Escolar Documentos
Profissional Documentos
Cultura Documentos
Withdrawal and
Inattendance
A Current Perspective
A Continuation Order Plan is available for this series. A continuation order will bring
delivery of each new volume immediately upon publication. Volumes are billed only upon
actual shipment. For further information please contact the publisher.
Voluntary Elllployee
Withdrawal and
Inattendance
A Current Perspective
Edited by
Meni Koslowsky
Bar-Ilan University
Ramat Can, Israel
and
Moshe Krausz
Bar-Ilan University
Ramat Can, Israel
ISBN 978-1-4613-5151-1
ISBN 978-1-4615-0599-0 (eBook)
DOI 10.1007/978-1-4615-0599-0
http://www.wkap.nl/
ro 9 8 7 6 5 4 3 2 1
A c.I.P. record for this book is available from the Library of Congress
AII rights reserved
No part of this book may be reproduced, stored in a retrieval system, or transmitted in
any form or by any means, electronic, mechanical, photocopying, microfilming,
recording, or otherwise, without written permission from the Publisher, with the
exception of any material supplied specificalIy for the purpose of being entered and
executed on a computer system, for exclusive use by the purchaser of the work.
Contributors
Helena Mensah Addae
University of the West Indies, St. Augustine Campus, Trinidad & Tobago
Vishwanath V. Baba
Michael G. DeGroote School of Business, McMaster University, Hamilton,
Ontario L8S 4M4, Canada
Julian Barling
School of Business, Queen's University, Kingston, Ontario K7L 3N6,
Canada
Gary Blau
Human Resource Administration Department, Temple University, Fox
School of Business Administration, Philadelphia, Pennsylvania 19122
Yair Amichai Hamburger
Department of Psychology, Bar-Han University, Ramat-Gan, Israel 52900
Kathy A. Hanisch
Iowa State University, Ames, Iowa 50011
David A. Harrison
Department of Management and Organization, Pennsylvania State
University, University Park, Pennsylvania 16802
Muhammad Jamal
Department of Management, Concordia University, Montreal, Quebec
H3G 1MB, Canada
Gary Johns
Department of Management, Concordia University, Montreal, Quebec
H3G 1M8, Canada
v
vi
Contributors
Rabindra N. Kanungo
Faculty of Management, McGill University, Montreal, Quebec. H3A IG5,
Canada
E. Kevin Kelloway
Department of Management, Saint Mary's University, Halifax, Nova
Scotia, B3H 3C3, Canada
Meni Koslowsky
Department of Psychology, Bar-Han University, Ramat-Gan, Israel 52900
Moshe Krausz
Department of Psychology, Bar-Han University, Ramat-Gan, Israel 52900
Manuel Mendonca
Faculty of Management, McGill University, Montreal, Quebec H3A IG5,
Canada
Abraham Sagie
School of Business Administration, Bar-Han University, Ramat-Gan, Israel
52100
Caroline Weber
School of Industrial Relations, Queen's University, Kingston, Ontario,
K7L 3N6, Canada
Preface
viii
Preface
world. During the current phase of the business cycle, companies can dictate the terms of employment; the pressure is on to cut many jobs and
severely curtail the choice now available to job searchers.
To borrow a term that has been one of the key concepts explaining
employee turnover-perceived ease of mobility (March & Simon, 1958)labor markets have moved, within one year, from a high and often insatiable quest for new workers to large scale downsizing, with some
companies near bankruptcy. This situation has created fewer labor market
alternatives for those beginning their careers as well as for "old-timers"
wishing to make a job change or midlife transition. The reader must be
cognizant of the fact that these recent developments will undoubtedly
have an effect on the issues dealt with in the current volume. The question may, therefore, be raised as to whether voluntary withdrawal behavior is at all an important issue under these circumstances. Moreover,
whereas economic and labor market status reflect situational trends in the
demand and supply of employees, affected by cyclical fluctuations, the
world has for quite some years been undergoing changes that appear to
be more enduring and stable.
A related question raised by Harrison in his chapter concerns the
present status of the milder or temporary forms of withdrawal, namely,
lateness and absence. Are they still viable and important organizational
measures in an era where work communication may be considered as
"continuous," i.e., not bound by time or distance? Thus, working at home,
on the road, or over the Internet does not require an office and is not limited to the hours between 9 and 5.
While current knowledge does not provide firm answers to such
questions, it is our belief that the various forms and modes of withdrawal
behavior will continue to prevail, though possibly on a somewhat smaller
scale. First, good workers and performers are still a premium and are still
in demand (Donovan, 2001) even during these times. Second, for the average worker, even in labor markets that are indeed more risky and uncertain for employees, individuals will continue to be absent voluntarily and
ponder the possibility of finding a "better job," even if chances of doing
so are slim. Such behaviors and attitudes are just natural reactions and
are, to a large extent, independent of the economy. For example, the oftenquoted turnover model by Mobley, Griffeth, Hand, and Meglino (1979)
does include job alternatives as an antecedent but also incorporates
within it many individual difference variables and company-related
measures. Thus, psychological reasons such as job dissatisfaction or the
need to alleviate the stress associated with working on dull and unsatisfying jobs would be expected to explain a large amount of turnover
variance. In addition, people will still arrive late or take days off, perhaps,
Preface
ix
Preface
Preface
xi
xii
Preface
REFERENCES
Donovan, A. (2001, April 22). StiU hearing ') quit' at work. The New York Times,
p.4(L).
Kurland, N. B., & Bailey, D. E. (1999). Telework: The advantages and disadvantages
of working here, there, anywhere, and anytime. Organizational Dynamics,
5,53-68.
March, J. G., & Simon, H. A. (1958). Organizations. New York: Wiley.
Mobley, W. H., Griffeth, R. w., Hand, H. H., & Meglino, B. M. (1979). Intermediate
linkages in the relationship between job satisfaction and employee turnover.
Psychological Bulletin, 86, 493-522.
Contents
Chapter 1 ................................................
Antecedents of Employee Lateness: A Multiple-Level Model
Abraham Sagie, Meni Koslowsky, and Yair Amichai Hamburger
Chapter 2 .......... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
21
Chapter 3 .......... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
53
Chapter 4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
71
Chapter 5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
95
Chapter 6 ....................... . . . . . . . . . . . . . . . . . . . . . . . . .
133
Chapter 7 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
161
xiv
Contents
Chapter 8 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
167
Chapter 9 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Job Involvement and Absence: The Role of Constraints as
Moderators
Vishwanath V. Baba and Muhammad Jamal
179
Chapter 10. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
193
213
Voluntary Employee
Withdrawal and
Inattendance
A Current Perspective
1
Antecedents of Employee Lateness
A Multiple-Level Model
Abraham Sagie, Meni Koslowsky, and
Yair Amichai Hamburger
Steers & Rhodes, 1978). In addition, late employees are likely to reflect low
motivation, dissatisfaction at work, and low commitment to the organization (Hanisch & Hulin, 1990), which may diffuse among their co-workers
(Jamal, 1984). From the late employee's point of view, lateness behavior
may indicate an initial withdrawal from work that may deteriorate toward
more severe forms of withdrawal such as absenteeism and turnover
(Herzberg et al., 1957; Koslowsky, Sagie, Krausz, & Dolman-Singer, 1997).
Additionally, management responses to late arrivals may include the
recording of the late occurrences in the individual's file, disciplinary
reviews, and outright punishment. In view of such responses, lateness
could indeed be considered a costly behavior for the involved employee(s).
Being a volitional, generally unexcused, dysfunctional behavior for
both the self and the organization, one may ask why the employee does
not avoid tardiness behavior, and why frequently he or she consciously
chooses to come late to work. As these questions relate to the individual
level of analysis, the answers provided in the literature generally adhere
to this level. Most of the explanations concentrate on the individual's personality (Blau, 1994; Knatz, Inwald, Brockwell, & Tran, 1992) or on his or
her work relationships, experiences, and expectations (Hanisch & Hulin,
1991; Koslowsky et al., 1997; Steers & Rhodes, 1978). Detailed explanations focusing on the individual are addressed below. The overall aim
of the present chapter is to suggest that individual-level variables are part
of a multiple-level model that also identifies and describes various
antecedents at the group (e.g., work-team), organizational, and extraorganizational (e.g., national culture) levels. The primary notion is that the
process of lateness behavior and its antecedents would better be understood if one considers all of these levels of analysis.
