanding the range of variation
in the modalities of collective sovereignty that a country may come to experience.
It would certainly be hard to project the trajectory of nationalist movements
into the future. Nonetheless, understanding the type of issues being contested,
considering that issues are often resolved in oppositional relation to the ideol-
ogy of the ruling political regime or the cultural institutions that are dominant in
society, and given that there are only finite ways in which issues may be resolved,
using the proposed conceptual schema may enable us to speculate more effec-
tively the types of modalities most likely to emerge in the society in the future.
Acknowledgment
I am grateful to Colter Mitchell for his assistance on AMOS, and to Ronald
Aminzade, Stuart Karabenick, and Julie de Jong for comments.
Note
1. Anderson offers a misguided rationale for his conceptualization of national-
ism. A nation, says Anderson (1983, p. 6), “is imagined because the members of even
the smallest nation will never know most of their fellow-members, meet them or even
hear of them, yet in the minds of each lives the image of their communion.” This
statement implies that any large political or economic organization can be considered
an imagined community because not all the members will ever know one another
whereas, in reality, such organizations in fact exist. The concept of “imagined,” how-
ever, is more in tune with the cognitive turn in the study of nationalism and ethnicity.
In fact, the notion that nation is invented or imagined indicates a departure from
substantialism and is consistent with a premise that nations are “not things in the
world, but perspectives on the world” (Brubaker et al., 2004, p. 45).
References
Abrahamian, E. (1982). Iran between Two Revolutions. Princeton: Princeton University
Press.
Adamiyat, F. (1955) [1334]. Amir Kabir va Iran. Tehran: Amir Kabir Publications.

96
———. (1970). Andishaha-yi Mirza Fath Ali Akhundzadah. Tehran: Kharazmi.

———. (1976) [1355]. Idi’olozhi-ye mahzat-i mashrutiyat-i Iran [The ideology of the
constitutional movement in Iran]. Tehran: Payam Publications.
Afary, J. 1996. The Iranian Constitutional Revolution, 1906– 1911: Grassroots
Democracy, Social Democracy, and the Origin of Feminism. New York: Columbia
University Press.
Ahmed, J. M. (1960). The intellectual origins of Egyptian nationalism. Oxford: Oxford
University Press.
Akhavi, S. (1992). The Clergy’s Concepts of Rule in Egypt and Iran. Annals of the
American Academy of Political and Social Science, 524(November), 92–102.
Amanat, A. (1997). Pivot of theUniverse: Nasir al-Din Shah Qajar and the Iranian
Monarchy, 1831–1896. Berkeley: University of California Press.
Aminzade, R. (2013). Race, nation, and citizenship in postcolonial Africa.
Anderson, B. (1983). Imagined communities: Reflections on the origin and spread of
nationalism. London: Verso.
Ansari, H. (1986). Egypt, the Stalled Society. Albany: State University of New York Press.
Armstrong, J. (1982). Nations before Nationalism. Chapel Hill: University of North
Carolina Press.
Brass, P. (1991). Ethnicity and nationalism. London: Sage.
Boroujerdi, M. (1996). The Iranian Intellectuals and the West: The Tormented Triumph
of Nativism. Syracuse: Syracuse University Press.
Breuilly, J. (1994). Nationalism and the State. Chicago: The University of Chicago press.
Brubaker, R., Loveman, M. and Stamatov, P. (2004). “Ethnicity as Cognition,” Theory
and Society, 33, 31–64.
Burnstein, E., Crandall, C., & Kitayama, S. (1994). Some neo-Darwinian decision
rules for altruism: Weighing cues for inclusive fitness as a function of the genetic
importance of the decision. Journal of Personality and Social Psychology, 67,
773–789.
Chartouni-Dubarry, M. (1993). “The Development of Internal Politics in Iraq from
1958 to the Present Day,” pp. 19–36, in D. Hopwood, H. Ishow, and T. Koszinowski,
eds., Iraq: Power and Society. Reading: Ithaca Press.
Che Guevara, E. (1969). Che Guevara Speaks: Selcted Speeches and Writings. Edited by
Steve Clark. Atlantia: Pathfinder Press.
Christelow, A. (1985). Muslim law courts and the French colonial state in Algeria.
Princeton, NJ: Princeton University Press.
———. (1987). Ritual, Culture and Politics of Islamic Reformism in Algeria. Middle
Eastern Studies, 23, no. 3 (July), 255–273.
Cleveland, W. L. (1971). The making of an Arab nationalist: Ottomanism and Arabism
in the life and thought of Sati’ al-Husri. Princeton, NJ: Princeton University Press.
Conner, W. (1994). Ethno- Nationalism: The Quest for Understanding. Princeton:
Princeton University Press.
Crecelius, D. (1972). “Nonideological Responses of the Egyptian Ulama to
Modernization,” pp. 167–210 in Scholars, Saints, and Sufis, edited by N. R. Keddie.

97
Cromer, E. B. (1908). Modern Egypt. New York: Macmillan.

Dawisha, A. (2003). Arab nationalism in the twentieth century: From triumph to despair.
Dawn, C. E. (1973). From Ottomanism to Arabism. Urbana: University of Illinois Press.
———. (1988). The formation of pan-Arab ideology in the interwar years. International
Deeb, M. (1997). Islam and national identity in Algeria. Muslim World, 87(2), 111–128.
———. (1979). Party Politics in Egypt: The Walf and Its Rivals, 1919– 1939.
London: Ithaca Press.
de Figueiredo, R. J. P., Jr., & Elkins, Z. (2003). Are patriots bigots? An inquiry into the
vices of in-group pride. American Journal of Political Science, 47(1), 171–188.
Fox, J. (2004). The Rise of Religious Nationalism and Conflict: Ethnic Conflict and
Revolutionary Wars, 1945–2001. Journal of Peace Research, 41, 6 (Nov.), 715–731.
Eller, J. D., & Coughlan, R. M. (1993). The poverty of primordialism: The demystifica-
tion of ethnic attachments. Ethnic and Racial Studies, 16(2), 183–202.
Eriksen, T. H. (1993). Ethnicity and nationalism. London: Pluto Press.
Fishman, J. A. (1972). Language and Nationalism: Two Integrative Essays. Rowley,
MA: Newbury House.
Fishman, J. A., & Gertner, M. H. (1985). The Rise and Fall of the Ethnic Revival: Perspectives
on Language and Ethnicity. New York: Mouton.
Floor, W. M. (1976). “The Merchants (tujjar) in Qajar Iran.” Zeitschrift der Deutschen
Morgenlandischen Gesellschaft, 126, 101–135.
Friedland, R. (2001). Religious nationalism and the problem of collective representa-
tion. Annual Review of Sociology, 2, 125–152.
———. (2002). Money, sex, and God: The erotic logic of religious nationalism.
Sociological Theory, 20(3), 381–425.
Geertz, C. (1973). The Interpretation of Cultures. London: Fontana.
Gellner, E. (1964). Thought and change. London: Weidenfeld and Nicolson.
———. (1983). Nations and nationalism. Ithaca, NY: Cornell University Press.
Gershoni, I., & Jankowski, J. P. (1995). Redefining the Egyptian nation, 1930–1945.
Gramsci, A. (2007). Prison notebooks (Vol. 3) (J. A. Buttigieg, Trans.). Columbia
University Press.
Haim, S. G. (Ed.). (1962). Arab nationalism: An anthology. Berkeley: University of
California Press.
Hamilton, W. D. (1964). The Genetical Evolution of Social Behavior, Part I and II.
Journal of Theoretical Biology, 7, 1–52.
Hechter, M. (1977). Internal Colonialism: The Celtic Fringe in British National
Development. London: Routledge Kegan Paul.
Heyworth-Dunne, J. (1968). An Introduction to the History of Education in Modern
Egypt. London: Frank Cass & Co.
Hobsbawm, E. (1977). Some Reflections on ‘The Break-up of Britain’. New Left Review,
105(September-October), 3–23.
Hobsbawm, E., & Ranger, T, (Eds.). (1983). The Invention of Tradition. Cambridge:
Cambridge University Press.

98
Hourani, A. (1983). Arabic thought in the liberal age, 1798–1939. Cambridge: Cambridge

University Press.
Ibn Khaldun, A. al-R. (1967) [1377]. The muqaddhimah: An introduction to history
(R. Rosenthal, Trans.). Princeton, NJ: Princeton University Press.
Issawi, C. (1954). Egypt at Mid- century: An Economy Survey. London: Oxford
University Press.
———. (1966). The Economic History of the Middle East, 1800–1914. Chicago: University
of Chicago Press.
Jahanbegloo, R. ed. (2004). Iran: Between Tradition and Modernity. Lexington Press.
Juergensmeyer, M. (1993). The new Cold War? Religious nationalism confronts the secu-
lar state. Berkeley: University of California Press.
———. (1996). The worldwide rise of religious nationalism. Journal of International
Affairs, 50(1), 1–20.
Kamrava, M. (2008). Iran’s Intellectual Revolution. Cambridge: Cambridge University
Press.
Kashani-Sabet, F. (1999). Frontier fictions: Shaping the Iranian nation, 1804–1946.
Kedourie, E., ed. (1971). Nationalism in Asia and Africa. London: Weidenfeld and
Nicholson.
Khalidi, R., Anderson, L., Muslih, M., & Simon, R. S. (Eds.). (1991). The origins of Arab
nationalism. New York: Columbia University Press.
Khomeini, R. (1981). Islam and revolution: Writings and declarations of Imam Khomeini
(H. Algar, Trans., annot.). Berkeley, CA: Mizan Press.
Khoury, P. S. (1983). Urban notable and Arab nationalism. Cambridge: Cambridge
University Press.
———. (1987). Syria and the French mandate: The politics of Arab nationalism, 1920–
1945. Princeton, NJ: Princeton University Press.
Kinross, P. (1977). The Ottoman Centuries: The Rise and Fall of the Turkish Empire.
London: Perennial.
Kohn, H. (1967) [1944]. The idea of nationalism (2nd ed.). New York: Collier-Macmillan.
Korchmaros, J. D., & Kenny, D. A. (2001). Emotional Closeness as a Mediator of the Effect
of Genetic Relatedness on Altruism. Psychological Science, 12, No. 3 (May), 262–265.
Lapidus, I. M. (1988). A History of Islamic Societies. Cambridge: Cambridge
University Press.
Mahdavi, M. (2011). Post-Islamist Trends in Post-revolutionary Iran. Comparative
Studies of South Asia, Africa and the Middle East, 31(1), 94–109.
Mann, M. (1993). The Sources of Social Power, Vol. 2: The Rise of Classes and Nation-
States, 1760–1914. New York: Cambridge University Press.
Ma’oz, M. (1968). Ottoman Reform in Syria and Palestine, 1840–1861: The Impact of the
Tanzimat on Politics and Society. Oxford: Oxford University Press.
Marsot, A. L. (1968). Egypt and Cromer: A study in Anglo- Egyptian relations.
New York: Praeger.
———. (1977). Egypt’s Liberal Experiment: 1922–1936. Los Angeles: University of
California Press.

99
———. (1984). Egypt in the Reign of Muhammad Ali. Cambridge: Cambridge

University Press.
Mateo, J. M. (1996). The development of alarm-call response behavior in free-living
juvenile Belding’s ground squirrels. Animal Behavior, 52, 489–505.
Milner, A. (1892). England in Egypt. London: Macmillan.
Mitchell, R. P. (1969). The Society of the Muslim Brothers. London: Oxford University
Press.
Moaddel, M. (1993). Class, politics, and ideology in the Iranian Revolution. New York:
Columbia University Press.
———. (2005). Episode and discourse: Islamic modernism, nationalism, and Islamic fun-
damentalism. Chicago, IL: The University of Chicago Press.
———. (2009). The Iranian Revolution and its nemesis: The rise of liberal values
among Iranians. Comparative Studies of South Asia, Africa, and the Middle East,
126–136.
Moaddel, M., & de Jong, J. (2013). Trends in values among Saudi youth: Findings
from values survey. Journal of the History of Childhood and Youth, 6(1), 151–164.
Moaddel, M., de Jong, J., & Dagher, M. (2011). Beyond Sectarianism in Iraq. Context,
10, 366–367.
Moaddel, M., & Karabenick, S. (2013). Religious Fundamentalism in the Middle
East: A Cross-National, Inter-Faith, and Inter-Ethnic Analysis. Leiden: Brill.
Nairn, T. (1977). The Break-up of Britain: Crisis and Neo-Nationalism. London: New
Left Book.
Özkirimli, U. (2000). Theories of Nationalism: A Critical Introduction. New York: Saint
Martin’s Press.
Quraishi, Z. M. (1967). Liberal Nationalism in Egypt: Rise and Fall of the Wafd Party.
Allahabad: Kitab Mahal.
Rajaee, F. (2007). Islamism and Modernism: The Changing Discourse in Iran.
Austin: University of Texas Press.
Reid, D. M. (1996). Cromer and the classics: Imperialism, nationalism and the Greco-
Roman past in modern Egypt. Middle Eastern Studies, 32(1), 1–29.
Robinson, F. (1993). Islam and the Impact of Print. Modern Asia Studies, 27(February),
229–251.
Ruedy, J. (1992). Modern Algeria: The origins and development of a nation.
Bloomington: Indiana University Press.
Safran, N. (1961). Egypt in search of political community. Cambridge, MA: Harvard
University Press.
Sageman, M. (2004). Understanding Terror Networks. Philadelphia: University of
Pennsylvania.
Seton-Watson, H. (1977). Nations and States: An Enquiry into the Origins of Nations and
the Politics of Nationalism. Boulder, CO: Westview Press.
Sivan, E. (1985). Radical Islam: Medieval theology and modern politics. New Haven,
CT: Yale University Press.
Smith, A. D. (1998). Nationalism and modernism. London: Routledge.
Vahdat, F. (2002). God and Juggernaut. Syracuse: Syracuse University Press.

100
—— — . (2003). Post- Revolutionary Islamic Discourses On Modernity In Iran:

Expansion and Contraction of Human Subjectivity. International Journal of Middle
East Studies, 35, 599–631.
van den Berghe, P. (1978). Race and ethnicity: A sociobiological perspective. Ethnic
and Racial Studies, 1(4), 401–411.
Vatikiotis, P. J. (1973). “Tradition and Political Leadership: The Example of Algeria.”
In Man, State, and Society in the Contemporary Maghirb, edited by I. William
Zartman. New York: Praeger.
———. (1980). The History of Egypt. 2nd ed. Baltimore: Johns Hopkins
University Press.
Viroli, M. (1995). For love of country: An essay on patriotism and nationalism.
Oxford: Clarendon Press.
Seton-Watson, H. (1977). Nations and States: An Enquiry into the Origins of Nations and
the Politics of Nationalism. Boulder, CO: Westview Press.
Wendell, C. (1972). The evolution of the Egyptian national image: From its origins to
Ahmad Lutfi al-Sayyid. Berkeley: University of California Press.
Zeine, Z. N. (1973). The emergence of Arab nationalism. Delmar, NY: Caravan Books.
Zürcher, E. J. (2004). Turkey: A modern history. London: I.B. Tauris.

101
CHAPTER 4 Modernization, World System, and

Clash of Civilization Perspectives
in Lay Views of the Development–
Morality Nexus in the United States
and the Middle East
Arland Thornton, Kathryn M. Yount,
Linda Young-DeMarco, and Mansoor Moaddel
In this chapter we examine the linkages between people’s perceptions of

national development and their perceptions of national morality. We are inter-
ested in whether development and morality are viewed as complementary or con-
flicting. We expect that this relationship varies across settings, with residents in
the United States more likely to see a positive relationship between morality and
development, and Middle Easterners more likely to express mixed views. Within
societies, we expect that certain personal and social attributes influence people’s
views of the morality–development nexus. We assess these propositions using
probability samples of ordinary people in Egypt, Lebanon, Saudi Arabia, and the
United States.
For centuries, many elites have believed that national development and moral-
ity are linked positively. They also have posited causal connections between devel-
opment and morality, suggesting these concepts are mutually causal (Friedman,
2005). The presumed linkages have informed governmental policies for develop-
ment that have affected populations worldwide (Thornton, 2005). Yet, we know
little about how ordinary people perceive the connections between morality and
development. This knowledge is needed in today’s globalized world. Varying
understandings of the relationship between morality and development may shape
laypeople’s political, economic, and social behavior, and contribute to intercultural
harmony or conflict.
We situate our research within three theoretical perspectives. The first is mod-
ernization theory, which has its roots in the Enlightenment and has dominated
the social sciences for centuries. Modernization theory has posited strong posi-
tive connections between development and morality. The second theoretical per-
spective is world system theory, which emerged during the late 20th century in
opposition to modernization theory. Instead of seeing morality and development
as linked positively, world system theory has viewed them as related negatively.
102
The third theoretical framework is an adaptation of the “clash of civilizations”

approach that has viewed different cultures in contention, with cross-cultural dif-
ferences affecting people’s views of morality. We modify this framework by argu-
ing that “clash of moralities” may produce either negative or positive views of
the relationship between development and morality, depending on the viewer’s
beliefs and values concerning morality.
Our goal is not to test the validity of these three perspectives in explaining
morality or social change. Rather, we assess which of these three perspectives pro-
vides a more effective prediction of the development–morality nexus in people’s
perceptions. This is a useful question because many elements of world culture
have been globalized and have permeated societies in many parts of the world
(Krücken & Drori, 2009; Meyer, Boli, Thomas, & Ramirez, 1997). Such evidence
leads us to ask whether the elements of modernization theory suggesting a posi-
tive connection between development and morality, world system theory sug-
gesting a negative relationship, or the clash of moralities perspective are most
accepted among ordinary people. Our aim is thus to evaluate how these diverse
perspectives are reflected in the sociological imaginations of lay people.
Egypt, Lebanon, Saudi Arabia, and the United States provide important con-
texts for assessing people’s perceptions of the morality–development relation-
ship. The United States is one of the world’s wealthiest countries, having long
experience with democracy, an emphasis on individual liberty, a foundation of
religious pluralism and tolerance, and the norm of separation of religion and poli-
tics. Egypt and Saudi Arabia have, for centuries, been two of the most important
centers of Islamic learning and culture. Economically, they are less wealthy than
the United States, with income levels being considerably greater in Saudi Arabia
than in Egypt (United Nations Development Programme, 2010). These two soci-
eties have had extensive contact during the past century with ideas of modernity
and have experienced substantial social change (see Yount and Rashad [2008] on
Egypt). Lebanon is one of the most religiously heterogeneous countries in the
Middle East, with deep roots in Islam and Christianity. Since the 19th century, it
has played a prominent role in cultural movements in the region and has expe-
rienced considerable conflict. The three Middle Eastern countries also have had
historically different relationships with the United States and Europe. We expect
that these national differences will be reflected in people’s views of the relation-
ship between morality and development.
We address these issues using two recent surveys from Egypt and single
surveys from Lebanon, Saudi Arabia, and the United States. In these surveys,
respondents were asked to rate several countries on morality and development.
We examine how respondents’ ratings of countries on morality relate to their rat-
ings of the same countries on development. We know from earlier research that
people from many countries, including Egypt, Lebanon, Saudi Arabia, and the
United States, are able to rate countries on development, and they do so similarly
to the United Nations (UN) Human Development Index (HDI) (Thornton et al.,
2012). We are, however, unaware of prior research in which people rate countries
on morality. We also investigate two questions asking Egyptians to rate directly the
relationship between development and morality. One question asked respondents

103
to compare the morality levels of developed and less developed places. Another
asked respondents to evaluate the effects of a development program on morality.
We expect that the data will reveal great heterogeneity across and within coun-
tries in perceptions of the relationship between morality and development. We
also expect that the views of individuals concerning the morality of countries will
depend on their religiosity, beliefs about the role of religion in public life, percep-
tions about the interventions of foreign countries in local affairs, and values and
beliefs concerning homosexuality, transactional sex, consumption of alcohol, and
styles of dress. We investigate these expectations using multivariate regression.
Our interest in studying development–morality linkages is not informed by any
belief about whether development is desirable or undesirable, moral or immoral.
Instead, we wish simply to understand how ordinary people view morality and its
relationship to development.
Our chapter has four sections. First, we focus on theoretical frameworks and
their implications for views of development and morality. Second, we describe the
surveys, the questions used to assess people’s views of development and moral-
ity, and our approach to data analysis. Third, we present our results. Finally, we
discuss our conclusions and their implications.
Concepts, Theories, and Hypotheses

Modernization Theory
Although modernization theory has many versions, its general dimensions can
be traced to the Enlightenment (Harris, 1968; Mandelbaum, 1971; Nisbet, 1975
[1969]). Among these general dimensions was the idea of a universal process of
national advancement through incremental stages (Levy, 1966; Parsons, 1964).
Countries were seen as progressing at different rates and were thus located at
different developmental stages at any given time. The most modern places were
believed to be in northwest Europe and its overseas populations, and other coun-
tries were believed to be lower on the development ladder. The model suggested
that developed countries represented the futures of the less developed and that, as
the latter advanced, they would resemble increasingly their more developed peers
(Berkhofer, 1978; Carniero, 1973; Harris, 1968).
The identification of development with western Europe and North America,
and of lesser development with other places specified what was and was not devel-
oped. In socioeconomic terms, modernity was believed to include widespread
affluence, technological innovation, roads and other public infrastructures, cit-
ies, education, free markets, and industrial economies. In political terms, moder-
nity was believed to include democracy, freedom, equality, and the separation of
church and state. Less developed societies were identified as having less of these
attributes. Modernization theory concluded that these attributes co-occurred geo-
graphically and were interrelated causally (Apter, 1965; Huntington, 1968).
Most Western proponents of modernization theory believed development
and morality went together (Friedman, 2005). Many thinkers of the 18th and
Modernization, World System, And Clash Of Civilization | 103

104
19th centuries were moral philosophers who believed morality and economic
well-being were related positively and causally. For example, Auguste Comte,
an important French scholar of the 19th century, said that “all human progress,
political, moral, or intellectual, is inseparable from material progression” (cited in
Friedman, 2005, p. 31). Such views have been expressed by a long line of writers
to this day (Friedman, 2005).
Many have seen modernization first emerging in northwest Europe and its
overseas populations and spreading elsewhere thereafter through multiple diffu-
sion mechanisms (Apter, 1965). Modernization theory has viewed these interna-
tional ties as beneficial in that they allow less developed countries to exploit their
“comparative advantages” of abundant labor and natural resources to increase
production of exports and to import manufactured goods and scarce means of
production from modernized countries (Kindleberger, 1965; Meier & Baldwin,
1966; Moaddel, 1994). The modernization approach suggests that such relations
may accelerate the progress of less developed countries. Thus, this view paints the
developmental system as natural and proper, and the countries at the top of this
hierarchy as good and moral.
During the 20th century, strong critiques have been mounted against modern-
ization theory (Böröcz, 2000; Mandelbaum, 1971; Nisbet, 1975 [1969]; Wallerstein,
1991). In addition to the critiques by world system theorists (discussed later), the
developmental framework is teleological, and its assumption of uniform and
directional social change is indefensible. As a result, modernization theory is
much less accepted in academia today than in earlier periods.
Nevertheless, the ideas of modernization theory are exceptionally powerful
internationally among many elites and ordinary people. Anthropologists and his-
torians have identified the presence and power of these ideas among ordinary peo-
ple and intellectual leaders in Sub-Saharan Africa, China, India, Nepal, and New
Guinea (Ahearn, 2001; Caldwell, Reddy, & Caldwell, 1988; Dahl & Rabo, 1992;
Ferguson, 1999; Liechty, 2003; Osella & Osella, 2006; Pigg, 1992; Wang, 1999).
More important for our research, these ideas have been documented among people
in Egypt, Lebanon, and the Arabian Peninsula (Abu-Lughod, 1998; Deeb, 2006;
Moaddel, 2005; Osella & Osella, 2006). Ferguson (1999, p. 84) has argued that
urban African workers’ conceptions were “not simply compatible with the modern-
ist narratives of social science; they were a local version of them. Modernization
theory had become a local tongue” (emphasis in original). Likewise, Osella and
Osella (2006), anthropologists of the Arabian Peninsula and South Asia, have
written that “concepts, ideals, and practices of something called ‘modernity’ …
exist and are continually appealed to in people’s economic endeavours, political
projects, and identity crafting” (pp. 570–571).
Similar evidence is accumulating from surveys showing that ordinary peo-
ple in diverse countries know and adhere to modernization theory and hier-
archical models of development (Binstock & Thornton, 2007; Mitchell, 2009;
Thornton, Binstock, & Ghimire, 2008; Thornton et al., 2012). The ideas of
modernization theory also are widely accepted among policymakers and gov-
ernmental and nongovernmental organizations, such as the UN, World Bank,
and International Monetary Fund (International Monetary Fund, 2009; Latham,

105
2000; Meyer et al., 1997; Nisbet, 1980; UN Development Programme, 2009;
World Bank, 2011).
Thornton (2001, 2005) has suggested that the ideas of modernization have
been so widely disseminated internationally that they form a package of ideas
called developmental idealism. This package of ideas suggests that the socioeco-
nomic, familial, and political elements defined as modern are themselves good,
will help to produce other good things, and should be strived for. As Friedman
(2005) asserted, the good things associated with modernity are also often judged
as moral.
World System Theory
World system theory rejects modernization theory, including the idea of a uni-
versal model of national development (Chase-Dunn & Rubinson, 1977; Moaddel,
1994; Wallerstein, 1974, 1997 [1979]). Instead, world system theory regards the
world economy as a hierarchy of asymmetric exchange relations that functions
for the benefits of the core economies at the expense of the peripheral countries.
According to the theory, economic exchange within this world hierarchy is based
on the unequal appropriation of surplus such that economic well-being in the
dominant, core high-income countries is gained at the expense of the subordinate,
peripheral low-income countries. Both upward and downward movement within
the world economic structure occurs, and the gains of any one country necessitate
losses for others (Wallerstein, 1997 [1979]).
According to this viewpoint, low-income countries are peripheralized in the
world economy as they are penetrated by interests originating from the core coun-
tries (Duvall, Jackson, Russett, Snidal, & Sylvan, 1981; de Janvry & Garramon,
1977; Wallerstein, 1997 [1979]). This process is espoused to involve the growing
dependence of peripheral countries on core countries because economic produc-
tion is organized increasingly according to the external dictates of the world mar-
ket and the international accumulation of capital. Ultimately, peripheralization
and dependence are theorized to limit capacity for economic growth (Bornschier,
1980; Bornschier, Chase-Dunn, & Rubinson, 1978; Duvall et al., 1981; Emmanuel,
1972; de Janvry & Garramon, 1977; Moaddel, 1994).
World system theory thus contradicts modernization theory, and challenges
its evaluation of the world’s economic system and high-income countries as good
and moral. World system theory asserts the system is built on unequal relations
and exploitation, and is therefore unjust. This assertion implies that rich coun-
tries of the core, which control and protect an unjust economic structure, are
themselves immoral.
Adherence to popular versions of world system theory has been docu-
mented. In 1964, many low-income states became associated with the Group of
77 (G77), a coalition that adopted an overt moral stance against Western domi-
nation (Furedi, 1977). The coalition criticized the exploitation of low-income
countries by wealthier ones and unequal exchange between rich and poor. The
group coalesced against a “universal model of development” and argued for the
restructuring of international trade to create a “New International Economic

106
Order” that would benefit low-income countries (Bhagwati, 1986; Lavelle, 2001).
During the 1970s, the G77 promulgated these ideas in the UN Conference on
Trade and Development, the 1974 World Population Conference at Bucharest
(Mauldin, Choucri, Notestein, & Teitelbaum, 1974), and broader diplomatic
circles. Parallel efforts emerged among some “like-minded” Western countries,
nongovernmental organizations, church groups, and academics who had an
interest in “third-world development” (Løvbrǽk, 1990). Later changes to the
international political economy weakened the position of the G77, and the UN
Conference on Trade and Development position began to disintegrate (Lavelle,
2001). Subsequently, many low-income states began to seek liberalized (as
opposed to protected) trade relations as a component of so-called development
schemes. Acceptance of liberalization, however, was not universal among the
G77. According to Lavelle (2001), many poorer states enacted the new road map
not because of a sincere change of belief, but rather because the “costs of …
defection [were] unsustainable” (pp. 45–46).
Clash of Moralities
In both modernization and world system theories, morality is linked to economic
development, but in different ways. A clash of moralities approach, adapted from
Huntington’s (1996) clash of civilizations model, provides another perspective.
In this perspective, a country’s economy, moral values, and social institutions are
shaped by the civilization of which it is a part. According to Huntington, the dif-
ferent civilizations are in contention.
However, it is not clear in Huntington’s framework what causes different world
civilizations to clash with one another or which of their aspects is central to inter-
civilizational conflict. For sure, economic interests, political rivalries, and military
power are important in the conflicts between civilizations. Power, for Huntington,
is crucial to cultural diffusion between civilizations insofar as he believes that
“culture … follows power” (Huntington, 1996, p. 310) and that as Western power
declines, the influence of Western culture declines. The inverse of this argument
also is true for Huntington: with growing power and self-confidence, other societ-
ies increasingly express and perhaps impose their “own cultural values and reject
those ‘imposed’ on them by the West” (Huntington, 1996, p. 28; Moaddel, 2010).
Nonetheless, Huntington (1996, p. 304) considers moral issues to be impor-
tant, as he believes that “moral decay” is “far more significant than economics and
demography” in the decline of Western civilization: “antisocial behavior,” “family
decay,” “a decline in ‘social capital’,” and “general weakening of the ‘work ethic’
and the rise of a cult of personal indulgence.” Thus, he argued that “the future
health of the West and its influence on other societies depends in considerable
measure on its success in coping with those trends, which, of course, give rise to
the assertions of moral superiority by Muslims and Asians” (p. 304).
We propose that an important difference between civilizations centers on peo-
ple’s conceptions of morality. Indeed, a clash of moralities may be one of the most
enduring aspects of conflict among people with different cultures. Added to the
significance of moral difference is research showing that differences in political

107
values and beliefs in democracy between people in predominantly Muslim and

Christian countries are quite small. Yet, differences are large between Muslims
and Christians in beliefs about gender equality and sexual norms, with certain
Muslim groups being less supportive of gender equality and more restrictive on
sexual practices (Esmer, 2002; Norris & Inglehart, 2002; Yount, 2004).
Given that gender relations and sexual practices are moral issues in many
societies, such differences between some Christian and Muslim groups may
reflect more a clash of moralities than a clash of civilizations. This idea is but-
tressed by the knowledge that gender relations and sexual practices are two main
areas of change in Western countries, with dramatic increases in many Western
societies in the acceptance and prevalence of abortion, same-sex partnerships,
nonmarital sex, cohabitation, and nonmarital childbearing (Inglehart & Welzel,
2005; Lesthaeghe & Neidert, 2006; Thornton, 1989; Thornton & Young-DeMarco,
2001). These behaviors also have become more prevalent in some places outside
the West (e.g., Ali, Cleland, & Shah 2003; Coltabiano & Castiglioni, 2008; Parish,
Laumann, & Mojola, 2007); however, if such changes are perceived to be the most
prevalent in countries associated with the peak of development, people may see
morality as related negatively to development.
Of course, gender relations, sexuality, family relations, and abortion are not the
only issues defining views of morality. Other issues defined by some as having
moral components include religious beliefs and practices (or lack thereof), pat-
terns of consumption, the use of violence, the use of drugs, respect for property
and laws, and the involvement of one country in the affairs of another. Differences
in views on such matters can influence people’s evaluations of a country’s morality.
Hypotheses from the Theoretical Frameworks

From the preceding discussion, we derive five general hypotheses about how lay
people in Egypt, Lebanon, Saudi Arabia, and the United States perceive develop-
ment, morality, and the interconnections between the two. We also discuss how
the five hypotheses combine and interact in the ways that they are reflected (or
not) in the beliefs of ordinary people.
Hypothesis 1 states that people’s view of development, morality, and their link-
ages vary by social context. We expect that such differences are especially marked
between the United States and the countries of the Middle East. We expect that
historical and cultural forces lead Americans to be more accepting of moderniza-
tion theory, less accepting of world system theory, and less cognizant of clash of
morality than people in the Middle East.
Hypothesis 2, derived from the clash of moralities framework, states that the
observed relationship between development and morality depends on people’s views of
morality. This framework suggests that people in countries or a region with a shared
cultural heritage have their own standards of moral conduct and schemas that affect
views of morality in various countries. If Hypothesis 2 is true, we would expect that
Middle Easterners would be especially critical of morality in Western countries.
Hypothesis 3, also drawn from the clash of moralities framework, states that
the views of people concerning the morality of a country are related directly to their

108
personal views of morality. This hypothesis recognizes that there is often extensive
heterogeneity within groups about what is moral. We expect that people who are
highly religious, and people who endorse strict interpretations of religion, gender
inequality, modesty in dress, families, cultural traditions, and sexuality will be
more likely to rate places in the West as less moral and places like Saudi Arabia,
with its strict orthodox Muslim society, as more moral than do others. Similarly,
we expect that people in the Middle East who see the influence of the United
States (or Saudi Arabia) in the affairs of their country as negative will view the
morality of the United States (or Saudi Arabia) less positively.
Hypothesis 4, drawn from modernization theory, states that there is a positive
link between people’s views about development and morality. If Hypothesis 4 is cor-
rect, people will rate countries they see as developed as more moral than countries
they see as less developed. People also would state that low levels of morality are
more common in less developed places than in developed places, and that a devel-
opment program would increase the morality of the people in a country.
Our final hypothesis, Hypothesis 5, is drawn from world system theory and
states there is a negative link between people’s views about development and morality.
If Hypothesis 5 holds, then people will rate countries they believe to be developed
as less moral than countries they see as undeveloped. People will also state that
low levels of morality are less common in undeveloped than in developed places,
and that a development program would decrease morality. Although world system
theory does not consider people living in wealthy core countries to be less moral
than people in the periphery, it considers the international economic hierarchy as
unjust and therefore immoral. Consequently, a person who is sympathetic to the
ideas of world system theory may reason that the citizens of the core countries are
benefiting from an unjust economic order, and thus would be less moral than the
people in the periphery, who are not benefiting from the normal operation of the
world economy.
We expect that many ordinary people will not have sophisticated and thorough
understandings of the theories of modernization, world system, and clash of
moralities. However, we hypothesize that the basic ideas of these frameworks are
widely disseminated. Our goal is to see how these frameworks have (or have not)
permeated to the grassroots.
Data and Methods

Our chapter is based on two surveys in Egypt and single surveys in Lebanon,
Saudi Arabia, and the United States conducted between 2005 and 2008. One
survey in Egypt was drawn from a district in Qaliubia Governorate north of Cairo
and a district in Fayoum Governorate south of Cairo. These districts represent
broadly the governates in Upper (southern) and Lower (northern) Egypt, rural
and urban areas, and various ethnic groups. We sampled women between the
ages of 18 years and 54 years. The other Egyptian survey interviewed young adults
age 18 to 25 years in the cities and rural surroundings of Alexandria, Cairo, and
El-Minya. The Saudi survey interviewed young adults age 18 to 25 years in the

109
cities and rural surroundings of Jeddah, Riyadh, and Dammam-Khobar. The sur-
veys in Lebanon and the United States are nationally representative samples of
all adults. The Middle Eastern surveys were conducted in face-to-face interviews
whereas the US survey was conducted by telephone.
Because of budgetary constraints and different methodological limitations in
each setting, we used different sampling and interviewing strategies in the vari-
ous surveys. Thus, strict comparability across settings is not possible. Yet, our
goals are not strict comparability, but observation of the general distribution of
responses across and within settings.
In each survey, people rated countries on development and morality from
either 0 or 1 to 10, with 10 points indicating the highest morality or development.
The countries rated are listed in Table 4.1 and vary across the various surveys.
We did not give respondents definitions of development or morality; yet, within
some of the surveys, immorality was equated with cultural decadence (as described
later). Thus, respondents in those surveys may have been led to associate morality
with cultural (vs. political or economic) meanings.
The US survey and the adult survey in Egypt used the following question to
ascertain ratings on development:
We would like you to think about development in different countries around the
world today. We’ll be talking about countries as varied as England and Mongolia.
Think of a development scale that rates countries from zero to ten. The least devel-
oped places in the world are rated zero and the most developed places in the world
are rated ten. You can use both of those numbers for rating countries plus all of
the numbers in between. Using this development scale, where would you put
[Country]?
For respondents who said they did not know the developmental level of a country,
we asked: Even if you don’t know exactly, about where would you put [Country]?
The measurement approach used in Lebanon and in the youth surveys in Egypt
and Saudi Arabia used somewhat different wording and asked respondents to rate
countries from 1 to 10. We also did not specify the specific probes to be used in
these surveys.
The same approach was used for asking respondents to rate countries on
morality. The Egyptian adult survey substituted the word morality for the word
development. The US survey indicated that “countries where people have the lowest
morality are rated zero, and countries where people have the highest morality are
rated ten.” The Egyptian and Saudi youth surveys asked respondents to rate coun-
tries on a scale of morality (in Egypt, “a scale of morality and manners”), where 1
means that society is culturally decadent with a very low level of morality and 10
means that people in that society have a high level of morality. In the Lebanese
survey, the endpoints of the morality scale were “society is culturally decadent and
people have a very low level of morality” and “people in that society have a high
level of morality.”
For each survey, we calculated for each country rated its average morality and
development ratings. These averages were computed by combining the ratings
given with and without probes. We also compared these average ratings with the

110
Table 4.1 Mean Country Scores on Development and Morality (Plus United Nations Human Development Index Scores)
Countries rated Egyptian Saudi youth Egyptian Lebanese World United United Nations
youth 2005a 2005 a women 2008b Values Survey States 2007d Human
2008c Development
Index ×10, 2004
develop morality develop morality develop morality develop morality develop morality
Japan — — — — 7.91 6.18 — — 8.56 7.34 9.53

Egypt 5.26 6.49 7.10 7.14 6.72 8.01 — — 5.50 5.81 7.08
India — — — — 5.69 5.51 — — 5.02 5.71 6.19
United States 9.22 2.14 9.54 4.15 8.52 2.53 8.85 4.73 8.73 6.37 9.51
China 8.62 4.69 8.55 5.95 8.03 5.73 8.11 5.80 6.92 5.72 7.77
Saudi Arabia 6.40 7.25 7.10 8.14 — — 5.03 4.00 5.53 4.74 8.12
France 8.57 3.11 9.38 5.21 7.43 5.25 8.36 5.72 7.36 5.82 9.52
Brazil — — — — 5.86 4.89 — — 5.63 5.38 8.00
Pakistan — — — — 5.39 6.44 — — 3.96 4.35 5.51
Russia — — — — — — — — 5.95 5.18 8.02
Central African Republic — — — — 4.33 4.48 — — 3.56 4.13 3.84
Bulgaria — — — — — — — — 4.50 4.93 8.24
Nigeria — — — — 5.08 4.93 — — — — 4.65
Iran 5.40 6.64 6.84 7.10 — — 6.62 4.62 — — 7.59
Yemen 4.83 5.89 4.20 6.71 — — — — — — 5.08
Somalia 2.36 3.98 2.98 4.90 — — — — — — —
United Arab Emirates 7.24 6.45 8.08 7.51 — — — — — — 8.68
Iraq — — — — — — — — — — —
Kuwait — — — — — — — — — — —
Lebanon — — — — — — 4.94 5.99 — — 7.74
Syria — — — — — — 4.49 4.36 — — 7.16
n 878 902 951 952 871 872 2,989 2,972 489 484 —
111
Correlation between country .80 .94 .90 .74 .84

development and UN
scores
(all countries rated)
Correlation between respon- –.39 –.11 –.11 .36 .88
dent development and
morality
scores (all countries rated)
Correlation between respon- .30 .56 .48 .43 .90
morality
scores (excluding the
United States and France)
Correlation between respon- –.38 –.09 –.15 .24 .88
morality
scores (excluding China)
Correlation between respon- –.19 .85
morality
scores (excluding Japan)
Correlation between respon- .89
morality
scores (excluding Saudi
Arabia)
a
Unweighted.
b
Weighted.
c
WVS unweighted.
d
Weighted.
112
UN HDI scores for the rated countries. We also estimated Pearson’s correlation
coefficients between the average ratings of respondents and the scores of the UN
HDI for the same countries. In addition, we calculated correlations between the
average ratings of countries on development and the average ratings on morality.
We also calculated the correlations between each individual’s ratings of devel-
opment and the UN HDI. In other words, in each survey with n respondents, we
calculated n correlations between those respondents’ development scores and the
HDI scores. We also calculated similar correlations between an individual’s rat-
ings on development and on morality.
We asked two questions in the Egyptian adult survey about the relationship
between morality and development. One question asked respondents to address
this issue as follows: Now, please think about countries that are not developed and
countries that are developed. Compare between the living conditions in these coun-
tries. Could you tell me whether the following things are more common, in gen-
eral, in the countries that are not developed or more common in the countries that
are developed. The item of interest is low levels of morality of the people. The second
question asked: If a development program is implemented successfully in a coun-
try where the standard of living is low, most people live in rural areas, and access
to healthcare is poor, can you tell me whether over time, these things will increase
or decrease in that country after the successful implementation of the develop-
ment program. The item of interest here is the morality of people.
We focused our analyses of the predictors of people’s views of morality on
Lebanon because views of morality vary considerably within Lebanon, our
Lebanese survey included a large national sample, and a wide range of predic-
tor variables was available. Two dependent variables were respondent ratings of
American and Saudi morality. To control for the way people use rating scales, we
subtracted from the individual’s US morality rating, the respondent’s own average
score on morality for all the countries except France and the United States. We
excluded France (as well as the United States) from this base average because, as
a Western country, France may be rated in a similar way as the United States by
at least some respondents. Similarly, we adjusted the ratings of morality for Saudi
Arabia by subtracting from the Saudi rating the average rating for all countries
except Saudi Arabia. We also did a similar analysis using ratings of US and Saudi
country ratings on development, with similar adjustments for people’s use of the
rating scale.
The Lebanese survey included several questions about socioeconomic and
demographic factors, along with beliefs and attitudes toward foreign interven-
tion, religion, gender, sexuality, and attire. We used data from these questions as
predictors in multivariate analyses. The wording of these questions is provided
at this website: http://developmentalidealism.org/projects/lebanon/Lebanon
Questionnaire2008_English.pdf.
For each of the four dependent variables, we estimated the effects of the
respondent’s religiosity, beliefs about the role of religion in public life, identities
concerning tradition and family, views of US and Saudi involvement in Lebanon,
and values and beliefs concerning homosexuality, prostitution, alcohol, and styles

113
of dress. We coded these variables so that high scores reflected less strict religious
views, less identity with tradition and family, positive views of American or Saudi
involvement in Lebanon, and tolerance of homosexuality, alcohol, and immod-
est dress. To avoid multicollinearity, we estimated separate equations for each of
these predictor variables, controlling for gender, education, age, and income rat-
ing decile in all equations. We estimated the equations separately for Muslims
and Christians.
Results
Rating Countries on Development
We first consider the average country ratings on development displayed in Table
4.1, along with the UN HDI scores for the rated countries.1 In all five surveys,
respondents’ average development ratings differ dramatically across the rated
countries, indicating images of developmental hierarchies. Furthermore, these
average respondent ratings closely match the UN conceptualization of develop-
ment hierarchies. Pearson’s correlation coefficients between UN HDI scores and
the average respondent scores range from .74 to .94, indicating high aggregate
matching of respondent scores with the HDI (bottom panel of Table 4.1).
Table 4.2 shows the percentile distribution of the individual correlations
between individual country development ratings and HDI scores. These data show
Table 4.2 Bivariate Correlations between Individual Respondent’s Ratings

of Development and the United Nations Ratings of Development (All
Countries Rated by Respondent)
Percentiles Egyptian Saudi Egypt Lebanese United

youth youth women WVS 2008 States
2005 2005 2008a 2007a
10th .17 .57 .12 .25 .36

20th .37 .70 .34 .44 .49
30th .50 .75 .47 .50 .56
40th .58 .79 .58 .55 .62
50th .65 .81 .65 .62 .67
60th .70 .84 .71 .67 .71
70th .75 .86 .77 .72 .75
80th .79 .88 .84 .76 .79
90th .85 .91 .89 .84 .85
Mean correlation .57 .76 .57 .57 .63
No. of cases 767 856 754 2,801 476
note: These correlations are Pearson’s correlation coefficients between the country
development scores given by an individual and the United Nations Human
Development Index scores for the same countries.
a
Weighted.

114
that most individuals have mental maps of developmental hierarchies that match
closely those of the UN. Only 10% or fewer people have correlations less than .12.
Moreover, in all countries, the 30th percentile ranges from .47 to .75—meaning,
only 30% or fewer have correlations of .47 or less. The 50th percentile reaches
.62 or higher in all surveys; the means are slightly lower than the medians. These
findings are consistent with other research showing that many ordinary people in
diverse countries have conceptualizations of development, and those conceptual-
izations match those of the UN (Thornton et al., 2012).
Rating Countries on Morality

We now examine the average country morality ratings shown in Table 4.1. A strik-
ing finding is that respondents generally have high estimations of their own coun-
try’s morality. This is seen among adult Egyptians, Saudi youth, and Lebanese
adults, where the highest average ratings were for the home country (at 8.0, 8.1,
and 6.0, respectively). Americans rated themselves second (at 6.4) behind Japan
(at 7.3). Although Egyptian youth did not rate their country the highest, the average
morality score for Egypt was 6.5.
Another striking finding is the low ratings given by Egyptians and Saudis to
morality in the United States. In each of these three Middle Eastern surveys,
respondents gave the United States the lowest morality rating, with the aver-
age ratings being as low as 2.1 and 2.5 (on a scale of 0/1–10) in the Egyptian
surveys. The Lebanese were more positive about American morality, giving
the United States an average score of 4.7, the middle score in a group of seven
countries.
France fared better than the United States in each of the Middle Eastern sur-
veys. Yet, with the exception of the Lebanese, France was rated relatively low in
each Middle Eastern country, and especially among Egyptian youth, who gave
France the second lowest morality rating of 3.1. Among Saudi youth, France
received an average morality rating of 5.2 (third lowest in this survey), and among
Egyptian adults, the average rating was 5.2. The differences between the morality
ratings of France and the United States suggest that people in the Middle East see
the United States as less moral than Europe, although ratings of other European
countries are required to confirm this extrapolation.
Comparing Ratings of Morality and Development

The high ratings of France and the United States on development and their low rat-
ings on morality almost guarantee negative Pearson’s correlations between moral-
ity and development scores in the Egyptian and Saudi surveys. These correlations
are confirmed in the bottom panel of Table 4.1. For Egypt and Saudi Arabia, the
correlations range from –.39 to –.11, which supports Hypothesis 5, associated with
the world system approach, and contradicts the predictions of Hypothesis 4 and
modernization theory. The low average morality ratings for France and the United
States by Egyptians and Saudis are also consistent with Hypothesis 2, and the idea
that people in these two countries see a clash of moralities, with France and the
United States being especially immoral. Thus, the low morality ratings given to the

115
United States and France suggest that these countries may be seen as representing
a particular immoral version of modernity rather than immoral modernity per se.
To shed more light on this issue, we removed France and the United States
from the analysis and recalculated the correlations between average morality and
development ratings. These exclusions increased the correlations to between .30
and .56 for the Egypt and Saudi surveys, respectively (bottom panel of Table 4.1).
These correlations provide further evidence that the form of modernity existing in
France and the United States is seen to be immoral and not modernity in general.
In fact, these results support Hypothesis 4 and the view that modernization the-
ory, with its positive view of the morality–development relationship, is widespread
in Egypt and Saudi Arabia. It also suggests that the negative correlations with the
United States and France in the analysis are more a result of the clash of morali-
ties perspective (Hypothesis 2) than the world system perspective (Hypothesis 5).
We also calculated correlations between country ratings of morality and devel-
opment with either Japan or China excluded. We did this because Japan and China
are perceived in the Middle East as quite highly developed, but may be perceived
to have quite different cultures and morality systems than the United States and
France. These analyses (bottom panel of Table 4.1) revealed that the exclusion of
Japan or China has very little effect on the correlations. This corroborates the
observation that in all Middle Eastern surveys, average morality ratings for Japan
and China exceed those for France and the United States. Again, it is the French
and American versions of modernity that are seen as immoral rather than moder-
nity in general, suggesting a clash of moralities (Hypothesis 2) rather than adher-
ence to world system theory (Hypothesis 5).
The correlations in Lebanon contrast with those in Egypt and Saudi Arabia.
The fact that the morality ratings for France and the United States are higher in
Lebanon than in the other two countries suggests that the Lebanese do not harbor
the overall disdain for French and American morality that exists in Egypt and
Saudi Arabia. Also, even with France and the United States included, the correla-
tion is .36—much higher than in the other two countries. Moreover, when France
and the United States are removed, the correlation only rises moderately from
.36 to .43. Thus, these correlations suggest that, on average, people in Lebanon
have assimilated modernization theory’s ideas of a positive relationship between
development and morality.
In the United States, the data indicate overwhelmingly that, on average, peo-
ple perceive a strong, positive relationship between morality and development.
There is a consistent increase in morality ratings as development ratings increase.
In fact, the correlation between average development and morality scores in the
United States is .88, slightly higher than the correlation between average develop-
ment scores and UN HDI scores. These data are thus consistent with Hypothesis
4 and modernization theory’s view of a positive morality–development relation-
ship, and are inconsistent with Hypothesis 5 and the world system theory’s view
of a negative relationship.
We checked to determine whether Americans might have a particularly nega-
tive view of the morality of Saudis. We found that Americans rate Saudi Arabia
similarly on morality and development. Furthermore, removing Saudi Arabia

116
from the countries rated leads to little change in the aggregate correlation between
morality and development (bottom row of Table 4.1).
We now turn to the individual level and the correlations in Table 4.3 between
each individual’s morality and development ratings for the countries rated. The
individual correlations for the United States further support Hypothesis 4—that
most Americans adhere to modernization theory, perceiving a positive relation-
ship between morality and development. More than 70% of American have cor-
relations of .32 or higher, with half having correlations of .6 or higher (Table 4.3,
column A). Only about 10% have negative correlations between morality and
development, suggesting that adherence to world system theory is uncommon.
We noted earlier that, for a large percentage of Americans, the ratings they gave
countries on development correlated closely with the UN HDI scores for the same
countries (Table 4.2). This reflects the fact that the countries are being rated on
the same dimension—development—by different raters (the UN and American
respondents). The individual-level correlations of ratings of development and
morality are only slightly lower than the individual-level correlations between
perceived ratings of development and HDI. For example, the median correlation
between individual morality and development ratings was .60, only somewhat
less than the .67 median correlation for the individual ratings of development
with HDI. Moreover, the 80th percentile for the development–morality correla-
tions is .82, a bit higher than the 80th percentile for the correlations between
individual development ratings and HDI (.79).
The Middle East presents a very different picture. Many individuals in these
countries accept the views of modernization theory (Hypothesis 4) whereas many
others accept the views of world system theory (Hypothesis 5). With all the coun-
tries rated included (Table 4.3, column A), respondents in the Saudi and Egyptian
adult surveys are split almost evenly between having positive and negative cor-
relations between country ratings on morality and development (median values
of –.09 and +.04, respectively). Also, substantial percentages in these two surveys
have either large negative correlations or large positive correlations. For example,
20% of Egyptian adults have correlations at –.3 or less whereas 20% have correla-
tions of .4 or more. The Saudi youth are even more split, with 20% having correla-
tions of –.50 or less and 20% having correlations of .51 or more. The individual
morality–development correlations are more negative in the Egyptian youth sur-
vey. The median correlation is –.22, and 70% of the Egyptian youths have correla-
tions at –.02 or less. Also, few Egyptian youth had high positive correlations—the
90th percentile being only .34. In contrast, the distribution of correlations in
Lebanon is tilted somewhat in the positive direction, with the median correlation
being .31. In Lebanon there are also relatively few large negative correlations and
a substantial number of large positive correlations, consistent with the idea that
the viewpoints of modernization theory are more widespread than the viewpoints
of world system theory.
Consistent with the aggregate-level results, the distribution of individual corre-
lations becomes much more positive with the exclusion of France and the United
States from the calculations (Table 4.3, column B). In this case, in all four Middle

117
Table 4.3 Bivariate Correlations between Individual Respondent’s Ratings of Development and Morality
Percentile Egyptian youth Saudi youth 2005 Egyptian women 2008a Lebanese WVS 2008 USA 2007a
2005
A B C A B C A B C D A B C A B C E
th
10 –.61 –.35 –.61 –.63 –.41 –.65 –.46 –.29 –.50 –.54 –.48 –.45 –.55 –.07 –.05 –.09 –.10
20th –.49 –.17 –.50 –.50 –.16 –.50 –.30 –.06 –.32 –.38 –.25 –.17 –.33 .14 .23 .13 .15
30th –.41 –.05 –.39 –.34 .05 –.33 –.19 .07 –.20 –.24 –.04 .08 –.13 .32 .42 .35 .31
40th –.30 .07 –.31 –.23 .21 –.22 –.05 .20 –.09 –.11 .15 .27 .08 .46 .50 .49 .50
50th –.22 .17 –.21 –.09 .35 –.09 .04 .32 .03 –.01 .31 .41 .28 .60 .60 .61 .60
60th –.12 .27 –.12 .07 .41 .07 .15 .40 .13 .09 .48 .55 .45 .67 .66 .68 .68
70th –.02 .36 –.02 .27 .61 .26 .26 .51 .24 .18 .62 .66 .60 .74 .74 .76 .75
80th .13 .52 .13 .51 .72 .57 .40 .63 .37 .36 .74 .78 .73 .82 .80 .84 .82
90th .34 .71 .37 .80 .87 .82 .54 .77 .58 .56 .86 .89 .87 .89 .88 .90 .91
Mean corr. –.17 .16 –.17 –.01 .29 .00 .05 .26 .03 .00 .24 .31 .20 .48 .51 .49 .49
No. of cases 725 730 733 680 674 705 725 727 726 730 2476 2497 2516 459 455 459 460
note: These estimates are Pearson’s correlation coefficients between the country scores given by an individual for development and for morality. A, all countries rated; B,
excluding the United States and France; C, excluding China; D, excluding Japan; E, excluding Saudi Arabia.
a
Weighted.
118
Eastern surveys, about 70% of the respondents have individual correlations that
are positive, and the median values range from .17 to .41. About 30% have high
positive correlations and only a few have large negative correlations. As with the
aggregate analysis, the removal of Japan and China has little effect on the distribu-
tion of individual correlations, providing additional support for a clash of morali-
ties rather than a world system perspective.
Predictors of Lebanese Ratings of American and Saudi Morality

We next turn to the predictors of the ratings of US and Saudi morality by
Lebanese. Christians, on average, rated all the predominantly Muslim countries—
Syria, Saudi Arabia, and Iran—lower on morality than Muslims (not shown in
the tables). Conversely, on average, Christians rated all the predominantly non-
Muslim countries—China, France, and the United States—higher on morality
than Muslims. This is consistent with Hypothesis 2, about the existence of differ-
ent morality systems among different social groups.
Table 4.4 displays the regression coefficients for a range of predictors of
(adjusted) ratings of American and Saudi morality, controlling in all cases for sex,
age, education, and a scale for subjective income. These data reveal much evi-
dence consistent with Hypothesis 3. Looking first at Muslims, we find an over-
whelmingly consistent pattern. Each indicator of low religious observance and
strictness predicts higher ratings of US morality and lower ratings of Saudi moral-
ity (although one coefficient is small and not statistically significant). All the items
measuring tolerance of frequently proscribed behaviors have significant positive
coefficients on US morality ratings and, with one exception, have negative signifi-
cant coefficients on Saudi morality ratings. People who believe that modest dress
is not important and that women should be able to dress any way they want, on
average, rate US morality higher and Saudi morality lower than people concerned
about women’s dress. People who say that traditionality is not like them (implying
that modernity is like them) and who reject more emphasis on family life have
significantly more positive views of American morality and less positive views of
Saudi morality. In addition, positive views about the justifiability of homosexu-
ality, prostitution, and alcohol are all related positively to higher ratings of US
morality and lower ratings of Saudi morality.
A similar pattern emerges from our question asking which of three things is
the most immoral: stealing, violence, or premarital sex. We coded answers into
two dummy variables—one for violence being the most important and one for
premarital sex being the most important—with our excluded category being steal-
ing. The choice of premarital sex being the most immoral is, as expected, related
negatively to high US morality ratings and is related positively to high ratings of
Saudi morality.
The last set of numbers reveals that views of foreign involvement in Lebanon
have substantial effects on how Muslim Lebanese rate Saudi and American moral-
ity. People who do not see Saudi involvement in Lebanon as a problem rate Saudi
morality higher than others. Similarly, people who are sanguine about American
involvement rate the United States higher on morality.

119
Table 4.4 Predicting Adjusted Morality Ratings for the United States and Saudi Arabia: 2008 Lebanese Muslim and Christian Respondents
(t Values in Parentheses)
Independent variables US morality Saudi Arabia morality
Muslim Christian Muslim Christian
Religion and religiosity

Composite: Less strict religious views .58 (3.48) .62 (3.22) –.03 (.20) –.84 (5.51)
Prayer frequency low (5) .30 (6.96) –.05 (.60) –.23 (5.24) –.14 (1.96)
God not important (10) .15 (6.77) –.02 (.69) –.08 (3.74) .02 (.69)
Not make laws according to religion (5) .27 (5.41) .00 (.02) –.12 (2.43) –.24 (4.01)
Mosque/church attendance low (7) .08 (2.60) –.15 (2.85) –.08 (2.75) –.08 (1.90)
Women’s dress
Modest dress for women not important (5) .61 (12.34) .44 (6.23) –.35 (6.85) –.40 (7.06)
Women dress any way they want (4) .31 (4.22) .50 (4.54) –.43 (5.91) –.27 (3.03)
Family life and tradition
Traditional not like me (6) .17 (3.61) –.06 (.98) –.22 (4.57) –.01 (.19)
More emphasis family life bad in future (3) .36 (1.78) .45 (1.65) –.24 (1.16) –.25 (1.15)
Values
Homosexuality justified (10) .14 (5.12) .04 (1.29) –.13 (4.78) –.04 (1.53)
Prostitution justified (10) .16 (5.43) .00 (.12) –.16 (5.18) –.06 (2.16)
Alcohol justified (10) .18 (7.47) .04 (2.30) –.14 (5.85) –.03 (1.35)
If stealing most immoral *REF* *REF* *REF* *REF*
If violence most immoral .33 (2.06) .48 (2.46) .07 (.43) .01 (.09)
If premarital sex most immoral –1.07 (6.41) –.81 (3.03) .47 (2.76) .67 (3.07)
Beliefs about foreign involvement
United States is not causing political .53 (8.31) .44 (6.15) — —
problems/violence in Lebanon (5)
Saudi Arabia is not causing political — — .68 (17.79) .40 (7.75)
n 1,746 996 1,797 1,002
note: Each independent variable coefficient and t value shown represents a single model that also includes sex, age, education, and subjective income. The latter
effects are not shown but are available on request.
120
The predictors of Saudi morality among Christian Lebanese are similar to the
predictors of Saudi morality among Muslim Lebanese. With one exception, the
coefficients for Christians are in the same direction as Muslims, although several
are not statistically significant. This suggests that the forces influencing views of
Saudi morality are similar among Christians and Muslims.
The ratings of American morality are less consistent among Christian
Lebanese. Less strict religious views, permissive views about women’s dress, more
emphasis on family life being bad, and being sanguine about American involve-
ment in Lebanon all have the expected positive influence on ratings of American
morality among Christian Lebanese (Table 4.4). In addition, Christians saying
that premarital sex is the most immoral have low evaluations of American moral-
ity. For other indicators, however, the coefficients are very small and statistically
insignificant, and for one indicator—church attendance—the coefficient is in
the opposite direction, suggesting that low church attendance decreases positive
views of American morality.
Predictors of Lebanese Ratings of American and

Saudi Development
The data discussed in the previous section are broadly consistent with Hypothesis
3 that many indicators of personal religiosity and morality standards influence
people’s views of the morality of various countries. This raises the possibility that
people’s personal religiosity and personal views of morality also influence percep-
tions of development.
This led us to analyze factors predicting Lebanese ratings of American and
Saudi development in ways similar to our analyses of predictors of morality. Our
first observation is that many of the same factors that predict Lebanese ratings
of country morality predict Lebanese ratings of country development, suggesting
that morality is intertwined in people’s minds with conceptions of modernity. Just
as Muslims rated the morality of predominantly Muslim countries—Syria, Saudi
Arabia, and Iran—higher than Christians, they also rated the same countries
higher on development than Christians. And, just as Muslims rated the morality
of predominantly non-Muslim countries—China, France, and the United States—
lower on morality than Christians, they also rated the same countries lower on
development.
More extensive data concerning the idea that conceptions of morality enter into
evaluations of development are provided in Table 4.5, where we predict adjusted
ratings of American and Saudi development using the same predictors used in
the morality analysis. These coefficients indicate that at least some of the morality
predictors also predict ratings of Saudi and American development.
Among Muslims, less strict religious views, low prayer frequency, permis-
sive attitudes about women’s dress, negative views about more family emphasis,
believing that alcohol use is justified, and being sanguine about US involvement
in Lebanon all have positive and statistically significant coefficients on ratings of
American development. Thus, these items indicating less Muslim orthodoxy all
increase positive views of American development.

121
Table 4.5 Predicting Adjusted Development Ratings for the United States and Saudi Arabia: 2008 Lebanese Muslim and Christian
Respondents (t Values in Parentheses)
Independent variables US development Saudi Arabia development
Muslim Christian Muslim Christian
Religion and religiosity

Composite: Less strict religious views .31 (3.02) .50 (4.53) –.34 (2.61) –.29 (1.97)
Prayer frequency low (5) .07 (2.62) .13 (2.70) –.07 (2.05) –.06 (.89)
God not important (10) .00 (.22) .04 (1.62) –.02 (1.35) –.02 (.84)
Not make laws according to religion (5) –.03 (1.12) .05 (1.13) –.27 (7.23) –.26 (4.65)
Mosque/church attendance low (7) .00 (.21) .03 (1.09) –.02 (.95) –.06 (1.56)
Women’s dress
Modest dress for women not important (5) .08 (2.41) .03 (.67) –.07 (1.67) –.20 (3.64)
Women dress any way they want (4) .17 (3.93) .08 (1.30) –.26 (4.66) –.05 (.58)
Family life and tradition
Traditional not like me (6) .03 (1.10) .05 (1.57) –.09 (2.49) .09 (2.05)
More emphasis family life bad in future (3) .28 (2.20) .11 (.70) –.35 (2.13) .04 (.19)
Values
Homosexuality justified (10) –.01 (.59) .00 (.31) .04 (2.03) .04 (1.77)
Prostitution justified (10) –.03 (1.83) .01 (.39) .03 (1.49) .00 (.08)
Alcohol justified (10) .06 (3.91) .02 (1.42) .05 (2.91) .11 (4.61)
If stealing most immoral *REF* *REF* *REF* *REF*
If violence most immoral .02 (.23) –.30 (2.68) .14 (1.10) .02 (.14)
If premarital sex most immoral –.13 (1.28) –.49 (3.14) .57 (4.33) .24 (1.16)
Beliefs about foreign involvement
United States is not causing political .09 (2.37) .09 (2.20) — —
Saudi Arabia is not causing political — — .09 (2.95) .16 (3.24)
n 1,804 1,021 1,803 1,022
note: Each independent variable coefficient and t value shown represents a single model that also includes sex, age, education, and subjective income.
122
At the same time, among Muslims, having less restrictive religious views, low
prayer attendance, not wanting religious laws, permissive attitudes about wom-
en’s dress, thinking of oneself as not traditional, and thinking that more emphasis
on family things is bad all have negative statistically significant coefficients on rat-
ings of Saudi development. In addition, thinking that premarital sex is the most
immoral of three factors and being sanguine about Saudi involvement in Lebanon
are related positively to ratings of Saudi development. Some of the other coef-
ficients for Muslims in Table 4.5 are not statistically significant, although in the
expected direction. In addition, some factors, including views of the justification
of homosexuality, prostitution, and alcohol use, are in the opposite direction and
are statistically significant.
Interestingly, for Lebanese Christians, all the coefficients predicting ratings of
US development are in the positive direction, indicating morality is relevant to the
way people view development. Most of these coefficients are not statistically sig-
nificant, but some are. In addition, several morality indicators have opposite rela-
tionships with ratings of Saudi development, although only a few are statistically
significant. Like the ratings of Saudi development by Muslim Lebanese, there is a
tendency among Christian Lebanese for values about homosexuality and alcohol
use to be in the opposite direction.
Although the results are mixed, they suggest that, for both Muslims and
Christians, there is a tendency for concerns about morality to spill over into rat-
ings of development. When the two religious groups are combined into a single
analysis, the statistical significance of several of the coefficients increases (not
shown in tables). These results thus suggest that some matters of morality are
affecting evaluations of modernity among at least some people.
Direct Questions Relating Morality and Development

We now examine two direct questions asking adult Egyptians about the relation-
ship between development and morality. Concerning the question about where
low levels of morality are more common, we find a split; 60.3% said that low
morality is more common in developed places, 30.4% said it is more common in
not developed places, and 9.3% volunteered that it was about the same in the two
places or that they did not know (not shown in tables). Thus, there is a 2:1 prepon-
derance of believing that low morality is more common in developed places than
in not developed places.
However, most Egyptians are optimistic about the consequences of a develop-
ment program for morality in a country where the standard of living is low, most
people live in rural areas, and access to healthcare is poor. We find that 79.1%
believe a development program will increase morality, 14.5% say it will decrease
morality, and 6.4% volunteer that it will not affect morality. Thus, in more than a
5:1 ratio, Egyptians view development leading to more, not less, morality.
At first inspection, the results from these two questions appear contradictory
in that the first question suggests a negative morality–development relationship
and the second question, a positive relationship. This apparent contradiction
likely results from the different contexts of the two questions. The first question

123
made no effort to separate the West from the issue of development, and many
respondents were probably thinking about Western places such as France and the
United States when they answered the question, making the comparison implic-
itly between not developed non-Western places and developed Western places.
Given the low opinions of many Egyptians concerning morality in France and the
United States, it is not surprising that twice as many perceived low levels of moral-
ity in developed places as in not developed places.
However, the second question asked explicitly about a development program in
a country where the standard of living is low, most people live in rural areas, and
access to healthcare is poor. Although we do not know where Egyptian respondents
supposed this country to be, we are confident they were not thinking of Western
places. Thus, respondents were comparing a low-income, rural non-Western place
before development with the same place after development. The respondents may
also have been thinking of a non-Western development program rather than a
Western one. With the element of the West removed from the question, Egyptians
said overwhelmingly that a development program would increase morality.
These results corroborate those from the Egyptian country ratings. For many
Egyptians, low morality is associated with Western countries such as France and
the United States, and when it is Western development that people have in mind,
many associate low morality with development. The clash of moralities framework
influences many Egyptians in this context to state that morality and development
are related negatively. However, when development and the West are separated,
thereby removing the clash of moralities perspective, most Egyptians associate
development and morality together positively—a result that is consistent with
Hypothesis 4 and the modernization framework.
Conclusions
This chapter was motivated by the idea that individuals around the world have
perceptions of societal development and morality. In our theory section, we dis-
cussed three social–scientific perspectives concerning the relationships between
development and morality: modernization, world system, and an adaptation of
clash of civilizations that focuses on a clash of moralities. We hypothesized that
many of these ideas—in both crude and sophisticated forms—have been dissemi-
nated widely around the world, where they influence individuals and their per-
ceptions of the world. Our empirical work investigated which conceptualization
of the development–morality linkage is most common in Egypt, Lebanon, Saudi
Arabia, and the United States. Our findings provide some support for each of the
conceptualizations.
Our first conclusion is that the ideas of developmental and morality hierarchies
are widespread in the four countries studied. People are able to rate countries on
development, and they rate them similarly in the four countries and similarly
to the ratings of the UN HDI. People also are able to rate countries on morality.
Thus, both development and morality are salient concepts that people in each
country can apply to various countries.

124
Our research found evidence supporting each of the five hypotheses con-
cerning development and morality that we discussed earlier, although the
strength of the evidence is greater for some hypotheses than others. The data
are very consistent with Hypothesis 1: people’s views of development, morality, and
their linkages vary by social context. Differences are especially marked between
the United States and the Middle East, but also appear within the countries of
the Middle East, underscoring ideational heterogeneity within the region. We
summarize these context differences later, as we review the evidence for each
of the other four hypotheses.
As noted in “Data and Methods,” there were differences in sample universes and
designs across countries that may influence comparisons. In addition, we noted
modest differences in wording of the questions, but we expect these wording effects
to be small. For these reasons, we interpret observed differences across contexts
tentatively and advocate for additional research to confirm them, including survey
experiments to explore potential response effects to variants in question wording.
We found considerable evidence for Hypothesis 2: the observed relationship
between development and morality depends on people’s views of morality. One piece of
this evidence is that, on average, people rate their own country highly on morality.
This undoubtedly reflects universal ethnocentrism, in that people tend to judge
their own country’s morality in terms of their own moral standards.
Additional evidence consistent with Hypothesis 2 and the clash of moralities
perspective is that Americans and Middle Easterners have very different views
concerning French and American morality. Although Americans rate both French
and American morality high, Egyptians and Saudis rate the two countries—
especially the United States—low. These data corroborate a clash of moralities
perspective.
We do not know whether this clash of moralities with France and the United
States represents a general clash of moralities with the West as a cultural region
or only with France and the United States. It may be that France and the United
States are seen as particularly immoral and other Western countries are seen as
moral. Our expectation, however, is that there is at least a partial generalization of
immorality to the West, with individual countries differing to some degree on the
perceived extent of their immorality. We also do not know whether this view of low
American and French morality exists in other Middle Eastern countries, although
we expect that it does, at least to some extent. More research is needed to answer
these questions.
Although the idea of the United States and France being immoral is wide-
spread in the three Middle Eastern countries studied here, there is also consider-
able heterogeneity. In each Middle Eastern country, there are a few people who
rate the United States and France relatively highly on morality, with this being
more common in Lebanon than in Egypt and Saudi Arabia, and more common
among Christians than among Muslims in Lebanon. For these people, there is
much less a clash of moralities than for many others.
We found considerable evidence for Hypothesis 3: the views of people concerning
the morality of a country are related directly to their personal views of morality. Our
multivariate analyses of the predictors of Lebanese ratings of US and Saudi moral-
ity confirm this. Lebanese who are less religious and have less strict religious

125
interpretations are more likely than others to rate the United States high on moral-
ity. Similarly, people who indicate acceptance for women dressing the way they
want, homosexuality, prostitution, and alcohol rate the United States higher, as do
people who do not identify with tradition or an emphasis on families. Similarly,
Lebanese who view negatively the influence of the United States in their country’s
affairs also view American morality less positively. These findings suggest that ele-
ments of religious position, tolerance of proscribed behaviors, views of Western
influence, and views on families and tradition versus modernity are all related to
the ways Lebanese people think about American morality. Interestingly, these vari-
ables predict ratings of Saudi morality in the opposite direction, again indicating
an overall clash of moralities.
These results do not mean that Lebanese evaluate morality only in terms of the
factors considered here. Many other factors could affect evaluations of a country’s
morality, including materialism, extreme individualism, consumerism, and drug
use. Further research concerning other dimensions affecting morality evaluations
would be valuable.
There is also considerable evidence for Hypothesis 4: there is a positive link
between people’s views about development and morality. One of the remarkable
aspects of the American ratings of countries on morality is that they follow closely
the ratings of countries on development. This accords especially strongly with the
modernization framework in that modernization and morality overlap closely in
the minds of Americans. This interpretation holds not only at the aggregate level,
but also for most individuals. However, there may be a few individual Americans
who reject this tight association between development and morality.
The evidence for Hypothesis 4, however, is mixed for the three Middle Eastern
countries. For Lebanon, at the aggregate level, the correlation between country
morality and development is positive, but of much smaller magnitude than for the
United States. In addition, only about 70% of individual Lebanese have positive
correlations between their country development and morality ratings compared
with almost 90% in the United States.
In both Egypt and Saudi Arabia, the aggregate-level correlations between national
morality and development scores are negative, which is in the opposite direction
predicted by Hypothesis 4. However, these results likely underestimate the sup-
port for Hypothesis 4 and the modernization perspective, because they are affected
strongly by the clash of moralities existing in the Middle East. When we control
for the clash of moralities perspective by removing France and the United States
from the analyses, about two thirds or more of Egyptian and Saudi respondents
linked development and morality positively. Our interpretation is that there is not
an overall denigration of the morality of developed places, but a denigration of the
morality of France and the United States. When this clash of moralities is removed,
there is an overall perception of a positive association of development and morality,
which is consistent with a modernization perspective. Nevertheless, the percentage
of Middle Easterners linking development and morality positively is considerably
smaller than in the United States, suggesting less commitment to the moderniza-
tion framework in these Middle Eastern countries than in the United States.
Further evidence for Hypothesis 4 and the influence of modernization theory
is provided by our question in Egypt asking about the effects of development in

126
a low- income agricultural place— presumably non- Western. Overwhelmingly,

Egyptians saw development bringing higher morality in such a place. The ques-
tion did not specify whether such a development program was Western-funded or
locally supported, and we do not know what the respondents were thinking about
when they answered the question. Thus, it is unclear how respondents would have
answered this question had the type of development program been specified as
either Western-funded or locally supported. This issue seems highly relevant and
worthy of future research.
The evidence for Hypothesis 5—there is a negative link between people’s views
about development and morality—is weaker and more mixed than for our other
hypotheses. In the United States, there is virtually no evidence for a negative
link between people’s views about development and morality. In the aggregate,
the correlation is strongly positive, and very few individuals display negative
correlations.
The negative aggregate and individual correlations between morality and devel-
opment ratings in Egypt and Saudi Arabia may reflect support for Hypothesis 5
and the world system perspective. However, support for this interpretation is weak-
ened by the observation that there is a strong clash of moralities in Egypt and Saudi
Arabia, which causes people there to rate France and the United States relatively
low on morality, while rating them high on development. As we mentioned earlier,
when we control this factor, the negative aggregate correlations disappear and even
become moderately positive, and the share of individuals with negative correlations
declines substantially. Nevertheless, even with this control, as many as one fourth to
one third link development and morality negatively. This suggests that some people
have views consistent with world system theory in these Middle Eastern countries.
A similar interpretation explains the fact that most Egyptians said that low
morality is more common in developed than not developed places. We believe
that, instead of this reflecting support for Hypothesis 5 and the world system per-
spective, such responses likely resulted from many Egyptians thinking of Western
developed places—rather than developed places in general—as having low moral-
ity, reflecting a clash of moralities perspective. Experimental research concerning
the influence of question wording on these results would be beneficial in confirm-
ing or not confirming this interpretation.
These findings lead to the insight that at least some people in these Middle
Eastern settings have alternative models of modernity. Egyptians appear to
recognize a Western and a non-Western model of modernity, with very differ-
ent moralities. Further evidence of this perspective is provided by the fact that
Lebanese Muslims evaluate development levels of various countries differently
than Lebanese Christians, with religious composition of the rated countries being
relevant to this differential. Also, several indicators of religiosity and attitudes
toward factors sometimes seen as indicators of morality themselves affect how
Lebanese evaluate American and Saudi development. This suggests that at least
some people in Lebanon have their views of US and Saudi development affected
by their stances on moral issues.
These findings are related to what Eisenstadt (2000) has called multiple moder-
nities, when different people or cultural groups define modernity differently, with

127
those differences being shaped by varying conceptions of morality. That is, some
people build their own sense of the moral into their definition of modernity,
resulting in alternative (e.g., moral, amoral, and immoral) versions of modernity.
This viewpoint of multiple and alternative modernities corroborates other find-
ings in the literature. For example, a recent ethnography argues that Nepalese
are committed predominantly to modernity, but only to “suitable” modernity
(Liechty 2003). For them, unsuitable modernity includes such immoral attributes
as materialism, consumerism, sexual excesses, and illicit drugs, which contradict
their notions of a moral modernity. Also, an ethnography of a Shi’i community
in Lebanon (Deeb, 2006) indicates that the popular conception of modernity in
this community is one that has both material and spiritual components, with
Western modernity being materially advanced but spiritually backward. For them,
the backward immorality of the West includes “atheism, violence, capitalism,
consumerism, materialism, sexual promiscuity, the objectification of women, an
emphasis on the individual to the detriment of social relations, and the collapse
of the family” (Deeb, 2006, p. 23–24). Such people sometimes use the label of
“westernized or ‘Americanized’ Lebanese … to explain moral laxity” (Deeb, 2006,
p. 24). As an alternative, they endorse a “pious modernity” that includes material
and spiritual progress.
Also, focus group discussions with Egyptians (Yount, Thornton, Mehanna, &
Patel 2016, this volume) showed that participants often associated morality with
religion, especially religious observance, as well as with the avoidance of non-
recognized forms of marriage, the exercise of sexual discretion, and modesty in
dress—especially for women. Many asserted that it was easier for a traditional
family or society to have good moral values and more difficult for a modern family
or society to retain such values. Nevertheless, participants believed that it was pos-
sible to achieve a culturally authentic, moral modernity, which they distinguished
from and preferred to the so-called moral chaos of Western modernity.
Distinguishing between Western modernity and morality, and suitable moder-
nity and morality is likely to be common in predominantly Muslim countries and
probably elsewhere as well. Indeed, the expressions of Muslim intellectual leaders
and activists in settings such as Algeria, Egypt, Iran, and Jordan connect sexual
practices and women’s dress to conceptions of morality (al-Banna 1978; Mitchell,
1969; Motahhari, 1969; Slyomovics, 1995; Taraki, 1996). Although both Western
and pious modernities endorse material prosperity, the two versions differ in their
treatment of piety, religious practice, sexual purity, modesty of dress, and family
ties. Also, those who espouse pious modernities may view most or all realms of
human behavior as morally constitutive, and therefore requiring limits. In this
broader sense, Western modernity is viewed as “excessive” (e.g., Yount et al., 2016).
The idea of alternative modernities further complicates our understanding of
the ways in which people view development and morality. It suggests that peo-
ple may not only see development and morality to be related—either positively
or negatively—but also that they use their views of morality to define what they
mean by development. That is, some people may think of development as a suit-
able modernity that excludes the factors they believe to be immoral. They may fol-
low standard definitions of modernity that include such things as wealth, health,

128
and education, but add to these criteria elements that can be described as moral.
Thus, the definition of modernity in the minds of individuals may have consider-
able influence on how they see the relationship between morality and develop-
ment. Further research about the factors entering into views of modernity would
be helpful.
We close with the observation that this research investigating views about moral-
ity, development, and their relationship is very new. It applies to only a few countries
with limited regional coverage. We need more research both in different settings in
these regions and in other world regions. In addition, although we have provided
plausible interpretations of people’s views of the relationship between development
and morality, we need more substantive and methodological research, including in-
depth qualitative work, concerning people’s meanings of morality. What do people
in various countries mean when they say a country is moral or immoral or that it
rates a 2 or 8 on an 11-point scale? And, what do people have in mind when they say
a development program will increase or decrease morality? Also important are addi-
tional analyses examining the factors influencing people’s views of morality. The
importance of our findings recommends such studies in many settings.
Acknowledgment
An earlier version of this chapter was presented at the annual meeting of the
American Sociological Association, Atlanta, Georgia, August 14–17, 2010; at the
Symposium on Globalization of Modernization Theory: Clashes of Modernities
and Moralities, June 8–10, 2010, Ann Arbor, Michigan; at the Workshop on
Theoretical and Methodological Issues in the Study of Values in Islamic Countries,
May 16–18, 2010, Cairo, Egypt; and at a seminar in the Department of Sociology,
University of Michigan. We appreciate the comments and suggestions made by
participants of the various meetings.
Note
1. The HDI scores are created by the UN as an index combining adult literacy
and school enrollment in primary, secondary, and tertiary school; life expectancy at
birth; and per capita gross domestic product. Although the UN scores range from 0
to 1, we multiplied them by 10 to make them more comparable with our survey scales.
References
Abu-Lughod, L. (1998). Feminist longings and postcolonial conditions.” In l. Abu-
Lughod (Ed.). Remaking women: Feminism and modernity in the Middle East (pp.
3–32). Princeton, NJ: Princeton University Press.
Ahearn, L. M. (2001). Invitations to love: Literacy, love letters, and social change in Nepal.
Ann Arbor: University of Michigan Press.