Figure 1 shows the proposed multiple-level model; causal linkages
within the same level of analysis are depicted by horizontal arrows connecting antecedents to outcomes (i.e., lateness behavior). Vertical arrows
represent causal linkages leading from a higher level (e.g., organizational)
to a lower one (e.g., group). Table 1 presents the antecedents at each level.
The discussion will be presented first for each level separately and then
for the links across levels. In addition to describing the model and its multiple levels, we will provide here the results of a survey of the causes of
lateness that were gathered from a sample of managers. The fit between
the empirical data and the underlying model will be assessed.
INDIVIDUAL LEVEL
In the literature, there are five main categories of factors used to explain
the individual-level lateness behavior: personality traits, organizational
Employee Lateness
Extraorganizational
Factors
....
Ex traorganizationalLevel
Lateness Behavior
,It
Organizational
Factors
Organizational-Level
Lateness Behavior
,It
Group
Factors
-"
Group-Level
Lateness Behavior
,It
Individual
Factors
Indi vidual-Level
Lateness Behavior
Group or organizational
level
Extraorganizational
level
exhibit either stable levels or, for various reasons, increasing chronic lateness behavior. The latecomers may "cut things too close," i.e., feel that
they have all the time in the world to get to work, and will be quite surprised when the train already left at the station or the traffic jam turns out
to be worse than expected. Blau's analysis clarifies that the delay time is
not the only component that should be taken into account when one calculates the impact of mismatched personalities on attendance behavior.
Consider a meeting of prompt and tardy participants. Whereas the former
may enter the meeting room a short time (e.g., 15 minutes) before the
scheduled time, the latter will be there some time (e.g., half an hour) after
the appointed time. Hence, the punctual person would wind up waiting
during the period between both arrivals (in our example, 45 minutes).
A personality dimension that is especially relevant to attendance
behavior is the Type A behavioral pattern. Baker, Dearborn, and
Hamberger (1984) found that Type A individuals arrive earlier for
appointments than their counterparts, Type B individuals. Also, Burman,
Pennebaker, and Glass (1975) have reported that Type A persons judged
the lapse of 1 minute sooner than did Type B persons. In a study by
Levine and Bartlett (1984), the authors found a moderate, though significant, relationship between Type A behavior and self-reported measures of
punctuality. One specific component of Type A pattern, time urgency,
holds considerable promise as a potential predictor of lateness behavior.
Time urgency can be defined as working at an "accelerated pace"
(Burman et al., 1975; Landy, Rastegary, Thayer, & Colvin, 1991). It is the
tendency to consider time as a scarce resource and to plan its use very
carefully. An individual with a high level of time urgency always feels
that he or she does not have enough time. Indeed, Bercovits (1996) found
that time urgency significantly discriminates between never-late and late
Employee Lateness
workers. Workers who showed higher levels of time urgency had fewer
lateness occurrences at work.
An entirely different approach to promptness and lateness in the
workplace considers these behaviors as a result of one's differential social
or organizational resources (e.g., role, status, prestige, and power) rather
than as an outcome of personality. From an organizational point of view,
the employee's working time is not considered to be under his or her
exclusive control, but is considered an organizational resource. Further,
the cost of employee time is not constant across hierarchy levels. Senior
workers' time is believed to be more expensive than that of junior workers. Hence, it is typically accepted that higher-ranked members (e.g.,
superiors) are allowed to control the time of lower-ranked workers.
Consequently, in many organizations, different informal rules define the
entitlements and obligations of superiors and their subordinates with
regard to punctuality and lateness behaviors, in general, and, more specifically, when both parties meet with each other. The inequality among the
various rules reflects the diversity of roles, power, and status in the workplace. In Table 2, we propose an informal protocol of conduct for two
roles: boss (or superior) and subordinate.
Table 2. An Informal Protocol of Attendance Behaviors for Superiors and
Subordinates in the Organization
Subordinate
Superior
Based on Table 2, one may see that different attendance rules control
the superior and subordinate during their meeting. The rules related to
the superior are more generous and those related to the subordinate are
stricter. The subordinate is expected to precede the superior at the meeting room. He or she should not come late to the meeting; if he or she does,
a detailed excuse is required. On the other hand, the tardy superior does
not need to provide a detailed explanation; often a few general words are
adequate. If the meeting takes place in the superior's room (as normally
happens), the subordinate has to wait in or out of the room until the superior finishes doing the current activities, such as interview, phone call, or
document writing. Conversely, if the meeting takes place in the subordinate's room, he or she must stop doing the current activities before the
meeting's onset.
Indeed, these differential rules are not universal. As described in
detail below, the organizational culture and the norms of the surrounding
society may affect the discrepancy in attendance rules between the
superior and subordinate. Additionally, the quality of the superiorsubordinate relationship and, particularly, their mutual trust (Bauer &
Green, 1996; Dienesch & Liden, 1986) are inversely related to the divergence between both sets of rules. Similarly, the higher the subordinate's
commitment to the organizational goals, the higher his or her performance, and the more opportunities for delegation, empowerment, and
participative decision making offered by the superior to the subordinate
(Sagie & Koslowsky, 2000), the smaller is the discrepancy in their respective attendance rules. The rationale is that achieving better relations
between both parties or some resemblance in their authorities (e.g., both
are involved in making work decisions) is incongruent with a large gap in
the protocols applying to their attendance and lateness. It appears, therefore, that various measures at the dyadic, organizational, or extraorganizational levels may interact with role, status, and power, and then
influence employees' lateness behavior. In the following sections, this
issue is described in further detail.
According to a personality-based framework, as well as an organizational resource one, it is difficult to attribute either an aim or an external
cause to one's lateness behavior. Alternatively, however, Blau himself
(Blau, 1985, 1986, and the reprinted chapter in this book) explained that
lateness might be a result of work-related situational variables. Rather
than internal (personality) causes, the focus here is on the causes or aims
of one's behavior, primarily concerned with his or her relationship with
the job or the company. Because of opportunities (e.g., important meeting,
proximity of raise discussions) the worker may come earlier to work; conversely, because of difficulties in the work setting (e.g., poor interpersonal
Employee Lateness
Employee Lateness
10
Proposition 3: Group or organizational factors, like rules, norms, reward system, leader's
self-examples, work arrangements, and organizational change, cumulatively account for
the variance in group- or organizational-level lateness.
Proposition 4: Group or organizational constant characteristics (e.g., rules, norms, reward
system, leader's self-example, work arrangements) primarily account for intergroup differences in lateness behavior; the situational factors (e.g., organizational change) primarily
account for intragroup variations over time in lateness behavior.
EXTRAORGANIZATIONAL LEVEL
Organizations do not exist within a vacuum; the impact of the environment within which the company operates on the staff attendance behavior is often crucial and can be more significant than the influence of the
specific company for which the employee works. The rightmost column
in Table 1 presents several extraorganizational antecedents of punctuality
and tardiness. Starting with industry type, it appears that as compared to
Employee Lateness
11
12
Employee Lateness
13
and prospects of alternative employment are low. Second, some employers use hard times to divest themselves of problematic employees. Both
explanations are equally relevant to lateness; hence, it could be proposed
that higher unemployment would be associated with a decrease in lateness. Another situational variable operating in the extraorganizational
environment is the political climate; it could be predicted that in times of
higher political stability lateness will decrease. Conversely, in times of
higher political tension or conflict, there will be more interference with
the employees' orderly attendance behaviors. Similar to our previous
analyses, we suggest that constant extraorganizational characteristics
account for differences between units (e.g., national cultures) in lateness
behavior and situational factors primarily account for within-unit variations over time in lateness behavior.
In summary, the following extra organizational-level propositions are
suggested:
14
Item
Cause of lateness
Personality dimensions
Situational variables
Nonwork life
Commuting
20
17
8
7
Group norms
Leadership
Rewards and sanctions
Flextime
13
7
7
3
Reward systems
Organizational norms
Organizational leadership
Discipline
10
7
2
2
Cultural differences
Industry type
Public transport
Political tension
22
7
Frequency
6
5
Employee Lateness
15
nonwork causes (e.g., home obligations, taking care of children), and difficulties in commuting.