129
al-Banna, H. (1978). Five tracts of Hasan al-Banna (1906–1949) (C. Wendell, Trans.).
Ali, M. M., Cleland, J., & Shah, I. H. (2003). Trends in reproductive behavior among
young single women in Colombia and Peru: 1985–1999. Demography, 40(4),
659–673.
Apter, D. E. (1965). The politics of modernization. Chicago, IL: University of Chicago.
Berkhofer, R. F. (1978). The white man’s Indian: Images of the American Indian from
Columbus to the present. New York: Knopf.
Bhagwati, J. (1986). Ideology and North–South relations. World Development, 14(6),
767–774.
Binstock, G., & Thornton, A. (2007). Knowledge and use of developmental thinking
about societies and families among teenagers in Argentina. Demografía, 50(5),
75–104.
Bornschier, V. (1980). Multinational corporations and economic growth: A cross-
national test of the decapitalization thesis. Journal of Development Economics, 7(2),
191–210.
Bornschier, V., Chase-Dunn, C., & Rubinson, R. (1978). Cross-national evidence of the
effects of foreign investment and aid on economic growth and inequality: A sur-
vey of findings and reanalysis. American Journal of Sociology, 84, 651–683.
Böröcz, J. (2000). The fox and the raven: The European Union and Hungary rene-
gotiate the margins of “Europe.” Comparative Studies in Society and History, 42,
847–875.
Caldwell, J. C., Reddy, P. H., & Caldwell, P. (1988). The causes of demographic
change: Experimental research in South India. Madison: University of
Wisconsin Press.
Carniero, R. L. (1973). Classical evolution. In R. Naroll & F. Naroll (Eds.). Main cur-
rents in cultural anthropology (pp. 57–121). Englewood Cliffs, NJ: Prentice-Hall.
Chase-Dunn, C., & Rubinson, R. (1977). Toward a structural perspective on the world-
system. Politics and Society, 7, 453–476.
Coltabiano, M., & Castiglioni, M. (2008). Changing family formation in
Nepal: Marriage, cohabitation and first sexual intercourse. International Family
Planning Perspectives, 34(1), 30–39.
Dahl, G., & Rabo, A. (Eds.). (1992). Kam-ap or take-off: Local notions of development.
Stockholm: Stockholm Studies in Social Anthropology.
Deeb, L. (2006). An enchanted modern: Gender and public piety in Shi’i Lebanon.
de Janvry, A., & Garramon, C. (1977). Laws of motion of capital in the center–periphery
structure. Review of Radical Political Economy, 9, 29–38.
Duvall, R., Jackson, S., Russett, B. M., Snidal, D., & Sylvan, D. (1981). A formal
model of “dependencia” theory: Structure and measurement. In R. Merritt & B.
M. Russett (Eds.). From national development to global community (pp. 312–350).
London: Allen and Unwin.
Eisenstadt, S. N. (2000). The reconstruction of religious arenas in the framework of
“multiple modernities.” Millennium: Journal of International Studies, 29, 591–611.
Emmanuel, A. (1972). Unequal exchange: A study in the imperialism of trade.
New York: Monthly Review Press.

130
Esmer, Y. (2002). Is there an Islamic civilization? Comparative Sociology, 1(3–4),

265–298.
Ferguson, J. (1999). Expectations of modernity: Myths and meanings of urban life on the
Zambian copperbelt. Berkeley: University of California Press.
Friedman, B. M. (2005). The moral consequences of economic growth.
New York: Vintage Books.
Furedi, F. (1977). The moral condemnation of the South. In C. Thomas & P. Wilkin
(Eds.). Globalization and the South (pp. 76–89). New York: St. Martin’s Press.
Harris, M. (1968). The rise of anthropological theory. New York: Thomas Y. Crowell.
Huntington, S. P. (1968). Political order in changing societies. New Haven, CT: Yale
University.
———. (1996). The clash of civilizations and the remaking of world order.
New York: Simon & Schuster.
Inglehart, R., & Welzel, C. (2005). Modernization, cultural change, and democracy: The
human development sequence. New York: Cambridge University Press.
International Monetary Fund. (2009). World economic outlook: Sustaining the recovery.
Washington, DC: International Monetary Fund Author.
Kindleberger, C. P. (1965). Economic development. New York: McGraw-Hill.
Krücken, G., & Drori, G. S. (Eds.). (2009). World society: The writings of John W. Meyer.
Oxford: Oxford University Press.
Latham, M. E. (2000). Modernization as ideology. Chapel Hill: University of North
Carolina Press.
Lavelle, K. C. (2001). Ideas within a context of power: The African group in an evolv-
ing UNCTAD. Journal of Modern African Studies, 39(1), 25–50.
Lesthaeghe, R. J., & Neidert, L. (2006). The second demographic transition in the
United States: Exception or textbook example? Population and Development
Review, 32(4), 669–698.
Levy, H. S. (1966). Chinese footbinding: The history of a curious erotic custom.
New York: Walton Rawls.
Liechty, M. (2003). Suitably modern: Making middle-class culture in a new consumer
society. Princeton, NJ: Princeton University Press.
Løvbrǽk, A. (1990). International reform and the like-minded countries in the North–
South dialogue 1975–1985. In C. Pratt (Ed.). Middle power internationalism: The
North–South dimension (pp. 25–68). McGill: Queen’s University Press.
Mandelbaum, M. (1971). History, man, and reason: A study in nineteenth-century thought.
Baltimore, MD: Johns Hopkins University Press.
Mauldin, W. P., Choucri, N., Notestein, F. W., & Teitelbaum, M. (1974). Population
change and economic and social development. Studies in Family Planning, 5(12),
363–365.
Meier, G. M., & Baldwin, R. E. (1966). Economic development: Theory, history, policy.
New York: Wiley.
Meyer, J. W., Boli, J., Thomas, G. M., & Ramirez, F. O. (1997). World society and the
nation-state. American Journal of Sociology, 103(1), 144–181.
Mitchell, R. P. (1969). The Society of the Muslim Brothers. London: Oxford University
Press.

131
Mitchell, C. (2009). Three essays on worldviews, autonomy and the family in Nepal.
Doctoral dissertation, University of Michigan.
Moaddel, M. (1994). Political conflict in the world economy: A cross-national analysis
of modernization and world-system theory. American Sociological Review, 59(2),
276–303.
———. (2005). Islamic modernism, nationalism, and fundamentalism: Episode and dis-
course. Chicago, IL: University of Chicago Press.
———. (2010). Religious regimes and prospects for liberal politics: Futures of Iran,
Iraq and Saudi Arabia. Futures, 42, 532–544.
Motahhari, M. (1969). Mas’aleh-ye hijab [The problem of veiling]. Tehran: Islamic
Society of Physicians.
Nisbet, R. A. (1975) [1969]. Social change and history. New York: Oxford
University Press.
———. (1980). History of the idea of progress. New York: Basic.
Norris, P., & Inglehart, R. (2002). Islamic culture and democracy: Testing the clash
of civilizations thesis. Comparative Sociology, 1(3/4), 235–264.
Osella, F., & Osella, C. (2006). Once upon a time in the West: Stories of migra-
tion and modernity from Kerala, South India. Journal of the Royal Anthropological
Institute, 12(3), 569–588.
Parish, W. L., Laumann, E. O., & Sanyu, A. M. (2007). Sexual behavior in China: Trends
and comparisons. Population and Development Review, 33(4), 729–756.
Parsons, T. (1964). Evolutionary universal. American Sociological Review, 29,
339–357.
Pigg, S. L. (1992). Inventing social categories through place: Social representations
and development in Nepal. Comparative Studies in Society and History, 34(3),
491–513.
Slyomovics, S. (1995). Hassiba Ben Bouali, if you could see our Algeria: Women and
public space in Algeria. Middle East Report, 25(192), 8–13.
Taraki, L. (1996). Jordanian Islamists and the agenda for women: Between discourse
and practice. Middle Eastern Studies, 32(1), 140–158.
Thornton, A. (1989). Changing attitudes toward family issues in the United States.
Journal of Marriage and the Family, 51(4), 873–893.
———. (2001). The developmental paradigm, reading history sideways, and family
change. Demography, 38(4), 449–465.
———. (2005). Reading history sideways: The fallacy and enduring impact of the develop-
mental paradigm on family life. Chicago, IL: University of Chicago Press.
Thornton, A., Binstock, G., Abbasi-Shavazi, M. J., Ghimire, D., Gjonça, A., Melegh,
A., et al. (2012). Knowledge and beliefs about National Development and
Developmental Hierarchies: the Viewpoints of Ordinary People in Thirteen
Countries. Social Science Research, 41(5), 1053–1068.
Thornton, A., Binstock, G., & Ghimire, D. (2008). International dissemina-
tion of ideas about developmental and family change. In R. Jayakody, A.
Thornton, & W. Axinn (Eds.). International family change: Ideational per-
spectives (pp. 19–4 4). New York: Lawrence Erlbaum Associates, Taylor &
Francis Group.

132
Thornton, A., & Young-DeMarco, L. (2001). Four decades of trends in attitudes

toward family issues in the United States: The 1960s through the 1990s. Journal
of Marriage and the Family, 63(4), 1009–1037.
United Nations Development Programme. (2009). Human development report:2009:
Overcoming Barriers: Human Mobility and Development. New York: Palgrave
Macmillan.
Wallerstein, I. (1974). The modern world-system: Capitalist agriculture and the origins of
the European world-economy in the sixteenth century. New York: Academic Press.
———. (1991). Unthinking social science: The limits of nineteenth century paradigms.
Cambridge: Polity.
———. (1997) [1979]. The capitalist world economy: Essays by Immanuel Wallerstein.
Wang, Z. (1999). Women in the Chinese Enlightenment: Oral and textual histories.
World Bank. (2011). A short history [On- line]. Available: http://go.worldbank.org/
U9BK7IA1J0. Retrieved August 29, 2016.
Yount, K. M. (2004). Symbolic gender politics, religious group identity, and the
decline in female genital cutting in Minya, Egypt. Social Forces, 82(3), 1063–1090.
Yount, K. M., & Rashad, H. (Eds.). (2008). Family in the Middle East: Ideational change
in Egypt, Iran, and Tunisia. Oxford: Routledge.
Yount, K. M., Thornton, A., Mehanna, S., & Patel, S. (2016). Lay accounts of “modern”
and “traditional” family in greater Cairo: A test of developmental models of family
life. Paper presented at the Symposium on the Globalization of Modernization
Theory: Clashes of Modernities and Moralities, Ann Arbor, MI, June 9–10, 2010.

133
CHAPTER 5 Lay Accounts of “Modern” and

“Traditional” Family in Greater Cairo
A Test of Developmental Models of Family Life
Kathryn M. Yount, Arland Thornton,
Sohair Mehanna, and Shilpa N. Patel
Schemas are important tools for people to understand and to evaluate the physi-
cal and social world around them. Modernization, or an evolutionary model of
socioeconomic development, has been one of the most important models for people
around the world. This model outlines a trajectory of socioeconomic change as
well as a framework for ranking countries, institutions, people, and families along
a continuum from less developed, or traditional, to more developed, or modern. This
model thus defines which kinds of societies and families are developed, or mod-
ern, and which kinds of societies and families are less developed, or traditional. It
also stipulates the societal and family factors that lead ostensibly to socioeconomic
development. In this way, developmental models provide familial, societal, and
economic goals to be attained and the mechanisms to achieve these goals. These
developmental models have been spread around the world through many mecha-
nisms, and the acceptance, rejection, or modification of such models can affect
many dimensions of family life (Thornton, 2001, 2005).
In this chapter, we assess the nature and extent to which Egyptians under-
stand and use developmental models to conceptualize the family. We explore the
meanings Egyptians attach to the concepts of traditional family and modern fam-
ily, the ways Egyptians use these terms, and whether these meanings and uses
conform to the perspective of the modernizing elites around the world during the
contemporary period (Thornton, 2001, 2005). We are interested, specifically, in
whether lay Egyptians follow the schema of scholars and policymakers in defining
traditional families as undesirable and backward, and as displaying stereotypical
features, such as extended households, arranged marriage, young age at marriage,
uncontrolled and high fertility, and gender inequality. We also are interested in
whether Egyptians follow the modernization schema in defining modern fami-
lies as desirable and civilized, and as displaying other attributes, such as nuclear
households, love marriages, older age at marriage, controlled and low fertility, and
gender equality.
134
We address these questions by analyzing data from 84 informants. Each infor-

mant took part in a freelisting (FL) exercise—a qualitative technique designed to
identify the elements of a cultural domain (Pelto, 2015). We asked informants to
list the attributes they associated with traditional families and those they associ-
ated with modern families. We then organized the informants into groups defined
by gender, age, and school-level attainment, and conducted 12 group discussions
that addressed similar topics (three age groups, each having a group with lower
schooling attainment and a group with higher schooling attainment, totaling six
groups each for men and women; see “Data and Methods”). Triangulating these
data, we assessed how local models of the family among lay Egyptians matched
developmental models of traditional and modern family. Our findings offer
insights into how lay Egyptians have received, resisted, and adapted these models
to resonate with local history, society, and culture.
Developmental Models and Their Global Dissemination

The ideas associated with “modernity” or a “developmental model of change”
have a long history. They have been influential in Western thought at least since
the age of the Enlightenment during the late 1700s and 1800s (Thornton, 2001,
2005). This model posits that a developmental or evolutionary model of historical
change, which assumes that all societies pass through similar stages of devel-
opmental trajectory (Harris, 1968; Mandelbaum, 1971; Nisbet, 1969). The speed
of change along this trajectory is seen to vary across societies, resulting in their
placement at different stages of development in any given period of time. Societies
that are believed to be at the low end of this trajectory are labeled undeveloped, tra-
ditional, or backward whereas societies believed to be at the high end are labeled
developed, modern, or advanced.
Advocates of this evolutionary perspective have claimed that the most modern
societies are found in northwest Europe and its overseas populations, and that
other societies are distributed across this developmental hierarchy, below the pin-
nacle of northwest Europe. Using this developmental model to rank the countries
of the world, Western scholars of the 1700s and 1800s created a methodology for
describing the process of development (Thornton, 2001, 2005). Instead of describ-
ing societal development by tracking a particular society over time, they compared
societies at the same time (Carneiro, 1973; Gordon, 1994; Harris, 1968; Thornton,
2001, 2005). Their method assumed that the conditions observed in a so-called
traditional contemporary society proxied the prior conditions of a so-called modern
society. In this way, the socioeconomic attributes of northwest European coun-
tries came to represent those of so-called modern societies whereas the attributes
of countries outside northwest Europe came to represent those of so-called tradi-
tional societies (Thornton, 2001, 2005).
One of the early and influential examples of developmental idealism in action
is James Mill’s History of British India. In providing “scientific” justification for
Westernizers’ policy in India, James Mill (1773–1836) used the criteria of civili-
zation drawn chiefly from “conjectural” history, the Benthamite principles of

135
utility, and the general intellectual heritage of the Enlightenment (Mill, 1848).
As Forbes (1951) explains, for Mill, the organizing principle of history was the
“scale of nations.” Conjecturing that the place of Indians on the scale was low,
he condemned every single aspect of their way of life as barbarous—not only
their science, but their philosophy, art, and manners as well. In Duncan Forbes’s
assessment, “given the ‘scale of nations’ as the organizing principle of history,
and the uniformitarianism of Rationalist historiography, the logical result is an
extreme form of Europocentricism, ‘scientifically’ established” (Forbes, 1951, p. 31;
Moaddel, 2005).
A similar scale was used in assessing cross-national variation in family life.
Crucial for our study of lay perceptions about the attributes of traditional and
modern families is that family life in northwest Europe during the 18th and 19th
centuries differed markedly from family life elsewhere (Malthus, 1986 [1803];
Millar, 1979 [1771]; Westermarck, 1894 [1891]; William, 1995 [1779]). Although
Western scholars of this era recognized diversity in family life within north-
western Europe, as well as within and across other countries of the world, they
stereotyped the differences between northwest Europe and elsewhere in terms
of the modern–traditional continuum. They generalized their observations that
families outside northwestern Europe had extensive solidarity, little individual-
ism, high parental control over adolescent children, young ages at marriage, uni-
versal marriage,1 marriages arranged by parents, polygamy, high levels of gender
inequality, and large and extended households. They focused further on their
observations that family life in northwestern Europe had less solidarity, greater
individualism, less parental control over adolescent children, older ages at mar-
riage, more people never marrying, more marriages arranged by couples through
courtship, monogamy, more gender equality, and smaller and more nuclear (or
stem) households.
With their developmental model and cross-cultural data, Western scholars of
the late 1700s and 1800s classified as modern the familial attributes observed in
northwestern Europe and classified as traditional the familial attributes observed
elsewhere (Thornton, 2001, 2005). Many of these scholars also concluded that
northwestern Europe had, at some time in the past, traditional-family patterns
like those observed elsewhere and that, with development, had adopted the
modern-family attributes observed in northwest Europe during the 1700s and
1800s. These scholars also assumed that the process of development in societ-
ies outside northwest Europe would cause them to move from having tradi-
tional family systems to having modern ones. Moreover, as northwest European
societies experienced greater control over childbearing and declines in fertility
during the 19th and 20th centuries, controlled and low fertility became associ-
ated with modernity; uncontrolled and high fertility became associated with
traditionality. The thinking, methodology, and conclusions of these scholars
have influenced academics, politicians, activists, and laypeople for centuries.
This developmental model and its corresponding conclusions offered a power-
ful means to judge societies and families. It prescribed a set of ideals and the
means for people to attain them. Arguably, this developmental model, which
Thornton (2001, 2005) has identified as developmental idealism—for better or
L ay Accounts Of “modern” And “traditional” Family In Greater cairo | 135

136
worse—became an ideological engine for many social, economic, and familial

changes. It has fueled an ethos favoring a modern society that is industrialized,
urbanized, highly educated, and has high levels of knowledge and technology.
It also has fostered a preference for so-called modern over so-called traditional
families and has proposed that a modern society is both a cause and an effect of
a modern family.
These ideas have spread globally through many means. Scholars of the 1700s
and 1800s distributed their treatises widely in both Europe and elsewhere. The
expansion of schools, mass media, industrialization and urbanization, European
colonization, international aid programs, and international organizations such as
the United Nations were other means of the diffusion of developmental idealism.
Social movements—such as those for political democracy, women’s rights, and
international family planning—were both energized by these ideas and influential
in spreading them (Thornton 2001, 2005).
Archival and ethnographic research from diverse settings in Africa, India,
China, Nepal, and New Guinea have documented examples of ordinary people
relying on developmental models and ideas to assess human attributes and behav-
ior (Ahearn, 2001; Amin, 1989; Blaut, 1993; Caldwell, Caldwell, & Quiggin, 1989;
Comaroff & Comaroff, 1997; Dahl & Rabo, 1992; LiPuma, 2000; Pigg 1992; Wang,
1999). Surveys from Argentina, China, Egypt, Iran, Malawi, Nepal, and the United
States have revealed widespread understanding and acceptance of developmental
models and their relationships to familial attributes (Abbasi-Shavazi, Nodoushan,
& Thornton 2012; Binstock & Thornton, 2007; Lai & Thornton 2014; Thornton,
Binstock, & Ghimire, 2008; Thornton et al., 2012; Thornton, Pierotti, Young-
DeMarco, & Watkins, 2014).
As these ideas spread globally, people who had longstanding local systems of
family life were exposed to them. Each local family system has had its own attri-
butes, some of which were unique to that society. Moreover, the precise mech-
anisms and contacts of local societies with external ideas of traditionality and
modernity have varied substantially across societies. For these reasons, the ideas
of traditionality and modernity have not been adopted without examination.
Instead, they are assessed, with one outcome being rejection of some elements
and modification of others. These reflections suggest that ideas about tradition-
ality and modernity can vary markedly across societies. Some individuals and
societies will not understand all the differences that Western scholars associ-
ated historically with traditionality and modernity. Others will identify certain
aspects of family life as traditional or modern that were not part of the models
that Western scholars espoused. As a result, it is useful to distinguish in any soci-
ety between elements that match this external model from those that may come
from within the society.
As we discuss in the next section, archival, ethnographic, and other evidence
from Middle Eastern countries, including Egypt, suggests that developmental
models and ideas have spread in this part of the world. Such evidence makes
Egypt an interesting place to explore how local people understand and value the
concepts of traditional and modern family, and to use these concepts in ways that
resemble or differ from the international communities discussed previously.

137
Egypt as a Context for the Flow of Ideas About Family

Historical Context
For much of its history, Egypt has been a “linchpin of regional trends … in
the Middle East” (Marr, 1999, p. xiii). Throughout the 19th century—the period
between Napoleon’s invasion and occupation of the country (1798–1801) and the
nationalist revolution of 1919—Egypt experienced considerable social transforma-
tion, including the emergence of capitalism; integration of a local economy with
a global economy; decline of the old social classes; emergence of new dominant
classes, including landowners, merchants, and a nascent middle class; and the
formation of a state with an extensive bureaucratic and military organization. The
Suez Canal opened in 1869, connecting the Mediterranean and Red seas. In so
doing, the Canal permitted travel by sea between Europe and South Asia without
navigating around Africa, adding to Egypt’s regional and global economic and
political import. Cultural transformation was underway concurrently, where the
grips of Islamic orthodoxy weakened and various cultural discourses, including
evolutionary thought, diversified Egypt’s intellectual market (Moaddel, 2005).
These changes placed Egypt at the forefront of cultural movements that many
other Middle Eastern countries experienced later. During the 20th century,
demography and geography added to Egypt’s regional centrality. Demographically,
with more than 83 million residents in 2014, Egypt is the most populous country
in the Arab world and Middle East (United Nations, Department of Economic and
Social Affairs, Population Division, 2014). Situated geographically at the intersec-
tion of Europe, Africa, and Asia, Egypt is linked to the flows of people and ideas
among the three continents. And, Egypt remains a hub for transport and commu-
nications. The Suez Canal remains a vital shipping channel between the Indian
Ocean and the Mediterranean Sea. Overland routes via the Sinai connect Egypt to
southwest Asia, and the Nile links Egypt to Sub-Saharan Africa.
Egypt’s regional centrality also facilitates its cultural and political influence. For
centuries, Egypt has been a cultural crossroads, contributing to cultures as diverse
as the Greeks, Nubians, Phoenicians, and Romans (Marr, 1999). Egypt also has
been and remains a hub for Islamic learning, with Al-Azhar—founded in Cairo
in 970 AD—being the oldest, continuously existing university in the world and a
center for Sunni law and theology. Egypt’s television and radio programs, films,
newspapers, magazines, and books have spread Egyptian influence within and
beyond the Arab world (Abu-Lughod, 2005). Egypt has been a leader in the rise
of Islamic modernism during the 19th century and Islamic fundamentalism of
the 20th century. The Society of the Muslim Brothers formed by Hasan aI-Banna
in 1928 is one of the strongest and most enduring Islamic fundamentalist move-
ments in the region.
Since World War II, the Arab world often has turned to Egypt for political guid-
ance; Egypt, for example, led the Arab League from 1945 to 1979, and again from
1989 to the present. The League, although a voluntary organization, embodies
collective Arab ideals and provides an institutional framework for collective action
(Marr, 1999). This historical influence in the Middle East has led the United States

138
and other Western countries to view Egypt strategically as a political and cultural
gateway to the region. As a result, the exposure of lay Egyptians to developmental
models of the family is undeniable, and the propagation of such models through
schools, the media, and family-planning programs is illustrative.
Social Change and a New Conception of Desirable Family
Depictions of Family through Schools

The first major and systematic push toward the modernization of Egyptian
educational system transpired under Khedive Ismail (1863–1879). The Khedive
increased the educational budget dramatically, expanding the available resources
for the production of what was defined as modern culture. This new educational
policy was directed and implemented by such formidable administrators as Ibrahim
Pasha and Ali Mubarak Pasha. The School of Languages and Administration was
reopened and, in 1886, it became the first secular law school under Vidal Pasha, a
French jurist. In 1872, the Dar al-Ulum teachers’ college was founded, which later
played a leading role in the revival of Arabic literature (Moaddel, 2005).
This process continued well into 20th-century Egypt, where schools and educa-
tional materials constituted a chief mechanism through which Egyptian officials
promoted what Thornton (2005) considers the developmental–idealistic values
and models of the family (Szyliowicz, 1969). Lisa Pollard (2003) has described
how Egyptian civil servants (the effendiyya) under British occupation linked such
idealistic family values to nationalist struggles for independence and dissemi-
nated these ideals through textbooks taught in public schools:
Early 20th-century textbooks were highly prescriptive about what kinds of fam-
ily relations were proper and fitting to the national struggle. First, children were
instructed that the proper home consisted of a father, mother, and their obedi-
ent children. Polygamy was very clearly discouraged, as was the habit of having
extended family share the domicile. Thus, the family was redefined to fit the
models of Victorian domesticity; Arabic readers often contained lessons … in
which very precise definitions for household relationships were laid out. … Thus
the process of liberating Egypt was cast as a family affair. (p. 23)
The presence of such values has been corroborated in contemporary public opin-
ion surveys, where schooling is said to be linked to accepting certain features of
modern family. For instance, Egyptian women with more secular schooling have
tended to favor activities associated with women’s autonomy (Kishor, 1995).
The textbooks of adult literacy classes during the 1990s also were designed
to produce similar effects. According to Laila Abu-Lughod (2005), specific les-
sons imparted to rural Egyptians exemplify the “efforts of state culture,” at times
funded by US agencies, to promote a specific set of modernist values, such as
those pertaining to women’s rights, consensual marriage, and small, nuclear fam-
ilies. One such reading lesson, for example, depicted the small nuclear family as
the happy and loving family: “Kamil loves his family. Habiba is Kamil’s wife. …
Tariq and Samah are Kamil’s children. Kamil’s family is small. Kamil’s family is
small and happy” (p. 62).

139
Arguably, however, schools have been a medium by which to disseminate com-

peting ideas about the family in Egypt (Szyliowicz, 1969). As early as the 19th
century, the introduction of European education to Egypt created a cultural duality
in the country; Egyptian society began to “divide into two. … One half retained
the traditional system imbued with Islamic teachings, while the other half mod-
eled itself after Europe” (Ali [1989] as cited in Cook [2000, p. 485]). Since then,
debate has ensued among the elites about the best balance between religion and
so-called modern culture, with contemporary Egyptian opinions split on how edu-
cation should be imparted (Cook, 2000). Critics of Egypt’s educational system
have stressed the “spiritual and moral cost” of the “West’s accent on technologi-
cal development” and have favored other “measures of progress” to the “material
standards proffered by the West” (Cook, 2000, p. 487). These critics have linked
directly the perceived lack of tarbiyya (ethical nurturing) in schools to questions
of Egyptian identity (Cook, 2000, p. 487). Such concerns have led many to turn
to other forms of education, such as private Islamic schools and free private les-
sons by Islamic charitable organizations. These mediums of instruction (as well
as some public schools [Cook, 2000]) have been conduits for alternative models of
the family, which Egyptian feminist scholars claim that their Islamic character has
stymied women’s equal participation in public life (Shukrallah, 1994). Egyptian
schools thus have become a channel to disseminate diverse ideals about modern
and traditional family.
Depictions of Family in the Media

Diverse portrayals of the family in schools have coincided with diverse por-
trayals of the family in the media. Several scholars have documented the use of
Western developmental models of the family in the press, film, and television.
During early 20th-century revolutionary Egypt, for example, effendi nationalists
used the press to promote developmental models of the family, nation building,
and their causal linkages (Pollard, 2003):
This equation between domestic behavior and the success of the nation appeared
in the Egyptian press as well. … Marital practices and the habits of couples inside
and outside of the home were common topics of discussion in the early national-
ist press. Proper ages for marriages, the pros and cons of arranged marriages,
choosing the “right” spouse, all found their way into the press. … In both the
classroom and the press, Egyptians were reminded that their bourgeois, modern
behavior had political ramifications. (pp. 24–27)
Similarly, portrayals of the southern, or Sa’idi, family as backward pervade
contemporary Egyptian television and film (Abu-Lughod, 2005). One example
includes the 1990s television series Dream of the Southerner. According to Abu-
Lughod (2005), the series writer “deployed … stock themes of rural backward-
ness and Upper Egyptian violence” as well as “arbitrary patriarchal authority …
expressed most often in the control over women and the institution of forced
arranged marriage” (p. 58). More generally, Abu-Lughod has argued that Egyptian
serials and films since the 1960s have pitted “ ‘ignorant,’ uneducated Upper
Egyptian clans … against patriotic educated modernizers” whose “tutelage”

140
eventually enables the Upper Egyptian protagonists to “break with the negative
traditions of their kin” (p. 58). The exposure of local Egyptians to such media is
reflected in contemporary national surveys, which show that Egyptian women
who have watched more hours of television have more often favored women’s
noncustomary autonomy (Kishor, 1995).
Egyptian state-sponsored serials also have depicted “good Sa’idis” who join
with “honest and patriotic urban Alexandrians in a black and white morality play”
(Abu-Lughod, 2005, p. 60). Such representations reflect a “modernist discourse
… in state culture” that dominated much of the 20th century: “the educated, cul-
tured individual” who represented “the good … and pride in the greatness of the
nation’s heritage” (p. 60). In this sense, state-sponsored educational curricula and
television serials were part of the same project of what Abu-Lughod (2005, p. 61)
has called “national pedagogy.”
Depictions of the Family in the State’s Family-Planning Program

The state’s deployment of a “modernist discourse” on the family also emerged
in its creation of a national family-planning program. Yet, the state’s initiation into
family planning during the 1950s met with popular reinterpretations of messages
about family planning and small family size (Bier, 2008):
[M]‌essages intended to promote family planning and small family sizes were fre-
quently interpreted by their target audiences in ways antithetical to the intentions
of program planners. At a rural social center in Menoufiyya, a poster featuring
the “ideal” family of three children and contrasted with a family of seven chil-
dren in ragged clothing was shown to local women during an outreach session
on family planning. When asked for their interpretation of the picture, women
pointed to the smaller family as the family that was “muzluma” (oppressed) and
the larger family as the happy family. [This] scene and ones like [it] were repeated
throughout the country. … [They] suggest that family planning discourses, and
the nationalist and state discourses on parenthood and domesticity that underlie
them, are read and interpreted through multiple lenses and may be deployed in
contradictory ways. (p. 70)
Even late-20th-century family-planning programs in Egypt met with some resis-
tance to certain messages about the family, economic development, and their pre-
sumed linkages (Ali, 1997):
The Egyptian state in collaboration with international donor agencies uses its
family planning program as a tool to modernize its population. … In this pro-
cess, the normalization of conjugal marriage and the nuclear family helps to
construct the modern categories of the “individual,” the “private,” and ultimately
“modern citizens.” … [H]‌owever, these constructions rest on slippery terrain . …
In intellectual circles as well as among the poor, questions are raised about the
assumed “natural” linkage of economic development and population control.
The increasing disparity in income levels that is quite independent of family
size subjects the emphasis on family planning to criticism. … Moreover, popular
notions of the body, healing, fertility, and sexuality undermine the legitimacy of
state-sponsored medical and social impositions. (p. 43)

141
Together, these accounts illustrate the diverse ideas about family that have cir-
culated through schools, the media, and family-planning programs during the
20th century. They suggest that Egyptians, indeed, have encountered a range of
developmental models and ideals about modern and traditional family, and that
Egyptians have reacted to these models in various ways. Given this theoretical
and contextual backdrop, we now explore the subjective meanings and attributes
that Egyptians in Greater Cairo attach to the concepts of traditional and modern
family. The results provide insights about how Western developmental models
for these concepts are received, evaluated, and adapted in one Middle Eastern
context.
Data and Methods

Sample
The study used 12 focus-group discussions (FGDs), each with six to eight par-
ticipants. These participants were selected from the metropolitan area of greater
Cairo and varied by age (from 15–54 years), gender, and schooling attainment.
Final participants in the study were selected from a list of prospective partici-
pants that local collaborators generated from older, lower-and middle-class areas
as well as industrial and commercial areas in greater Cairo, including Cairo gov-
ernorate, the urban parts of Giza and Qaliubia governorates, and Helwan and
Sixth of October cities. Participants were selected systematically from this list
to create a diverse and balanced sample across gender, age (ranges, 15–24 years,
25–44 years, 45–54 years), schooling attainment (less than secondary, secondary
or more), and occupation. The final sample included 84 participants (41 women,
43 men) in the 12 focus groups, who varied in age from 15 to 54 years and in
schooling from none or illiterate to postgraduate training (Table 5.1). Several
Table 5.1 Study Participants
Gender Age range, y Schooling n
Men, n = 41 18–22 None, illiterate, primary 7

15–20 Preparatory, postsecondary 6
25–42 None, illiterate, primary 7
25–33 Postsecondary, university 7
45–54 None, illiterate 8
45–54 Secondary, postsecondary 6
Women, n = 43 17–21 None, illiterate, primary 8
16–23 Primary, preparatory, postsecondary 8
29–40 Postsecondary, university 7
46–53 Postsecondary 8
Total 15–54 None to university 84

142
occupational groups were also represented, including unskilled laborers, small-

business owners, students and teachers, tailors and fashion designers, com-
puter programmers, social workers, government employees, salespeople, and
homemakers.
Individual Interviews and Focus-Group Discussions

Each participant was interviewed separately before taking part in an FGD so that
personal perceptions could be recorded before exposure to ideas from the group.
The separate interviews covered a range of topics, including sociodemographics,
perceptions of the attributes of traditional and modern marriages and families,
subjective evaluations of specific attributes of families, and perceptions about the
effects of family change on society and of economic development on families.
These interviews also included questions about the attributes of developed and
less developed societies, subjective evaluations of specific attributes of society,
and ratings of countries on scales of development, education, wealth, and gender
equity. Here, we focus on participants’ perceptions and evaluations of the concepts
of traditional and modern family.
To understand the attributes and meanings that participants assigned to
these concepts, interviewers administered two FL exercises after asking par-
ticipants about their demographic background, and before asking any other
questions about family and development. FL is a qualitative technique used
to identify the elements of a cultural domain (Weller & Romney, 1988). In the
most basic form of FL, interviewers ask the members of a homogeneous group
one open-ended question, inviting each member to list all the attributes of the
cultural domain of interest. Interviewers probe in a nonleading fashion to elicit
a complete list of responses from each participant, and all responses are poten-
tial elements of the cultural domain. In general, the total number of items
that participants mention reflects the salience of the cultural domain. The
relative frequency and average rank order in which an item on the list is men-
tioned indicate the salience of the item. In this study, interviewers asked each
participant: In your opinion, what are all of the characteristics of traditional
marriages and families? Interviewers then recorded verbatim the attributes
mentioned, and probed for additional attributes until no others were men-
tioned. Interviewers conducted a similar FL to ascertain the locally assigned
attributes of modern marriage or family.
After completing the individual interviews, participants were organized into 12
focus groups, as described earlier and as depicted in Table 5.1. A local collaborator
moderated these discussions. The discussion guide opened with a question about
the groups to which participants felt they belonged—to which all focus groups
immediately mentioned family. A follow-up question asked each group about the
types of families they could imagine. The moderator then presented four photos
depicting (a) a poor Egyptian family, (b) a wealthier Egyptian family, (c) an Anglo-
European family, and (d) a family from a Persian Gulf Arab country (Figure 5.1).
Each group described each photo in their own words. All discussions were tape-
recorded, transcribed into Egyptian colloquial Arabic, and translated into English.

143
Figure 5.1 (A–D) Photos used in the 12 focus group discussions.
Each transcript was read, “memoed,” or annotated for relevant themes, and coded
in NVivo 7 (QSR International, Melbourne, Australia).
Data Analysis
The FL responses formed the basis of this analysis, and quotes from the group
discussions were used to illustrate and supplement the findings from the FL. In
theory, the attributes listed in an FL are individual words that are aggregated eas-
ily. For more abstract cultural domains, however, respondents may describe its
elements in phrases that interviewers record verbatim. The latter response format
was typical in this study, and so the initial stages of analysis required a careful
review and recoding of the FL responses to permit aggregation. One researcher
assigned to the original FL responses preliminary emic codes, which were intended
to capture in words or short phrases the attributes or characteristics communi-
cated in the response. A second researcher reviewed and revised these prelimi-
nary codes and added others. To ensure intercoder concordance, both researchers
reviewed and agreed on the final coding of each response. Similar procedures
were used to code the FL responses pertaining to traditional and modern families,
respectively. A researcher then computed relative frequency distributions of the
final codes depicting the attributes of traditional and modern families, respec-
tively. Three relative frequency distributions were computed for each family type
to assess the salience of each attribute: (a) the percentage of respondents (n = 84)
mentioning each attribute, (b) the relative frequency of each attribute among all
coded responses (even if sequential responses within a respondent were assigned
the same code), and (c) the relative frequency of each attribute using only nondu-
plicative attributes for each respondent.

144
Results
The next sections provide results and a window into lay understandings of “tradi-
tional marriage and family” and “modern marriage and family” in Greater Cairo.
Perceptions of “Traditional Marriage and Family” in Greater

Cairo: Multiple Competing Models?
Assigned Attributes of Traditional Marriage and Family

As shown in Table 5.2, participants identified 37 attributes of traditional mar-
riage or traditional family. Eight attributes were mentioned by at least 20% of the
participants. These attributes included preparations for marriage, family bonds,
procedures for mate selection, premarital courtship, morality and religion in fam-
ily life, the relatedness of spouses, social and material lifestyle, and the compatibil-
ity of the match between spouses and their families. Between 10% and 20% of the
participants mentioned each of the six other attributes pertaining to the extent of
adherence to customs, marital power, levels of fertility, gender roles, the stability
of marriage, and other aspects of marital and family bonds. Less than 10% of the
participants mentioned each of the remaining 23 attributes.
Of the 14 most commonly mentioned attributes, two clearly emerged as the
most salient. Seventy-nine percent of the participants mentioned the first of these,
which related to the bride’s and groom’s financial obligations in preparation for
marriage. These participants delineated distinct financial obligations of the bride
and her family to contribute specific consumer durables as part of the bride’s
trousseau:
MMA: [Until] now, … the traditional family insist[s]‌on getting a dowry for the
bride and furnishing the bride’s house. … (FGD, men, 15–24 years, second-
ary schooling)
AMR: The bride’s needs and the kitchen necessities are her responsi
bility . … (FL, woman, 25 years, secondary schooling)
These participants also described the distinct financial obligations of the groom
and his family.
AAF: The apartment is the responsibility of the groom. (FL, man, 19 years, pri-
mary schooling)
AAA: The groom has to buy a wedding gift, a gift of gold that the family shows
off. (FL, man, 38 years, postsecondary schooling)
We should clarify what the participants saw as uniquely traditional about these
preparations for marriage. Namely, our participants did not describe as uniquely
traditional the general process of preparing financially for marriage; rather, they
stressed the specific and gendered obligations in this process, such as the groom’s
duty to buy a gold wedding gift and the apartment, and the bride’s duty to buy the
kitchen utensils and other items in her trousseau. The high frequency of this attri-
bute reflects our expectation that lay populations include elements of local mar-
riage systems into local models of traditional family. Our participants’ conception

145
Table 5.2 Freelisting Results for the Attributes of Traditional Family/Marriage
Attribute % (N = 84)a
[Bride’s/groom’s] financial obligations in preparation for marriage 78.6

Strong family bond, closeness, without problems, affection, love, 56.0
care for one another, compassion
Family has role in partner selection, arranged marriages 31.0
Limited courtship, acquaintance before marriage 28.6
Incorporating morals, values, religion into one’s life 23.8
Relative marriages 22.6
Simple lifestyle, content with life, less attention on monetary matters 21.4
Social, educational, cultural, financial compatibility of bride/groom, 21.4
their families
Upholding traditions and customs in the context of marriage 19.0
Husband in control, wife obedient 17.6
Many, average number of children 14.3
Gender complementarity, husband provides financially; wife 14.3
cares for home, id
Stability, success in marriage 11.9
Marital conflict, weak marital bonds and family bonds 11.9
Low income, simple means 9.5
Parents care for children, supervise children, are happy for children, 9.5
raise children well
Mutual respect, honesty, trust, contentment, romantic feelings 9.5
between spouses (after marriage)
Acquaintance, affection before marriage; not arranged marriage 8.3
Early marriage (for girl) 7.1
Children deal well, respect their parents 7.1
No informal (orphi), announced marriages 6.0
Higher education is not a priority 6.0
No love in marriage 6.0
Parents set limits, rules for children 4.8
Few children 3.6
A street, “low-class” wedding party 2.4
Living within one’s means 2.4
Living with husband’s family after marriage 2.4
Negative perceptions of traditional families, marriage 2.4
Keen knowledge 1.2
Son preference 1.2
Respect for neighbors 1.2
Length of engagement depends on groom’s finances 1.2
Discreet in bride’s appearance 1.2
Wedding celebration and wedding party 1.2
Groom is able to go on with his life after marriage preparations 1.2
No care for children and their needs 1.2
note: Orphi refers to informal, temporary marriage.

a
Percentage of respondents mentioning this attribute.

146
of local marriage systems as part of a so-called traditional family corroborates

findings from ethnographic research on marriage and the family in Egypt, which
describes a widespread preoccupation with the costs of marriage (e.g., Singerman
& Ibrahim, 2003). Accordingly, marriage is the “occasion for a major intergenera-
tional transfer of wealth,” often “larger than the inheritance following a parental
death” (Singerman & Ibrahim, 2003, p. 80). For this reason, marriage demands
substantial assets from both families, which are often accumulated over a long
time and in gendered ways.
The second most salient attribute of traditional marriage or family related to its
affective and instrumental bonds. A majority (56%) of participants mentioned this
attribute, stressing the affection, care, cohesion, and reciprocity that traditional
family members experience with one another:
AAF: There are cohesion and strong attachment among family members espe-
cially at times of crises. … (FGD, men, 15–24 years, less than secondary
schooling)
DAH: In the traditional family, they take care of each other. Family members
are attached to one another. (FGD, women, 15–24 years, less than secondary
schooling)
AAM: The [traditional] family sticks together; they work hand in hand. (FL,
woman, 45 years, no schooling)
These quotes echo the characteristic of “familism” in developmental models of
traditional family, where familism is seen as prioritizing family relationships
and commitments over individual needs and concerns (Thornton, 2001, 2005).
Correspondingly, ethnographers have identified norms of reciprocity, mutual
exchange, and “connectivity” to be basic elements of Arab models of kinship
(Joseph, 1996). Thus, the depiction by our participants of traditional families as
having strong bonds echoes developmental models and ethnographies of ideal
Arab kinship in the Middle East.
Between one fifth (21%) and one third (31%) of the participants mentioned six
other attributes of traditional marriage or family. These attributes, in order of
salience, were arranged marriage or the strong role of family in mate selection
(31%); the limited courtship or acquaintance of the bride and groom before mar-
riage (29%); the presence of morals, values, and religion in family life (24%); the
marriage of blood relatives (23%); the simple lifestyle, contentment with life, and
lesser attention to monetary matters (or nonmaterialistic, 21%); and the social,
educational, cultural, and financial compatibility of the bride, groom, and their
respective families (21%).
AMMR: Parental approval is necessary for the marriage [and] their opinion is
important. (FL, woman, 26 years, secondary schooling)
NRN: There is not much dialogue between the boy and girl prior to the engage-
ment. (FL, woman, 25 years, more than secondary schooling)
ASS: [There is a] commitment to moral rules and religious practices such as
praying regularly. (FL, woman, 50 years, more than secondary schooling)
ASS: [There is] stability inside the family since they know each other, they are
relatives. (FL, woman, 50 years, more than secondary schooling)

147
SSA: Life is within the means of their income. A simple and basic lifestyle. (FL,
women, 39 years, primary schooling)
NAS: There is social compatibility between the husband and wife. (FL, woman,
54 years, more than secondary schooling)
The next six attributes of traditional marriage or family were mentioned by 12%
to 19% of the participants. These attributes were more diverse, but among them,
two concerned the roles and relative power of husbands and wives, and two con-
cerned the nature of the marital and family bond. The specific attributes, in order
of salience, were the upholding of traditions and customs in marriage (19%); the
husband’s control and power in decisions and the wife’s obedience (18%); the bearing
and raising of an average to a large number of children (14%); the complementarity
of husband–wife roles, in which the husband is the financial provider and the wife is
the homemaker and primary caretaker of the children (14%); the stability and success
of marriages (12%); and, contrarily, the weak bonds of such marriages/families (12%).
SSA: [They get married] exactly … how their parents got married. (FL, woman,
25 years, secondary schooling)
AAA: The nature of the relationship dictates that man has the upper hand. (FL,
man, 50 years, more than secondary schooling)
RAH: The number of children is usually big (FL, man, 29 years, more than
secondary schooling)
SSA: The man is responsible for the house expenses. (FL, man, 46 years, more
than secondary schooling)
AMR: T]he mother is responsible for the children, food, and drink. (FL, woman,
25 years, secondary schooling)
MAA: A stable marriage that has no conflicts. (FL, man, 17 years secondary
schooling)
NVN: The lack of mutual understanding between the husband and wife. (FL,
woman, 27 years, more than secondary schooling)
Lay Model Versus Developmental Model of Traditional

Family: Congruence and Contextual Specificity
It is notable that some of the most salient attributes on this freelist corresponded
to those highlighted in developmental models of family life. According to
Thornton (2001, p. 456), “the traditional end” of this continuum is characterized
by familism, young and parentally arranged marriage, high parental control, a low
status for females, limited control of fertility, and extended households. Some of
the most frequent related responses in this sample included the traditional fam-
ily’s tendency to have a strong bond (second most frequent response), to practice
some form of arranged marriage (third) and to limit courtship before marriage
(fourth), to expect wives to obey their husbands (10th), and to have an average to
high number of children (11th). Although, in the freelists, less than 3% of partici-
pants mentioned co-residence with the husband’s family as characteristic of tradi-
tional families, family structure and living arrangements across family types were
discussed spontaneously in several focus groups (findings available on request).

148
Although many mentioned spontaneously attributes of traditional marriage

and family overlapped with those characterizing traditional family in domi-
nant developmental models (Thornton 2001, 2005), other salient attributes
did not overlap. These attributes included the distinct financial obligations of
brides and grooms in preparation for marriage (first), consanguineous or rela-
tive marriage (sixth), and the preservation of morals, values, and religion in
family life (fifth). We postpone discussion of the moral character of traditional
families and focus on participants’ references to certain local marital practices
as traditional.
The ascription of consanguinity and gendered marital preparations to the
construct of traditional family suggests that local marriage systems are power-
ful influences on lay beliefs about this construct. Consanguineous marriages are
common in the Middle East, constituting about one third of marriages in Egypt
and between 23% and 66% of marriages in the region (El-Zanaty & Way, 2006;
Tabutin & Schoumaker, 2005). Although rates of relative marriage vary across
countries in the region, national rates have remained quite stable since the 1970s
(Tabutin & Schoumaker, 2005). Thus, consanguineous marriage, although not
universal, has been a prominent feature of Middle Eastern marriage systems.
Thus, lay Egyptians may interpret such features as customary, and in this sense
are considered part of a traditional family model.
Subjective Meanings of Traditional Family:

Backward or Venerated?
The traditional family, as depicted in developmental models, has distinct character-
istics and a particular subjective meaning, as “undesirable,” “backward,” “uncivi-
lized,” or “less developed.” Some attributes of traditional family that emerged in
the freelist had similarly negative connotations, including the low priority given to
schooling and the failure of parents to care for their children:
NSE: [There is a] lack of persistence in having the children continue their educa-
tion (FL, woman, 31 years, more than secondary schooling)
MEH: [There is] no care for children and their needs (FL, man, 20 years, sec-
ondary schooling)
These attributes, however, were mentioned infrequently (e.g., by 6% and 3%
of participants, respectively), and tended to be mentioned by the same study
participants. Namely, 4 of the 84 participants mentioned only neutral or nega-
tive attributes. The male participant just cited, for example, described tradi-
tional families and marriages in these terms: “There are more children …
no mutual love … [and a] multitude of problems after marriage. There is no
prior acquaintance before the marriage … [and] no care for children and their
needs” (FL, man, 20 years, secondary schooling). Thus, the participants in this
sample rarely described the traditional family categorically as “undesirable” or
“backward.” Corroborating these findings were those from the focus groups,
in which the traditional family was discussed much less often in negative (five
times) than positive (19 times) terms.

149
Instead, many salient attributes that participants assigned to traditional mar-

riage or family had distinctly positive connotations. Examples included its strong
bond (ranked second); its preservation of moral behavior, esteemed values, and/or
adherence to religion (ranked fifth); its simple, content, and nonmaterialistic life-
style (ranked seventh); the social, educational, cultural, and financial compatibility
of spouses (ranked seventh); and its overall stability and success (ranked 13th).
Less frequently mentioned but positive attributes pertained to the parental care
and supervision of children (ranked 15th), the mutual respect and bond between
spouses after marriage (ranked 16th), and the tendency for children to respect
their parents (ranked 19th).
MAA: Families are bound by respect, customs, and traditions and social codes.
(FL, man, 50 years, more than secondary schooling).
FEAF: Parents watch out for their children and take care of them. (FL, woman,
48 years, primary schooling)
In addition to the high frequency of positive attributes accorded to traditional
families in the sample, 25 of 84 participants described traditional families in
exclusively positive terms. For example, the previously cited 48-year-old woman
with primary schooling described the traditional family in these terms: “There are
love and affection between them. Good moral and religious values. Parents watch
out for their children and take care of them. The parents bring up their children
on religious values and rules. The husband and wife share life expenses” (FL,
woman, 48 years, primary schooling).
This tendency to describe traditional families auspiciously contradicts its nega-
tive meaning in developmental models and in related images of the “backward”
Sa’idi in Egyptian television dramas (Abu-Lughod, 2005). Moreover, the common
ascription of attributes reflecting local marriage systems suggests that the par-
ticipants viewed the traditional family as culturally “authentic,” and perhaps a
model of resistance against corrupt imports. According to two participants in the
younger, educated women’s focus group:
RMM: If [traditional families] become open to the outside world with all its good
and bad, the bad will become more prominent than the good. They have to
be careful not to take everything.
SMH: Like when we see homosexuality abroad, and we start finding it here in
Egypt, that will bring us more damage, like AIDS. We should not imitate
the wrong doings of other societies. When the society becomes infected,
each cares only for himself, and people lead wrong lives. How will that bring
about a progressive society? (FGD, women, 16–23 years, primary or higher
schooling)
These quotes echo the words of Yuval-Davis (1997, p. 63), who argues that
“[t]‌he … essentialization of identities … is a defensive reaction to … global-
ization.” In other words, external threats to cultural identity may induce people
to construct a myth of some homogenous family of the past, which comes to
represent pure and authentic collective culture. The “good traditional Egyptian
family” may be one such myth, which, paradoxically, may include appropriated

150
features of developmental models of traditional family. Overall, such findings

corroborate the idea that Caireans have encountered multiple, competing mod-
els of traditional family, including but not only that in Western models of devel-
opment. Laypeople, then, may select and evaluate separately elements from
these competing models to create a meaningful hybrid.
However, the attachment of contradictory meanings to traditional family sug-
gests that some individuals are conflicted in their evaluation or that subgroups
assessed the concept differently. Both of these interpretations are plausible. First,
20 of 84 participants assign both positive and negative attributes to traditional
families. Second, older participants (35–54 years) and those with many children
(six to seven) more often depicted traditional families as having strong bonds,
and older participants depicted traditional family members as sharing mutual
respect. In contrast, younger participants more often depicted traditional families
as neglecting education and lacking love in marriage, and participants with three
or fewer children depicted traditional families as having marital conflict and weak
family bonds.
Thus, our findings suggest that the term traditional family has substantial
salience among people in Greater Cairo. The term reflects an amalgam of attri-
butes that coincide with developmental and cultural models of the family, as well
as prominent features of local marriage systems. This amalgam, along with fre-
quent positive evaluations of traditional family, contradicts developmental models
and signifies the presence of substantial resistance, at least among certain sub-
groups, to what they believe as Western cultural intrusion. The ascription by indi-
viduals of contradictory meanings to traditional family also suggests this concept
is contested and in flux. Now we turn our discussion to perceptions of modern
marriage and family in Greater Cairo.
“Modern” Marriage and Family in Greater Cairo: Disintegration

and Imitation of the West?
Assigned Attributes of Modern Marriage and Family

As reported in Table 5.3, participants mentioned 47 attributes of modern mar-
riage and family. Seven of these attributes were mentioned by at least 20% of par-
ticipants. These attributes included the cost of the wedding, the nature of family
bonds, attitudes toward money and material goods, premarital courtship, fertility
and family size, public appearances, and procedures for mate selection. Between
10% and 20% mentioned one of the nine other attributes concerning the standard
of living and education, nature of the marital bond, morality and religion in family
life, stability of marriage, parent–child relations, distribution of financial obliga-
tions for marriage, nature of the match between spouses, other aspects of the
marital bond, and gender roles in marriage. Less than 10% mentioned the other
31 attributes.
By far, the most salient attribute of modern marriage and family is related
to the high cost and materialistic nature of the wedding and “requirements” for
marriage. The 48% of the participants who mentioned this attribute typically

151
Table 5.3 Freelisting Results for Open-Ended Question on the Attributes of a Modern Family/Marriage
Expensive, fancy wedding celebration (in club with wealthy guests); big or expensive wedding gift or dowry 47.6
Weak family bonds, disintegration of family, lack of closeness, self-interest 38.1
Concern for finances, money, materialism 28.6
Love, romance, courtship, acquaintance before marriage 28.6
Few children, planning family, delayed childbearing after marriage 21.4
Concern for appearances 20.2
Selecting partner and engagement arrangements without consulting family, no forced-arranged marriage, Internet marriage 20.2
Standard of living is high, high income, higher education 19.0
Love, trust, understanding, respect in marriage 16.7
Do not abide by moral values, religious beliefs 16.7
Lack of stability of marriage, separation and divorce, marriage fails fast, marriage is unsuccessful 15.5
Parents give children excessive freedom to do what they want, parents have little control over children, children are corrupt 14.3
Shared, flexible financial responsibilities in preparation for marriage 13.1
Social and financial compatibility of couple (at a high level) 11.9
No love, trust, mutual understanding in marriage, conflict in marriage 10.7
Husband and wife share family responsibilities, work, house chores, management, childrearing; or wife manages house 10.7
Attributes of modern women 9.5
Informal (orphi) marriage 9.5
Wealthy husband, husband with prestigious job, high financial demands on husband 9.5
Marriage at older age (especially for girl), age does not matter w.r.t. marriage 8.3
Parents do not care or pay attention to children, children are not a priority 8.3
Husband and wife share decisions, opinions; husband not in control 7.1
Family members exercise freedom, independence 6.0
Parents love, care for, educate their children 6.0
Open-mindedness, freedom of opinion 6.0
(continued)
152
Table 5.3 Continued
Success, stability in marriage 4.8
No dowry, wedding gift 3.6
Youthful autonomy, freedom of children 3.6
Strong family bond, closeness 2.4
Leisure time, activities 2.4
Family has role in partner selection 2.4
Use of technology, computer 2.4
Difference in treatment of modern families 1.2
“Premarital sex” 1.2
No relative marriage 1.2
No son preference 1.2
Similar to traditional (marriage) 1.2
Prevalent today 1.2
Evolved from traditional (marriage) 1.2
Short engagement 1.2
Total dependence on family 1.2
People do not get married in a traditional way 1.2
Good behavior and decency are higher 1.2
They travel abroad for their honeymoon 1.2
They aspire for more, are very ambitious 1.2
Long engagement 1.2
People have tough demands in the marriage 1.2
note: Orphi refers to informal, temporary marriage.

a
Percentage of respondents mentioning this attribute.
153
described a fancy wedding celebration in a club with wealthy guests, and a large or
expensive wedding gift and dowry.
MAA: Wedding parties take place in luxurious hotels and the couple lives in
luxury. (FL, woman, 50 years, postsecondary schooling)
SHE: [There is] exaggeration in marriage requirements like the apartment, car,
and cell phone. (FL, woman, 42 years, primary schooling)
Although preparations for the wedding and marriage had certain financial require-
ments for traditional and modern families, the excessive and materialistic nature
of these expenses was attributed exclusively to modern families.
The second most salient attribute assigned to modern family was its weak
bonds, which 38% of participants mentioned (Table 5.3). These participants typi-
cally stressed the lack of closeness, infrequent contact, and self-centeredness of
the members of modern families.
ARA: Each of the family members is on his own; there is no closeness in the
family. (FL, woman, 18 years, postsecondary school)
IAA: [There is the] absence of family attachments and closeness. For example,
in a modern family, the siblings or family members may see one another by
chance. The family only gets together on special occasions. (FGD, women,
29–40 years, postsecondary schooling)
Between 19% and 29% of participants mentioned six other attributes of modern
marriage or family. These attributes were (a) concern for finances, money, and
materialism (29%); (b) love, romance, courtship, and acquaintance before mar-
riage (29%); (c) having few children and using family planning (21%); (d) concern
for appearances (20%); (e) selecting one’s partner and making arrangements for
the engagement without consulting family (20%), and (f) having a high standard
of living or income (19%). The following are quoted examples of each of these
attributes:
SMA: He has to have the latest car model. (FL, man, 22 years, primary schooling)
AMMR: Marriage mostly happens after a love affair. (FL, man, 26 years, primary
schooling)
LAM: [They have] few children, one or two, family planning, and delayed child-
bearing after marriage. (FL, man, 27 years, university schooling)
ARS: [They care about] keeping appearances in front of the extended family. (FL,
woman, 18 years, primary schooling)
KMH: [They] select the partner and engagement arrangements without con-
sulting the family; no forced-arranged marriage, Internet marriage. (FL,
woman, 52 years, no schooling)
NRA: The economic level is high and money is abundant. (FL, man, 20 years,
primary schooling)
Between 11% and 17% of participants mentioned the next eight attributes of mod-
ern marriage or family. These attributes were more diverse, and in some cases,
contradictory, including (a) love, trust, understanding, and respect in marriage
(17%); (b); not abiding by moral values or religious beliefs (17%); (c) instability
in marriage, separation, and divorce (16%); (d) parents giving children excessive

154
freedom (14%); (e) shared, flexible financial responsibility in preparation for mar-

riage (13%); (f) the social and economic compatibility of the couple (12%); (g) the
absence of love, trust, or mutual understanding in marriage (11%); and (h) equi-
table spousal roles, in which the husband and wife share financial and domestic
responsibilities (11%).
NVN: There is more understanding between the husband and wife, they know
each other. (FL, man, 27 years, postsecondary schooling)
LAM: Morality is lacking to some extent. The children do not dress modestly
and do not practice religious rituals such as prayers on a regular basis. This
is very scarce in modern families. (FGD, women, 29–40 years, postsecond-
ary schooling)
ZSL: [There is] a high percentage of separation and divorce due to a lack of
agreement. (FL, woman, 54 years, no schooling)
NRA: The children could take drugs, pills, and shots. (FL, man, 20 years, pri-
mary schooling)
SMH: There is flexibility on issues of the marriage gift, dowry, and marriage
expenses. (FL, woman, 29 years, postsecondary schooling)
RMM: It is very improbable that they would marry from a lower class than
theirs. (FL, man, 20 years, postsecondary schooling)
EMA: The smallest problem causes a big drift between them. This marriage has
lots of [problems]. (FL, woman, 49 years, postsecondary schooling)
RRM: They equally work. They share house chores. (FL, woman, 21 years, pri-
mary schooling)
Lay Attributes of Modern Family vis-à-vis Developmental

Models: Congruence and Divergence
Given these attributes of modern marriage or family, 6 of the 10 most
salient ones aligned with developmental models of modern family. According
to Thornton (2001, p. 454), elites described such families historically as having
high individualism, nuclear households, marriages arranged by mature couples,
youthful autonomy, courtship preceding marriage, a high valuation of women,
and controlled and low fertility. Similar attributes of modern family among our
participants included the practice of premarital courtship (fourth), weak family
bonds and self-centeredness of its members (second), low fertility and the use of
family planning (fifth), selection of spouses without family involvement (seventh),
and more egalitarian gender roles in marriage (16th).
Despite these similarities, 4 of the 10 most salient attributes of modern
marriage or family did not overlap with those in developmental models. These
included the expense of the wedding celebration, gift, or dowry (first); extent of
materialism or (over)concern with money (third); concern for appearances (sixth);
and the absence of moral values and religious beliefs (10th). These attributes
reflected neither local marriage systems nor local cultural models of the family.
As we discuss next, many participants perceived these practices as arising from
cultural intrusion from the West.

155
Subjective Meaning of Modern Family: Progressive or Corrupt?

The modern family and its attributes, as depicted in developmental models, carry
the positive subjective meanings of being “good,” “progressive,” “civilized,” or
“more developed.” Some attributes of modern family that participants mentioned
in the freelist also had positive connotations, including the presence of love, trust,
and understanding in marriage; the shared or flexible financial preparations
for marriage; the shared roles and responsibilities of spouses; the presence of
open-mindedness and freedom of opinion; and the stability and success of such
marriages:
MAA: There is more trust and honesty between the two [spouses]. (FL, woman,
17 years, primary schooling)
MEH: There is cooperation in setting up the house. (FL, woman, 20 years, pri-
mary schooling)
EYM: [The spouses] help each other financially and stand by each other. (FL,
man, 17 years, postsecondary schooling)
RRM: There is freedom of opinion and contentment. (FL, woman, 21 years, pri-
mary schooling)
SMH: Marriage lasts for a long time. (FL, man, 18 years, postsecondary
schooling)
The last two attributes, however, were mentioned infrequently (e.g., by 6%, and
5% of participants, respectively) and tended to be mentioned by the same study
participants. Only 11 participants, who were younger (seven younger than 25 years,
four who were 30–40 years) but varied in gender (five men, six women) and school-
ing (five primary or less, six postsecondary), mentioned only neutral or positive
attributes. One of these 11 participants described modern families and marriages
in these terms: “Prior acquaintance between the boy and girl before marriage.
There is more trust and understanding between the two. There is higher educa-
tion, they are more civilized. Good behavior and decency are higher” (FL, woman,
17 years, primary schooling). Thus, few participants described the modern family
as unequivocally “good.”
Instead, 10 of the 16 most salient attributes assigned to modern family had dis-
tinctly negative connotations. Examples included the materialism of the wedding
(first); weak family bonds (second); general materialism and concern for money
(third); concern for appearances (fifth); disregard for moral values and religious
beliefs (10th); frequent separation, divorce, and failed marriages (11th); excessive
freedom of children and their corrupt/immoral behavior (12th); and absence of
love, trust, and understanding in marriage (16th). Less common but still nega-
tive attributes concerned the practice of orphi or temporary marriage (18th), the
low priority given to children (21st), and the occurrence of premarital sex (34th).
Likewise, in the focus groups, the modern family was described less often in posi-
tive (six times) than negative (28 times) terms.
In addition to the frequent ascription of negative attributes to modern family
in the sample, 35 of 84 participants described modern families only in neutral
or negative terms. A 48-year-old man with no schooling, for example, described

156
the modern family as having “excessive freedom in everything. No commitment

to morals or religion. There is disintegration in the family and lack of closeness.
The father doesn’t have a say in family affairs” (FL, man, 48 years, no schooling).
This tendency to describe the modern family inauspiciously contradicts its
positive meaning in developmental models (Thornton, 2001, 2005) and in pop-
ular images of the enlightened northern2 Egyptian (Abu-Lughod, 2005). Also,
the frequent statements of participants that modern families “lacked” certain
attributes—such as morals or religious beliefs, love or understanding, care for
children, dowry or a wedding gift, relative marriage, son preference, and custom-
ary marriage—hinted that participants saw this model as foreign. Indeed, accord-
ing to the discussion of older men with less schooling, the modern Egyptian
family was an “imitation” of the Western family:
MA: The modern family wants to be like the foreign family.

AEG: There is no difference between them.
KMH: There is no difference between them
Moderator: The modern family wants to be like the foreign family?
Everyone: Yes
AEG: And there are no differences between them.
…
Moderator: OK, what about customs and characteristics? Maybe they look
alike in appearances. But what about in their customs and characteristics?
AEG: There are no differences in that too. They are the same in their customs.
ZSL: They are very similar because the modern family is trying to imitate the
West. (FGD, men, 45–54, no schooling or illiterate)
This attachment of contradictory meanings to modern family suggests that some

participants were ambivalent in their evaluations or that subgroups assessed the
concept differently. Indeed, 18 of 84 participants assigned positive and negative
attributes to modern families. And, compared with those with more schooling,
those with no schooling more often depicted modern families as having extrav-
agant weddings, spending excessively on preparations for marriage, being con-
cerned with appearances, and lacking moral and religious values (data available
on request). The younger, less-schooled women, for example, contrasted the reli-
giosity and morality of traditional and modern families in this way:
HAM: [In the traditional family, the father] teaches the children moral values
and about religion. Maybe this [traditional] family would know its religion
and its morals better than this [modern] one. (FG, women, 17–21 years, illit-
erate, no schooling, or primary schooling)
Still, negative views of modern family existed among the more schooled. The
older, more-schooled men, for example, contrasted the bonds of traditional and
modern families in this way:
MAA: In the traditional family, there is more intimacy and stronger family ties,
whereas in the modern family the closeness is below average. (FGD, men,
45–54 years, secondary or postsecondary schooling)

157
And, the middle-aged, more-schooled women specified the immoral behaviors of

modern families:
LAM: Morality is lacking. … The children do not dress modestly and do not
practice religious rituals such as prayers on a regular basis. This is very
scarce in modern families. … That is why you find orphi [informal] mar-
riage and inappropriate sexual relations. A lot of bad things are found in
modern families. (FGD, women, 29–40 years, postsecondary schooling)
Thus, in Greater Cairo, the term modern family appears to be salient and reflects a
mix of attributes that coincide with developmental models of modern family and
local models of the Western family. The ascription by individuals of conflicting
meanings to modern family suggests that this concept is contested and in flux.
Also, its frequent negative assessment exposes resistance among some groups
to models of modern family that are seen as Western. At this point, we discuss
three models of family change that lay constructs of traditional and modern fam-
ily may imply.
Three Models of Family Change in Greater Cairo

The models in Figure 5.2 are derived from group discussions and present
three models of family change linking lay constructs of traditional, modern,
and Western families. Models A and B reflect a linear process of family change
from traditional to modern, and, ultimately, to Western; however, these mod-
els differ in the perceived causes and meanings of such change. Model C also
describes a linear process of family change; yet, in this case, the transition
from traditional to modern is distinctly Arab. We first describe the tendency
for participants to rank family types on a linear continuum of “development.”
We then present evidence from group discussions for these three models of
family change.
Linear Ranking of Traditional, Modern, and Foreign/Western Families

The tendency to rank family types was common across focus groups. For
example, older men with more schooling described the traditional roots of mod-
ern families:
KIM: [T]‌his modern family that we’re talking about could be the outcome of this
[traditional] family.
AAA: Exactly.
KIM: When the children grew up, got educated, got more freedom, and started
to know more.
AAA: This traditional family is basically a rural one that lives in one of the vil-
lages of Egypt. This modern family has probably stemmed from rural roots.
They moved from the countryside to the city. (FGD, men, 45–54 years, sec-
ondary or postsecondary schooling)
Accordingly, modern families were seen to be born in the span of a generation
(“When the children grew up”) by acquiring education, more freedom, and knowl-
edge, and by migrating from villages to cities. Several focus groups extended this

158
model of change by ranking Western families with respect to modern ones. The
middle-age group of more-schooled men clarified:
SMH: No, [modern and Western families] have similar standards of living
AAA: It is almost the same.
SMH: And the way of thinking is similar because now we have taken the path
towards modernization.
AAA: To become modernized.
SMH: Or becoming more like Western people. So, the modern family here may
look like the foreign family in its homeland.
AMA: No, this picture presents a family type that is in between the traditional
type and the Western type
Moderator: You mean the modern family?
AMA: Yes, in the freedom they have, for example (FGD, men, 25–33 years, post-
secondary schooling)
These participants thus placed Western families at the completion of a modern-
izing process, with modern Egyptian families intermediate to traditional Egyptian
and Western families. In this sense, modernization was equated (or conflated)
with westernization.
The tendency of the focus groups to rank family types generally in a linear
progression from traditional to modern to western also appeared in the tendency
of the focus groups to rank specific attributes of families on the same linear scale.
In the last and more structured portion of the discussions, each group was asked
to describe traditional, modern, and foreign or Western families in terms of 15
prespecified attributes. These included, among others, the typical age at marriage,
extent of parental involvement in marriage, distribution of spousal roles, status of
the wife, extent of divorce, number of children, and extent of parental control over
children. Table 5.4 summarizes the responses of each focus group to questions
about the number of children and the extent of parental control over children that
the focus groups associated with traditional, modern, and foreign or Western fam-
ilies. All focus groups, regardless of their gender, age, and schooling, perceived
that levels of fertility and parental control over children were highest in traditional
families, intermediate in modern families, and lowest in foreign or Western fami-
lies. With the exception of premarital sex, extramarital childbearing, and same-
sex marriage, the focus groups ranked all the other family attributes in a similar
way from traditional to modern to Western (results available on request). For the
three exceptional attributes, all focus groups set apart Western families as the only
type in which these behaviors occurred. From the perspective of these informants,
then, Western families are characterized by distinct qualities of decadence.
“It Would Be Great If We Follow Their Path”: Family Change as Part of

Economic Development and Westernization
As might be expected, participants attached different subjective meanings to
processes of family change in which modernization was conflated with western-
ization. The focus group of middle-age women with postsecondary schooling

Table 5.4 Scalability of Selected Attributes of Family, Structured Portion of Focus Group Discussions
Attribute Gender Age, y Schooling Traditional Modern Foreign
Number of children? M 18–22 None, illiterate, primary 5–6 2–3 2–3

M 15–20 Preparatory, postsecondary 3–5 2–4 1–2
M 25–42 None, illiterate, primary 2–7 1–2 1–2
M 25–33 Postsecondary, university 5–7 2–3 0–2
M 45–54 None, illiterate 2–3 1–3 1–2
M 45–54 Secondary, postsecondary 2–4 1–2 1
F 17–21 None, illiterate, primary 5–10 2–3 0–2
F 16–23 Primary, preparatory, postsecondary 4–12 2–3 1–2
F 29–40 Postsecondary, university 4–6 2 2–3
F 46–53 Postsecondary 2–5 1–3 1–2
Parental control over children? M 18–22 None, illiterate, primary 100% low None
M 15–20 Preparatory, postsecondary 100% 50% 1%
M 25–42 None, illiterate, primary 100% Little None
M 25–33 Postsecondary, university 50%–99% 30% Not at all
M 45–54 None, illiterate 70% 50%–60% None
M 45–54 Secondary, postsecondary 90% 50% None
F 17–21 None, illiterate, primary 90% Some None
F 16–23 Primary, preparatory, postsecondary 60T–65% 50% ≤40%
F 26–42 None, illiterate, primary ≤80% ≤50% 40%
F 29–40 Postsecondary, university 100% Less traditional None
F 45–54 None, illiterate, primary 95% 50%–60% None
F 46–53 Postsecondary 100% 50% None
note: F, female; M, male.

160
Family change = economic Family change =

development and westernization westernization and disintegration
(and this is good) (and this is bad)
Model A Model B
W
est
,
rel med
ion
“D on a /scho
ern
igi
isi nd
dia cat Western family Traditional family
nte cu oling
me du
wealthy, advanced, good, moral, and
gra ltur , a
nd h e
ia
civilized authentic
l, a ug
tio al c nd
ve hro
n”
fro rrup avel
tra ent t
m
o tr
em
Modern family
los ion b o W
nc
s o y est
trying to imitate
t
va
Modern
f
Ad
the West
t
Traditional family Western family
bad and backward corrupt and immoral
Model C s
igiou
l, rel
ora
,m
tal Modern
ien
Or
Traditional
Western
• Veneration of traditional and modern “oriental” families, which have a common cultural,
religious, and moral base, in opposition to Western families.
• Local family types are “innate” and therefore not easily penetrated by external influence.
Western types have always been corrupt.
• With their distinctive religious and moral base, Arab families can undergo a culturally
authentic process of modernization.
Figure 5.2 Competing models of family change in Greater Cairo.
described such change positively, as “progress” that resulted from “economic

development” and the “interplay” of cultures (Figure 5.2, model A):
NRN: The progress is due to the economic development.
NRN: [The modern family is] like the West. Just a step behind in everything.
…
NRN: This will be the next step. With more wealth, the modern family will be
exactly like the Western family.
…

161
IHO: It is also the cultures … the interplay between the different cultures.
NRN: Because the poor family wants to be like the middle-class family and the
middle-class family wants to be—
NSE: Like the modern family.
NRN: And the modern wants to be like the foreign family. Each one is aspiring
to the one above it.
NSE: To the higher level.
NRN: To the higher level of course. If I am a low-level employee, I will want to
be a mid-level employee. And if I am a mid-level employee, I will want to
be a senior employee, and so on until I quit. (FGD, women, 29–40 years,
postsecondary schooling)
Family types in this depiction were analogous to company employees who aspired
to climb the corporate ladder of success. Poor families occupied the lowest rung,
below middle-class and modern families, all of whom were motivated to achieve
(and exceed) the “higher level” of Western families. Although this depiction
of change was construed largely in economic terms, it almost conflated seam-
lessly economic modernization with westernization. Across all focus groups,
such changes were assigned most often to increased education (157 times) and
standards of living (135 times), but also to general changes in the environment
(35 times), as well as exposure to the media (60 times), increases in technology
(31 times), and rural-to-urban migration (27 times).
“I’m Talking about the Disintegration of the Family”: Decline Resulting

from Western Intrusion
Not all participants viewed westernization as a favorable aspect of family
change. For some, modern families embodied a precarious step in the pro-
cess of family change, in which “family disintegration” was a possible outcome
(Figure 5.2, model B). In this case, modern families abandoned the virtues of
traditional families and attempted to imitate a Western family icon. The weaker
bonds and more provocative dress of modern families, for example, resulted
ostensibly from their exposure to Western culture, their consequent abandon-
ment of morality and religious practices, and their vacuous imitation of corrupt
Western families. One 30-year-old woman with university schooling depicted
the deviance of one American family, the history of which she had learned at
university:
IAA: Last year in the university we studied this book called Events in America.
It had a story about a family whose members were criminals. They studied
different generations of this family. They had some of the members of this
family live in a good environment where the material conditions were good.
But, they found that criminal behavior was passed on to different genera-
tions in the family; seven generations in the family. So this shows that even
when they changed the surrounding environment of this family, they could
not change who they really were (FGD, women, 29–40 years, postsecondary
schooling)

162
As the deviance of Western families was seen as deeply rooted, efforts to emulate
them were seen to reflect “blind imitation” by Egyptian families who had surren-
dered their religious and cultural roots:
KIM: I’m talking about the disintegration of the family … it’s exactly like you
said, the foreign family is an icon that the modern family wishes to imitate.
…
HAS: Blind imitation.
…
KSH: [T]‌he foreign family has been brought up to be like that. This is how
they originally are; they have nothing to bind them together, but this [mod-
ern] family wants to relinquish its religion and customs to be like the other
family.
…
MAA: The modern family is trying to imitate the foreign family. (FGD, men,
45–54 years, secondary or postsecondary schooling)
For these people, this transition in no way reflected “progress” or “development,”
but instead reflected “disintegration” from a venerated traditional family (Figure
5.2, model B).
“We Have Our Own Society” and “We Should … Follow Our Path”: An
Arab Developmental Model of Family Change
For some participants, the path of family disintegration was avoidable and, in a
few cases, was not even a possibility. These participants adhered to a third model
of family change (Figure 5.2, model C). In this model, participants stressed the
(near) universal goodness of Egyptians, as well as the similarities of traditional
and modern Egyptian families. Participants juxtaposed these positive images with
the innately immoral Westerner and Western family. According to a 42-year old
woman with some primary schooling:
SAM: A bad person is one in a thousand here, but a good person there is one in
a thousand. (FGD, women, 45–54 years, illiterate, no schooling, or primary
schooling)
All focus groups, moreover, distinguished the sexual behavior of traditional and
modern families from Western families. Neither traditional nor modern Egyptian
families were depicted as having premarital sexual relations, extramarital child-
bearing, or same-sex marriage. In contrast, the sexual behavior of Western fami-
lies was “limitless”:
Moderator: In a traditional family, what are the relations between men and
women who are not married … ?
NSE: None exist between men and women from traditional families.
…
NVN: Work relations. They can be colleagues at work
Moderator: What other kind of relations are acceptable … ?
IAA: Work.
NSE: Buying, selling, neighborly relations.