Ouring the preparatory interviews with managers (they were not
included in the present sample), the interviewees rarely mentioned variables at the group and organizational levels as antecedents of lateness. The
questionnaire included, therefore, a direct question requiring the respondents to focus on the interteam factors. At these levels, the most important
causes of lateness were group norms, leadership, rewards and sanctions,
and flexible time. At the interorganizational level the respondents raised
such causes as reward systems, company norms, leadership, and discipline. Finally, when the respondents focused on national origin, cultural
differences were the most frequent explanation of tardiness (22 out of 40).
Other responses to this question were concerned with typical industry
characteristics (especially the existence of flextime arrangements), public
transport system, and political tension. It appears that our respondents
believed that an inadequate public transportation system increases the frequency of lateness occurrences. Also, they assumed that the higher the
political tension in a given society the more numerous are the interferences
with one's orderly attendance behavior. In summary, although not
designed to test the aforementioned propositions, these results supported
the notion that individual, group, organizational, and extraorganizational
factors may explain the phenomenon of employee lateness.
CONCLUSIONS
16
Employee Lateness
17
18
Employee Lateness
19
Hill, J. M., & Trist, E. L. (1955). Changes in accidents and other absences with
length of service: A further study of their incidence and relation to each other
in an iron and steel works. Human Relations, 8, 121-152.
Hofstede, G. (1980). Culture's consequences: International differences in work-related
values. Beverly Hills: Sage.
Jamal, M. (1984). Job stress and job performance controversy: An empirical assessment. Organizational Behavior and Human Performance, 33, 1-21.
Johns, G. (1997). Contemporary research on absence from work: Correlates,
causes, and consequences. In C. I. Cooper & I. T. Robertson (Eds.), International review of industrial and organizational psychology (Vol. 12, pp. 113-173).
New York: Wiley.
Kanekar, S., & Vaz, L. (1993). Effects of gender and status upon punctuality norms.
Journal of Social Psychology, 133, 377-384.
Kerr, S. (1995). On the folly of rewarding A, while hoping for B. Academy of
Management Executive, 9 (1), 7-16.
Klein, K. J., Dansereau, E, & Hall, R. J. (1994). Levels issues in theory development,
data collection, and analysis. Academy of Management Review, 19, 195-229.
Knatz, H. E, Inwald, R. E., Brockwell, A L., & Tran, L. N. (1992). IPI and MMPI
predictions of counterproductive job behaviors by racial groups. Journal of
Business and Psychology, 7, 189-200.
Koslowsky, M., Kluger, AN., & Reich, M. (1995). Commuting stress: Causes, effects,
and methods of coping. New York: Plenum Press.
Koslowsky, M., Sagie, A, Krausz, M., & Dolman-Singer, A (1997). Correlates of
employee lateness: Some theoretical considerations. Journal of Applied
Psychology, 82, 79-88.
Krausz, M., Sagie, A., & Bidermann, Y. (2000). Actual and preferred work schedules and scheduling control as determinants of job-related attitudes. Journal of
Vocational Behavior, 56,1-11.
Landy, E J., Rastegary, H., Thayer, J., & Colvin, C. (1991). Time urgency: The construct and its measurement. Journal of Applied Psychology, 76, 644-657.
Leigh, J. P. (1985). The effects of unemployment and the business cycle on absenteeism. Journal of Economics and Business, 37,159-170.
Leigh, J. P., & Lust, J. (1988). Determinants of employee tardiness. Work and
Occupations, 15, 78-95.
Levine, R. V., & Bartlett, K. (1984). Pace of life, punctuality, and coronary heart disease in six countries. Journal of Cross-Cultural Psychology, 15, 233-255.
Levine, R. V., West, L. J., & Reis, H. T. (1980). Perceptions of time and punctuality
in the United Sates and Brazil. Journal of Personality and Social Psychology, 38,
541-555.
Manrai, L. A, & Manrai, A K. (1995). Effects of cultural-context, gender, and
acculturation on perceptions of work versus social/leisure time usage. Journal
of Business Research, 32,115-128.
March, J. G., & Simon, H. A (1958). Organizations. New York: Wiley.
Mobley, W. H. (1987). Intermediate linkages in the relationship between job satisfaction and turnover. Journal of Applied Psychology, 62, 237-240.
20
Nicholson, N., & Goodge, P. (1976). The influence of social, organizational, and
biographical factors on female absence. Journal of Management Studies, 13,
234-254.
Ralston, D. (1989). The benefits of flextime: Real or imagined? Journal of
Organizational Behavior, 10, 369-373.
Richard, D., & Slane, S. (1990). Punctuality as a personality characteristic: Issues of
measurement. Journal of Psychology, 124, 397-402.
Sagie, A. (1998). Employee absenteeism, organizational commitment, and job satisfaction: Another look. Journal of Vocational Behavior, 52, 156-171.
Sagie, A., Birati, A., & Tziner, A. (2002). Assessing the costs of behavioral and psychological withdrawal: A new model and an empirical illustration. Applied
Psychology: An International Review, 51, 67-89.
Sagie, A., Elizur, D., & Koslowsky, M. (1996). Work values: A theoretical overview
and a model of their effects. Journal of Organizational Behavior, 17, 503-514.
Sagie, A., & Koslowsky, M. (2000). Participation and empowerment in organizations:
Modeling, effectiveness, and applications. Beverly Hills: Sage.
Shafritz, J. M. (1980). Dictionary of personnel management and labor relations. Oak
Park, IL: Moore.
Spink, K. S., & Carron, A. V. (1992). Group cohesion and adherence in exercise
classes. Journal of Sport and Exercise Psychology, 14, 78-86.
Steers, R. M., & Rhodes, S. R. (1978). Major influences on employee attendance: A
process model. Journal of Applied Psychology, 63, 391-407.
Trompenaars, E, & Hampden-Turner, C. (1998). Riding the waves of culture:
Understanding cultural diversity in global business (2nd ed.). New York:
McGraw-Hill.
2
National Culture and Perceptions
of Absence Legitimacy
Helena Mensah Addae and Gary Johns
In recent years there has been increasing interest in the cross-cultural aspects
of organizational behavior. This interest is a joint effect of the globalization
of business, increasing cosmopolitanism among researchers, and recognition that cultural contrasts can provide a new optic on traditional domains
of organizational research. Despite the growth in cross-cultural organizational research, one area has remained almost immune to its influence-the
study of so-called work withdrawal behaviors such as lateness, absenteeism, and turnover. This is especially curious when it is recognized that
these behavioral manifestations of withdrawal are themselves culture-free,
unlike many of the hypothetical constructs favored by organizational
researchers. Indeed, in the domain of absenteeism research, Martocchio and
Harrison (1993:295) allude to a "gaping cross-cultural hole." This chapter is
an attempt to repair this gap by proposing a model of some factors that
influence the perception of absence legitimacy across national cultures.
EFFECTS OF NATIONAL CULTURE ON ABSENTEEISM:
THE BASIC ARGUMENT
There are at least three reasons to suspect that cross-national cultural
differences will have an impact on attitudes toward absenteeism and
HELENA MENSAH ADDAE University of the West Indies, St. Augustine Campus,
Trinidad & Tobago
GARY JOHNS Department of Management, Concordia University,
Montreal, Quebec H3G 1MS, Canada
21
22
attendance, and ultimately on attendance behavior. One is growing evidence for the impact of social dynamics on absenteeism within cultures,
or at least within Western cultures where the extant research has been
conducted. Given the marked differences in values that have been
observed between cultures (Earley & Gibson, 1998; Hofstede, 1980, 1991;
Triandis, 1995) and the impact that values have on norms and other social
influence mechanisms, there is every reason to expect that views concerning absenteeism might differ across cultures.
Absenteeism is a behavior that has at least two qualities that would
seem to make it particularly susceptible to social influence, and by extension, cultural influence. For one thing, in the abstract, it is far from clear
just what constitutes a reasonable, legitimate level of absence. This follows from the rather large differences in absence rates that have been
observed between social units such as work groups, departments, plants,
industries, and nations (reviewed by Johns, 1997). This inherent ambiguity provides entree for social influence. At the same time, absenteeism is
perceived as mildly deviant behavior (Robinson & Bennett, 1995). At least
in the West, people make negative attributions about the absence of others (Nicholson, 1975), feel guilty about being absent (Hackett, Bycio, &
Guion, 1989), and get into disputes with employers about absenteeism
(e.g., Clay & Stephens, 1994). In both the East and the West, people underreport their own absence behavior and see their attendance as superior to
that of their work colleagues (Johns & Xie, 1998). These negative connotations do not deny individual differences in the perceived legitimacy of
absenteeism, but they do again suggest that people may be socially sensitive to how others view the behavior, thus setting the stage for social
influence (see Gellatly & Luchak, 1998).