163
Moderator: Any other relations?

Everyone: No.
Moderator: What about in a modern family?
NSE: Friendship, being colleagues
NVN: Friendship between families, observing social etiquettes of decency.
…
Moderator: Can there be romantic relations?
LAM: Yes there are.
NSE: Platonic romantic relations.
Moderator: What about in a foreign family?
IAA: You can find everything there. (FGD, women, 29–40 years, postsecondary
or university schooling)
The “goodness” and morality of traditional and modern Egyptian families is
depicted partly as “natural” or innate, but also is said to be anchored in religion:
Moderator: Would I change them if I changed their financial status?
SSM: They would only change superficially, but their core. …
Moderator: How do you mean superficially?
SSM: Meaning their economic level.
SAS: They would dress up nicely, and that kind of thing.
SSM: Yes, their general appearance would improve, but their core, their nature
and habits and customs, will remain the same.
SAS: That’s it; they have been brought up like this. (FGD, women, 46–53 years,
postsecondary schooling)
MEE: Religion anchors us to something good. But if there is no religion, no
prayers, and no God, then it would be different, like it is in the West. (FGD,
men, 45–54 years, illiterate or no schooling).
The religious, moral, and cultural distinctiveness of Egyptian families provides
the basis for an alternative path to modernization. This path, according to three
older women with postsecondary schooling, involved “liberation” with the reten-
tion of culturally authentic values and customs. This status of moral and culturally
authentic “liberation” distinguishes modern Egyptian families from Western ones.
SSM: Here [in the modern family] you find more values and principles although
they have more liberation.
AAS: The modern family has these things too, stepping out of traditions and
liberation.
NAS: We have customs and values in the modern family, but the European one
completely lacks them.
SSM: Complete moral corruption. Some families here have that, but the major-
ity of them still keep their traditions and principles. (FGD, women, 46–
53 years, postsecondary schooling)
Thus, Arab modernization was a process of “moral liberation,” which included
increases in the “economic level” (SSM), “personal freedom” (WEB), and “inde-
pendence” (AEG). These changes, however, had “limits” or “restrictions” (e.g.,
SHE, IYM, SMH) that were determined by religious, moral, and cultural codes

164
of “decency,” “politeness,” “etiquette,” and “respect” (e.g., AMA, FAM, FYA, HAS,
NAE, NVN, SAM). Ultimately, these restrictions avoided a “fall” into the “limit-
less,” “excessive,” and “absolute” state of the Western family (e.g., AAH, EYM,
FRM, IAA, SAM, SHE).
AEG: We have our own society and we should not be like other societies.
KMH: We have nothing to do with other countries.
AEG: We should focus on—look up to—those who follow our path not those
who follow the West. (FGD, men, 45–54 years, illiterate or no schooling)
Discussion
This chapter was inspired by the premise that the ideas connected to the develop-
mental evolutionary model of change have been widely disseminated around the
world and have shaped the beliefs and values of ordinary people (Thornton 2001,
2005). Such ideas have also influenced people’s perception of what constitutes tra-
ditional and modern family. We expected that ordinary people would have models
of traditional and modern families that were in many ways consistent with the
centuries-old models of scholars and policymakers. We also expected that, in other
ways, ordinary people would bring their own local historical culture to what they
defined as attributes of traditional and modern families. We investigated these ideas
in Egypt using data from individual freelists and focus group discussions involving
84 men and women varying in age, school-level attainment, and occupation.
In many ways, our findings are consistent with these expectations. Our
Egyptian study participants clearly understood the concepts of traditional and
modern families. Also, their ideas of traditional and modern families were largely
consistent with the models of traditionality and modernity that have been formu-
lated and disseminated for centuries by social scientists and other elites. Despite
these similarities, the specifics of Egyptian culture and history clearly affected the
perspectives of lay Egyptians on traditional and modern families. Egyptians used
their own cultural and historical experiences to define traditional and modern
families, with the result being that Egyptian conceptualizations were hybrids of
the international and the local.
Interestingly, unlike several generations of social scientists and other elites who
have interrelated or conflated Western and modern families, a large percentage of
our Egyptian respondents distinguished modern families sharply from Western
families. For them, modernity and Western were two different things. Many of
these Egyptians took the developmental model a step further and suggested a
developmental trajectory from traditional to modern to Western. That is, Western
is a step beyond modern on the developmental pathway. Yet, others placed the
West outside the pathway of development, at least as it relates to families, and said
there is no need for Egypt to follow the Western model of family change. In effect,
they were constructing an alternative modernity.
When it comes to evaluating traditional, modern, and Western family attri-
butes, some study participants followed the models of generations of social

165
scientists and other elites by denigrating the traditional and valuing the modern.
These participants saw the developmental trajectory as being upward in terms of
goodness. Such participants, however, were in the minority, as many of them saw
the traditional as generally good and the modern as generally bad. A number of
participants also expressed contradictory views about modern families, suggest-
ing a substantial degree of cognitive dissonance among those exposed to com-
peting models of the modern family. The majority of participants tended to see
any trajectory from traditional to modern to Western family types as tenuous or
generally negative. These results indicate that, although the traditional–modern
approach for conceptualizing and categorizing families was widespread among
our study participants, the assertion of developmental idealism that modern fam-
ily attributes are good has much opposition in this sample.
Some limitations of this analysis are noteworthy and suggest avenues for future
research. First, the three lay models of family change presented in this chapter do
not reflect all the opinions that study participants expressed; rather, they summa-
rize three of the most prominent typologies of family change. Other typologies of
the family may be emergent, however. In this vein, these data and other qualitative
studies may be useful to expose secondary and tertiary competing models of tra-
ditional and modern family. Second, our sample was drawn from neighborhoods
in and around Cairo, the capital city. The residents of this highly urbanized part of
Egypt are likely to have been exposed to more diverse concepts of the family from
the media, some of which may conflict sharply with the cultural beliefs of many
Egyptians. This purposive sample from greater Cairo also may not be representa-
tive of the residents of Greater Cairo or Egypt as a whole. Survey research using
a representative probability sample that asks about the attributes of traditional,
modern, Western, and Arab families would be a useful next step to validate our
findings. Third, this qualitative analysis is cross-sectional, and so the investiga-
tion of hypotheses about ideational change over time is not possible. Longitudinal
qualitative studies would be useful to explore in detail how schemas of the family
evolve over time. Finally, this qualitative study was focused on understanding how
lay Egyptians in Greater Cairo understand and view models of traditional and
modern family. We did not undertake any exploration of the influences of various
beliefs on actual changes in the family. Such research would require longitudinal
surveys in which researchers collected repeated measures for the models of family
depicted here and repeated measures of family behavior.
Despite these limitations, this study has several overriding strengths. First, the
collection of rich, qualitative data permitted a detailed exploration of the subjective
meanings that ordinary people in one local context attached to the constructs of
traditional, modern, Western, and Arab families. Second, the qualitative methods
used in this study clearly helped to identify local elements of these constructs
of family, which may be included in population-based surveys and which should
enhance the content validity of any quantitative analyses of these constructs.
Third, these qualitative data will provide a richness of detail that facilitates the
construction of quantitative variables and the interpretation of the findings from
statistical analyses.

166
Taken together, our findings suggest that lay Egyptians in Greater Cairo do
have clear conceptualizations of traditional and modern family, and hold strong
views about the merits of and pitfalls of specific pathways of family change. Many
participants in this study expressed concern about the influx of corrupting influ-
ences of the West, and many recommended an “alternative modernity” that com-
bined certain socioeconomic and demographic elements of the so-called modern
family ideal with the venerated moral, religious roots of the so-called traditional
Egyptian family.
Notes
1. Meaning, almost everyone eventually married.
2. Meaning, an Egyptian who lives in Cairo or to the north of Cairo—a region
where people have higher schooling, on average, and a higher relative standard of
living.
References
Abbasi-Shavazi, M. J., Nodoushan, A. A., & Thornton, A. (2012). Family life and devel-
opmental idealism in Yazd, Iran. Demographic Research, 26, 207–238.
Abu-Lughod, L. (2005). Dramas of nationhood: The politics of television in Egypt (1st ed.).
Chicago, IL: University of Chicago Press.
Ahearn, L. M. (2001). Language and agency. Annual Review of Anthropology, 30,
109–137.
Ali, K. A. (1997). Modernization and family planning programs in Egypt. Middle East
Report, 27(4), 4–40.
Amin, S. (1989). Eurocentrism (1st ed.). New York: Monthly Review Press.
Bier, L. (2008). From birth control to family planning: Population, gender, and the
politics of reproduction in Egypt. In K. M. Yount & H. Rashad (Eds.). Family
in the Middle East: Ideational change in Egypt, Iran, and Tunisia (pp. 55–79).
New York: Routledge.
Binstock, G., & Thornton, A. (2007). Knowledge and use of developmental thinking
about societies and families among teenagers in Argentina. Demográfia, 50(5),
74–104.
Blaut, J. M. (1993). The colonizer’s model of the world: Geographical diffusionism and
Eurocentric history. New York: Guilford Press.
Caldwell, J. C., Caldwell, P., & Quiggin, P. (1989). The social context of AIDS in Sub-
Saharan Africa. Population and Development Review, 15(2), 185–234.
Carneiro, R. L. (1973). Structure, function, and equilibrium in the evolutionism of
Herbert Spencer. Journal of Anthropological Research, 29(2), 77–95.
Comaroff, J. L., & Comaroff, J. (1997). Of revelation and revolution: The dialec-
tics of modernity on a South African frontier (Vol. 2). Chicago, IL: University of
Chicago Press.

167
Cook, B. J. (2000). Egypt’s national education debate. Comparative Education, 36(4),

477–490.
Dahl, G., & Rabo, A. (Eds.). (1992). Kam-ap or take-off: Local notions of development.
Stockholm: Stockholm Studies in Social Anthropology.
El-Zanaty, F., & Way, A. (2006). Egypt demographic and health survey 2005. Cairo,
Egypt: Ministry of Health and Population, National Population Council, El-
Zanaty and Associates, and ORC Macro.
Forbes, D. (1951). James Mill and India. Cambridge Journal, 5(1): 19–33.
Gordon, D. (1994). Citizens without sovereignty. Princeton, NJ: Princeton
University Press.
Harris, M. (1968). The rise of anthropological theory. New York: Thomas Y. Crowell.
Joseph, S. (1996). Patriarchy and development in the Arab world. Gender and
Development, 4(2), 14–19.
Kishor, S. (1995). Autonomy and Egyptian women: Findings from the 1988 Egypt demo-
graphic and health survey. Calverton, MD: Macro International, Inc.
Lai, Q., & Thornton, A. (2014). The making of family values: Developmental idealism
in Gansu, China [On-line]. Social Science Research, 51. Available: http://dx.doi.org/
10.1016/j.ssresearch.2014.09.012. Retrieved December 27, 2014.
LiPuma, E. (2000). Encompassing others: The magic of modernity in Melanesia. Ann
Arbor: University of Michigan Press.
Malthus, T. R. (1986) [1803]. An essay on the principle of population. In E. A. Wrigley
& D. Souden (Eds.). The works of Thomas Robert Malthus (Vol. II). London: William
Pickering.
Mandelbaum, M. (1971). History, man and reason. Baltimore, MD: Johns Hopkins Press.
Marr, P. (1999). Egypt at the crossroads: Domestic stability and regional role. Washington,
DC: National Defense University Press.
Mill, James. (1848). The history of British India (Vol. 1). London: James Madden.
Millar, J. (1979) [1771]. The origin of the distinction of ranks. In W. C. Lehmann (Ed.).
John Millar of Glasgow: 1735—1801. New York: Arno Press.
Moaddel, M. (2005). Islamic modernism, nationalism, and fundamentalism: Episode and
discourse. Chicago, IL: University of Chicago Press.
Nisbet, R. A. (1969). Social change and history. New York: Oxford University Press.
Pelto, P. J. (2015). What is so new about mixed methods? Qualitative Health Research,
25(6), 734–745.
Pigg, S. L. (1992). Inventing social categories through place: Social representations and
development in Nepal. Comparative Studies in Society and History, 34(3), 491–513.
Pollard, L. (2003). The promise of things to come: The image of the modern fam-
ily in state building, colonial occupation, and revolution in Egypt 1805–1922. In
L. Haney & L. Pollard (Eds.). Families of a new world: Gender, politics, and state
development in a global context (pp. 17–39). New York: Routledge.
Shukrallah, H. (1994). The impact of the Islamic movement in Egypt. Feminist
Review, 47, 15–32.
Singerman, D., & Ibrahim, B. (2003). The costs of marriage in Egypt: A hidden
dimension in the new Arab demography. In N. Hopkins (Ed.). The new Arab fam-
ily: Cairo papers in social science (Vol. 24, pp. 80–116). Cairo: Cairo University Press.

168
Szyliowicz, J. S. (1969). Education and political development in Turkey, Egypt, and

Iran. Comparative Education Review, 13(2), 150–166.
Tabutin, D., & Schoumaker, B. (2005). The demography of the Arab world and the
Middle East from the 1950s to the 2000s: A survey of changes and a statistical
assessment. Population, 60(5/6), 505–616.
Thornton, A. (2001). The developmental paradigm, reading history sideways, and
family change. Demography, 38(4), 449–465.
developmental paradigm on family life. Chicago, IL: University of Chicago Press.
Thornton, A., Binstock, G., & Ghimire, D. (2008). International dissemination of
ideas about development and family change. In R. Jayakody, A. Thornton, &
W. Axinn (Eds.). International family change ideational perspectives (pp. 19–44).
New York: Lawrence Erlbaum.
Thornton, A., Binstock, G., Yount, K. M., Abbasi-Shavazi, M. J., Ghimire, D., & Xie, Y.
2012. International fertility change: New data and insights from the developmen-
tal idealism framework. Demography, 49, 677–698.
Thornton, A., Pierotti, R. S., Young-DeMarco, L., & Watkins, S. (2014). Developmental
idealism and cultural models of the family in Malawi. Population Research and
Policy Review, 33, 693–716.
United Nations, Department of Economic and Social Affairs, Population Division.
(2014). World urbanization prospects: The 2014 revision, highlights (ST/ESA/SER.A/
352). New York, United Nations.
Wang, Z. (1999). Women in the Chinese Enlightenment: Oral and textual histories.
Weller, S. C., & Romney, A. K. (1988). Systematic data collection. Beverly Hills, CA: Sage
Publications.
Westermarck, E. A. (1894) [1891]. The history of human marriage. London: Macmillan.
William, A. (1995) [1779]. The history of women from the earliest antiquity to the present
time. Bristol: Thoemmes Press.
Yuval-Davis, N. (1997). Gender and nation. London: Sage Publications, Ltd.

169
SECTION III Social-Scientific Perspectives

on Collective Action, Political
Engagement, and Voting
Behavior
170
171
CHAPTER 6 The Roots of Political Activism

in Six Muslim-Majority Nations
Nancy J. Davis, Robert V. Robinson,
and Tom VanHeuvelen
The antigovernment uprisings in Tunisia, Egypt, Libya, and other pre-

dominantly Muslim nations that shook the world in 2010 to 2012 took most
academics, media pundits, and heads of state by surprise. Although Muslim
political activism has been a subject of much scholarly attention, most quanti-
tative research, especially since the events of September 11, 2001, has focused
on transnational terrorism—violence originating in one country that targets
citizens or interests of another country, and has used country-level variables
(e.g., gross national product per capita, inequality, state repression) to explain
this (Abadie, 2006; Krueger & Malečková, 2003; Li & Schaub, 2004). Rarely
have quantitative studies of Muslim-majority countries focused on domestic
protest, the bulk of which, as we shall show, is nonviolent (but, see Crenshaw
and Robison [2010]). Qualitative studies and observational accounts of move-
ments in the Muslim world report on the characteristics and motivations of
those involved (e.g., Wickham, 2002; Woltering, 2002); but, lacking systematic
samples, they cannot compare participants with nonparticipants. Important
as these studies are, they cannot tell us what leads Muslim citizens in pre-
dominantly Muslim countries—where democracy is rarely fully realized and
repression against opponents of the government is not uncommon—to involve
themselves in nonelectoral protest.
As Osa and Corduneanu-Huci (2003) note:
Since data on repressive regimes are scarce, social action that takes place despite
the risks imposed by a stable authoritarian state is not commonly studied. Yet …
the preconditions of a successful transition to democracy are found in courageous,
often abortive, attempts by social actors to challenge non-democratic institutions
and repressive practices. (p. 606)
In this chapter, through analyses of surveys gathered in 2000 through 2002 in
the fourth wave of the World Values Survey (WVS), we study political engagement
among Muslims living in six Muslim-majority countries (Algeria, Bangladesh, Egypt,
Indonesia, Jordan, and Pakistan), where just less than half the world’s Muslims live.
172
We explore the individual-level and contextual bases of a full range of nonelectoral

political engagement, from discussing politics with friends, through signing peti-
tions, joining boycotts, and participating in lawful demonstrations, to undertak-
ing extralegal actions such as joining unofficial strikes and occupying buildings or
factories. Such activism, which Inglehart and Catterberg (2002, p. 301) call “elite-
challenging,” differs from “elite-directed” engagement, such as voting, and may be
especially important in societies in which opportunities for political expression are
limited. Although signing a petition, marching in a demonstration, or “talking poli-
tics” with friends (Johnston, 2005, 2006) are arguably low risk, commonplace, and
even institutionalized in advanced democracies (Inglehart & Catterberg, 2002; Meyer
& Tarrow, 1998), they can be far more risky and less routine in relatively authoritarian
states. Because, in such societies, much of the impetus to engagement may be pent-
up, not yet fully expressed, and waiting for an opportune moment, we explore—in
addition to the actual engagement in a range of activism—the willingness to engage
in such activism.
Social Movement Theories and Muslim Political Engagement

Writing in 2004, Kurzman observed, “Over the past generation the fields of
social movement theory and Islamic studies have followed parallel trajectories,
with few glances across the chasm that has separated them” (p. 289). Although
several studies have appeared in recent years (e.g., Beck, 2009; Kurzman, 2004;
Wiktorowicz, 2004), none has brought social movement theory and quantita-
tive analyses of individual-level data to bear in explaining the willingness of
ordinary Muslims to engage in nonelectoral protest. Moreover, in qualitative
and quantitative studies of activists in predominantly Muslim countries, there
is no attempt to compare the demographic characteristics of activists (e.g., high
education, unemployment) with the incidence of the same characteristics in
the general population. Thus, we do not know which characteristics lead some
people to engage in protest whereas others do not. In this chapter, we apply
to Muslim political engagement in Muslim-majority nations social movement
theories positing that engagement is affected by differential recruitment, rela-
tive deprivation, social dislocation, and embeddedness in preexisting networks
of activists.
Differential Recruitment
Social movement theories of differential recruitment focus on the questions of
“who join[s]‌social movements, the characteristics and circumstances that pre-
dispose them to become activists, the mechanisms that mobilize some, and the
barriers that deter others from participation” (Nepstad & Smith, 1999, p. 26).
Educated people are argued to be more inclined to engage politically, reflecting
their greater economic and social resources, civic skills, and political sophistica-
tion (e.g., Martinez, 2005; Schussman & Soule, 2005). Also, movement recruit-
ers may prefer those with greater education because they are seen as having the
172 | Social-Scientific Perspectives on Collective Action

173
determination and willingness to see a task through to completion (Krueger &

Malečková, 2003, p.122).
Differential recruitment may also reflect “biographical availability”— “the
absence of personal constraints that may increase the costs and risks of move-
ment participation” (McAdam, 1986, p. 70). By this argument, young people—
who may have more time, are less invested in careers, and have fewer obligations
to family—are more available for engagement than older people (Beyerlein &
Hipp, 2006, p. 301). Similar arguments can be made about older people (relative
to the middle-aged), so age could instead have a U-shaped relationship to activ-
ism (Nepstad & Smith, 1999, p. 28). Single people may have more time available
because they have no obligation or commitment to a spouse that could compete
with their loyalty to a movement, and they face no sanctions from a spouse for
engaging in protest. People without children may have more time and energy to
participate and have less at stake if they were to be arrested or injured. And people
who are not employed or employed only part-time may have more free time and
less to lose financially than those working full-time (Beyerlein & Hipp, 2006,
p. 301).
Although most studies find mixed support for these arguments (Beyerlein &
Hipp, 2006; Martinez, 2005; Nepstad & Smith, 1999), qualitative/observational
accounts of Islamist movements in the Muslim world report that participants
are young, single, students or recent graduates, unemployed or underemployed,
and highly educated (e.g., Wickham, 2002; Woltering, 2002; but see Sageman,
2004). We test for effects of differential recruitment (education, employment
status, age, marital status, and having children) in settings where the risks and
costs of political engagement may be weighed carefully in view of the possible
consequences.
Relative Deprivation and Grievance

Although grievance is rarely seen by contemporary social movement scholars
as the sole basis of political mobilization, we consider it here as one of several
factors. Relative deprivation—the perception that one is underrewarded relative
to others with comparable skills, abilities, or education—was a primary fac-
tor identified by Gurr (1970) in an early effort to explain rebellion. Meyer and
Tarrow (1998, p. 13) note that, although access to higher education is increasing
in Western democracies, this has not always meant an increase in the avail-
ability of jobs in the professions, resulting in many more educated people who
are underemployed, potentially experiencing relative deprivation and who are
possibly open to political engagement. “Blocked aspirations” is the most com-
mon source of grievance mentioned by observers of the Muslim world (Ayubi,
1991, p. 162; Ghadbian, 2000, p. 51; Sageman, 2004, p. 95; Woltering, 2002,
p. 139). By this argument, it is not education that causes university graduates to
become activists, but the failure of state and society to reward their efforts with
good-paying jobs. For example, in Egypt during the 1960s, President Gamal
Abdul Nasser guaranteed a white-collar state job to every university graduate.
During the years of President Hosni Mubarak (1981–2011), many graduates
The Roots Of Political Activism In Six Muslim-majority Nations | 173

174
felt that this “contract” was still in effect, despite the government’s claim that
it could no longer afford to provide a good job to every graduate (Slackman,
2008; Wickham, 2002, pp. 24–35). We expect highly educated people with low
incomes to be especially open to political engagement, and test this with the
interaction term education × (low) income. We also expect that low income,
which makes survival precarious and increases vulnerability to rising food
prices, unemployment, and so on, is a source of grievance and hence creates a
greater willingness to engage politically.
Social Dislocation
Social dislocation may also affect openness to engagement. In the Muslim world,
rapid urbanization has meant that large numbers of newcomers to the city face
challenges to traditional ways of life and the breakdown of family, kin, and friend-
ship networks. Megacities such as Cairo, Algiers, and Tehran have seen massive
influxes of migrants from rural areas, making it difficult or impossible to employ
and house them (Wiktorowicz, 2004, pp. 6–7). It is widely agreed by observers of
the Muslim world that city dwellers are especially likely to be mobilized (Dessouki,
1982, p. 23), and we will test this expectation.
Interest in Politics
We also test Schussman and Soule’s (2005, p. 1086) argument that having a
strong interest in politics motivates individuals to become politically engaged.
These authors see having knowledge about politics and a sense of political efficacy
as going hand in hand with interest, but the WVS data for these countries contain
no measures of these.
Gendered Expectations
Gender role socialization and gendered social structures may also affect the like-
lihood of engaging in political protest. Early studies in the United States found
that women were less likely than men to engage in protest (Schlozman, Burns, &
Verba, 1994; Verba, Schlozman, Brady, & Nei, 1995), but more recent studies find
that gender has no effect on protest behavior (Beyerlein & Hipp, 2006, p. 310;
Martinez, 2005, p. 140). Although some women in predominantly Muslim coun-
tries are active in Islamist, labor, and democracy movements, and in struggles for
greater social, economic, and political equality for women (Moghadam, 2002), the
historical and cultural importance ascribed to domestic or private-sphere roles for
women and to employment or to public-sphere positions for men should mean
that men are more open to political engagement than women. Most studies of
participants in Islamist movements find them to be overwhelmingly male (e.g.,
Ibrahim, 1980; Sageman, 2004), although a recent study of Muslim activism in
the Netherlands and New York finds no effect of gender (Klandermans, van der
Toorn, & van Stekelenburg, 2008).

175
Embeddedness
McAdam (1982, 1986, 1988a, 1988b) argues that biographical availability is a nec-
essary but not sufficient cause of activism. Individuals who are available must also
be “pulled into” political engagement through their connections to “micromobili-
zation contexts”—preexisting social networks of activists. Schussman and Soule
(2006) find that the key factor in individual protest participation is being asked
to protest. In the Muslim world, preexisting networks of activists can sometimes
be found in mosques (Wiktorowicz, 2004, p. 10). Mosques in many Muslim-
majority countries are also hubs of vast networks of Islamic social welfare services
(Esposito, 2003; Wickham, 2002, 2004)—a form of nonelectoral activism not
captured in our analyses.
Although the WVSs of Muslim-majority countries contain no information on
organizational memberships, on whether respondents were asked to protest by
someone already in a movement organization (which might have raised concerns
about the purposes of the survey), or on Internet usage, they do include questions
on how often respondents spend time with friends, with colleagues from work, and
with people at sports, cultural, or communal organizations. Such informal ties may
be even more important than formal organizational memberships in the contexts we
study. We expect that people who spend more time socializing with others will have
more opportunity to form ties in networks where political talk and proto-political
activities (Johnston, 2005) are common, and will be more politically engaged.
Religion
Because some scholars have argued that Islam is a major basis of mobilization
in Muslim-majority countries (Beck, 2009; Fuller, 2003), we test whether three
dimensions of religion are related to nonelectoral political engagement. One
dimension that we consider is orthodoxy/modernism, which is defined as the
extent to which Muslims believe that the Quran and other sacred texts of Islam
should be the sole basis of the legal system and the state, as opposed to deriving
law from secular, as well as religious, sources. The establishment of the shari’a or
Islamic law, which is based on sacred Islamic texts, is the fulfillment of the ortho-
dox belief that it is the responsibility of the community to uphold timeless divine
law (Davis & Robinson, 2006).
We also test whether two other dimensions of religion affect political engage-
ment: mosque attendance and religiosity. Frequent mosque attendance may signal
greater devotion to the faith. Moreover, frequent attendees have greater exposure
to imams, sacred texts, and fellow worshipers who may strengthen their religious
beliefs and identity. As we discussed earlier, frequent attendance at the mosque
may also embed worshippers in a preexisting network of believers that facilitates
individual political activism and mobilization by political movements (McAdam
1982, 1988a). We note, however, that, although for men mosque attendance—
at least at Friday afternoon prayers—is a strong expectation (but not a pillar) of
Islam, this is not as strong an expectation for women. With an interaction term,

176
mosque attendance × gender, we test whether frequent attendance matters less

to the political engagement of women than it does for men. We also test whether
those who profess to be religious are more engaged politically.
To summarize our arguments thus far, we expect people who are male, young,
single, city dwelling, highly educated, poorly paid (and especially both highly edu-
cated and poorly paid), not employed or working part-time rather than full-time,
students, those who are well connected to others, those who favor establishment
of the shari’a as the sole law of the land, who attend the mosque frequently, who
identify as religious, and who are strongly interested in politics to be most willing
to take nonelectoral political action.
Contextual Bases of Muslim Political Engagement

Economic and political characteristics of Muslim-majority countries may also
spur political engagement. Area scholars have located the country-level roots of
Muslim political activism in poverty, inequality, foreign economic domination,
state repression, and human rights abuses against opposition groups (Anderson,
1997; Esposito, 2003; Ghadbian, 2000; Hafez, 2003). Although we are limited in
our analyses at the country level to six nations, we can explore whether a coun-
try’s political opportunity structure (openness of its political system, absence of
state repression) and economic characteristics (low standards of living, higher
inequality, greater foreign investment in the economy) are associated with non-
electoral activism.
Poverty
Poverty is believed to be a key source of grievance in the Muslim world. Most
research on the effects of poverty or low standards of living—and of country-level
variables more generally—has focused on whether these spark transnational ter-
rorism. Of course, terrorism may have very different causes from the nonelectoral
engagement we study here, but such studies are the only clue in the literature
regarding the effects of country-level variables on political engagement. The find-
ings on standards of living have been mixed, with some studies reporting that
poor living standards are associated positively with terrorism (Li, 2005, p. 287;
Li & Schaub, 2004, p. 254), others that these are unrelated (Abadie, 2006, p. 55;
Krueger & Malečková, 2003, p. 142), and another that low standards of living are
related negatively to terrorism (Muller, 1985). Beck (2009, p. 351) finds no consis-
tent effect of gross domestic product (GDP) per capita on the number of political
and militant organizations in a country.
Inglehart and Catterberg (2002) offer another approach to the effect of standard
of living on political engagement. They argue that higher standards of living in the
advanced democracies allow citizens the luxury of focusing less on elite-directed
actions (voting) in support of their material needs and more on elite-challenging
(petitioning, demonstrating) actions in support of noneconomic concerns (e.g.,
the environment, globalization). Inglehart and Catterberg find that, among the

177
established democracies, GDP per capita is associated positively with nonelectoral

political engagement, to the point that petitions, demonstrations, and boycotts
“are no longer unconventional, but have become more or less normal actions
for a large part of the citizenry of post-industrial societies” (p. 301). Here, we can
explore whether higher standards of living are associated with greater nonelec-
toral engagement, as Inglehart and Catterberg’s argument suggests, or with less,
as our grievance-based argument suggests.
Inequality
Economic inequality may be another source of grievance leading to greater polit-
ical involvement. Findings on the effect of inequality are mixed. Muller (1985,
p. 60) finds that inequality increases deaths from political violence, whereas three
more recent studies (Abadie, 2006, p. 53; Li, 2005, p. 288; Li & Schaub, 2004,
p. 247) find no effect of inequality on terrorism. Here, we test the hypothesis that
economic inequality is related positively to the level of nonelectoral engagement.
Foreign Economic Domination

Another macro-level economic variable that could affect political engagement is
foreign penetration of the national economy. On the one hand, economic global-
ization, such as foreign direct investment (FDI), could generate a backlash in the
form of protest or terrorism (Hoffmann, 2002). Foreign corporations in a coun-
try may also provide targets for political action or violence (Li & Schaub, 2004,
p. 236). On the other hand, some studies show that foreign investment is linked
to higher standards of living (e.g., de Soysa & Oneal, 1999), which may (for the
reasons given earlier) reduce protest. Findings on the effect of foreign investment
are mixed. Li and Schaub (2004, p. 244) find that FDI is unrelated to terrorism,
whereas Robison, Crenshaw, and Jenkins (2006, p. 2022) find that FDI/GDP actu-
ally reduces Islamist—and especially Leftist—terrorist attacks. Because Marxists,
nationalists, and Islamists have objected to the extent to which their countries’
economies are dominated by foreign interests (Hoffmann, 2002, p. 112), we test
the hypothesis that FDI/GDP increases willingness to engage politically.
Political Opportunity
The openness of the political system to dissent and the repressiveness of the state
may also be factors in citizen nonelectoral activism. The political process model
(McAdam, 1982; McAdam, McCarthy, & Zald, 1996; Tarrow, 1998) posits that open
and inclusive political systems that allow citizens to voice their opposition and vote
for candidates of their choice should present greater opportunities for mobiliza-
tion than more restrictive systems. Closed political systems and the use of impris-
onment, torture, and murder against government opposition should tamp down
political engagement. Meyer and Tarrow (1998) argue that advanced democracies,
such as the United States, are becoming increasingly “social movement societies,”
where the openness of the political system means that protest is increasing to the
point that it is not only commonplace but also is even institutionalized. This is a

178
similar conclusion to that of Inglehart and Catterberg (2002) for countries with
high living standards (see Soule and Earl [2005] for a test of the “social move-
ment society” argument). A straightforward extension of this logic would be to
expect that the more open and inclusive the political system, the more citizens will
engage in nonelectoral protest.
A more complex argument, also building on the political process model, pos-
its an inverted U-shaped relationship between repression/exclusion and protest.
Muller (1985, p. 48) argues that when political systems are open and there is no
repression, it is easy for dissident groups to organize electorally, they are optimistic
about their chances of success, and there are plenty of options, including electoral
avenues, open to them, so they will be less likely to engage in nonelectoral protest.
When political systems are closed and repression is high, it is difficult for dissident
organizations to mobilize, they are pessimistic about their chances for success, and
the risks/costs of political protest are too high. Muller argues that it is at intermedi-
ate levels of openness and repression that nonelectoral protest is high because the
regime structure is not so restrictive that dissident groups cannot organize and the
cost/risk of collective action is not prohibitive, but opportunities for electoral partici-
pation are limited. Because this argument can be tested by including both an addi-
tive and a quadratic term for exclusion/repression (squaring the original variable,
less its mean), we can assess which of the applications of the political process model
best fits the six countries we consider here—the expectation of a linear decrease in
nonelectoral protest with exclusion/repression or an increase followed by a decline.
Research on the effects of closed political systems and/or repression on pro-
test/terrorism has yielded mixed results. Muller (1985, p. 60) and Abadie (2006, p.
55) find an inverted U-shaped relationship, where middle levels of repression are
associated with the most terrorism. Similarly, Crenshaw and Robison (2010) find
that antigovernment protests rise with democratization, “but subsequently reach
a tipping point in which they soon begin to decline, likely because citizens are
granted satisfying and co-opting levels of political access/participation and civil
liberties” (pp. 244–245). Testas (2004, p. 263), however, found a U-shaped rela-
tionship in which middle levels of repression have the least terrorism. Other stud-
ies find a negative relationship (Opp & Roehl, 1990) or no relationship (Krueger &
Malečková, 2003, p. 139) between repression and activism, whereas one of the few
studies of protest in authoritarian states found that repression under some cir-
cumstances decreases and under other circumstances increases collective action
(Osa & Corduneanu-Huci, 2003, p. 622). Still other studies uncovered a posi-
tive relationship between repression and protest (Francisco, 2005; Hafez, 2003;
Kurzman, 2001; Li, 2005; Moaddel, 2002). We note that the relationship between
repression and protest is undoubtedly reciprocal to some extent (Carey, 2006)
and that there is sound theory and research that focus on repression as a response
to citizen mobilization (e.g., Ayoub, 2010; Davenport, Soule, & Armstrong, 2011;
Earl, 2003; Earl, Soule, & McCarthy, 2003). Unfortunately, with only six countries
and several country-level variables to consider, we are unable to model reciprocal
effects.

179
The Setting: Six Muslim-Majority Nations

There are roughly 1.6 billion Muslims in the world and more than 50 coun-
tries in which Islam is the predominant faith. The six countries examined here
are among the most populous and influential Muslim nations. They include
Indonesia, with the world’s largest Muslim population; Pakistan with the sec-
ond largest; and Bangladesh, Egypt, and Algeria with the fourth, fifth (and
largest in the Arab world), and ninth largest Muslim populations, respectively.
Adding Jordan, just under half of the world’s Muslims live in these six coun-
tries (Central Intelligence Agency, 2004; SBS World Guide, 2003, p. 644).
Although we would have liked also to include Saudi Arabia and Iran, both of
which are highly influential countries in the Muslim world, the questionnaires
for these countries did not include all the items we use to measure nonelectoral
political engagement.
The six nations in our study are overwhelmingly Muslim, ranging from
88% in Bangladesh to 99% in Algeria (see Appendix). They generally apply
the shari’a only to family matters (e.g., divorce, marriage, child custody),
whereas criminal and civil law are based on non-Islamic legal codes (SBS
World Guide, 2003).
Standards of living in these countries range from low to moderate. The United
Nations (2001) Human Development Index (HDI) ranges from a low of .47 in
Bangladesh (the same as Haiti, the poorest country in the western hemisphere)
to a high of .71 in Jordan (roughly the same as Peru). The Gini coefficient, an indi-
cator of inequality that ranges from 0 to 100, with high scores indicating more
inequality, varies from 31.8 in Bangladesh to 36.4 in Jordan, higher than the post-
Communist countries of eastern Europe or the social democracies of Scandinavia,
but lower than most countries of the Americas or Africa (World Bank, 2004,
pp. 60–62). FDI as a percentage of GDP varies from 0.1% in Bangladesh to 2.1%
in Indonesia—all somewhat low relative to other low-or middle-income countries
(World Bank, 2004, pp. 306–308) (see Appendix).
Political democracies do not exist or are compromised in these countries.
Freedom House (2001, Table 6.1), which rates political rights (openness of elec-
tions, universality of voting rights, and so on) worldwide, reports that political
exclusion varies from moderate in countries such as Bangladesh and Indonesia
(3 points on a 1-to 7-point scale, where 7 points represents few political rights),
through slightly higher in Jordan (4 points), to quite high (6 points) in Algeria,
Egypt, and Pakistan. Repression and human rights abuses, such as political
imprisonment, torture, or murder of dissidents, are also present. The Political
Terror Scale (Gibney, Cornett, & Wood, 2009) ranks countries from 1 to 5 points,
with 5 points indicating countries “where levels of terror are population-wide and
decision makers do not limit the means by which they pursue private or ideo-
logical goals” (Hafner-Burton & Tsutsui, 2005, p. 1392). On this scale, Jordan is
rated 2.5 points; Bangladesh and Egypt, 3 points; Pakistan, 3.5 points; Indonesia,
4 points; and Algeria, 4.5 points (Appendix).

180
Table 6.1 Means on Items Measuring Nonelectoral Political Engagement and the Six-Item Index of Nonelectoral
Political Engagement, Muslims Only, Separately by Country and Pooled Sample
Form of nonelectoral political engagement
Discussing Signing Joining Joining legal Joining Taking Index of

politics petition boycott demonstration illegal over nonelectoral
strike building political
Country engagement n
Algeria 1.82 1.59 1.46 1.70 1.25 1.07 8.83 1,034

Bangladesh 1.89 1.71 1.52 1.54 1.31 1.08 9.12 1,092
Egypt 1.65 1.77 2.04 1.20 1.03 1.07 8.75 2,831
Indonesia 1.79 1.32 1.26 1.49 1.10 1.18 8.05 769
Jordan 1.57 1.25 1.11 1.13 1.06 1.00 7.10 1,136
Pakistan 1.66 1.34 1.16 1.20 1.03 1.01 7.41 1,750
Pooled 1.76 1.47 1.39 1.40 1.12 1.09 8.21 8,177
note: Missing values were not imputed. n Values for each country and the pooled sample are for the index of nonelectoral
political engagement. n Values for each form of protest range from 1,136 to 1,255 in Algeria, 1,195 to 1,340 in Bangladesh, 2,831
(no missing values) in Egypt, 820 to 910 in Indonesia, 1,147 to 1,161 in Jordan, 1,863 to 1,967 in Pakistan, and 8,744 to 9,337 for
the pooled sample.
source: World Values Survey (2000–2004).
181
Data and Methods

Data
The data analyzed here were gathered in 2000 through 2002 from the fourth wave
(World Values Survey 1999-2004) of the WVS, which included a unique battery of
questions on nonelectoral activism, several of which were not asked again in more
recent waves of the WVS. The countries included Algeria (n = 1,282, surveyed in
2002), Bangladesh (n = 1,499, surveyed in 2000), Egypt (n = 3,000, surveyed in
2001), Indonesia (n = 1,004, surveyed in 2001), Jordan (n = 1,233, surveyed in 2001),
and Pakistan (n = 2,000, surveyed 2002). The surveys of other Muslim-majority
countries conducted through the WVS in this wave do not include either the ques-
tion on support for the shari’a or many of the questions on political engagement
and hence are not analyzed here. Analyses were limited to respondents age 18 years
and older who identify as Muslim. In Algeria, which is 99% Muslim, religion was
not asked; thus we include every respondent. In some of our analyses, we pooled
the samples for all six countries into a single sample (n = 8,177 without imputation
of missing values; n = 9,588 with imputation), weighting each sample proportional
to the size of the country’s Muslim population.
Measures
Dependent Variable: Nonelectoral Political Engagement
We measure nonelectoral political engagement in a way that is well suited to
authoritarian states. Discussing politics is how often the respondent discusses polit-
ical matters with friends: never (1), occasionally (2), or frequently (3). Although in
democratic countries, discussing politics may hardly seem to constitute political
activism, Johnston (2005, 2006) argues that, in authoritarian regimes, “talking
politics”—often taking place in mundane settings such as kitchens, coffee houses,
or book clubs—is the “ ‘smallest’ form of contention” (2006, p. 208) and cannot
be dismissed as “mere grumbling … because complaints in authoritarian sys-
tems carry consequences” (2005, p. 115) As with the other items we used to mea-
sure engagement, respondents were not asked if their discussion was directed
against the government, which might have raised concerns about the purposes
of the survey.
The next five items indicate whether respondents would never under any cir-
cumstances do (1), might do (2), or have done (3) the following (arranged roughly
by the risk/cost involved): signing a petition, joining in boycotts, attending law-
ful demonstrations, joining unofficial strikes, and occupying buildings or factories.
Most scholars of protest analyze the distinction between participating and not
participating in protest. Yet, because these five questions include additional infor-
mation on the willingness to engage, analyzing only actual engagement misses
important information. In authoritarian settings, where people may be reluctant
to take action, information on willingness to engage may capture the potential
for individual action, given the right circumstances. Beyerlein and Hipp (2006)

182
argue that mobilization is best seen as a two-stage process: willingness to engage

in protest and actually participating in it. Although it would be ideal to model this
process, as they do, with two variables, each of our items combined the two stages
in a single continuum from unwillingness to protest (“would never do”), through
willingness (“might do”), to actual participation (“have done”). In our regression
analyses, we combined these six items in an index of nonelectoral political engage-
ment. In each country, factor analyses of the items identified a predominant factor
with an eigenvalue greater than 2.50. The loadings of the items on this factor are
generally more than .50. Cronbach’s alpha, an indicator of reliability, ranges from
.68 in Jordan and Pakistan to .83 in Bangladesh. The index ranges from 6 to 18
points, with high scores indicating actual participation in a full range of activism;
middle scores, primarily willingness to engage in a narrower range of actions; and
low scores, unwillingness to engage in even less risky/costly actions.
Did the politically repressive climate in many of these countries affect the
candor with which respondents answered these questions? Although it may
have affected some respondents, Ross (2009) finds that repression, measured
in two ways, was unrelated to willingness to express support for democracy in 79
countries surveyed by the WVS. All interviews were conducted face-to-face in the
respondent’s home to minimize concerns about privacy.
Independent Variables
In our regression analyses, gender is coded as male (1) and female (0). Age is in
years. To capture the possibility that young and old people may be more involved
than middle-aged people, we include age2 (subtracting the mean before squaring).
Marital status is coded as single (1) versus all other marital statuses (0). Children is
the number of children the respondent has—from 0 to 8 or more. Urban is the
size of the respondent’s town, ranging from under 2,000 people (1) to 500,000 or
more (8). Education is coded in nine ordered categories ranging from no formal
education (1) to university degree (9). Household income before taxes, counting all
wages, salaries, pensions, and other income, is coded in approximate deciles in
each country (with 1 as the lowest decile and 10 as the highest). To test the blocked
aspirations thesis, we include the interaction of education × (low) income, subtract-
ing the means for each variable and reversing income before creating the interac-
tion term. Employment status is a dummy variable series identifying not employed,
employed part-time, and student, with employed full-time as the reference category.
We included three variables indicating network ties beyond attendance at the
mosque: spend time with friends, spend time with colleagues, and spend time with
people at organizations are measured by whether the respondent reports doing so
“never” (1), “only a few times a year” (2), “once or twice a month” (3), “weekly”
(4), or “every day” (5).1 Interest in politics is respondents’ profession of being “not
at all interested” (1), “not very interested” (2), “somewhat interested” (3) or “very
interested” (4) in politics.
Shari’a is based on responses to the following: “I would like to know your views
about a good government. Which of these traits is (5) very important, (4) important,

183
(3) somewhat important, (2) least important, or (1) not important for a good govern-
ment to have?” It should implement only the laws of the shari’a.
Mosque attendance refers to how often, apart from weddings and funerals, the
respondent attends religious services, ranging from never (1) to more than once
a week (7).2 Following Brooks (2002), we test with an interaction term, shari’a ×
attendance, whether exposure to imams, sacred texts, and networks of fellow
worshipers intensifies the effects of orthodox cosmology (see also Starks and
Robinson [2005]). To reduce multicollinearity, we center each variable in the
interaction term on its country mean (or the pooled mean for analyses of the
pooled sample) before creating the interaction term (Aiken & West, 1991, p. 35).
Self-professed religiosity is based on responses to the following: Independently of
whether you go to the mosque, would you say that you are “not a religious person”
(0) or “a religious person” (1). Correlations among the religion variables raise no
concerns about multicollinearity: the highest is .277 between mosque attendance
and religiosity in Algeria.
Country-Level Variables
In analyses of the pooled sample of the six countries, we tested for the effect of
several country-level variables: The United Nation’s (2001) HDI measures stan-
dard of living, incorporating three dimensions: income, education, and health.
Economic inequality is measured using the Gini coefficient in the closest year pre-
ceding the survey (World Bank, 2004, pp. 60–62). We include, as a measure of
foreign economic investment, gross FDI as a percentage of GDP in 2002 (the
closest year available to the survey years).3
Finally, following Beck (2009), we included measures of political exclusion and
repression. Political exclusion is the country’s rating on Freedom House’s (2001)
Political Rights Index, based on a checklist of 10 political rights.4 Although the
Freedom House ratings have been criticized for tending to rank Muslim countries
unfairly as less free (Bollen & Paxton, 2000), because all our countries are pre-
dominantly Muslim, any such bias is unlikely to affect our comparisons among
these countries. Giannone (2010) argues that a neoliberal bent, emphasizing free-
dom from government regulation and deemphasizing the value of equality, colors
the Freedom House ratings. Although this may be true for Freedom House’s Civil
Rights Index, we do not see bias of this sort in the 10 criteria used in the Freedom
House Political Rights Index.
We measure repression, another indicator of lack of political opportunity,
using the Political Terror Scale (Gibney et al., 2009), a widely used indicator of
human rights abuses by the state (Beck, 2009; Hafner-Burton & Tsutsui, 2005,
2007). The Political Terror Scale is based on separate content analyses of annual
reports published by Amnesty International and the US State Department,
which are combined into a single score. Political exclusion and repression are cor-
related only modestly (.202 in our pooled sample of countries). To capture a pos-
sible curvilinear relationship, we included both additive and quadratic (squared)
terms for exclusion and repression (after subtracting the means on each) in
some equations.

184
Methods
We used ordinary least squares (OLS) regression to estimate the effect of
individual-level characteristics on willingness to engage in nonelectoral political
activism within countries. To handle missing values, we used multiple imputation
modeling (mi in Stata 11), pooling 20 imputations. A sensitivity analysis compar-
ing the findings using multiple imputation with those using listwise deletion of
missing values without multiple imputation showed that a few more effects were
significant in the latter analyses. We report the more conservative analyses using
multiple imputation here (details available on request).
Our analysis of the pooled sample of all countries proceeded in two steps. First,
we estimated a country fixed-effect model that included individual-level variables
and dummy variables for countries. This model is more conservative in estimating
effects than the random-effects model when the number of cases (countries) is small
(Borenstein, Hedges, Higgins, & Rothstein, 2009, p. 84) and when one cannot be
certain that the key assumption of the random-effects model (that the random effect
error term is uncorrelated with covariates) holds (Lee, Kim, & Shim 2011, pp. 110–111).
Second, in estimating the effects of country- level variables, we cannot use
hierarchical linear modeling because of the small number of countries (Bryk &
Raudenbush, 1992, pp. 222–224). Yet, OLS regression produces biased estimates
of coefficients, especially of standard errors if key assumptions regarding normality,
heteroskedasticity, or large residuals are not met. Robust regression analysis helps
to address these issues (Brooks & Manza, 2006). Moreover, because our data are
sampled within six countries, we cannot assume that errors are uncorrelated across
observations. Thus, we clustered our data by country, which assumes that observa-
tions are independent across countries but not necessarily within them. We used
the robust cluster command in Stata 11 to produce robust (Huber-White) standard
errors. With our small number of countries (n = 6), these standard errors are con-
siderably larger—and thus more conservative (i.e., fewer effects are significant)—
than those produced by OLS regression. In addition, Stata takes into account the
fact that missing values are imputed by lowering the degrees of freedom associated
with the t-scores for independent variables, making the tests even more conserva-
tive. Because direction is predicted for all variables, we use one-tailed tests (positive
for male; age2 [expecting younger and older to be more engaged than middle-aged
people]; single; urban; education; education × (low) income; shari’a; mosque atten-
dance; shari’a × mosque attendance; religiosity; not employed; employed part-time;
student; time spent with friends, colleagues, and people in organizations; interest in
politics; economic inequality; and FDI/GDP; and negative for age, children, income,
HDI, political exclusion, repression, political exclusion2, and repression2 [expecting
low and high exclusion/repression to have less engagement than middle levels]).
Results
Nonelectoral Political Engagement
In Table 6.1, we show means (without imputation) by country on the six items
that make up our index of nonelectoral political engagement. The likelihood of

185
engaging in political activism decreases in most countries as one moves in the

table from left (discussing politics) to right (taking over a building), no doubt
reflecting the greater risk/cost. Engagement is highest in discussing politics; the
means in most countries indicate people report discussing politics occasionally (a
score of approximately 2), although this is lower in Jordan (1.57), Egypt (1.65), and
Pakistan (1.66). Johnston (2005) notes, “In the repressive context, oppositional
speech acts are a less demanding and less risky form of collective action, standing
for a part of the iceberg below the waterline but linked to those above it by net-
works of oppositional speech” (p. 117).
Fewer respondents are willing to engage or have engaged in signing a petition,
joining a boycott, or participating in a lawful demonstration; the means for these
are between “would never do this” (1) and “might do this” (2). Not surprisingly,
relatively few respondents are willing to engage or have engaged in an unlawful
strike, and fewer still in taking over a building or factory. The means for these two
extralegal forms of protest indicate that most respondents say they would never do
these (a score of approximately 1).
The most engaged Muslims are to befound in Bangladesh (9.12), followed
closely by Algeria (8.83) and Egypt (8.75). Still, even in the most engaged pop-
ulations, openness to nonelectoral activism is fairly low. For most respondents,
the level of engagement being registered, apart from discussing politics, is the
potential for involvement—willingness to take political action, rather than actually
doing so.
This pent-up possibility for action, however, may require only the proper
spark to explode. Table 6.2 shows the numbers of antigovernment demonstra-
tions involving more than 100 citizens, and of general strikes involving more than
1,000 citizens aimed at changing government policies or authority from 1995 to
2006 (the most recent year available). The data were gathered through Banks’
(2008) Cross-national Time-Series Data Archive and are based on reports in the
New York Times. As Davenport, Soule, and Armstrong (2011, p. 157) note, newspa-
per accounts of protest suffer from description bias (misreporting the “hard facts”
of an event, such as event size) and selection bias (failing to report every event),
both of which, but especially the latter, may be problematic in coverage of events
in other countries. Nonetheless, because the events being counted are fairly large
and are directed against the government, they may have attracted the attention of
reporters, especially in a Western world concerned with Islamist and other forms
of political mobilization by Muslims.
Bearing these limitations in mind, we can see there were three important epi-
sodes of protest, all of them during the mid to late 1990s. In Bangladesh in 1995
to 1996, a 2-year general strike and demonstrations by opposition forces claiming
rigged elections brought down the government of Prime Minister Khaleda Zia
(Cooper, 1999). In Pakistan, demonstrations against government corruption and
mismanagement in 1996 led the country’s president to depose Prime Minister
Benazir Bhutto (Burns, 1996). And protesters in Indonesia in 1998, spurred by
the Asian financial crisis, succeeded in removing President Suharto, ending his
32-year rule (Hafez, 2003). Although it is beyond the scope of our chapter to
explore these episodes further, they suggest that, although for most of this 12-year
period, citizens of these six countries engaged rarely in nonelectoral protest, given

Table 6.2 Antigovernment Demonstrations and General Strikes by Country, 1995 to 2006
Algeria Bangladesh Egypt Indonesia Jordan Pakistan
Demonstrations General Demonstrations General Demonstrations General Demonstrations General Demonstrations General Demonstrations General
Year strikes strikes strikes strikes strikes strikes
1995 1 0 4 4 0 0 3 0 1 0 2 1
1996 0 0 5 4 0 0 1 0 1 0 8 2
1997 0 0 0 0 0 0 2 0 0 0 0 0
1998 0 0 0 0 0 0 24 1 0 0 0 0
1999 0 0 0 2 0 0 3 0 0 0 4 1
2000 0 0 0 0 0 0 5 0 0 0 0 0
2001 2 0 1 1 0 0 7 0 0 0 2 0
2002 2 0 1 1 0 0 0 0 0 0 3 0
2003 0 0 2 2 1 0 0 0 0 0 0 0
2004 0 0 0 1 0 0 0 0 0 0 0 0
2005 — — 0 0 2 0 0 0 0 0 0 0
2006 0 0 4 1 0 0 0 0 0 0 4 0
Total 4 0 13 12 3 0 42 1 1 0 21 3
source: Banks, A. S. (2008). Banks’ cross-national time-series data archive: Variables and variable locations. Binghamton, NY: Databanks International.
187
the right set of grievances and opportunities, their potential, but unexpressed,
frustrations could erupt, as we saw in the Arab Spring activism of 2010 to 2012 in
Tunisia, Egypt, Libya, and other Muslim-majority nations.
To put into context our figures on nonelectoral political engagement from
the WVSs of Muslim countries, we show in Table 6.3 the means for the index of
nonelectoral political engagement for established democracies in the fourth wave
of the WVS. Although engagement in Spain (9.29) and Portugal (9.32), perhaps
because of their relatively recent authoritarian pasts, nearly overlaps that in the
most engaged Muslim countries, participation in most democracies is consider-
ably greater than that in our six Muslim countries. People in Sweden (12.15), the
United States (11.48), New Zealand (11.39), Greece (11.29), and France (11.27) are the
most engaged politically, lending support to the arguments of Meyer and Tarrow
Table 6.3 Means on Index of

Nonelectoral Political Engagement
for Established Democracies in the World
Values Survey, 1999 to 2004
Country Mean n
Australia 10.94 1,592

Austria 9.79 1,407
Belgium 10.81 1,710
Canada 10.65 1,826
Denmark 11.24 918
Finland 10.27 885
France 11.27 1,393
Germany 10.40 1,728
Greece 11.29 995
Iceland 10.83 923
Ireland 10.21 922
Israel 9.97 1,080
Italy 10.50 1,691
Japan 9.90 741
Luxembourg 10.44 997
Netherlands 11.02 997
New Zealand 11.39 812
Northern Ireland 9.87 879
Norway 11.10 1,103
Portugal 9.32 912
Spain 9.29 1,858
Sweden 12.15 938
Switzerland 10.07 1,017
United Kingdom 10.43 892
United States 11.48 1,106
note: Missing values not imputed.


188
(1998) that participation in nonelectoral protest is becoming commonplace in

advanced democracies.
The Social Roots of Nonelectoral Political Engagement

In Table 6.4, we show unstandardized coefficients in OLS regressions of the index
of nonelectoral political engagement on the religious and sociodemographic char-
acteristics of respondents in each country. Three variables have consistent effects
across countries: gender, education, and interest in politics. As expected, men are
Table 6.4 Unstandardized Coefficients in Ordinary Least-Squares Models Regressing

Nonelectoral Political Engagement on Social and Religious Characteristics, Muslims Only,
Separately by Country
Country
Variable Algeria Bangladesh Egypt Indonesia Jordan Pakistan
Male 0.341a 0.571a –0.182 0.344a 0.437a 0.707a

Age 0.011 –0.011 0.002 –0.016a 0.009 0.009
Age2 (×10) 0.004 0.008a 0.003 0.015a –0.003 0.002
Single –0.075 –0.035 –0.151 0.444 0.312a –0.141
Children –0.002 –0.054 –0.054a –0.007 –0.017 –0.034
Urban 0.170a –0.016 –0.004 0.041 –0.011 –0.025
Education 0.161a 0.089a –0.012 0.063a 0.094a 0.059a
Income –0.022 –0.032 –0.031 –0.036 0.029 0.022
Not employed –0.085 –0.489b –0.097 –0.289 0.091 –0.087
Employed part time 0.438a –0.609b –0.219 –0.172 0.154 0.075
Student 0.102 –0.345 0.149 0.044 –0.003 –0.059
Shari’a –0.092 0.203a –0.010 0.070 –0.033 –0.059
Mosque attendance 0.033 0.032 0.266a –0.083 0.011 0.014
Religiosity 0.050 –0.808 b –0.412 –0.141 –0.077 –0.109
Spend time, friends 0.244a 0.004 0.030 0.044 0.049 0.025
Spend time, colleagues 0.033 –0.054 0.009 0.024 0.039 0.013
Spend time, people at 0.092 0.237a –0.041 0.145a 0.051 0.125a
organizations
Interest in politics 1.047a 1.076a 0.020 0.731a 0.694a 0.844a
Shari’a × attendance 0.110a 0.043a
Education × (low) –0.021a
income
Constant 2.590 5.697 8.547 5.811 3.921 4.180
Adjusted R2 0.258 0.330 0.074 0.253 0.257 0.272
n 1,282 1,378 2,831 929 1,168 2,000
note: Multiple imputation modeling, pooling 20 imputations.

a
p < .05 (one-tailed tests because direction is predicted to be positive for male; age2; single; urban;
education; education × (low) income; support for the shari’a; mosque attendance; shari’a × mosque
attendance; religiosity; not employed; employed part-time; student; time spent with friends, colleagues,
and people in organizations; and interest in politics; and negative for age, children, and income).
b
p < .05 (two-tailed test for coefficients in the opposite direction from prediction).
Source: World Values Survey (2000–2004).

189
significantly more engaged politically than women in every country except Egypt.
The interaction of gender-by-mosque attendance has no effect in any country (not
shown in Table 6.4), indicating that attendance has similar religious importance
and/or consequences for women as it does for men.
The differential recruitment argument expects that highly educated people are
engaged more politically, reflecting their greater civic skills and knowledge of poli-
tics, and/or the interest of recruiters in attracting people who have demonstrated
they can apply themselves and carry out a task. We found that education is associ-
ated with greater engagement in every country except Egypt. Although blocked
aspirations was cited widely as a key source of grievance in the Muslim world, we
found that high education, coupled with low income, leads to exceptional engage-
ment in Pakistan only. Moreover, low income—contrary to accounts that those
who are poor are more likely to rebel—has no effect in any country.
Interest in politics is the third variable that has strong and consistent effects
across countries. Only in Egypt does such interest have no effect on political engage-
ment. Because interest in politics could, arguably, mediate between the other inde-
pendent variables and political engagement, we also estimated models excluding
this variable. Essentially, these yielded the same effects as those in Table 6.4, indicat-
ing this strong predictor of engagement is not rendering the effects of other inde-
pendent variables nonsignificant.
Biographical availability arguments expect that younger (and possibly older as
opposed to middle-aged) people, single people, those without children, students,
and those employed part-time or not employed (vs. full-time workers) would
have more spare time and fewer other obligations to prevent their participation.
Reflecting the equivocal findings of previous research testing these arguments, we
found only spotty support for them in our analyses of Muslim-majority countries.
In Bangladesh and Indonesia, younger and older people are, as expected, more
likely than middle-aged people to be politically active (as indicated by the positive
effect of age2 in both countries and by the negative effect of age in Bangladesh),
but the age variables have no effect in the other countries. Only in Jordan are sin-
gle people more likely to engage politically. Having children reduces engagement
only in Egypt. Part-time workers are more engaged than full-time workers only in
Algeria, whereas in Bangladesh they, along with people who are not employed, are
significantly less likely than full-time workers to be politically active.
Social dislocation arising from urbanization has been mentioned by some
scholars as fueling discontent and activism in Muslim countries. Yet, city dwellers
are exceptionally engaged only in Algeria.
We expected embeddedness in (preexisting) social networks to make people
more open to engagement. We found some support for this argument. Given the
centrality of the mosque in networks of political activism and in Islamist social wel-
fare networks, our finding that political engagement is heightened in Bangladesh,
Egypt, and Jordan by frequent attendance at the mosque, either alone or in con-
junction with support for making the shari’a the sole legal foundation, is consis-
tent with arguments on embeddedness. In four countries—Algeria, Bangladesh,
Indonesia, and Pakistan—frequent contact with friends and/or people in orga-
nizations is associated with higher engagement, lending further support to this

190
argument, although we lack direct measures of contact with organizations encour-

aging political activism, or of whether the respondent was asked by others to
become involved politically.
We found that religion is related to engagement in only three of the six coun-
tries. Support for establishing Islamic law—the principal end-goal of Islamists—
is associated with greater engagement in Bangladesh, and coupled with frequent
mosque attendance (shari’a × attendance), is associated with activism in Bangladesh
and Jordan. Frequent attendance at the mosque is associated very strongly with
political activism in Egypt. Yet, in Algeria, Indonesia, and Pakistan, none of the
three religion variables is significant. In Bangladesh, self-professed religiosity is
actually related negatively to activism (if we had predicted no direction and used
a two-tailed test of significance). Although, in the zero-order, mosque attendance
is related significantly to political engagement in Algeria (.153) and Pakistan (.231),
these effects disappear with controls.
Notable about the findings of Table 6.4 is how often Egypt is the exception to
the rule. Although being male, highly educated, and strongly interested in poli-
tics matter for political engagement in most countries, none of these is important
in Egypt. The adjusted R2 value for Egypt is also markedly lower than those for the
other countries.5 Frequent attendance at the mosque is the key variable affecting
willingness to engage politically in Egypt. The Muslim Brotherhood and other
Islamist groups in Egypt were highly successful in organizing their civic and
political activities around local mosques they had built and/or then control led
(Lia, 1998, pp. 33, 40–41, 54). Centering such political activities on local mosques
ensures that those who come to listen to services in the mosque are already
somewhat predisposed theologically to the message. Situating activities around
mosques lends respectability to the Islamist cause and makes local Islamist activ-
ities less subject to government repression because the Egyptian state, like others
in the Muslim world, often has been reluctant to intervene in religious activities
(Berman, 2003, p. 260; Munson, 2001, p. 502). Mosques in Egypt may thus have
served as a site of preexisting networks of Islamist activists, fostering the engage-
ment of frequent attendees (Esposito, 2003; Lia, 1998; Wickham, 2002, 2004).
Although the democracy protests in Egypt in early 2011 were organized initially by
secularists, some of the largest demonstrations occurred when people marched
from mosques to Tahrir Square after Friday afternoon prayers. Moreover, the
Muslim Brotherhood helped to organize services (garbage collection, bullhorns,
and weapons checks) for the demonstrations, and some local mosques were used
as field hospitals for wounded demonstrators (Kirkpatrick 2011).
In Egypt, willingness to engage politically, although among the highest in these
six countries (see Table 6.1), is unaffected by social characteristics that prior obser-
vational accounts have suggested are important: gender, youth, unemployment,
high education coupled with low income, and so on. Consistent with this finding,
the New York Times reported that, during the demonstrations that brought down
President Hosni Mubarak in early 2011, “[the] protests appeared to represent a
nearly universal cross section of the public” (Kirkpatrick, 201, p. A1). The most
distinguishing characteristic mentioned about those engaging in protest was their
youth (Kirkpatrick & Slackman, 2011). Although those who rebelled against the

191
Mubarak regime—and especially those who were mobilized via Facebook and
Twitter—may well have been exceptionally young, in a country where more than
62% of the population is younger than age 30 (Nawar, 2010), one would expect
that many of the protestors would be young.
In Table 6.5, we show the country fixed-effect model for the pooled sample of
six countries, in which samples were weighted proportionally to each country’s
Table 6.5 Country Fixed-Effect Model Explaining Nonelectoral Political

Engagement, Pooled Sample of Six Muslim-Majority Countries
Variable
Male 0.320a
Age –0.012a
Age2 (×10) 0.012a
Single –0.055
Children –0.038a
Urban 0.017
Education 0.057a
Income –0.033a
Not employed –0.353b
Employed part time –0.117
Student –0.059
Shari’a 0.060a
Mosque attendance 0.048a
Religiosity –0.201
Spend time, friends –0.009
Spend time, colleagues 0.002
Spend time, people at organizations 0.167a
Interest in politics 0.735a
Shari’a × attendance 0.072a
Education × (low) income 0.013a
Bangladesh 0.452a
Egypt 0.449a
Indonesia –0.563a
Jordan –1.309a
Pakistan –0.869a
Constant 6.169
Adjusted R2 0.259
n 9,588
note: Multiple imputation modeling, pooling 20 imputations.

a
p < .05 (one-tailed tests because direction is predicted to be positive for male; age2; single;
urban; education; education × (low) income; support for the shari’a; mosque attendance;
shari’a × mosque attendance; religiosity; not employed; employed part-time; student; time
spent with friends, colleagues, and people in organizations; and interest in politics; and
negative for age, children, and income). Two-tailed tests for country dummy variables.
b

192
population. Note that these are exploratory analyses, combining as they do samples
collected by different organizations, with somewhat different sampling designs,
in countries with different populations and different political and economic con-
texts. Nonetheless, the original researchers in these countries, gathering the data
for the WVSs, made every effort to make the samples, procedures, and measures
comparable. The pooled sample allowed us to explore the factors affecting non-
electoral political engagement in the Muslim world more broadly.
Some of the effects in Table 6.5 are consistent with expectations derived from
social movement theory and research. From the differential recruitment argu-
ment, we expected educated people to be more inclined to involvement—and
they are. Low income, often cited as a source of grievance, increases willingness
to engage politically. The blocked aspirations argument that highly educated but
poorly paid people are exceptionally open to engagement is supported by the
positive effect of education × (low) income. The positive effects of mosque atten-
dance (alone and in interaction with support for the shari’a) and time spent with
people in organizations are consistent with the argument that embeddedness
in networks is important for protest. Men are significantly more willing to take
action than women, consistent with gendered expectations. And as we found in
the within-country analyses, interest in politics has a strong effect on openness
to engagement.
Other effects that we see in Table 6.5 are not in accord with theory and research
on social movements. Biographical availability may affect engagement in that
younger and older people are more engaged than middle-aged people, and hav-
ing no children is associated with greater openness to engagement, but we also
found that single people, students, and part-time workers are not exceptionally
inclined to activism, whereas the unemployed are actually less inclined to involve-
ment than full-time workers. And contrary to the social dislocation argument,
city dwellers are no more open to engagement than people living in small towns
or rural areas.
Do Economic and Political Contexts Matter?

In Table 6.6, we show robust cluster regressions for country-level factors that are
argued to affect nonelectoral political engagement. Models were estimated with mul-
tiple imputation of missing values, pooling 20 imputations. All models controlled
for the individual-level variables in Tables 6.4 and 6.5 (not shown in Table 6.6). In
models 1 through 3, we introduce separately the HDI6, economic inequality (Gini),
and FDI/GDP. Models 1 and 2 show that, contrary to the findings of at least one
study (Li & Schaub, 2004), but consistent with many others (Abadie, 2006; Beck,
2009; Hafez, 2003; Krueger & Malečková, 2003), low standards of living and high
inequality are unrelated to activism. As we see in model 3, direct foreign investment
affects engagement, but in the opposite direction from what we expected under the
assumption that penetration of the economy by foreign interests is a source of griev-
ance spurring political activism.

193
Table 6.6 Robust Cluster Regressions Explaining Nonelectoral Political Engagement,

Pooled Sample of Six Muslim-Majority Countries
Variable Model 1 Model 2 Model 3 Model 4 Model 5
Human development index –0.526

Economic inequality (Gini) –0.123
Foreign direct investment (FDI/GDP) –0.630b
Political exclusion –0.364a
Political exclusion2 0.611b
Repression –.700a
Repression2 1.378
Constant 6.402a 10.152a 6.284a 6.506a 7.952a
Adjusted R2 0.208 0.211 0.240 .213 .238
n 9,588 9,588 9,588 9,588 9,598
note. Multiple imputation modeling, pooling 20 imputations. All models include controls
for male; age; age2; single; children; urban; education; income; education × (low) income; not
employed; employed part-time; student; shari’a; mosque attendance; shari’a × attendance;
religiosity; spend time with friends, colleagues, people at organizations; and interest in politics.
a
p < .05 (one-tailed tests because direction is predicted to be positive for economic inequality
and direct foreign investment (FDI/gross domestic product [GDP]), and negative for human
development index, political exclusion, political exclusion2, repression, and repression2).
b
We have two indicators of the political climate in which Muslims in these six
countries live: political exclusion and repression. We entered each measure, along
with a squared term, in models 4 and 5. Recall that we are testing two theories
derived from the political process model—one that closed political systems dis-
courage protest and the other that there is an inverted U-shaped relationship
where middle levels of exclusion/repression generate the most protest. For both
indicators of political climate, we found support for the first theory—that exclu-
sion (lack of political rights) or repression (human rights abuses against the oppo-
sition) squelch political activism—in the negative effect of the additive variables.
Yet, contrary to the second theory, we found, for political exclusion and nearly so
for repression (p = .065), a U-shaped relationship with engagement, as indicated
by the positive effect of the quadratic (squared) term for exclusion. This means
activism is highest at low and high levels of political exclusion; relatively open
political states have high levels of nonelectoral engagement. Political engagement
declines up to a point as states begin to limit political rights, but then activism
increases, perhaps in a backlash, as states become highly exclusionary. As we
noted in our previous discussion, prior research is mixed on the effect of exclu-
sion, but Testas (2004) found a U-shaped pattern between exclusion and terror-
ism. Given our small number of countries, we cannot present models in which all
the significant country-level variables are entered together.

194
Discussion and Conclusion

Through analyses of nationally representative surveys in the WVS, we explored
the individual-level and contextual bases of nonelectoral political engagement
among Muslims living in six predominantly Muslim nations. We used a mea-
sure of engagement sensitive to the often-authoritarian political contexts in
which citizens of these countries assess the costs and risks of activism, one
that takes into account a range of actions—from discussing politics to taking
over buildings—and a range of involvement from none, through willingness
to undertake a specific action, to taking the action. In these countries in 2000
through 2002, we found that the level of nonelectoral engagement is relatively
low. Many Muslims discuss politics with their friends—a risky action in itself
in some of these contexts (Johnston, 2005, 2006). Fewer are willing to go
beyond this to involve themselves in signing petitions, joining boycotts, or par-
ticipating in officially sanctioned demonstrations. Fewer still would consider
joining an illegal strike or taking over a building or factory. Moreover, apart
from discussing politics, much of the political engagement we found is the
potential for engagement—what respondents might do, not what they actually
have done.
Putting this engagement into context, we found that, in advanced democra-
cies, there is considerably greater involvement in nonelectoral political activism.
This supports Meyer and Tarrow’s (1998) argument, using the political process
model, that advanced democracies are becoming “social movement societies”
in which protest is commonplace and even institutionalized. We found further
corroboration of this argument among the predominantly Muslim countries in
that political inclusion and the absence of repression are associated positively
with nonelectoral engagement. Yet, we also found a curvilinear relationship in
which both poles of the political inclusion/exclusion continuum are associated
with greater nonelectoral engagement, whereas middle levels have the least
engagement.
Other country-level variables have unexpected effects on political engage-
ment. It has been argued that low standards of living (HDI) and high inequality
are sources of grievance that spur popular protest, but we found no such tenden-
cies in our analyses. Inglehart and Catterberg’s (2002) arguments about postma-
terialism would suggest that a high HDI is associated with greater nonelectoral
engagement, but most citizens of the countries considered here are not at the
point where economic/material concerns have become unimportant to them.
Another commonly cited source of grievance is high penetration of the country’s
economy by foreign interests. We found the reverse—that greater levels of FDI
are associated with lower levels of political engagement. Robison et al. (2006)
found that FDI/GDP reduces the incidence of Islamist and especially Leftist ter-
rorist attacks. de Soysa and Oneal (1999) identified one mechanism by which
this could occur, finding that foreign investment is not detrimental to growth
and, in fact, stimulates domestic investment. Although in our analyses economic
development (HDI) had no effect on protest, FDI/gross domestic product may

195
be capturing beneficial effects of foreign investment, such as job growth, not

measured by the HDI.
Drawing on social movement theory and research, we considered how demo-
graphic and socioeconomic characteristics on the individual level affect willing-
ness to engage in nonelectoral political activism. The biographical availability
argument posits that younger (and older) people, single people, the unemployed,
those working part-time and students (vs. full-time workers), and people without
children are more available for engagement. Consistent with the equivocal find-
ings of earlier research, we found that younger and older people are more willing
than middle-aged people to become engaged politically, as are people who have
no children, but beyond this none of the other characteristics had an effect in
the predicted direction. This is surprising because Beyerlein and Hipp (2006)
argued that biographical availability should affect willingness to engage politi-
cally more than actual engagement, and this is the distinction (between unwill-
ingness and willingness) that our measure of engagement captured primarily.
Nepstad and Smith (1999), who studied the effect of biographical availability on
high-risk/cost activism (as we studied here), also found mixed support for this
argument. They suggest that some of the people who engage in such activism,
despite their seeming biographical unavailability, may have “constraint manage-
ment” skills (p. 38) that allow them to juggle their responsibilities so they can fit
in political activism.
Under the expectation that relative deprivation, an important concept in early
accounts of rebellion (Gurr, 1970) and widely cited as a source of grievance for
people in the Muslim world, is a basis for nonelectoral political involvement, we
tested the blocked aspirations thesis that the highly educated but poorly paid are
exceptionally open to political activism. We found this to be true in Pakistan and
in the pooled sample. Another potential source of grievance—low income—is also
associated positively with engagement, although we found no parallel effect of
standard of living (HDI) at the country level. Education, which differential recruit-
ment arguments assume imparts political knowledge and civic skills, and is attrac-
tive to movement recruiters, affects engagement positively in most countries and
in the pooled sample.
We expected social dislocation experienced by city dwellers, some of whom
would have moved from small towns and rural areas, to increase nonelectoral
political engagement, but we found no tendency for this. Of course, because we
lack information on the origins of current urban residents, we are unable to ascer-
tain whether recent arrivals to the city are more open to activism than long-time
residents. It is also possible that rural residents face their own sources of dislo-
cation or grievance in the shift from subsistence to cash economies, droughts,
floods, famine, the outmigration of men, and so on.
We found some support, albeit indirect, for the argument that embedded-
ness in social networks is important in mobilization. Frequent attendees at
the mosque are more open to political engagement, and frequent attendees
who also want to establish Islamic law are exceptionally willing to engage.
Although contact with friends or with coworkers has no effect on willingness

196
to engage, contact with people at a sport, cultural, or communal organization

is associated with greater openness to activism. Of course, embeddedness can
also take a more technological turn, as in the 2011 uprising in Egypt, where
connections through social media such as Facebook and Twitter are credited
with having mobilized tens of thousands of people, most of whom had never
met in person.
Interest in politics has been argued to be associated with activism, and we
found strong evidence of this in most countries and in the pooled samples.
Possibly reflecting gendered institutions and expectations of women and men,
men are more likely than women to express willingness to engage in nonelectoral
activism.
We see our study as having important policy implications. Most research on
activism in the Muslim world since 9/11 has focused on the causes of transna-
tional terrorism, often with the aim of understanding how this can be stopped.
Although we had to draw on research on terrorism as a clue to what might
affect nonelectoral political engagement, protest should not be conflated with
terrorism and, given the often-autocratic states and endemic poverty and unem-
ployment in many predominantly Muslim countries, our purpose here is not
to suggest how domestic protest can be curbed. As Meyer and Tarrow (1998)
have argued and our analyses corroborate, protest is a commonplace—and we
would argue a vital—feature of open societies. Taking steps to make the gov-
ernment more open and to address poverty and joblessness may not reduce
nonelectoral protest, but are worthy ends in their own right. With these steps,
perhaps the focus of protest might change from toppling regimes, as evidenced
in Bangladesh and Pakistan in 1996, Indonesia in 1998, and Tunisia, Egypt,
Yemen, Libya, and other Muslim nations in 2010 to 2012, to the pursuit of more
reformist/incremental changes.
We note several important limitations of our study. First, some variables identi-
fied by social movement theorists and researchers are missing in the WVS data
for these countries. We would have liked to have direct measures of contact with
preexisting networks of activists, and information about whether anyone had
asked the respondent to participate in protests, although such questions might
have raised respondents’ concerns about the purposes of the interview in these
political contexts. We also lack information on respondents’ political knowledge
or sense of efficacy, which have been found to be important in earlier studies
(Schussman & Soule, 2006). A greater sense of efficacy may have been a criti-
cal factor in the spread of protest in early 2011 from Tunisia to Egypt to Bahrain,
Libya, and other Muslim countries in the Middle East, as citizens of countries with
repressive regimes, inspired by the success of movements in other countries in
overturning authoritarian rule, came to believe they might be able to do the same
in their countries.
Moreover, as useful as the indicators of nonelectoral activism in the WVS
data have been for our purposes, they suffer from an important omission. In
our earlier, qualitative research, we found that a primary strategy that Islamist

197
and other movements in the Muslim world have adopted for accomplishing
their goals is a patient, beneath-the-radar capture of civil society that we call
“bypassing the state” (Davis & Robinson, 2009, 2012). One institution at a
time, Islamist movements have built massive, grassroots networks, usually
centered on local mosques, of autonomous, religion-based social-service agen-
cies, hospitals and clinics, schools, charitable organizations, and businesses.
The Muslim Brotherhood in Egypt, for example, as early as the mid 1930s,
began a bottom-up, institution-building strategy that eventually became a
vast, decentralized “state within a state” (Esposito 2003, p. 71). Side-stepping
the state, rather than confronting it directly and immediately, allows Islamist
movements to accomplish their broad theological, cultural, and economic
agendas across the nation; addresses local needs not being met by the state;
empowers followers as they work toward the movements’ goals; and estab-
lishes a base of popular support from which to push their agendas in the arena
of party politics. This model has been adopted by many of the Brotherhood’s
branches and many independent Islamist movements throughout the world
(Esposito, 2003, p. 71; Fuller, 2003, p. 27; Ghadbian, 2000, p. 80; Marty
& Appleby, 1992, p. 150; Woltering, 2002, p. 1134), and may well have been
responsible for the Brotherhood’s victories in Egypt’s first open parliamen-
tary and presidential elections following Mubarak’s resignation. Working in
such networks is missed by the WVS questions; thus, they may understate the
extent of citizen nonelectoral engagement.
We conclude by noting that some of our findings—especially those regarding
the propensity for engagement among the young, males, and those with blocked
career aspirations—are in accord with prior first-hand, but often nonsystematic,
accounts of the Muslim world. Yet, our other findings differ from what many
observers would expect. Older people are also willing to engage politically, city
dwellers are no more open to engagement than small-town or rural residents,
and students and the unemployed are not exceptionally inclined to nonelectoral
engagement. And at the country level, low standards of living and inequality
do not foster political activism, and foreign involvement in the economy low-
ers rather than raises openness to taking political action. As we have shown, in
explaining nonelectoral political engagement, first-hand observations, although
important, are no substitute for systematic quantitative analyses of surveys of
ordinary Muslims.
Acknowledgment
The authors are listed alphabetically. A revised version of this chapter was
presented at the Workshop on Theoretical and Methodological Issues in the
Study of Values in Islamic Countries, Cairo, Egypt, May 16–18, 2010. We thank
Mansoor Moaddel and the participants in this conference for their advice and
support on this work; Jeff Dixon, Josh Klugman, and Brian Starks for their
comments; and Scott Long and Patricia McManus for their statistical advice.