Johns (1997, 2001) reviews the growing evidence concerning the
impact of social influence on absenteeism within national cultures. For
instance, a wide variety of operationalizations of absence norms tend to
be correlated with actual absence behavior (e.g., Baba & Harris, 1989;
Gale, 1993; Gellatly, 1995; Gellatly & Luchak, 1998; Harrison, 1995; Xie &
Johns, 2000). Similarly, work group structure and process variables, in
particular group cohesiveness, have been shown to predict absence.
Although cohesive work groups tend toward less absence (Johns, 1997),
moderators such as job dissatisfaction and collusive tendencies can stimulate absence in cohesive groups (Drago & Wooden, 1992; Xie & Johns,
2000). Finally, cross- and multi-level research has detected the existence of
distinctive absence cultures (Chadwick-Jones, Nicholson, & Brown, 1982;
Johns & Nicholson, 1982; Nicholson & Johns, 1985) at the unit (usually
work group) level of analysis (e.g., Iverson, Buttigieg, & Maguire, 1999;
Johns, 1994; Markham & McKee, 1995; Martocchio, 1994; Mathieu &
23
Kohler, 1990; Xie & Johns, 2000). Such research frequently reveals the
impact of the group on individual absenteeism or related perceptions.
A second reason why absenteeism might be expected to vary across
cultures rests in cross-national differences in factors such as economic
development, infrastructure support for families, social welfare provisions
concerning absenteeism from work, and so on. For example, Griindemann,
de Winter, and Smulders (1994) documented a range of legislative differences and a corresponding range of absenteeism rates between nations of
the European Union. Kaiser (1998) interpreted such differences in terms
of the economic implications of Hofstede's (1980) cross-cultural typology of
values. If the social influence described above constitutes discretionary
stimuli on the part of the culture that might shape absence attitudes and
behavior fairly directly, the influence being described here is ambient stimulation (Hackman, 1992), background factors that condition views about
the legitimacy of absenteeism and stem indirectly from cultural values.
A third reason why perceptions and attitudes about absenteeism
might be expected to vary across cultures rests in a small body of research
that has actually examined culture and job withdrawal. In the domain of
turnover, Abrams, Ando, and Hinkle (1998) found that low identification
with one's organization predicted intentions to quit in both Japan and
Britain. However, subjective norms regarding turnover had a much more
potent influence in collectivistic Japan than individualistic Britain. In a
case study of a General Dynamics plant in Arizona staffed by Navajo
Native Americans, Winfield (1995) explored how Navajo culture and tradition influenced work attendance. In particular, she described how participation in a nine-day healing ceremony led to absence and how plant
management accommodated this ritual. Kuzmits (1995) studied absenteeism among employees in a midwest U.S. food processing plant, some
of whom were Vietnamese refugees. Based on Confucian values, he predicted and confirmed that the Vietnamese exhibited less voluntary
absence and less absence-related discipline than non-Vietnamese, while
not differing on involuntary absence. Johns and Xie (1998) predicted and
found that Chinese and Canadians were equally self-serving in underestimating their own absence from work and seeing their attendance record
as better than that of their peers. However, in line with collective values,
the Chinese were also more inclined to group-serve, seeing the attendance
of their work peers as much superior to the occupational norm. The
authors also found cultural differences in the perceived legitimacy of various reasons for absenteeism. Finally, at the same Chinese research site,
Xie and Johns (2000) found that work group cohesiveness and absence
culture salience interacted to predict absence at the individual, work
group, and cross-levels of analyses. Although this was a uniculture study,
24
25
upset stomach, or mild backache were absent from work more frequently.
As they note, these are high-discretion illnesses in terms of attendance.
Directly relevant to the cross-cultural thesis, Johns and Xie (1998) found
similarities and differences between Chinese and Canadians in the perceived legitimacy of a wide variety of causes of absenteeism. For example,
the Chinese were less likely to endorse illness, stress, and depression as
reasons for absence and more likely to endorse house maintenance and
personal business. No difference was observed regarding problems with
bosses or co-workers. This limited evidence suggests promise for a theoretically sound model of cross-cultural differences in absence legitimacy.
In the model to be presented below, it will be argued that the following variables influence perceptions of absence legitimacy at both the cultural and individual levels of analysis: work centrality; time orientation;
locus of control; perceptions of gender role differentiation; perceptions of
the efficacy of social support systems. These variables were chosen for
consideration because of their documented or suspected variation both
between individuals and between cultures. This choice was intended to
accommodate cross-national variation while recognizing individual differences in values and perceptions within cultures. As such, the proposed
model is a multilevel model (Klein, Dansereau, & Hall, 1994; Klein,
Tosi, & Cannella, 1999; Rousseau, 1985) in that the proposed effects for
absence legitimacy are expected to occur at both the cultural and
individual levels. As Klein et al. (1994:223) explain, such models are
"uniquely powerful and parsimonious."
WORK CENTRALITY
26
27
28
their school activities and regard these as important in their lives and central to their conceptions of self." This definition is consistent with that of
work centrality and Kanungo's (1982) conceptualization of work involvement. However, the domain investigated here is different from the work
environment, which may suggest that the dynamics of academic involvement may be different from those of work involvement. These findings
provide some evidence that work centrality may not strongly influence
absenteeism. Perceptions of absence legitimacy may therefore be a viable
mediator in the relationship between work centrality and employees'
absenteeism.
In spite of the paucity of research linking work centrality with absenteeism, most studies of absenteeism have implied that work is central to
the lives of individuals and have ignored other life spheres that may be of
more importance and influential in explaining absence. This is seen in the
large amount of research that links job variables to absence (Harrison &
Martocchio, 1998; Johns, 1997) and in the relative lack of research on nonwork factors.
It is postulated that, in general, if work is a central life interest for individuals, their work involvement will be high and consequently they will
engage more in work and work-related activities than those whose involvement in work is low. As such, people with high work centrality would not
view absenteeism as a legitimate behavior and would exhibit lower levels
of absenteeism. Based on the preceding we offer the following proposition:
Proposition 1: There will be a negative relationship between employees' work centrality and their perceptions of absence legitimacy at the individual and national
levels of analysis.
TIME ORIENTATION
Time is one of the most important dimensions of human and organizational behavior. Despite its inevitability and pervasiveness, the dimension of time and its effect on employee behavior have not been accorded
appropriate attention in research in organizational behavior. There are
cultural differences in people's concepts of time, and societies can be
differentiated as either monochronic or poly chronic (Hall, 1966).
Monochronic societies view time as money, take time commitments such
as deadlines and schedules seriously, and emphasize promptness. On the
other hand, in polychronic societies people perceive time as a "free" good,
view time commitments as a goal to strive for if possible, and are likely to
change their plans easily and often (Hall, 1966, 1983).
29
30
United States and Canada for example, the general tendency is for shops
to close later than offices. Furthermore, a number of services such as
banking and airline reservations can be done via 1-800 numbers or the
Internet, and some grocery stores are open 24 hours a day.
In nations where there is uniformity of work schedules across institutions, this will lead to greater constraints on individuals' ability to
accomplish both work and nonwork activities. Absence from work to
accomplish nonwork activities that coincide with work hours may therefore be seen as a legitimate work behavior. In contrast, where there are
differences in work schedules, individuals will have relatively more flexibility in allocating their work and nonwork time, making it more difficult
to legitimize absenteeism as a means of performing nonwork activities.
For example, Johns and Xie (1998) found that while the Chinese perceived
house maintenance and personal business as legitimate reasons for absenteeism, their Canadian counterparts did not.
In spite of the fact that differing or uniform work schedules within a
nation will lead to disparate constraining effects on the allocation of time
cross-nationally, individuals' general values will also influence the
amount of time they allocate to various activities and subsequently influence their absences and their perception of absence legitimacy.
Johns and Nicholson (1982) posited that although there is an undeniable negative relationship between work and nonwork time, there is a
paucity of research in the area. Absence is a constrained behavior (Johns,
1991); thus, it temporarily redraws the boundaries between work and nonwork. If inflexible work schedules impose constraints on individuals'
behavior, non-conventional work schedules, such as flextime and the compressed workweek, should in theory improve employee attendance behavior. This is because individuals are expected to have more liberty in the
allocation of their time to meet competing demands from their various
social spheres. For example, both flextime and the compressed workweek
permit employees to engage in personal business or family activities during standard business hours. Therefore, to be absent in order to meet competing demands from nonwork spheres may not be perceived as legitimate.
Research findings on the relationship between the various nonconventional work schedules and absenteeism are mixed. The evidence for
flextime suggests that it stimulates attendance. For example, in a longitudinal field experiment, Dalton and Mesch (1990) found the introduction
of a flextime program in a large utility company led to a considerable
decrease in absenteeism, but absence levels returned to baseline rates with
the termination of the program. Kim and Campagna (1981) found that flextime reduced absenteeism of employees in a public sector organization.
31
Interviews with employees revealed that they felt they had more control
over the use of their time, and this allowed them to balance work and nonwork demands.
On the other hand, McGuire and Liro (1987) reported that traditional
flextime did not reduce absenteeism. Instead, they found that a staggered
fixed schedule reduced absenteeism among employees. Pierce,
Newstrom, Dunham, and Barber (1989) reported that researchers have
generally found a reduction in absenteeism with the introduction of flextime in organizations, a result confirmed in a recent meta-analysis by
Baltes, Briggs, Huff, Wright, and Neuman (1999).
These findings suggest that since flextime attempts to minimize or
eliminate constraints to attendance, employees on such work schedules
may not be able to legitimize their absences. A lack of justification for
absenteeism under a flextime schedule may be even more pronounced in
nations where there are differential work schedules across institutions.
This is because flextime allows them considerably more flexibility than
employees working under standard hours. For example, recent interviews with employees in a large multinational organization that introduced a flextime work schedule in Trinidad and Tobago suggest that
whereas employees may have viewed absenteeism as a legitimate behavior before the introduction of the system, they can no longer justify their
absences as a means of meeting nonwork demands.
Results for the effect of the compressed workweek on absence have
been less favorable. Ivancevich and Lyon (1977) found no change in
absence among operating employees in a manufacturing company. Nord
and Costigan (1973), however, found decreased absence in a pharmaceutical company after the introduction of a compressed workweek. Based on
five such studies, Baltes et al. (1999) concluded that the compressed workweek does not reduce absenteeism. It is likely that the compressed
workweek may have more effect on employees' nonwork activities than
work activities.
Another stream of research that is relevant to an individual's ability
to distribute his or her time accordingly is research on shift work. It is postulated that shift work allows employees discretion in the use of their time
to meet extra work responsibilities.
Taylor, Pocock, and Sergean (1972a) reported that absenteeism
among shift workers was lower than that of day workers. However, they
speculated that although shift work influences absence, certain factors
other than shift work may playa role in the effect of shift work on absenteeism. In another study (Taylor, Pocock, & Sergean, 1972b), in which they
examined factors such as sick pay schemes, weekly hours of work, and
32
33
Proposition 2: The flexibility of the boundaries between work and nonwork time
will influence perceptions of absence legitimacy at the individual and national
levels of analysis.
If competing demands on individuals' time have an effect on how
individuals apportion their time between their work and nonwork life
spheres, their time perspective is also expected to influence their behavior.
Landy, Rastegary, Thayer, and Colvin (1991) indicated that people differ in
their concern about time and its passage. Thus, people with a time urgency
perspective will function at a more accelerated pace than those who do not.
Central to this line of research are Type A and Type B behavior patterns.
Type A individuals have a heightened sense of time urgency, while Type Bs
have a more relaxed view of time. Although Landy et al. (1991) conceptualized time urgency as an individual-difference construct, it can be
extended to the national level of analysis. For example, Jones (1988) linked
Type A and B behaviors to temporal perspectives associated with different
cultural groupings. He advocated that in cultures where future time orientation is predominant, Type A behavior outcomes are expected to prevail.
Levine and Wolffe (1985) demonstrated that the pace of life differs
across nations. In a six-nation study, Japan was the most time-conscious
nation, while Indonesia was quite leisurely on the indicators of accuracy
of bank clocks, walking speed, and post office service speed. In a followup study of 31 countries, Levine (1997) found differences in the pace of
life among the countries surveyed. It can be argued that in nations where
the pace of life is fast, and time management is emphasized, individuals
would more likely have a time-urgent orientation. For example, in Latin
America, one might expect to wait hours even in the face of appointments, since appointment does not have the same meaning as it does in
the United States (Hall, 1973). It is therefore safe to assert that in general
Latin Americans would be more Type B than North Americans. Moreover,
it is estimated that 50% of North Americans are Type As (Friedman &
Rosenman, 1974).
Conte, Mathieu, and Landy (1998) posited that time urgency is
related to organizational outcomes such as performance (Friend, 1982;
Glass, Snyder, & Hollis, 1974). Explicitly relevant to current purposes,
Conte and Jacobs (1999) found that time awareness was negatively related
to absenteeism. They posited that employees who schedule their work
and are more aware of the passage of time are expected to be more punctual and would therefore be less absent. Furthermore, consistent with Hall
and Hall's (1990) assertion that polychronics would be less concerned
about the passage of time, they found a positive relationship between
34
35
36
roles. On the other hand, in places of low-role-differentiation, the separation of roles according to gender will be at a minimum.
Beach (1989) asserted that ideological perspectives on the societal
roles of males and females led to a gender-based model of work.
Accordingly, these ideological viewpoints have been sustained by a theory of structural functionalism that has upheld the need for such a division of labor. Traditionally, females have been charged with taking care of
the family and attending to household responsibilities. In comparison,
men have worked outside the home and have been expected to be the
dominant income earners in the family (Lewis, 1992; Major, 1993).
This is consistent with Hofstede's (1980) stipulation that in masculine
societies, gender roles are clearly differentiated and males are expected to
dominate society and females are expected to be nurturing. In contrast,
feminine societies believe that gender roles are more fluid and therefore
they espouse gender equality. As such, like females, males can also take
on nurturing roles. Hofstede's views are congruent with those of
Trompenaars (1993), who also asserted that societies differ on how roles
are ascribed to males and females. It is thus possible to locate nations
along a gender role continuum.
Scott, Alwin, and Braun (1996) found evidence of cross-national differences in gender role differentiation. In a comparative study of Britain,
the United States, West Germany, the Netherlands, Italy, and Ireland they
found that the Netherlands was the least differentiated according to gender. Attitudes toward gender role differentiation in the United States and
Britain were on par and ranked second. This was followed by Ireland and
Italy, respectively, and it was found that West Germany was the most
gender-role-differentiated nation.
The demographic composition of the work force is changing internationally, and this should have some impact on gender role differentiation.
For example, in the United States, females constitute one of the fastestgrowing sectors of the work force. The proportion of working females
increased from 36% to 54% between 1960 and 1991 (U.S. Department of
Commerce, 1992). While the rate of growth of participation in the work
force is different in other nations, the trend is similar. In Trinidad and
Tobago, there was a steady increase of females in the work force between
1990 and 1993 (Central Statistical Office, 1993). The Statistical Institute of
Jamaica (1994) reported a total population of about 2.5 million in 1993,
comprising 50% females. In the same year, employment of female workers increased by 7400; however, there was a decline of 6800 in male
employment (Economic and Social Survey, 1995).
With the influx of females into the work force, changes in the
compartmentalization of gender roles have occurred. For example, the
37
division of roles between females and males in the United States has
changed over time due to economic and historical reasons (Beach, 1989),
and men have continued to increase their family roles (Pleck, 1985).
Notwithstanding the changes in role differentiation, females continue to be
more involved in traditional roles than their male counterparts. In addition,
females in high-role-differentiated nations will probably be employed in
lower-paying jobs than their counterparts in low-role-differentiated
nations. This is because low-role-differentiated nations will tend to be more
egalitarian. As such, it is likely that females will be employed in jobs that
are traditionally considered male jobs, whereas in high-role-differentiated
nations, certain jobs remain exclusively reserved for males.
In Japan for example, there is high gender role differentiation. In fact,
there is a ban on females engaging in overtime, midnight, and holiday
work. Females are also rarely promoted to higher-level positions in organizations. Despite the fact that Japan has the largest number of employed
women in the developed world (Friedland, 1994), only 2% of them are in
managerial and professional positions (Nihon Horei Kyokai, 1989).