198
Notes
1. We do not include the WVS variable on time spent with people at your mosque
because it has no effect beyond mosque attendance on political engagement.
2. As a result of a translation error, mosque attendance was not asked in Pakistan
(the question instead referred to frequency of prayer). The closest proxy to mosque
attendance in Pakistan is the question: How often do you spend time with people at
your mosque? The average correlation of this variable with mosque attendance in the
remaining five countries is .395. Although not ideal, we use this question in place of
mosque attendance in Pakistan.
3. Because no information on FDI is available for Algeria, we use multiple impu-
tation to estimate this (see Methods).
4. Fair elections of (a) head of state; (b) legislature; (c) fair electoral laws; (d) rep-
resentatives have real power; (e) right to organize into parties; (f) significant opposi-
tion vote; (g) freedom from domination by the military, foreign powers, economic
oligarchies, and so on; (h) self-determination of minority groups; (i) for monarchies,
consultation with the people and right to petition; and (j) no repression of ethnic
groups (Freedom House, 2001, pp. 650–651).
5. We examined variation in the individual items making up the index of non-
electoral engagement and the overall index itself and found that Egypt does not have
exceptionally low variation in these. Egypt had the lowest standard deviation among
these countries only on engaging in an illegal strike, and standard deviations for
other items and the index in Egypt are similar to those for other countries.
6. Analyses substituting log HDI yielded similar results (–.304, not significant).
References
Abadie, A. (2006). Poverty, political freedom, and the roots of terrorism. American
Economic Review, 96, 50–56.
Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and interpreting interactions.
Newbury Park, CA: Sage.
Anderson, L. (1997). Fulfilling prophecies: State policy and Islamist radicalism. In
J. L. Esposito (Ed.). Political Islam: Revolution, radicalism or reform (pp. 17–31)?
Boulder, CO: Lynne Rienner.
Ayoub, P. M. (2010). Repressing protest: Threat and weakness in the European con-
text, 1975–1989. Mobilization, 15, 465–488.
Ayubi, N. N. M. (1991). Political Islam: Religion and politics in the Arab world.
London: Routledge.
Banks, A. S. (2008). Banks’ cross-national time-series data archive: Variables and variable
locations. Binghamton, NY: Databanks International.
Beck, C. J. (2009). State building as a source of Islamic political organization.
Sociological Forum, 24, 337–356.
Berman, S. (2003). Islam, revolution, and civil society. Perspectives on Politics, 1,
257–272.

199
Beyerlein, K., & Hipp, J. R. (2006). A two-stage model for a two-stage process: How
biographical availability matters for social movement mobilization. Mobilization,
11, 299–320.
Bollen, K. A., & Paxton, P. (2000). Subjective measures of liberal democracy.
Comparative Political Studies, 33, 58–86.
Borenstein, M., Hedges, L. V., Higgens, J. P. T, & Rothstein, H. R. (2009). Introduction
to meta-analysis. New York: Wiley.
Brooks, C. (2002). Religious influence and the politics of family decline concern: Trends,
sources, and U.S political behavior. American Sociological Review, 67, 191–211.
Brooks, C., & Manza, J. (2006). Social policy responsiveness in developed democra-
cies. American Sociological Review, 71, 474–494.
Bryk, A. S., & Raudenbush, S. W. (1992). Hierarchical linear models for social and behav-
ioral research: Applications and data analysis methods. Newbury Park, CA: Sage
Publications.
Burns, J. F. (1996). With goats and gunfire, Pakistanis cheer Bhutto’s fall. New York
Times, November 6.
Carey, S. C. (2006). The dynamic relationship between protest and repression.
Political Research Quarterly, 59, 1–11.
Central Intelligence Agency. (2004). The CIA world factbook [On-line]. Available:
https:// w ww.cia.gov/ l ibrary/ p ublications/ t he- w orld- f actbook/ i ndex.html.
Retrieved February 21, 2011.
Cooper, K. J. (1999). General strikes cripple Bangladesh economy. Washington Post,
March 4.
Crenshaw, E., & Robison, K. (2010). Political violence as an object of study: The need
for taxonomic clarity. In K. T. Leicht & J. C. Jenkins (Eds.). Handbook of poli-
tics: State and society in global perspective (pp. 235–246). New York: Springer.
Davenport, C., Soule, S. A., & Armstrong, D. A., II. (2011). Protesting while black?
Differential policing of American activism, 1960 to 1990. American Sociological
Review, 76, 152–178.
Davis, N. J., & Robinson, R. V. (2006). The egalitarian face of Islamic ortho-
doxy: Support for Islamic law and economic justice in seven Muslim-majority
nations. American Sociological Review, 71, 167–190.
———. (2009). Overcoming movement obstacles by the religiously orthodox: The
Muslim Brotherhood in Egypt, Shas in Israel, Comunione e Liberazione in Italy
and the Salvation Army in the United States. American Journal of Sociology, 114,
1302–1349.
———. (2012). Claiming society for God: Religious movements and social welfare in
Egypt, Israel, Italy, and the United States. Bloomington: Indiana University Press.
de Soysa, I., & Oneal, J. R. (1999). Boon or bane? Reassessing the productivity of for-
eign direct investment. American Sociological Review, 64, 766–782.
Dessouki, A. E. H. (1982). The Islamic resurgence: Sources, dynamics, and implica-
tions. In A. E. H. Dessouki (Ed.). Islamic resurgence in the Arab world (pp. 3–31).
New York: Praeger.
Earl, J. (2003). Tanks, tear gas, and taxes: Toward a theory of movement repression.
Sociological Theory, 21, 44–68.

200
Earl, J., Soule, S. A., & McCarthy, J. D. (2003). Protest under fire? Explaining the
policing of protest. American Sociological Review, 68, 561–606.
Esposito, J. L. (2003). Islam and civil society. In J. L. Esposito & F. Burgat (Eds.).
Modernizing Islam: Religion in the public sphere in the Middle East and Europe
(pp. 69–98). New Brunswick, NJ: Rutgers University Press.
Francisco, R. A. (2005). The dictator’s dilemma. In C. Davenport, H. Johnston, &
C. Mueller (Eds.). Repression and mobilization (pp. 58–81). Minneapolis: University
of Minnesota Press.
Freedom House. (2001). Freedom in the world, 2000–2001. New York: Transaction Publishers.
Fuller, G. (2003). The future of political Islam. London: Palgrave MacMillan.
Ghadbian, N. (2000). Political Islam and violence. New Political Science, 22, 77–88.
Giannone, D. (2010). Political and ideological aspects in the measurement of democ-
racy: The Freedom House case. Democratization, 17, 68–97.
Gibney, M., Cornett, L., & Wood, R. (2009). Political Terror Scale, 1976–2009 [On-line].
Available: http://www.politicalterrorscale.org/. Retrieved February 11, 2011.
Gurr, T. (1970). Why men rebel. Princeton, NJ: Princeton University Press.
Hafez, M. (2003). Why Muslims rebel: Repression and resistance in the Islamic world.
Boulder, CO: Lynne Rienner.
Hafner-Burton, E. M., & Tsutsui, K. (2005). Human rights in a globalizing world: The
paradox of empty promises. American Journal of Sociology, 110, 1373–1411.
———. (2007). Justice lost! The failure of International Human Rights Law to matter
where needed most. Journal of Peace Research, 44, 407–425.
Hoffman, S. (2002). Clash of globalizations. Foreign Affairs, 81, 104–115.
Ibrahim, S. E. (1980) Anatomy of Egypt’s militant Islamic groups. International
Inglehart, R., & Catterberg, G. (2002). Trends in political action: The development
trend and the post- honeymoon decline. International Journal of Comparative
Sociology, 43, 300–316.
Johnston, H. (2005). Talking the walk: Speech acts and resistance in authoritarian
regimes. In C. Davenport, H. Johnston, & C. Mueller (Eds.). Repression and rebel-
lion (pp. 108–137). Minneapolis: University of Minnesota Press.
———. (2006). “Let’s get small”: The dynamics of (small) contention in repressive
states. Mobilization, 11, 195–212.
Kirkpatrick, D. D. (2011). Mubarak’s grip on power is shaken. New York Times, January
31, p. A1.
Kirkpatrick, D. D., & Slackman, M. (2011). Egyptian youths drive the revolt against
Mubarak. New York Times, January 26, p. A1.
Klandermans, B., van der Toorn, J., & van Stekelenburg, J. (2008). Embeddedness
and identity: How immigrants turn grievances into action. American Sociological
Review, 73, 992–1012.
Krueger, A. B., & Malečková, J. (2003). Education, poverty and terrorism: Is there a
causal connection? Journal of Economic Perspectives, 17, 119–144.
Kurzman, C. (2001). Critics within: Islamic scholars’ protests against the Islamic
state in Iran. International Journal of Politics, 15, 341–359.

201
———. (2004). Social movement theory and Islamic studies. In Q. Wiktorowicz

(Ed.). Islamic activism: A social movement theory approach (pp. 289– 303).
Lee, C.S., Kim, Y.M., & Shim, J.M. (2011). The limit of equality projects: Public-sec-
tor expansion, sectoral conflicts, and income inequality in postindustrial nations.
American Sociological Review, 76, 100–124.
Li, Q. (2005). Does democracy promote or reduce transnational terrorist incidents?
Journal of Conflict Resolution, 49, 278–297.
Li, Q., & Schaub, D. (2004). Economic globalization and transnational ter-
rorism: A pooled time-series analysis. Journal of Conflict Resolution, 48,
230–2 58.
Lia, B. (1998). The Society of the Muslim Brothers in Egypt: The rise of an Islamic mass
movement, 1928–1942. Beirut: Ithaca Press.
Martinez, L. M. (2005). Yes we can: Latino participation in unconventional politics.
Social Forces, 84, 135–155.
Marty, M. E., & Appleby, R. S. (1992). The glory and the power: The fundamentalist chal-
lenge to the modern world. Boston, MA: Beacon Press.
McAdam, D. (1982). Political process and the development of black insurgency, 1930–1970.
———. (1986). Recruitment to high-risk activism: The case of Freedom Summer.
American Journal of Sociology, 92, 64–90.
———. (1988a). Freedom summer. New York: Oxford University Press.
———. (1988b). Micromobilization contexts and recruitment to activism. International
Social Movement Research, 1, 125–154.
McAdam, D., McCarthy, J. D., & Zald, M. N. (1996). Opportunities, mobilizing
structures, and framing processes: Toward a synthetic, comparative perspective
on social movements. In D. McAdam, J. D. McCarthy, and M. N. Zald (Eds.).
Comparative perspectives on social movements (pp. 1–20). New York: Cambridge
University Press.
Meyer, D. S., & Tarrow, S. (Eds.). (1998). The social movement society: Contentious poli-
tics for a new century. New York: Rowman & Littlefield.
Moaddel, M. (2002). Jordanian exceptionalism: A comparative analysis of state–religion
relationships in Egypt, Iran, Jordan and Syria. New York: Palgrave Macmillan.
Moghadam, V. M. (2002). Islamic feminism and its discontents: Toward a resolution
of the debate. Signs: Journal of Women in Culture and Society, 27, 1135–1171.
Muller, E. N. (1985). Income inequality, regime repressiveness, and political violence.
Munson, Z. (2001). Islamic mobilization: Social movement theory and the Egyptian
Muslim Brotherhood. The Sociological Quarterly, 42, 487–510.
Nawar, L. (2010). Socio- economic characteristics of the population, Egypt [On-line].
Available: http://www.statssa.gov.za/assd2010/Presentations/6th_ASSD/Dr%20
Laila%20Nawar-Egypt.pdf. Retrieved February 13, 2011.
Nepstad, S. E., & Smith, C. (1999). Rethinking recruitment to high risk/cost activ-
ism: The case of Nicaragua exchange. Mobilization, 4, 40–51.

202
Opp, K.-D., & Roehl, W. (1990). Repression, micromobilization, and political protest.
Osa, M., & Corduneanu-Huci, C. (2003). Running uphill: Political opportunities in
non-democracies. Comparative Sociology, 2, 605–629.
Robison, K. K., Crenshaw, E. M., & Jenkins, J. C. (2006). Ideologies of violence: The
social origins of Islamist and Leftist transnational terrorism. Social Forces, 84,
2009–2026.
Ross, M. (2009). Oil and democracy revisited. Unpublished manuscript, University of
California at Los Angeles.
Sageman, M. (2004). Understanding terror networks. Philadelphia: University of
Pennsylvania Press.
SBS World Guide. (2003). SBS world guide: The complete fact file on every country (11th
ed.). South Yarra, Australia: Hardie Grant Books.
Schlozman, K. L., Burns, N., & Verba, S. (1994). Gender and the pathways to partici-
pation: The role of resources. Journal of Politics, 56, 963–990.
Schussman, A., & Soule, S. A. (2005). Process and protest: Accounting for individual
protest participation. Social Forces, 84, 1083–1108.
Slackman, M. (2008). Stifled, Egypt’s young turn to Islamic fervor. New York Times,
February 17, p. A1.
Soule, S. A., & Earl, J. (2005). A movement society evaluated: Collective protest in the
United States, 1960–1986. Mobilization, 10, 345–364.
Starks, B., & Robinson, R. V. (2005). Who values the obedient child now? The
religious factor in adult values for children, 1986– 2002. Social Forces, 84,
343–359.
Tarrow, S. (1998). Power in movement: Social movements and contentious politics (2nd
ed.) New York: Cambridge University Press.
Testas, A. (2004). Determinants of terrorism in the Muslim world: An empirical
cross-sectional analysis. Terrorism and Political Violence, 16, 253–273.
United Nations. 2001. Human development indicators: 2001 report [On-line]. Available: http://
hdr.undp.org/en/media/completenew1.pdf. Retrieved February 14, 2011.
Verba, S., Schlozman, K. L., Brady, H. E., & Nei, N. (1995). American citizen participa-
tion study 1990. Ann Arbor, MI: ICPSR.
Wickham, C. R. (2002). Mobilizing Islam: Religion, activism and political change in
Egypt. New York: Columbia University Press.
———. (2004). Interests, ideas, and Islamist outreach in Egypt. In Q. Wiktorowicz
(Ed.). Islamic activism: A social movement theory approach (pp. 231– 249).
Wiktorowicz, Q. (2004). Islamic activism and social movement theory. In Q. Wiktorowicz
(Ed.). Islamic activism: A social movement theory approach (pp. 1–33). Bloomington: Indiana
University Press.
Woltering, R. A. F. L. (2002). The roots of Islamic popularity. Third World Quarterly,
23, 1133–1143.
World Bank. (2004). World development indicators, 2004. Washington, DC: Author.
World Values Survey (1999-2004). Official Aggregate, v. 20140429. World Values
Survey Association (www.worldvaluessurvey.org). Aggregate File Producer: Asep/
JDS, Madrid SPAIN.

203
APPENDIX
Table A6.1 Characteristics of Surveys and Countries for Six Muslim-Majority Nations
Country-level characteristics
Country HDI Inequality FDI/GDP Political exclusion Repression
Algeria .69 35.3 0.9 6 4.5

Bangladesh .47 31.8 0.1 3 3.0
Egypt .64 34.4 0.8 6 3.0
Indonesia .68 34.3 2.1 3 4.0
Jordan .71 36.4 0.9 4 2.5
Pakistan .49 33.0 1.4 6 3.5
note: “HDI” (human development index) is from United Nations. (2001). Human development
indicators: 2001 report [On-line]. Available: http://hdr.undp.org/en/media/completenew1.pdf.
Retrieved February 14, 2011. “Inequality” is the Gini coefficient in the closest year preceding the
survey from World Bank. (2004). World development indicators, 2004 (pp. 60–62). Washington,
DC: Author. “FDI/GDP” is gross foreign investment as a percentage of gross domestic
product (GDP) in the year closest to the survey from World Bank. (2004). World development
indicators, 2004 (pp. 306–308). Washington, DC: Author. “Political exclusion” (in 2000),
which ranges from 1 (low) to 7 (high), is from Freedom House. (2001). Freedom in the world,
2000–2001 (Table 1, pp. 655–656) New York: Transaction Publishers. “Repression” (in 2000),
which ranges from 1 (low) to 5 (high), is the Political Terror Scale from Gibney, M., Cornett,
L., & Wood, R. (2009). Political Terror Scale, 1976–2009 [On-line]. Available: http://www.
politicalterrorscale.org/. Retrieved February 11, 2011.

204
205
CHAPTER 7 The Arab Spring and Egyptian

Revolution Makers
Predictors of Participation
Mansoor Moaddel
On December 17, 2010, street vendor Mohammed Bouazizi set himself on fire
in Sidi Bouzid, Tunisia, in protest against the mistreatment he received from
municipal agents. His death, days later, turned into a national tragedy and
unleashed such an uproar that President Zein al-Abedin Ben Ali was forced to
flee the country. Like wildfire, popular protests then spread from one Arab coun-
try to another until the entire region was engulfed, forcing out dictators in Egypt,
Libya, and Yemen. Now, Syrian President Bashar al-Assad, too, seems near the
end of his rule.
These events, dubbed the Arab Spring, raise questions about the dominant
sociological theories of rebellion and revolution, including political conflict,
resource mobilization, organizational, and political opportunity structure (PROP)
theories. Although in every case discontented individuals appeared to have over-
come fear, defied the repressive regimes, and poured into the public space in large
numbers, it is not clear how these individuals organized themselves into collective
action, planned rallies, designed visual rhetoric, produced slogans they chanted
harmoniously, and mobilized emotional energies. More challenging, however,
are the socioeconomic and cultural factors that predict participation. Are these
participants predominantly self-empowered, cognitively liberated, linked in orga-
nized networks, and morally outraged as PROP theories tend to predict? Or, alter-
natively, are they predominantly fatalistic, insecure, and powerless as forecasted
by mass–society, structural–functional, and relative–deprivation (MSR) theories?
What were the facilitators of and inhibitors to participation? What is the relation-
ship between socioeconomic status and participation? How did the discontented
individuals manage to connect to one another and form the critical mass that
tipped the balance of forces vis-à-vis the government to their favor? To what extent
is participation linked to liberal values, religiosity, morality, organization, sources
of information, and attitudes toward outsiders?
An analysis of the series of events that contributed to the making of the Arab
Spring is useful for empirical interrogations of the dominant theories of revo-
lutions and rebellions. In the case of Egypt, however, nationally representative
206
sample data are available that provide insights into the socioeconomic and cultural
profiles of the revolutionary actors. By analyzing data from a full-scale national
survey carried out in the country in summer 2011, this chapter (a) assesses alter-
native hypotheses derived from PROP and MSR theories, (b) reveals limitations
in both clusters of theories, (c) suggests additional predictors of participation,
(d) offers an alternative conceptualization of revolutionary actors, (e) suggests the
notion of the monolithic state as the unifying factor in revolutionary development,
and, thus, (f) contributes to the social–scientific understanding of revolutionary
participation.
Theoretical Development: Profile of Revolution Makers

People who play leading roles in a revolutionary movement have been conceptu-
alized as “professional revolutionists” (Arendt, 1963, pp. 262–263),1 “power con-
tenders” (Tilly, 1978), or “state builders” (Skocpol, 1979, p. 172). These concepts,
however, are too restrictive to be applicable to the ordinary people involved in
a revolutionary movement. As an alternative, the individual participants may be
conceptualized as revolution makers. This concept is used as a heuristic device to
apply to the category of diverse individuals committed in varying degrees to a revo-
lutionary overthrow of the ruling regime. These individuals seek out resources,
networks, and organizations; make use of the available opportunities; and invent
new strategies of political action that render the state’s repressive technologies
ineffective, catching the repressive forces by surprise. Although PROP and MSR
theories are juxtaposed to assess their explanatory power in predicting the pro-
files of revolution makers, an interdisciplinary perspective is adopted to arrive at a
fuller understanding of the range of sociological and psychological factors linked
to revolutionary actions.
Revolution Makers: Mobilizing Ideas and Attitudes, Efficacy,

and Morality
In MSR theories, revolutions are outcomes of unresolved tensions generated
by rapid social change, a sudden economic downturn, or the outbreak of a war.
These tensions are produced through (a) social-system disequilibrium that causes
the breakdown of social ties (Johnson, 1964, 1966), (b) the collapse of interme-
diate social organizations that link individuals to political system (Arendt, 1958,
1963; Kornhauser, 1959), (c) frustration caused by rising expectations that out-
pace achievements (Davies, 1969; Gurr, 1970), and (d) the widening gap between
modernization and institutionalization (Huntington, 1968). PROP theorists reject
this vision of the revolutionary process, arguing that revolutions are forms of
collective action shaped by the structure of political opportunity and the avail-
ability of resources. They arise out of a revolutionary situation characterized by
(a) the emergence of contenders making exclusive alternative claims to control
of the government, (b) commitments of a significant section of the population
to these claims, and (c) the failure of the government to repress (Tilly, 1978,

207
p. 202). Revolutions are manifestations of political conflict; they are “politics by

other means” (Gamson, 1975, p. 3; McAdam, 1982, 1986; Oberschall, 1973; Zald &
McCarthy, 1987).
As described in the following pages, MSR and PROP theories predict different
profiles of revolutionary participants.
Mobilizing Ideas and Attitudes

Which types of ideas and subjective orientations are linked to participation,
and how such linkages are understood depend on one’s perspective. For MSR
theories, the concept of anomie features prominently; anomic individuals are
thought to be vulnerable to the appeals of revolutionary movements (Arendt, 1958;
Davies, 1969; Gurr, 1970; Johnson, 1964, 1966). Although anomie is an outcome
of deregulation caused by societal breakdown, insofar as one theorizes about the
mental pathological consequences of dysfunctional social structure, the explana-
tory power of MSR theories may be enhanced if the etiological range of these
theories is expanded to include also the opposite extreme of an underregulated
society—namely, an overly regulated social order. Durkheim coined the concept of
fatalism as a mental pathology produced by “excessive regulations,” such that, as
he noted, “futures [are] pitilessly blocked and passions violently choked by oppres-
sive discipline” (cited in Dohrenwend, 1959, p. 467; see also Acevedo, 2005;
Besnard, 1993; Lockwood, 1992; Lukes, 1967; Moaddel & Karabenick, 2008).
Although overlooked by MSR theorists, fatalism signifies that excessive regula-
tions also undermine individual integration into the social system. Revolutionary
movements may thus be even more appealing to the people who feel overcome
by oppressive regulations because such movements provide a conduit for their
deliverance from overbearing authoritarian domination.
PROP theorists, on the other hand, not only reject the notion of psychological-
cum-revolutionary consequences of dysfunctional social structure, but also
overlook the significance of grievances in shaping revolutionary mobilization.
This is because, as McCarthy and Zald have argued, “there is always enough
discontent (grievance) in any society to supply the grass-roots [sic] support for
a movement” (cited in Snow, 2004, p. 382; see also McCarthy & Zald, 1977,
pp. 1214–1215). Likewise, Tilly (1975) ignores the significance of mobilizing ideas
and attitudes by criticizing MSR theories for resting on the fallacious prem-
ise “that rebellion is an individual act intimately dependent on a certain …
rebellious attitude … toward some or all authorities” (p. 487). For sure, he
considers people accepting alternative claims as a proximate cause of revo-
lution. Nonetheless, this acceptance is not an effect of the mobilizing ideas
these claims encompass. Rather, for him, mobilization occurs because of “the
sudden failure of the government to meet specific obligations” or “unexpected
increase in the government’s demand for surrender of resources by its sub-
ject population” (Tilly, 1978, pp. 204–205). Mobilizing ideas and attitudes are
thus relatively inconsequential in the emergence and dynamics of revolution-
ary movements.
This chapter tests this proposition by assessing the extent to which grievances
are linked to revolutionary mobilization. It focuses on such factors as low trust
The Arab Spring And Egyptian Revolution makers | 207

208
in government, dissatisfaction with its performance, and the perception that the
government is not working for their benefit. When people have high trust in their
government, and consider public officials trustworthy and civil, they tend to sup-
port government policies and are less likely to engage in collective political pro-
test. They may participate in political upheavals when this trust is shattered and
political cynicism increases, and when the belief in the prevalence of corruption
in high places is widely shared (see Levi and Stoker [2000] for a review; Freitag &
Bühlmann, 2009; Gamson, 1968; You, 2012). The empirical evidence on the link-
age between mistrust and system-changing or unconventional political behavior,
however, have been inconclusive (Citrin, 1977; Muller, Jukam, & Selligson, 1982;
Sigelman & Feldman, 1983). Other research findings, on the other hand, have sup-
ported the connection between mistrust, efficacy, and protest participation (Craig &
Maggiotto, 1981; Lee, 1992; Paige, 1971; Sears & McConahay, 1973; Shingles, 1981).
None of these studies, however, assesses these linkages in a successful case of a
revolutionary movement similar to Egypt.
Grievances, however, are problematic; they cannot be taken for granted.
They must be defined in terms that not only reflect the movement’s interests,
but are also broad enough to be acceptable to the wider public, contributing to
their participation in the revolutionary movement. It is one thing for discon-
tented individuals to make effective use of the available resources and oppor-
tunity structure to realize their goals. It is quite another thing to legitimize
their goals and actions vis-à-vis a larger public, particularly when they altercate
with the authorities, scuffle with security forces, and damage property. To be
effective in rallying the public to their side, such acts must be framed not as
misconducts subject to criminal prosecution, but rather as “expressions of out-
rage against injustice” (Turner, 1969, p. 816). Framing theorists thus amend
PROP theories by viewing movements as “signifying agents … they frame, or
assign meaning to and interpret relevant events and conditions in ways that are
intended to mobilize potential adherents and constituents, to garner bystander
support, and to demobilize antagonists” (Snow, 2004, p. 384; Snow & Benford,
1988, p. 198).
This chapter uses the concept of ideological framing, but with one important
caveat. Because revolution makers are constituted by the same ideological fram-
ing in terms of which events and issues are interpreted (Furet, 1981; Moaddel,
1992), they cannot frame or assign meaning to these events at will to broaden
the appeal of their movements. It would be quite difficult for liberal-cum-secular
actors to frame issues in religious terms or religious fundamentalists in terms of
the equality of all faiths in their mobilization efforts. This study also argues that,
in a revolutionary movement, there may be more than one framing that shapes
political action. In the case of contemporary Egypt, of the three ideological fram-
ings of liberalism, pan-Arab nationalism, and religious fundamentalism that have
informed politics in the country (Moaddel, 2005), pan-Arab nationalism has been
identified closely with the authoritarian regimes (particularly in Egypt, Iraq, and
Syria) and thus has little support among the ordinary public.2 Liberal outlook and
religious fundamentalism, on the other hand, have remained two competing ide-
ologies of revolutionary movement in the country in recent decades. This chapter

209
assesses the linkages between people’s adherence to these two diverse framings
and revolutionary participation.
Individual Efficacy Versus Dysphoric Emotions

MSR and PROP theories offer contrasting hypotheses on the individual par-
ticipants in revolution movements. MSR theorists consider such individuals as
those who feel powerless, fatalist, insecure, or xenophobic. PROP theorists, on
the other hand, portray an opposite conception of the individual revolutionary pre-
disposition. Revolution makers are portrayed as well-integrated members of their
communities, self-empowered, capable of displaying considerable enthusiasm
and energy, and cognitively liberated. For them, favorable structural conditions
and political opportunities are translated into action only when dissatisfied indi-
viduals consider themselves powerful enough to alter their lot in life (McAdam,
1982; Nepstad, 1997; Piven & Cloward, 1977). Inefficacious individuals, according
to Schwartz (1973, pp. 12–13), do not believe themselves to be entitled to more
power and are happy with the way things are. It is the feeling of being empowered
enough—politically efficacious—to demand a more responsive government that
prompts individuals to participate in protest movements (see also Citrin, 1974;
Craig & Maggiotto, 1981; Gamson, 1968; Miller, Miller, & Schneider, 1980; Pierce
& Converse, 1989; Sigelman & Feldman, 1983).
Moral Values
Although morality has been either an explicit or implicit component of vari-
ous theories of revolution, its linkages with participation has rarely been tested
empirically. How do moral commitments motivate participation in a revolutionary
movement? To what extent are revolution makers committed to the same moral
ideals of nonviolence, honesty, truth, and social justice that they accuse the ruling
regime of violating? Are the more active participants driven by a stronger attach-
ment to morality than the less active or nonparticipants? Revolution makers may
be in consensus in considering immoral such behaviors as the regime’s violence
against peaceful demonstrators, economic corruption, and deception of the public
through the spread of misinformation. Does the feeling of moral outrage at these
behaviors activate people with a stronger moral commitment to participate in pro-
tests against the regime?
The literature on revolution offers opposing interpretations of the relation of
moral commitment to revolutionary action. In Marxism, morality has been an
integral component of its theory of revolution. Marxists have been explicit about
the moral superiority of workers in socialist revolutionary movements. As a uni-
versal class, their struggle against the particularistic interests of the bourgeoisie
serves the broader humanity by creating the condition for the ultimate elimina-
tion of class exploitation and the formation of a just social order. In PROP theo-
ries, on the other hand, although morality is not as central, allusions are often
made to such expressions as “moral outrage,” “moral shock,” “social injustice
and corruption in high places,” “the spread of misinformation and deception by
the ruling regime,” and “violation of human rights” as factors motivating par-
ticipation in collective protests (Collins, 2001; Gamson, 1992; Goodwin, Jasper, &

210
Polletta, 2001; Jasper, 1999; Nepstad & Smith, 2001). Nepstad and Smith (2001)
suggest that “moral outrage is a logical reaction to the torture, disappearances,
and assassinations of innocent civilians and to the lies disseminated by a govern-
ment to cover its role as an accomplice to these atrocities” (p. 173). Focusing on
such popularly labeled progressive movements as those for racial equality, nuclear
disarmament and peace, poor people, and gender equality, these theorists tend to
presume higher morality of the individual participants in such movements than
nonparticipants.
MSR theorists do not address the role of morality in revolutionary develop-
ments; but, because—for them—revolutions flow from societal breakdown, ano-
mie, or individual frustrations caused by unmet expectations, they do not consider
moral outrage as a motivating factor for the participants. On the contrary, their
theoretical framework predicts just an opposite relationship between morality and
participation. Among MSR theorists, however, Arendt (1963) is explicit in point-
ing out the immorality of professional revolutionists. She identifies a series of
social, psychological, and historical factors to explain the revolutionists’ retreat
from the ideal of freedom and liberty on the one hand, and adoption of violence
on the other, during the French Revolution and the subsequent revolutions that
it inspired.
For her, the French revolutionaries failed to draw the correct lesson from the
American Revolution. They were affected little by “the thoughtful and erudite
political theories of the Founding Fathers” (Arendt, 1963, p. 17). Rather, “ ‘the
lovely equality,’ in Jefferson’s words, ‘which the poor enjoy with the rich’ [in
America] revolutionized the spirit of men … to such an extent that … it appeared
to revolutionary men more important to change the fabric of society, as it had
been changed in America prior to its Revolution, than to change the structure of
political realm” (p. 17). This shift to the social was also dictated by France’s dire
economic conditions, which in turn gave rise to a range of emotions that was
first noted by Rousseau’s discovery of compassion. This sensibility, however, was
flawed not just because of Rousseau’s “fantastic irresponsibility and unreliability”
(p. 85), but because it was not realized “out of suffering with others” and was thus
insensitive to reality. Understood as an emotion, delight of intimacy, and pity, com-
passion turned into a political tool in Robespierre’s glorification of the poor and
praising of suffering as the spring of virtue—all constituting the subjective under-
pinning for his “surprising faithlessness that foreshadowed the greater perfidy
which was to play such a monstrous role in the revolutionary tradition” (p. 85, see
also pp. 83–86). As Arendt elaborates further, “the more bloodthirsty the speaker
[in revolutionary France] the more likely that he will insist on ces tendres affections
de l’âme—on the tenderness of his soul” (p. 295, note 30). In addition, this expe-
rience became a historical factor that perpetuated violence in the revolutionary
traditions, for the revolutionists who took their cues from the French Revolution
“learned and knew beforehand the course a revolution must take. … They knew
that a revolution must devour its own children, just as they knew that a revolution
would take its course in a sequence of revolutions, or that the open enemy was
followed by the hidden enemy under the mask of the ‘suspects’ ” (p. 51).

211
These contradictory interpretations of the relationship between morality and

participation notwithstanding, these theories commonly misconceive the role of
morality in revolutionary development. Both Marxist and Arendt’s conception of
morality is deductive and formalistic. In Marxism, it rests on whether one agrees
with its theory of history. Likewise, for Arendt, professional revolutionists turn
to violence for the very reason of believing in a theory of history projected by the
French experience. PROP theorists, on the other hand, do not offer a theory of his-
tory, but their recourse to moral outrage as galvanizing revolutionary mobilization
is often self-referential—based on the participants’ assertion.
In either case— whether one presumes commitments to moral values as
an essential feature of revolution makers or revolutions as an immoralizing
process—the relationship between morality and revolutionary participation must
be tested empirically. This test, however, must be based on independent measures
of morality and participation, which entails evaluating how people’s ratings of
the immorality of behaviors such as violence against other people, stealing other
people’s property, and telling lies to protect one’s interests (not whether they are
outraged morally at the behavior of the regime) are related to the extent of revolu-
tionary participation.
Mediums of Communicative Power
Revolutionary action is not just an outcome of participants’ self-reflections on the
social conditions shaping their lives; people generally rely on others in forming
opinions on issues (Kinder, 1998; Popkin, 1991; Sniderman, Brody, & Tetlock,
1991) and in validating the authenticity of the available information (Kruglanski,
1989). However, the information that serves a revolutionary end is hard to come
by. What is more, useful information is not just a one-time deal that, when or
if it is present, individuals begin to mobilize. In a revolutionary development,
there must be a continuous replenishing of information about situational changes
wrought by unfolding events. On the other hand, because information and com-
munication are among the fundamental sources of power and domination (Bell,
1973; Castells, 2007; Salvaggio, 1989), an unrestrained communication system
and unfettered distribution of information are potentially subversive of state
power. Naturally, the ruling regimes see it in their interests to control the sources
of news information and invalidate the authenticity of the alternative opposition
sources.
A medium of communication thus plays a transformative role when it facili-
tates exchanges of information among revolutionary activists. For example, a lan-
guage is said to have served as a medium for turning social conflict into ethnic
antagonism and nationalist revolutionary movement under ethnically divided
empires, where economic change has caused social dislocations (Gellner, 1965);
print capitalism—by generating unified fields of communication, giving a new
fixity to language and creating languages-of-power—is claimed to have contrib-
uted to the rise of nationalism in Europe (Anderson, 1983, pp. 44–45) and in mod-
ern cultural movements in the 19th-century Muslim world (Robinson, 1993); and a
network of discursive literacy—“that is, the ability to read and write non-formulaic

212
text” (Mann, 1992, p. 141)—is believed to have created a medium that gave rise to
cross-class national communities. In these instances, language, print capitalism,
and discursive literacy constituted new mediums of communicative power that
contributed to the emergence of different forms of nationalism.
An innovative use of a medium of communication by the opposition may thus
enhance significantly the process of revolutionary mobilization. The state’s loss of
communicative control—the decline in its ability to manage and manipulate the
flow and content of communication—on the other hand, may undermine its abil-
ity to defeat the revolutionary movement. This loss may occur in a variety of ways,
such as when the alternative sources of communicative power are located beyond
the national borders, as occurred in Iran, for example, before the Constitutional
Revolution (Abrahamian, 1982; Bayat, 1991); when revolution makers use cultural
or religious networks and rituals as a vehicle for communicating with discon-
tented individuals (Moaddel, 1993; Tehranian, 1980); when the ruling dictator is
immersed in the silent medium of whispers and gossips about his vulnerability
and impending downfall (Chen, Lu, & Suen, 2012); or when inventions in commu-
nication technology facilitate the spread of revolutionary messages in a manner
either beyond the state’s control, or the ruling elite are unable to develop an effec-
tive strategy to control the new medium. For example, anecdotal evidence pointed
to the significance of Facebook, Google, and Twitter as outlets for outrage expres-
sions during the Egyptian revolutionary movement (Gross, 2011; Preston, 2011).
This chapter assesses the role of cyberspace and wireless networks in shap-
ing a new medium of communicative power and thus predicting participation in
revolutionary movements.
Demography, Organization, and Differential Recruitment

PROP and MSR theories relate different demographic characteristics to par-
ticipation. In PROP theories, access to social capital—networks, communities,
voluntary associations, frequency of face- to-
face interactions—
provides the
infrastructure for collective action (Davis & Robinson, 2010; Minkoff, 1997;
Paxton, 2002; Portes, 1998). Likewise, education is said to lower cognitive
barriers to political enlightenment; educated people are believed to be more
engrossed and concerned politically, have greater skills and sophistication in
analyzing and understanding issues and events, and display a stronger deter-
mination to see a task completed than those less educated (Davis & Robinson,
2010; Kinder, 1998; Krueger & Malečková, 2003; Martinez, 2005; Schussman
& Soule, 2005). Also consistent with PROP theories are the influence of gender,
urban living, and ethnicity on participation. Whereas men enjoy more resources
and social networks than women, as in the contemporary Middle East, they
tend to participate in social movements more often than women (Ibrahim,
1980; Sageman, 2004). Urban living promotes participation because it is con-
ducive to more resources and access to more information and organization
than living in the rural areas (Dessouki, 1982; Moaddel, 1993; Wiktorowicz,
2004). Finally, ethnic belonging may be linked to a greater level of participa-
tion because ethnicity provides a basis for shared understanding and effective

213
communication among members, particularly where an ethnic group is dis-

criminated against (Armstrong, 1982; Burnstein, Crandall, & Kitayama, 1994;
Connor, 1994; Fishman, 1972; Fishman, Gertner, Lowy, & Milan, 1985; Hechter,
1975; Korchmaros & Kenny, 2001; Moaddel, Tessler, & Inglehart, 2008a, 2008b;
Smith, 1998; van den Berghe, 1978).
Consistent with MSR theories, on the other hand, are the influences of such
other factors as income and social class, where lower income compared with
one’s relatively higher education or self-assigned class membership tends to
enhance the feeling of relative deprivation and unmet expectations (Ayubi,
1991; Davis & Robinson, 2010; Ghadbian, 2000; Gurr, 1970; Woltering, 2002),
as are the effects of being younger and unmarried, which indicate the signifi-
cance of weaker social ties on participation.3 Both clusters of theories, however,
appear to be silent on the significance of experience during formative years on
revolutionary participation. It has been argued that attitudes and behavior are
often shaped by the common experience of a cohort (Schuman & Rieger, 1992).
People’s impressionable years under a particular regime are said to shape their
orientation toward that regime. South Korean youths, for example, have grown
critical of the United States, but those 50 years and older, having experienced
the Korean War, consider “the U.S. as a savior that stopped communist take-
over” (Lee, 2004, p. 17). Instead of age, this chapter uses formative years under
President Mubarak as a better predictor of participation. Those age 12 years or
younger when Hosni Mubarak became president have probably weaker memo-
ries of political conditions under his predecessors than those who were older
than 12. Their view of politics being formed under President Mubarak may be
different from that of the older cohort and thus may be linked to revolutionary
participation.
Methods: Measurement and Hypotheses

These theoretical issues are used to specify and measure the variables of this study
and develop hypotheses on the predictors of participation.
Dependent Variable
The dependent variable is the respondents’ self-rating of the extent of their partici-
pation. They were asked:
Regarding the revolutionary movement against former President Hosni Mubarak,
on the scale of 1 to 10, with 1 indicating no participation and 10 the utmost partici-
pation, where do you put the extent of your participation on this scale?
Correlates of Participation
The correlates of participation fall into three broad categories: (a) attitudes, effi-
cacy, and morality; (b) mediums of communicative power; and (c) demographics
and organizational membership.

214
Mobilizing Ideas and Attitudes, Efficacy, and Morality

More detailed information on how the indicators used in this study are con-
structed is presented in the Appendix.
Several indicators are used to measure anomie, fatalism, grievances, ideologi-
cal framings, personal efficacy and dysphoric emotions, and moral flexibility. The
first two focus on governmental rules and regulations and are intended to mea-
sure Durkheim’s concepts of fatalism, too many rules (A1); and anomie, two few
rules (A2). Three indicators measure grievances: mistrust of government (A3), dis-
satisfaction with government (A4), and the perception that the economy runs for the
benefits of a special few (A5). Higher values on these variables indicate stronger
fatalism, anomie, mistrust in government, dissatisfaction with government, and
perception that a special few are benefiting from the economy. Three interaction
variables are also included in the analysis: mistrust × dissatisfaction (A6), special
few × dissatisfaction (A7), and special few × mistrust (A8) interactions.
To measure ideological framings, indices of liberalism and fundamentalism
are constructed. Liberalism is a multidimensional concept that captures orien-
tation toward recognizing the autonomy of the individual, gender equality, and
secular democracy. A composite index of social individualism is constructed by
averaging responses to questions on the basis for marriage (A9), whether inde-
pendence (A10) is considered, and whether obedience (A11) is not considered
among favorable qualities for children to have. A gender–equality index is also
constructed by averaging four Likert-scale measures of attitudes toward gender
relations: A wife must always obey her husband (A12), men make better political
leaders (A13), a university education is more important for a boy (A14), and men
should have more right to a job than women (A15). Finally, the third component is
secular democracy, which is also a composite measure of attitudes toward secular
politics (A16) and democracy (A17). The liberalism index is the average of indices
of social individualism, gender equality, and secular democracy:
Social individualism + Gender equality + Liberal government

Liberalism index =
3
Higher values on this index are associated with stronger liberal attitudes—that is,
more favorable orientations toward social individualism, gender equality, democ-
racy, and secular politics.
The liberalism index is normative. A cognitive indicator of secularism is also
used to assess how understanding that development causes a decline in religious
belief is linked to participation. This variable measures whether respondents
believe the belief in God declines as Egypt becomes more developed in the future
(A18). Higher values on this measure indicate a stronger belief in secularism.
Religious fundamentalism is defined in general terms as a set of beliefs about
and attitudes toward whatever religious beliefs one has. For example, believ-
ing that one’s religion is superior to the religion of others, that one’s religion
is the only true religion, and that one’s religion is closer to God than other reli-
gions does not address religious tenets per se, but rather attitudes toward the

215
significance of one’s religion. Fundamentalism also projects an image of the deity

that is disciplinarian; advances a reading of the scriptures that is literal, inerrant,
and infallible; and bestows a standing to its religion that is superior to, intoler-
ant of, and more exclusivist (i.e., closer to God) than other religions (Altemeyer
& Hunsberger, 2004; Moaddel & Karabenick, 2008). Religious fundamentalism
is thus a multidimensional concept, consisting of an authoritarian conception of
the deity (authoritarian deity factor based on items A19–A23), literalism/inerrancy
(literalism factor based on items A24–A26), exclusivity (exclusivism factor based
on items A27 and A28), and intolerance (intolerance factor based on A29 and
A30). A single fundamentalism index is constructed by averaging the four factors.
Authoritarian deity + Literalism + Exclusivity + Intolerance

Fundamentalism index =
4
A higher value on this measure indicates a stronger fundamentalist orientation.

To underscore the significance of fundamentalist framing, not religiosity, in
shaping people’s desire to participate in the revolutionary movement, several mea-
sures of religiosity are included in the analysis as control variables: daily prayer
(A31), self-described as a religious person (A32), the importance of God in life (A33),
and mosque attendance (A34). Mosque attendance may also reflect organization
and a source of communicative power. Mosques have been a setting for com-
munications between religious leaders-cum-activists and followers, exchange of
information among the participants, and networking.
Political efficacy refers to individuals’ belief in their ability to affect the scope
and content of government policy. Although this conceptualization may apply to
conditions of liberal democracy, for political activists under an authoritarian state
the question may not be whether they are capable of affecting state policies, but
rather how to get rid of the authoritarian ruler. Thus, the belief that one cannot
affect government policy may not be indicative of inefficacy. As an alternative, a
broader conception of efficacy is used; one that focused on perceived control (A35)
and the belief in freewill versus fate (A36). That is, those who give themselves a
higher mark of “having control over their lives” or “deciding their own destiny” are
more efficacious individuals, whereas those who give themselves a lower score on
these measures are less so. Because the linkages between one’s belief in free will
and participation may depend on whether one perceives the presence of too many
rules (see A1), a variable measuring the interaction between these two indicators
is also included in the analysis (A37).
Furthermore, measures of insecurity and xenophobia are included in the analy-
sis to assess the predictive power of MSR theories. A general measure of insecu-
rity is used in this study (A38). Because the feeling of insecurity may be reinforced
by the perception that there are too many rules (A1), a measure of the interaction
between these two variables is included in the analysis (A39). The measure of
xenophobia, however, may depend on time and space. In the current European
context, the construct is measured in terms of attitudes toward immigrants (de
Figueiredo & Elkins, 2003), but immigration is not an issue among Egyptians.

216
In 20th-century Egypt, on the other hand, xenophobia was often expressed in

terms of anti-Western attitudes, particularly in such radical ideologies as pan-Arab
nationalism and Islamic fundamentalism (Gershoni & Jankowski, 1995; Moaddel,
2005). Therefore, xenophobia is measured as a factor based on the preference not
to have Americans (A40), British (A41), or French (A42) as neighbors.
Finally, morality is measured in terms of respondents’ rating three indica-
tors of immoral behavior: violence against other people, stealing other people’s
property, and telling lies to protect one’s interests. Because these variables are
correlated they are combined linearly to make a single moral–flexibility index
(A43–A45):
Violence + Stealing + Telling Lies

Morality =
3
Lower values indicate stronger moral commitment, while higher values moral
flexibility.
Radio (A46), domestic TV (A47), satellite TV (48), newspapers (A49), the
Internet (A50), and mobile phones (A51) are considered mediums of commu-
nicative power. Respondents’ reliance on these mediums as sources of news
information was measured. It is proposed that these mediums provide varied
communicative power to the revolution makers. Although the Egyptian govern-
ment had a more effective control over the content and flow of information,
such as in cases of domestic TV and radio, the news information broadcast
through these mediums naturally portrayed the government’s performance in
a positive light and primed more favorable features of the ruling elite. It is thus
proposed that domestic TV and radio enhanced the communicative power of
the government and weakened that of the opposition, and as a result, people
who relied on these sources tended to participate less often in the revolutionary
movement. The government, on the other hand, having far less control over
satellite TV, the Internet, or people’s conversation on mobile phones, could not
have manipulated easily the flow of information traveling through these medi-
ums. Being accessible to opposition activists, these mediums tended to increase
these activists’ communicative power, serving as useful vehicles to spread anti-
government messages among the broader public. As a result, people relying on
these sources tended to participate more often in antiregime demonstrations.
The linkages between relying on newspapers as sources of information and
participation depend on the specific newspaper on which respondents relied.
It may be proposed that people who relied on the opposition’s newspapers
tended to participate in the revolutionary movement more often than those who
relied on proregime newspapers or those who did not rely on newspapers as a
source of information. A dummy variable was thus created based on whether
the respondents mentioned relying on opposition newspapers (A52). Finally, it is
not possible to consider newspapers, the Internet, and mobile phones as totally

217
separate sources of news information for individual respondents. People who

mentioned a newspaper as a source might have accessed it via the Internet. Or
people who used mobile phones as a source of information might have also
used the Internet or accessed the Internet from their phone. As a result, and
given that these three measures correlate significantly, a paper–Internet–mobile
phone factor was constructed based on these indicators.
Demographics
Several measures of demographic attributes of the respondents are included
in this analysis: education (A53), self-assigned class membership (A54), and
income (A55). These three variables, which are correlated significantly, are used
to create a socioeconomic status factor. To assess whether the members of the
middle class participated more frequently in the revolutionary movements than
the members of lower or higher classes, a square of the socioeconomic factor
(A56) was included in the analysis to assess an inverted-U pattern of class and
participation.
Other demographic variables are included to measure their effects on par-
ticipation of living one’s impressionable years under Mubarak (A57), being male
(A58), residing in an urban area (A59), and being single (A60). Finally, an orga-
nizational membership factor is constructed based on membership in a religious
organization (A61), political party (A62), and humanitarian or charitable orga-
nization (A63).
The relationships among all these variables and participation are considered
linear, except for the effect of socioeconomic status, which is proposed to be curvi-
linear, following an inverted U shape.
Hypotheses and the Limitations of the Study

Based on the measurements of the theoretically specified variables, the following
hypotheses are advanced: The participation rate is higher among Egyptians who
• Feel there were too many rules and regulations (fatalistic) (hypothesis 1)
• Feel there were too few rules and regulations (anomic) (hypothesis 2)
• Have a greater mistrust of government (hypothesis 3)
• Display a greater level of dissatisfaction with the government’s performance.
(hypothesis 4)
• Believe the economy runs for big interests rather than for all the people.
(hypothesis 5)
• Have stronger liberal attitudes (hypothesis 6)
• Associate development with a decline in faith in Allah (hypothesis 7)
• Have stronger fundamentalist attitudes (hypothesis 8)
• Hypothesis 8 remains true, controlling for mosque attendance, daily prayer,
self-description as a religious person, and the importance of God in life
(hypothesis 9)

218
• Attend mosque more frequently (as a proxy measure of organization or tra-

ditional source of communicative power) (hypothesis 10)
• Have higher perceived control (hypothesis 11)
• Have a stronger belief in free will (hypothesis 12)
• Feel more secure (hypothesis 13)
• Feel more xenophobic (hypothesis 14)
• Display a stronger moral commitment (less moral flexibility) (hypothesis 15)
• Rely less often on radio as a source of information (hypothesis 16)
• Rely less often on domestic TV as a source of information (hypothesis 17)
• Rely more often on satellite TV as a source of information (hypothesis 18)
• Rely more often on newspapers, the Internet, and mobile phones as sources
of information (hypothesis 19)
• Rely more often on opposition newspapers as a source of information
(hypothesis 20)
• Are male (hypothesis 21)
• Are single (hypothesis 22)
• Lived their impressionable years under President Hosni Mubarak
(hypothesis 23)
• Are Christian (hypothesis 24)
• Were members of the middle class (There is an inverted-U relationship
between socioeconomic status and participation.) (hypothesis 25)
• Live in urban areas (hypothesis 26)
• Belong to an organization (hypothesis 27)
The analysis presented in this chapter is based on data collected at one point in
time. Although one may still advance causal analysis of the proposed relationship
between, for example, socioeconomic status or age and participation in the revo-
lutionary movement, it is acknowledged that with cross-sectional data, the direc-
tion of causality cannot be determined empirically for all the hypotheses. It is also
acknowledged that it may not be possible to rule out causality in a direction opposite
to what is proposed or that causality flows in both directions. Consequently, the
relationships specified in the following analysis can only be concluded that the data
are consistent with the proposed hypotheses. Nonetheless, by determining whether
the hypothesized relationships exists, the data are useful in evaluating the explana-
tory power of PROP and MSR theories, uncovering the limitations of both theories,
offering theoretically informed insights about the political dynamics that seem likely
to account for variation in participation among different individuals, and suggesting
alternative interpretations of the predictors of revolutionary actions that may tran-
scend the status of the existing sociological theories of revolutionary movements.
Analysis and Results

Data
The data are from a full-scale survey of a nationally representative sample of
3,143 Egyptian adults (18 years and older) conducted between June 6, 2011, and
August 8, 2011. The survey used multistage random sampling procedures,

219
stratified according to urban and rural areas of the country in proportion to their
size, with roughly equal proportions of male and female respondents. The inter-
views took approximately 1 hour to complete and represented 87% of the house-
holds contacted.
The sample included 95.8% Muslims and 4.2% Christian Copts. Just less
than half the sample (45.3%) was from urban areas and the rest was from rural
areas of the country. The respondents had an average age of 39 years, 48.2% were
male, 17.8% had never been married, 71.7% were married, 1.2% divorced, 0.4%
separated, 8.8% widowed, 13.6% had college degrees; and, in terms of class, 0.6%
of the respondents described themselves as members of the upper class; 16.9%,
upper middle class; 33%, lower middle class; 23.1%, working class; and 26.1%,
lower class. Only 44.5% of the respondents indicated they had jobs.
Descriptive Statistics
Dependent Variable: Participation Rate
According to Table 7.1, 59.4% reported they did not participate in the revolu-
tionary movement at all, and the rest indicated varying rates of participation. The
mean rate of participation is 2.71 and standard deviation is 2.56.
This study considers respondents’ self-rated participation to be a valid indicator
of their actual participation in the revolutionary movement. For sure, they may dif-
fer in understanding the operational meaning of the number they selected to indi-
cate the extent of their participation; some may have exaggerated or understated
their participation. It is unlikely, however, that there would be more than 1 or 2
point differences between the number they selected and the number correspond-
ing to their actual participation. For example, the actual participation rate of the
respondents who selected 3 may in fact be closer to 2 or 4. One way to compensate
Table 7.1 Self-Rated Participation in the Revolutionary Movement

against President Hosni Mubarak
Rating Frequency Percent
1 (no participation) 1,866 59.5

2 199 6.3
3 178 5.7
4 179 5.7
5 158 5.0
6 130 4.1
7 172 5.5
8 145 4.6
9 55 1.8
10 (utmost participation) 56 1.8
Total 3,138 100.0
Mean (SD) 2.71 (2.56)
DK/NA 5 .2
note: DK/NA, XXXX; SD, standard deviation.

220
for the effect of such biases is to group responses into four categories: 1 (no partici-
pation), 2 (2–4), 3 (5–7), and 4 (8–10). The regression estimates, however, were not
significantly different from the estimates produced using the original categories
of the dependent variable (not reported). Another test that also supports the stabil-
ity of the findings is to differentiate between those who participated and those who
did not. Again, the estimates from logistic regression (not reported) were remark-
ably consistent with the findings presented in this chapter.
Furthermore, a factor that also reinforces the validity of the measure is that the
cities that reported higher incidents of collective action against the ruling regime
had also on average higher self-rated participation. Respondents from the cities
deemed most active based on anecdotal reports of antiregime demonstrations
had self-rated averages higher than the averages from other cities: Greater Cairo,
3.18; Alexandria, 4.06; Port Said, 4.03; Suez, 4.72; Asyut, 4.09; and Al-Sharqia.
3.72 (Al-Jazeera, 2011; El-Ghobashy, 2011; Kirkpatrick, 2011; Reuters, 2011). On the
other hand, places that experienced the lowest incidents of protest demonstra-
tions also had the lowest on average self-rated participation: Kafr el-Sheikh, 1.69;
El-Menoufia (President Mubarak’s birthplace), 1.87; Beni-Suef, 1.83; and the tour-
ist areas of Aswan, Luxor, and Qena, 1.65. Other cities were between these two
extremes (Table 7.2).4 Finally, although other measures of political activism were
Table 7.2 Region Where the Interview Was Conducted
Region n Mean SD Incidents of

participation antiregime
protest
demonstrations
Greater Cairo 515 3.17 2.30 High

Alexandria 202 4.06 2.72 High
Port Said 31 4.03 3.26 High
Suez 18 4.72 3.43 High
Damiate 52 2.06 2.30 Low
Dakhalia 227 2.09 2.20 Medium
Al-Sharqia 218 3.72 2.87 Medium
Qalioubia 198 3.21 2.93 High
Kafr El-Sheikh 114 1.69 2.07 Low
El-Garbia (Tanta) 198 1.41 1.33 Medium
El-Menoufia 158 1.87 1.84 Low
Baharia (Damnhour) 215 2.31 2.64 Medium
Ismailia 41 4.02 2.75 High
Giza 139 3.03 2.67 Medium
Beni-Suef 101 1.83 2.17 Low
El-Fayoum 100 1.54 1.30 Low
El-Menia 169 1.91 1.85 Low
Asyut 128 4.09 2.78 High
Sohag 140 3.46 2.70 High
Aswan, Luxor, Qena, 128 1.65 1.50 Low
SD, standard deviation.

221
not included in Egypt’s survey, findings from a pilot survey in Tunisia indicated a
strong correlation between self-rated participation in the country’s revolutionary
movement and participation in other form of political activities such as attending
nonviolent protests, marches, or sit-ins.
Predictors of Participation
Table 7.3 shows descriptive statistics for predictors of participation rate. The
first category consists of variables measuring fatalism, anomie, grievances, ideo-
logical framing, perception of personal efficacy, dysphoric emotions, and morality.
Accordingly, about 20% and 36% of the respondents expressed feelings that were
consistent with Durkheim’s conception of fatalism (“too many rules”) and ano-
mie (“two few rules”), respectively. The majority of respondents, expressed “not
very much” or “no confidence at all” in their government (mean, 2.55) or showed
more than average dissatisfaction with its performance (mean, 5.47), and less than
47% indicated the economy runs for big interests. The mean score for the liberal-
ism index (mean, 1.97) indicates that Egyptians, on average, were not liberal. The
illiberalism of the respondents is a result primarily of their unfavorable attitudes
toward gender equality. The mean composite measure of gender equality is 1.84
(minimum–maximum range, 1–4), indicating that a majority of the respondents
favored gender inequality, whereas on the measure of the appropriate form of
government for Egypt, the majority were in favor of secular democracy (mean,
3.05; minimum–maximum range, 1–4; individual composite measures not shown
in Table 7.3). The great majority of Egyptians rejected that with development there
will be a decline in faith in God (mean, 1.87). On the contrary, they believed that
as Egypt becomes more developed, faith in Allah either increases a lot (51%) or
increases a little (23%).
Fundamentalist attitudes, in contrast, were strong among respondents. The
combined measure of fundamentalism based on a lineal combination of com-
posite measures of deity, literalism, exclusivity, and intolerance had a mean of 3.5
(individual composite measures not shown in Table 7.3). However, constructing a
fundamentalism measure based on factor analysis or a measure of fundamental-
ism based on a lineal combination of composite measures of the four components
of fundamentalism did not affect the results of the analysis. The mean score mea-
sures of religiosity indicated Egyptians were high on all measures of religiosity
(4.86 for mosque attendance, 5.62 for daily prayer, 7.28 for defining self as a reli-
gious person, and 9.75 for the important of God in life).
On the first measure of personal efficacy—perceived control—respondents,
on average, felt they had considerable control over their lives (mean score for
perceived control, 6.67). On the second measure of the construct, however, on
average, more Egyptians were fatalists than believers in free will (mean, 3.59).
Likewise, they felt quite insecure (mean, 3.04) and xenophobic, in that 87% of
respondents did not like to have Americans, British, or French as neighbors. On
morality, respondents considered overwhelmingly violence against other people,
stealing other people’s property, and telling lies to protect one’s interests highly
immoral (mean, 1.44).
On average, more respondents used domestic TV as a source of news informa-
tion than other sources (mean, 2.49), followed by satellite TV (2.32), radio (1.43),

222
Table 7.3 Descriptive Statistics (listwise; n = 1,948)
Variable n Minimum Maximum Mean SD
Mobilizing ideas and attitudes, efficacy, and morality

Fatalism, anomie, grievances
Too many rules and 2,972 0 1 0.20 0.40
regulations
Too few rules and regulations 2,972 0 1 0.36 0.48
Mistrust of government 3,044 1 4 2.55 0.99
Dissatisfaction with govern- 2,996 1 10 5.47 2.19
ment performance
Economy run for big interests 2,963 0 1 0.47 0.50
Ideological framings
Liberalism index 2,746 0.89 3.03 1.97 0.31
Development reduces faith 3,078 1 4 1.87 1.05
in Allah
Fundamentalism 2,560 –4.19 1.21 0.00 1.00
Religiosity
Mosque attendance 3,140 1 7 4.87 2.43
Prayer daily 3,127 1 6 5.62 0.99
Self-rated religiosity (10, very 3,057 1 10 7.28 1.72
religious)
Importance of God (10, very 3,138 1 10 9.75 0.98
important)
Personal efficacy and dysphoric emotions
Perceived control 2,996 1 10 6.67 2.17
Freewill (10) versus fate (1) 3,069 1 10 3.59 2.80
Insecurity 3,073 1 4 3.04 0.98
Xenophobia 3,037 0 1 0.90 0.29
Morality
Moral flexibility 3,139 1 9.33 1.44 0.76
Mediums of communicative power
Source of informa- 3,139 1 4 2.49 1.16
tion: domestic TV
Source of information: radio 3,137 1 4 1.43 0.93
Source of information: 3,138 1 4 2.32 1.20
satellite TV
Paper, Internet, mobile phone 3,083 1 4 1.39 0.63
Antigovernment newspapers 3,143 0 1 0.11 0.32
Demographics
Male 3,143 0 1 0.48 0.50
Never Married 3,143 0 1 0.18 0.38
Impressionable years 3,143 0 1 0.64 0.48
Christian 3,143 0 1 0.04 0.20
Socioeconomic status 3,053 –1.90 3.10 0.0 1.00
Urban 3,143 0.00 1.00 0.45 0.50
Organizational membership 3,136 –0.24 9.25 0.00 1.00
SD, standard deviation.
223
and newspapers, Internet, or mobile phones (1.39). Only 11% reported using anti-
government newspapers over progovernment or none as a source of information.
64% lived their impressionable years under Mubarak. Finally, between 95% and
97% reported not belonging to a religious organization, political party, or humani-
tarian organization (individual indicators not shown in Table 7.3).
Table 7.4 shows a correlation matrix between all variables in the analysis.
Almost all the attitudinal variables are linked significantly to participation rate,
except for the measure of anomie (too few rules), the perception that the economy
runs for big interests, daily prayer, and self-described as religious. Among the
mediums of communicative power, only reliance on domestic TV is not linked to
participation. Reliance on radio is linked negatively to participation, but satellite
TV, newspapers, the Internet, mobile phones, and opposition newspapers corre-
lated positively with participation. Finally, among demographics, only being male,
not married, socioeconomic status, impressionable years under Mubarak, and liv-
ing in urban area correlated significantly with participation.
Findings that may enhance validity of the measures are that the relationships
among some of the variables are in directions that either make intuitive sense
or are consistent with the literature. For example, the liberalism index and fun-
damentalism are linked negatively (r = –.191). Furthermore, the liberalism index
correlated positively with perceived control (r = .087) and free will (r = .277), nega-
tively with insecurity (r = –.105) and xenophobia (r = –.070), and positively with
reliance on newspapers–Internet–mobile phones or opposition papers as sources
of information (r = .068 and r = .114, respectively); more Christians than Muslims
(r = .115), socioeconomic status (r = .137), and urban living. Fundamentalism, in
contrast, displays just the opposite relationships with these variables: negatively
with perceived control (r = –.042, not significant) and free will (r = –.083), posi-
tively with insecurity (r = .103) and xenophobia (r = .142), negatively with reliance
on newspapers–Internet–mobile phones or opposition newspapers as sources of
information (r = –.150 and r = –.081, respectively), fewer Christians than Muslims
(r = –.084), and negatively with socioeconomic status (r = –.098) and urban living.
These relationships are consistent with previous research on fundamentalism
(Moaddel & Karabenick, 2008, 2013) and with profiles of people with extremist
orientations as proposed in MSR theories.
To assess the significance of each of the predictors of participation, four mul-
tiple regression models are estimated and the results are reported in Table 7.5. The
first three models assess the linkages of (a) mobilizing ideas and attitudes, efficacy,
and morality; (b) mediums of communicative power; and (c) demographic and
organizational variables with participation separately. The fourth model includes
all these variables.
In the first model, the analysis of variance (ANOVA) shows a strong main
effect (F22, 1989 = 25.49; adjusted R2 = 0.21). The regression estimates indicate
that the measures of fatalism, anomie, and mistrust of government are linked
positively to participation. Among the interaction variables, only the interaction
between government dissatisfaction and mistrust of government is linked posi-
tively to participation. Measures of alternative ideological framings—liberalism

Table 7.4 Correlation Coefficients between Participation Rate and Predictors (Listwise Deletion, n = 1,945)
Variable Participation rate
1 2 3 4 5 6 7 8 9 10 11 12 13 14
c
Too many rules .144
Too few rules .011 –.374c
Mistrust of .130c .141c .133c
government
Government .067b .173c .039 .164c
dissatisfaction
Economy for big .032 .182c –.042 .224c .206c
interests
Liberalism index .146c .022 –.093c .046a –.048a –.005
Development .232c .097c –.066b .004b .018 .004 .063b
reduces faith
Fundamentalism .045a .022 –.022 –.053a –.039 –.213c –.191c .054a
Mosque attendance .104c .055b .092c .052a .017 .020 –.115b .005 .008
c
Pray daily –.022 –.076 .056b –.015 –.025 –.104c –.026 –.043 .067b .085c
b
Religious person –.035 –.058 .030 –.025 –.118b –.008 .030 –.098c .007 –.011 .290c
c
Importance of God –.048a .030 .022 .083c .003 .112c –.045a –.011 .080 –.002 –.002 .026
Perceived control .106c –.094c .061b –.046a –.144c –.076c .087c –.029 –.042 .058b .014 .112c .040
Freewill vs. fate .230a .004 –.014 .018 –.047a –.008 .277c .115c –.083c –.083c –.020 .035 –.086c .147c
225
Insecurity .130c .173c .068b .086c .194c .111c –.105c .205c .103c –.021 –.020 –.100c –.002 –.198c
c
Xenophobia –.153c –.077 .006 –.129c –.111c –.066b –.070b –.025 .142c –.047a .063b .088c –.026 .050a
Moral flexibility .255c .041 –.058b .081c .051a .116c .024 .231c .077c .047a –.039 –.108c –.004 –.025
Domestic TV –.020 –.043 .063b –.106c –.012 –.081c –.022 –.025 –.018 –.004 .093c .064b .024 .024
a c
Radio –079c .051 .090c –.046a .083c –.021 –.106c –.094c –.027 .077 .053a –.016 –.032 –.034
c
Satellite TV .054a .037 –.011 .046 .095c –.019 .047a .056b –.031 .088 .097c –.041 .006 .016
c
Paper–Internet– .219c .114c .086c .120c .128c .056b .068b –.001 –.150c .170 .032 –.019 –.041 .006
mobile phone
Against government .199c .128c –.007 .116c .066b .045a .114c .034 –.081c .108c .021 –.034 .024 .012
paper
Male .149c .000 .037 .052a .034 .024 –.026 –.004 –.026 .583c –.041 –.038 –.054a .063b
a a
Not married .105c .029 .034 .076c .014 .053 .024 –.005 –.044 .072c –.122c –.096c –.019 .020
Impressionable .094c .020 .022 .097c –.034 –.008 .037 .026 –.010 –.034 –.067b –.081c .042 .053a
years
Christian .034 .006 .052a .056b –.052b .020 .115c –.050a –.084c .107c –.334c –.004 .019 .048a
Socioeconomic .236c .083c .061b .131c .038 .063b .137c .117c –.098c .089c .059b –.024 .074c .131c
status
Urban .144c .077c –.051a .059b .094c .062b .193c .014 –.102c .036 –.063b –.094c –.058b –.049a
Organization .007 –.014 –.019 .078c –.009 –.013 .034 –.103c .078c .001 .031 –.033 –.029 .008
member
(continued)
226
Table 7.4 Continued
Variable Participation rate
15 16 17 18 19 20 21 22 23 24 25 26 27 28 29
Insecurity .038
Xenophobia –.050a –.084c
Moral flexibility .050a .023 –.065b
Domestic TV .012 .104c –.021 –.064b
Radio –.136a .001 –.039 –.058b .072c
Satellite TV .020 .002 –.034 –.036 –.018 .018
Paper–Internet– .006 .056b –.190c –.037 –.013 .061b .197c
mobile phone
Against government .041 .021 –.151c .002 –.057b .056b .153c .490c
paper
Male .011 .014 –.015 .008 –.046a .026 .083c .169c .125c
Not married .032 .004 –.079c –.007 –.055a –.038 .032 .194c .067b .133c
Impressionable .027 .013 .008 .026 –.058b –.120c .047a .095c .050a –.106c .314c
years
Christian .039 –.002 –.054a .001 –.023 –.060b –.022 .018 –.016 .009 .006 –.016
Socioeconomic .112c .069b –.138c .008 –.033 –.054a .221c .471c .316c .072c .157c .173c .032
status
Urban .094c . 013 –.217c .011 –.047a .004 .060a .176c .176c –.009 .033 –.094c .074c .178c
Organization .113c –.001 .015 –.001 –.111c –.011 –.065b –.007 –.030 .043 –.024 .026 –.001 –.003 –.056b
member
a
p < .05.
b
p < .01.
c
p < .001.
227
Table 7.5 Results of the Regression Estimates of the Predictors of Participation Rate
Model 1 Model 2 Model 3 Model 4
Predictors B value SE t Value B value SE t Value B value SE t Value B value SE t Value
Constant –1.41 0.83 –1.70 1.59 0.17 9.33c 1.80 0.11 16.44c –1.57 0.88 –1.79
Mobilizing ideas and attitudes, efficacy, and morality
Fatalism, anomie, grievances
Too many rules 0.60 0.15 3.90c 0.52 0.15 3.46c
Too few rules 0.32 0.13 2.53b 0.25 0.13 1.94a
Mistrust of 0.18 0.06 3.12b 0.12 0.06 2.10a
government
Government 0.04 0.02 1.57 0.03 0.02 1.14
dissatisfaction
Mistrust × dis- 0.05 0.02 2.00a 0.04 0.02 1.68
satisfaction
interaction
Economy for spe- –0.19 0.12 –1.58 –0.20 0.12 –1.67
cial few
Special few × mis- 0.23 0.12 1.90 0.18 0.12 1.54
trust interaction
Special few × 0.31 0.05 0.60 0.04 0.05 0.72
dissatisfaction
interaction
Ideological Framings
Liberalism index 0.85 0.18 4.75c 0.50 0.18 2.75b
Development 0.34 0.05 6.25c 0.30 0.05 5.49c
reduces faith
Fundamentalism 0.17 0.06 2.99b 0.23 0.06 3.97c
(continued)
228
Table 7.5 Continued
Religiosity
Mosque 0.11 0.02 4.55c –0.01 0.03 –0.42
attendance
Prayer daily –0.01 0.06 –0.07 0.03 0.06 0.46
Self–described –0.01 0.03 –0.38 0.02 0.03 0.65
religious
Importance of –0.15 0.06 –2.70b –0.13 0.05 –2.48b
God
Personal efficacy and dysphoric emotions
Perceived control 0.14 0.03 5.34c 0.12 0.03 4.44c
(fatalism)
Freewill vs. fate 0.15 0.02 7.58c 0.14 0.02 7.15c
Fatalism × freewill 0.18 0.05 3.40c 0.16 0.05 3.24c
interaction
Insecurity 0.20 0.06 3.45c 0.16 0.06 2.90b
Fatalism × insecu- 0.24 0.15 1.66 0.26 0.14 1.84
rity interaction
Xenophobia –1.06 0.19 –5.50c –0.79 0.20 –4.00c
Moral flexibility 0.66 0.07 9.37c 0.69 0.07 9.97c
Mediums of communicative power
Domestic TV 0.03 0.04 0.87 0.02 0.05 0.45
Radio –0.16 0.05 –3.38c –0.11 0.06 –1.88
Satellite TV 0.07 0.04 1.96a –0.04 0.05 –0.91
Paper–Internet– 0.72 0.08 8.79c 0.46 0.10 4.41c
mobile phone
Against govern- 0.88 0.16 5.42c 0.49 0.18 2.71b
ment papers
229

Demographics
Male 0.64 0.09 7.06c 0.61 0.14 4.53c
Never married 0.37 0.12 2.98b 0.16 0.15 1.10
Impressionable 0.35 0.10 3.53c 0.30 0.12 2.43a
years
Christian 0.24 0.22 1.12 0.19 0.30 0.65
Socioeconomic 0.45 0.05 9.55c 0.19 0.06 2.98b
status
Socioeconomic –.03 .04 –0.96 –0.12 0.04 –2.96b
status2
Urban 0.71 0.09 7.84c 0.41 0.11 3.54c
Organizational 0.06 0.04 1.33 –0.02 0.05 –0.34
membership
R values R = 0.47 R2= 0.22 R2ad = 0.21 R = .263 R2 = 0.07 R2ad = 0.07 R = 0.32 R2 = 0.10 R2ad = 0.10 R = 0.53 R2 = 0.28 R2ad = 0.27
F values (ѵ1, ѵ2) 22,1989 = 25.49 53,064 = 45.51 83,032 = 43.34 351,909 = 21.07
a
p < .05.
b
p < .01.
c
p < .001.
230
index, development reduces faith in God (secular cognition), and fundamentalism

index—are also linked positively to participation. Among religiosity variables, only
two have significant effects: mosque attendance having a positive relationship and
the importance of God having a negative relationship with participation.
Both measures of personal efficacy—perceived control and the belief in free
will—are related positively to participation. The interaction between believing in
free will and the perception of too many rules (fatalism) is also linked positively
to participation. Measures of dysphoric emotions, on the other hand, had oppo-
site connections with participation; the feeling of insecurity is linked positively
whereas xenophobia is linked negatively to participation. Moral flexibility is also
linked significantly to participation.
The ANOVA for the second model also shows the strong main effect of the
mediums of communicative power (F5, 3064 = 45.51; adjusted R2 = 0.07). Among
the variables, reliance on radio as a source of information is linked negatively
to participation, whereas reliance on satellite TV, newspapers–Internet–mobile
phones, and opposition newspapers are connected positively to participation.
Finally, the third model estimates the regression coefficients of demographic and
organizational predictors of participation. To assess the inverted-U relationship
between socioeconomic status and participation, a variable measuring the square
of socioeconomic status is also included in the model. The ANOVA shows the
strong main effects of demographics on participation (F8, 3032 = 43.34; adjusted
R2 = 0.10). Regression estimates also show that people who were male, single, or
age 12 years or younger when Hosni Mubarak became president, had to a higher
socioeconomic status, and lived in urban areas participated in the revolutionary
movement more often than other respondents in the sample.5 Because the square
of the effect of socioeconomic status on participation was insignificant, this model
does not support the proposed inverted-U relationship between this variable and
participation.
These analyses show that attitudinal, personal efficacy, and morality variables
explain the largest proportion of variations in the dependent variable (21%), fol-
lowed by demographic (10%), and then mediums of communicative power (7%).
These differences may be attributed to the fact that the first categories of predictors
included more indicators than the other two. Nonetheless, they do demonstrate
that attitudinal and perceptual variables are important in predicting revolutionary
action.6
Finally, for the full model, the ANOVA indicates a strong main effect on par-
ticipation (F35, 1909 = 21.07; adjusted R2 = 0.27). The regression estimates also show
that all the coefficients but five that were significant in the three separate models
have remained significant in the full model as well. These five variables are the
interaction between dissatisfaction with government and mistrust of government,
mosque attendance (which correlated highly with male status), reliance on domes-
tic TV and radio as sources of news information, and being single. On the other
hand, the full model estimates show that the square of the socioeconomic sta-
tus measure is linked negatively to participation, hence supporting an inverted-U
relationship between participation and socioeconomic status.