It is anticipated that females will generally be more absent than males
in most nations. Females in high-role-differentiated nations, however, are
expected to be relatively more absent than those in low-role-differentiated
nations. However, absence levels for dominant income earners may not
be affected by gender role differentiation, so both males and females who
are dominant income earners may generally be less absent than those who
are not. For example, females are reported to be less absent than males in
Jamaica (Employers' Consultative Association of Trinidad and Tobago,
1993). This is because a majority of females in the survey headed their
households.
A meta-analysis by Cote and Haccoun (1991) determined that women
exhibit higher absence levels than men. For example, in the United States,
the Bureau of Labor Statistics (1982) reported that the incidence of absenteeism for women was 58% higher than that for men. For instance,
Meisenheimer (1990) reported that during an average week in 1989, the
U.S. absence rate for women was 6.6% compared to 4% for their male
counterparts. Furthermore, females had higher rates of absenteeism than
males in each age group. Due to the widespread acceptance of women's
additional nonwork roles, there seems to be a psychological contract that
has allowed for the institutionalization of higher absenteeism among
females. This institutionalization of female absenteeism is also prevalent
in other nations. For example, Hein (1984) posited that employers in
Mauritius perceived that women who were absent from work had legitimate reasons such as family responsibilities or illness. However, the managers were not equally tolerant of male absenteeism and perceived that
38
males who were absent did so for "absolutely no good reason." Thus, we
offer the following proposition:
39
40
was abolished. Watkins (1994) noted that these changes were precipitated
by an increasing concern about the effects of statutory sick pay on
absenteeism.
In a comparative study of sickness insurance on absenteeism, Prins
(1990) as well as Prins and de Graaf (1986) asserted that unlike the United
States and Canada, employees in Belgium, Germany, and the Netherlands
are entitled to state-supported sickness insurance. They found that the
Netherlands had both a higher frequency of sickness absence and longerduration spells than either Belgium or Germany. These differences were
attributed to institutional disparity among the countries in the provision
of social insurance for sickness absence. Benefits in the Netherlands were
more generous than in the other countries. Moreover, eligibility criteria
were less stringent.
These findings concerning how national absence rates correspond to
the details of state-mediated sick leave plans parallel within-country findings that absenteeism is extremely sensitive to firm-level variations in sick
leave policies, control policies, and other incentives and disincentives
(Dalton & Mesch, 1991; Dalton & Perry, 1981; Ehrenberg, Ehrenberg, Rees,
& Ehrenberg, 1991; Kenyon & Dawkins, 1989; Ng, 1989). At both the
national and firm levels, it is likely that legitimization is a critical mediating variable. As economic and social provisions for absenteeism increase,
a signal is sent that absenteeism is a legitimate work behavior, not only
tolerated but also provided for (Kaiser, 1998).
Thus, while social support systems may serve to either increase or
decrease absenteeism, employees' perception of the efficacy of these
systems will influence their perceptions and behavior. In light of the
preceding, the following proposition is advocated:
41
(Klein et al., 1994; Klein & Kozlowski, 2000; Rousseau, 1985) of the impact
of culture, and we agree. Such models would posit main effects of culture
on individual absence or locate culture as a moderator of the relationship
between individual-level causes and absenteeism. In fact, we alluded to
one such moderator when we speculated how the tendency for women to
be absent more than men might be exacerbated in cultures with high
gender role differentiation.
Johns and Xie (1998) found several differences between Canadians
and Chinese in the People's Republic in terms of endorsing what constituted legitimate reasons for absenteeism. An extension of this finding is
that culture should moderate the relationship between potential causes
and the occurrence of absenteeism. For example, the needs of a distant
family member might be more likely to stimulate absenteeism in a more
collective culture than in an individualistic culture. Such more specific
hypotheses are not at odds with our more general propositions that
national cultures influence absence legitimacy through the mechanisms of
work centrality, self-control, time allocation, gender dynamics, and social
support.
Moving from theory to research, we have been conducting absence
research in Canada, the United States, China, Mexico, Trinidad, Nigeria,
Ghana, India, Pakistan, and Japan. It is well known that both cross-cultural
research and absenteeism research have special challenges associated
with them, and we will not review these here. One unique challenge when
the two are combined is accessing sites with relatively clean absenteeism
data that are otherwise reasonably comparable in terms of occupational
mix, gender mix, organization of work, and so on. We have not had great
luck in getting multinational firms to provide comparative cross-national
research sites; such access is almost always under local control. This
means that statistical control may be necessary to equate research sites
(e.g., Johns & Xie, 1998), a less palatable option than procedural control
via ideal sampling opportunities. In fact, Klein et al. (1999:244) posited
that "the analysis of multilevel data has been the topic of considerable
debate ... but perhaps even more daunting than multilevel data analysis is
the task of multilevel data collection." In a related vein, we have found
managers in less developed countries to be more receptive to absenteeism
research than North American managers. The reader can speculate on the
reasons for this.
A subtext to this chapter is that absence may have different meanings
in different cultures (see Johns & Nicholson, 1982). This having been
acknowledged, we would like to strongly emphasize one point: We have
not had particular difficulties asking respondents from various cultures to
reflect on absenteeism in questionnaires. We simply begin by defining
42
absence in a way that seems to be cross-culturally transportable (not showing up for scheduled work), adjusting the basic definition to reflect local
customs. For example, time off for jury duty is not absenteeism in Canada,
and time off to celebrate one's marriage is not absenteeism in China. Other
than these simple adjustments, we avoid attributions about cause in the
definition or measurement of absenteeism. Such attributions should be
restricted to predictor variables, not incorporated into the criterion.
CONCLUSION
The study of variations in the perceptions of the legitimacy of absenteeism across national cultures would seem to benefit both absenteeism
research and cross-cultural research. At the same time, it may have some
practical advantages for the management of attendance.
In the domain of absenteeism research, a cross-cultural perspective
virtually forces researchers to examine in detail the nonwork factors that
might affect attendance. As Johns (1997) notes, extant absence research is
highly deficient in this regard. Looking at absence cross-culturally, we
must confront off-the-job social influence and infrastructure differences
that surely affect absenteeism within our own cultures but have seldom
been studied. In addition, the cross-cultural approach highlights the more
general role that social influence may play in affecting attendance.
Although the value-added benefits of this theoretical position are now
well recognized (Johns, 1997,2001), the cross-cultural approach extends
and refines the role of social influence on absenteeism.
A cross-cultural approach has at least two other benefits for absenteeism research. The typical single-site research study is often dealing
with absence behavior that is highly constrained by social forces, industry norms, local absence culture, and organizational control systems (see
Johns, 1991). Studying absence cross-culturally provides requisite variety
in these ambient causal mechanisms such that one is able to examine less
constrained behavior. This provides both theoretical and methodological
leverage. In addition, and relatedly, cross-cultural research forces the
withdrawal researcher to attend more carefully to contextual forces on
withdrawal. It is our impression that such forces have been particularly
ignored in withdrawal research, including that on absenteeism.
In terms of cross-cultural research per se, the study of absenteeism
(and other so-called work withdrawal behaviors) may provide an interesting bridge between anthropological studies of mundane human behaviors (e.g., time usage) and psychological studies focused on hypothetical
constructs (e.g., time urgency). Because the act of absenting oneself from
work can be measured in a culture-free way, it is an attractive basis from
43
which to study organizational behavior cross-culturally. This can be contrasted with the study of concepts such as leadership, the meaning of
which is debated even within North American culture.
The cross-cultural study of absenteeism holds some relevance for the
practice of managing absence. One issue concerns the absence behavior of
migrants in an increasingly mobile global work force. How does an individual who has been socialized in a nation where absence is generally
viewed as a more legitimate behavior behave in a nation where it is
viewed as less so? The answer to this question is far from clear. Harrison,
Johns, and Martocchio (2000) speculated how ethnic diversity might
reveal normative differences about what constitutes a reasonable level of
absence, differences that might be especially salient in a team-oriented
work environment. For instance, employees with origins in collective cultures might be quite accepting of absence due to problems arising in their
extended families, a reason not acceptable to their individualistic work
colleagues or manager.
On the other hand, according to the cultural persistence and adaptation hypothesis (Joy, 1989), although individuals may adapt their cultural
values to suit different cultural settings, their original cultural values tend
to persist. It is therefore anticipated that in the short term, while individuals would adapt their absence behavior to suit the prevailing ethos in the
current culture in which they are employed, their perceptions of absence
legitimacy may not change. Thus, when such individuals return to their
country of origin, they would revert to their original absence behavior.