231
That the significance of the rest of the variables remains the same in sepa-
rate and full models is indicative of the robustness of the analytical framework
advanced in this chapter.
Discussion
The foregoing empirical assessments of the hypotheses drawn from both PROP
and MSR theories open up a new insight into the process of revolutionary develop-
ment. This insight suggests rethinking (a) the predictors of participation, (b) the
conception of the individual participants that has informed these theories, and
(c) the unifying element in the revolutionary process.
Mobilizing Ideas and Attitudes, Personal Efficacy, and Morality

As PROP theories state, the more efficacious individuals—those who expressed
a greater level of perceived control or believe more strongly in free will—tend to
participate more often in the revolutionary movement. Alternatively, people who
feel more powerless or believe more strongly in fate participate less often. Lower
is also the participation rate of xenophobic individuals. Consistent with MSR theo-
ries, fatalistic individuals (those who feel constrained by too many governmen-
tal rules and regulations) and anomic individuals (those who feel there are not
enough rules and regulations) participate in revolutionary movements more often
than those who believe there is the right amount of rules and regulations, as do
those who have a stronger feeling of insecurity. Supporting both clusters of theo-
ries is the interaction between fatalism and the belief in free will. That is, people
who feel constrained by governmental rules and regulations (fatalists) and are
stronger believers in free will tend to participate more often than those who have
the same beliefs but are not fatalists. This indicates that efficacious individuals
rebel when they feel constrained by the government more often than when they
do not have such a feeling.
The linkages between grievances and revolutionary action provide inconsistent
support for PROP theories. That dissatisfaction with government performance
and the perception that the economy runs for the benefits of a special few have
no significance linkage with participation supports the PROP theories’ premise
that grievances by themselves are not enough for mobilization to occur. Mistrust
of government, on the other hand, is linked positively to participation, which is
not consistent with these theories. This linkage can be interpreted differently,
however. Mistrust of government may not indicate just discontent; it may also
suggest the framing of the government as untrustworthy and thus illegitimate. If
this interpretation is correct and given the positive linkage of liberalism and fun-
damentalism indices with participation, this study found considerable evidence
in support of the framing perspective. In addition, it showed that multiple fram-
ings may be at work in a revolutionary movement, as people with stronger lib-
eral or fundamentalist attitudes participated more often than those who displayed
weaker attitudes. What is more, because such contradictory value orientations as

232
liberalism, secular cognition, and fundamentalism are linked to participation, this

finding lends credence to the notion that revolution has different meanings for
different participants (Furet 1981; Moaddel 1992).
One of the interesting findings of this study is the negative relationship
between participation and the importance of God in life. This relationship holds
even if the variables that are linked significantly to the latter (fundamentalism
index, mistrust of government, and economy runs for a special few) are removed
from the regression model. It may thus be speculated that people who attach a
greater importance to God in life tend to delegate the punishment of an authori-
tarian ruler (i.e., evil-doer) to God and, as a result, opt to participate less often in a
revolutionary movement against him.
Finally, contrary to PROP theories, which hypothesized a positive connection
between moral commitments and participation, the current analysis revealed just
the opposite relationship; moral flexibility is linked positively (or morality, nega-
tively) to participation. Although this finding is consistent with MSR theories,
the mechanism that explains this linkage, however, may be related to the mun-
dane process of political dynamics than to the belief in a teleology of revolution-
ary developments, as Arendt (1963) claims. Accordingly, moral flexibility may be
an outcome of the experience of revolutionary activism. Revolution makers often
face the regime’s atrocities; comrades are arrested, beaten, tortured, or even killed
by security forces. Compared with the general public, they may display stronger
awareness of the prevalence of corruption among high-ranking officials and may
have more extensive knowledge of the government’s deceptive tactics in spread-
ing misinformation to cover up dishonesties. At the same time, repression tends
to arouse the feeling of revenge for the fallen comrade; economic corruption, the
desire to seize or destroy forcibly the ruling elite’s property; and deception, the
need to launch propaganda campaigns against the government. All these in turn
may relax or modify the moral standards of the revolution makers. The more
extensively the state engages in acts of violence against the opposition, for exam-
ple, the less the opposition activists remain committed to the immorality of vio-
lence. It may be true, as Nepstad and Smith (2001, p. 173) suggested, that “moral
outrage is a logical reaction to the torture, disappearances, and assassinations of
innocent civilians and to the lies disseminated by a government to cover its role
as an accomplice to these atrocities.” It may also be true that such immoral acts
undermine the morality of the revolutionaries. As the current analysis showed,
those who participated more often in the revolutionary movement tended to be
less strict in their moral standards than those who did so less often.
This relationship is consistent with the notion that politics tends to generate
wickedness, providing insights into why many revolutionary leaders end up prac-
ticing the same immoral behaviors against which they fight so passionately. One
may surmise that, by the time they seize power, these leaders have resolved that it
is acceptable, for the “sake of the revolution,” to lie, deceive, steal, or even to kill.7
The dictum no organization = lost revolution may be labeled a hallmark of PROP
theories. Yet, neither organizational membership nor mosque attendance (as a

233
proxy for organization) was linked significantly to participation. Also failing

to play a leading role in the revolution were groups such as the well-organized
Society of the Muslim Brothers or dedicated Muslim extremists, who had tried to
overthrow the government in past decades. Moreover, even if these groups had
managed to assume a leadership role in the revolutionary movement, this would
have alerted the regime and its international allies, including the United States,
about the prospect of the rise of an anti-West Islamic government and thus uni-
fied a counterrevolutionary coalition, providing justifications for a more forceful
repression of the opposition. On the other hand, the image of a movement being
initiated by a group of computer-savvy Egyptians, displaying a nonideological and
pragmatic orientation, calling for the formation of a democratic and transparent
government, and demanding a dictator to step down was far removed from any-
thing resembling bearded Islamic extremists trying to establish an anti-Western
totalitarian theocracy. Only after the fall of the regime were the Muslim Brothers
able to use their vast organizational resources to succeed in the post-Mubarak elec-
tions. It may thus be argued that an organized network of any opposition group is
useful in seizing power only under the constraints set by the existing political and
discursive structures. The Muslim Brothers were unable to lead the revolutionary
movement against the regime, but after it was overthrown, they could reap sub-
stantial political benefits, thanks to their organizational prowess.
On a more abstract level, one may suggest, it is not the mere presence of an
organization per se that is vital for revolutionary mobilization; rather, it is an orga-
nization’s function as a medium of communicative power or a channel for trans-
fer of resources among revolution makers that underlines its importance. In the
absence of an organization, these functions may be taken over by other means.
In the case of Egypt, for example, the wireless network appears to have enabled
revolution maker A to communicate with revolution maker B relatively uncon-
strained by state power. Moreover, the innovative use of the Internet reveals the
significance of an element of surprise in the success of a revolutionary movement.
Accordingly, the victory of revolution makers may depend on their ability to invent
new mobilizational tactics that cannot be combated easily by the government’s
repressive routines or that pose problems that cannot be readily understood in
terms of the categories known to the regime’s counterrevolutionary strategists.
Organizations, whether preexisting or formed by revolutionary activists, are use-
ful insofar as they facilitate the flow of information among activists, create and
reinforce connectivity among the potential revolution makers, mobilize emotional
energy against the ruling regime, and make it possible to launch attacks on the
regime where it appears most vulnerable. Not all organizations—even those that
are decidedly committed to revolutionary change—are capable of performing
such functions.
Demographic Variables
The linkages between demographics and participation support some of the hypoth-
eses drawn from PROP theories. The greater participation of men compared with
women, people living in the urban areas compared with those in the rural areas,
and members of the middle class compared with the lower classes are consistent

234
with the notions that being male, living in urban areas, and having a middle-
class background are conducive to having access to more resources, information,
and knowledge about oppositional perspectives. Thus, insofar as the resource
availability–participation nexus is concerned, PROP theories are supported.
These findings may also be interpreted differently and lend support to MSR
theories. With regard to the extent to which one considers a lower degree of social
integration leading to a higher rate of participation, men may be less integrated
in societal customs or constrained by daily routines than women who tend to be
stronger observers of social norms and more immersed in the daily operations
of their household—cleaning, buying groceries, cooking, and taking care of the
children. Likewise, people living in urban areas may be characterized by weaker
bonds of social solidarity than people living in the rural areas. Finally, the higher
rate of revolutionary participation by the members of the middle class compared
with the members of the upper classes may be indicative of a feeling of relative
deprivation—a factor consistent with MSR theories.
None of these theories, however, account for the higher participation of the
individuals who lived their impressionable years under the ruling regime. Nor do
they explain why these individuals, as well as men, urban residents, and members
of the middle class, should be more revolutionary than others. Greater resources
and access to varied sources of information are important in enhancing the capac-
ity to act, but these factors do not explain one’s desire or willingness to partici-
pate in a revolutionary movement. It can be argued that people with a stronger
tendency to participate in a revolutionary movement are those who have a stron-
ger claim of ownership of their country and consider themselves more forcefully
entitled to social distinction. It may thus be proposed that stronger claimants of
national politics are more often men than women in a patriarchal social context,
urban residents than rural in a nonegalitarian rural–urban division, members of
the middle class who consider themselves more the backbone of the country’s
economy than other classes, and people who come of age under the ruling regime
than those belonging to previous generations.
Reconciling Contradictory Explanations: The Notion

of Manifold Individuals
Some of this study’s findings are consistent with PROP theories, whereas others
align with MSR theories. Individuals with a stronger perception of self-efficacy, as
proposed in PROP theories, and those with a stronger feeling of fatalism and inse-
curity, as proposed in MSR theories, for example, are linked to a greater participa-
tion rate. The significance of the Internet as a network of relatively anonymous
and isolated individuals in shaping revolutionary action is not quite consistent
with a nonanonymous vision of empowered, mobilized, and interconnected
individuals portrayed by PROP theories. Nor is this significance consistent with
MSR theories. Finally, the importance of such diverse framings as liberalism and
fundamentalism in predicting participation is consistent with both progressive
and retrogressive visions of revolution portrayed by PROP and MSR theories,
respectively.

235
However, the empirical data provide an image of revolutionary participants

who are conflicted ideologically and psychologically, are perceptually diverse, and
are organized only anonymously. This image does not quite conform to the one
projected by either theory. This is true because both PROP and MSR theories
tend to presume that revolution makers are monolithic actors with a coherent,
homogeneous, and consistent set of beliefs and value orientations. The difference
is that, in PROP theories, such individuals are cognitively liberated, motivated
by feelings of empowerment and moral outrage, and relieved of any dysphoric
emotion, whereas, in MSR theories, they are disoriented, fatalist, and insecure
individuals who have gained full assurance of their righteousness by commitment
to a revolutionary ideology and a reliance on totalitarian leaders for the verification
of their values and attitudes.
The Egyptian case demonstrates that revolution makers, in contrast, display
manifold and heterogeneous characteristics and even contradictory orientations.
Some are self-actualized, efficacious, and cognitively liberated, whereas others
feel insecure, anomic, and fatalist. The same individuals may even carry a mul-
titude of contradictory attitudes toward the ruling regime, perceptions of self-
empowerment, and assessments of security. Thus, the notion of the monolithic
individual cannot be sustained empirically and must be abandoned. The alterna-
tive conceptualization, however, by introducing a manifold image of revolutionary
participants, cannot be reconciled easily with the existing theories of revolution.
This is true because both PROP and MSR theories remain remarkably lucid and
comprehensible insofar as their explanations rest on a unified set of interrelated
attributes of monolithic individuals. But, the notion of manifold, heterogeneous,
or contradictory actors begs the question about the source of their unity. After
all, Egyptian revolutionary actors did display a remarkable unity and appear as a
homogeneous whole in mass demonstrations in Cairo’s Tahrir Square and many
other parts of the country.
If the driving ideology is diverse; individual efficacy coupled with fatalism,
anomie, and insecurity; the revolutionary role of the existing organizations of
the opposition secondary; and a cohesive culture of resistance wanting; then the
centripetal force that generates the unity of purpose among the diverse partici-
pants in calling for the overthrow of the ruling regime must be located not within
but outside the dynamics of the revolutionary movements. Considering Stepan’s
(1985, p. 338) statement that the character of the state affects “the evolution of
opposition politics,” this unifying force may be an outcome of the interactions
between civil society and a monolithic state. A pluralistic political structure is
manifested in the presence of elite diversity, crisscrossing political institutions,
and differentiated positions of power that functions as a mechanism for checks
and balances. This form of the state is said to be amenable to public demand
and malleable to reform as a result of pressures from below. Well-anchored in
the society, a pluralistic state would be less likely to face a de facto revolutionary
coalition against it.
A monolithic state, on the other hand, having a homogeneous pyramidal politi-
cal structure controlled by the all-powerful dictator, is more vulnerable to revolu-
tionary challenges from below. The authoritarianism of the ruler undermines the

236
autonomy of office holders, filters out the more independent-minded politicians

from gaining access to positions of power, promotes yes-men in state bureaucracy,
and propagates a conformist culture. Furthermore, with state officials lacking
autonomy to make important political decisions, the ruler is faced with micro-
managing the daily affairs of the state—a key factor that makes him or her even
more vulnerable to revolutionary challenges from below. Having a uniform struc-
ture, the monolithic state is ill-equipped to forge alliances with or reflect diverse
interests emanating from civil society. The state ideology being closely identified
with the ruling elite loses its broader universalistic appeals and thus fails to tie
the state to the broader public. Incapable of receiving and processing information
autonomously through its mediums of communicative power, the ruling authori-
ties tend to underestimate the extent of mass discontent in society. As a result,
through time, the monolithic state becomes increasingly detached from the popu-
lation it attempts to control and regulate, with the only mediating institutions
being the military and ever-present secret police. As the revolutionary challenge
from below gains critical mass, even the military often opts to stay neutral rather
than risking a civil war defending an increasingly isolated and unpopular dictator.
One may thus suggest that variation along a pluralistic–monolithic continuum
explains the resilience of Arab monarchies in places such as Jordan and Morocco,
and the relative vulnerability of the republican systems in other Arab-majority
countries. Although both systems are repressive and nonegalitarian, the monar-
chies appear to have a more differentiated political structure than republican sys-
tems, which rest decidedly on the whim of a single dictator and have thus become
a clear target of attack from diverse revolutionary movements.
Acknowledgment
This chapter is based a research project supported by a grant from the Office of
Naval Research (N00014-09-1-0985).
Notes
1. According to Arendt (1963, pp. 262–263):
[T]‌the French Revolution also gave rise to a new figure of professional revo-
lutionist, who spent his life not in revolutionary agitation, for which there
existed but few opportunities, but in study and thought, in theory and debate
… [His] essentially theoretical way of life was spent in the famous libraries
of London and Paris, or in the coffee house of Vienna and Zurich, or in the
relatively comfortable and undisturbed jails of the various anciens regimes.
2. According to 2011 survey data used in this chapter, less than 2% of Egyptians
defined themselves as “Arabs, above all,” whereas 51% defined themselves as
Egyptians and about 47% as Muslims, above all.
3. PROP theorists have conceptualized this variable in terms of “biographical
availability … as the absence of personal constraints that may increase the costs

237
and risks of movement participation, such as full-time employment, marriage, and

family responsibilities” (McAdam, 1986, p. 70; see also Beyerlein & Hipp, 2006).
The “absence of personal constraints” means, practically, in this context “weaker
social ties.”
4. In writing this interpretation, I benefited by comments from Abdul-Hamid
Abdul-Latif, Director of Egyptian Research and Training Center, and Asma Ahmed,
Cairo, Egypt.
5. Although not shown in this Table 8.5, the youth bulge (measured as a dummy
variable differently for ages 18–34 years = 1, 18–28 years = 1, or 18–24 years = 1) had no
significant effect on participation, but age had a negative effect (significant at 0.05),
indicating that being younger is conducive to participation. Thus, there was little sup-
port for McAdam’s (1986) notion of biographical availability.
6. If the three sets of predictors are entered into analysis in a stepwise fashion—
(a) demographics, (b) demographics and measures of mediums of communicative
power, and (c) demographics, mediums of communicative power, and attitudes, effi-
cacy, and framing—the adjusted R2 values would be 0.10, 0. 12, and 0.27, respectively.
7. Psychologists have related support for political violence to greater moral cer-
tainty produced by higher religiosity (Shaw & Zárate, 2011).
References
Abrahamian, E. 1982. Iran between two revolutions. Princeton, NJ: Princeton
University Press.
Acevedo, G. A. (2005). Turning anomie on its head: Fatalism as Durkheim’s con-
cealed and multidimensional alienation theory. Sociological Theory, 23(1), 75–85.
Al-Jazeera, February 14, 2011. Timeline: Egypt’s revolution: A chronicle of the revolu-
tion that ended the three-decade-long presidency of Hosni Mubarak [On-line].
February 14. Retrieved January 16, 2013.
Altemeyer, B., & Hunsberger, B. (2004). A revised religious fundamentalism
scale: The short and sweet of it. International Journal for the Psychology of Religion,
14, 47–54.
Anderson, B. (1983). Imagined communities: Reflections on the origin and spread of
nationalism. London: Verso.
Arendt, H. (1958). The origin of totalitarianism. Cleveland, OH: Meridian Books.
———. (1963). On revolution. New York: Viking Press.
Armstrong, J. (1982). Nations before nationalism. Chapel Hill: University of North
Carolina Press.
Ayubi, N. N. M. (1991). Political Islam: Religion and politics in the Arab world.
London: Routledge.
Bayat, M. (1991). Iran’s first revolution: Shi’ism and the constitutional revolution of 1905–
1909. New York: Oxford University Press.
Bell, D. (1973). The coming of post-industrial society. New York: Basic Books.
Besnard, P. (1993). Anomie and fatalism in Durkheim’s theory of regulation.
In S. P. Turner (Ed.). Emile Durkheim: Sociologist and moralist (pp. 169–190).
New York: Routledge.

238
Beyerlein, K., & Hipp, J. R. (2006). A two-stage model for a two-stage process: How
biographical availability matters for social movement mobilization. Mobilization,
11, 219–240.
Burnstein, E., Crandall, C., & Kitayama, S. (1994). Some neo-Darwinian decision
rules for altruism: Weighing cues for inclusive fitness as a function of the bio-
logical importance of the decision. Journal of Personality and Social Psychology, 67,
773–789.
Castells, M. (2007). Communication, power and counter-power in the network soci-
ety. International Journal of Communication, 1, 238–266.
Chen, H., Lu, Y. K., Suen, W. (2012). The power of whispers: A theory of rumor, commu-
nication and revolution [On-line]. A working paper. Available: http://www.sv.uio.
no/esop/english/research/news-and-events/events/conferences-and-seminars/
normac-papers/chen.pdf. Retrieved September 13, 2012.
Citrin, J. (1974). Comment: The political relevance of trust in government. American
Political Science Review, 68(3), 973–988.
———. (1977). Political alienation as a social indicator. Social Indicators Research, 4,
381–419.
Collins, R. (2001). Social movements and the focus of emotional attention. In J.
Goodwin, J. M. Jasper, & F. Polletta (Eds.). Passionate politics (pp. 27–45). Chicago,
IL: University of Chicago Press.
Connor, W. (1994). Ethno- nationalism: The quest for understanding. Princeton,
NJ: Princeton University Press.
Craig, S. C., & Maggiotto, M. A. (1981). Political discontent and political action. Journal
of Politics, 43(2), 514–522.
Davies, J. C. (1969). The J-curve of rising and declining satisfaction as a cause of
some great revolutions and a contained rebellion. In H. D. Graham & T. R. Gurr
(Eds.). Violence in America (pp. 671–709). New York: Signet.
Davis, N. J., Robinson, R. V. (2010). The roots of political activism in six Muslim-majority
nations. Paper presented at the Workshop on Theoretical and Methodological
Issues in the Study of Values in Islamic Countries, Cairo, Egypt, May 16–18, 2010.
de Figueiredo, R. J. P., Jr., & Elkins, Z. (2003). Are patriots bigots? An inquiry into the
vices of in-group pride. American Journal of Political Science, 47(1), 171–188.
Dessouki, A. E. H. (1982). The Islamic resurgence: Sources, dynamics, and implica-
tions. In A. E. H. Dessouki (Ed.). Islamic resurgence in the Arab world (pp. 3–31).
New York: Praeger.
Dohrenwend, B. P. (1959). Egoism, altruism, anomie, and fatalism: A conceptual
analysis of Durkheim’s types. American Sociological Review, 24(4), 466–473.
El-Ghobashy, M. (2011). The praxis of the Egyptian revolution [On-line]. Middle East
Research and Information Project, 41. Available: http://www.merip.org/mer/
mer258. Retrieved January 16, 2013.
Fishman, J. A. (1972). Language and nationalism. Rowley, MA: Newbury House.
Fishman, J. A., Gertner, M., Lowy, E. G., & Milan, W. G. (1985). The rise and fall of the
ethnic revival: Perspectives on language and ethnicity. Berlin: Mouton.
Freitag, M., & Bühlmann, M. B. (2009). Crafting trust: The role of political institu-
tions in a comparative perspective. Comparative Political Studies, 42(12), 1537–1566.
Furet, F. (1981). Interpreting the French Revolution. New York: Cambridge
University Press

239
Gamson, W. A. (1968). Power and discontent. Homewood, IL: Dorsey Press.

———. (1975). The strategy of social protest. Homewood, IL: Dorsey Press.
———. (1992). Talking politics. New York: Cambridge University Press.
Gellner, E. (1965). Thought and change. Chicago, IL: University of Chicago Press.
Gershoni, I., & Jankowski, J. P. (1995). Redefining the Egyptian nation, 1930–1945.
Ghadbian, N. (2000). Political Islam and violence. New Political Science, 22, 77–88.
Goodwin, J., Jasper, J. M., & Polletta, F. (Eds.). 2001. Passionate politics: Emotions and
social movements. Chicago, IL: The University of Chicago Press.
Gross, D. (2011). Google, Twitter help give voice to Egyptians [On- line]. February
1. Available: http://articles.cnn.com/2011-02-01/tech/google.egypt_1_twitter-
users-Internet-access-google?. Retrieved August 27, 2012.
Gurr, T. (1970). Why men rebel. Princeton, NJ: Princeton University Press.
Hechter, M. (1975). Internal colonialism: The Celtic fringe in British national development
1536–1966. Berkeley: University of California Press.
Huntington, S. (1968). Political order in changing societies. New Haven, CT: Yale
University Press.
Ibrahim, S. E. (1980). Anatomy of Egypt’s militant Islamic groups. International
Jasper, J. M. (1999). The art of moral protest: Culture, biography and creativity in social
movements. Chicago, IL: The University of Chicago Press.
Johnson, C. (1964). Revolution and the social system. Stanford, CA: Stanford
University Press.
———. (1966). Revolutionary change. Boston, MA: Little Brown.
Kinder, D. R. (1998). Communication and opinion. Annual Review of Political Science,
1, 167–197.
Kirkpatrick, D. D. (2011). As Egypt protest swells, U.S. sends specific demands [On-
line]. New York Times. February 8. Retrieved January 16, 2013.
Korchmaros, J. D., & Kenny, D. A. (2001). Emotional closeness as a mediator of the
effect of genetic relatedness on altruism. Psychological Science, 12, 262–265.
Kornhauser, W. (1959). The politics of mass society. Glencoe, IL: Free Press.
Krueger, A. B., & Malečková, J. (2003). Education, poverty and terrorism: Is there a
causal connection? Journal of Economic Perspectives, 17, 119–144.
Kruglanski, A. W. (1989). Lay epistemics and human knowledge: Cognitive and motiva-
tion bases. New York: Plenum Press.
Lee, A.-R. (1992). Values, government performance, and protest in South Korea.
Asian Affairs, 18(4), 240–253.
Lee, S.-J. (2004). The rise of Korean youth as a political force: Implications for the
U.S.–Korea alliance. In R. C. Bush, S, Yanagi, and K. Scott (Eds.). Brookings
Northeast Asia Survey (pp. 15–30). Washington, DC: The Brookings Instituion.
Levi, M., & Stoker, L. (2000). Political trust and trustworthiness. Annual Review of
Political Science, 3, 475–507.
Lockwood, D. (1992). Solidarity and schism: The problem of disorder in Durkheimian and
Marxist sociology. Oxford: Oxford University Press.
Lukes, S. (1967). Alienation and anomie. In P. Laslett & W. G. Runciman
(Eds.). Philosophy, politics, and society (3rd series) (pp. 134–156). Oxford: Basil
Blackwell.

240
Mann, M. (1992). The emergence of modern European nationalism. In J. A. Hall &

I. C. Jarvie (Eds.). Transition to modernity: Essays in power, wealth, and belief (pp.
137–166). Cambridge: Cambridge University Press.
Martinez, L. M. (2005). Yes we can: Latino participation in unconventional politics.
McAdam, D. (1982). Political process and the development of black insurgency, 1930–1970.
———. (1986). Recruitment to high-risk activism: The case of Freedom Summer.
American Journal of Sociology, 92(1), 64–90.
McCarthy, J. D., & Zald, M. N. (1977). Resource Mobilization and Social
Movements: A Partial Theory. American Journal of Sociology, 82, 1212–1241.
Miller, W. E., Miller, A. H., & Schneider, E. J. (1980). American national election
studies data sourcebook, 1952–1978. Cambridge, MA: Harvard University Press,
p. 253.
Minkoff, D. (1997). Producing social capital: National social movements and civil
society. American Behavioral Scientist, 40(5), 606–619.
Moaddel, M. 1992. Ideology as episodic discourse: The case of the Iranian revolution.
American Sociological Review, 57(3), 353–379.
———. (1993). Class, politics, and ideology in the Iranian revolution. New York: Columbia
University Press.
———. (2005). Islamic modernism, nationalism, and fundamentalism: Episode and dis-
course. Chicago, IL: University of Chicago Press.
Moaddel, M., & Karabenick, S. (2008). Religious fundamentalism among young
Muslims in Egypt and Saudi Arabia. Social Forces, 86(4), 1675–1710.
———. (2013). Religious Fundamentalism in the Middle East: A Cross-National, Inter-
Faith, and Inter-Ethnic Analysis. Leiden: Brill.
Moaddel, M., Tessler, M., & Inglehart, R. (2008a). Foreign occupation and national
pride. Public Opinion Quarterly, 72(4), 677–705.
———. (2008b). Saddam Hussein and the Sunni insurgency: Findings from Iraqi
values surveys. Political Science Quarterly, 123(4), 623–644.
Muller E. N., Jukam, T. O., & Selligson, M. A. (1982). Diffuse political support and
anti-system political behavior: A comparative analysis. American Journal of
Political Science, 26, 240–264.
Nepstad, S. E. (1997). The process of cognitive liberation: Cultural synapses, links, and
frame contradictions in the U.S.–Central America peace movement. Sociological
Inquiry, 67, 470–487.
Nepstad, S. E., & Smith, C. (2001). The social structure of moral outrage in recruit-
ment to the U.S. Central America peace movement. In J. Goodwin, J. M. Jasper,
& F. Polletta (Eds.). Passionate politics: Emotions and social movements (pp. 158–175).
Oberschall, A. (1973). Social Conflicts and Social Movements. Englewood
Cliffs: Prentice-Hall.
Paige, J. (1971). Political orientation and riot participation. American Sociological
Review, 36, 810–820.
Paxton, P. (2002). Social capital and democracy: An interdependent relationship.

241
Pierce, R., & Converse, P. E. (1989). Attitudinal roots of popular protest: The French
upheaval of May 1968. International Journal of Public Opinion Research, 1, 221–241.
Piven, F., & Cloward, R. (1977). Poor people’s movements: Why they succeed and how they
fail. New York: Pantheon.
Popkin, S. L. (Ed.). (1991). The reasoning voter: Communication and persuasion in presi-
dential campaigns. Chicago, IL: University of Chicago Press.
Portes, A. (1998). Social capital: Its origins and applications in modern sociology.
Annual Review of Sociology, 22, 1–24.
Preston, J. (2011). Movement began with outrage and a Facebook page that give it an
outlet [On-line]. New York Times. February 5. Available: http://www.nytimes.com/
2011/02/06/world/middleeast/06face.html?pagewanted=all. Retrieved August
27, 2012.
Reuters. (2011). Timeline-protests in Egypt [Online]. February 3. Retrieved January
16, 2013.
Robinson, F. (1993). Islam and the impact of print. Modern Asia Studies, 27, 229–251.
Sachdeva, S., Iliev, R., & Medin, D. (2009). Shining saints and saintly sinners: The
paradox of moral self-regulation. Psychological Science, 20, 523–528.
Sageman, M. (2004). Understanding terror networks. Philadelphia: University of
Pennsylvania Press.
Salvaggio, J. L. (Ed.). (1989). The information society: Economic, social, and structural
issues. Hillsdale, NJ: Erlbaum.
Schuman, H., & Rieger, C. (1992). Historical analogies, generational effects, and atti-
tudes toward war. American Sociological Review, 57, 315–326.
Schussman, A., & Soule, S. (2005). Process and protest: Accounting for individual
protest participation. Social Forces, 1083.
Schwartz, D. C. (1973). Political alienation and political behavior. Chicago, IL: Aldine,
pp. 12–13.
Sears, D. O., & McConahay, J. B. (1973). The politics of violence; the new urban Blacks
and the Watts Riot. Boston: Houghton Mifflin.
Sewell, W. H., Jr. (1985). Ideologies and social revolutions: Reflections on the French
case. Journal of Modern History, 57(1), 57–85.
Shaw, M., Quezada, S. A., & Zárate, M. A. (2011). Violence with a conscience: Religiosity
and moral certainty as predictors of support for violent warfare. Psychology of
Violence, 1(4), 275–286.
Shingles, R. D. (1981). Black consciousness and political participation: The missing
link. American Political Science Review, 75, 76–91.
Sigelman, L., & Feldman, S. (1983). Efficacy, mistrust, and political mobiliza-
tion: A cross-national analysis. Comparative Political Studies, 16(1), 118–143.
Skocpol, T. (1979). States and social revolutions: A comparative analysis of France, Russia,
and China. Cambridge: Cambridge University Press.
Smith, A. D. (1998). Nationalism and modernism. London: Routledge.
Sniderman, P. M., Brody, R. A., & Tetlock, P. E. (1991). Reasoning and choice: Explorations
in political psychology. New York: Cambridge University Press.
Snow, D. A. (2004). “Framing Processes, Ideology, and Discursive Field.” In D. A.
Snow, S. A. Soule, and H. Kriesi, The Blackwell Companion to Social Movements
(pp. 380–412). Malden: Blackwell Publishing.

242
Snow, D. A., & Benford, R. D. (1988). “Ideology, Frame Resonance, and Participant
Mobilization.” In From Structure to Action, edited by Bert Klandermans, Hanspter
Kriesi, and Sidney Tarrow, vol. 1 of International Social Movement Research (pp.
197–217). Greenwich, Conn.: JAI Press.
Stepan, A. (1985). State power and the strength of civil society in the southern cone of
Latin America. In P. B. Evans, D. Rueschemeyer, & T. Skocpol. Bringing the state
back in (pp. 317–343). Cambridge: Cambridge University Press.
Swidler, A. (1986). Culture in action: Symbols and strategies. American Sociological
Review, 51(2), 273–286.
Tehranian, M. (1980). Communication and revolution in Iran: The passing of a para-
digm. Iranian Studies, 13(1–4), 5–30.
Tilly, C. (Ed). (1975). The formation of nation states in Western Europe. Princeton,
NJ: Princeton University Press.
———. (1978). From mobilization to revolution. Reading, MA: Addison-Wesley.
Turner, R. H. (1969). The public perception of protest. American Sociological Review,
34(6), 815–831.
van den Berghe, P. L. (1978). Race and ethnicity: A sociobiological perspective. Ethnic
and Racial Studies, 1(4), 401–411.
Wiktorowicz, Q. (Ed.). (2004). Islamic activism: A social movement theory approach.
Woltering, R. A. F. L. (2002). The roots of Islamic popularity. Third World Quarterly,
23, 1133–1143.
You, J.-S. (2012). Social trust: Fairness matters more than homogeneity. Political
Psychology, 33(5), 701–721.
Zald, M. N., & McCarthy, J. D. (Eds.). (1987). Social movements in an organizational
society. New Brunswick, NJ: Transaction.

243
APPENDIX Predictive Variables of the study
Attitudes, Efficacy, and Morality
Fatalism, Anomie, and Grievances

Fatalism versus anomie: These measures are constructed using two dummy vari-
ables based on the following question: Which of the following statements comes
closer to your sentiment: (1) there are just about the right amount of governmental
laws and regulations, (2) there are not enough governmental laws and regula-
tions to guide me, (3) I feel constrained by the number of governmental laws and
regulations.
A1 Fatalism: Too many rules. I feel constrained by the number of
governmental laws and regulations (= 1, 0 = otherwise).
A2 Anomie: Too few rules. There are not enough governmental laws and
regulations to guide me (= 1, 0 = otherwise).
A3 Mistrust of government: Do you have (1) a great deal of confidence, (2) quite
a lot of confidence, (3) not very much confidence, or (4) no confidence at
all in the government?
A4 Dissatisfaction with the government: People have different views about the
ideal way of governing this country. Here is a scale for rating how well the
government is doing: 1 means very good; 10 means very bad.
A5 Special few: Generally speaking, would you say that this country’s economy
is run for the benefit of all the people (coded as 0) or is run by a few big
influential people and organizations who are looking out for themselves
only (coded as 1)?
A6 Mistrust × dissatisfaction interaction = (mistrust –mean)
(dissatisfaction – mean)
A7 Special few × dissatisfaction interaction = (special few –mean)
(dissatisfaction – mean)
A8 Special few × mistrust interaction = (special few –mean)
(mistrust – mean)
Ideological Framings: Liberalism and Fundamentalism

The social–individualism index is a composite measure of three indicators:
A9 Love as the basis for marriage: In your view, which of the following is the
more important basis for marriage: (coded as 1) parental approval or (coded
as 2) love?
Reponses to two of the items regarding child qualities are selected as measures of
individualism. The following is a list of 10 qualities that children can be encouraged
to learn at home. Which, if any, do you consider to be especially important? (Please
choose up to five.)

244
A10 Independence: (conded as 2) selected, (coded as 1) not selected

A11 Obedience: (coded as 2) not selected, (coded as 1) selected
Social� individualism index = (A9 + A10 + A11) /3
The gender–equality index is a composite measure of four indicators measuring atti-

tudes toward women, all in Likert-scale format. Do you (1) strongly agree, (2) agree,
(3) disagree, or (4) strongly disagree that
A12 A wife must always obey her husband
A13 On the whole, men make better political leaders than women
A14 A university education is more important for a boy than for a girl
A15 When jobs are scarce, men should have more right to a job than women
Gender� equality index = (A12 + A13 + A14 + A15) /4
The secular–democracy index is a composite measure of two indicators, both in Likert-

scale format: Do you (4) agree strongly, (3) agree, (2) disagree, or (1) strongly disagree
with the following statements:
A16 Secular politics: Egypt will be a better place if religion and politics are
separated.
A17 Democracy: Democracy may have problems but it’s better than any other
form of government
Liberal� government index = ( A16 + A17) /2
A18 If Egypt became more developed, would faith in Allah (1) increase a lot,
(2) increase a little, (3) decrease a little, or (4) decrease a lot?
Islamic fundamentalism is a multidimensional concept consisting of components
such as an authoritarian conception of the deity, exclusivity, intolerance, and literal-
ism. These concepts are measured in Likert-scale format. Do you (4) strongly agree,
(3) agree, (2) disagree, or (1) strongly disagree with the following statements:
A19 Any infraction of religious instruction will bring about Allah’s severe
punishment.
A20 Only the fear of Allah keeps people on the right path.
A21 Allah requires his slaves to repent (tobbah).
A22 Satan is behind any attempt to undermine belief in Allah.
A23 People stay on the right path only because they expect to be rewarded in
heaven.
The authoritarian–deity factor is constructed based on these five components (A19–
A23), with an eigenvalue of 1.93 and Cronbach’s alpha of .59.
The literalism–inerrancy factor is constructed using the following three measures
(A24–A26), with an eigenvalue of 1.52 and Cronbach’s alpha of .49:

245
A24 Whenever there is a conflict between religion and science, religion is

always right.
A25 In the presence of the Quran, there is no need for man-made laws.
A26 Different interpretations of the Quran are equally valid.
The exclusivity factor is constructed using the following two measures (A27 and A28),
with an eigenvalue of 1.22 and Cronbach’s alpha of .34:
A27 Only Islam provides comprehensive truth about Allah.
A28 Only Muslims will go to heaven.
The intolerance factor is constructed using the following two measures (A29 andA30),
with an eigenvalue of 1.59 and Cronbach’s alpha of .68:
A29 Criticism of Islam should not be tolerated.
A30 Criticism of Muslim religious leaders should not be tolerated.
Religiosity Indicators
A31 How often do you pray: (1) never, (2) once or twice a month, (3) once a day,
(4) once or twice a week, (4) two to four times a day, or (6) five times daily?
A32 To what extent do you consider yourself a religious person, on a scale
from 1, not at all religious, to 10, very religious?
A33 How important is the presence of God in your life? Please use this scale
to reply: 10 means very important and 1 means not at all important.
A34 Apart from funerals, how often do you go to a mosque these
days: (1) rarely, (2) once a year, (3) only in the religious events, (4) once a
month, (5) once a week, or (6) more than once a week?
Personal Efficacy and Dysphoric Emotions

A35 Perceived control: Some people feel they have completely free choice and
control over their lives whereas other people feel what they do has no
real effect on what happens to them. Please use this scale, where 1 means
“none at all” and 10 means “a great deal,” to indicate how much freedom
of choice and control you feel you have over the way your life turns out.
A36 Freewill: Some people believe that individuals can decide their
own destiny whereas others think that it is impossible to escape a
predetermined fate. Please tell me which comes closest to your view on
this using a scale for which 1 means “everything in life is determined by
fate” and 10 means “people shape their fate themselves.”
A37 Fatalism × free will interaction = (too many regulations –mean)(free
will – mean)
A38 Insecurity: In Egypt these days, life is unpredictable and dangerous. Do
you (4) agree strongly, (3) agree, (2) disagree, or (1) strongly disagree?
A39 Insecurity × fatalism interaction = (insecurity –mean)(too many
regulations – mean)

246
Xenophobia is a factor based on three indicators, all asking respondents if they would
like to have Americans, British, or French as neighbors (eigenvalue, 2.81’ Cronbach’s
alpha, 0.97). On this list are various groups of people. Please indicate who you would
not like to have as neighbors: (2) would not like, (1) like:
A40 Americans
A41 British
A42 French
Moral flexibility, whether constructed as a factor (eigenvalue, 2.281; Cronbach’s alpha,
0.839) or a linear combination of three interrelated variables, has almost identical
relationship with participation. Respondents were told the following: I will read you a
list of behaviors. Behaviors that are immoral are rated 1 and behaviors that are moral
are rated 10. You can use both of these numbers for rating behaviors plus all the num-
bers in between. Using this scale, where 1 is immoral and 10 is moral, how would you
rate the morality of the following:
A43 Stealing other people’s property
A44 Violence against other people
A45 Telling lies to other people to protect one’s own interests
A46 How much do you rely on radio as a source of information: (4) a great

deal, (3) some, (2) not very much, or (1) not at all?
A47 How much do you rely on domestic television as a source of
information: (4) a great deal, (3) some, (2) not very much, or (1) not at all?
A48 How much do you rely on satellite television as a source of
information: (4) a great deal, (3) some, (2) not very much, or (1) not at all?
A49 How much do you rely on newspapers as a source of information: (4) a
great deal, (3) some, (2) not very much, or (1) not at all?
A50 How much do you rely on the Internet as a source of information: (4) a
great deal, (3) some, (2) not very much, or (1) not at all?
A51 How much do you rely on a mobile phone as a source of
information: (4) a great deal, (3) some, (2) not very much, or (1) not
at all?
A52 Which newspaper is the most reliable source of information for you?
(precoded open-ended question)
Opposition paper: Repondents were asked to mention the newspapers they read as
their sources of information, those who mentioned an opposition paper is coded as
1 and government-controlled paper or none is coded as 0. Paper–Internet–mobile: a
single factor constructed based on a linear combination of reliance on the Internet,
mobile phones, and newspapers as sources of information (eigenvalue, 1.66;
Cronbach’s alpha, .59).

247
Demographics
A53 Education: What is the highest educational level you have attained? (1.)
no formal education, (2) incomplete primary school, (3) complete primary
school, (4) incomplete secondary school: technical/vocational type,
(5) complete secondary school: technical/vocational type, (6) incomplete
secondary: university preparatory type, (7) complete secondary: university
preparatory type, (8) some university-level education, without degree, and
(9) university-level education, with degree.
A54 Self-assigned class membership: People sometimes describe themselves
as belonging to the working class, the middle class, or the upper or
lower class. Would you describe yourself as belonging to (5) the upper
class, (4) upper middle class, (3) lower middle class, (2) working class, or
(1) lower class?
A55 Income: Annual household income before taxes, counting all wages,
salaries, pensions, and other income, was coded in deciles by the local
investigators in each country, with 1 as the lowest decile and 10 as the
highest.
Socioeconomic status: Because social class includes only five categories, for the sake
of consistency and construction of a single factor based on education, income, and
social class, education and income variables were recoded to reduce the number of
categories to five. For education, these are 1 (=1), 2 (= 2 + 3), 3 (=4 + 5), 4 (=7 + 8), and
5 (=8 + 9); for income, they are 1 (=1 + 2), 2 (=3 + 4), 3 (=5 + 6), 4 (=7 + 8), and 5 (=9
+ 10). This factor’s eigenvalue was 1.73 and Cronbach’s alpha was .61. This recoding,
however, did not affect the results of the data analysis significantly.
A56 Socioeconomic status2 = (socioeconomic status –means)2
A57 Impressionable years under Mubarak: A dummy variable coded as 1 for
those between the ages of 18 and 42 (those who were 12 in 1981, when
Mubarak became the president of Egypt) and 0 for other.
A58 Gender (male = 1, female = 0)
A59 Urban residence (urban = 1, rural = 0)
A60 Marital status (single = 1, other = 0; also used as a dummy variable
Finally, three questions measured organizational membership: Now I am going to
read a list of voluntary organizations (associations, communities, networks). For each
one, could you tell me whether you are (2) an active member, (1) an inactive member,
or (0) not a member of that type of organization?
A61 Religious organization
A62 Political party
A63 Humanitarian or charitable organization
Organizational membership is based on A54 to A56 (eigenvalue, 2.81; Cronbach’s
alpha, .97).

248
249
CHAPTER 8 Change and Continuity in Arab

Attitudes toward Political Islam
The Impact of Political Transitions in Tunisia
and Egypt from 2011 to 2013
Mark Tessler
Tunisia and Egypt were the first countries to experience the spontaneous popular
uprisings that burst forth and shook the Arab world at the end of 2010 and the first
months of 2011—events popularly known as the Arab Spring. The two countries
were also distinguished by the rapidity with which the ruling regimes collapsed
and a political transition was set in motion. After thinking initially that the sus-
tained peaceful protests could be contained, Tunisian President Zine al-Abidine
Ben Ali fled the country on January 14; Egyptian President Hosni Mubarak sur-
rendered power and was arrested on February 11. Inspired by these developments,
although played out in different ways and with different results, protests and—in
some cases—major and sustained uprisings also broke out in Bahrain, Yemen,
Libya, Syria, and elsewhere in the months that followed.
Fueled by anger at decades of misrule by governing regimes that were cor-
rupt, authoritarian,1 and, in the minds of ordinary citizens, concerned only with
their own privilege and that of their well-connected friends, the rallying cry of
many of the protesters and the populations for which they claimed to speak was
“dignity” (karama). Expressed in this context, an insistence on dignity signified
a refusal to be led by a regime that did not consider its citizens worthy of atten-
tion. The demand was for leaders who would respect the people they govern,
who would work on their behalf and be concerned about their welfare. These
demands were both articulated and translated into action by young people and
many others who came into the streets and, as it is often described, “crossed the
barrier of fear” in their determination that there would be no return to business
as usual.2
Tunisia and Egypt were distinguished not only by the rapidity with which the
ruling regimes fell, but also by the fact that, first, each quickly proceeded to free and
fair elections and, second, in both countries these elections were won by parties
that campaigned under the banner of Islam. Advocating what is usually referred
to as political Islam, and frequently as Islamism, the Ennahda Party was victorious
in Tunisia’s legislative elections, and the Muslim Brotherhood’s political wing, the
250
Freedom and Justice Party (FJP), received a plurality in Egypt’s parliamentary elec-
tions. The Muslim Brotherhood candidate, Mohammed Morsi, also subsequently
won the country’s presidential election. Although the history and circumstances
of the two movements differ in significant respects, the ideology and platform of
Ennahda and the Muslim Brotherhood are based on the principle that Islamic
authorities, institutions, and legal codes should play an important role, perhaps
even a leading role, in government and political affairs.3
Against the background of these developments in Tunisia and Egypt, this chap-
ter uses public opinion data from surveys conducted in 2011 and 2013 to assess
the attitudes held by ordinary Tunisians and Egyptians about the place that Islam
should occupy in political life. First, using data collected in 2011, before the elec-
toral victories of Ennahda and the FJP, we examine the degree to which positive
attitudes toward these parties are associated with support for a system of govern-
ment in which Islam plays an important role or, alternatively, whether Ennahda
and the FJP also drew significant support from men and women who do not favor
an Islamist government. Second, we compare findings from the 2011 and 2013
surveys—the latter collected at a time when Ennahda and the FJP had been in
power for more than a year—to assess whether and how the experience of living
under an Islamist government affected Tunisian and Egyptian judgments about
the place Islam should occupy in the political affairs of their countries. Finally, this
chapter uses survey data collected in four other Arab countries—Algeria, Palestine
(West Bank and Gaza), Jordan, and Iraq—during the same 2-year period to inves-
tigate the possibility that any changes in Tunisian and Egyptian attitudes toward
political Islam also occurred elsewhere. Men and women in these countries did
not experience life directly under a democratically elected Islamist government,
but rather witnessed and observed from a distance the way that Tunisia and Egypt
fared when Islamists were in power.
In the next section of this chapter, before presenting public opinion data and
considering the insights they offer, events leading up to the fall of the Ben Ali and
Mubarak regimes are discussed briefly. The purpose of this description is to give a
sense of the mood in Tunisia and Egypt during the first months of the Arab Spring
and at the time when the 2011 surveys were conducted. Following the data-based
sections, the chapter concludes with some thoughts about the path going forward
on which Tunisia and Egypt appear to have embarked. This concluding section
describes and considers the implications of the collapse of the Islamist-led govern-
ments in both countries in 2013.
Tunisia and Egypt during the Arab Spring

The uprisings in Tunisia, known as the Jasmine Revolution, were sparked by the
self-immolation of a street vendor, Mohamed Bouazizi, in the central town of Sidi
Bouzid on December 17, 2010. Bouazizi’s act followed the confiscation of his goods
by local officials, who claimed he did not have the necessary permits, and then,
according to news reports at the time, further humiliation when he was slapped by
a female municipal inspector. After taking his case to the local governor and being

251
denied a meeting, Bouazizi set himself on fire in front of the governor’s office.
Reports of Bouazizi’s action emphasized his struggles with underemployment,
the absence of opportunity, and state arrogance and corruption. Although possibly
exaggerated in some accounts,4 these widely reported events told a story that reso-
nated with many Tunisians and catalyzed widespread protests. Demonstrators,
including middle-class as well as disadvantaged Tunisians, denounced the politi-
cal and economic status quo and demanded an end to Ben Ali’s corrupt and
authoritarian regime.
Tunisian authorities initially met the protesters with force, believing it would
be possible to suppress the spreading demonstrations. The BBC’s account of
Mahjoub El Harbaoui, a government worker from the economically depressed
Siliana region, illustrates the early confrontations. El Harbaoui joined a peace-
ful and legal demonstration, marching from the trade union office to the local
governor’s house. The protesters were not calling for removal of the governor,
but merely for more jobs and better government services. Security forces fired on
them with birdshot, however, hitting El Harbaoui in the face and severely injur-
ing one of his eyes. When asked a year later about the incident and his reason for
protesting, he told a BBC interviewer:
I can’t believe this happened. I’m not a casseur [hooligan]. I was demonstrating
peacefully. … If a Tunisian government worker cannot express his disappoint-
ments with his authorities without it resulting in violence, what hope does that
give the ordinary citizens of Tunisia … ? There is a feeling that we are just not
paid attention to, that no-one cares about our problems here, that no-one really
respects us (“Three Faces of the New Tunisia,” 2012, http://www.bbc.com/news/
world-africa-20755296).
Despite these efforts at containment, protests continued and, indeed, amplified
in the days that followed,5 and in less than a month the Ben Ali regime collapsed.
Ben Ali, age 74 and in power since 1987, fled the country with his family and took
refuge in Saudi Arabia. The Prime Minister, Mohammed Ghannoushi, then told
the country in a speech from the presidential palace in Carthage that he would
carry out the functions of the president during an interim period. He called for
“Tunisians of all political and regional tendencies to show patriotism and unity.”
Soon thereafter, in accordance with the Constitution, leadership passed to the
speaker of parliament and the country began to prepare for elections.
The complaints of ordinary Egyptians mirrored those that were fueling protests
in Tunisia. Under the authoritarian and corrupt Mubarak regime, conspicuous
and sometimes ostentatious consumption characterized a wealthy and politically
connected class, whereas more than 40% of population lived below the poverty
line and had little hope that their situation would be better in the future. Those
who challenged the status quo were suppressed regularly, including not only those
whom the government might accuse of terrorism and political violence, but also
intellectuals, human rights advocates, democratic activists, and others who criti-
cized the regime.6 Suppression and force remained the regime’s response when
massive anti-Mubarak protests, inspired by developments in Tunisia, erupted on
January 25, the country’s National Police Day. Violent clashes between security
Change And Continuity In Arab Attitudes Toward Political islam | 251

252
forces and protesters resulted in more than 800 people killed and more than
5,000 wounded. In the end, however, after trying unsuccessfully for 18 days to
suppress the demonstrations in Cairo, Alexandria, and other cities, Mubarak
resigned and turned over power to a military council: the Supreme Council of the
Armed Forces.
Mubarak, who had been in power for almost three decades and was 82 at
the time of his ousting, was arrested, along with his two sons, in April 2011.
This ended speculation, probably unfounded, that the former president had
made a secret deal with the military and was working to engineer a return to
power. It also boosted confidence in the ruling military council, which had
increasingly come under criticism for not going after Mubarak, himself a
former air force commander. With the public thus assured that the demise
of the Mubarak regime was permanent, Egypt turned its attention to the
next phase of its transition, known locally as the January 25 Revolution, and
began to gear up for elections in which both Islamists and others would field
candidates.
The tenor of this period—understood, as noted earlier, as a struggle for dignity,
or karama—is nicely captured in the following recollections of one of the protest-
ers, Ahmed Raafat Amin, a 22-year-old student in Cairo (Shah 2011):
I was a normal college guy—just focusing on my studies and hanging out with
my friends. I was also like many other Egyptians who were dreaming of change
but never believed it could happen. However, our country’s condition was get-
ting worse and worse. There was corruption, torture, injustice, inequality and no
freedom. Someone had to stand up and say, “Enough is enough,” and that is why
I decided to take part in the revolution.
At first I was afraid to take part. But, as I realized the demands of the revolu-
tion were my own demands, I was willing to pay whatever price our freedom
would cost. I saw all kinds of people—rich, poor, young, old, men, and women—
losing their lives for something they believed in. Tahrir Square, the focus of the
protests in Cairo, was like heaven. It was how you wanted Egypt to be. In the past,
I only focused on personal dreams, but now I’m focusing on a national dream
that we all share.
Egypt and Tunisia were not the only countries that experienced antiregime pro-
tests during this period. In some cases these protests subsided after govern-
ments responded by implementing modest reforms or by distributing additional
resources. Governments fell in other cases, most notably in Yemen and Libya,
and violence continues still in other countries, including Syria, Iraq, and, more
recently, Yemen, and, to a lesser extent, Bahrain. But, Tunisia and Egypt stand
out—not only because they set in motion the diffusion of protests and challenges
to the status quo that characterized the Arab Spring but also, and in some ways
even more, because of the path they followed after the demise of regimes that had
been in power for decades. Free and fair elections were held in 2011 in Tunisia and
in 2011 and 2012 in Egypt, and in both countries these elections brought to power a
party that, while declaring its commitment to democracy, advocated the centrality
of Islam in government and political affairs.

253
In less than 2 years, however, these Islamist parties had lost power in both
countries. In Egypt, economic conditions deteriorated further under the Islamist
government, and there was also growing concern about some of President Morsi’s
political appointments and his authoritarian turn more generally, particularly
his efforts to undermine the authority of the judiciary. Thus, he faced massive
and repeated protests during the first half of 2013, which gave the military an
excuse to intervene and depose him. In Tunisia, Ennahda also lost popularity as
the country’s economic situation worsened, and there was anger as well at the
party’s inability, or, according to some, its unwillingness, to restrain the attacks of
Islamic extremists on liberal institutions and leaders. Among the latter were the
assassinations of two prominent liberal politicians. As a result, there were also
massive protests in Tunisia in summer 2013 and, in the wake of these demonstra-
tions, the Ennahda government resigned and surrendered power to a caretaker
government.
Against the background of these developments during the critical period of
2011 to 2013, the remainder of this chapter uses public opinion data to examine
the nature and evolution of Tunisian and Egyptian attitudes toward the role that
Islam should play in government and political affairs. The analysis seeks first to
assess the degree to which the electoral victories of Ennahda and the FJP reflected
a desire for government in accordance with Islam and, then, whether more than a
year under an Islamist government changed the way that ordinary Tunisians and
Egyptians were thinking about the place Islam should occupy in the political life
of their countries.
Data and Measures

Public opinion data collected as part of the second and third waves of the Arab
Barometer7 are used to examine the Tunisian and Egyptian attitude patterns out-
lined earlier. These include (a) a comparison of attitudes toward Ennahda and the
FJP on the one hand, and toward Islam playing an important role in government
and political affairs on the other; and (b) a comparison of attitudes toward political
Islam in 2011 and 2013, after the two democratically elected Islamist parties had
been in power for more than a year. As noted, the analysis also includes a compari-
son of attitudes during the same 2-year period in four Arab countries in which an
Islamist party did not come to power after the Arab Spring.
The Tunisian sample sizes are N = 1,196 in 2011 and N = 1,119 in 2013; the
Egyptian sample sizes are N = 1,220 in 2011 and N = 1,196 in 2013. The four other
Arab countries for which attitudes are compared, based again on the second and
third waves of Arab Barometer surveys, are Algeria, Palestine (West Bank and
Gaza), Jordan, and Iraq. In Palestine and Jordan, the first-wave survey was actually
carried out in late December 2010, and in both countries the third-wave survey
was conducted in late December 2102. The sample sizes for these countries are
Algeria 2011, N = 1,261; Algeria 2013, N = 1,220; Palestine 2010, N = 1,200; Palestine
2012, N = 1,200; Jordan 2010, N = 1,188; Jordan 2103, N = 1,795; Iraq 2011, N = 1,234;
and Iraq 2013, N = 1,215. All 12 of the surveys included in the analysis—six in

254
2010/2011 and six in 2012/2013—are nationally representative and were carried

out through face-to-face interviews with probability-based samples of adults age
18 years and older. Before conducting fieldwork, sampling plans were prepared
by in-country teams that were then reviewed by specialists in the United States
and a unit working with the Arab Barometer at the Center for Strategic Studies
(CSS) at the University of Jordan. Specialists from CSS, together with a team in
the United States, provided guidance and oversight on other methodological con-
siderations, and members of the CSS team visited several of the countries, most
notably Tunisia and Algeria, to assist in interviewer training and implementa-
tion of the surveys. On completion of the surveys, data were reviewed both in the
United States and at CSS for possible falsification, and the validity and reliability
of key items were also assessed. There was follow-up with local in-country teams
when needed.
The Arab Barometer survey instrument contains a number of significantly
intercorrelated items measuring attitudes toward political Islam. These include
the following items, which were asked in both the second-and third-wave surveys:
To what extent do you agree or disagree with each of the following statements?
Indicate whether you strongly agree, agree, disagree, or strongly disagree.
• [Country name] would be better off if religious people hold public positions
in the state.
• Religious leaders (imams, preachers, priests) should have influence over
government decisions.
• Religious practices are private and should be separated from social and
political life.
The item that asks whether religious leaders should have influence over govern-
ment decisions is used in Figures 8.1, 8.2, 8.3, and 8.4 to measure attitudes toward
political Islam. This item consistently has very high loadings on the single factor
that emerges from a factor analysis based on the three items—with a loading of
.792 in one instance and more than .800 in all others—and more often than not it
has the highest loading of the three items. Accordingly, the analysis not only offers
evidence of reliability and a basis for inferring validity, it also indicates that the
item is a robust indicator of the concept measured by the battery of items. Factor
loadings are presented in the Appendix, and sources are available for those wish-
ing additional information about measures and findings pertaining to views about
political Islam based on the analysis of Arab Barometer data.8
Islamist Electoral Victories in Tunisia and Egypt

The political transitions in Tunisia and Egypt brought elections in which
Islamist candidates were victorious. This was not entirely unexpected given the
organizational strength of Islamist political parties and their ability to mobi-
lize their supporters. In Tunisia and Egypt, as in many other Arab counties,
these parties disseminated their message that “Islam is the solution” through
a network of religious, social, and cultural institutions. Given their religious

255
character, the state had only limited control over most of these institutions,
many of which did charitable work or delivered social services that also helped
them to win adherents, and this in turn freed them from at least some of the
government interference that constrained the organizational opportunities
available to other potential opponents of the regime. In this way, Islamist par-
ties and movements were often able to build a sizeable core constituency.9 In
Tunisia and Egypt, although they were neither early joiners nor subsequent
leaders of the protests that brought down the Ben Ali and Mubarak regimes,
both Ennahda and the Muslim Brotherhood thus had a block of supporters on
whom they could draw in the electoral contests that took place in the two coun-
tries. Their organizational network also proved effective in mobilizing support-
ers and getting these individuals to the voting stations.10
But, Islamist opposition parties in the Arab world have very often punched
above their ideological weight in electoral contests, gaining enough votes beyond
those provided by their core constituencies to prevail at the ballot box. Their elec-
toral success has been the result, in part, of superior mobilization capacity and the
related fragmentation or disorganization of secular opposition parties. Indeed, in
some instances there has been no credible secular opposition party. In addition,
however, a vote for an Islamist opposition party has often been seen as the best
way, and sometimes the only way, to cast a consequential vote against the regime
in power. A vote for an Islamist party is, for some, and perhaps many, a vote
against corruption or authoritarianism—one in favor of fairness and accountable
government, rather than a vote for Islam to play a role in government and political
affairs. This kind of “strategic voting,” in which support is based on considerations
other than a preference for the party’s platform, is well known in elections all over
the world, and it has produced votes for Islamist parties in many Arab elections
over the years. Although somewhat dated, a good illustration of the reasoning that
leads some who do not favor political Islam to cast their votes, nonetheless, for
an Islamist party is provided by the way in which a young man explained his vote
for the Islamic Salvation Front in the critical Algerian election of May 1990 (Y.
Ibrahim 1990, http://www.nytimes.com/1990/06/25/world/militant-muslims-
grow-stronger-as-algeria-s-economy-weakens.html?pagewanted=all):
In this country, if you are a young man … you have only four choices: you can
remain unemployed and celibate because there are no jobs and no apartments to
live in; you can work in the black market and risk being arrested; you can try to
emigrate to France to sweep the streets of Paris or Marseilles; or you can join the
Islamic Salvation Front and vote for Islam.
Suggesting that many of the same factors were at work in Ennahda’s victory in
the October 2011 election in Tunisia, a contemporary journalistic account reported
that “most voters who supported Ennahda didn’t do so because of a yearning for a
rigid, religious regime. They were instead attracted by the party’s credibility—and
its perceived proximity to the common people” (von Rohr, 2012).
As noted, a strategic vote involves casting a ballot for a party or candidate with
which one does not necessarily agree, or that at least does not hold one’s preferred
positions, but the electoral success of which is judged by voters to be preferable

256
to the probable outcome should they not support that party or candidate. Electoral
dynamics and the specific choices available to a strategic voter depend on the char-
acter of the electoral system, the ideological arrangement of the various parties or
candidates, and many other context-specific considerations. At its core, however,
as reflected in a complex, multifaceted, and wide-ranging body of political science
literature on the subject, strategic voting involves basing one’s decision about how
to vote on factors other than the alignment of one’s own views and policy prefer-
ences with those of the entity for which one does in fact vote.11
Evidence from the 2011 Arab Barometer survey described earlier makes clear
that this kind of strategic voting was an important contributor to Ennahda’s victory
during the Tunisian parliamentary election of October 2011, the first genuinely
contested free and fair election in the country’s history. Although Ennahda won
the election easily with a plurality of around 40% of the votes, the Arab Barometer
survey, conducted less than a month before voting, found that many who were
prepared to support the party, and presumably did support it on election day, did
not endorse an Islamist platform. Almost half the Tunisians interviewed said they
trusted Ennahda. In contrast, only one quarter agreed strongly or even agreed
with the proposition that religious leaders should have influence over government
decisions. These findings are shown in Figure 8.1.
The Egyptian parliamentary elections, held in three stages between late
November 2011 and mid January 2012, were won by the FJP of the Muslim
Brotherhood. The party captured about 38% of the votes. Again, however, data
from 2011 Arab Barometer survey in Egypt indicate there is a significant gap
between the proportion of Egyptians who, at the time, had a positive attitude
toward the party and the proportion who supported an Islamist agenda.
The Egyptian case is more complicated given the presence of a second Islamist
party, An-Nour, and also because the Arab Barometer survey in Egypt was con-
ducted in June and July 2011, fully 5 months before the first round of parliamen-
tary voting. Nevertheless, the same insights about strategic voting emerge from
the Egyptian case. Forty-seven percent of the respondents in the Arab Barometer
survey expressed either great or at least medium trust in the party. In contrast,
again, the proportion agreeing that political life should be influenced by Islam was
substantially less. Only 37% agreed strongly or agreed that religious leaders should
have influence over government decisions. These findings are shown in Figure 8.2.
49
Great or medium trust in Ennahda
51
Yes
No
Agree that religious leaders should 25
have influence over government 75
decisions
0 10 20 30 40 50 60 70 80
Percentage
Figure 8.1 Tunisia in 2011: Attitudes toward Islamism before Ennahda came
to Power.

257
Great or medium trust in Freedom 47

and Justice Party 53
Yes
Agree that religious leaders should No
37
have influence over government
63
decisions
0 10 20 30 40 50 60 70
Percentage
Figure 8.2 Egypt in 2011: Attitudes toward Islamism before the Muslim Brotherhood
came to Power.
The lesson to be drawn from these data about the post-Arab Spring experience
of Tunisia and Egypt is not only that many ordinary citizens engage in strategic
voting, but also, with respect to Islam in particular, that the electoral victories of
Ennahda and the Muslim Brotherhood’s FJP do not provide evidence that most
ordinary citizens want to be governed by a political system in which religion
exercises significant influence. In both cases, the parties came to power on the
strength of votes not only from a core constituency, but also from a large number
of ordinary citizens who desire an alternative to the political and economic status
quo but apparently have little or no sympathy for an Islamist political agenda.
Attitudes toward Political Islam Two Years Later

The governments led by Ennahda in Tunisia and the Muslim Brotherhood in
Egypt collapsed in mid 2013. The details of the two cases differ, and in both coun-
tries there were, and continue to be, debates about responsibility for the shortcom-
ings and subsequent fall of the two democratically elected Islamist governments.
Nonetheless, the Tunisian and Egyptian experiences resemble one another in sev-
eral important respects.
First, the economic situation, already in a dismal state, continued to deteriorate
in both countries. There was also growing unhappiness with the religious agenda
of the governments in power, including policies that were shaped by religious
ideology rather than economic exigencies, political appointments that were based
more on religious affiliation than competence, and tacit support or at least toler-
ance of Islamic extremism. Although Islamists argued that economic and many
other problems were the result, at least in part, of efforts by the former regime to
subvert a transition that threatened their privileges, popular discontent and disap-
pointment with the new government became increasingly intense and widespread
during 2012 and into 2013.
Second, by mid 2013 there were massive protests in both Tunisia and Egypt
against the governments that had come to power less than 2 years earlier on a
wave of anger at the pre-Arab Spring authoritarian regimes and an intense desire
for a more democratic political system. This led to the voluntary resignation of
the government in Tunisia, which was replaced by a caretaker government of

258
independents and technocrats. In Egypt, claiming a popular mandate to intervene,

the army removed and arrested the elected Islamist president, Mohammed Morsi.
These developments suggest that Ennahda and the Muslim Brotherhood had
lost the support of many and perhaps most of their strategic voters, those who had
helped to bring them to power not because of, but rather despite, their Islamist
credentials and commitments. Many who, in 2011, considered these parties to be
desirable, or at least acceptable, alternatives to the Ben Ali or Mubarak regimes
now appeared to conclude they were not qualified to lead the country. However,
Islamists had come to power not only with support from strategic voters, but also
based on the strength of their sizeable, and mobilizable, core constituencies. Were
these men and women who had endorsed political Islam in 2011 still in favor of an
Islamist platform in 2013?
Arab Barometer surveys conducted in January 2013 in Tunisia and April 2013
in Egypt make it possible to determine whether and how attitudes toward politi-
cal Islam changed as a result of living for more than a year under an Islamist
government. Figure 8.3 compares findings from the 2011 and 2013 surveys in each
country.
In Tunisia, the percentage of respondents who strongly agreed or agreed that
religious leaders should have influence over government decisions remained con-
stant, at 26% in both years. The size of Tunisia’s core Islamist constituency—those
who favor a political formula that gives influence to Islam—although generally
low, apparently was not reduced by 15 months of life under an increasingly unpop-
ular Islamist-led government.
In Egypt, in contrast, the percentage agreeing strongly or agreeing that reli-
gious leaders should have influence over government decisions declined substan-
tially: from 37% to 20% during the nearly 20 months between the 2011 and 2013
surveys. The differing patterns in Tunisia and Egypt may be a result, in part, of
the fact that the 2013 Tunisian survey was conducted earlier, when the events that
brought down the Ennahda government later that year were only beginning to
unfold. In any event, as shown in Figure 8.3, there remained core constituencies
Religious Leaders Should Have Influence over Government Decisions

60
50 56
Percentage
40 43 44
30 36 38 37 36
20 26 26
19 20 20
10
0
Tunisia Tunisia Egypt Egypt
2011 2013 2011 2013
Strongly agree/agree Disagree Strongly disagree
Figure 8.3 Tunisian and Egyptian Attitudes toward Islamism after Living under an
Islamist Government.

259
for political Islam in both countries, constituting, despite its decline in Egypt,
roughly one fifth to one quarter of the adult population.
There is an additional conclusion to be drawn from the comparisons of Tunisian
and Egyptian attitudes in 2011 and 2013. Among those who did not agree that reli-
gious leaders should have influence over government decisions, some of whom
presumably voted for an Islamist party in 2011, there was a substantial shift in
attitudes. In both Tunisia and Egypt, the proportion of those who disagreed strongly
with the proposition increased significantly, both in absolute terms and relative to
those who merely disagree. In Tunisia, as shown in Figure 8.3, the proportion of
respondents who disagree strongly that Islam should play a role in government and
political affairs increased by 19% between 2011 and 2013, and in Egypt the increase
was 16%. Thus, although between the two surveys a minority in both countries
continued to favor a political role for Islam, there was a remarkable increase in the
intensity of the views held by those who believe that Islam, however important it
might be in one’s personal life, should be kept apart from the affairs of government.
A broader proposition that may be deduced from these findings is that much of
the support for political Islam that comes from those who do not favor an Islamist
platform—those who may be described as strategic voters—is based on assump-
tions about how well the Islamists will perform in office that are uninformed
by any actual experience with Islamist government. After these strategic voters
had for a time withnessed the performance of an Islamist government, however,
their assumptions no longer held and their support disappeared—or even became
opposition. Tunisia and Egypt provide only two examples, of course, and support-
ers of political Islam can and do argue that circumstances in each country denied
Ennahda and the Muslim Brotherhood the time and opportunity to deliver the
good governance they are fully capable of providing. However, although such
arguments are not unreasonable and will continue to be debated, it remains the
case that public opinion data from Tunisia and Egypt lend support to the proposi-
tion that the appeal of government led by an Islamist party is much greater when
anticipated than when actually experienced.
Impact on Attitudes in Other Arab Countries

The citizens of other Arab countries did not experience the rise and fall of Islamist-
led governments that marked the 2011 to 2013 period in Tunisia and Egypt. But,
these developments were presumably witnessed and watched, and to that extent
they were probably viewed and evaluated against the background of the antistatus
quo sentiments that characterized these other Arab countries. Was there, as a
result, some diffusion of the attitude changes pertaining to political Islam that
took place in Tunisia and Egypt? Or were these shifts largely absent among the
citizens of Arab countries who did not live through the rise and decline of an
Islamist-led government? It is not possible to know with certainty whether and to
what degree ordinary citizens in Algeria, Jordan, Palestine, and Iraq were aware
of developments in Tunisia and Egypt. It is likely, however, that large segments of
these populations were at least broadly familiar with events unfolding in the two

260
countries and, accordingly, it is reasonable to assume there was enough aware-

ness to give importance to the question of whether witnessing the performance
of an Islamist government in a nearby Arab country, as opposed to experiencing
it directly at home, had an impact on popular views about the place Islam should
occupy in government and political affairs.
This question can be addressed using data from the second and third waves of
Arab Barometer surveys conducted in Algeria, Jordan, the Palestinian Territories,
and Iraq. The first survey in each country was conducted immediately before
or very shortly after the fall of the secular authoritarian regimes in Tunisia and
Egypt—at the end of 2010 or early in 2011; the second survey was conducted
approximately 2 years later—at the end of 2012 or during the first half of 2013.12
The intervals thus differ only slightly from the period covered in the Tunisian and
Egyptian surveys.
In all four countries, patterns are very similar to one another but different
from those observed in Tunisia and Egypt. In each case, as shown in Figure 8.4,
responses to the question about whether religious leaders should have influ-
ence over government decisions, the question to which Tunisian and Egyptian
responses are presented in Figure 8.3, are distributed in very similar ways in the
second-and third-wave Arab Barometer surveys.
Support for political Islam is greater in Jordan, Palestine, and Iraq than in
Algeria. This may be because the first two societies are more conservative, or pos-
sibly because the immediacy of the Israel–Palestine conflict adds a dimension to
the political landscape that fosters a favorable attitude toward political Islam. And
in Iraq, this may be because attitudes toward political Islam during this period
were shaped, to a substantial degree, by sectarian considerations. Whatever the
reason, however, the important point for the current analysis is that, whether
skewed in one direction or another, the distributions in Algeria, Jordan, Palestine,
Religious Leaders Should Have Influence over Government Decisions

60
50 54
51
48 47 47 47
46
40 42 43
41 40 41 41
Percentage
39
30 34
27
20
19 18
14 16
10 12 12 12
9
0
Algeria Algeria Jordan Jordan Palestine Palestine Iraq Iraq
Wave 2 Wave 3 Wave 2 Wave 3 Wave 2 Wave 3 Wave 2 Wave 3
Strongly agree/agree Disagree Strongly disagree
Figure 8.4 Algerian, Jordanian, Palestinian and Iraqi Attitudes toward Islamism in

2010–2011 and 2012–2013.

261
and Iraq did not change very much during the 2 years that experiments with an
Islamist government were playing out in Tunisia and Egypt.
Algerians, Jordanians, Palestinians, and Iraqis may have looked at the Tunisian
and Egyptian experience and reached any one of a number of conclusions: that
Islamist parties are not ready to govern, that former regimes sabotaged the transi-
tions to regain their privilege, or that political transitions inevitably bring unrest
and even violence. Regardless of the relative weight of these and any other les-
sons that might have been learned from events in Tunisia and Egypt, however,
these perceptions did not modify—neither increase nor decrease—the degree of
popular support for political Islam in Algeria, Jordan, Palestine, and Iraq, and
very probably not in other Arab countries as well. Thus, given the contrast with
the attitude changes observed in Tunisia and Egypt, the conclusion to be drawn
with respect to attitudes toward Islam’s place in political life is that witnessing the
rise and fall of a democratically elected Islamist-led government in the wake of the
Arab Spring did not have the same impact as actually living for a year or two under
a government led by an Islamist political party.
Tunisia and Egypt have pursued different paths since the third wave of Arab
Barometer surveys were conducted. The military has returned to power in
Egypt and the prospects for a genuine democratic transition are not promis-
ing. In Tunisia, in contrast, Islamist and secular leaders reached agreement
on a new Constitution and on the path forward toward democratic governance.
Although it remains to be seen how events will play out in these two countries
during next few years, and although it is not out of the question that there will
be some surprises, for better or worse, the experience of Tunisia and Egypt—
the leaders, symbols, and exemplars of what the Arab Spring produced—has
apparently left unchanged the divisions and debates within broader Arab popu-
lations about whether Islam should play an important role in government and
political affairs.
Questions Going Forward

The uprisings of the Arab Spring were not motivated by ideological consider-
ations. Rather, the primary grievances that brought masses of ordinary citizens
into the streets were economic and political: unemployment and growing pov-
erty, censorship and governmental control of the media, and especially govern-
ment corruption and a perceived lack of responsiveness to the needs of the
people. Many believed that religion is a private matter and should be kept apart
from government and politics. Others held the view that Islam should play
an important role in political affairs. In either case, however, Tunisians and
Egyptians who held these differing perspectives were able to agree in late 2010
and early 2011 that the status quo was unacceptable and a change of regime
was necessary. Equally important, they were also able to agree on the need for
political space that would allow to be expressed the will of the people about how
and by whom their country should be governed. In this environment, Islamist
parties were the best organized and also had the best record of opposition to

262
the regime in power, and thus, as most observers predicted, Tunisia’s Ennahda
Party and Egypt’s FJP won the democratic elections that the events of the Arab
Spring had made possible.
By mid 2013, the governments led by each of these Islamist parties had become
extremely unpopular. Unresolved economic problems were paramount, although
there were also complaints relating to government appointments and to policies
in other areas. Islamists and some others argued that many of the difficulties
were the result of sabotage by supporters of the Ben Ali and Mubarak regimes. In
Egypt, for example, those taking this view pointed out that fuel shortages during
the Morsi presidency disappeared almost overnight when Morsi was no longer in
office. Nevertheless, regardless of the accuracy of the charges and countercharges
about responsibility for the problems in Tunisia and Egypt, disappointment and
even anger brought about new popular protests and led to the demise of the dem-
ocratically elected governments in both countries. In Tunisia, the government
resigned and entered into negotiations with its secular opponents, including
individuals who had been part of the Ben Ali government. In Egypt, the military
launched what was, in effect, a military coup and arrested the leaders (and many
supporters) of the Muslim Brotherhood.
Since that time, the political trajectories of Tunisia and Egypt have been very
different, illustrating, respectively, at least thus far, both the hopes and the dis-
appointments that the events of the Arab Spring fostered. Tunisia stands out,
not only in comparison with Egypt, but also in comparison with the rest of the
Arab world, as the one country that is on track to fulfill the promise of the Arab
Spring. Significant economic, security, and other problems remain, and it is pos-
sible at the time of this writing that the Tunisian experience is generating more
optimism than is warranted. Nevertheless, the country’s accomplishments have
been impressive. Recognizing the need for pluralism, inclusion, and compro-
mise, secular and Islamist leaders drafted a new Constitution—adopted January 1,
2014—and, despite falling short in the view of some Tunisians, this constituted a
significant democratic advance.
Competitive and highly contested parliamentary and presidential elec-
tions took place in Tunisia in late 2014. A new and essentially secular political
party, Nidaa Tunes (Tunisia’s Call), emerged and prevailed in both contests,
but Ennahda and the candidates it supported also did very well. Other par-
ties, particularly on the political Left, added further to political and ideological
pluralism taking shape in Tunisia. And most important of all, fringe elements
notwithstanding, all sides appeared to be committed to political inclusion. As
expressed by a prominent member of Ennahda’s Shura Council at the time
of the parliamentary elections, “It is in the benefit of the country to include
all the political players” (Al Amraoui, 2014). It thus appeared, at the end of
2014, despite unanswered questions about the long-term relationship between
those who do and those who do not support political Islam, that Tunisians had
established a functioning democracy, with Islamists and secularists participat-
ing in a political process that accepts the legitimacy of each, gives meaningful
choices to the population, and resolves differences through compromise and
the ballot box.

263
Egypt’s post-Arab Spring political trajectory has been very different. It has also
been much more problematic from the perspective of those calling for democ-
racy and political inclusion. The poor performance of the government led by the
FJP brought a loss of support not only among strategic voters, but also among
some whose favorable attitude toward the party in 2011 reflected a belief that Islam
should play a meaningful role in political affairs. By mid 2013, discontent with
continuing economic and other problems was broad and deep, and this brought
new public demonstrations. Then, with tens of thousands massing in Tahrir
Square and other locations across Cairo, as well as in a number of other cities, the
army intervened on July 3, removing the president, suspending the Constitution,
and appointing a military-backed caretaker government.
In contrast to Tunisia, this was not to be the occasion for a political reset in
which Islamists and their opponents sought common ground based on a formula
that emphasized inclusion and compromise. On the contrary, led by the military,
authorities moved quickly not only to arrest the leaders and many supporters of
the Muslim Brotherhood, but also to suppress protest and dissent more broadly.
Then, in November 2013, the military-backed government passed the Protest Law,
which effectively banned, under penalty of arrest, the kind of street demonstra-
tions that had brought down the Mubarak regime and also led to the coup of July
2013. According to the Human Rights Watch, the law showed clearly that the new
government’s goal was “to sharply restrict peaceful assembly and to let security
[ forces] shut down protests at will” (Human Rights Watch, 2013). Thus, the state-
ment by the Human Rights Watch continued, the law “will reverse the freedom to
demonstrate that Egyptians seized in January 2011, and risks putting that freedom,
which brought about momentous change, into reverse.”
The Egyptian military defended the Protest Law in the name of security, and it
was not necessarily unpopular given that Egypt had, and continues to have, legiti-
mate security concerns. But popular or not, at least for now, the military-backed
government has continued its crackdown on all forms of dissent. Journalists have
been arrested and security forces also moved against students, including women,
in an effort suppress what the Human Rights Watch called “the peaceful exercise
of the right of free speech.” In October 2014, noting that more than 110 students
had been arrested since the start of the school year, and pointing to other arrests
as well, the Human Rights Watch concluded that Egypt has “vastly extended the
reach of the country’s military courts and risks militarizing the prosecution of
protesters and other government opponents” (Human Rights Watch, 2014). Thus,
with the leaders and many supporters of the Muslim Brotherhood either in jail,
in exile, or driven underground, and with independent voices on the Left, at uni-
versities, and in the media increasingly suppressed, Egypt returned, in the view of
many, to the kind of police state that it had become under Hosni Mubarak.
In Egypt, as in Tunisia, it remains to be seen how public attitudes toward politi-
cal Islam will evolve in this context. On the one hand, there are reasons to believe
that support for political Islam will remain at the low level it reached in mid 2013,
or perhaps may decline even further. Not only was the country’s experience with
Islamist government judged to be unsatisfactory by most Egyptians, the network
of institutions through which the Muslim Brotherhood disseminated its message

264
and gained adherents has been largely destroyed, or at least driven underground.
On the other hand, if the economic situation fails to improve and more Egyptians
begin to chafe under the political restrictions imposed by the military-backed gov-
ernment, and especially if these are accompanied by significant and visible cor-
ruption, the pressures for political change that drove Mubarak from power may
reemerge and strategic voters may once again come to see Islamists as an accept-
able and the best available alternative to the status quo. Indeed, raising the possi-
bility that there may already be some movement in this direction, a national survey
carried out in January 2015 by the Egyptian Center for Public Opinion Research
(Baseera) found that 35% of the 2,020 respondents supported the participation
of “religious parties” in Egyptian elections, an additional 18% were undecided,
and 47% said religious parties should not participate (Egyptian Center for Public
Opinion Research, January 27, 2015).13
Beyond Tunisia and Egypt—iconic symbols of what the Arab Spring has and
has not produced, respectively—a diverse array of experiences characterized the
Arab world during early 2011 and late 2014. In some countries, such as Syria,
Libya, and Iraq, and more recently in Yemen, there was sustained violence,
fueled in most instances by sectarian as well as political antagonisms. In oth-
ers, such as Jordan and Morocco, meaningful, albeit modest, reforms addressed
some grievances while leaving the status quo basically unchanged. And in still
other countries, such as Algeria, Saudi Arabia, Palestine, and a few others, the
domestic political situation looked, in December 2014, about the same as it had
in December 2010. All these countries are important, and in no case was the situ-
ation on the ground as straightforward as the preceding simplified characteriza-
tions might suggest. Nevertheless, in light of their common experience with an
Islamist government and the diverging courses they charted following that experi-
ence, Tunisia and Egypt stand out and offer particularly instructive insights about
the opportunities and difficulties associated with democratization and its connec-
tion to political Islam.
Concluding Thoughts
Although the focus of this chapter is on views about political Islam held by ordi-
nary men and women, the political trajectories of Tunisia and Egypt during the
critical period of 2011 to 2013 invite reflection about the possibilities for democrati-
zation in the Arab world. One take-away from this period, reflecting the Egyptian
experience, is that the path to democratization is neither easy nor ensured, and
that overturning an entrenched authoritarian regime, important and exhilarating
as that may be, is only the beginning of a difficult and uncertain journey. Forces
of the old order may remain and push back, reflecting what is sometimes called
authoritarian learning, and divisions based on ideology, region, and/or sectarian-
ism may quickly undermine the unity of purpose that for a time enabled people
to come together, resist, and eventually topple the previous regime. Furthermore,
the poor performance of a new and perhaps inexperienced government, as well as
the absence of short-term solutions to underlying economic and social problems

265
and/or the instability and perhaps even violence that can accompany a political
transition, may lead many citizens to wonder whether a democratic political sys-
tem is indeed the best way for their country to be governed.
There is, however, an additional take-away, and this—reflecting lessons derived
from the Tunisian experience—is that failure in the struggle for democracy is not
inevitable. Rather, significant progress can be made if the country and its lead-
ers, both secular and Islamist, are prepared to embrace a political formula based
on pluralism and inclusion. This requires the institutionalized representation of
differing political or ideological tendencies, normally though political parties that
offer citizens a choice in free and fair elections, and the acceptance by these par-
ties and their supporters of the principle that governing requires compromise and
the support of policies that address the needs of all citizens. For the Arab world, or
at least for most of it, this means there must be room for both Islamist and non-
Islamist political parties, with each recognizing the legitimacy of the other, and
that all factions must be committed to a political formula that respects diversity,
encourages and facilitates compromise, and ensures that the will of the people
can, without interference, be expressed regularly and freely.
Acknowledgment
The author acknowledges with appreciation the valuable research and editorial
assistance of Sarah Rusinowski and Lilly Morrison.
Notes
1. Freedom House consistently rated both Tunisia and Egypt as “not free” in the
years leading up to the ousting of Ben Ali and Mubarak. For a short assessment of cor-
ruption in Tunisia and Egypt at the time of the Arab Spring, see Gelvin (2012, pp. 40–
42). Gelvin quotes from US embassy cables made public by WikiLeaks. He notes that in
Egypt some of the protesters chanted, “O Mubarak, tell us where you get $70 billion!”
2. These assessments are based on personal communications with a number of
Tunisian and Egyptian scholars and are echoed in numerous media reports of the
mood in the two countries during this period. Writing about the meaning of dignity,
for example, one analyst wrote: “Authoritarian regimes have many failings. … But
their greatest weakness is moral: They do not recognize the basic dignity of their
citizens and therefore can and do treat ordinary people with at best indifference and
at worst with contempt” (Fukuyama, 2012). More generally, according to an American
scholar of the Arab world, “The Arab uprising unfolded as a single, unified narrative
of protest with shared heroes and villains, common stakes, and a deeply felt sense of
shared destiny” (Lynch, 2012, p. 8).
3. For a useful discussion of Islamist party platforms, see Kurzman and Naqui
(2010). They note that, relative to their starting point, some Islamic parties have lib-
eralized their stances in recent years. But, see also Hamid (2010) for a somewhat
different perspective; for a fuller discussion, see Wright (2012), who reports that most
Islamist parties today do not embrace theocratic rule, even if they are not willing

266
to adopt a completely Western model of democracy. For details about the Muslim
Brotherhood’s platform during the latter years of the Mubarak period, see Rutherford
(2008, especially pp. 179–183).
4. Later accounts raised doubts that the inspector, Fedia Hamdi, had actually
slapped Bouazizi. Hamdi was charged and spent nearly 4 months in jail but was
subsequently released and all charges were dropped. For a fuller account, see “Fedia
Hamdi’s Slap” in The Guardian (April 23, 2011). The article states, “What is indisput-
able is that when Bouazizi tried to retrieve his cart from the police station, he was
turned away. He then asked to see the local governor, but was also refused entry. At
about 1 pm he set himself alight … [and the story of his] self-immolation rapidly
became the stuff of legend in the early days of the [J]‌asmine [R]evolution. It was
reported in media outlets across the globe.”
5. For a useful account, see Schielke (2015). Schielke writes that the “the wave of
politicization, with people participating in protests” was strongest in urban areas “but
it was also felt villages” (p. 191ff).
6. For an editorial expression of these complaints by a liberal Egyptian human
rights activist who had been jailed by the Mubarak government, see S. E. Ibrahim
(2011). Writing during the first weeks of the uprising in Egypt, Ibrahim noted Mubarak
that “has systematically eliminated, harassed or defamed any potential alternative to his
leadership. He has rigged elections, silenced dissidents and prohibited even small polit-
ical rallies. He has beaten and jailed democracy activists and sent civilians to military
courts. He tightly controls the media and has cracked down hard on young bloggers.”
7. For methodological details, reports on specific countries, and downloadable
data files, see the Arab Barometer website: arabbarmeter.org.
8. For an additional discussion of attitudes toward political Islam based on Arab
Barometer data, see Tessler (2010, 2015b) and Tessler, Jamal, and Robbins (2012). For
findings and methodology based on data from a larger number of surveys conducted
in the Middle East, see Tessler (2015a).
9. For a general overview, see Cesari (2014, pp. 123–167). Cesari writes, “Precisely
because they weren’t political parties, Islamic institutions were able to survive under
authoritarian regimes” (p. 123). Useful accounts of factors that contributed to the
strength of the Islamist movements in Tunisia, Egypt and other Arab countries
include Alexander (2000), Wickham (2002), and Schwedler (2006).
10. As the Washington Post (November 30, 2011) reported during the first stage
of the parliamentary elections in Egypt, “The Brotherhood’s unrivaled army of vol-
unteers and campaigners appears to have given it an edge in the first election since
Mubarak was ousted in February” (Londono). For an interesting first-hand report of
the FJP’s mobilization efforts, see Trager (2011).
11. To examine one of the most important early studies of strategic voting, see
Cox (1997). For a valuable study that examines strategic voting in electoral contests
involving competition between establishment “big-tent” parties and those that are
more ideological in character, see Kedar (2005). Other examples of systemic studies
of strategic voting in various contexts include Gschwend (2007); Blais, Nadeau, and
Nevitte (2001); and Alvarez and Nagler (2000).
12. The second-wave Arab Barometer surveys were carried out in Algeria in April
and May 2011, in Iraq in February and March 2011, in Jordan in December 2010, and
in Palestine in December 2010. The third-wave surveys were carried out in Algeria

267
in March and April 2013, in Iraq in June 2013, in Jordan in December 2012, and in
Palestine in December 2012.
13. The poll was conducted on January 13–15, 2015.
References
Al Amraoui, A. (2014). Tunisia’s Ennahda “faces defeat” in elections [On-line]. Al
Jazeera Online, October 28. Available: http://www.aljazeera.com/news/africa/
2014/1 0/t unisia-e nnahda-s ays-s ecular-r ivals-a head-2 0141027141737474289.
html.
Alexander, C. (2000). Opportunities, organizations and ideas: Islamists and work-
ers in Tunisia and Algeria. International Journal of Middle East Studies, 32,
465–490.
Alvarez, M., & Nagler. J. (2000). A new approach for modeling strategic voting in
multiparty elections. British Journal of Political Science, 30(1), 57–75.
Blais, A., Nadeau, R., & Nevitte, N. (2001). Measuring strategic voting in multiparty
plurality elections. Electoral Studies, 20(3), 343–352.
Cesari, J. (2014). The awakening of Muslim democracy: Religion, modernity, and the state.
Cox, G. W. (1997). Strategic coordination in the world’s electoral systems.
Egyptian Center for Public Opinion Research (2015). Poll Results on Political
Conditions in Egypt [On- line]. Cairo: Egyptian Center for Public Opinion
Research, January 27, 2015. Available: https://mail.google.com/mail/u/0/
#inbox/14b2a955881e7ae5.
Fedia Hamdi’s slap which sparked a revolution didn’t happen. (2011). The Guardian,
April 23. Available: https://www.theguardian.com/world/2011/apr/23/fedia-
hamdi-slap-revolution-tunisia
Fukuyama, F. (2012). The drive for dignity [On-line]. Foreign Policy Online, January 12.
Available: http://foreignpolicy.com/2012/01/12/the-drive-for-dignity.
Gelvin, J. (2012). The Arab uprisings: What everyone needs to know. New York: Oxford
University Press.
Gschwend, T. (2007). Ticket-splitting and strategic voting under mixed electoral
rules: Evidence from Germany. European Journal of Political Research, 46(1), 1–23.
Hamid, S. (2010, August). The Islamist response to repression: Are mainstream Islamist
groups radicalizing? Brookings Doha Center policy briefing. Doha: Brookings
Doha Center.
Human Rights Watch. (2013). Egypt: Deeply restrictive new assembly law [On-line].
Human Rights Watch Online, November 26. Available: http://www.hrw.org/news/
2013/11/26/egypt-deeply-restrictive-new-assembly-law.
Human Rights Watch. (2014). Egypt: Unprecedented expansions of military
courts: Decree broadens jurisdiction over civilians [On- line]. Human Rights
Watch Online, November 17. Available: http://www.hrw.org/news/2014/11/17/
egypt-unprecedented-expansion-military-courts.
Ibrahim, Y. (1990). Militant Muslims grow stronger as Algeria’s economy grows
weaker [On-line]. New York Times, June 25, Available: http://www.nytimes.com/

268
1990/ 0 6/ 2 5/ w orld/ m ilitant- m uslims- g row- s tronger- a s- a lgeria- s - e conomy-

weakens.html?pagewanted=all.
Ibrahim, S. E. (2011). Mubarak must leave to save Egypt. The Washington Post,
February 4. Available: http://www.washingtonpost.com/wp-dyn/content/article/
2011/02/03/AR2011020305174.html, p. XX.
Kedar, O. (2005). When moderates prefer extreme parties: Policy balancing in parlia-
mentary elections. American Political Science Review, 99(2), 185–199.
Kurzman, C., & Naqui, I. (2010). Do Muslims vote Islamic? Journal of Democracy,
21, 50–63.
Londono, E. (2011). Egypt’s Muslim Brotherhood claims strong showing in vote. The
Washington Post, November 30. Available: https://www.washingtonpost.com/
world/middle_east/egypts-muslim-brotherhood-claims-strong-showing-in-vote/
2011/11/30/gIQA1TRQEO_story.html.
Lynch, M. (2012). The Arab uprisings: The unfinished revolutions in the new Middle East.
New York: Public Affairs.
Rutherford, B. (2008). Egypt after Mubarak: Liberalism, Islam, and democracy in the
Arab world. Princeton: Princeton University Press.
Schielke, S. (2015). Egypt in the future tense: Hope, frustration and ambivalence before
and after 2011. Bloomington: Indiana University Press.
Schwedler, J. (2006). Faith in moderation: Islamists parties in Jordan and Yemen.
Shah, D. (2011). Arab Spring [On- line]. BBC News Online, December 26.
Available: http://www.bbc.com/news/world-middle-east-16275176.
Tessler, M. (2010). Religion, religiosity and the place of Islam in politi-
cal life: Insights from the Arab Barometer surveys. Middle East Law and
Governance, 2(2), 221–252.
Tessler, M. (2015a). Islam and politics in the Middle East: Explaining the views of ordinary
citizens. Bloomington: Indiana University Press.
Tessler, M. (2015b). Mapping and explaining attitudes toward political Islam among
ordinary citizens in the Middle East and North Africa. Economic Research Forum
Working Paper Series. Cairo: Economic Research Forum.
Tessler, M., Jamal, A., & Robbins, M. (2012). Arab attitudes toward democracy and
governance: Findings from the 2010–2011 Arab Barometer. Journal of Democracy,
23(4), 89–103.
Three Faces of the new Tunisia [On-line]. (2012). BBC News Online, December 17.
Available: http://www.bbc.com/news/world-africa-20755296.
Trager, E. (2011). Why an undemocratic party is going to win Egypt’s first democratic
election. The New Republic, November 29. Available: https://newrepublic.com/
article/97879/muslim-brotherhood-democracy-egypt-elections.
von Rohr, M. (2011). Victory for Ennahda: Why Tunisians voted for the Islamists [On-
line]. Spiegel Online International, October 26. Available: http://www.spiegel.de/
international/world/victory-for-ennahda-why-tunisians-voted-for-the-islamists-a-
794133.html.
Wickham, C. (2002). Mobilizing Islam: Religion, activism and political change in Egypt.
New York: Columbia University Press.
Wright, R. (Ed.). (2012). The Islamists are coming: Who they really are. Washington,
DC: Woodrow Wilson Center Press.

269
APPENDIX Factor Loadings for Items Asking about the Role of

Islam in Government and Political Affairs
Country and Item factor loadings
year [Country name] Religious leaders Religious
would be better (imams, preachers, practices are
off if religious priests) should private and
people hold have influence should be
public positions over government separated from
in the state. decisions. social and
political life.
Tunisia 2011 .833 .863 –.582
Tunisia 2013 .840 .792 –.505
Egypt 2011 .777 .810 –.533
Egypt 2013 .850 .820 –.393
Algeria 2011 .831 .844 –.637
Algeria 2013 .873 .844 –.669
Palestine 2010 .854 .831 –.456
Palestine 2012 .847 .815 –.415
Jordan 2010 .858 .861 –.241
Jordan 2013 .847 .846 –.151
Iraq 2011 .749 .802 –.629
Iraq 2013 .845 .869 –.650

270
271
CHAPTER 9 Autocratic Recidivism

Computational Models of Why Revolutions Fail
Andrzej Nowak, Michele J. Gelfand, Wojciech
Borkowski, and Arie Kruglanski
In this chapter, we introduce a new construct—autocratic recidivism—and

develop a theory of why groups return to autocracy after concerted efforts to
overthrow autocratic regimes. We observe that, throughout the past century, a
similar scenario is repeating itself in different parts of the world. Autocratic
regimes are first brought down by a popular uprising, followed by a short period
of political openness, when diverse contenders for power compete freely for
control of the government. In many cases, these attempts at establishing democ-
racy fail and the government is replaced either by the formation of a similar
autocratic government or by a brand new and even more repressive regime—a
phenomenon termed autocratic recidivism. Adopting a dynamical systems per-
spective (Nowak & Vallacher, 1998), we (a) argue that the capacity to synchro-
nize actions is critical for efficient functioning of society; (b) examine the notion
that meso-level institutions provide the conditions that allow societies to achieve
and maintain synchronization during a transition to democracy, and the lack
of meso-level institutions produces a lack of synchronization that can result in
frustration of basic needs and facilitate the return to autocracy; and (c) imple-
ment computer simulations that illustrate some initial support of the theory.
Our models also suggest reasons for why some nations will return to another
autocracy versus total chaos and civil war, as in the case of Egypt and Syria,
respectively.
Computational Model of Autocratic Recidivism

As just stated above, we have witnessed a similar scenario repeat itself I differ-
ent parts of the world. Also as just noted, in these scenarios, autocratic regimes
fall as the result of a popular uprising, and there is a short period of political
openness when diverse contenders for power compete freely for the control of
the government. In many cases, such as the Iranian Revolution of 1979 and the
Egyptian Arab Spring of 2011, when authoritarian rulers were overthrown through
mass demonstrations, the attempts at establishing democracy did not succeed and
272
eventually resulted either by the formation of a similar autocratic government or

by an even more repressive one. What explains such “pendulum shifts?” How can
we predict when and why revolutions that overthrow autocratic governments will
fail and be replaced with another autocracy or when they will succeed and ulti-
mately result in a democratic government? We know from history, for example,
that peaceful, democratic transitions can be successful. As a case in point, in east-
ern Europe after Poland transitioned successfully from Communism to democ-
racy in 1989, numerous countries followed suit and transitioned from autocratic
governments and centrally controlled economies to democratic and decentralized
market economies (Blanchard, Froot, & Sachs, 2007).
In this chapter, we develop a theory of autocratic recidivism by drawing on
dynamical systems theory (Nowak & Vallacher, 1998; Vallacher & Nowak, 1994).
We implement computer simulations to test some of its assumptions. We first
discuss general factors that precipitate revolutions, followed by a discussion of
factors that we theorize are necessary for successful transitions to democracy. We
then present several initial simulations to test different scenarios derived from
the theory. We conclude with theoretical and practical implications and future
directions.
Scenarios Following the Collapse of Autocracy

Autocratic governments, by definition, control society tightly through the enforce-
ment of strict rules, harsh punishments, and hierarchical top-down decision mak-
ing. Top-down control minimizes the influence of individuals on national policy,
any expression of dissatisfaction is suppressed, and self-organizing structures
are dismantled. The strong control of information leads to a large discrepancy
between public perceptions steered by state-controlled propaganda and reality.
Issues such as government inefficiency, injustice, and possible corruption typi-
cally do not reach the level of public discourse, although some instances of it may
be suspected or known by individuals. Although individuals may dislike the gov-
ernment, society is fragmented, and members express their attitudes only among
families or friends. This fragmentation prevents any coordinated efforts directed
against the government. Autocratic governments not only exercise control over
individuals, but also over radical movements and extreme religious organizations.
In some societies, however, as we have seen in the Arab Spring, autocratic
structures become challenged through collective action. Such collective action can
be stimulated by a number of factors that destabilize the autocratic regime, includ-
ing allegations of corruption, perceptions of social inequality, the inability to meet
basic needs, limitation of freedom, perception of civil unrest in neighboring coun-
tries, or pressures from democratic government or even public opinion abroad.
This stage is characterized by an increase in communication among individuals
that escapes governmental efforts to control it, and the emergence of horizontal
communication networks. Public protests often take the form of civil resistance,
such as strikes, demonstrations, and sit-ins, but sometimes also riots and even

273
civil wars (e.g., in Syria). In most cases, the common factors are the rise of demo-
cratic ideology, rise of antigovernment attitudes, and weakening of the control
of the government, all of which facilitate a shared collective belief that the gov-
ernment can be overthrown. These processes are mutually synergetic—meaning,
each facilitates the other.
For instance, during the Arab Spring, the availability of social media facili-
tated collective action against autocratic regimes. Although social media are not
the cause of the process, they helped to spread information in a horizontal way,
not controlled by government, and provided a tool for organizing social move-
ments (Howard et al., 2011; Stepanova, 2011). Social media provided a channel by
which individuals could communicate in peer-to-peer fashion, and informal lead-
ers could coordinate an outreach. Individuals using these channels could share
information suppressed from the public debate, such as examples of corruption,
which undermined the legitimacy of the government. Individuals could also share
democratic values and coordinate their protests. More generally, in this prototypi-
cal scenario, new democratic structures are created quickly and start to provide
alternatives to governmental structures of top-down control. In addition, informa-
tion about democratic uprisings in other countries of the region spurs a belief
that democratic transition is possible. In a positive feedback loop, the government
begins to lose control, people are less afraid of it, which in turn increases protests
and civil disobedience. During social uprisings, the old hierarchical structures
providing coordination weakens or falls apart, and is replaced temporarily, at least
in part, by self-organizing structures.
At this point, the collapse of autocratic regimes can result in a new self-
organized order or at, the opposite extreme, chaos and disorder. As such, the
final outcome of this process can be transition to a democratic political order,
the return of the old autocratic regime, or an even more autocratic regime (e.g.,
a radical religious organization). The final outcome may also be the outbreak of
a civil war. We hypothesize that an important factor that affects whether a society
will continue on the road to democracy or result in returned autocratic gover-
nance is the degree to which the emerging self-organized structures can maintain
social order and functioning of the society. For a democratic transition to succeed,
the emerging structures must be capable of coordinating basic functions such as
building an infrastructure, providing security, and satisfying basic needs on a soci-
etal, not a local, scale. It must, in our theory, provide societal-level synchronization.
At a minimum, it must provide enough synchronization to organize individuals
to a sufficient degree so they can cooperate to satisfy their basic needs. A lack of
such synchronization leads to mass discontent. After all, democracy is not simply
an end, but a mechanism for more effective social provisions and satisfaction of
people’s needs.
Next, we discuss the construct of synchronization and its basis in dynamical
systems theory (Nowak & Vallacher, 1998; Vallacher & Nowak, 1994). We then dis-
cuss what factors may cause the emergence of low versus high societal synchro-
nization after revolutions, and their implications for autocratic recidivism versus
democracy. Following that, we provide an initial test of our predictions using
Autocratic Recidivism | 273

274
computer simulations, which are the tool of choice for investigating the dynam-
ics of social processes, particularly for addressing how complex social processes
emerge from interactions among individuals (Epstein, 2006; Gilbert & Troizsch,
2005; Nowak, 2004; Nowak et al., 2013).
The Role of Synchronization in Societal Systems

The dynamical systems approach offers a formal perspective to understand syn-
chronization in societies (Nowak &Vallacher, 1994, 1998; Vallacher & Nowak,
1997; Vallacher, Reed, & Nowak, 2004). Generally speaking, a dynamical system
is a set of interacting elements that change in time. Psychological and social phe-
nomena at different levels of analysis can be understood as dynamical systems.
Neurons fire at a given rate because they receive activating and inhibiting signals
from other neurons and, by firing, they change the firing rate of the neurons to
which they are connected. The functioning of the mind can be considered to be a
dynamical system (Port & Van Gelder, 1995) in which activation of some mental
elements (thoughts and emotions) leads to activation of some other elements, and
suppression of the level of activation of other elements (Lewis, 1995). At the level
of a dyad, the state of one individual in the next moment in time depends to some
degree on the current state of this individual and to some degree on the state of
the partner (e.g., Gottman, 2002). At the level of groups and societies, the state
and action of each individual depends both on the previous state of these individu-
als and also on the influence of others.
Within this perspective, we propose that one critical variable that affects auto-
cratic recidivism after the fall of autocratic regimes is the capacity of the society to
achieve societal-level synchronization among individuals and social groups. From
a dynamical systems perspective, synchronization is the coordination among enti-
ties in time, or temporal dependence of the elements in a system (Delaherche
et al., 2012) Synchronization has been studied across many different levels of
analysis. In dyads, synchrony refers to the degree of congruence, or coordination
in time of the behavior between two individuals (Condon & Ogston, 1967; Baron
et al., 1994; Newtson, 1994; Schmidt & Richardson, 2008). Movement and speech
of one individual is related in time to the movement and speech of another. In a
conversation, for example, when one partner speaks, the other is silent, communi-
cating understanding nonverbally through nods, gestures, and facial expressions,
and then the partners in the conversation switch the active and the passive roles.
With an increase in tempo or stress, more complex forms of synchronization tend
to break down (Kelso, 1997). In an argument, for example, as partners start to
interrupt each other, the synchronization in a conversation breaks down.
In social groups, synchronization refers to the temporal interdependency of the
states and actions of some individuals on the states and actions of other individu-
als. The tendency to synchronize causes individual behaviors to coalesce into reg-
ular patterns of individuals acting together (Guastello & Guastello, 1998; Marsh,
Richardson & Schmidt, 2009). The social group is thus defined as not only a
set of individuals, but also as a constant process of synchronization of gestures,
gazes, communication, and actions (Arrow, McGrath & Berdahl, 2000; Nowak &

275
Vallacher, 1998; Nowak et al., 2002). Synchronization leads to coherence in social

systems, and the capacity for achieving synchronization is critical to the effective
functioning of groups and organizations (Araújo, Silva, & Davids, 2015; Davis,
Brooks, & Dixon, 2016; Richardson, Garcia, Frank, Gergor, & Marsh, 2012; Fowler
et al., 2008).
Finally, at the societal level, synchronization among individuals and social
groups is critically important to the ability of a society to perform its functions
and satisfy the basic needs of individuals (e.g., food, safety, education). To be
sure, different social functions require synchronization of different magnitudes.
Building a home requires several individuals to synchronize, whereas providing
security across an entire nation requires synchronization on a much larger scale.
Synchronization satisfies basic needs both within levels in a society (e.g., between
members of a family or players on a sports team) as well as between levels in a
society (e.g., between families and the local government, and between local gov-
ernments and the central government), and both are important for the efficient
functioning of societies. Observing synchronized actions in one’s social environ-
ment instills the feeling of coherence and order, and a feeling of belonging and
significance, because a system that allows individuals to have control and attain
their goals gives them a sense of meaning. In contrast, a lack of synchronization
across individuals and groups in a society interferes strongly with the govern-
ment’s ability to perform its basic functions and to satisfy individuals’ basic needs.
The lack of capacity to synchronize within and across social groups leads to a feel-
ing of alienation and loneliness, confusion, fear, and anger. In turn, the lack of
satisfaction of needs can also lead to further desynchronization. Ultimately, a lack
of synchronization leads to increasing disorder, anomie, and chaos. When society-
level synchronization is low, as we discuss next, we predict that individuals will
tend to support any structure that can provide coordination.
Predicting Synchronization in Societal Systems Following

Revolutions: The Critical Role of Meso-Level Institutions
With this conceptual background in mind, we now turn to our central question—
namely, how can we predict when revolutions that overthrow autocratic govern-
ments will fail and be replaced with another autocracy, or succeed and ultimately
result in a democratic government?
Before uprisings, autocratic governments can usually provide quite a high
level of synchronization among individuals. In these contexts, even if individuals
are dissatisfied, it is typically not expressed because of high synchronization and
repression. If there are factors, however, that increase dissatisfaction to very high
levels, such as sudden changes in economic conditions or events that cause the
disruption of social order, individuals may rebel against authority.
From the perspective of our synchronization theory, democratic uprisings
destabilize and disassemble autocratic governments that had formerly provided
high levels of societal-level synchronization by top-down control. Destabilizing the
autocratic government thus destroys the hierarchical structure that had formerly
provided synchronization. The success of an attempt to establish democracy thus

276
depends critically on whether groups and individuals can maintain a sufficient

degree of synchronization during the transition period after the collapse of autoc-
racy in their effort to provide basic resources and, ultimately, to establish demo-
cratic structures of governance that provide more top-down control. However, if
the social organization that was sufficient to bring down the autocratic regime is
not sufficient to provide enough synchrony, and thus basic needs for the function-
ing of the society, such as food, justice, conflict resolution, and security, this likely
results in a failed democratic transition. In this case, although initially enthu-
siastic about their new-found freedoms and the collapse of despotic regimes,
when there is low synchronization, individuals sense the increasing chaos and
feel increasingly insecure and frustrated by society to satisfy their basic needs.
Synchronization may also collapse because of the failure of convergence of dif-
ferent interests, as was the case in Egypt, where concerted efforts were made by
power interests groups within and outside the old regime to cause economic prob-
lems (e.g., scarcity of gas and electricity) that led to the inability to meet basic
needs and mass dissatisfaction with Mohammed Morsi.
Fundamentally, when synchronization is very low, individuals may be more
willing to sacrifice their freedom in exchange for security and may, ironically, pre-
fer strong top-down control that can provide synchronization and the satisfaction
of basic needs. This paves the way for the return of the old autocracy or for radi-
calization in which a new, even more autocratic government—often formed by an
extreme religious group—can take power in a vacuum of low synchronization.
A critical question, then, is: What factors affect the ability of individuals and
groups to coordinate at necessary levels after the collapse of autocratic govern-
ments? We hypothesize that the existence of strong meso-level structures can pro-
vide the scaffolding that can help sustain a sufficient level of synchronization
during the interim period before the new democratically elected administration
can function effectively, as long as it is not controlled by autocratic forces. By
meso-level structures, we refer to all the stable and temporary structures that exist
between the micro level (the individual) and the macro level (society) (Bourdon,
2002; Cappelli & Sherer, 1991; Lin & Li, 2003). Meso-level structures are com-
posed of formal structures including firms, organizations, and associations, and
also informal structures such as neighborhood groups, volunteer and grassroots
organizations, friendship circles or coalitions, and the like (Faist, 2010).
From a network perspective, meso-level structures involve a heterogeneous
and changing set of ties that connect individuals to individuals, individuals to
groups and organizations, and organizations to other organizations. It allows
for individuals to have weak social ties and numerous connections to others in
a social system in addition to strong local (often family/household) ties. As such,
meso-level structures often result in higher collective levels of trust, potency, and
commitment (Bligh, Pearce, & Kohles, 2006), or what has also been termed social
capital (Fukuyama, 2001; Putnam, 1993). More generally, meso-level structures, to
a large degree, determine the flow of information in society and provide an alter-
native to the top-down control of autocratic governments. They allow synchroniza-
tion to occur through bottom-up organizing.

277
Importantly, it should be noted that meso-level structures can provide synchro-

nization during democratic transition only if they are, at least to some degree,
autonomous of the autocratic regime. In essence, meso-level structures can pro-
vide synchronization for democratic processes only to the degree they can carry
content that is needed for the synchronization outside the autocratic power net-
work. In a society, many meso-level structures are not neutral politically; they are
conductive selectively to special types of information (e.g., pro or anti government
or religion). They can coordinate on some type of messages and actions, and are
opaque to others. Thus, at the societal level, some meso-level structures are not
neutral politically. This is especially true for autocratic regimes. Autocratic regimes
try to control as much of the meso-level structures as possible. For example, in
autocracies one may need government permission to start any type of organiza-
tion or association. From this perspective, is it not just how rich the meso-level
structure is, but rather how much the meso-level structure is autonomous from
the autocratic government that helps to pave the way for democracy.
In sum, the main hypothesis that we sought to test is that strong meso-level
structures play a crucial role in successful transitions to democratic society as long
as they are autonomous from the autocratic network. Autocratic regimes rely on
top-down control and repress grassroots organization through the enhancement
of top-down asymmetric relations and destruction of horizontal links. Put differ-
ently, they tend to destroy the independent meso-level and instead rely on very
rigid hierarchical networks to create synchronization across the society. In con-
trast, democratic societies rely, to a large extent, on self-organization stemming
from meso-level bottom-up processes (e.g., lobbying, public discussion, citizen
initiatives). These processes rely on individuals communicating with other indi-
viduals or, in other words, on symmetric horizontal links. Strong meso-level struc-
tures thus provide a way for individuals in a society to synchronize without the
coordination provided in a top-down manner by a central authority. Accordingly,
we hypothesize that during the transition period following the collapse of auto-
cratic governments, strong meso-level structures can enable the maintenance of
social order and provide societal-level synchronization essential for societal func-
tioning. In contrast, when meso-level structures are weak, and ties are largely local
(e.g., within the family or local neighborhood), the decomposition of the auto-
cratic structures of power can result in low societal synchronization and chaos,
and the inability to provide basic needs for the functioning of the society. This,
in turn, can result in sentiments for strong government that pave the way for the
return of autocracy.
Computer Simulations of Synchronization in Societies

Following Revolutions
We have argued that the capacity to maintain basic societal functions is one of the
critical factors that determine the success of revolutions. As we argued earlier, cru-
cial societal functions can be recast in terms of synchronization and, accordingly,

278
in our computational modeling we have thus focused on investigating synchroni-

zation in social groups.
Historically, formal models of synchronization were developed in physics,
which illustrated that when systems that are similar influence each other—even
weakly—they tend to synchronize (Strogatz, 2014). For example if several pen-
dulums hang on the same string, such that the movement of one pendulum, by
the movement of the string, influences the others, the pendulums tend to start
swinging in synch. Synchronization has since then been shown to be a power-
ful phenomenon that occurs in many physical and biological systems (Pikovsky,
Rosenblum, & Kurths, 2003; Strogatz, 2003), including flashing fireflies, sing-
ing crickets, cardiac pacemakers, firing neurons, human intestines, clocks, and
applauding audiences. To our knowledge, this is the first work to apply formal
models of synchronization to the study of cultural dynamics. In the following
pages, we first discuss how synchronization is defined conceptually in compu-
tational models, before turning to how we model it formally in our simulations.
We then discuss how we modeled different network structures—autocratic and
meso-level structures—and describe the results of our simulations of scenarios
that follow the collapse of autocratic governments.
Formal Models of Synchronization

In physics, synchronization has been studied by computer simulations in the
framework of coupled dynamical systems (e.g., Pecora & Carroll, 1990; Kaneko,
1991; Gade & Hu, 2000; Pikovsky et al., 2003). A dynamical system is a system that
evolves in time (Nowak & Lewenstein, 1994; Strogatz, 2014). Coupling means the
state of each system in the next moment depends, to some degree, on the current
state of the system and, to some extent, on the state of the other system. This
general principle has been also used to model social systems (Nowak & Vallacher,
1998). For example, when modeling interactions of married couples, it has been
shown that the emotions of one partner in the next moment depend, to some
degree, on a person’s current emotions (e.g., a person who is happy tends to be
happy in the next time moment) and, to some degree, on the emotions of their
partner (Gottman, 2002). Synchronization has been also demonstrated in cou-
pled dynamical systems capable of much more complex dynamics, such as when
the relation between the previous and the next values is nonlinear, in what has
been termed nonlinear dynamical systems (Pecora & Carroll, 1990; Pecora, Carroll,
Johnson, Mar, & Heagy, 1997; Strogatz, 2014). Coupled nonlinear dynamical sys-
tems, for example, have been used to model synchronization among partners in a
dyad in which the rules for behavior are more complex—such as when one’s state
might have the opposite effect on one’s partner (Nowak, Vallacher, & Borkowski,
2000; Nowak, Vallacher, Zochowski, 2005).
Following the principles of dynamical minimalism, our goal was to construct
the simplest possible model that could represent the crucial phenomena of inter-
est: synchronization in social systems. To model behavioral synchronization in the
simplest way, we modeled synchronization as the coordination of rhythms among
agents. In particular, we examined pulsating agents in which each pulsation is

279
interpreted as a cycle of behavior of an individual, and we investigated under what

conditions and to what degree the pulsations of different agents synchronize.
Key Parameters of the Model

The agent in our simulation corresponded to an individual. The agent is repre-
sented as a dynamical system consisting of a single dynamical variable, x, and a
control parameter, r. In this representation, x corresponds to behavior. This is the
simplest possible representation of behavior: x-like behavior changes in time. It
could be interpreted as the intensity of behavior (i.e., the number of body parts
that are moving at the moment [e.g., Newtson, 1994]). The control parameter, r,
represents individual differences; r has a specific value for an agent and it does
not change in time. It controls the dynamic of the behavior of an agent (i.e., how
x changes in time). Parameter r, our control parameter, thus decides how fast the
changes of x occur in the system. Put simply, the magnitude of r determines the
complexity of the dynamics of x. For low values of r, the dynamics are simple and
regular; for greater values of r, the dynamics become complex and irregular. For
low values of r, after a relatively short amount of time, variable x stabilizes on
some value, and it returns to this value if x is perturbed by external influence. In
dynamical systems parlance, the system converges on a fixed-point attractor. For
greater values of r, the dynamic variable, x, oscillates regularly between a set of
values; the system converges on a periodic attractor. For even greater values of r,
x evolves in a very irregular, complex, seemingly chaotic way; the system follows
a chaotic attractor. In sum, in our simulations, r represents individual differences
among agents and describes the complexity of individual behavior. Psychologically
speaking, low values of r induce dynamics of x corresponding to a simple, regular
way of life. Greater values of r induce the dynamics of x corresponding to active,
irregular patterns of actions full of unpredictability and change.
Adopting the models developed in biology and physics (Kaneko, 1991), we use
a network of logistic maps to simulate synchronization in social systems. A logistic
map (May, 1976; see also Nowak & Vallacher, 1998; Nowak et al., 2000, 2005) is
a specific equation describing the state of the system in the next time moment,
depending on the current state. It is called a map because it maps the next state of
the system on the current state. In general, for lower values of r, a system described
by logistic maps converges on a specific value or pulsates in a more regular way;
for greater values of r, it pulsates in a more complex way. For more technical detail
on the mathematics of logistical maps, please see the Appendix.
Simulation Dynamics
In our simulation, each agent is connected in a network to some other agents. The
next state of an agent depends, to some degree, on the current state of the agent
and, to some degree, on the influences of the others to whom the agent is con-
nected in the network. Agents thus influence others to whom they are connected
in the network. The strength of this influence is proportional to the strength of
the social relations represented as connections in the network between the agents.

280
At the beginning of the simulation, we defined the network of connections

between the agents by programming who is connected to whom and the strength
of their connection. Each agent is assigned a value of parameter r, which was—for
most of the agents—assigned randomly from a flat distribution in the specified
range. Agents representing local leaders in the autocratic power network were
assigned the same value of r as the autocratic leader to represent that their dynam-
ics in the power network is governed by the same principles, and their behavior
is controlled by principles set by the leader. All the agents were assigned, at the
beginning, a random value of variable x; in others words, they were not synchro-
nized at the beginning.
In successive simulation steps, the state of each agent was calculated depend-
ing on the current state of the agent and the sum of influences of others to whom
the agent is connected in the network. Then, the new state, x, was assigned to the
agent and the procedure was repeated in the next simulation step. During each
step, the degree of synchronization was assessed by applying the entropy mea-
sure, which calculates the degree of synchronization versus chaos (this measure
is discussed in more detail later).
Manipulating Social Network Structure

Social networks represent the strongest available tool for the description of social
structures (e.g., Freeman, 1979; Girvan & Newman, 2002). In social networks,
individuals are represented as nodes; social relations are represented as links (i.e.,
ties between agents). We used two architectures of social networks: the hierarchi-
cal autocratic network and the meso-level network.
Autocratic Networks
In an autocracy, the flow of decisions is from the top—from the leader—
downward through administrative figures to ordinary citizens. Influence in an auto-
cratic society is thus asymmetric; the leader is linked with strong connections to
his or her subordinates (e.g., regional governors) and each subordinate is linked to
ordinary citizens by asymmetric connections. The influence of the leader on citizens
is very strong whereas the influence of ordinary citizens on the leader is minimal.
In our model, the autocratic network had a recurrent structure with three levels
of hierarchy in which the groups were composed of nine individuals. At the high-
est level there was only the ruler and eight governors controlled by the ruler. This
represents the highest structure of power. The ruler was connected by strong out-
going connections to all the members of his group (i.e., governors). Each governor
controlled (had strong connection to) the leader of one of eight regional groups.
As such, the regional leaders represented the middle level of the power structure.
In the lowest level, there were 64 groups, each composed of a local leader and
eight subordinates. In total, there were 729 agents in the simulation. At each level,
the leader controlled all the members of his or her group.
The structure of the hierarchical network was generated by a two-dimensional
lattice resembling a chessboard. Specifically, each nine-person group can be visual-
ized as a square chessboard of three by three. The agent in the middle represented

281
the leader; the other eight individuals represented subordinates. The leader was
connected to all the subordinates and exerted a much stronger influence on them
than the subordinates exerted on the leader, and this was the case at each level of
the group. Because the structure of influence in autocracy is top-down, the con-
nection strength is much stronger from the higher level of the hierarchy to lower
levels than the reverse.1 We assumed that all the agents representing autocratic
government follow the same relatively simple rules (i.e., low levels of r, the control
parameter). All individuals representing the power network had the same value of
the control parameter, which was somewhat different from the agents correspond-
ing to ordinary citizens outside of autocratic network. This represented a situation
in which those in the autocratic network of power share the preference for the
same dynamics as the autocratic leader, which is different from the preferences of
the ordinary citizens.
Meso-Level Networks
In contrast to the autocratic network, the meso-level network structure was
represented by adding horizontal, symmetric, distant random connections to the
hierarchical network. This network represented social relations outside the auto-
cratic network. These links represented connections to others of similar power
(i.e., both connected agents’ influence on each other to a similar degree2). In a
meso-level network, individuals may form social ties with more distant individ-
uals, beyond immediate family and close friends. These connections represent
friendships, acquaintances, informal ties, and common membership in formal
and informal organizations. The meso-level network thus adds symmetric dis-
tant connections, where the probability of a being connected to an agent does not
depend on the distance in a similar way as in constructing a small world network
(Watt, & Strogatz (1998) In a meso-level network, even a small number of distant
connections, also called weak ties, may be of crucial importance to the functioning
of an individual and developing social capital (Granovetter, 1973). In the condi-
tion of a strong meso-level structure, it was assumed that an agent, on average,
was connected to two other agents by horizontal, distant connections in addition
to local connections. The strength of the meso-level structure is one of the most
important variables manipulated by our simulation. This allowed us to examine
how the richness of the meso-level structures affects synchronization with the
presence of the autocratic network and how it can provide synchronization when
the autocratic network of power is destroyed.
Simulating Scenarios after the Collapse of Autocratic Regimes

Two factors—our independent variables—were varied between simulations.
First, we manipulated the strength of the meso-level structures by whether
agents have distant connections in the horizontal network (representing the
presence of a rich meso-level structure) or whether hierarchical connections
representing only the autocratic power structure were present (weak meso-level
structure). Second, we manipulated whether the society was relatively homo-
geneous with respect to how individuals function or very heterogeneous with

282
respect to the functioning. This was achieved by varying the diversity of the
values of the control parameter, r, in the society—from highly homogeneous
to highly heterogeneous. In our simulations, we thus varied two independent
variables in a two-by-two design: whether a meso-level network is present and
whether a society is homogeneous or heterogeneous.
Our main dependent variable was the synchronization between the states of
agents. Synchronization was measured by entropy. Low entropy described the
highly ordered state of the systems in which the agents were in similar states at
the same moment in time. High entropy, in contrast, described a highly disor-
dered state—a society in a state of very low synchrony. Thus, the more synchrony
in the society, the lower the entropy. To construct a measure that was more intui-
tive, instead of a raw value of the entropy, we provided the amount of entropy as
a percentage of maximal entropy possible in the system. Entropy of 0% describes
a perfectly synchronized system whereas entropy of 100% describes a fully disor-
dered system. To get a more stable measure of synchronization, we averaged the
measure of entropy over the last 100 steps of the simulation.
Based on the theory presented, after the collapse of autocracy that had pro-
vided a high degree of top-down synchronization, we expected the highest level
of entropy (i.e., the most disorder and lowest synchrony) would occur in systems
that have low levels of meso-level structures and which have a high degree of
heterogeneity (composed of individuals who represent different values, such as
different identities, ethnic groups, lifestyles, and so on). In contrast, we expected
that the lowest level of entropy (the most order and highest synchrony) would
occur in systems that have high levels of meso-level structures and high levels of
homogeneity.
Simulating the Removal of the Autocrat from Power

All simulations started with the preset top-down autocratic network structure,
and, in the condition of a strong meso-level structure, distant connections. For
every simulation, there also was a preset heterogeneity controlled by the diversity
of control parameter r. The initial values of x were set to a random value between
0 and 1. After less than 50 simulation steps, the society achieved relatively high
synchronization. All the agents at the same moment were in the same state (i.e.,
the agents in the whole network were highly synchronized).
To model the collapse of the autocracy, after 1,500 simulation steps, the dictator
was removed but the rest of the autocratic apparatus stayed in power. This may
correspond to removing a dictator who has lost legitimization. The results showed
that removal of the dictator led to the breakdown of global synchronization. The
society then consisted of several locally synchronizing groups. Thus, the removal
of the ruler broke down the overall synchronization in society, yet the remaining
autocratic power structure provided high synchronization at the local level. Put
differently, the society broke down into fractions controlled by local leaders. The
lack of synchronization among the local leaders may indicate conflict emerging
between the local leaders. The highest disruption of synchronization occurs at the
highest level (i.e., among the governors who had been controlled by the ruler).

283
Simulating the Removal of the Autocratic Structure of Power

The next phase of the simulation represented the process by which democratic
movements are able to gain support at the expense of traditional government
systems after the removal of an autocratic leader. As our next manipulation, the
whole power structure of the formerly autocratic society was removed. To simulate
the removal of the power structure, we reset all the strong connections that gave
asymmetric power to local leaders by decreasing their strength to the strength
equal to connections in the horizontal networks.3 At the same time, we raised the
strength of the connection from the citizens to the leaders to make them symmet-
ric, such that citizens now had the same influence on the former leaders as the
former leaders had on them.
Simulating the Role of Meso-Level Structure
The aim of the final stage of the simulation was to investigate whether the
capacity to synchronize after the removal of the autocratic structure depends on
the strength of the meso-level structures. We thus investigated the conditions
under which a predemocratic transitional society—namely, one that had enough
capacity to overthrow autocratic structures—could maintain sufficient societal
synchronization to fulfill its base functions and satisfy the basic needs of individu-
als. We assumed that if the society maintained a sufficient level of synchronization
for an extended period of time, then it would be much easier to establish a stable
democracy. The collapse of synchronization, in contrast, paves the way for the
reestablishment of a new autocratic structure.
The next simulations investigated the joint effects the meso-levels structures
and the homogeneity of the population control parameter in the society on the
ability to synchronize after the collapse of autocratic structures. The four condi-
tions included (a) weak meso-level structures with a heterogeneous population,
(b) weak meso-level structures with a homogeneous population, (c) strong meso-
level structures with a heterogeneous society, and (d) strong meso-level structures
with a homogeneous society. The four conditions correspond to the combinations
of the strong versus weak meso-level structures and homogeneous versus hetero-
geneous society.
The simulations revealed that, in the network without many horizontal, dis-
tant connections (weak meso-level structures) and a highly heterogeneous con-
trol parameter, synchronization is very weak; it occurred only at a very local level.
In this condition, local groups were completely desynchronized with each other;
the society was fragmented and unable to function as a whole. Entropy reached
the highest value of all the conditions (>80%). The simulations also showed that,
even with a small number of horizontal, distant connections (weak meso-level
structures), when the control parameter was homogenous, local synchronization
dominated and the society maintained some degree of synchronization, although
it was primarily local. In this case, entropy had the intermediate value (~50%).
When there was a rich meso-level structure (i.e., many horizontal, distant con-
nections) and high heterogeneity, some synchronization started to emerge at the
global level, but local synchronization was weak. In this case, entropy increased

284
(>60%). Finally, and most important, the strongest synchronization after the col-
lapse of autocracy was achieved when the meso-level structure was strong and
when the society was homogeneous. High synchronization was evidenced by low
entropy in this case (<50%).
Both the removal of the authority figure and the downfall of the autocratic
structure decreased synchronization, as evidenced by the sharp increases of
entropy in the beginning of the simulation for all societies. Over time, the best
synchronization can be achieved by the presence of rich meso-level structures and
high homogeneity of the control parameters (i.e., low individual variation). This
condition can attain and maintain levels of synchronization comparable with the
synchronization realized by the formerly autocratic hierarchical control structures.
Discussion
In this era of conflict and instability, we are witnessing the collapse of autocratic
governments—from Egypt to Tunisia to Syria and beyond—at unprecedented
rates, and there is a great need for theory and research to understand the possible
scenarios that can ensue thereafter. How can we predict whether the collapse of
autocracy will lead to democratic transitions or to the return of autocracy in full
force, what we have termed autocratic recidivism? There are, to be sure, many fac-
tors that can affect such scenarios. In this chapter, we introduced a new perspec-
tive on the importance of the achievement of societal-level synchrony, which is
dismantled after the collapse of autocracies, and the factors that affect the reestab-
lishment of synchrony, which we argue is necessary for effective societal function-
ing and, ultimately, critical to the transition to democratic structures.
The results illustrate that after the collapse of autocratic structures, societies
that have low levels of distant connections (i.e., have very weak meso-level struc-
tures) and that are highly heterogeneous experience a high degree of disorder
and low synchrony. Put simply, although collective empowerment to overthrow
governments is key to early stages of democratic transitions, in societies in which
there are few distant connections and in which there is high heterogeneity, the
breakdown of autocratic top-down control is likely to lead to local synchroniza-
tion within specific groups rather than on the scale of the whole society. This
level of synchronization, however, is not capable of sustaining large-scale national
efforts such as the coordination of the creation of a new government along with
other basic functions such as the building of roads and infrastructures. As we
argued, it is precisely these conditions in which we expect that new autocratic
governments—and those that might even be more extreme—will be necessary to
reestablish social order and provide synchrony on a large scale. It is important to
note that autocratic governments thrive on top-down control and on the destruc-
tion of meso-level institutions. Thus, our results suggest that, after the collapse of
autocratic governments, with no meso-level institutions, it will be very difficult to
establish societal synchrony, particularly in contexts where there is high diversity,
resulting in pressures toward autocratic recidivism.

285
We note that the low levels of synchronization in such scenarios also raises
another even more extreme possibility. If the democratic transition fails and no
autocracy is strong enough to rise to power in the situation of chaos and the lack of
synchronization, this may result in a fragmented society—with segmented groups
that synchronize internally and remain in conflict with each other. Moreover, it
is precisely in situations of low societal synchronization when participation in
radical networks may provide the synchronization necessary for achieving basic
needs. Radical groups such as ISIS not only disrupt societal synchronization,
but also they attract support not because people agree with their values, but
because these groups are able to provide strong synchronization and fulfill basic
needs: food, shelter, and safety. From the perspective of synchronization, in the
long run this may generate support for ISIS, even among individuals who initially
opposed their ideology strongly. Put simply, they may forfeit their values in favor
of supporting a force that provides synchronization, order, and better functioning
of the society in place of low synchrony and chaos.
From the perspective of our model, the difference between Egypt, which expe-
rienced the return to autocracy, and Syria, which experienced civil war, may have
been the result of the strength of the military. Egypt has a strong military, which
prevented chaos and the breakdown of the country into fragments, providing syn-
chronization when the power of the government had weakened. In Syria, a weaker
military was fully controlled by the autocrat, who was the subject of much dissat-
isfaction, so the country became divided. Moreover, in Syria, the opposition was
heterogeneous, with a more democratically oriented opposition but also several
radical factions. Accordingly, after the autocratic network lost power, it still con-
trolled a fragment of the country, yet the opposition could not synchronize and
broke down into several groups in conflict. We plan to model such scenarios in
the future.
Our simulations also showed an alternative scenario— namely, that those
societies that have horizontal networks in which there are distant connections,
or strong meso-level institutions, can achieve and maintain level of synchroniza-
tion comparable with the synchronization provided by autocracies, particularly in
contexts when there is homogeneity (when people agree on basic values). In this
respect, networks with rich meso-level structures that provide distant links pro-
vide the structure necessary for self-organization during the transition period. If
a sufficient level of synchronization is achieved, a society can perform basic func-
tions. Accordingly, individuals’ basic needs are satisfied, there is a feeling of safety
and order, and the meso-level structures can provide the scaffolding for the emer-
gence of a democratic government. Although our data were simulated, we can see
real-world examples—such as during Poland in the 1980s, where there was much
meso-level organizing—that follow our results in some ways (Blanchard, Froot, &
Sachs, 2007).
More generally, our results contribute to a growing body of literature on synchro-
nization across multiple levels of analysis and entropy in societal systems. When
systems have very low synchrony and high levels of disorder, they are not func-
tional. At the opposite extreme, extreme autocratic contexts—in which top-down

286
control provides total synchrony and very little latitude—are also not functional.
This is consistent with recent data that show that, across 32 nations, societies that
had extreme tightness (strong rules and very strong punishments for deviance)
as well as societies that had extreme looseness (few rules and little punishment of
deviance) exhibited poor psychosocial outcomes (lower levels of happiness, greater
dysthymia, higher suicide rates) and poor health (lower life expectancy, greater
mortality rates from cardiovascular disease and diabetes), and were worse off eco-
nomically and politically (lower gross domestic product per capita, greater risk for
political instability) compared with societies that had a balance of constraint and
latitude (Harrington, Boski, & Gelfand, 2015; see also Gelfand et al., 2011).
Methodologically, we complement the other chapters in this volume by illustrat-
ing the value of computer simulations to understand societal dynamics after revolu-
tions such as those that occurred during the Arab Spring. Computer simulations
allow for understanding how simple rules among individuals cause the emergence
of complex patterns at the system level (Nowak, 2004). Computer simulations,
however, are limited in that they are decontextualized simplifications of real-world
systems. Clearly, our results are preliminary and one should be cautious about their
generality and robustness. Many more parameters can be varied beyond those at
we manipulated in our simulation. For example, we assumed all the agents in the
autocratic control network had identical and low values of the control parameters.
We could examine how variation of the control parameters influences the effective-
ness of control in the hierarchical network. Moreover, in our model we assumed
that, if synchronization is low, individuals are dissatisfied and are likely to favor an
autocratic regime. In our future work, we plan to test this assumption explicitly in
the simulation model. Finally, in our future research, we will vary all the parameters
systematically, and investigate their relative importance and how they interact.
In conclusion, as we have seen in the Arab Spring, even when a populace
can synchronize to cause the collapse of autocratic governments, the transition
to a new and different social order is a complex and uncertain process, which
can result in various scenarios—from new democratic structures, as we saw in
Poland; to the return of autocratic governments, as we saw in Egypt; to outright
civil war, as in Syria. In this chapter, we outlined a theory of autocratic recidivism
and illustrated through initial computational models that successful transitions
are more likely to occur when there is a well-developed meso-level structure and
individuals have some degree of consensus over values—factors that help soci-
eties achieve synchronization after the collapse of autocracy. In contrast, weak
meso-level institutions, which are fostered typically by autocratic regimes, coupled
with high heterogeneity result in low societal synchronization and the risk for
autocratic recidivism or even radicalization. In the long term, with more simula-
tions coupled with real-world data, we will, ideally, be able to predict such sce-
narios with more certainty, with important implications for policy.
Acknowledgment
This work has been supported in part by the Office of Naval Research, grant no.
N000141310054 and by the Polish Science Centre NCN 2011/03/B/HS6/05084.

287
Notes
1. The strength of the link from a leader to a subordinate was set to 1; the strength
of the connection from the subordinate to the leader was set to .015.
2. The strength of the connection between two agents in the horizontal network
was set to .15.
3. The connection strength in both directions was set at .15, the same value as in
the horizontal network.
References
Araújo, D., Silva, P., & Davids, K. (2015). Capturing group tactical behaviors in expert
team players. Rotledge Handbook of Sport Expertise: Routledge.
Arrow, H., McGrath, J. E., & Berdahl, J. L. (2000). Small groups as complex sys-
tems: Formation, coordination, development, and adaptation. London: Sage
Publications.
Baron, R. M., Amazeen, P. G., & Beek, P. J. (1994). Local and global dynamics of
social relations. In R. R. Vallacher & A. Nowak (Eds.). Dynamical systems in social
psychology. Academic Press.
Blanchard, O. J., Froot, K. A., & Sachs, J. D. (Eds.). (2007). The Transition in Eastern
Europe, Volume 2: Restructuring. University of Chicago Press.
Bligh, M. C., Pearce, C. L., & Kohles, J. C. (2006). The importance of self-and shared
leadership in team based knowledge work: A meso-level model of leadership
dynamics. Journal of managerial Psychology, 21(4), 296–318.
Bourdon, S. (2002). The integration of qualitative data analysis software in research
strategies: Resistances and possibilities. Forum Qualitative Sozialforschung/
Forum: Qualitative Social Research, 3(2).
Cappelli, P., & Sherer, P. D. (1991). The missing role of context in OB: The need for a
meso-level approach. Research in Organizational Behavior, 13, 55–110.
Condon, W., & Ogston, W. (1967) A segmentation of behavior. Journal of Psychiatric
Research, 5, 221–235.
Davis, T. J., Brooks, T. R., & Dixon, J. A. (2016). Multi-scale interactions in interper-
sonal coordination. Journal of Sport and Health Science, 5(1), 25–34.
Delaherche, E., Chetouani, M., Mahdhaoui, A., Saint-Georges, C., Viaux, S., & Cohen,
D. (2012). Interpersonal synchrony: A survey of evaluation methods across disci-
plines. Affective Computing, IEEE Transactions, 3(3), 349–365.
Epstein, J. M. (2006). Generative social science: Studies in agent-based computational
modeling. Princeton, NJ: Princeton University Press.
Faist, T. (2010). The crucial meso level. Selected Studies in International Migration and
Immigrant Incorporation. Amsterdam: AUP, 59–90.
Fowler, C. A., Richardson, M. J., Marsh, K. L., & Shockley, K. (2008). Language use,
coordination, and the emergence of cooperative action. In W. A. Fuchs & V. K.
Jirsa (Eds.). Coordination: Neural, behavioral and social dynamics (pp. 261–279).
Berlin: Springer-Verlag.
Freeman, L. C. (1979). Centrality in social networks conceptual clarification. Social
Networks, 1(3), 215–239.

288
Fukuyama, F. (2001). Social capital, civil society and development. Third World
Quarterly, 22(1), 7–20.
Gade, P. M., & Hu, C. K. (2000). Synchronous chaos in coupled map lattices with
small-world interactions. Physical Review E, 62(5), 6409.
Gelfand, M. J., Raver, J. L., Nishii, L., Leslie, L. M., Lun, J., Lim, B. C., et al. (2011).
Differences between tight and loose cultures: A 33-nation study. Science, 332(6033),
1100–1104.
Gilbert, N., & Troitzsch, K. (2005). Simulation for the social scientist. McGraw-Hill
Education.
Girvan, M., & Newman, M. E. (2002). Community structure in social and biological
networks. Proceedings of the National Academy of Sciences, 99(12), 7821–7826.
Gottman, J. M. (2002). The mathematics of marriage: Dynamic nonlinear models.
MIT Press.
Guastello, S. J., & Guastello, D. D. (1998). Origins of coordination and team effective-
ness: A perspective from game theory and nonlinear dynamics. Journal of Applied
Psychology, 83, 423–437.
Harrington, J. R., Boski, P., & Gelfand, M. J. (2015). Culture and national well-
being: Should societies emphasize freedom or constraint? PLoS One, 10(6).
Howard, P. N., Duffy, A., Freelon, D., Hussain, M. M., Mari, W., & Mazaid, M. (2011).
Opening closed regimes: What was the role of social media during the Arab
Spring? Available: SSRN 2595096.
Kaneko, K. (1991). Coupled map lattice. In Chaos, order, and patterns (pp. 237–247).
Springer US.
Kelso, J. S. (1997). Dynamic patterns: The self-organization of brain and behavior.
MIT Press.
Lewis, M. D. (1995). Cognition–emotion feedback and the self-organization of devel-
opmental paths. Human Development, 38(2), 71–102.
Lin, Z., & Li, D. (2003), Does competency-based trust benefit organizational perfor-
mance? A meso exploration using computer modeling. Journal of Mathematical
Sociology, 27(4), 219–261.
Marsh, K. L., Richardson, M. J., & Schmidt, R. C. (2009). Social connection through
joint action and interpersonal coordination. Topics in Cognitive Science, 1(2),
320‒339.
May, R. M. (1976). Simple mathematical models with very complicated dynamics.
Nature, 261(5560), 459–467.
Miller, N. E. (1959). Liberalization of Basic SR Concepts: Extensions to Conflict Behavior,
Motivation, and Social Learning. McGraw-Hill Company.
Newtson, D. (1994) The perception and coupling of behavior waves. In R. R. Vallacher
& A. Nowak (Eds.). Dynamical systems in social psychology (pp. 139–167). San
Diego: Academic Press.
Nowak, A. (2004). Dynamical minimalism: Why less is more in psychology.
Personality and Social Psychology Review, 8(2), 183–193.
Nowak, A., & Lewenstein, M. (1994). Dynamical systems: A tool for social psychology?.
Academic Press.
Nowak, A., Rychwalska, A., & Borkowski, W. (2013). Why simulate? to develop a men-
tal model. Journal of artificial societies and social simulation, 16(3), 12.

289
Nowak, A., & Vallacher, R. R. (1998). Dynamical social psychology (Vol. 647).
Guilford Press.
Nowak, A., Vallacher, R. R., & Borkowski, W. (2000). Modeling the temporal coor-
dination of behavior and internal states. Advances in Complex Systems, 3, 67–86.
Nowak, A., Vallacher, R. R., & Zochowski, M. (2005) The emergence of personal-
ity: Dynamic foundations of individual variation. Developmental Review, 25,
351–385.
Pecora, L. M., & Carroll, T. L. (1990). Synchronization in chaotic systems. Physical
Review Letters, 64(8), 821.
Pecora, L. M., Carroll, T. L., Johnson, G. A., Mar, D. J., & Heagy, J. F. (1997).
Fundamentals of synchronization in chaotic systems: Concepts and applications.
Chaos, 7, 520–543.
Pikovsky, A., Rosenblum, M., & Kurths, J. (2003). Synchronization: A universal concept
in nonlinear sciences (Vol. 12). Cambridge University Press.
Port, R. F., & Van Gelder, T. (1995). Mind as motion: Explorations in the dynamics of
cognition. MIT Press.
Richardson, M. J., Garcia, R. L., Frank, T. D., Gergor, M., & Marsh, K. L. (2012).
Measuring group synchrony: A cluster-phase method for analyzing multivariate
movement time-series. Frontiers in Physiology, 3, 405.
Schmidt, R. C., & Richardson, M. J. (2008). Dynamics of interpersonal coordination.
In W. A. Fuchs & V. K. Jirsa (Eds.). Coordination: Neural, behavioral and social
dynamics (pp. 281–307). Berlin: Springer-Verlag.
Stepanova, E. (2011). The role of information communication technologies in the
“Arab Spring.” Ponars Eurasia, 15, 1–6.
Strogatz, S. (2003). Sync: The emerging science of spontaneous order. Hyperion.
Strogatz, S. H. (2014). Nonlinear dynamics and chaos: With applications to physics, biol-
ogy, chemistry, and engineering. Boulder, CO: Westview Press.
Vallacher, R. R., & Nowak, A. E. (1994). Dynamical systems in social psychology.
Academic Press.
Vallacher, R. R., & Nowak, A. (1997). The emergence of dynamical social psychology.
Psychological Inquiry, 8(2), 73–99.
Vallacher, R. R., Read, S. J., & Nowak, A. (2002). The dynamical perspective in per-
sonality and social psychology. Personality and Social Psychology Review, 6(4),
264–273.
Watts, D. J., & Strogatz, S. H. (1998). Collective dynamics of ”small-world” networks.
Nature, 393(6684), 440–442.

290
APPENDIX Technical Details of the Model
A logistic map is the simplest equation capable of generating complex patterns of

pulsating of a system. The equation in our logical map describes the competition of
two opposite tendencies: “as one thing increases, the other increases” (i.e., the-more-
leads-to-more principle) and “as one thing increases, the other decreases” (i.e., the-more-
leads-to-less principle). In a logistic map the next state of a system (i.e., the next value
of the variable, xn+1) depends on the previous value of the variable, xn, in two opposite
ways. The greater the previous value of x, the greater is the next value of x. At the
same time, the opposite relation holds in that the greater the previous value of x, the
lower the next value of x. The combination of these two opposing tendencies results
in pulsations—the volatility and regularity of which depend on the control parameter,
r, as described earlier.
The logistic map is one of the most important equations describing the dynam-
ics of biological systems. For example, it describes the time evolution of a colony of
bacteria in a closed space (May, 1976). It has been shown that the more bacteria cur-
rently present in a colony, the more bacteria will divide, thus creating new bacteria
(representing the-more-the-more rule). On the other hand, the more bacteria in a
colony, the less the space there is to occupy and the less food (representing the-more-
the-less rule). Indeed, in psychology, there are many examples of when two opposing
tendencies govern human behavior. For example, in approach–avoidance conflicts
(Miller, 1959), the motivation to be closer to the goal competes with the motivation
to avoid the goal.
Think, for example, of an individual on a diet. The most relevant feature of the
behavior is the distance from the refrigerator, which we can represent as the dynam-
ics variable, x. Following the rules of the approach–avoid conflict, as the individ-
ual approaches the refrigerator, the tendency to approach (i.e., to get even closer)
increases (the-more-the-more principle), yet at the same moment the tendency to
avoid increases as well (the-more-the-less principle). In this example, r corresponds
to hunger, with a high r value, the individual overshoots and then corrects, overshoot-
ing again, which results in complex chaotic dynamic. With a lower value of r, the
individual oscillates in less agitated way or just settles at equilibrium.
For those of you interested in the mathematical translation of this theory, these
two rules are expressed in the equation for a logistic map:
xn+1 = rxn(1 – xn). (A.1)
In Eq. (A.1), xn + 1 = rxt describes the-more-the-more rule. The greater the current
value of x, the greater it is in the next moment. The control parameter, r, decides
how much greater the next value of x is in the next moment according to this
part of the equation. One possible interpretation of x is the intensity of behavior,
where 0 means the individual is motionless and 1 denotes the maximal intensity
of behavior. t represents time, counted as discrete time moments (e.g., seconds),
where the symbol t corresponds to the current time moment and t + 1 is the next
time moment. Because x can take values between 0 and 1, xn+1 = r(1 – rxt) describes
the rule “the more the less.” The greater the current value of x, the less it is in the

291
next moment. For example, if the current value of x is maximal (i.e., 1), the next
value is minimal (i.e., 0). The competition of the two opposite rules leads to very
complex dynamics of x, which may be very irregular, chaotic, and unpredictable
over time.
Logistic maps can evolve in time following patterns of different complexity
depending on the value of the control parameter, r (Shuster, 1984). For our agents
to represent complexity of human behavior, we chose to use in our simulations the
values of the control parameter, r, for which logistic maps are known to behave in a
highly complex, seemingly chaotic way (i.e., r > 3.6 and r < 4). For high heterogeneity
conditions, we assumed that r for each agent is a random number from the interval
3.6 to 3.99. To represent a more homogenous society, we restricted the range of r
to be 25% of the diversity of the heterogeneous society and to be a random number
between 3.75 and 3.85. Note that the mean value of the control parameter is the same
in both conditions and is equal to 3.8.

292
293
SECTION IV Basic Methodological Issues

in the Study of Values
294
295
CHAPTER 10 Best Practices

Lessons from a Middle East Survey Research
Program1
Julie de Jong and Linda Young-DeMarco
During the past several decades, the practice of administering multicountry

survey projects to collect comparative data on a wide array of topics has become
ubiquitous. Indeed, organizations throughout the world, including the United
Nations, the World Bank, and the World Health Organization, along with govern-
ments worldwide, have become reliant on the availability of multinational com-
parative data to inform and guide their political and program agendas (Harkness,
Edwards, Hansen, Miller, & Villar, 2010b). One commonly held belief is that if the
research design process is developed with attention to standardization, the result-
ing data will be appropriate for use in both inter-and intracountry comparisons.
Yet, despite a researcher’s best attempts to achieve standardization, the reliability
and validity of comparisons may remain questionable as a result of differences
in cultural sensitivities, infrastructure, census data availability, and survey meth-
ods training. And although each region of the world poses its own unique set of
challenges, for many researchers the Middle East has proved to be a particularly
difficult place in which to achieve cross-national comparative research objectives.
In light of the societal and political instability experienced in the region in recent
years, the Middle East has become home to a public whose voice is crucial to
understanding cultural, political, and ideational phenomena in that part of the
world. This underscores further the need to identify and promote strategies to
help resolve differences that may persist even after standardized survey research
techniques are applied.
In a volume of collected articles published after a 1983 conference on Middle
East scholarly research in Bellagio, Italy, scholars addressed some of the chal-
lenges faced in conducting survey research in the region (Tessler, Palmer, Farah,
& Ibrahim, 1987). Included were such within-country issues as the difficulty devel-
oping sample frames, the lack of adherence to quality control procedures, the dif-
ficulty of obtaining necessary permits from study country officials, and the issues
in translating Western-centric concepts into Arabic and other regional vernaculars
(Tessler et al., 1987). Unfortunately a quarter century later, contemporary schol-
ars working in the region continue to struggle with many of these same issues.
296
And, if a researcher is desirous of cross-national comparative data collection, the

challenge is only multiplied. Recognizing that achievement of the ideal may not
be fully possible, what steps can researchers today take to produce high-quality
survey data comparable across multiple countries?
In this chapter, we provide an overview of which protocols are most critical
to the design and administration of multicountry survey data collection in the
Middle East when the research goals call for cross-national comparative analyses.
Because we are constrained by chapter length limitations, it is not possible for our
discussion to be truly comprehensive; we do attempt, however, to touch on those
areas we believe to be most important relative to cross-national survey research in
the Middle East. We also make the assumption that readers have a rudimentary
understanding of survey research principles and thus do not discuss basic tenets
of the discipline.2 First, we examine the specific challenges faced by researchers
in the region. Next, we discuss the importance of establishing a cross-national
research team. We then turn to a discussion of sampling procedures, drawing
on experiences of other cross-national comparative survey researchers to identify
design aspects most crucial to the standardization process. Next, we discuss ques-
tionnaire development and its inherent challenges brought on by both regional
and cultural differences, and then we discuss standardization, particularly across
Arabic-speaking countries. Last, we discuss other survey research protocols,
including pretests, interviewer training, and data entry.
The research strategies discussed in this chapter are based on both theoreti-
cal methods and our own practical experience in the region. The project from
which much of our experience is derived was motivated by the desire to iden-
tify the social, cultural, and perceptual factors associated with certain worldviews
or streams of thought. The project focused on Egypt, Iraq, Lebanon, Pakistan,
Saudi Arabia, and Turkey (a) to inspect the extent of the penetration of Western
values in Islamic countries and to assess how this penetration is linked to the
nature of these countries’ relationships with the West; (b) to explain cross-national
variations in values; (c) to construct attitudinal conceptions of religious funda-
mentalism, developmental idealism, and morality, and explain variations in such
attitudes cross-nationally; and (d) to construct a series of indicators useful in pre-
dicting trends in values among the citizens of these countries.
Each collaborating country had been a study site for one or more of the fol-
lowing: (a) World Values Survey (WVS), (b) surveys on youth in Egypt and Saudi
Arabia, and/or (c) developmental idealism studies. Although these past studies
collected roughly equivalent questionnaire data across research sites, other pro-
tocols such as respondent sampling, questionnaire design and layout, transla-
tion, interviewer training, project administration, and data entry procedures were
implemented at the discretion of those contracted to collect the data. For the cur-
rent project, standardization of as many protocols as possible was essential to
achieve the ambitious goals set forth in our research agenda. We were also aware
that strict adherence to our ideal standardization specifications would not always
be possible given the differences among geographies, cultures, and norms—not
only between our Western and our collaborators’ Middle Eastern viewpoints, but
296 | Basic Methodological Issues in the Study of Values

297
also among the study countries themselves. Nevertheless, we identified a num-

ber of methodological areas where we believe the application of documented
best practices enhance overall data quality and lead to increased comparability
across study sites. Occasionally, these best practices were not entirely suited to our
particular situation and needed some adjustment to satisfy our objectives; these
adaptations are discussed as well. Although each challenge is described separately,
anyone undertaking such a project must keep in mind that, in practice, many of
these areas are best addressed simultaneously.
Challenges of Conducting Research in the Middle East

In general, survey research in most of the Middle East is still in its infancy, with
business-initiated market research the only source of empirically collected opin-
ion data until the commencement, around the year 2000, of the WVS, the Arab
Barometer, and Gallup International. This paucity is a result of the lack of avail-
able infrastructure and trained professionals, as well as challenges inherent to
conducting research under repressive political conditions (Tessler 2011). And,
certainly, these two issues are entwined historically. For example, the Iraqi col-
laborator in our cross-national project was among the very first in his country to
establish a survey research firm after the fall of dictator Sadaam Hussein in 2003.
Until then, there had been no surveys firms in Iraq. In other countries in the
region, survey activities have, historically, been severely curtailed and focused on
a narrow band of topics. For example, until Tunisia’s dictator Zine El-Abidine Ben
Ali was deposed in 2011, surveys that included any questions about politics were
not permitted in Tunisia. Survey items asking explicitly about democracy, among
other issues, may still not be included in questionnaires administered in Saudi
Arabia. Indeed, in a post-9/11 survey of political scientists conducting research in
the Middle East, authoritarian political conditions were cited as the primary chal-
lenge to working in the region (Clark 2006).
The lack of infrastructure and support by regional governments also impacts
a host of factors related to survey research. Because few survey research centers
are established in the region, interviewing capacity is generally limited to paper-
and-pencil interviewing rather than the ideal: computer-assisted personal inter-
views. This limitation both affects the ability to administer complex interviews
and, potentially, diminishes data quality (de Leeuw, Hox, & Snijkers, 1998). Lack
of infrastructure and unstable political conditions may also affect the ability to
identify a sampling universe—a concern discussed in detail later in this chapter.
In addition to the difficulties of conducting research in an authoritarian regime,
several countries in the Middle East are also currently armed conflict zones, add-
ing an additional layer of complexity to the research process. Mneimneh, Axinn,
Ghimire, Cibelli, and Salih Alkaisy (2014) provide guidance to researchers work-
ing in areas of conflict, focusing particularly on interviewing techniques designed
to enhance both the security of the interviewer and the respondents. Maintaining
security, confidentiality, and protection of human subjects is essential, and it is
Lessons From A Middle East Survey Research Program | 297

298
imperative that researchers have a full understanding of current conditions. For

example, in our own cross-national project, Syria was included initially as a study
country, but the outbreak of unrest and, eventually, civil war in the country caused
us to drop Syria from the protocol. As conditions in Syria deteriorated in 2011
following the Arab Spring, it was determined that the volatile political climate
made administration of the survey questionnaire too unsafe for interviewers and,
moreover, that the reliability of the data would be in question because of the cir-
cumstances. Therefore, as a result of the ever-changing nature of conditions in
the region at large, researchers must demonstrate flexibility, and recognize that
some adjustments to the overall protocol may be necessary. Regarding conducting
research in conflict zones in the Middle East, Romano (2006) wrote: “Particularly
in unpredictable conflict zones, the resource I needed the most was time. You
should multiply the amount of time you think a research project will take you in
a conflict zone by at least three or four” (p. 441). Our experience is indicative of
this sentiment, when not only data collection itself, but also postdata collection
processes have been delayed because of instability in the region.
Establishing a Cross-National Research Team

Harkness et al. (2010b) cite the lack of readily available direction and lessons from
past experience as one of the greatest challenges confronting scholars who under-
take cross-national comparative research. This dearth of knowledge has the poten-
tial to affect nearly every aspect of the project negatively. The authors suggest the
formation of teams that specialize in certain areas, such as questionnaire design,
project management, cultural norms, language translation, and so on, to mini-
mize the loss of standardization or at least lessen the negative impact of it. Their
model calls for teams that can expand or contract easily, depending on the needs
of the core group at a given stage of the research.
Our cross-national six-country survey in the Middle East followed sug-
gested best practices in this area but for one; we deliberately avoided creating
a team that would expand and contract according to the needs of the project.
Instead, we sought input of collaborators from each study country throughout
the entire process and, by doing so, fostered a spirit of cooperation among
representatives of all the participating countries. This approach was impor-
tant because, historically, Middle Easterners have viewed the motivations of
foreign researchers with suspicion (Harik, 1987). Therefore, we went to great
lengths to demonstrate that our research was transparent, that the data would
be made publicly available, and that knowledge gained from the surveys would
benefit the study countries as well. The data collection collaborators were a
unique mix of participants from a wide range of organizations and included
university scholars, for-profit market research company managers, individuals
from academic-led for-profit survey research companies, and consultants from
nonprofit program facilitation agencies. Team members based in the United
States were led by a political sociologist and a family sociologist/social demog-
rapher, and included a cognitive psychologist, two research staff members

299
with expertise in cross-national survey design and administration, and a visit-

ing predoctoral Fulbright scholar from Turkey. The US-based political sociolo-
gist and family sociologist initiated the collaborative study because of a shared
interest in the relationships among religious fundamentalism, developmental
idealism, morality, and the penetration of Western values. The cognitive psy-
chologist was invited to collaborate because of his expertise in psychometric
scales, particularly those focusing on religion. The research staff members
were instrumental in the methodological and management components of the
project, and the Fulbright scholar was crucial to the evaluation of survey items
using her own cultural perspective.
The US-based team began by building a working relationship with our
research site colleagues well before fieldwork activities were scheduled to
begin. We believed their cooperation to be imperative if we were to create suc-
cessful protocols that were culturally sensitive while simultaneously satisfy-
ing our research objectives. The collaboration began by actively seeking input
from our study site collaborators during item construct development and
initial survey measure draft phases. After the US-based members developed
an initial draft of the survey instrument, a 3-day workshop was conducted in
Cairo with the entire group in order (a) to provide the opportunity for every-
one to get to know one another, (b) to evaluate the draft questionnaire topic by
topic and question by question, (c) to provide input collectively for reworking
or rewording sections and questions, (d) to discuss problems and possible
solutions to ensure overall standardization of the project, (e) to seek culturally
specific advice from individual study site team members, and, perhaps most
important, (f ) to demonstrate clearly that ours was a unified research project
rather than six unrelated data collection efforts. Second and third workshops
were held in Cairo and Istanbul; the purpose of the second workshop was to
discuss pretest results and possible alterations/improvements to the produc-
tion questionnaire, and the third reunited the entire group along with field
supervisors from their respective research sites for study specific interviewer
training.
By far, the decision to build a collaborative working relationship with our
data collection contractors throughout the entire design process was one of the
best decisions we made for our project. Not only did the relationship facilitate
the realization of project objectives, but also the process of working collabora-
tively toward a common goal allowed the entire group to become invested in
the research. By implementing procedural steps slowly and by asking for input
at each stage of project development, we were able to alleviate any fears study
site collaborators might have had concerning our motives, consequently develop-
ing a relationship built on trust. Moreover, the time spent together brainstorm-
ing, problem solving, and getting to know one another after the workday gave
everyone a sense of project ownership. And, relationships formed during the
workshops facilitated the data processing stage and provided the opportunity for
continued collaboration in analyses. Participants left the workshops with a uni-
fied sense of purpose and the realization that all involved must work in tandem
to achieve the project’s success.

300
Sample Design
As multinational surveys have become more widespread in recent years, there
has been an associated increase in the literature regarding the particular chal-
lenges of comparable sample designs across study countries (Heeringa et al.,
2008; Heeringa & O’Muircheartaigh, 2010; Lynn, 2003; Lynn, Häder, Gabler, &
Laaksonen, 2007; Verma & O’Muircheartaigh, 1980). To compare populations in
different countries accurately during the analysis stage, one must start with the
premise that survey respondents come from like populations, with “survey esti-
mates … unbiased (or nearly so)” and with “the precision of sample estimates …
sufficiently high to permit useful probability statements concerning the true
population value that the survey sample is designed to estimate” (Heeringa
et al., 2008, p. 15). And, comparable samples from each study country in the
project must ensure “that every member of the relevant population in the coun-
try in question has a known, non-zero probability of selection” (Häder & Gabler,
2003, p. 127). In our cross-national research, we have used a combination of
protocols from the World Mental Health surveys and European Social Survey
(ESS), but have adapted portions of these protocols specifically for research in
the unique Middle Eastern context.
Before turning to the specifics of our protocol, however, we first discuss why
sample design in the Middle Eastern context is unique and presents a particular
challenge. As noted earlier in this chapter, the Middle East survey research cul-
ture is in its infancy and has been hampered historically by political, economic,
and cultural forces. Restrictions have also hindered the development of stringent
survey sampling protocols and innovation in sample design—in a region where
innovation is particularly necessary.
In some Western countries, such as the Netherlands, regional government
authorities compile and update routinely lists of their citizens, providing the
potential to develop a sampling frame in which elements of the population have
a known, estimable probability of selection (Bijl, Van Zessen, Ravelli, De Rijk,
& Langendoen, 1998). In other countries, such as the United States, routinely
updated census data combined with conscientious block listing provides a usable
sample frame. However, in much of the Middle East, researchers must rely on out-
of-date census data at every stage of the sample design. For example, in Iraq, the
most recent national census was conducted in 1997 and it excluded three Kurdish
autonomous governorates, which in 1987, the last time a complete census of Iraq
occurred, compromised about 12% of the total population (United Nations, 2003).
The decade between 1987 and 1997 witnessed portions of the Iran–Iraq War and
the Gulf War, leading to potential unreliability of using 1987 estimates to impute
census data in 1997. And, of course, the 14 years between 1997, the most recent
partial census, and our most recent survey in 2011 witnessed the fall of Saddam
Hussein, the invasion of US forces, and the displacement of millions of Iraqis,
rendering even more obsolete already incomplete census data from 1997. Plans to
administer a census most recently in Iraq have been delayed because of concerns
of stoking further sectarian violence (Myers, 2010).