However, it is suggested that over time due to socialization in the new
culture, individuals would undergo a permanent change in their perceptions of absence legitimacy as well as their absence behavior.
The cross-cultural aspects of absence are also relevant to firms that
conduct global business. If our speculations are correct, a "global corporate attendance policy" is probably an ill-advised human resources initiative. Rather, expatriate managers need to be sensitized to indigenous
views about the legitimacy of absence. Knowing those views, attendance
management systems can be devised that correspond to and take advantage of the causal mechanisms behind them.
As we noted at the beginning of this chapter, work withdrawal
behaviors have been all but ignored in cross-cultural research on work.
We hope that we have demonstrated the value of examining the cultural
manifestations of absenteeism.
The preparation of this chapter was supported by
grant OO-ER-0506 from Quebec's Fonds pour la Formation de Chercheurs
et l' Aide a la Recherche and grant 410-99-1491 from the Social Sciences
and Humanities Research Council of Canada.
ACKNOWLEDGMENTS.
44
REFERENCES
Abrams, D., Ando, K., & Hinkle, S. (1998). Psychological attachment to the group:
Cross-cultural differences in organizational identification and subjective
norms as predictors of workers' turnover intentions. Personality and Social
Aycan, Z., Kanungo, R N., Mendonca, M., Yu, K., Deller, J., Stahl, G., and Kurshid,
A. (2000). Impact of culture on human resource management practices: A tencountry comparison. Applied Psychology: An International Review, 49,192-220.
Baba, V. V. (1989). Central life interests and job involvement: An exploratory study
in the developing world. International Journal of Comparative Sociology, 3D,
181-194.
Baba, V. v., & Harris, M. J. (1989). Stress and absence: A cross-cultural perspective.
Research in Personnel and Human Resources Management, Suppl.1, 317-337.
Baltes, B. B., Briggs, T. E., Huff, J. w., Wright, J. A., & Neuman, G. A. (1999).
Flexible and compressed workweek schedules: A meta-analysis of their
effects on work-related criteria. Journal of Applied Psychology, 84, 496-513.
Beach, B. (1989). Integrating work and family life. The home-working family. Albany:
State University of New York Press.
Blau, G. J. (1987). Locus of control as a potential moderator of the turnover
process. Journal of Occupational Psychology, 60, 21-29.
Brooke, P. P., & Price, J. L. (1989). The determinants of employee absenteeism: An
empirical test of a causal model. Journal of Occupational Psychology, 62, 1-19.
Byrne, N. (1982). Time construction in everyday life. Paper presented at the meeting
of the American Sociological Association, San Francisco.
Central Statistical Office. (1993). Social indicators of the Republic of Trinidad and
Tobago. Port of Spain: Office of the Prime Minister.
Chadwick-Jones, J. K., Nicholson, N., & Brown, C. (1982). Social psychology of absenteeism. New York: Praeger.
Clay, J. M., & Stephens, E. C. (1994). An analysis of absenteeism arbitration cases:
Factors used by arbitrators in making decisions. International Journal of
Conflict Management,S, 130-142.
Coleman, D. E, Irving, G. P., & Cooper, C. L. (1999). Another look at the locus of
control-organizational commitment relationship: It depends on the form of
commitment. Journal of Organizational Behavior, 20, 995-1001.
Conte, J. M., & Jacobs, R R (1999). Temporal and personality predictors of absence and
lateness. Paper presented at the annual convention of the Society for Industrial
and Organizational Psychology, Atlanta.
Conte, J. M., Mathieu, J. E., & Landy, E J. (1998). The nomological and predictive
validity of time urgency. Journal of Organizational Behavior, 9,1-13.
Cote, D., & Haccoun, R R (1991). L'absenteism des femmes et des hommes: Une
meta-analyse. Canadian Journal of Administrative Sciences, 8,130-139.
Dalton, D. R, & Mesch, D. J. (1990). The impact of flexible scheduling on employee
attendance and turnover. Administrative Science Quarterly, 35, 370-387.
45
Dalton, D. R., & Mesch, D. J. (1991). On the extent and reduction of avoidable
absenteeism: An assessment of absence policy provisions. Journal of Applied
Earley, c., & Gibson, C. B. (1998). Taking stock in our progress on individualismcollectivism: 100 years of solidarity and community. Journal of Management,
24, 265-304.
Economic and Social Survey of Jamaica. (1995). Planning Institute of Jamaica. Kingston.
Ehrenberg, R G., Ehrenberg, R A., Rees, D. I., & Ehrenberg, E. L. (1991). School
district leave policies, teacher absenteeism, and student achievement. Journal
of Human Resources, 26, 72-105.
Eisenberger, R, Fasolo, P., & Davis-LaMastro, V. (1990). Perceived organizational
support and employee diligence, commitment, and innovation. Journal of
46
Friend, K (1982). Stress and performance: Effects of subjective workload and time
urgency. Personnel Psychology, 35, 623--633.
Gale, E. K (1993). Social influences on absenteeism. Unpublished doctoral dissertation, Purdue University.
Gellatly, I. R (1995). Individual and group determinants of employee absenteeism:
Test of a causal model. Journal of Organizational Behavior, 16, 469-485.
Gellatly, I. R, & Luchak, A A (1998). Personal and organizational determinants of
perceived absence norms. Human Relations, 51,1085-1102.
Glass, D. c., Snyder, M. L., & Hollis, J. F. (1974). Time urgency and the Type A
behavior pattern. Journal of Applied Social Psychology, 4, 125-140.
Goff, S. J., Mount, M. K, & Jamison, R L. (1990). Employer supported child care,
work/ family conflict, and absenteeism: A field study. Personnel Psychology, 43,
793-809.
Griindemann, R W. M., de Winter, C. R, & Smulders, P. G. W. (1994). Absenteeism in
the European Union (Working paperWjPj94j29jEN). Shankill, Co. Dublin, Ireland:
European Foundation for the Improvement of Living and Working Conditions.
Hackett, RD., Bycio, R M., & Guion, R M. (1989). Absenteeism among hospital
nurses: An idiographic longitudinal analysis. Academy of Management Journal,
32, 424-453.
Hackman, J. R (1992). Group influences on individuals in organizations. In
M.D. Dunnette & L.M. Hough (Eds.), Handbook of industrial and organizational
psychology (2nd ed., Vol. 3, pp.199-267). Palo Alto, CA: Consulting
Psychologists Press.
Hall, E. T. (1966). The hidden dimension. Garden City, NY: Doubleday.
Hall, E. T. (1973). The silent language. Garden City, NY: Anchor Press.
Hall, E. T. (1983). The dance of life: The other dimension of time. Garden City, NY:
Anchor Press/Doubleday.
Hall, E. T., & Hall, M. R. (1990). Understanding cultural differences. Yarmouth, ME:
Intercultural Press.
Hamid, P. (1994). Self-monitoring, locus of control, and social encounters of Chinese
and New Zealand students. Journal of Cross-Cultural Psychology, 25, 353-368.
Harari, H., Jones, c. A, & Sek, H. (1988). Stress syndromes and stress predictors
in American and Polish college students. Journal of Cross-Cultural Psychology,
19,243-255.
Harrison, D. A. (1995). Volunteer motivation and attendance decisions:
Competitive theory testing in multiple samples from a homeless shelter.
Journal of Applied Psychology, 80, 371-385.
Harrison, D. A, Johns, G., & Martocchio, J. J. (2000). Changes in technology, teamwork, and diversity: New directions for a new century of absenteeism
research. Research in Personnel and Human Resources Management, 18, 43-91.
Harrison, D. A., & Martocchio, J. J. (1998). Time for absenteeism: A20-year review
of origins, offshoots, and outcomes. Journal of Management, 24, 305-350.
Harvey,J., & Nicholson, N. (1999). Minor illness as a legitimate reason for absence.
Journal of Organizational Behavior, 20, 979-993.
Hein, C. (1984). Jobs for the girls: Export manufacturing. International Labour
Review, 123, 251-265.
47
48
Klein, K. J., Dansereau, E, & Hall, R J. (1994). Levels issues in theory development, data collection, and analysis. Academy of Management Review, 19,
195-229.
Klein, K. J., & Kozlowski, S. W. J. (Eds.). (2000). Multilevel theory, research, and methods in organizations: Foundations, extensions, and new directions. San Francisco:
Jossey-Bass.