301
And, although national census population estimates for sample design use are
poor in Iraq, they are, unfortunately, nonexistent in Lebanon—a country that last
completed a national census in 1932 as a result of the desire of the government
not to inflame political sensitivities (Abdulrahim, Ajrouch, Jammal, & Antonucci,
2012). Nor does the government maintain any other sort of population registry—
meaning, population distribution data are not readily available to allocate a
national sample using area probability methods. It is in this milieu that survey
researchers must develop a sample design that leads to a nationally representative
sample.
Until very recently, one of the only cross-national survey research projects
underway in the Middle East was conducted under the auspices of the WVS,
which has now collected data from 97 societies, with some countries experiencing
multiple waves of data collections. Inglehart, Basanez, Diez-Medrano, Halman,
and Luijkx (2004) discuss the limitations faced by the WVS team in carrying out
surveys in non-Western settings—particularly in countries where there have been
few, if any, nationally representative data collections, and where interviews must
be carried out face-to-face because of limited telephone coverage. These authors
argue that, although full probability samples are the gold standard, when doing
face-to-face interviews, repeated callbacks to contact sampled respondents are
often necessary, and this additional cost is frequently prohibitive. Therefore, for
a number of study countries, researchers used a probability model down to the
household level, tapping in to whatever demographic data were available for popu-
lation estimates, and then used quota sampling. In addition, because of the budget
constraints of face-to-face interviewing, WVSs have often underrepresented illiter-
ate, rural respondents (Heath, Fisher, & Smith, 2005). Methodological documen-
tation has been inconsistent and has made it very difficult or even impossible to
calculate response rates accurately using such well-regarded methods as detailed
by the American Association for Public Opinion Research (2008). And, a number
of WVS study countries permitted the replacement of sampled respondents in
cases of refusals or noncontacts, according to publicly available methodological
documentation (World Values Survey, 2015).
Such practices may inflate response rates artificially, do not meet the require-
ments for probability sampling, and, consequently, do not create comparable data
sets cross countries (Groves 2006). In instances such as these, respondents do
not have an equal, nonzero probability of selection, and any sampling error attrib-
uted to sample design is impossible to calculate. Because of the multifaceted topic
of our cross-national data collection effort and the necessity to obtain attitudinal
data from all segments of a country’s population, it was crucial—at the imple-
mentation of our project—we improve on previous data collection efforts in the
region. In our new project, our intention was to overcome previous limitations by
requiring strict criteria to be met in nearly all aspects of the sample design and
associated fieldwork.
Heeringa and O’Muircheartaigh (2010) argue that when planning multina-
tional data collection, absolute comparability at all stages of the sampling process
is not a requirement for ensuring ultimate cross-national comparability at the data

302
analysis stage. Even when multicountry projects are organized and funded cen-
trally, circumstances at the local level dictate that flexibility in at least some aspects
of sample design is necessary. In discussing cross-national survey practices, Lynn
et al. (2007) suggest, rather, that at least two criteria be met: (a) study populations
in all countries must be equivalent and (b) “sample-based estimates must have
known and appropriate precision in each nation” (p. 108)—meaning, the sample
design is transparent, with sample detail available at each selection stage, and
that, hopefully, a minimum and comparable precision requirement is met in each
study country. To meet these criteria, the population sample and expected cover-
age must be defined clearly and nonresponse rates must be similarly low.
The ESS is one example of a large-scale cross-national comparative study that
implemented guidelines that all study countries were obligated to follow; much of
their protocol guided our project in the Middle East. First, we defined the target
population clearly to ensure all of its members had an equal, nonzero probability
of selection as well as equivalency in demographic characteristics. In the context
of the Middle East, we deemed age, citizenship, and certain residency require-
ments as critical to all study countries.
We defined the target population in each study country as all citizens age
18 years and older to avoid the burden of the additional human subjects require-
ments inherent in sampling and interviewing minors. To aid in analytical com-
parability, the target population was also limited to include citizens of each study
country only. This action was necessary to maintain cross-national sample com-
parability because some of our study countries exclude migrant worker popu-
lations routinely from survey sample eligibility whereas others may not. Many
countries throughout the Middle East have an ever-present anxiety toward the
growing proportion of nonnationals in the region. Believing that migrants pose a
threat both economically and culturally to the national populations, some coun-
tries have implemented a number of restrictions throughout the years that serve
to reinforce the transient nature of nonnationals, including limits on duration of
stay, the practice of sequestering migrants in labor camps for their duration of
stay, strict limitations on naturalization, and limited citizenship rights for those
who do manage to become naturalized (Kapizewski, 2006; Williams, Thornton,
& Young-DeMarco, 2013).
We also excluded from our target population those members of study country
populations living in institutional settings, such as prisons, nursing homes, mili-
tary bases, student dormitories, and other group homes. Obtaining government
permission to do survey research in Middle Eastern countries can be challeng-
ing, particularly because of the sensitive nature of some of our survey questions.
Requesting access to many of these institutionalized populations in our study
countries would likely meet with negligible success and may have made collabora-
tion with our study countries even more taxing.
Finally, security concerns and budget restraints required that, in some coun-
tries, certain geographic areas be eliminated from the sample design, with all
exclusions being documented carefully. For example, in Egypt, the governorates
in the most remote desert regions were not included in the sample population for

303
budgetary reasons, resulting in the exclusion of about 1.8% of the total popula-
tion (Red Sea, New Valley, and Matrouh). And in Saudi Arabia, very rural areas
and military zones were eliminated because of both fiscal constraints and security
concerns.
In addition to a shared definition of target population, sample designs in our
study countries met three additional criteria. First, each country’s sample design
consisted of a multistage area probability sample of households, which provided
estimable probability to all members of the target population but allowed for cost
control through stratification and clustering of sampled households into primary
sampling units (PSUs). In all countries, the probability sample was designed using
population distributions from the most recent census or, in the case of Lebanon,
the most recent national survey, with first-stage, second-stage, and, if necessary,
third-stage sampling phases drafted and shared with us for approval. Our experi-
ence in the region has demonstrated that when discussing sample design with our
study country collaborators, sampling terminology is not understood consistently
across cultures and particular care must be taken to ensure everyone involved
understands the principles behind particular techniques in the sample design. For
example, the sampling phrase probability proportionate to size indicates a design in
which “each cluster is given a chance of selection proportionate to its size. Thus,
a city block with 200 households has twice the chance of selection as one with
only 100 households. Within each cluster, however, a fixed number of elements is
selected … [resulting] in each household having the same probability of selection
overall” (Babbie, 2013, p. 238). However, in our project, one study country collabo-
rator proposed a sample design using probability proportionate to size, in which he
intended for the phrase to indicate only that the number of clusters chosen from
each province would be proportional to the country’s population, but without an
intention of selecting the clusters themselves in accordance to their size. In this
case, the misunderstanding was resolved through further discussion and agree-
ment about terminology. However, it is crucial to be aware of how differences in
terminology, especially when exacerbated by language barriers, have the potential
to lead to incomparability of sample designs.
The manner in which households were selected from PSUs differed across
countries as a result of the availability of household data. Team members in Egypt
and Pakistan were able to draw a sample of household addresses from available
census data or other government registries and assign the addresses to coversheets
before releasing interviewers to the field. In Turkey, area probability sampling was
used to stratify the sample by governorates, and then voter registration lists were
used to select households. The other study countries used area probability sam-
pling before data collection began in the field only to the level of the PSU, and
then used random route sampling in the field to identify eligible households in
each sampled PSU (Lynn et al., 2007). Although a random route sample design is
not generally used in those projects with sampling designs we strove to emulate—
indeed, only one of the ESS’s 22 participant countries was permitted to use random
route sampling—it is more straightforward and budget friendly than conventional
field block listing before data collection, which was the other viable option available

304
in our study countries. And, in this region of the world, it may prove to be an advan-
tageous sampling strategy when compared with drawing households from preex-
isting lists, for it prevents the possibility of missing those dwellings that are not
present or available on registries or other listings. Indeed, collaborators in Pakistan
used data that are more than a decade out of date to compile a sample frame of
household addresses. Last, although field block listing is the gold standard in lieu
of the availability of government registries of household addresses, the method has
not been widely used in the Middle East as of now, with the notable exception of
Lebanon (Karam et al., 2008). Certainly, this is a challenge of survey research in
the region, and future research efforts should focus, at least in part, in training field
staff to carry out block listings to obtain higher quality survey samples.
The second criterion common to all countries pertains to the selection of the
respondent at the sampled household. Best practice in the survey research field
would dictate the use of the Kish table when selecting respondents, and, indeed, in
Egypt, Pakistan, and Saudi Arabia, respondents were selected using this method
(Groves et al., 2009). However, our overseas collaborators in Iraq, Lebanon, and
Turkey were resistant to the use of the Kish table because of interviewer inexpe-
rience and concerns about the burden on the household contact in completing
the Kish table. They instead requested that interviewers be permitted to use the
“next birthday” method, in which interviewers list all eligible household mem-
bers and their birthdays, and then select as the interviewee that person whose
birthday occurs next. Comparative research in the two methods in the United
States shows very little difference in response rates, but also very little difference
in demographic distribution of selected respondents between the two methods—
meaning, the choice of method likely has little impact on the overall sample dis-
tribution (Oldendick, Bishop, Sorenson, & Tuchfarber, 1988). However, no similar
experiments have been completed in the Middle East, and differences in methods
are unknown. Our experience has given us new insights into differing levels of
resistance to various methods across study countries, such as the Kish table, and
although in the current study the criteria for selection of respondents at the house-
hold level was the use of either the Kish table or the next-birthday method, in the
future more attention needs to be given to questions of comparability between the
two methods.
The final criterion common to all countries again pertains to the selection of
the respondent at the household level. As noted earlier in this chapter, respondent
and/or household replacement has occurred in previous data collection efforts
in Middle Eastern countries, which renders impossible the calculation of sam-
pling statistics via American Association for Public Opinion Research guidelines.
In addition, there is some evidence that permitting replacement sampling in the
field leads to decreased interviewer motivation with regard to obtaining an inter-
view at the originally sampled household (Chapman, 1983; Elliot, 1993). For both
of these reasons, replacement at the household or individual respondent level was
not permitted in any of our study countries under any circumstances.
In addition to specifying these aspects of sample design, we required that each
participating country provide written sample documentation according to our

305
specifications at the end of the field period, with countries obligated to provide
indicators of sampling stratum and/or clusters at each sampling stage as well as
probability of selection at each stage, allowing for calculation of response rates.
Construct Development
For any theoretically driven survey research project, beginning with a clear con-
ceptual framework and definitions of underlying constructs for effective measure-
ment is essential to the creation of common indicators in the survey questionnaire.
When the project requires cross-national surveys to fulfill specific aims of the
research, such clarity takes on particular importance, especially if common indi-
cators are to be used across all survey questionnaires. Determinations of whether
a construct can be measured comparably across settings must be made early in
the project. Therefore, it is in the best interest of the researcher to provide writ-
ten descriptions of the key concepts and underlying constructs for review by the
team members from each participating survey site. Failure to define comparable
concepts and constructs adequately may blur the effect of the social context in
creating the observed differences, because it would not be clear whether such dif-
ferences are contextual effects or whether they reflect differences in the meanings
of the concepts or the manner in which they are operationalized across countries
(Harkness et al., 2010a). Moreover, the concepts researchers want to measure
in the Middle Eastern context have often been imported from the West, rather
than originating from indigenous roots; many past studies of Arab populations
have implemented measurement of Western models without first determining
whether the constructs are relevant in the context of Arab society (Ibrahim, 1987;
Zurayk, 1987).
Our challenge in this project was to measure constructs that tap into social
support for indigenous movements and Western models. Adding to this challenge
was the desire to design comparable constructs for a wide range of respondents
living across the six Middle Eastern study countries. These constructs included
such diverse questions as those related to the veil, polygamy, and religious fun-
damentalism on the one hand, and those that assess adherence to developmental
idealism—a Western evolutionary perspective on social change but hypothesized
to have been disseminated worldwide—on the other. The challenge of measuring
religious fundamentalism was to formulate a series of items that were not rooted
necessarily in one particular religion, but indicate particular orientations toward
whatever religious belief one has. In this manner, similar standard items can be
formulated that measure fundamentalism among both Muslims and Christians.
We reasoned that fundamentalists have similar orientations toward the deity,
their scripture, their own religious community, and the faith of others. Given this
reasoning, we identified four components of religious fundamentalism: (a) the
nature of deity, (b) inerrancy and literalism, (c) exclusivity and religious centrism,
and (d) intolerance/tolerance (Moaddel & Karabenick, 2008, 2013). We also faced
similar challenges in measuring the components of the package of ideas that
Thornton (2001, 2005) describes as developmental idealism. These components

306
were identified as (a) modern society is good and attainable, (b) modern family is

good and attainable, (c) modern political systems are good and attainable, (d) a
modern family is a cause as well as an effect of a modern society, and (e) individu-
als have the right to be free and equal with social relationships based on consent.
The problem, however, was that, in the West and for Westerners, development and
westernization are synonymous. But, for people from cultural traditions outside
the West, development was one thing and westernization quite another. We thus
needed to separate these two components to assess the kind of attributes people
attach to development and the kind they attach to westernization.
The challenge of finding an indicator on veil preference was no less telling. We
were not able to ask a single cross-national comparable question about women’s
dress and head-covering preference because of the variation is dress-style prefer-
ence across the six countries. Therefore, we decided to use pictures of women in
different styles of dress and head coverings to ensure all respondents were exposed
to the same stimuli, and we then asked respondents to identify which woman was
dressed most appropriately. Other key constructs identified were religious, family,
and economic values; perceptions of well-being and trust; attitudes toward violence;
perceptions of westernization versus development; international development and
morality hierarchies; sources of information; and epistemic authority.
Cross-national research calls for comparable measurement of common con-
structs across research sites. We went a step further and stressed the importance
of determining whether constructs are understood comparably among project ini-
tiators before attempting to assess their commonality across study sites. We imple-
mented this strategy during our cross-national values project; it removed any
ambiguities between the primary team before attempting to convey the constructs
to our research site collaborators. As constructs were defined and components
were identified and described in written form, they were circulated among the US
group, discussed, refined, and revised as needed. When constructs were defined
completely and agreed on internally, the descriptions were circulated to our study
site collaborator teams for feedback, clarification, and improvement. Back-and-
forth iterations continued until group members felt confident the constructs were
understood comparably by all.
Creating Indicators
After the existence of common constructs between target countries is confirmed,
researchers must take steps to ensure the components are measured comparably
across data collection sites. It is important to take advantage of the multicultural
resources available so that the resulting instrument will be appropriate for use
in all study countries (Harkness, Van de Vijver, & Johnson, 2003). Seeking input
from collaborators to help identify which target language words and phrases will
produce comparable native language translations effectively is critical as well.
Building time and monetary resources into the project budget for cross-national
collaborators to discuss the resulting draft questionnaire together as a group is
another key element of the process.

307
In cross-national research there are a number design approaches to use,

singularly or in combination. One approach, asking identical questions across
sites, begins by developing a source questionnaire in one language and later
translating it into the languages spoken in the target areas. This strategy is
most often applied when researchers want to reuse existing questions from
past surveys, although protocols containing newly written questions may follow
this approach as well. A variation to the standard asking-identical-questions-
across-sites approach is called decentering and it involves developing questions
in the source language, translating them into another language, and then
iterating back and forth until the questions have the same meaning in both
languages. In the context of Middle East cross-national research, decenter-
ing works quite well; it is particularly helpful in resolving within-and across-
country differences in question interpretation caused by variations in Arabic
dialect and in other vernaculars. Best practices suggest decentering techniques
are most effective when applied to questionnaires administered in a maximum
of two languages (Harkness et al., 2010b, 2016), yet at least one cross-national
project has been successful using decentering to create questionnaires for nine
languages (Thornton et al., 2010). The ask-a-different-question method identi-
fies or creates appropriate questions for the source questionnaire, and then
adapts them as needed to fit the remaining target countries. In practice, most
cross-national comparative studies use a combination of approaches to pro-
duce project questionnaires; collectively this method is known as parallel design
(Harkness et al., 2003, 2010b, 2016).
There are several approaches to measurement selection; for example, there
are existing questions from surveys that can be used “as is,” there are existing
questions that call for some type of refinement before fielding, and there is the
option to compose entirely new questions. When evaluating existing measures, it
is best not to assume a question is acceptable simply because it has been used in
prior surveys. Sometimes existing questions are unclear or worded poorly; more-
over some poorly designed measures appear in even well-respected long-running
surveys.
Our project serves to illustrate one way successful selection of indicators may
be achieved. We chose to use all three tactics for selecting indicators to measure
the construct of religious fundamentalism. The group agreed during the con-
struct development phase that fundamentalism consisted of the four factors men-
tioned earlier in this section. Questions designed to measure levels of religious
fundamentalism had been collected in past values surveys. Adjustments to those
questions were required if they were to be considered for inclusion in our study
because many were double-barreled, had unclear components, or measured over-
lapping components within the same question. New questions were developed
for missing or incomplete components within the construct. We also took care to
develop a mixture of questions worded positively and negatively to guard against
acquiescence bias. As we discovered during the pretest, religion is a highly contro-
versial topic for many respondents, and social desirability bias and/or tendencies
to answer at extreme ends of the scale became issues that were more difficult
to address. To minimize this bias before a lengthy set of items measuring the

308
different components of religious fundamentalism was administered to respon-

dents, we included the following sentences to explain the international flavor of
this research and the scientific intent of the questions, which were read immedi-
ately before asking the set of items:
The next set of questions is about religion. Please be assured they are in no way
intended to reveal doubts about anyone’s faith in their religion. These ques-
tions are being asked to people of different religions all over the world, includ-
ing Muslims, Christians, Hindus, Buddhists, and so on. We compare people’s
responses in different countries in order to understand the similarities and dif-
ferences in their religious faiths. We appreciate everyone’s generous contribu-
tions to our research.
By providing context to the series of questions, we intended to decrease bias by
affording respondents the freedom to express their own opinions rather than
those dictated by social norms. We also hoped to give respondents a greater sense
of importance and purpose in the research process, thereby leading to increased
attention and thoughtful response to the survey items. This introduction was also
beneficial to the interviewers because it provided them with a rationale to present
to the respondents about why such a question was being asked.
Other revisions to the questionnaire included some wording adjustments and,
on a country-specific basis, the exclusion of certain items from the production
questionnaire. We were also advised by our collaborators to reorder some of the
items within the questionnaire, placing the more sensitive items toward the end.
A fair amount of preparatory work had already occurred before the start of
our project to measure the concept of developmental idealism on a cross-national
level in Argentina, China, Egypt, Nepal, Saudi Arabia, Vietnam, Taiwan, and the
United States (Thornton et al., 2010). Although budget and methodological con-
straints limited the ability to achieve strict comparability relative to sampling and
interviewer administration, basic comparisons of these data were possible. These
comparisons convinced us there was considerable international overlap in terms
of how ordinary people defined development and developmental hierarchies rela-
tive to United Nation ratings (Thornton et al., 2010), and that ordinary people
believed causal relationships among development, morality, family change, and
societal change existed as well.
With that in mind, our current group selected some existing measures that
had worked particularly well, and then broadened and expanded our components
to include relationships among development and religion, politics, and western-
ization. We also embedded cognitive interviewing techniques into the pretest
interview to assist us with learning how respondents defined particular concepts
that had not yet been studied systematically on a cross-national level, but were
important aspects of our research agenda. For example, following a list of coun-
tries for which respondents were asked to rate the “morality” of each country on
a scale from 1 to 10 points, we included the cognitive probe: As you were rating
countries using this morality scale, what kinds of things were you including in
your definition of “morality?” It was important to verify whether respondents were
thinking similarly relative to the meaning of morality as we were asking them

309
to rate morality levels of people in various countries around the world (Beatty &
Willis, 2007; Bradburn, Sudman, & Wansink, 2004; Willis, 2005). Results from
the probes in the pretest indicated that respondents across countries had a shared
understanding of the term morality as it was translated, and increased our confi-
dence of the term’s veracity in the survey. Scholars are well advised to use such
techniques to help establish parameters for topics that have not been studied sys-
tematically in prior research.
Subject matter sensitivities are a bit more challenging because they vary from
country to country. For example, most respondents in Lebanon can be asked
almost anything, whereas in Saudi Arabia, many topics—such as those pertaining
to democracy, for example—are off limits. We were surprised during the pretest
debriefing by which topics were labeled sensitive by our study countries. We were
not expecting our questions concerning religious beliefs to be extremely contro-
versial, but respondent reaction to them during the pretest told us otherwise. For
example, one problematic survey question was “Do you believe in God?” with
response categories “yes” and “no.” Interviewers in Egypt and Saudi Arabia, in
particular, reported that during the pretest, respondents were very upset by these
questions, responding with such phrases as, “Of course I believe in God and it is
insulting to ask otherwise” and “I am Egyptian and Egyptians believe in God; it is
blasphemous to suggest that we do not believe in God, simply by asking the ques-
tion.” To minimize these types of reactions, interviewers read the following phrase
immediately before asking the question: “Please keep in mind that we ask the next
set of questions because we will compare the results from [Study Country] with
many other countries.” This led to decreased negative reactions by respondents,
presumably because, by providing a context for the question, respondents felt less
defensive. This introduction was also beneficial to the interviewers because it pro-
vided them with a rationale to present to the respondents a reason for why such
a question was being asked. On the hand, we expected our series of questions
gauging respondent level of approval toward attacks on Americans and other for-
eigners to be quite controversial; to our astonishment, these questions were not
at all problematic for our respondents, although they felt we should have added
questions concerning attacks on their specific countrymen as well.
Language Translation
One of the most difficult and challenging aspects of questionnaire develop-
ment is the accurate translation of concepts and dimensions from the source
language questionnaire to the languages of the individual study countries.
Literal translations of questions may prove to be insufficient because often they
do not convey the intended meaning or idea accurately. Idiomatic translations
of the material may not produce comparable results either. In cross-national
research, the potential for translation inaccuracies is compounded by the num-
ber of separate translations required. Regional differences within and across
countries must be addressed as well. Although there is agreement among both
survey researchers and translation science scholars that the need for a better

310
melding of the two disciplines exists, so far that melding is in its early stages.
Best practices, however, suggest that having an organized team approach for
language translation helps to navigate through this complicated exercise, and
recent work suggests that language translation considerations be built into the
design process itself, rather than waiting until the survey questionnaire has
been finalized in the source language (Harkness, 2003; Harkness, Villar, &
Edwards, 2010c).
The difficulty in producing translations of questionnaires across the Middle
East is great given the numerous languages spoken in addition to Arabic, plus
regional dialectic differences found among Arab-speaking populations. Some
cross-national surveys in the Middle East have required each study country to pro-
vide an Arabic translation of the English questionnaire, and have subsequently
used that translation for production interviewing. This protocol results in a
unique Arabic translation from each country, leading to decreased possibilities for
cross-national comparison. Moreover, implementation of the protocol implies that
regional differences in Arabic dialect will not be addressed fully. Consequently,
during cross-national as well as single-country data analyses, it may be difficult to
separate true differences of opinion from differences generated by inadequate lan-
guage translation. Obviously, this generates serious concern for any survey work
planned in the region. Because the issue of language and meaning is so complex
in the Middle East, it is imperative for preparation to begin as early as possible—
for example, when exploring whether the constructs one wants to measure actu-
ally exist across countries.
Again, we turn to our experiences in the Middle East to demonstrate one prac-
tical approach for translation of the survey questionnaire from the source lan-
guage to the target languages. In our cross-national research project, up to seven
separate translations were required: Arabic, Kurdish, Turkish, Punjabi, Pushto,
Sindhi, and Urdu/Hindi. Because many of the measures we planned to ask in our
survey had been administered during past surveys, if we replicated the practice
of one questionnaire translation per Arabic-speaking country we would have pro-
duced four Arabic interpretations of the same questionnaire. The complexities of
questionnaire translation were discussed at length with our collaborators during
the first workshop.
Because we knew we would also be introducing some very complex concepts to
our respondents, and comparing both repeated and newly created measures cross-
nationally, it was imperative we field, to the best of our ability, one comparable
survey instrument across countries. To aid researchers in making cross-national
comparisons, everyone agreed it was essential to translate just one version of an
English questionnaire into Modern Standard Arabic to be fielded across countries,
and that all remaining language questionnaires go through the same rigorous
translation protocols put in place by the study team. To facilitate ease in language
translation, we would construct our measures as clearly and definitively as possi-
ble, seeking the advice of our collaborators during every stage of the measurement
design. We believed this strategy would allow for the concepts to be understood by
our respondents regardless of study site location.

311
For all intents and purposes, language translation actually began during our
first workshop, when we made available to our collaborators an initial draft of
the pretest questionnaire. As we went through each question in detail, our col-
laborators noted measures where our English wording failed to convey our objec-
tives successfully. For example, the survey included a question asking respondents
about the relationship between development and “individualism.” In Arabic,
there is not an adequate phrase to convey this term, so the English was modified
to read “freedom of choice over one’s own life”—a phrase possible to translate into
comprehensible Arabic. Conversely, our team was able to detect instances when
collaborators believed they understood our intended meaning but had, in fact, mis-
interpreted it. In a set of questions asking respondents to rate the “development”
of a series of countries on a scale from 1 to 10 points, the term development was
understood during the first iteration of translation to be evolution and was trans-
lated as such into Arabic, resulting in a different question than we had intended.
In all such occurrences, we iterated back and forth as a group to formulate the
English phrasing required to create a measure that was understood comparably
by all. In the end, we hoped this initial “pretranslation” exercise would result in
far fewer translation issues when we were ready to finalize the pretest instrument
than would have otherwise been the case had we skipped this important step.
Before formal questionnaire translation, we developed a rigorous translation
protocol that consisted of steps each participating country was required to fol-
low. Step 1, of course, was for the English questionnaire to be translated into the
language of the study site. Step 2 was to have two independent translators from
each country (and for each language represented within that country) produce
a back-translated English questionnaire derived from the study site language
questionnaire(s). Back-translators were required to have had no previous contact
with the original English draft. Step 3 required the two back-translators to review
and reconcile into one instrument the two separate back-translations. Step 4 was
to send the single back-translation to the US team.
As previously discussed, the US team and the study site collaborators agreed
the collaborators would produce just one Arabic version of the questionnaire.
This meant an identical Arabic questionnaire would be used in Egypt, Iraq,
Lebanon, and Saudi Arabia. All four countries participated in the first round of
back-translations. Initially, our Saudi collaborators took responsibility for produc-
ing the English-to-Arabic translation; when the complexity of it became obvious,
the task quickly became a joint venture among all Arabic-speaking study sites.
Thus, problems relative to regional differences and dialects were expected to be
reduced, because working as a group forced each participating country to trans-
late, from English to Arabic, words and phrases that would be understood compa-
rably by respondents regardless of study site. Then, each study country produced
an Arabic-to-English back-translation per the protocol. After the US research team
received the back-translated Arabic-to-English questionnaires, each was compared
carefully with the original English version as well as with the back-translations
of the remaining countries providing the Arabic-to-English translations. As we
compared each back-translated question with what was in our original English

312
version, we only gave approval to each question in the English-to-Arabic transla-

tion if at least three of the four country back-translations agreed with the origi-
nal English version. Whenever two or more countries disagreed with the original
English version, we flagged the item and asked our collaborators to resolve the
differences jointly. We then reviewed the revised translation and either approved
it or sent it back to our collaborators for further work.
The translation protocol was repeated for each remaining study-site language.
For all participating countries, some translation “errors” turned out to be improve-
ments over our original English questions. In the previous example regarding
the difficulty in translating the English term individualism, the rephrasing of the
concept into “freedom of choice over one’s own life” specifically states what we
intended when using the otherwise possibly ambiguous term. In those cases,
we changed our English to match the English back-translation. We have found
translation refinement to be an evolving process, beginning during measurement
formulation and continuing through finalization of the production questionnaire.
Some Arabic-to-English translations uncovered potential differences between
how respondents in the Middle East and respondents in Western countries visual-
ize and process rating scales mentally. Tanzer (2005) cautions against administer-
ing visual representations to right-to-left readers that are meant to be processed
from left-to-right. In studies comparing the results of a pictorial inductive reason-
ing exercise administered to Arabic-educated Nigerian and Togolese high school
students with that of an Austrian calibration sample, researchers found the Arab-
educated students exhibited far more difficulty using the left-to-right processing
format required by the test than the Austrians, because Arabic is read from right
to left.
This finding should be of great concern to all research teams planning to use
“showcards” as an interviewing aid. Showcards are visual depictions shown to
the respondent to help facilitate ease of response to difficult concepts or to ques-
tions with numerous response categories (Bradburn, Sudman, & Wansink, 2004).
Figure 10.1 and Figure 10.2 are examples of the types of showcards respondents
might expect to see during the course of the interview.
Child Qualities of Importance

Please choose up to five qualities
• Independence
• Hard work
• Feeling of responsibility
• Imagination
• Tolerance and respect for other people
• Thrift
• Determination, perseverance
• Religious faith
• Unselfishness
• Obedience
Figure 10.1 Showcard Example for Survey Questions Containing Numerous

Response Categories.

313
How democratic is your government today?
Not at all Completely

democratic democratic
1 2 3 4 5 6 7 8 9 10
Figure 10.2 Showcard Example for Survey Questions Containing Response

Categories Constructed Along a Left-to-right Reading Linear Scale (Smallest
Value=Least, Largest Value=Most).
Figure 10.1 is an example of a showcard that can be used as an aid to respon-

dents when the question has numerous response categories. In this particular
case, the respondent can choose from 10 possible answer categories—far too
many to remember from the interviewer’s verbal reading of the question. The
card is straightforward and unambiguous. Figure 10.2 displays a pictorially rep-
resented 1-to 10-point democracy scale, with the extreme ends labeled as “Not at
all democratic” and “Completely democratic.” The card is meant to be processed
from left to right, with 1 depicting the lowest anchor and 10 depicting the highest.
The card contains a picture of the scale along with written content at each end of
it describing the anchor points.
After our team received the Arabic-to-English back-translations of showcards of
the type depicted in Figure 10.2, we were alerted to potential problems relative to
the design. Within individual back-translated questionnaires, scales were labeled
inconsistently; for some measures the visual appeared as we had intended it, but
for other measures the written content describing the anchor points appeared
over the wrong (opposite) anchor. Figure 10.3 is an example of such a discrepancy;
the anchoring labels have been reversed so that 1 is labeled “Completely demo-
cratic” and 10 is labeled “Not at all democratic.”
How democratic is your government today?
Completely Not at all

democratic democratic
1 2 3 4 5 6 7 8 9 10
Figure 10.3 Example of Showcard Error that may Occur During Translation of

a Left-to-right Reading Linear Scale by Right-to-left Reading Translators (Smallest
Value=Most, Largest Value=Least).
Not only did this scale translation discrepancy have enormous implications
for our measurement validity, it cast potential doubt on the validity of scales
administered during past surveys. We immediately sought input and advice
from our collaborators and learned the error had occurred during the back-
translation from Arabic-to-English, rather than during the initial translation

314
from English-to-Arabic. Although we were relieved the English to Arabic had

been translated correctly and that the left-to-right processing difference became
a point of confusion only during the back-translation, this incident underscores
just how critical the translation process is to the overall success of an interna-
tional research project.
The back-translation exercise also revealed a lack of uniformity across the
questionnaire relative to the translation of words such as development or morality.
For example, the Arabic terms for morality and ethics were used interchangeably
throughout the question sets, although the word morality was used consistently in
the original English version of the questionnaire. The same was true for the word
development, which, in addition to other phrases, was translated as evolution in at
least one instance, as mentioned earlier, although the English questionnaire used
the same word throughout. In general, we found that abstract concepts, such as
development and morality, as well as individualism, were particularly vulnerable
to mistranslation. We asked our collaborators to resolve this practice and to come
up with consistent terms throughout the survey instrument. We also noticed that
some English words used to convey subtle differences in strength were omitted
in many back-translations (for example, phrases such as “or more,” “if any,” or “if
only”). Because these omissions sometimes resulted in inaccuracies relative to
the intended meanings of questions, it was important to find a way to make these
qualifiers obvious in the translations. Here, again, we sought the advice and input
of our study-site collaborators to rework the translations to reflect these important
subtleties.
As the research team evaluated each of the study countries’ back-translated
questionnaires, we compiled a list of general observations along with any question-
by-question concerns arising from the back-translation exercise. We then sent this
document to those participating in the first round of back-translation and asked
they meet and resolve each of the issues and concerns jointly. We also requested
they provide us with one written summary from their group explaining how each
question/concern had been resolved, along with the corrected original English
version and the corrected study-site language version(s).
Our collaborators displayed an amazing spirit of cooperation and worked
together diligently to address our concerns. Communicating via Skype and e-mail,
our overseas colleagues dealt with each concern systematically, came to an accept-
able resolution together, and, ultimately, were responsible for one revised Arabic
and English questionnaire for our approval. Their version contained corrections
and comments as well as suggestions for improvement. The US group then final-
ized the questionnaire and gave the go-ahead for Arabic questionnaire pretesting
to begin. We then applied the same procedures to produce questionnaires written
in the primary languages of remaining the non-Arabic-speaking study countries.
At the end of this iterative process, we were very pleased with the final trans-
lations of the survey instrument. We fully believe our advance discussions
during the first workshop, along with the commitment of our collaborators to
work together to produce one common instrument, and our previous interna-
tional question design experience contributed to a relatively smooth translation
phase. Yet, despite our efforts from the very beginning of the project to design

315
comparable measures across countries, it is crucial to note that discrepancies

between the English and study-country languages did occur. Anyone undertak-
ing an international project such as this should devote considerable resources to
English-to-study-site translation because it is critical to accurate measurement of
the constructs and hypotheses put forth in the project goals.
Questionnaire Design and Layout

As with sampling protocol and questionnaire translation, many values studies in
the region have delegated questionnaire design and layout to each country par-
ticipating in the project, resulting in a unique product that is not comparable
cross-nationally. Undoubtedly, this is the most convenient approach, but the lack
of uniformity in questionnaire appearance, particularly among countries sharing
a common language or text direction, may lead to differences in questionnaire
administration. Obviously, this contributes to measurement error as well as com-
parability among countries. To avoid this problem, it is advisable for all question-
naires to have the same physical layout regardless of which language is used. In
our project, after translations were completed, we reviewed the final versions and
verified that all Arabic questionnaires in particular were identical in layout across
study site countries.
Pretesting
One of the costliest mistakes a researcher can make before launching production
interviewing is to neglect a thorough pretest of the questionnaire. No matter how
many “experts” have pronounced the instrument suitable, the true test is realized
through its administration to everyday, ordinary people. In both single-country
and cross-national surveys, a pretest serves a number of important functions
related to production interviewing; it tests (a) the ease of navigability through the
instrument; (b) the time it takes to administer it; (c) the flow of subject matter,
whether questions are ordered properly, and whether the questions are being per-
ceived properly (both cognitively and translation-wise); and (d) overall respondent
reaction to the questionnaire. For cross-national surveys, pretesting takes on the
added value of testing how well the instrument suits the comparative aims of the
project—in other words, whether construct and measurement equivalence have
been achieved. Therefore, after all the advantages of pretesting have been taken
into account, it behooves the research team to allocate adequate resources to the
budget and project timeline to accommodate it properly.
There is a variety of ways to pretest a cross- national instrument (Kasper,
Peytcheva, Yan, Lee, Liu, & Hu, 2016), but the most comprehensive method is
to test the full questionnaire on a small number of respondents located through-
out the research sites, taking care to replicate as many elements of the production
protocol as possible, including mode of administration, population composition,
and interviewer characteristics. By doing so, the researcher is in a good position

316
to evaluate all aspects of the study—from the physical aspect of the questionnaire,
to interviewer evaluation of it and the protocol, to respondent reaction relative to
the subject matter. Interviewer and respondent feedback should be taken seri-
ously, and steps should be taken to improve the questionnaire in ways that make
it more effective without compromising the aims of the research. For example,
our project’s pretest revealed the sensitive nature of the question regarding belief
in God in Saudi Arabia and in Egypt, which was discussed earlier in this chapter.
Several pretest respondents were so upset they refused to continue the interview.
Without the addition of the introduction to this question, which offered context and
decreased respondent discomfort, it is likely our response rates in the subsequent
national survey would have suffered. Gelfand (2012) reported similar results in
a pretest for a project in the United States and Japan, in which the research pro-
tocol was successful among American pretest respondents, but a failure among
Japanese respondents, resulting in a complete revision of the protocol before the
larger subsequent study.
Interviewing Techniques
Efforts to achieve questionnaire comparability may be lost if interviewer training
does not receive proper attention. Poorly trained interviewers affect the quality of
data collection by contributing to multiple sources of error, including sampling
error, nonresponse error, and measurement error, that threaten cross-national
data comparability between countries. The questionnaire can be safeguarded in
some ways to aid in standardization of administration. Much can be done within
the physical layout of the instrument itself, such as including standardized built-
in probes and definitions. Answer categories the respondent should not hear, such
as “Don’t know,” can be clearly flagged as DO NOT READ OUT. Standardized
transitions can be inserted at the beginning of new sections or at subject matter
changes within sections, thereby eliminating any perceived need for the inter-
viewer to invent them. Question-by-question objectives and allowable definitions
must be produced for production interviewing and, to the extent possible, inter-
viewer training should be standardized across study sites.
Although there are exceptions, in the Middle East it is common for the data col-
lection contractor to subcontract interviewers temporarily in lieu of maintaining a
permanent interviewing staff. This practice may result in a combined cross-national
interviewing staff whose quality of training and skill levels vary greatly, particularly
between agencies who subcontract and agencies who maintain a permanent staff.
A potential disparity of this magnitude should be enough to encourage researchers to
allow time and resources to conduct interviewer training that includes both standard-
ized interviewing techniques and study-specific training. Many projects choose the
“train-the-trainer” method for interviewer training before field production activities,
often because of budget constraints. This style of training calls for each study country
to send several individuals to one common training session to receive instruction;
they then return to the study sites to train the rest of the field staff (Alcser, Clemens,
Holland, Guyer, & Hu, 2016; Pennell, Harkness, Levenstien, & Quaqlia, 2010).

317
Holding a joint “train-the-trainer” workshop for study-site collaborators and

their field supervisors before the beginning of the data collection effort is an effec-
tive strategy for cross-national studies in the Middle East. The advantages of hav-
ing representatives from each participating country trained together at the same
time far outweigh the disadvantages. The meeting should serve both to provide
standardized general interviewer training and to introduce the questionnaire and
conduct study-specific training of study-site field supervisors so they, in turn, may
train their interviewing staff.
Using well-established interviewer training protocols for general interviewer
training is imperative to reduce differences in delivery of questions as well as bias
from differential unit response.3 Differential unit response rates across study coun-
tries can affect data quality in cross-national survey projects. Evidence suggests
there is a current trend of decreasing response rates at the target respondent level
in primarily Western multinational survey data collection projects (de Leeuw &
de Heer, 2002), but there is much less literature on response rates from multi-
country data collections with non-Western populations. And, argue Couper and de
Leeuw (2003), “in cross-national or cross-cultural survey research, non-response
issues have been largely ignored” (p. 157). The severity of the problem of nonre-
sponse can differ greatly among countries, and when there are large differences,
the validity of cross-cultural comparative analyses is severely decreased because of
differential nonresponse bias.
Similar differential bias can arise as a result of survey item nonresponse.
To this end, research teams should be prepared to offer study-s pecific train-
ing as well. Study-specific training introduces the purpose of the research
and allows researchers to answer questions about the overall study, respond
to interviewer queries concerning specific questions or sections, and
address any questionnaire content concerns. The training agenda should
also devote significant time for interviewer–r espondent role-p laying activi-
ties for the supervisors to receive constructive evaluative feedback from the
research teams.
Subject matter sensitivities should be discussed during the joint training
workshop as well. Extensive interviewer training should be provided about strate-
gies helpful to use when asking about sensitive topics, such as reading questions
without hesitation and in a confident neutral tone of voice, accepting responses
without reaction, having the field staff role-play interviewer–upset respondent
scenarios, and providing interviewers with information that will reassure respon-
dents the interview is voluntary and confidential (Fowler & Mangione, 1990). This
topic is particularly important in the Middle East, where, as discussed earlier in
the chapter, there is potential for surveys to occur in armed conflict zones or in
conditions that are otherwise unstable.
Last, our interviewer training included a discussion of standardized field-
work protocols to which interviewers must adhere to limit differential bias
resulting from interviewer behavior both within and among study countries.
Because all interviews were conducted face-to-face, we specified a minimum
of three visits to the sampled household to obtain an interview with the target
respondent.

318
The shared venue of the training site offers an opportunity to instill a sense
of excitement and ownership toward the project among the interviewing teams,
as was done for the study site collaborators early during our project. And, impor-
tantly, the meeting has the additional benefit of allowing all the trainees to hear
the same messages from the research team at the same time, rather than through
individual country trainings or second-hand instructions delivered through our
collaborators. To maximize efficiency before and during the training, subject
materials should be distributed well ahead of time so they may be translated into
the appropriate language, and it is useful to have professional translators available
throughout the training.
Cross-National Data Entry
Analytical data sets disseminated from past Middle Eastern values survey projects
have been produced on a country-to-country basis with little to no coordination
among research sites. Differences among data sets are numerous; for example,
variables appear across data sets that are often out of sync between the data records,
responses to cross-nationally collected variables do not always use identical codes
for measures that require it, missing data codes may not be handled consistently
across countries, and country-specific variables and codes are often poorly docu-
mented. Difficulties with past Middle East cross- national values comparison
analyses have been compounded by the fact that each individual data set and corre-
sponding documentation produced by participating countries has had differences
in universally asked questions ranging from extremely obvious, to extremely sub-
tle. The differences have hindered efficient and timely analysis of the data because
an enormous amount of data management harmonization work has been required
before any scientific comparison of the countries can be conducted. Of course,
these issues are not just limited to cross-national surveys in the Middle East; com-
parative survey projects elsewhere struggle with many of the same concerns.
If harmonization principles are applied consistently throughout each phase of
the cross-national research project, the burden of data set harmonization related to
scientific analyses activities will be greatly reduced. Then, regardless of the inter-
viewing mode—be it paper–pencil or computer-assisted personal interviews—one
common, predesigned coding shell can be used to create data records for inter-
views collected from each respondent across countries. The existence of one com-
mon protocol helps avoid any procedural incongruence, such as differences in
code categories in variables that should have been identical among all countries,
variables appearing out of order in the data set, varying record lengths, extra and
uncalled-for variables, and so forth.
Discussion
This chapter has highlighted many areas that are important to address when plan-
ning cross-national studies of values and values change in the Middle East. We
urge anyone undertaking such a venture to allow plenty of preproduction time

319
and to anticipate areas where inconsistencies among countries can occur. These
areas include construct interpretation, measurement design, language transla-
tion, sampling, questionnaire design, interviewing practices, and data entry/
data management tasks. Our team was fortunate in that each of us had learned
important lessons from prior international experience and was able to anticipate
and address effectively issues of comparability. Based on our own experiences,
we recommend taking steps early during the design process to set standards and
control for standardization to the extent possible. In our case, this was achieved by
producing one version of the questionnaire per language and implementing com-
mon sampling specifications, interviewer training specifications, and data entry
specifications, along with developing standardized strategies for dealing with sub-
ject matter sensitivities. It is also important to engage data collection collabora-
tors throughout project development and give them a sense of project ownership.
Seek their advice actively on certain issues and implement their suggestions when
appropriate. Keep the lines of communication open to collaborators and, if need
be, allow them to vent their frustrations. Be patient throughout each phase of the
process and flexible during occasions that call for it.
In addition, try to anticipate early during questionnaire development which top-
ics may be most culturally sensitive and make an attempt, when constructing the
instrument, to address those concerns. Emergent technology allows researchers and
collaborators to keep in touch with far greater ease than in the past. Take advantage
of communication software such as Skype to have regular face-to-face meetings with
research site collaborators. And finally, remember that you will surely encounter dif-
ficulties during production interviewing for which you did not prepare in advance.
However, the good relationships built with collaborators will help to ameliorate such
difficulties, and the data collected at each research site will prove to be of extremely
high quality as a result of everyone’s dedication to the success of the project.
Notes
1. Both authors contributed equally to this chapter and therefore have been listed
alphabetically.
2. See Groves et al. (2009) for a comprehensive overview of the survey life cycle.
3. See Fowler and Mangione (1990) and Groves et al. (2009, Chap. 9) for a com-
prehensive overview of general interviewer training and standardized interviewing
techniques.
References
Abdulrahim, S., Ajrouch, K. J., Jammal, A., & Antonucci, T. C. (2012). Survey methods
and aging research in an Arab sociocultural context: A case study from Beirut,
Lebanon. Journal of Gerontology Series B, 67(6), 775–782.
Alcser, K., Clemens, J., Holland, L., Guyer, H., & Hu, M. (2016). Interviewer recruit-
ment, selection, and training. In Guidelines for best practice in cross-cultural surveys.
Ann Arbor, MI: Survey Research Center, Institute for Social Research, University

320
of Michigan. Retrieved August 29, 2016, from http://ccsg.isr.umich.edu/index.

php/chapters/interviewer-recruitment-selection-and-training-chapter/.
American Association for Public Opinion Research. (2008). Standard definitions: Final
dispositions of case codes and outcome rates for surveys [On-line]. Available: http://
www.aapor.org/AAPORKentico/AAPOR_Main/media/MainSiteFiles/Standard_
Definitions_07_08_Final.pdf. Retrieved August 27, 2015.
Babbie, E. (2013). The basics of social research. Belmont, CA: Wadsworth, Cengage
Learning.
Beatty, P. C., & Willis, G. (2007). Research synthesis: The practice of cognitive inter-
viewing. Public Opinion Quarterly, 72(2), 287–311.
Bijl, R. V., Van Zessen, G., Ravelli, A., De Rijk, C., & Langendoen, Y. (1998). The
Netherlands Mental Health Survey and Incidence Study (NEMESIS): Objectives
and design. Social Psychiatry and Psychiatric Epidemiology, 33(12), 581–586.
Bradburn, N. M., Sudman, S., & Wansink, B. (2004). Asking questions: The definitive
guide to questionnaire design. New York: Wiley.
Chapman, D. W. (1983). The impact of substitution on survey estimates. In W.
G. Madow, I. Olkin, & D. B. Rubin (Eds.). Incomplete data in sample surveys, 2
(pp. 45–61). New York: Academic Press.
Clark, J. (2006). Field research methods in the Middle East. PS: Political Science and
Politics, 39(3), 417–424.
Couper, M. P., & de Leeuw, E. D. (2003). Nonresponse in cross-cultural and cross-
national surveys. In J. A. Harkness, F. J. R. Van de Vijer, & P. H. Mohler (Eds.).
Cross-cultural survey methods (pp. 117–137). Hoboken, NJ: Wiley.
de Leeuw, E. D., & de Heer, W. (2002). Trends in household survey quality: A lon-
gitudinal and international comparison. In R. M Groves, D. A. Dillman, J.
L. Eltinge, & R. J. A. Little (Eds.). Survey Nonresponse (pp. 41–54). New York:
Wiley.
de Leeuw, E. D., Hox, J. J., & Snijkers, G. (1998). The effect of computer-assisted
interviewing on data quality: A review. In B. Blyth (Ed.). Market research and infor-
mation technology (pp. 173–198). Esomar monograph. Amsterdam: Esomar.
Elliot, D. (1993). The use of substitution in sampling. Survey Methodology Bulletin,
33, 8–11.
Fowler, F. J., & Mangione, T. W. (1990). Standardized survey interviewing: Minimizing
interviewer-related error. Beverly Hills, CA: Sage Publications.
Gelfand, M. J. (2012). The trials and tribulations of cross-cultural research. In F.
Leong & A. Ryan (Eds.). Conducting multinational research projects in organiza-
tional psychology: Challenges and opportunities (pp. 179–199). Baltimore: United
Book Press.
Groves, R. M. (2006). Nonresponse rates and nonresponse bias in household sur-
veys. Public Opinion Quarterly, 70(5), 646–675.
Groves, R. M., Fowler, Jr., F. J., Couper, M. P., Lepkowski, J. M., Singer, E., &
Tourangeau, R. (2009). Survey methodology. Hoboken, NJ: Wiley.
Häder, S., & Gabler, S. (2003). Sampling and estimation. In J. A. Harkness, F. J. R.
Van de Vijer, & P. H. Mohler (Eds.). Cross-cultural survey methods (pp. 117–137).
Hoboken, NJ: Wiley.

321
Harik, I. (1987). Some political and cultural considerations bearing on survey research in
the Arab world. In M. A. Tessler, M. Palmer, T. E. Farah, & B. L. Ibrahim (Eds.). The eval-
uation of survey research in the Arab world (pp. 66–74). Boulder, CO: Westview Press.
Harkness, J. (2003). Questionnaire translation. In J. Harkness, F. J. R. Van de Vijver, &
P. Mohler (Eds.). Cross-cultural survey methods (pp. 35–56). Hoboken, NJ: Wiley.
Harkness, J., Bilgen, I., Cazar, A. C., Hu, M., Huang, L., Lee, S., et al. (2016).
Questionnaire design. In Guidelines for best practice in cross-cultural surveys. Ann
Arbor, MI: Survey Research Center, Institute for Social Research, University of
Michigan. Retrieved August 29, 2016, from http://ccsg.isr.umich.edu/index.
php/chapters/questionnaire-design-chapter/.
Harkness, J., Braun, M., Edwards, B. Johnson, T., Lyberg, L., Mohler, P., et al. (2010a).
Comparative survey methodology. In M. Braun, B. Edwards, J. Harkness, T. Johnson,
L. Lyberg, P. Mohler, B. E. Pennell, & T. W. Smith (Eds.). Survey methods in multina-
tional, multiregional and multicultural contexts (pp. 1–16). Hoboken, NJ: Wiley.
Harkness, J. A., Edwards, B., Hansen. S. E., Miller, D. R., & Villar, A. (2010b).
Designing questionnaires for multipopulation research. In M. Braun, B.
Edwards, J. Harkness, T. Johnson, L. Lyberg, P. Mohler, B. E. Pennell, & T. W.
Smith (Eds.). Survey methods in multinational, multiregional and multicultural con-
texts (pp. 33–58). Hoboken, NJ: Wiley.
Harkness, J., Van de Vijver, F. J. R., & Johnson, T. J. (2003). Questionnaire design in
comparative research. In J. Harkness, F. J. R. Van de Vijver, & P. Mohler (Eds.).
Cross-cultural survey methods (pp. 19–34). Hoboken, NJ: Wiley.
Harkness, J., Villar, A., & Edwards, B. (2010c). Translation, adaptation, and design.
In M. Braun, B. Edwards, J. Harkness, T. Johnson, L. Lyberg, P. Mohler, B. E.
Pennell, & T. W. Smith (Eds.). Survey methods in multinational, multiregional and
multicultural contexts (pp. 117–140). Hoboken, NJ: Wiley.
Heath, A., Fisher, S. & Smith, S. (2005). The globalization of public opinion research.
Annual Review Political Science, 8, 297–333.
Heeringa, S. G., & O’Muircheartaigh, C. O. (2010). Sampling designs for cross-
cultural and cross- national survey programs. In M. Braun, B. Edwards, J.
Harkness, T. Johnson, L. Lyberg, P. Mohler, B. E. Pennell, & T. W. Smith
(Eds.). Survey methods in multinational, multiregional and multicultural contexts
(pp. 251–267). Hoboken, NJ: Wiley.
Heeringa, S. G., Wells, J. E., Hubbard, F., Mneimneh, Z. N., Chiu, W. T., Sampson,
N. A., et al. (2008). Sample designs and sampling procedures. In R. C. Kessler &
T. B. Ustun (Eds.). World mental health surveys (pp. 14–32). New York: Cambridge
University Press.
Ibrahim, S. E. (1987). The agony and the ecstasy of social research in the Arab world.
In M. A. Tessler, M. Palmer, T. E. Farah, & B. L. Ibrahim (Eds.). The evaluation of
survey research in the Arab world (pp. 27–34). Boulder, CO: Westview Press.
Inglehart, R., Basanez, M., Diez-Medrano, J., Halman, L., & Luijkx, R. (2004). Human
beliefs and values. Mexico City, Mexico: Siglo XXI Editores.
Kapiszewski, A. (2006). Arab versus Asian migrant workers in the GCC countries.
Paper presented at the United Nations Expert Group Meeting on International
Migration and Development in the Arab Region, Beirut, Lebanon, May 15–17,

322
Karam, E. G., Mneimneh, Z. N., Dimassi H., Fayyad, J. A., Karam, A. N., Nasser, S. C.,
et al. (2008). Lifetime prevalence of mental disorders in Lebanon: First onset,
treatment, and exposure to war. PLoS Med, 5(4), e61.
Kasper, R., Peytcheva, E., Yan, T., Lee, S., Liu, M., & Hu, M. (2016). Pretesting. In
Guidelines for best practice in cross-cultural surveys. Ann Arbor, MI: Survey Research
Center, Institute for Social Research, University of Michigan. Retrieved August 29,
2016, from http://ccsg.isr.umich.edu/index.php/chapters/pretesting-chapter/.
Lynn, P. (2003). Developing quality standards for cross-national survey research: Five
approaches. International Journal of Social Research Methodology, 6(4), 323–336.
Lynn, P., Häder, S., Gabler, S., & Laaksonen, S. (2007). Methods for achieving equiva-
lence of samples in cross-national surveys: The European Social Survey experi-
ence. Journal of Official Statistics, 23, 107–124.
Mneimneh, Z., Axinn, W. G., Ghimire, D., Cibelli, K. L., & Salih Alkaisy, M. (2014).
Conducting surveys in areas of armed conflict. In R. Tourangeau, B. Edwards,
T. P. Johnson, K. W. Wolter, & N. Bates (Eds.). Hard-to-survey populations
(pp. 134–156). Cambridge: Cambridge University Press.
Moaddel, M., & Karabenick, S. A. (2008). Religious fundamentalism among young
Muslims in Egypt and Saudi Arabia. Social Forces, 86, 1675–1710.
Moaddel, M., & Karabenick, S. A. (2013). Religious fundamentalism in the Middle
East: A cross-national, inter-faith, and inter-ethnic analysis. Leiden: Brill.
Myers, S. L. (2010). Politics delay an Iraqi census [On-line]. The New York Times.
Available: http://www.nytimes.com/2010/12/07/world/middleeast/07iraq.html.
Retrieved November 27, 2013.
Oldendick, R. W., Bishop, G. F., Sorenson, S. B., & Tuchfarber, A. J. (1988). A compar-
ison of the Kish and last birthday methods of respondent selection in telephone
surveys. Journal of Official Statistics, 4(4), 307–318.
Pennell, B. E., Harkness, J., Levenstien, R., & Quaqlia, M. (2010). Challenges in cross-
national data collection. In M. Braun, B. Edwards, J. Harkness, T. Johnson, L.
Lyberg, P. Mohler, B. E. Pennell, & T. W. Smith (Eds.). Survey methods in multina-
tional, multiregional and multicultural contexts (pp. 269–298). Hoboken, NJ: Wiley.
Romano, D. (2006). Conducting research in the Middle East’s conflict zones.
PS: Political Science and Politics, 39(3), 439–441.
Tanzer, N. K. (2005). Developing tests for use in multiple languages and cul-
tures: A plea for simultaneous development. In R. K. Hambleton, P. F. Merenda,
& C. C. Spielberger (Eds.). Adapting educational and psychological tests for cross-
cultural assessment (pp. 235–263). Hillsdale, NJ: Erlbaum.
Tessler, M. (2011). Public opinion in the Middle East: Survey research and the political
orientations of ordinary citizens. Bloomington: Indiana University Press.
Tessler, M. A., Palmer, M., Farah, T. E., & Ibrahim, B. L. (1987). The evaluation and
application of survey research in the Arab world. Boulder, CO: Westview Pres.
Thornton, A. (2001). The developmental paradigm, reading history sideways, and
family change. Demography, 38(4), 449–465.
developmental paradigm on family life. Chicago: University of Chicago Press.

323
Thornton, A., Achen, A., Barber, J., Binstock, G., Garrison, W., Ghimire, D., et al.
(2010). Creating questions and protocols for an international study of ideas
about development and family life. In M. Braun, B. Edwards, J. Harkness,
T. Johnson, L. Lyberg, P. Mohler, B. E. Pennell, & T. W. Smith (Eds.). Survey
methods in multinational, multiregional and multicultural contexts (pp. 59–74).
Hoboken, NJ: Wiley.
United Nations. (2003). Press briefing on Iraq demographics [On-line]. Available: http://
www.un.org/News/briefings/docs/2003/iraqdemobrf.doc.htm.RetrievedNovember
27, 2013.
Verma, V., Scott, C., & O’Muircheartaigh, C. (1980). Sample designs and sampling
errors for the World Fertility Survey. Journal of the Royal Statistical Society, Series
A, 143(4), 431–473.
Williams, N. E., Thornton, A., & Young-DeMarco, L. C. (2013). Migrant values and
beliefs: How are they different and how do they change? Journal of Ethnic and
Migration Studies, 40(5), 796–813.
Willis, G. (2005). Cognitive interviewing: A tool for improving questionnaire design.
Thousand Oaks, CA: Sage Publications.
World Values Survey. (2015). Documentation/downloads. Available: http://www.world-
valuessurvey.org/WVSContents.jsp. Retrieved August 27, 2015.
Zurayk, H. (1987). The question of measurement in survey research in the Arab
world. In M. A. Tessler, M. Palmer, T. E. Farah, & B. L. Ibrahim (Eds.). The evalu-
ation of survey research in the Arab world (pp. 51–56). Boulder, CO: Westview Press.

324
325
CHAPTER 11 An Analysis of Subjective Culture

in the Middle East
Lessons Learned from a Qualitative Research
Program
Janetta Lun, Michele J. Gelfand, C. Bayan Bruss,
Lily Assaad, Zeynep Aycan, Munqith M. Dagher,
and Abdel-Hamid Abdel-Latif
As this volume attests, during the past several decades, research on values and
value change has increased dramatically in the Middle East and North Africa.
New theoretical models, drawn from multiple disciplines—from political science,
sociology and psychology, and even computer science—have shed new light on
the dynamics of the Arab Spring, and many rigorous multicountry surveys have
helped to track values and value change across the region. Perhaps most impor-
tant, this collective body of research has given voice to people in a region that has
been largely neglected in the social sciences.
In this chapter, we complement the other efforts in this volume and show that
to understand the region fully, we need (a) to broaden the values we study in the
Middle East and North Africa to include important culture-specific focal concerns,
including honor, wasta (intermediary), fatalism, and modesty, among others, and
(b) to understand how these values are understood subjectively and represented
by individuals in the region. Within this context, we discuss a large-scale quali-
tative interview of more than 400 individuals we conducted in the region. Not
surprisingly, the challenges one experiences in carrying out qualitative research
are highly distinct from those experienced with survey research. Accordingly, just
as de Jong and Young-De Marco (Chapter 10) advance best practices for survey
research, we highlight the important challenges and lessons learned that arise
when conducting qualitative research and strategies to deal with them in this final
chapter of the volume.
Qualitative research complements quantitative modes of investigation in
affording a great degree of depth to understand the values of people in a partic-
ular region. Although surveys allow for context-free judgments, interviews allow
investigators to understand the subjective meanings people associate with val-
ues. For example, consider this question asked frequently in the World Values
326
Surveys cross-nationally: What are the most favorable qualities for children to have?
Hypothetically speaking, if representative samples from Saudi Arabia and China
mention obedience at equivalent rates, can one conclude that people from these
two countries value this quality equally? An unqualified affirmative answer may
produce knowledge that may be incomplete, if not misleading. Common notions
of obedience in Saudi Arabia may imply a greater level of constraint and regula-
tion than those of the Chinese citizen. Without a greater understanding of people’s
subjective representation or mental picture of what obedience means across cultures,
our understanding of the cultural differences between the two countries would be
limited. This example illustrates the importance of taking a qualitative approach
to investigate how people understand and make the meaning of values as guiding
principles of their life.
Moreover, qualitative research methods not only allow us to analyze values
that have received little attention, but also provide critical information needed
to develop valid scales to assess their relevance in shaping human interactions.
Although it would be useful to develop quantitative scales on such values as honor,
it would be much more useful to examine how people understand the construct of
honor and modesty across different Middle Eastern and North African countries
before engaging in large-scale survey research. This understanding is particularly
important because what honor and modesty might mean in the United States,
where many surveys are developed and then exported to other countries, is likely
to be different from the meanings of the constructs in the Middle East and North
Africa, notwithstanding variations within the region. Accordingly, qualitative
research is often recommended to explore constructs that have received little atten-
tion in prior quantitative survey research (Gelfand, Raver, & Holcombe, 2002).
To complement the survey research discussed in this volume, and to broaden
the values that we typically investigate with the World Value Survey, we conducted
more than 400 in- depth interviews across Afghanistan, Egypt, Iraq, Jordan,
Lebanon, Pakistan, Turkey, the United Arab Emirate, and the United States. Our
qualitative research program on values in the Middle East and northern Africa
draws on the famous analysis of subjective culture (ASC) method, pioneered
by cross-cultural psychologist Harry Triandis (1972). The ASC method involves
designing a set of questions that asks informants systematically about the mean-
ings, cognitive associations, antecedents, and consequences of constructs of
interest. In this chapter, we discuss issues that arose from the beginning of our
qualitative research process, including decisions on the constructs to be included
in the interviews; interview design, sampling, and implementation; and data pro-
cessing, data analysis, and interpretation. At each step, we discuss challenges that
arose in our own research program and strategies to deal with these challenges.
Determining the Constructs for Inclusion in the Interviews

Determining the focal cultural constructs in our research is an important first step
of the qualitative research process. At this stage, it is critical that the determina-
tion of what should be included should not be determined solely by one particular

327
cultural perspective (e.g., the Western researchers/funders). Our multinational

research team deliberated for many months on which theoretical constructs
should be the focus of our interviews and how they should be structured. What
are the core cultural constructs or value principles in Middle Eastern and North
African countries that guide social interactions and transcend different social
domains that have received little attention? After consulting the literature and
having many in-depth discussions, we arrived collectively at a list of constructs for
the interviews, which included honor, dignity, public image (i.e., face), fatalism,
modesty, and respect. We were also interested in social processes, such as negotia-
tion, collaboration, and conflict and forgiveness, that have important implications
for cross-cultural interactions. All of these constructs are potentially universal, but
we were interested in how people across the Middle East and North Africa make
meaning of these constructs, and their antecedents and consequences.
Another important consideration at the beginning of the research process
was whether any safety and security issues would arise as a result of research
collaboration between Americans and researchers from the region. Given the
context of fairly strong anti-American sentiment among some religious groups
and movements, and the prevalence of conspiratorial thinking that views the
US government with suspicion, there was the concern that collaboration with
members of US academia and a project funded indirectly by institutions con-
nected to the US government may harm local collaborators. Our team discussed
these issues and strategized solutions to minimize any ethical and reputational
issues. As a result, in some cases, some collaborators elected not to be an author
on research papers. More generally, it is important to consult with team mem-
bers not only about the content of the work, but also the implications of the col-
laboration for them in the local context (Tapp, Kelman, Triandis, Wrightsman,
& Coelho, 1974).
Interview Design and Sampling

Two semistructured interview protocols were designed to probe the meaning
and antecedents and consequences of important values in the region based on
the ASC method. Each interview protocol began with a general question about
what a particular construct meant to them and what words they associated with
it (e.g., What words come to mind when we say honor?). The interviewees were
then asked to describe the antecedents and consequences of the construct (e.g.,
How does one demonstrate honor? What makes a person lose honor? What
happen if one’s honor is threatened?), as well as how these antecedents or con-
sequences varied according to different social contexts (e.g., How do men ver-
sus women demonstrate their honor? Does it make a difference if the person
who threatens your honor is someone you know well or someone you do not
know very well?). All interview protocols were translated and back-translated
by independent translators who had not seen the original English version in
each country (Brislin, 1980), and the lead investigator then compared them to
identify any discrepancies. Numerous iterations based on this process occurred
An Analysis Of Subjective Culture In The Middle east | 327

328
over several months before the interviews were ready to be piloted extensively
in each country. Appendix A contains a sample interview protocol that includes
questions on fatalism, public image and face, modesty, dignity, respect, core
values, and honor.
After we designed our interviews, we then worked for several months
to standardize the interview questions and protocols across countries.
Theoretically, standardization maximizes the validity of cross-country com-
parison. As we learned from our experience, however, standardization is
beyond making sure that the interviews are conducted strictly in the same
fashion across countries and that words are translated in the same way. In
fact, being sensitive and accommodating to local understandings of the inter-
view questions are critical to making final decisions of the wording of the
questions. As a case in point, many unexpected interpretations came up in our
pilot tests regarding the questions on modesty. Upon examining our data, we
found that some people interpreted the question on modesty in terms of not
being arrogant, but others interpreted it as being physically modest. Likewise,
the translation of social connection/networking was more consistent with our
intended meaning in the Arabic-speaking country (by using the term Wasta).
Wasta or wasata (Arabic‫ هطساو‬wāsiṭah) is an Arabic word that translates loosely
to “nepotism” or “who one knows.” However, the intended construct was
sometimes misunderstood in Turkey, Pakistan, and the United States, where
some people thought it was about “forming friendships,” although the trans-
lation was technically correct. As yet another example, in our questions on
negotiation, we learned we had to use two terms if we were to study the con-
struct in the region properly—the formal term (mufawada) and informal term
(musawama) of negotiation, which equate to negotiation versus bargaining in
English—to cover the range of possible responses and examples. More gen-
erally, ensuring interviewers’ understanding of the intended construct and
their meaning, and relying on local knowledge are critical when doing any
qualitative research.
After we standardized the interview protocols, we discussed sampling con-
siderations at length. Unlike quantitative research, the goal of using an inter-
view method is to cover as many different perspectives on the constructs as
possible, rather than generalizing the findings to the population. Therefore,
the sampling strategy was aimed at recruiting a wide variety of people with
different demographic backgrounds (e.g., gender, age, socioeconomic status)
to collect a wide range of responses in each country. Our stratified sampling
strategy for each country included male and female interviewees in different
age ranges (from 20s–60s) who had different levels of socioeconomic status
and who came from different ethnic and religious backgrounds. Although it
was difficult to obtain data in rural areas, we also sampled people who were the
first generation to move to the city to cover potential differences between rural
and urban regions. The interviews were conducted in the local language and
tape-recorded, and each interview typically lasted 1.5 to 2 hours. This process
resulted in the collection of more than 400 interviews, each with about 12 to 15
pages of transcripts from our nine countries.

329
Implementation of the Interviews

There are several important considerations that arise in the implementation of
cross-cultural qualitative research that warrant attention based on our experience.
Our research team discussed at length issues such as how best to introduce the
interview, how to describe the goal of the research, and how to build rapport with
the interviewees in ways that were standardized but also appropriate across the
countries. After many discussions, we arrived at the final script for the introduc-
tion of the interviews:
Hello, thank you so much for coming here today to talk with me. I really appreci-
ate your kindness and your generosity in meeting with me today. Before we begin
the actual interview, I’d like to tell you more about myself and about this research
project. I am [name] and I work at as a [researcher/teacher/other role] at [institu-
tion] doing [research on/teaching classes on/etc.]. This research project is part of
a large global social psychology study that looks at communication in different
parts of the world. As you know, people from different cultures have different val-
ues and attitudes. They place different levels of importance on relationships and
family. And they vary in the ways they communicate with others. A lot of research
has focused on how people in the United States and Europe communicate and
feel. Some research has looked at these issues in Asia (like Japan and China). But,
very little is known about these issues in the Arab world, so we are conducting
this study with other researchers in other Arab countries to see how people in our
part of the world fare when it comes to social psychological concepts.
As in other interview-based research, we included sets of interview probes to
help facilitate gaining as much information as possible about the meaning and
antecedents and consequences of the constructs. The basic function of probes
is to ask the interviewees to elaborate on their answers. For example, when ask-
ing about the social context of honor, we were interested in the question: Is your
Sharaf related to the Sharaf of other people, and whom? To probe their answers,
the interviewers then followed up with the questions: How contagious is Sharaf?
How interrelated is Sharaf and among whom? Probing questions provide the
interviewers a relatively standardized way to help interviewees elaborate on their
answers. We found that probes also serve the function of rephrasing a question
to help a diverse background of interviewees comprehend the questions in ways
that were intended and helps prevent interviewers from asking leading questions.
Leading questions are questions in which the interviewers suggest intentionally
or unintentionally an answer to the questions. For example, an interviewee in
one case said, “If a person’s honor is attacked, his worth is severely diminished.”
The interviewer then asked, “Will this result in revenge?” The interviewee stated,
“Possibly.” The idea of revenge was suggested to the interviewee in this context
and his responses had to be omitted from the analyses. See Appendix A for more
examples of interview probes.
Extensive interviewer training is essential for the implementation of the inter-
view protocols. At the time, we could not find any published manuals or toolkits
for conducting cross-cultural interviews. As a result, we did extensive research

330
about norms for communication across cultures and compiled a manual for
cross-cultural qualitative interviewing. The manual provided concrete informa-
tion about how to build rapport with the interviewees with considerations of local
communication norms (e.g., how to respectfully address interviewees, approach
interviewees in a kind and nonthreatening manner, be neutral and yet show inter-
est), how to follow up on interviewees’ responses (e.g., seek clarification, ask for
more details and examples, ask about thoughts and feelings, avoid leading ques-
tions), and how to deal with interview problems (e.g., talk about sensitive topics,
interviewee silence, and interviewees appearing anxious).
As a general principle, providing an open, friendly, nonjudgmental and
respectful environment where interviewees feel comfortable to share their experi-
ences, thoughts, and feelings is critical to getting in-depth responses. It is also
critical to obtaining valid data, or information about what the interviewees truly
think (Pareek & Rao, 1980). In this respect, considerations of cultural norms and
communication styles are important to consider in delivering interviews, and it is
critical to work with local collaborators to create a culturally appropriate interview
manual. Indeed, this might also affect the structure of the interview questions. For
example, we discovered that the topic of honor and honor loss could be a very sen-
sitive subject, and thus we asked those questions toward the end of the interview.
In addition to the content of the interviews, there are other factors that can
affect the interview process, including the gender and background of the inter-
viewer, relationships between the interviewer and interviewee, and locations of
the interviews (Pareek & Rao, 1980). In cross-cultural projects, these issues should
be considered according to the norms of the local culture (Liamputtong, 2010).
Confidentiality is also a primary concern for many participants. Like other types
of research, participants should be provided with full informed consent, and con-
fidentiality of their identity should be ensured. In our case, we discussed at length
with our participants whether they were comfortable with us tape-recording the
interviews before doing so. Finally, it is also important to consider local issues in
the compensation given to interviewers for participating in the research. In some
cultures, monetary compensation might not be appropriate or even may be com-
pletely unacceptable. For example, for our interviews in Iraq, participants received
blankets as a gift for participating because receiving money for such a task was
considered insulting to the participants. In all, it is critical to work with local col-
laborators to design and implement the interview protocol.
Data Processing
After the completion of the interviews, we next needed to prepare the data for
analysis. Unlike quantitative research, this is a highly involved process. The
audio-taped interviews were first transcribed into text in the local language, then
responses were extracted from the transcripts, and finally they were translated
into English. In our case, we were dealing with more than 400 interviews in four
different languages and five Arabic dialects. Because each local collaborators pro-
cess their own data, we first established a standardized extraction procedure, then

331
supplemented it with regular online and phone meetings to align the process
across teams. In the following pages, we describe some of the lessons we learned
during this process.
Transcription Considerations
In qualitative research, recorded interviews are typically transcribed into text for
further analysis, and the best practice is to transcribe the interview recordings in
the original language before translating to another language (Liamputtong, 2010).
For our project, transcription was completed by local collaborators, which helped
us retain as much of the cultural-specific meanings as possible. One lesson that
we have learned from working with collaborators across different countries in this
process is the importance of standardizing transcription instructions and prac-
tices. For example, we later learned that we should determine at the beginning
whether to include nonfluencies such as “um,” “uh,” and “ah,” for example in the
transcriptions. Although the use of nonfluencies might not appear to be impor-
tant to the understanding of an interviewee’s responses, nonfluencies could be
an indicator of psychological processes (e.g., uncertainty), which could become
important for our later discussion of text analysis. Another issue to be resolved
is the consistency with which the transcribers used punctuation. We had some
transcripts, for example, that lacked appropriate punctuation to indicate pauses
and fragments. This lack of clarity increases difficulty in comprehending the tran-
scripts and also affected translation time and accuracy. As yet another example,
with the different transcription styles and dialects in our project, translators who
work later on the transcripts can also experience difficulty in deciphering an unfa-
miliar word as a result of typographical errors. For example, we were particularly
confused by the word hissa “‫ةسه‬,” which in almost all dictionaries means “noth-
ing.” However, when sounding it out and resorting to other techniques such
as Web searches, we discovered that this was shorthand for hai-as-saa’a, which
means “at that time.” It had been written as hissa “‫ ”ةسه‬as a form of shorthand for
the longer word. The straightforward solution is to check the original recording
even if it takes more time to complete the translation.
Finally, differences between literal versus intended meanings can introduce
confusion. As an example, an interviewee from one country was telling a story
about a girl who risked losing her, honor and the interviewee made a sarcastic
remark at the end, saying that “she wasn’t in a bad situation.” The statement might
be understood literally as that the girl in the story was not in a bad situation, but
actually the interviewee intended to say that the girl was in a lot of trouble. Written
transcripts that have no information regarding the intonation and nonverbal cues
are particularly vulnerable to such misunderstandings. There are strategies at the
transcription stage that can minimize these issues. Transcribers can make note
of possible alternative meanings and tone nuances in the transcriptions. They
can also fix or highlight any logical inconsistencies of the speakers. Specific to
Arabic language, it is also helpful to have transcribers translate colloquial terms
and phrases unique to a particular dialect into Modern Standard Arabic.

332
Data Extraction
We devote this section to describing a process we called extraction. The large
amount of data (with more than 400 interviews and 85 questions per interview)
presented many challenges for our later analyses. Although our interview pro-
tocols were structured, in reality, responses to a particular question might be
discussed at various times during the interview. For example, in one question,
we asked people to describe how a person gains honor; yet, in some cases, they
sometimes began to discuss how honor is lost, a question that was asked later
in the protocol. Put simply, unlike quantitative data, even structured interviews
are messy, with information to questions coming up at many different points of
the interview. To address this issue, our country collaborators extracted all inter-
viewees’ responses from transcripts and organized them by questions in a data-
base format. In the previous example regarding honor loss, we would take the
answer about honor loss and put it under the honor loss question even though it
came up during the honor gain question. In the end, we had an Excel database
of all responses by question for each construct across more than 400 interviews.
Extraction also provided an efficient way for us to be able to analyze the responses
to the questions without the need to peruse the entire collection of transcripts, as
invariably there were many responses that were not related directly to the inter-
views. It also provided a much better platform to compare responses and perform
text analysis, a topic we address later. Finally, extraction also reduced translation
cost by decreasing translation time so that only the extracted answers to the ques-
tions are translated.
Aside from all the advantages of extraction, it requires an additional step before
translation and analysis. Like transcription, this text preparation should be com-
pleted in the local language and it should follow a standardized protocol so that
members across a cross-cultural team complete this task in a consistent manner.
In our case, as a team, we decided that all responses would be extracted verbatim,
unless it was a meaningless utterance or a repeated response. Furthermore, any
responses that did not fit into existing question were extracted to an “other” cat-
egory. Finally, we conducted a reliability check on a subset of the interview tran-
scripts for each construct to ensure consistency across the team before extracting
all the data.
Translation
After we transcribed and extracted the responses, the next step was to translate the
responses from the local languages to English. During this stage, our project faced
unique challenges with interviews completed in many dialects of Arabic. As dis-
cussed earlier, the same words might yield different meanings depending on the
dialect. Arabic is a vast language, and the use of which spans two continents and
is subdivided by regional, national, and even municipal dialectal differences. It is
extremely rare to find a translator or even a team of translators with such diverse
language skills. In addition to consulting local collaborators, dialect dictionaries

333
and Internet websites that offer translations of common colloquial words into
Modern Standard Arabic can be helpful resources. We also used Web searches
for unfamiliar words and used the results to inform possible meanings of the
responses. In all, notes taken during transcription and extraction also helped the
translators to understand the context of the interview and to increase the ease and
accuracy of translations.
There are other practices we developed to facilitate the translation process.
Having the audio-recordings of the interviews available to the translators was
helpful for the translators to check unclear segments of the interviews. Building
in translation review time is also critical. Because we encouraged translators to try
to stick to the original transcription as much as possible, word choice, sentence,
or syntax structure in the translated text might sound awkward or confusing at
first. Rereading the translated text provided opportunities to see new perspectives
and fix any unclear wording and phrasing. If possible, it is very helpful to have
another bilingual person review the translation to ensure the meanings are trans-
lated appropriately. Finally, keeping the original language and translation side by
side in the storage of data is also helpful for review and later analysis.
Data Analysis
One of the main advantages of using qualitative interviews in cross-cultural
research is the ability to discover unknown dimensions and processes through
in-depth analysis. The data generated from our interviews based on the ASC
approach provided many social narratives that could be used to facilitate a deeper
understanding of the psychological motivation evoked by cultural value-laden con-
structs. Our purpose of the research was to use these data to help generate psy-
chological theories of these cultural value-laden constructs that we can use later to
compare across cultures with survey methods.
To maximize our understanding of the responses, our team tested various ana-
lytical methods to inform our theory formulation. In the remaining part of this
chapter, we illustrate a grounded theory approach that we used to discover univer-
sal and cultural-specific dimensions of the constructs using honor (Sharaf, Irdh)
as an example. We then describe how we incorporated other methods, including
linguistic analyses, coding, and semantic network analysis to analyze the inter-
view data to validate and extend the theoretical framework. As we describe these
processes, we highlight the challenges we faced as well as potential solutions and
considerations for other similar projects.
A Grounded Theory Approach

To construct theories about the meaning and social processes of honor in the
Middle East, it is important for us to examine the data devoid of any preconceived
notions of what honor should be. The grounded theory method provides an ana-
lytical approach to discovering universal and cultural specific elements of the con-
structs examined during the interviews. Grounded theory is an inductive method

334
of generating theories from systematic research. Since the publication of The

Discovery of Grounded Theory by Glaser and Strauss (1967), researchers from vari-
ous disciplines have used this method and different approaches to do grounded
theory (Bryant & Charmaz, 2007). Researchers can read the SAGE Handbook of
Grounded Theory (Bryant & Charmaz, 2007) for more in-depth discussions of the
history and development of the grounded theory method. Although researchers
have different views about the influence of researchers’ interpretations, they all
agree fundamentally that this method allows the data to “tell the story” instead
of imposing an existing knowledge framework onto the data (Charmaz, 2006).
As such, researchers can explore various meanings and underlying processes of
interested phenomena. The goal here is not to test a priori hypotheses and gen-
eralize the findings to a population, but to explore and discover new themes and
processes that emerge from the data free from preexisting assumptions.
Theoretically, the grounded theory method resonates with what we aim to
achieve with the data—namely, to understand how individuals make meaning of
the constructs, and their antecedents and consequences. In our analysis of honor,
for example, we developed a model of honor gain and maintenance by organizing
and sorting codes generated across teams. We created a three-phase procedure
based on the coding process of grounded theory. During the first phase, we asked
collaborators from each country to read and generate “codes” or themes from the
interview responses of their own country. The codes are “close” to the data at an
appropriate level of interpretation according to the questions. This is similar to
the substantive coding process of grounded theory, during which researchers read
through the interviews and take notes about the meanings and processes through
constant comparisons among responses.
One of the challenges for us is to calibrate the appropriate level of abstrac-
tion across collaborators during this process, because it could vary from ques-
tion to question, and some questions have more layers of abstraction than others.
For questions that ask for the meaning of a construct directly (e.g., What is the
meaning of honor?), codes are likely to mirror what the interviewees state explic-
itly (e.g., morality, purity, chastity, family reputation). However, some degree of
abstraction may be needed when the meaning is conveyed through illustrations of
behavior. Consider this response to how honor might be insulted or lost: “Honor
can be insulted if I’m walking on the street and someone pulls my scarf or grabs
my hand.” Naming this under a code called “insult” might be too general and we
might miss important details. However, naming this under a code called “physical
contact” or “walking on the street” is too piecemeal or concrete that it fails to cap-
ture the important meaning. An appropriate level of abstraction for this response
could be “unwanted physical attention.” To facilitate mutual understanding on the
generated codes, we asked country collaborators also to provide explanations and
examples directly from the transcripts for their codes.
It was very important to us that our country collaborators were involved closely
in this coding process because misinterpretation might result from lack of local
cultural knowledge. To illustrate such importance, we asked the American team
to complete this process on the English translation of responses from the Middle

335
East interviews to examine whether there was convergence in themes noted by the
Middle East collaborators. For the construct of honor, for example, 1,769 codes (an
average of 103 codes per question) were generated across the team. Not surpris-
ingly, clearly described and explained responses generally yielded greater conver-
gence between codes generated by the local Middle East collaborators and those
generated by American collaborators. The more difficult cases were responses in
which interviewees used a very specific example without much elaboration, or
in which the essence of the example was related highly to local cultural knowl-
edge. For example, in one interview from the Middle East, a female respondent
described how she maintained her honor by saying, “Firstly, a person distances
himself from any suspicions and bad places. I will tell you that it goes so far that
sometimes when my brothers visit me and they stay in a hotel I am not able to visit
that hotel. Entering the lobby and not even doing anything and I expose myself
to suspicions.” From the American perspective, it was not clear why she would
avoid going to a hotel, especially when her brother was staying there, or to what
extent going to a hotel would cause suspicion. Our collaborator later explained
that, in the Middle East, women going to hotels are often thought to be associated
with prostitution. Therefore, even by being seen in a hotel lobby or near a hotel,
it could raise suspicion of one’s purity. As this example illustrates, local cultural
knowledge is critical to understanding the implications of the codes and examples
described in the interviews, which are often understood implicitly between the
interviewer and interviewee.
The second phase of our code generation process involved sorting and orga-
nizing the codes from all country collaborators at a conceptual level. This pro-
cess also marked the beginning of theoretical coding during which we begun
to establish relationships between conceptual categories and to identify key
psychological processes. Consistent with earlier writings on honor (e.g., Pitt-
Rivers, 1966), the interviews illustrated that honor is “the value of a person in
his own eyes, but also in the eyes of his society” (p. 21). It is also a commodity—a
valued possession that can be felt, claimed, and ultimately which must be paid or
recognized by others (Pitt-River, 1966). Our framework and interview analyses
expanded this earlier work and illustrated that one’s sense of felt honor (e.g.,
self-esteem, self -respect, pride) translates into honor claimed through two inter-
related self-regulatory processes: (a) the promotion of honorable behavior (e.g.,
projecting an image of an honorable self by measuring up to well-defined social
norms and obligations, such as codes of loyalty, honesty, fulfilling obligations,
steadfastness) and (b) the prevention of dishonorable behavior (e.g., by avoiding
harmful or suspicious situations and by projecting a sense of strength and
bravery so as not to appear vulnerable or weak [men], and/or projecting a sense
of modesty [women] to avoid inviting honor violations. The interviews also
illustrated there are numerous ways in which honor can be lost or stolen. Honor
can be given away through one’s own social errors and actions, including mis-
conduct (breaking the law, participating in sexual mistreatment, failing to meet
social–moral obligations, and committing slander, among other behaviors),
acts that are careless (e.g., being in suspicious places), and acts that signal one

336
is weak (e.g., failing to respond to assaults on honor). Honor can also be easily
stolen by others’ actions (such as when one is the victim of slander, gossip, and
insults, or being exposed or assaulted). The results also illustrated that honor
loss affects a wide range of cognitive, emotional, and behavioral consequences.
In all, as a result of this data-driven theory development, we built a grounded
theory of honor gain, honor maintenance, and honor loss, each of which had
numerous codes that comprised them. This also resulted in a coding manual
that entailed a detail documentation of the codes and conceptual categories
of the codes for each interview question, which we used in the third phase of
validation coding.
The third phase of our qualitative analysis involved validation coding, which is
a common method to validate themes observed during interviews or other types of
qualitative data. To substantiate the applicability of the identified codes or themes,
researchers are asked to apply a set of predetermined codes to the interview
responses following specific guidelines. We learned some essential lessons from
our pilot coding regarding the challenges of achieving coding reliability. We iden-
tified three factors that influenced our coding reliability: quality and clarity of the
data, clarity and specificity of codes, and coders’ understanding of the responses.
At this stage of the research, there is little we could do to the quality and clarity of
the data unless the confusion stemmed from translations. Ideally, we would have
probed what exactly interviewees meant when they said honor is demonstrated
by “good behavior,” “positive interaction,” or “being in harmony with yourself.”
Having a standardized procedure to deal with these ambiguous terms can reduce
its impact on coding reliability. For example, we decided that coders should set
ambiguous responses aside rather than fitting them to a specific code.
We also found that the level of specificity of the codes affected reliability. If
the codes are too abstract and general, they lack distinctiveness and potentially
miss culturally specific behavior or meanings. However, if the codes are too spe-
cific, they might be redundant (e.g., talking nicely vs. being polite). To strike a
balance on code specificity, we implemented a coding system that incorporated
specificity and inclusivity, which allowed us to have the flexibility to identify ideas
and behavior close to the data without losing sight of the conceptual ideas. This
system involves having complex codes or themes to have two level of code catego-
ries: one at a higher level that includes a set of sub-codes that are more specific.
To illustrate, we have a relatively higher code category called prosocial orientation
that defines gaining honor through personal traits and behavior that contributes
to positive social interaction. This code category includes a number of specific
subcodes, such as positive communication (e.g., talking kindly to others, being
polite), being cooperative (e.g., willing to work with others), maintaining good
relationships with others, showing care for others, and so forth. Based on our
pilot coding, we found that coders’ reliability at the high code category level (i.e.,
the higher order prosocial orientation codes) was much greater than at the specific
code level.
The remaining factor that contributed to coders’ reliability was coders’ under-
standing of the interview responses. We found that coding completed by bilingual

337
coders familiar with local or regional cultures tended to be more reliable. This
conclusion is based empirically on a small experiment we did comparing the
agreement of coding completed by nonbilingual American coders and bilingual
coders (fluent in English and Arabic) on the Arabic interviews. In all, there was
a clear advantage of having bilingual coders familiar with the local language and
culture. This familiarity affects understanding of interviewees’ intended mes-
sages, and bilingual coders also help communicate nuances to researchers. An
example of this is the understanding of the Arabic term Amana (‫)ةنامألا‬, which is
translated as “sincerity” and it encompasses the meaning of loyalty, honesty, and
trustworthiness. Another example a Lebanon respondent said in response to how
a person demonstrates honor was, “Being in harmony with yourself and acting
with morals and a conscience.” The nonbilingual American coders coded the first
part of the answer as “respecting oneself,” but the bilingual coders suggested that
it actually means adhering to one’s ethics and moral values.
Learning from our pilot coding, we developed a few general practices that
helped increase coders’ reliability. Extensive coding training, explicit coding rules
and guidelines, and pilot coding are important ways to align researchers’ and cod-
ers’ perspectives on the codes and coding procedure. Pilot coding is highly useful
to spot potential misunderstandings. Examples and counterexamples of codes are
helpful to clarify what specific codes mean and how they differ from one another.
Having regular and open communication between the researchers and the cod-
ers is also helpful for developing a shared understanding of the coding process.
When there is a large amount of coding, an intermediate reliability check can
also be implemented to realign different perspectives or add clarifications. Our
coding procedure encouraged coders to suggest new codes if they could not find
any fitting code, and coders could also opt out from applying a code when they felt
unsure about a response segment. Finally, we also found that coders with research
experience and an understanding of the research process were also more likely to
bring up questions and problematic areas. In all, these recommendations aim at
aligning researchers’ and coders’ perspectives.
Other Analytical Strategies

With technological advancements, we are now equipped with tools that provide
alternative ways to analyze qualitative data beyond coding. We mention briefly two
kinds of text analyses we are currently using and then discuss the advantages and
disadvantages of them in cross-cultural analyses.
Linguistic Analyses
Jaime Pennebaker and colleagues (Pennebaker, Mehl, & Niederhoffer, 2003) devel-
oped a text analysis approach to tap psychological processes through analyzing the
way people use language in a variety of contexts (e.g., newspaper articles, liter-
ary writing, personal diaries, and speech transcripts). A large body of empirical
studies has now shown that in addition to what people say explicitly, how people

338
convey ideas through writing and speech also reveals underlying psychological
dimensions and processes (e.g., affect, cognition). Pennebaker and his research
team developed a text analysis program called the Linguistic Inquiry and Word
Count (LIWC) that processes a large amount of text and sorts words according
to their well-validated linguistic and psychological categories, including emotion,
cognition, and time, among many others (Pennebaker, Chung, Ireland, Gonzales,
& Booth, 2007). Although its use for cross-cultural analyses is rare, previous
research has shown, for example, the use of first and second-person pronouns
(e.g., I, you, me) reflects cultural differences on the dimension of individualism
and collectivism (Kashima & Kashima, 1998; Na & Choi, 2009). In the context
of our interview project, the LIWC analyses illuminated new insights that we did
not find with the coding. For example, we found that Middle Eastern interview-
ees discussed honor with more emotional words whereas American interviewees
discussed honor with more cognitive words. Providing convergent evidence of the
two regulatory approaches outlined in the process model of honor gain and main-
tenance mentioned earlier, we also found that words related to inhibition (e.g.,
abstain, control, careful, discipline) were used more often in the discussion of
honor in the Middle Eastern interviews than in the American interviews.
Aside from the interesting results LIWC unveils, there are some limitations
to this technique. One challenge for using the program to perform cross-cultural
analysis is that LIWC is developed primarily based on English lexicon and lan-
guage use. Although LIWC is available in other languages (e.g., Spanish, German,
Dutch, French, Arabic), we found that the word categories, at least for the Arabic
version, were not as developed as the English version. For example, we used the
beta Arabic version to analyze our Arabic transcripts and it captured less than 50%
of words in our interviews. That said, it limited our text analysis only to using the
English translation. To minimize confounding mere language differences with
psychological differences, we also limited our analyses to substantive word catego-
ries such as affect, cognitive mechanism, religion, and achievement, and avoided
word categories of linguistic structure. For example, comparisons of linguistic
dimensions such as pronouns, articles, and tenses may not be useful because of
the differences in grammar and syntactic structure.
We also recommend researchers who use LIWC to review the word list in the
category of interest to determine whether it matches researchers’ conceptual-
ization. For example, the achievement category in the existing LIWC dictionary
includes words related to lack of achievement, such as failure, mistakes, and quit.
In the context of our interviews, achievement words are related to honor gain and
maintenance, but lack of achievement words are related to honor loss. Therefore,
we would want to separate them in our analysis. Finally, the reliability of this
type of text analysis is highly contingent on the quality of the text input itself. We
mentioned earlier the transcribing variations of including or excluding nonfluen-
cies. Such variations limit our ability to compare such paralinguistic characteris-
tic, which could be an indicator of certainty or accessibility of associated concepts.
Researchers should also be aware that the text needs to be reviewed and edited
accordingly to ensure proper processing (e.g., marking filler words such as like
in colloquial English as opposed to like as liking or like as being similar to). If the

339
language is not supported by LIWC, simple word frequency could be an alterna-

tive tool to analyze word use and generate word categories according to theoretical
interests. There are word frequency count software packages available for produc-
ing word frequency for different languages.
Using Interview Data to Create New Linguistic Dictionaries

Given that existing linguistic dictionaries are based primarily on Western
sources and have limited ability to tap into constructs relevant in the Middle
East, we used our own interview data to develop a new multidimensional dic-
tionary which can now be used to diagnose honor talk in text data such as news-
papers, blogs, social media, and the like (Gelfand et al., 2015). In order to do
so, we first pooled all Arabic responses to the honor questions together and
then conducted a word frequency count to generate a comprehensive word list
on honor. We removed commonly used non-essential words such as pronouns
(e.g., they, him, we, you), positions (e.g. of, to, in), and conjunctions (e.g., and,
but, or). Additionally two Arabic speaking researchers each independently
removed the words they deemed unnecessary and not related to honor, and
if there were disagreements, they were discussed and resolved. Thereafter the
list of honor-related words were carefully translated back into English. A paral-
lel process was also conducted with responses from other non-Arabic speak-
ing countries with English texts and the sets of honor-related words were then
combined.
The final list of honor-related words was further categorized into theoreti-
cal groups based on the honor model discussed above and prevailing literature.
These categories included honor gain including sub-categories of moral integrity
(words related to integrity, morals, and commitment); status (words related to sta-
tus gains, prestige, social standing); and achievement (words related to work and
education); honor protection and avoiding negative circumstances (including sub-
categories of word related to prevention, public image, strength/bravery); honor
loss (including sub-categories of harm, aggression, and wrongdoing); faith (words
related to religion, beliefs, and faith), sex/body, and etiquette (appropriateness),
along with numerous contexts in which honor talk is relevant (e.g., military, fam-
ily, tribes, among others).
The honor dictionary was validated with numerous sources based on estab-
lished theories of honor (e.g., Nisbett & Cohen, 1996; Cohen, Nisbett, Bowdle, &
Schwarz, 1996). For example, we found that there is a greater degree of honor talk
in newspaper articles in the historic and contemporary U.S. South than the North,
in the constitutional texts of countries that are thought to have a stronger honor
orientations than those that do not, and in social media discussions of Palestinian
writers than Israelis writers. Honor talk (e.g., moral integrity, prevention) also pre-
dicted better negotiated agreements in Egypt. In all, our cross-cultural interviews
provided useful input into new linguistic tools that can now be used to analyze
honor talk in any text. More details of of the honor dictionary development and
validation results can be found in Gelfand et al. (2015) and the dictionary can be
obtained from the 2nd author.

340
Semantic Network Analysis

In addition to the linguistic analyses, the interview data we collected also enable
us to use a new technique—network analysis—to examine “mental models” of
honor. Network analysis allows us to examine which concepts and ideas peo-
ple associated with honor. It also allowed us to ask questions such as: How are
these concepts organized? How do the concept structures differ across countries?
We addressed these questions through generating semantic networks based on
thoughts and ideas people associate with honor (Carley & Kaufer, 1993) based
on the one question in our interview protocol: What does honor mean to you?
As with linguistic word count analysis, there are important challenges in doing
network analysis in cross-cultural research. To generate the concept networks,
we had to develop a standardized procedure first to reduce the interview data to
create these networks. Colloquial responses to the question “What does honor
mean to you” were first reduced into statements of ideas. Each statement con-
tains key concepts discussed. For example, responses such as, “Honor is purity;
it also suggests masculinity” were reduced to “honor purity/honor masculinity.”
In doing so, we retained the original wording as much as possible. Moreover, to
allow meaningful cross-cultural comparison, we developed a procedure to create
concept categories that could be applied across countries. This method groups
words and phrases by concept categories that are created inductively from the
data. For example, mother, father, sisters, and brothers were grouped under fam-
ily. Likewise, bravery, not run away, vigor, strength, not coward, and courage,
among other phrases, were grouped under bravery/strength. Words/phrases ini-
tially categorized under honor loss were divided further into the subcategories
of wrongdoing, norm violations, and physical aggression to capture nuances of the
notion of honor loss. Using Automap1, we then coded all concepts from the honor
interviews with the original coding categories. Thus, the networks maintained
the concepts from the original interviews and also had a set of common nodes
to compare them across countries. After this data preparation stage, concept net-
works were generated for each individual respondent and were then aggregated
by country.
In constructing these networks, we treated concepts mentioned during the
interviews as nodes in a network of associations; when two concepts were men-
tioned in conjunction (i.e., in sequence), they were considered to be linked. The
number of times a concept is mentioned can be represented by the frequency of
the node; the strength of the association is indicated by the frequency by which a
pair of concepts was mentioned relatively closely temporally. In its simplest form,
if a person states that honor is honesty, then the construct of honor is linked
conceptually to honesty. In network analysis, various mathematical indices can
be used to analyze the structure of a semantic network as well as the centrality
of concepts. For example, one could interpret that a node/concept that has many
links with other nodes/concepts is a central concept in relation to honor. Interested
readers who wish to use this type of analysis for their qualitative data can read the
work by Carley and Kaufer (1993), Collins and Loftus (1975), Shapiro (1979), Sowa
(1991), and Sieck (2010).

341
Examples of semantic maps of honor from Jordan and the United States are
included in Appendix B. The size of the node (concept) reflects the number of
times the concept was mentioned. The bigger the node, the more times it was
mentioned. The width of the lines indicates the frequency with which the two
concepts were mentioned simultaneously. The thicker the link, the more times
the two concepts were connected, or appeared in the same statement. These net-
works reveal some interesting cross-country differences. For example, the concept
of women or women honor was important in Jordan (Figure B11.1), but it was not
mentioned at all in US word associations of honor (at least in the northern United
States, where the interviews took place, Figure B11.2). Another interesting differ-
ence is the emphasis on protection (e.g., honor is something to protect and safe-
guard). This theme was prominent in Jordan but not at all in the United States
The emphasis on protection is consistent with the importance of avoiding or pre-
venting honor loss—a theme delineated in our grounded theory model of honor
gain and maintenance. Similarly, concepts regarding honor loss were much more
prominent in Jordan than the United States. This suggests that the possibility of
losing or giving up honor is highly accessible when honor is brought to mind in
the Middle East. On the other hand, the concept of respect has a very central role in
defining honor in the US data. As shown in the concept network of the US data,
respect is highly centralized—meaning, this concept category was much more con-
nected to other concept categories in the semantic network.
In sum, this section offers some insights about other analytical techniques
researchers may consider when analyzing qualitative data. It is important to note
that there is very little work on applying these techniques to study cultural differ-
ences, and further research is much needed to examine the artifacts they might
create. Nevertheless, recent research is beginning to recognize the importance
and potential contribution of network analysis in understanding cultural con-
strual of the social world (Sieck, 2010). This is an exciting new direction in cross-
cultural research.
Conclusions
Although interviewing and qualitative data analysis are widely used methods in
many social science disciplines, application to cross-cultural research has been
limited. Our current project is one of the first large-scale research projects to use
such methods to study key cultural constructs in the Middle East. In closing, we
describe two final reflections from the “trenches” when conducting such research.
Although standardization of protocols and procedures was important, it is equally
important to be flexible in accommodating local norms, especially during the
design and data collection stage. We had to strike a balance constantly between
standardizing across a cross-cultural team and maintaining cultural sensitivity.
Another important point in the success of conducting such research rests on the
collaborative effort with local researchers. Effective team communication is highly
important to conduct such research efficiently and accurately. We hope this chap-
ter has brought the various issues and challenges of conducting cross-cultural

342
interviews to light, and facilitates future efforts in doing this type of research in
the Middle East and North Africa.
Acknowledgment
This work was funded in part by the US Army Research Laboratory and the
US Army Research Office (grant W911NF- 08-
1-
0144) and Federal Bureau of
Investigation contract J-FBI-10-009 (subaward 2014054568).
Note
1. Automap is a text mining tool developed by the Center for Computational
Analysis of Social and Organizational System (CASOS) at Carnegie Mellon University.
http://www.casos.cs.cmu.edu/index.php
References
Brislin, R. (1980). Translation and content analysis of oral and written material. In H.
C. Trandis & J. W. Berry (Eds.). Handbook of cross-cultural psychology: Methodology
(Vol. 2, pp. 389–444). Boston, MA: Ally and Bacon.
Bryant, A., & Charmaz, K. (Eds.). (2007). The SAGE handbook of grounded theory.
London: Sage.
Carley, K. M., & Kaufer D. S. (1993). Semantic connectivity: An approach for analyz-
ing symbols in semantic networks. Communication Theory, 3(3), 183–213.
Charmaz, K. (2006). Constructing grounded theory: A practical guide through quali-
tative analysis (Introducing Qualitative Methods Series).
Cohen, D., Nisbett, R. E., Bowdle, B. F., & Schwarz, N. (1996). Insult, aggression,
and the southern culture of honor: An “experimental ethnography.” Journal of
Personality and Social Psychology, 70(5), 945–960.
Collins, A. M., & Loftus, E. F. (1975). A spreading–activation theory of semantic pro-
cessing. Psychological Review, 82(6), 407–428.
Gelfand, M. J., Raver, J. L., & Holcombe Ehrhart, K. (2002). Methodological issues
in cross-cultural organizational research. In S. Rogelberg (Ed.). Handbook of
industrial and organizational psychology research methods (pp. 216–241). New York,
NY: Blackwell.
Gelfand, M. J., Severance, L., Lee, T., Bruss, C. B., Lun, J., Abdel-Latif, A. H., … &
Moustafa Ahmed, S. (2015). Culture and getting to yes: The linguistic signature
of creative agreements in the United States and Egypt. Journal of Organizational
Behavior, 36(7), 967–989.
Glaser, B. G., & Strauss, A. L. (1967). The discovery of grounded theory. Chicago,
IL: Aldine.
Kashima, E. S., & Kashima, Y. (1998). Culture and language: The case of cultural
dimensions and personal pronoun use. Journal of Cross-Cultural Psychology, 29(3),
461–486.

343
Liamputtong, P. (2010). Performing qualitative cross- cultural research. New York:

Cambridge University Press.
Na, J., & Choi, I. (2009). Culture and first-person pronouns. Personality and Social
Psychology Bulletin, 35(11), 1492–1499.
Nisbett, R. E., & Cohen, D. (1996). Culture of honor: The psychology of violence in the
South. Boulder, CO: Westview Press.
Pareek, U., & Rao, V. T. (1980). Cross-cultural surveys and interviewing. In H. C.
Trandis & J. W. Berry (Eds.). Handbook of cross-cultural psychology: Methodology
(Vol. 2, pp. 127–180). Boston, MA: Ally and Bacon.
Pennebaker, J. W., Chung, C. K., Ireland, M., Gonzales, A., & Booth, R. J. (2007).
The development and psychometric properties of LIWC2007. Austin, TX: LIWC.net.
Pennebaker, J. W., Mehl, M. R., & Niederhoffer, K. G. (2003). Psychological aspects of
natural language use: Our words, our selves. Annual Review of Psychology, 54, 547–577.
Pitt-River, J. (1966). Honour and status. In J. G. Peristiany (Ed.). Honour and shame
(pp. 19–77). London: Weidenfeld and Nicholson.
Shapiro, S. C. (1979). The SNePS semantic network processing system. In N. V.
Findler (Ed.). Associative networks: Representation and use of knowledge by computers
(pp. 263–315). New York: Academic Press.
Sieck, W. R. (2010). Cultural network analysis: Method and application. In D.
Schmorrow & D. Nicholson (Eds.). Advances in cross-cultural decision making, (pp.
260–269). Boca Raton, FL: CRC Press/Taylor & Francis.
Sowa, J. F. (Ed.). (1991). Principles of semantic networks: Explorations in the representa-
tion of knowledge. San Mateo, CA: Morgan Kaufmann.
Tapp, J. L., Kelman, H., Triandis, H. C., Wrightsman, L. S., & Coelho, G. V. (1974).
Continuing concerns with cross-cultural ethics: A report. International Journal of
Psychology, 9, 231–249.
Triandis, H. C. (1972). The analysis of subjective culture. Oxford, UK: Wiley-Interscience.

344
APPENDIX A Sample Interview Protocol
Q1. Fate/control
1A. What words come to mind when we say fate?
How much do you believe that important things that happen in your life are in
your control?
Please answer on a scale from 1 to 5, where 1 is not at all, 3 is somewhat, and 5 is
very much.
Who or what do you think controls the events in your life?
Do you feel you are able to influence important decisions in your own life?
(Probe: Personal, interpersonal, and sociopolitical domains of life—for example, education
and work, relationships, and larger issues affecting your life such as how the economy is run
or other regional and country wide decisions).
Do you think people in your country feel they are able to influence important deci-
sions in their own life? How important is that for them?
very much.
1B. Can you tell me about the situations that occur because of fate (qadar) and
why?
(Probe: ________ situations are subject to fate.)
Which situations are not as attributable to fate?
(Probe: _________ situations are not subject to fate.)
Do you think a person can change his or her fate? Why or why not?
1C. Would you say that you focus more on the past, the present, or the future?
How important is it to plan for the future? Why or why not?
1D. Do you believe in fate? Do people in your country believe in fate?
(Probe: Meaning and importance of fate)
Q2. Public image and face

2A. What does Wajh mean to you?
What words come to mind when we say Wajh?
How important is it to you to protect your Wajh so you do not lose Wajh?
very much.
How important is it to people in your country not to lose face? (Probe: Associations
and importance of Wajh to self and in the country)

345

very much.
2B. Wajh can be lost. Can you give me some examples of ways in which Wajh can
be lost, from relatively mild examples, to more moderate, and more extreme ways
in which Wajh can be lost? You can use personal examples or those that you have
heard of, in work and nonwork contexts. (Probe: Antecedents of loss of Wajh; Wajh
would be lost if ____ __; keep mild, moderate, and extreme examples classified.)
What happens when face is lost? What kind of feelings do you or others experi-
ence in these situations? (Probe: The emotions that are felt: When my Wajh is lost,
I or others feel _______.)
What do you or others do in response to the loss of Wajh?
(Probe: The behavioral reactions of loss of Wajh: When my Wajh is lost, I or others
_______.)
2C. Does it make a difference if the person who causes your Wajh to be lost is
someone you know well or is someone you do not know well? (Probe: In-group–
outgroup effects: How is the loss of face experienced if it occurs within one’s family or
group versus someone you do not know well?)
Does your Wajh loss affect others? If yes, whom? (Probe: How contagious is face
loss? How interrelated is face?)
Does others’ Wajh loss affect you? If so, whose Wajh is most important to you? In
what ways does their Wajh affect you personally if it is lost? (Probe: How contagious
is face loss? How interrelated is face?)
Q3. Modesty
3A. What does modesty mean to you?
What words come to mind when we say modesty?
(Probe: Associations with modesty.)
How important is being modest to you?
very much.
How important is being modest to people in your country? (Probe: Importance of
modesty for oneself and in the country at large.)
very much.
3B. How does one demonstrate one is modest in everyday contexts, in work, and
nonwork contexts? How might this vary depending if you are of high status or of
low status?
(Probe: Behaviors that are modest; probe differences depending on status.)

346
3C. Sometimes we expect people to act in a modest way, but they do not. Can you
give me an example of situations when a person was not modest even though you
were expecting them to be?
(Probe: Antecedents of modesty violations: A person acts immodestly when they _____
__.)
What happens when a person is immodest in a situation when they should
be modest? What kind of feelings do you or others experience in these
situations?
(Probe: Emotional consequences of modesty violations: When a person is
immodest, I or others feel _______.)
What do you or others do in response to a person being immodest?
(Probe: The behavioral reactions of modesty violations; When a person is
immodest, I or others _______.)
3D. Does it make a difference if the person who acts immodestly is a family mem-
ber/close friend or is someone you do not know well?
(Probe: In-group–outgroup effects: How are modesty violations experienced if it
occurs within ones group versus someone you do not know well?)
3E. When you act modestly or immodestly, does that behavior reflect on other
people in your life? Who? Does their ability to act modestly in the appropriate situ-
ation reflect on you? Can you give examples? (Probe: How contagious is modesty?
How interrelated is modesty and among whom?)
Q4. Respect/dignity
4A. What does dignity mean to you?
What words come to mind when we say dignity?
(Probe: Associations with dignity.)
How important is maintaining your dignity to you?
very much.
How important is maintaining your dignity to people in your country? (Probe:
Importance of dignity to oneself and to others in the country.)
very much.
4B. Some people believe that all human beings have a certain worth, no matter
what their circumstances and no matter what other people think of them. Do you
agree? If so, what is the basis for this worth? (Probe: What makes a person have
worth?)

347
Do you think that others in your country would agree? What would they think is
the basis of this worth?
4C. Can this worth be taken away by other people? Can you give an example? Can
people give up this worth themselves so they no longer have it? Can you give an
example? (Probe: Associations with dignity loss. Ask for as many examples as possible.)
Do people in your country think worth can be taken away?
What happens if someone’s worth is taken away? (Probe: Consequences of worth
being taken away.)
If worth is taken away, does it affect others? (Probe: The extent to which dignity loss
is contagious and affects others: Who does it affect?)
Can dignity be regained? How might it be regained?
4D. Sometimes we expect people to act in a respectful manner, but they do not.
Can you give me some examples of ways in which people can behave in a disre-
spectful manner?
You can use personal examples or those that you have heard of, in work and non-
work contexts.
(Probe: Antecedents of respect violations: A person acts disrespectfully when they
_______.)
What happens when a person is disrespectful in a situation when they should be
respectful? What kind of feelings do you or others experience in these situations
or what do you or others do in these situations?
(Probe: Consequences of respect violations: When a person is disrespectful, I or others
feel _______.)
4E. Does it make a difference if the person who does not show respect is a family
member/close friend or is someone you do not know well? (Probe: In-group–out-
group effects: How is being disrespectful experienced if it occurs within one’s group versus
someone you do not know well?)
4F. Is there a difference between showing disrespect to a woman compared with
showing disrespect to a man?
Q5. General Basic Values

5A. What are some core values and personal virtues or characteristics that you
think are important to have as guiding principles in your life?
What are some core values and personal virtues or characteristics that people in
your country think are important to have as guiding principles in life?
5B. What values and virtues would you recommend that children be raised to
have? (Probe: Children should be raised to _______.) What would you recommend
that one raise one’s children to avoid? (Probe: Children should not _______.)

348
5C. Are there any values that need to be taught to girls and boys differently?
Do you think that, in general, men and women should have different roles in
society?
Do people in your country believe that men and women should have different
roles in society? (Probe: Different roles at work, politics, the home, in leadership
positions.)
Q6. Sharaf and ird

6A. What does Sharaf mean to you?
What words come to mind when we say Sharaf?
(Probe: Associations with Sharaf.)
How important is protecting your Sharaf to you?
very much.
How important is protecting Sharaf to people in your country? (Probe: Importance
of this to self and others in one’s country.)
very much.
6B. How does one demonstrate or prove one’s Sharaf or in everyday contexts, both
in work and nonwork settings? How might this vary depending if you are of high
status or of low status? (Probe: Behaviors that are honorable, differences depending on
status.)
6C. Sharaf can be insulted or threatened. Can you give me some examples of ways
in which Sharaf can be insulted or threatened—from relatively mild examples,
to more moderate, and to more extreme ways in which Sharaf can be insulted or
threatened? You can use personal examples or those you have heard of, in work
and nonwork contexts.
(Probe: Antecedents of Sharaf violations: Sharaf would be threatened/insulted if _____
__; keep mild, moderate, and more extreme examples classified.)
What happens when Sharaf is insulted or threatened? What kind of feelings do
you or others experience in these situations?
(Probe: Emotional consequences of Sharaf violations: When Sharaf is threat-
ened/insulted, I or others feel _______.)
What do you or others do in response to insults or threats to Sharaf?
(Probe: The behavioral reactions of Sharaf violations: When Sharaf is threat-
ened/insulted, I or others _______.)

349
6D. Does it make a difference if the person who threatens your Sharaf is someone
you know well, such as a family member, or is a someone you do not know very
well?
(Probe: In-group–outgroup effects: How are Sharaf violations experienced if it
occurs within the family versus with someone you do not know very well?)
Is your Sharaf related to the Sharaf of other people, and whom? (Probe: Family,
others.) How does something affecting your Sharaf affect the Sharaf of others? Can
you give an example?
(Probe: How contagious is Sharaf? How interrelated is Sharaf and among whom?)
Likewise does the loss of the Sharaf of others affect your Sharaf? Whose Sharaf is
most important to you? How does it affect you? Can you give an example?
6E. How is ird different than Sharaf, if at all? What are the ways in which ird is
threatened or insulted that haven’t been captured in the previous questions? What
happens when ird is threatened or insulted? (Probe: Examples of antecedents and
consequences as with previous questions.)

350
APPENDIX B Sample Semantic Network Maps of Honor
country
avoidance
modesty
marriage
self_esteem
public_
religion imagepos
dignity
self integritypos
encompassing
behavior
public_image protection
women/women_honor
morality_values
important
family/family_honor
work
men/men_honor wrong doing
non_family relationships honor_loss

societal_duties_
treatment to others loss obligations
general_positive physaggress
public_imageneg
change
women_honor_loss
Figure B11.1 Jordan.
important
avoidance
family_honor
work
body
normviolations
country
encompassing military
strength
public_imageneg self
loss
family honoring personal_duties_obligations
social_status respect religion

treatment_to_others
self_esteem
modesty general_positive
dignity public_imagepos
achievement
integritypos non-family relationships
public_image morality_values
Figure B11.2 The United States.

351
INDEX
Note: Page numbers followed by “f” and “t” refer to figures and tables, respectively.
Abu-Lughod, Laila, 138, 139–40 country fixed-effect model explaining

access to social capital, 212 engagement, 191–2, 191t
activism, political. See political national identity versus national
engagement pride, 83, 84t–9t, 90–1
affective bonds in family, 146, 153 nonelectoral political engagement,
affective dimension of 180t, 184–8
nationalism, 66, 69 social roots of nonelectoral political
Afghanistan, 32, 67, 326 engagement, 188–92, 188t
age. See also intergenerational territorial nationalism in, 68–9, 70
value change Ali, K.A., 140
basis of identity, 44, 44f, 50, 51t Amin, Ahmed Raafat, 252
gender equality values, 39, 39f, analysis of subjective culture (ASC)
47, 48t method, 326. See also subjective
liberalism values, 42, 42f culture analysis
national identity versus national analysis of variance (ANOVA), 223, 230
pride, 92 Anderson, B., 95n1
nationalist views and, 91–2 anomie, 207, 214, 221, 222t, 243
participation in revolution, 213 ANOVA (analysis of variance), 223, 230
political engagement, 188t, 189 antigovernment demonstrations,
secularism values, 47, 49t, 50 185, 186t
social individualism values, 45, 46t approach–avoidance conflicts, 290
trends in values, 44–51 Arab Barometer surveys, 254, 266n12
alcohol use, 118, 119t, 121t, 122 Arab developmental model of family
Algeria, 179, 203t change, 162–4
antigovernment demonstrations and Arab identity, 71–2
general strikes in, 185, 186t Arab Island, 72
attitudes toward political Islam, Arab League, 137
253–4, 257f, 259–60 Arab models of kinship, 146
351
352
Arab Spring, 205. See also autocratic Banna, Hasan al-, 73

recidivism; revolution makers; Ben Ali, Zine al-Abidine, 249, 251, 262
youth values during Arab Spring best practices of survey research,
collective action during, 272–3 295–319
electoral victories in Tunisia and challenges of conducting research in
Egypt, 254–59 Middle East, 297–8
failure of democracy after, 14–23 conducting research in conflict
Tunisia and Egypt during, 250–3 zones, 297–8
Arendt, H., 210, 211, 236n1 construct development, 305–6
ASC (analysis of subjective culture) creating indicators, 306–9
method, 326. See also subjective cross-national data entry, 318
culture analysis establishing cross-national research
attitudes, mobilizing ideas and. See teams, 298–9
mobilizing ideas and attitudes interviewing techniques, 316–8
attitudes toward political Islam, 249–69 language translation, 309–15
data and measures, 253–6 pretesting, 315–6
electoral victories in Tunisia and questionnaire design and layout, 315
Egypt, 254–7 sample design, 300–5
impact in other Arab bias, differential, 317
countries, 259–61 Bier, L., 140
item factor loadings, 269 biographical availability, 173, 175, 189,
after living under, 257–9 192, 195
questions going forward, 261–4 birth cohort versus life cycle effects in
Tunisia and Egypt during Arab value change, 7–11
spring, 250–3 blocked aspirations, 173, 192
authoritarian–deity factor, 244 Bouazizi, Mohamed, 205, 250–1, 266n4
authoritarian learning, 264 bourgeoisification of nationalist
authoritarian rule, sense of security and leaders, 70–1
submission to, 12 boycotts, 180t, 181
autocracy. See autocratic recidivism
autocratic networks, 280–1 Catterberg, G., 176–7
census data, 300–1
Bangladesh, 179, 203t Center for Strategic Studies (CSS),
antigovernment demonstrations and University of Jordan, 254
general strikes in, 185, 186t change in values. See intergenerational
country fixed-effect model explaining value change
engagement, 191–2, 191t chaotic attractor, 279
forms of nonelectoral political child qualities, 34, 35–6, 36f, 53n3
engagement, 180t China, 18, 115
index of nonelectoral political Christian Lebanese
engagement, 184–8 development ratings among, 120–2, 121t
national identity versus national morality ratings among, 118–20, 119t
pride, 83, 84t–9t, 90–1 Christian values, laws inspired on,
social roots of nonelectoral political 40–2, 40f
engagement, 188–92, 188t clash of civilizations model, 106
352 | INDEX
353
clash of moralities, 102, 106–7, coupling, 278

115, 124–6 cross-national survey research, 295–319
closed political systems, effect on challenges of conducting research in
political engagement, 177–8 Middle East, 297–8
coding process of grounded construct development, 305–6
theory, 334–7 creating indicators, 306–9
coherence in social systems, 275 cross-national data entry, 318
cohort analysis, intergenerational value establishing cross-national research
change, 7–11, 9f teams, 298–9
collapse of autocracy. See autocratic interviewing techniques, 316–8
recidivism language translation, 309–15
collective action during Arab pretesting, 315–6
Spring, 272–3. See also political questionnaire design and layout, 315
engagement; revolution makers sample design, 300–5
collective sovereignty, modalities of, cross-national variation in values, 33–44
62, 64–7. See also modalities of gender equality, 37–9
nationalism identity measurement, 43–4
colonial rule, rise of territorial liberalism index, 42
nationalism linked to, 68 secularism, 40–2
communicative power, mediums of, 228t social individualism, 34–7
correlates of participation, 216–7 CSS (Center for Strategic Studies),
descriptive statistics, 222t University of Jordan, 254
linked to participation in revolution, culture. See also subjective culture
230, 232–3 analysis
predictive variables of study, 243 disease, effect on values in, 10
role in revolutionary local cultural knowledge, importance
development, 211–2 in research, 334–5
compensation for interviewers, 330 shift from survival values to self-
computational model of autocratic expression values, 4–5, 14
recidivism. See autocratic recidivism tight versus loose, 10
Comte, Auguste, 104
confidentiality in research, 330 Daesh. See ISIL
conflict zones, conducting research Darwaza, Muhammad Izzat, 72
in, 297–8 data processing and analysis, subjective
consanguineous marriages, 148 culture analysis, 330–3
constitutional government, 72–3 decentering, 307
construct development Deeb, L., 127
cross-national survey research, 305–6 democracy. See also autocratic recidivism
subjective culture analysis, 326–7 failure after Arab spring, 14–23
contextual bases of political gender equality in Islamic world, 16–20
engagement, 176–8 modalities of nationalism and
Corduneanu-Huci, C., 171 view of, 79
cost of weddings, 150, 153 nonelectoral political engagement
country-level variables, 183 under, 187, 187f
coupled dynamical systems, 278 world image of, 15–6, 16f
Index | 353
354
demographics of revolution makers Dream of the Southerner (television

linkages between participation series), 139
support and, 229t, 233–4 dress, freedom of women to choose, 34,
measures included in study, 217, 35, 35f
222t, 246–7 dyads, synchronization in, 274, 278
PROP and MSR theories, 212–3 dynamical systems theory, 272, 274, 278
demonstrations, political, 180t, 181, 186t dysphoric emotions versus individual
development. See developmental efficacy, 209
idealism; development–morality
nexus; family life, developmental economic development
models of attitude changes due to, 20–1
development, socioeconomic, 133 effect on postmaterialist values, 10–1
developmental idealism, 105 family change as part of, 158–61, 160f
developmental models of family world system theory, 105–6
life, 134–6 economic inequality, effect on political
Middle East survey research program engagement
measure of, 305–6, 308 country-level variables, 183, 203t
secularism, 32 linkages, 192–3, 193t
development–morality nexus, 101–32 source of grievance, 176–7
clash of moralities, 106–7 economic problems, and collapse of
comparing ratings, 114–8 political Islamic governments, 257
data and methods, 108–13 education
direct questions related to, 122–3 as indicator of value shift, 6
hypotheses, 107–8, 124 national identity and national
Lebanese ratings of development, pride, 79–80
118–22, 119t and political engagement, 172–3, 182,
modernization theory, 103–5 188t, 189
multiple modernities, 126–7 of revolution makers, 212
rating countries on efficacy, personal. See personal efficacy
development, 113–4 egalitarian orientation toward
rating countries on morality, 114 women, 39, 47
world system theory, 105–6 Egypt, 179, 203t. See also cross-national
development programs, 122–3 survey research; development–
Devotees of Islam (Fedaiyan-i morality nexus; Muslim
Islam), 74 Brotherhood; revolution makers;
differential bias, 317 subjective culture analysis
differential recruitment, 172–3, 189, antigovernment demonstrations and
192, 214–5 general strikes in, 185, 186t
dignity, research on, 265n2, 346–7 during Arab spring, 249–53
discursive literacy, 211–2 attitudes toward secularism,
discussions, political, 180t, 181 40–2, 47–50
disease, effect on values, 10 autocratic recidivism, 271–2, 285
disintegration of family life, 161–2 collapse of synchronization, 276
domestic TV, communicative power country fixed-effect model explaining
through, 216–7, 221 engagement, 191–2, 191t
354 | INDEX
355
cultural and political influence, 137–8 European countries

family life in, 137–41 cohort versus life cycle effects, 7–11
gender equality, 37–9, 45–7 effect of colonial rule, 32, 68
January 25 Revolution, 251–2 postmaterialist values, 11–3
liberalism index, 42 shift from survival values toward
models of family change in, 157–64 self-expression values, 14
modern marriage and family, 150–7 shift to postmaterialist values, 10–1
national identity, 43–4, 50–1 survey about value change, 6, 7f,
nonelectoral political engagement, 8–10, 9f
180t, 184–92, 188t transition from autocratic rule to
under political Islam, 255f, 254–9 democracy, 272
post-Arab Spring political European Social Surveys (ESSs), 302
trajectory, 263–4 exclusivity factor, 245
relative deprivation in university extraction, data, 332
graduates, 173–4
rise of Islamic nationalism, 72–4 fair elections, 198n4
social individualism, 34–7, 45 familism, 146, 147
strategic voting, 255–6 family bonds, 146, 153
territorial nationalism in, 68–9, 70 family life, developmental models
traditional marriage and family, 144–50 of, 133–68
value changes, 52 data and methods, 141–3
Eisenstadt, S.N., 126–7 depictions of family in state
electoral victories in Tunisia and family-planning program, 140–1
Egypt, 254–59 depictions of family in the
El Harbaoui, Mahjoub, 251 media, 139–40
elite-challenging actions, 176–7 depictions of family through
embeddedness, effect on political schools, 138–9
engagement, 175, 189–90, 192, 195–6 effect of external threats to cultural
employment, effect on political identity, 149–50
engagement, 182, 188t, 189 Egypt as context for discussion, 137–41
Enlightenment, influence of global dissemination of
ideology of, 31 developmental models, 134–6
Ennahda Party, Tunisia historical context, 137–8
attitudes toward political Islam before lay model versus developmental
and after elections, 256f models, 147–8, 154
collapse of, 253, 257–9 linear ranking of family
electoral victory, 249–50 models, 157–64
strategic voting, 256 modern marriage and family, 150–7
support for, 256 subjective meanings of modern
entropy, synchronization measured family, 155–7
by, 282 subjective meanings of traditional
Erdoğan, Recep Tayyip, 54n6 family, 148–50
ESSs (European Social Surveys), 302 three models of family
ethnicity, 33, 212 change, 157–64
ethnic nationalists, 27 traditional marriage and family, 144–7
Index | 355
356
family life and tradition, 118, 119t, strategic voting, 255

121t, 122 victory of, 249–50
family-planning program, Egypt, 140–1 Freedom House ratings, 14–23, 179,
Faraj, Muhammad Abd al-Salam, 74 183, 265n1
fatalism freelisting (FL) exercise, 134, 142–3
versus anomie, 243 freewill versus fate, 215
participation in revolution linked to, French Revolution, 210
207, 214, 221, 222t, 231 Fukuyama, F., 265n2
subjective culture analysis, 344
FDI (foreign direct investment), 177, G77 (Group of 77), 105–6
192–3, 193t, 194 Gabler, S., 300
Fedaiyan-i Islam (Devotees of Islam), 74 GDP (gross domestic product) per
feminism, Islamic, 30 capita, 176–7, 192–4, 193t, 203t
feminism, state, 30 Gelfand, M J., 10
Ferguson, J., 104 gender
FGDs (focus-group discussions), 141, and political engagement, 174,
142–3, 143f 188–9, 188t
films, depictions of family in of revolution makers, 212, 214, 221
Egyptian, 139–40 gender equality
financial obligations in preparation for clash of moralities, 107
marriage, 144, 146, 153 cross-national variation in
fixed-point attractor, 279 values, 37–9
FJP. See Freedom and Justice Party cultural difference between Islamic
FL (freelisting) exercise, 134, 142–3 and Western worlds, 16–20, 18f, 19f
focus-group discussions (FGDs), 141, modalities of nationalism and view
142–3, 143f of, 78, 91–2
Forbes, Duncan, 135 national identity versus national
foreign direct investment (FDI), 177, pride, 84t, 87t–9t, 90–1
192–3, 193t, 194 trends in values, 45–7, 48t
foreign powers youth values during Arab Spring,
economic domination linked to 27, 29–30
political engagement, 177 gender–equality index, 243
effect on nationalism, 65 general strikes, 185, 186t
modalities of nationalism and view of Ghannoush, Mohammed, 251
Western culture, 78–9 Gini coefficient, 179, 192–3, 193t, 203t
rise of territorial nationalism, 68, 76–7 globalization, effect on political
formative years, values fixed during, 5–6 engagement, 177
framing, ideological, 208, 214, government permission for survey
222t, 243–4 research, 302
France, morality rating of, 114–5, 124 grievances
Freedom and Justice Party (FJP) leading to Arab Spring, 261–2
attitudes toward, 257f link to revolutionary mobilization,
collapse of, 257–59 207–8, 214
loss of popularity, 253 and participation in revolution, 222t,
mobilization efforts, 266n10 231, 243
356 | INDEX
357
and political engagement, 173–4, ideological framing, 208, 214, 222t, 243

176, 195 ideological target of intellectual
gross domestic product (GDP) per leaders, 65, 71
capita, 176–7, 192–4, 193t, 203t impressionable years, 213, 217, 234
grounded theory approach, 333–7 income
Group of 77 (G77), 105–6 effect on postmaterialist values, 10–1
Guevara, Che, 65–6 nationalism and, 80
and participation in revolution, 213
Häder, S., 300 and political engagement, 182
Hamdi, Fedia, 266n4 source of grievance, 174
HDI. See Human Development Index independent variables, 182–3
Heeringa, S.G., 300 India, gender equality in, 18–9
hierarchical autocratic network, 280–1 individualism, social, 28–9, 34–7, 45
homosexuality, 20, 21f, 118, 119t, 121t, 122 Indonesia, 179, 203t
honor, research on antigovernment demonstrations and
coding process, 334–6 general strikes in, 185, 186t
interview design and sampling, 327 country fixed-effect model explaining
interview probes, 329 engagement, 191–2, 191t
linguistic dictionaries, creating, 339 forms of nonelectoral political
LIWC analyses, 338 engagement, 180t
network analysis, 340–1 index of nonelectoral political
sample interview protocol, 348–9 engagement, 184–8
semantic network maps of, 350f national identity versus national
hostility toward outsiders, 79, 84t, 90–1 pride, 83, 84t–9t, 90–1
Human Development Index (HDI), 128n1 social roots of nonelectoral political
correlation between country engagement, 188–92, 188t
development scores and, 113–4, 116 industrial societies, values in, 12
development-morality nexus, inegalitarian attitudes toward
110t–1t, 112 women, 39, 47
political engagement and, 192–3, inequality, political engagement and,
193t, 194 177. See also gender equality
for six Muslim-majority nations, 203t infectious disease, effect on values, 10
Human Rights Watch, 263 Inglehart, R., 26, 176–7
Huntington, S., 14–5, 106 insecurity, and participation in
Husri, Sati’ al-, 71 revolution, 215, 221
interest in politics, and political
Ibrahim, S.E., 266n6 engagement, 174, 182, 188t, 189, 196
identity intergenerational value change, 3–24
basis of, 27, 33 cohort versus life cycle effects, 7–11
cross-national variation in values, in Islamic world, 14–23
43–4, 43f, 44f key hypotheses, 5–6
nationalism and, 79 postmaterialist values in culture
national pride versus national, change, 11–3
77, 83–91 shift from survival values toward
trends in values, 50–1 self-expression values, 14
Index | 357
358
Internet, communicative power in family models, 163

through, 212, 216–7 modalities of nationalism and
interviewing techniques view of, 78
cross-national survey research, 316–8 and political engagement, 175–6,
Egyptian family study, 142 182–3, 188t, 190, 192
subjective culture analysis, separation of politics and religion,
326–30, 344–9 31–3, 47, 49t, 50
interview probes, 329 Islamic feminism, 30
intolerance factor, 245 Islamic fundamentalism. See also
Iran religious nationalism
Constitutional Revolution in 1906, 31 assault on modernity, 30
Islamic versus liberal nationalism in, emergence of, 70–1
76–7, 94 ideological framing, 208, 214–5, 221
means of revolutionary ISIL, 54n6, 285
communication in, 212 Middle East survey research program,
national identity versus national 305, 307–8
pride, 83, 84t–9t, 90–1 and participation in revolution, 214,
Revolution of 1979, 271–2 221, 223, 243–4
rise of Islamic nationalism, 72–4 political context favorable to rise
territorial nationalism in, 68–70 of, 32–3
Iraq. See also cross-national survey religion as basis of identity, 27, 33
research; subjective culture analysis Islamic modernists, 27
attitudes toward political Islam, Islamic nationalism
253–4, 257f, 259–60 general discussion, 61–2
attitudes toward secularism, versus liberal nationalism, 76–7
40–2, 47–50 rise of, 72–6
census data from, 300 Islamic world. See also political
emergence of pan-Arab engagement
nationalism, 70 gender equality, 16–20, 18f, 19f
gender equality, 37–9, 45–7 historic separation of religion and
Islamic nationalism versus liberal state, 31
nationalism in, 76–7 intergenerational value change
liberalism index, 42 in, 14–23
national identity, 43–4, 50–1 sexual liberalization, 16–7
national identity versus national shift between materialist and
pride, 83, 84t–9t, 90–1 postmaterialist values in,
separation of politics and 22–3, 22f
religion, 54n6 Islamism. See political Islam
social individualism, 34–7, 45
survey research in, 297 January 25 Revolution, 251–2
value changes, 52 Japan, morality rating of, 115
ISIL, 54n6, 285 Jasmine Revolution, 250–1
Islam. See also Islamic fundamentalism; Jihad Organization (Tanzim al-jihad), 74
political Islam; religiosity job market, women in, 38, 38t
basis of identity, 33, 43–4, 43f, 50, 51t Johnston, H., 185
358 | INDEX
359
Jordan, 179, 203t. See also subjective social individualism, 34–7

culture analysis liberalism
antigovernment demonstrations and cross-national variation in values, 42
general strikes in, 185, 186t gender equality, 29–30
attitudes toward political Islam, ideological framings, 208
253–4, 257f, 259–60 and participation in revolution, 214,
country fixed-effect model explaining 221, 223, 243–4
engagement, 191–2, 191t secularism, 31–3
forms of nonelectoral political social individualism, 28–9
engagement, 180t liberal nationalism
index of nonelectoral political decline of, 69–70
engagement, 184–8 historical analysis of, 68–9
national identity versus national versus Islamic nationalism, 76–7
pride, 83, 84t–9t, 90–1 modernity and, 78
semantic network map of honor, 350f national identity versus national
social roots of nonelectoral political pride, 77, 83–95, 87t–9t
engagement, 188–92, 188t national pride and, 66
value orientations, 77
Kamil, Mustafa, 69 life cycle effects, 4, 7–11
Khalkhali, Sadiqh, 74 linguistic analyses, 337–9
Khomeini, Ayatollah, 32, 74, 75 linguistic dictionaries, using interview
kinship, Arab models of, 146 data to create, 339
Kish table, 304 Linguistic Inquiry and Word Count
Kurzman, C., 172 (LIWC), 338–9
literal versus intended meanings in
language, role in revolution, 211–2 interviews, 331
language translation local cultural knowledge, importance in
cross-national survey research, 309–15 research, 334–5
subjective culture analysis, 332–3 logistic maps, 279, 290–91
Lapidus, I.M., 29 loose cultures, 10
Lebanon, 33–44. See also cross-national love as basis for marriage, 34–5, 35f, 45,
survey research; development– 46t, 53n3
morality nexus; subjective culture Lynch, M., 265n2
analysis Lynn, P., 302
census data from, 301
gender equality, 37–9 Maliki, Nouri al-, 54n6
identity measurement, 43–4 manifold individuals, notion of, 234–6
liberalism index, 42 marriage. See also family life,
national identity versus national developmental models of
pride, 83, 84t–9t, 90–1 basis for, 34–5, 35f, 45, 46t, 53n3
ratings of American and Saudi consanguineous, 148
development, 120–2, 121t modern, Egyptian perceptions of, 150–7
ratings of American and Saudi and participation in revolution, 213
morality, 118–20, 119t traditional, Egyptian perceptions
secularism, 40–2 of, 144–50
Index | 359
360
Marxism, 209, 211 material issues in, 28

mass–society, structural–functional, rise of Islamic nationalism, 72–4
and relative–deprivation (MSR) rise of territorial nationalism in, 68–9
theories, 205 secularism, 31–3, 40–2, 47–50
demographics, 213 social individualism, 28–9, 34–7, 45
individual participants, 209, 231 sociopolitical and cultural
morality, 210 issues in, 27
tensions leading to revolution, 206–7 trends in values, 44–51
materialist values migration, and political
birth cohort versus life cycle engagement, 174
effects, 7–11 Mill, James, 134–5
general discussion, 3–4 Millar, John, 29–30
polarization between postmaterialist mobile phones, communicative power
and, 12, 13t through, 216–7
survey about, 6, 7f mobilizing ideas and attitudes
McAdam, D., 173 indices of measurement, 214–5
measures, 181–2 predictive variables of study, 243
media in Egypt, depictions of family predictors of participation in
through, 139–40 revolution, 227t–8t, 231–2
mediums of communicative power. See PROP and MSR theories, 207–9
communicative power, mediums of modalities of collective
men. See also gender equality sovereignty, 64–7
participation in revolution, 212, modalities of nationalism, 61–100
214, 221 dependent variables, 78–9
political engagement, 174, form of government, 79
188–9, 188t gender and age, 91–2
MENA. See Middle East and gender equality, 78, 81f, 82f,
North Africa 84t, 87t–9t
meso-level structures hostility toward outsiders, 79,
manipulating social network 81f, 84t
structure, 281 independent variables, 79–80
role of, 275–7, 285 Islamic nationalism, 72–6
simulating role of, 283–4 Islamic versus liberal nationalism in
Middle East and North Africa (MENA). Muslim-majority countries, 76–7
See also cross-national survey modalities of collective
research; political engagement; sovereignty, 64–7
specific countries by name; subjective modernity and liberal
culture analysis nationalism, 78
basis of identity, 33, 43–4, 50–1 national identity versus national
cross-national variation in pride, 77, 83–91, 84t–9t
values, 33–44 national versus religious identity, 79
decline of territorial nationalism pan-Arab nationalism, 67–9, 71–2
in, 69–70 religious nationalism, 67–9
gender equality, 29–30, 37–9, 45–7 secular orientation, 78, 81f, 82f,
liberalism, 42 84t, 87t–9t
360 | INDEX
361
socioeconomic status, 79–80, 86t, Morsi, Mohammed, 249–50, 253, 262

87t–9t, 91 Mosaddiq, Mohammad, 70
structural equations models and mosque attendance, effect on political
data, 80–3 engagement, 175–6, 183, 190,
territorial nationalism, 67–9 192, 198n2
theoretical development, 63–4 MSR theories. See mass–society,
value orientations, 77 structural–functional, and
view of Western culture, 78–9, relative–deprivation theories
81f, 84t Mubarak, Hosni
modernization age of protesters against, 53n1,
development–morality nexus, 101, 190–1, 213
103–5, 115 arrest of, 249
general discussion, 133, 134 complaints against, 266n6
and liberal nationalism, 78 January 25 Revolution, 251–2
linked to cultural attitudes, 20–2 multinational surveys. See cross-
multiple modernities, 126–7 national survey research
perspective on nationalism, 63–4 multiple imputation
spiritual component of, 127 modeling, 184
modern marriage and family multiple modernities, 126–7
assigned attributes of, 150–4, 151t–2t multiple regression models,
lay model versus developmental 223, 224–9
model of, 154 Muslim Brotherhood
linear process of family change, arrest of leaders, 262
157–8, 159t, 160f attitudes toward political Islam after
subjective meanings of, 155–7 elections, 257f
modern societies, 134, 135–6 attitudes toward political Islam before
modesty, research on, 328, 345–6 elections, 257f
monarchical absolutism, 68 bottom-up strategy of, 197
morality. See also development– collapse of FJP government, 257–59
morality nexus; moral values, of electoral victory of FJP, 249–50
revolution makers formation of, 73
decay in, 106 mobilization efforts, 266n10
in family models, 162–3 organizing activities around
Middle East survey research program mosques, 190
measure of, 308–9 strategic voting, 256
moral flexibility, 246
moral values, of revolution makers nation
descriptive statistics, 222t concept of, 65, 95n1
indices of measurement, 214–5 in pan-Arab–nationalist modality, 71
predictive variables of study, 243 national development. See
predictors of participation rate, development–morality nexus
227t–8t national identity
PROP and MSR theories, 209–11 nationalism and, 79
results of study on, 231–2 versus national pride, 77,
Morocco, 83, 84t–9t, 90–1 83–95, 87t–9t
Index | 361
362
nationalism. See also modalities of Osella, C., 104

nationalism Osella, F., 104
affective dimension of, 66, 69
diversity of, 63–4 Pakistan, 179, 203t. See also cross-
generalization of, 61–2 national survey research; subjective
genesis of, 65–6 culture analysis; youth values
modernist perspective on, 63–4 during Arab Spring
perennial perspective on, 63–4 antigovernment demonstrations and
primordialism perspective on, 63–4 general strikes in, 185, 186t
national pride country fixed-effect model explaining
affective side of nationalism, 66 engagement, 191–2, 191t
versus national identity, 77, forms of nonelectoral political
83–95, 87t–9t engagement, 180t
nationalism and, 79 gender equality, 37–9
negotiation, research on, 328 identity measurement, 43–4
Nepal, modernity in, 127 index of nonelectoral political
Nepstad, S. E., 172, 210 engagement, 184–8
network analysis, 340–1, 350f liberalism index, 42
network of logistic maps, 279 public opinion of extremist
newspapers, communicative power groups, 54n7
through, 216–7 secularism, 40–2
next-birthday method, 304 social individualism, 34–7
Nidaa Tunes, 262 social roots of nonelectoral
nonelectoral political political engagement,
engagement, 194–7 188–92, 188t
country fixed-effect model Palestine, attitudes toward political
explaining, 191t Islam in, 253–4, 260f, 260–1
in established democracies, 187, 187f pan-Arab nationalism
forms of, by country, 180t Arab identity, 71–2
measures of, 181–2 emergence of, 33, 70–1
results of study on, 184–8 general discussion, 61–2
robust cluster regressions, historical analysis of, 67–9, 71–2
192–3, 193t ideological framing, 208
social roots of, 188–92, 188t parallel design, 307
nonlinear dynamical systems, 278 Parsons, T., 26
participation in Arab Spring.
obedience of wife, 38, 38t See political engagement;
occupying buildings, 180t, 181 revolution makers
opposition papers, 246 Pasha, Isma'il, 138
ordinary least squares (OLS) patriarchy, 29
regression, 184 patriotism versus nationalism, 66
organization, of revolution Pennebaker, Jaime, 337–8
makers, 212–3 per-capita income, as indicator of value
organizational membership, 247 shift, 6
Osa, M., 171 perceived control, 215
362 | INDEX
363
perennial perspective on political opportunity, and political

nationalism, 63–4 engagement, 177–8, 192–3
period effects, 8 Political Terror Scale, 179, 183
periodic attractor, 279 Pollard, Lisa, 138, 139
personal efficacy, 231–2 polygamy, 38, 38t
descriptive statistics, 222t postindustrial societies, values in, 12
versus dysphoric emotions, 209, 245 postmaterialist values
indices of measurement, 214–5 birth cohort versus life cycle
predictive variables of study, 243 effects, 7–11
predictors of participation rate, 221, general discussion, 3–4
227t–8t part of broader cultural change, 11–3
petitions, 180t, 181 polarization between materialist and,
pious modernity, 127 12, 13t
Poland, 272 scarcity hypothesis, 5
political conflict, resource mobilization, shift to, 10–1
organizational, and political survey about, 6, 7f
opportunity structure (PROP) poverty, and political engagement, 176–7,
theories, 205, 206–7 182, 192
demographics, 212 premarital sex, 17, 118–22, 119t, 121t
individual participants, 209, 231 preparation for marriage, 144, 146, 153
morality, 209–10 press in Egypt, depictions of family
political efficacy, 215 through, 139–40
political engagement, 171–203 pretesting, 315–6
contextual bases of, 176–8 primary sampling units (PSUs), 303
data and methods, 181–4 primordialism perspective on
economic and political nationalism, 63–4
contexts, 192–3 print capitalism, 211
in Muslim-majority nations, 179–80 PROP theories. See political
nonelectoral, 184–8 conflict, resource mobilization,
social movement theories and, 172–6 organizational, and political
social roots of nonelectoral, 188–92 opportunity structure theories
political exclusion, 183, 192–3, 193t, 203t prostitution, 118, 119t, 121t, 122
political Islam Protest Law, Egypt, 263
attitudes after living under, 254–6 protest movements. See political
electoral victories in Tunisia and engagement; revolution makers;
Egypt, 254–6 youth values during Arab Spring
future of, 261–4 PSUs (primary sampling units), 303
impact in other Arab public image and face, research on, 344–5
countries, 259–60 public opinion surveys. See also best
party platforms, 265n3 practices of survey research; specific
survey on, 40–2, 40f topics by name
Tunisia and Egypt during Arab about democracy, 15–6, 16f
spring, 250–3 about Egyptian families, 141–3
political leadership, women in positions about materialist and post materialist
of, 38, 38t, 47, 48t values, 6, 7f, 8–10, 9f
Index | 363
364
public opinion surveys (cont.) in family models, 163

on attitudes toward political modalities of nationalism and
Islam, 253–6 view of, 78
on basis of identity, 43–4 and political engagement, 175–6,
on gender equality, 37–9 182–3, 188t, 190, 192
on liberalism, 42 separation of politics and, 31–3, 47,
modalities of nationalism, 77–92 49t, 50
on secularism, 40–2 religiosity
on social individualism, 34–7, 35f, 36f and political engagement, 183, 190
in revolution makers, 215, 221,
qualitative research program, 325–43 222t, 245–6
data analysis, 333 religious fundamentalism. See Islamic
data extraction, 332 fundamentalism
data processing, 330–1 religious government, 40–2, 40f
determining constructs for religious nationalism
interviews, 326–7 emergence of, 70–1
grounded theory approach, 333–7 general discussion, 61–2
implementation of interviews, 329–30 historical analysis of, 67–9
interview design and sampling, 327–8 national identity versus national
lessons learned from, 325–6 pride, 77, 83–95, 87t–9t
other analytical strategies, 337–41 value orientations, 77
sample interview protocol, 344–9 replacement sampling, 304
transcription considerations, 331 repressiveness of state, effect on
translation, 332–3 political engagement, 177–8, 183,
qualities in children, 34, 35–6, 36f, 53n3 192–3, 193t, 203t
questionnaires respect, research on, 347
creating, 307–9 review time, translation, 333
design and layout, 315 revolution makers, 205–47. See also
interviewing techniques, 316–8 communicative power, mediums
language translation, 309–15 of; mobilizing ideas and attitudes
pretesting, 315–6 correlates of participation, 213–6
Qutb, Sayyid, 32–3, 73–5 data, 218–9
demographic variables, 212–3,
radio, communicative power 217, 233–4
through, 216–7 dependent variable, 213
recruitment, differential, 172–3, 189, differential recruitment, 214–5
192, 214–5 hypotheses and limitations of
regression analyses, 181–2, 184 study, 217–8
relative deprivation, and political individual efficacy versus dysphoric
engagement, 173–4, 195 emotions, 209
relative marriage, 148 moral values, 209–11, 214–5, 231–2
religion. See also Islamic multiple regression models, 224–9
fundamentalism notion of manifold individuals, 234–6
as basis of identity, 33, 43–4, 43f, organization, 212–3
50, 51t participation rate, 219–21, 219t
364 | INDEX
365
personal efficacy, 214–5, 231–2 modalities of nationalism and view

predictive variables of study, 243–47 of, 78, 79
predictors of participation, 221–31 national identity versus national
region where interviews were pride, 84t, 87t–9t, 90–1
conducted, 220t orientation toward secular state, 31–3
revolutions, failure of. See autocratic rise of ISIL in Iraq despite favor
recidivism toward, 54n6
robust cluster regressions, 192–3, 193t trends in values, 47–50
Roman law, 29 secular–rational values, 12
Romano, D., 298 security, sense of, 12
Romeo and Juliet revolution, 34–5, 35f, self-expression values, 12, 13t, 14
45, 46t semantic network analysis, 340–1, 350f
separation of religion and politics in
Sadat, Anwar al-, 74 Islamic world
safety issues in research, 297–8, 327 history of, 32–3
Sa’idi family, media portrayal of, 139–40 survey on, 40–2, 40f, 41f, 47, 49t, 50
sample design sexual liberalization, 16–7
cross-national survey research, 300–5 sexual norms, 107
subjective culture analysis, 328 Sharaf, 348–9
satellite TV, communicative power shari’a
through, 216–7, 221 attitudes toward, 40–2, 40f
Saudi Arabia. See also cross- and political engagement, 182–3,
national survey research; 188t, 190
development–morality nexus political engagement by supporters
attitudes toward secularism, of, 175
40–2, 47–50 Shi’i fundamentalism, 32, 74
gender equality, 37–9, 45–7 showcards, 312–3, 312f, 313f
Islamic nationalism versus liberal Smith, Anthony D., 61
nationalism in, 76–7 Smith, C., 172, 210
liberalism index, 42 Snow, D.A., 208
national identity, 43–4, 50–1 social capital, 212, 276, 281
national identity versus national social classes
pride, 83, 84t–9t, 90–1 and decline of territorial
social individualism, 34–7, 45 nationalism, 69–70
survey research in, 297 national identity and national
value changes, 52 pride, 80
scarcity hypothesis, 5–6 and participation in revolution, 213
Schielke, S., 266n5 social dislocation, 174, 189, 192, 195
schools in Egypt, depictions of family social groups, synchronization in, 274
through, 138–9 social individualism, 27, 28–9,
Schwartz, S.H., 26 34–7, 45
secular–democracy index, 244 social–individualism index, 243
secularism, 27 socialization hypothesis, 5–6
cross-national variation in social media, role of, 212, 273
values, 40–2 social movement theory, 172–6, 192, 195
Index | 365
366
social networks, and political surveys. See best practices of survey

engagement, 175, 189–90, research; public opinion surveys
192, 196–7 survival values, 4, 12, 13t, 14
social tolerance, 11–4 synchronization in societal
societal-level synchronization. See systems, 284–6
synchronization in societal formal models of
systems synchronization, 278–9
Society of the Muslim Brothers. See key parameters of model, 279
Muslim Brotherhood manipulating social network
socioeconomic development, 133 structure, 280–1
socioeconomic status meso-level structures, 275–7
national identity versus national role of, 274–5
pride, 79–80, 86t, 87t–9t, 90–2 simulating scenarios after collapse of
of revolution makers, 217, 247–8 autocratic regimes, 281–4
southern family, media portrayal simulation dynamics, 279–80
of, 139–40 Syria
South Korea, 213 civil war in, 285
spiritual component of modernity, 127 survey challenges in, 298
standard of living, effect on political territorial nationalism in, 68–70
engagement, 176–7
Stata 11, 184 Tabataba’ie, Sayyid Muhammad, 72
state family-planning program, Tanzim al-jihad (Jihad Organization), 74
Egypt, 140–1 target population, 303
state feminism, 30 technology, as means of revolutionary
stealing, 118, 119t, 121t communication, 212
strategic voting, 255–7, 257–8, 266n11 television in Egypt, depictions of family
subjective culture analysis, 325–43 through, 139–40
data analysis, 333 territorial nationalism
data extraction, 332 basis of identity, 27
data processing, 330–1 decline of, 69–70
determining constructs for gender equality movements, 30
interviews, 326–7 general discussion, 61–2
grounded theory approach, 333–7 historical analysis of, 67–9
implementation of interviews, 329–30 territory as basis of identity, 33
interview design and sampling, 327–8 value orientations, 77
lessons learned from qualitative territory as basis of identity, 33, 43–4,
research program, 325–6 43f, 50, 51t
other analytical strategies, 337–41 terrorism, 176, 177, 178
sample interview protocol, 344–9 Thornton, A., 32, 105
transcription considerations, 331 tight cultures, 10
translation, 332–3 Tilly, C., 207
subject matter sensitivities, 308–9, 317 traditional marriage and family
Suez Canal, 137 assigned attributes of, 144–7, 145t
sultan, political power of, 31 lay model versus developmental
Sunni fundamentalism, 32–3, 70–1, 73 model of, 147–8
366 | INDEX
367
linear process of family change, ulama, 31, 68

157–8, 159t, 160f UN Conference on Trade and
subjective meanings of, 148–50 Development, 106
traditional societies, 134, 135–6 United Arab Emirate. See subjective
traditional values, 12 culture analysis
training interviewers, 316–7, 329–30 United Nations Human Development
transcription of recorded interviews, 331 Index (HDI), 128n1
translation correlation between country
cross-national survey research, 309–15 development scores and, 113–4, 116
subjective culture analysis, 332–3 development-morality nexus,
trends in values, 44–51 110t–1t, 112
gender equality, 45–7 political engagement and, 192–3,
national identity, 50–1 193t, 194
secular politics, 47–50 for six Muslim-majority
social individualism, 45 nations, 203t
Tunisia. See also youth values during United States. See also development–
Arab Spring morality nexus; subjective culture
during Arab spring, 249–53 analysis
attitudes toward political Islam, 256f, role in rise of territorial
257f, 257–59 nationalism, 76–7
electoral victories for political semantic network map of honor, 350f
Islam, 256–8 university education for women,
gender equality, 37–9 38–9, 38t
identity measurement, 43–4 unofficial strikes, 180t, 181
liberalism index, 42 urban living, and participation in
post-Arab Spring political trajectory, revolution, 212
262, 265
secularism, 40–2 values. See also cross-national survey
social individualism, 34–7 research; intergenerational
survey research in, 297 value change; materialist values;
Turkey. See also cross-national survey postmaterialist values; subjective
research; subjective culture analysis culture analysis; youth values
attitudes toward secularism, during Arab Spring
40–2, 47–50 effect of national identity versus
gender equality, 37–9, 45–7 national pride, 83, 87t–9t, 90–1
liberalism index, 42 intergenerational shift from survival
national identity, 43–4, 50–1 toward self-expression, 14
national identity versus national laws inspired in Christian, 40–2, 40f
pride, 83, 84t–9t, 90–1 moral, of revolution makers, 209–11,
rise of Islamic nationalism, 73 214–5, 231–2
rise of territorial nationalism, 68–9 self-expression, 12, 13t, 14
social individualism, 34–7, 45 significance of, 26–8
value changes, 53 survival, 4, 12, 13t, 14
TV, communicative power through, traditional, 12
216–7, 221 variation in, 26
Index | 367
368
violence, 118, 119t, 121t dress, 34–5, 35f, 119t, 121t, 306

voting, strategic, 255–7, 257–8, 266n11 egalitarian orientation
toward, 39, 47
Wajh, 344–5 in job market, 38, 38t
Washington Post, 266n10 in positions of political leadership, 38,
wasta, 328 38t, 47, 48t
weak ties, 281 sexual liberalization, 16–7
weddings, cost of, 150, 153 university education for, 38–9, 38t
Welzel, C., 26 world system theory, 101–2, 105–6,
Western countries 115, 126
cohort versus life cycle effects, 7–11 World Values Surveys (WVSs)
gender equality, 16–20, 18f, 19f challenges in research, 301
modalities of nationalism and view of intergenerational value change study,
culture, 78–9 14, 15, 16f
modern family viewed as intrusion political engagement study, 181,
from, 155–7, 161–2 187, 187f
nationalism and view of Western World War II, value change after, 4
culture, 84t, 90–1 WVSs. See World Values Surveys
postmaterialist values, 11–3
role decline of liberalism, 70 xenophobia, 215–6, 221, 246
role in rise of territorial
nationalism, 76–7 youth, value change in. See
sexual liberalization, 16–7 intergenerational value change
shift from survival values toward youth values during Arab
self-expression values, 14 Spring, 25–57
shift to postmaterialist values, 10–1 basis of identity, 33
survey about materialist and post cross-national variation in
materialist values, 6, 7f, 8–10, 9f values, 33–44
Western families compared to gender equality, 29–30
modern families, 158, 159t secularism and orientation toward
westernization, 158–61, 160f, 306 secular state, 31–3
wife obedience, 38, 38t significance of values, 26–8
wireless networks, revolutionary social individualism, 28–9
communication over, 212 trends in values, 44–51
women. See also gender equality;
marriage za’im, 29
368 | INDEX

Gelfand, Michele J. - Moaddel, Mansoor-Values, Political Action, and Change in The Middle East and The Arab Spring-Oxford University Press (2017)

Enviado por

Dados do documento

Título original

Direitos autorais

Formatos disponíveis

Compartilhar este documento

Compartilhar ou incorporar documento

Opções de compartilhamento

Você considera este documento útil?

Este conteúdo é inapropriado?

Direitos autorais:

Formatos disponíveis

Gelfand, Michele J. - Moaddel, Mansoor-Values, Political Action, and Change in The Middle East and The Arab Spring-Oxford University Press (2017)

Enviado por

Direitos autorais:

Formatos disponíveis

i

Values, Political Action, and Change in

Values, Political Action, and

Published in the United States of America by Oxford University Press

© Oxford University Press 2017

All rights reserved. No part of this publication may be reproduced, stored in

You must not circulate this work in any other form

Library of Congress Cataloging-​in-​Publication Data

Printed by Sheridan Books, Inc., United States of America

SECTION I | Changes in Values and the Arab Spring

CHAPTER 2  outh Perceptions and Values During the Arab

SECTION II | Perspectives on Change: Development

CHAPTER 4 Modernization, World System, and Clash of

CHAPTER 5  ay Accounts of “Modern” and “Traditional” Family in

SECTION III | Social-​Scientific

CHAPTER 7  he Arab Spring and Egyptian Revolution

CHAPTER 8  hange and Continuity in Arab Attitudes toward

CHAPTER 9  utocratic Recidivism: Computational Models

SECTION IV | BasicMethodological Issues in the Study

CHAPTER 11  n Analysis of Subjective Culture in the Middle

W e thank the US Air Force Office of Scientific Research for funding a

Abdel-​Hamid Abdel-​Latif Nancy J. Davis

Sohair Mehanna Mark Tessler

Mansoor Moaddel and Michele J. Gelfand

T he anti-​authoritarian movements known as the Arab Spring signify as one

economic development in the perceptions of Middle Eastern and American peo-

sensitivities, infrastructure, census data availability, and survey methods training.

Makdisi, U. (2000). The culture of sectarianism: Community, history, and violence in

SECTION I Changes in Values and

CHAPTER 1 Changing Values in the Islamic World

About 45 years ago, I suggested “a transformation may be taking place in the

freedom of expression, to support for environmental protection, gender equality,

4 | Changes in Values and the Arab Spring

Changing Values In The Islamic World And The west | 5

up under different conditions from those that shaped earlier generations, so

6 | Changes in Values and the Arab Spring

Percentage of Age Group in Given Value Type

Cohort Effects Versus Life Cycle Effects

Changing Values In The Islamic World And The west | 7

8 | Changes in Values and the Arab Spring

Postmaterialists less materialists (%)

10 | Values, Political Action, and Change in the Middle East

Postmaterialist Values: Part of a Broader Cultural Change

Changing Values In The Islamic World And The west | 11

that emerged reflected (a) a polarization between traditional and secular–​rational

12 | Changes in Values and the Arab Spring

Table 1.1 Orientations linked with survival versus self-​expression values

Survival values emphasize the following:

Changing Values In The Islamic World And The west | 13

Intergenerational Shift from Survival Values

How Different Is the Islamic World?

14 | Changes in Values and the Arab Spring

democratization fail in most Arab countries? Huntington’s response would be

Changing Values In The Islamic World And The west | 15

16 | Changes in Values and the Arab Spring

and opportunities for women—​measured by such questions as whether men

Changing Values In The Islamic World And The west | 17

as a whole, high-​income countries show much stronger evidence of an inter-

18 | Changes in Values and the Arab Spring

Changing Values In The Islamic World And The west | 19

20 | Changes in Values and the Arab Spring

Changing Values In The Islamic World And The west | 21

22 | Changes in Values and the Arab Spring

Library of Congress Cataloging-in-Publication Data

CHAPTER 2 outh Perceptions and Values During the Arab

SECTION II | Perspectives on Change: Development

CHAPTER 4 Modernization, World System, and Clash of

CHAPTER 5 ay Accounts of “Modern” and “Traditional” Family in

SECTION III | Social-Scientific

CHAPTER 7 he Arab Spring and Egyptian Revolution

CHAPTER 8 hange and Continuity in Arab Attitudes toward

CHAPTER 9 utocratic Recidivism: Computational Models

SECTION IV | BasicMethodological Issues in the Study

CHAPTER 11 n Analysis of Subjective Culture in the Middle

Abdel-Hamid Abdel-Latif Nancy J. Davis

T he anti-authoritarian movements known as the Arab Spring signify as one

that emerged reflected (a) a polarization between traditional and secular–rational

Table 1.1 Orientations linked with survival versus self-expression values

and opportunities for women—measured by such questions as whether men

as a whole, high-income countries show much stronger evidence of an inter-

The bifurcation of Muslim-majority countries into two groups—one of which

Cross-national Variation in Values: Egypt, Iraq, Lebanon,

Separation of 51% 69% 80% 9% — 72% 76%

18–24 25–34 35–44 45–54 55+

18–24 25–34 35–44 45–54 55+