Klein, K. J., Tosi, H., & Cannella, A. A. (1999). Multilevel theory building: Benefits,
barriers, and new developments. Academy of Management Review, 24, 248-254.
Kluckhohn, E R., & Strodtbeck, E L. (1961). Variations in value orientations. New
York: Row, Peterson.
Kossek, E. E., & Nichol, V. (1992). The effects of on-site child care on employee attitudes and performance. Personnel Psychology, 45, 485-509.
Kuzmits, E E. (1995). Differences in incidence of absenteeism and discipline
between Vietnamese and non-Vietnamese employees. The International Journal
of Organizational Analysis, 3, 303-313.
Landy, J. E, Rastegary, H., Thayer, J., & Colvin, C. (1991). Time urgency: The construct and its measurement. Journal of Applied Psychology, 76,644-657.
Lefcourt, M. M. (1992). Durability and impact of the locus of control construct.
McGuire, J. B., & Liro, J. R (1987). Absenteeism and flexible work schedules. Public
Major, B. (1993). Gender, entitlement, and the distribution of family labour. Journal
Markham, S. E., & McKee, G. H. (1995). Group absence behavior and standards: A
multilevel analysis. Academy of Management Journal, 38,1174-1190.
Martocchio,
J. J.
49
MOW International Research Team. (1987). The meaning of work. New York:
Academic Press.
Ng, I. (1989). The effect of vacation and sickleave policies on absenteeism.
Pierce, J. L., Newstrom, J. W., Dunham, R. B., & Barber, A E. (1989). Alternative
work schedules. Boston: Allyn & Bacon.
Pleck, J. H. (1985). Working wives/working husbands. Beverley Hills: Sage.
Pocock, S. J. (1973). Daily variations in sickness absence. Applied Statistics, 22,
375-392.
Prins, R. (1990). Sickness absence in Belgium, Germany and the Netherlands: A comparative study. Amsterdam: Nederlands Instituut voor Arbeidsomstandigheden.
Prins, R., & deGraaf, A (1986). Comparison of sickness absence in Belgian,
German and Dutch firms. British Journal of Industrial Medicine, 43, 529-536.
Rahman, M. A, & Kumar, N. J. (1984). Absenteeism as related to locus of control.
50
51
3
The Many Faces of Voluntary
Employee Turnover
A Multifacet and Multilevel Perspective
Moshe Krausz
53
54
Moshe Krausz
55
56
Moshe Krausz
contract between the employee and the organization (formerly, the client
company of the external agency).
For individuals working under triangular arrangements, the
turnover process may be more complex than for employees on more traditional arrangements. It is likely that the cognitive processes that have
been proposed in Mobley's (1977) withdrawal model are duplicated:
Dissatisfaction, thoughts of quitting, evaluation of alternatives (jobs?
client organizations? agencies?), and withdrawal intentions need to be
considered for both entities-the client company and the external agency.
The above analysis focused on formal and tangible losses and gains
for the involved parties. Turnover, however, also incorporates psychological losses and gains. The decision whether termination relates to the
client company alone or to the agency as well, has implication for the individuals, for their supervisors from the client organization and from the
external agency, and for two social networks-that which prevails in the
client company and the other involving colleagues employed by the same
agency who may be working for different client companies.
The implication of the above for researchers is that both a conceptual
and operational definitions of turnover need to be explicitly formulated
for each study of employee turnover. Only in that way can valid comparisons be made between findings of various studies. One implication for
management is that a more differentiated outlook is needed when considering recruitment and retention of employees. Whereas the financial'
cost of both recruitment and turnover of contingent employees is lower
than for core employees, the psychological costs and benefits that are
involved are not yet sufficiently understood due to the scarcity of research
on behavioral aspects involved in the employment of workers on atypical
contracts. Are the psychological losses associated with the departure of
agency workers, who often work side by side with core employees, similar to those involved in turnover of core colleagues?
57
turnover process starts with evaluation of one's job and the ensuing level
of satisfaction. Perceptions of work alternatives, search of such alternatives, and their evaluation are further links in the chain that eventually
leads to turnover. Mobley acknowledges the possibility of certain variations in the temporal order of the various steps but in general, the process
is assumed to be developmental and linear.
Herzberg, Mausner, Peterson, and Capwell (1957) proposed another
outlook on progression. In contrast to Mobley's (1977) intrapsychic model,
their idea suggests that withdrawal from an unsatisfying work environment progresses from mild to more severe modes of behavior. The progression starts with lateness, continues to absenteeism, and eventually
culminates in turnover. Several reasons for such a progression may be proposed. First, actual turnover in the sense of dissociation of formal ties with
the organization is deemed by the individual as risky, difficult, or unrealizable. In that case, the employee may resort to easier solutions in the form of
lower and less complicated forms of withdrawal that are used as substitutes
for actual departure. A second rationale for such progression is the level of
discontent with the job. Mild dissatisfaction does not warrant turnover
from the organization. Only when satisfaction has further deteriorated and,
in addition, the mild withdrawal forms no longer provide effective coping
mechanisms, a higher (more severe) form of withdrawal may be sought.
Alternative job opportunities are not necessarily a factor that employees
consider in the initial and milder levels of dissatisfaction. This factor may
be probed only when the lower withdrawal forms have failed. Empirical
literature adopting the progression notion has been quite scant. Some
studies, however, found evidence of withdrawal progression (Krausz,
Koslowsky, & Eiser, 1998; Wolpin, Burke, Krausz, & Freibach, 1988).
Differential Importance of Satisfaction and Job Alternatives
An important conceptual treatise presented by Hulin, Roznowski,
and Hachiya (1985) proposed that the two constructs-satisfaction
and employment opportunities-may operate differently among different
populations. Thus, for example, turnover of marginal and temporary
employees may be more influenced by perceived opportunities than by
job satisfaction whereas the two variables will play an important role
among permanent and core workers. Peters, Jackofsky, and Slater (1981)
also found that elements of Mobley'S (1977) model significantly predicted
turnover among full-time but not part-time workers.
Hulin (1991) later suggested that turnover should be seen as one
out of several behavioral manifestations of an underlying adaptation/
withdrawal construct.
58
Moshe Krausz
Whereas dissatisfaction without turnover is a very common phenomenon, it seems that turnover without prior dissatisfaction is a less
prevalent situation. Three types of such turnover are proposed.
Impulsive/N on-Satisfaction-Dependent Turnover
Mobley (1977) acknowledged the possibility that the process
leading to turnover may be less rational and calculated than implied by
59
60
Moshe Krausz
61
stay. One interpretation given by the authors was that colleagues' departure may have acted like a torch that indicated the "light"-the availability of job alternatives. Such alternatives were either unnoticed prior to the
co-worker's departure or were deemed too difficult, risky, or unrealizable. The same study also found that the stayers' own work satisfaction
was not associated with the intention to leave, illuminating another variety of turnover that is unrelated to satisfaction.
If a contagion effect operates on turnover, then new labor market
realities such as the heterogeneity of employment contracts within organizations may have a substantial effect on turnover. A reality where core
and contingent employees work side by side and often collaborate within
work teams may imply contagion from the contingent to the core employees concerning loyalty and continuity of affiliation. As witnesses to the
mobility of contingent workers, core workers may react in one of two
ways. Some may bless their good luck for having a stable job whereas others may envy the opportunities for flexibility and variety that are afforded
to the contingent employees who are employed through external agencies
and who as part of that arrangement experience frequent changes of work
environment.
A research implication is the need to add co-workers' intention or
actual turnover as an antecedent of an employee's turnover. One way to
do that is by including items inquiring into employees' perceptions of coworkers' turnover intention. Management may use a similar approach by
including such questions in exit interviews.
Imposed Voluntary Turnover: The Difficulty of Disentangling
Voluntary from Involuntary
In certain situations or under some conditions, the distinction
between voluntary and involuntary turnover is not clear and in some
cases it is a mixture of both.
Moving employees from one site to another is an increasing phenomenon (Munton, Forster, Altman, & Greenbury, 1993). Most of the literature on relocation has dealt with individual relocation, namely,
situations when employees, most often in managerial positions, are
requested to relocate to another site of the same organization such as
another industrial plant or bank branch. In most cases, this is an individual move. In other instances, an entire plant, or a whole unit of an organization, relocates its premises to a different geographical location, often
requiring that employees move their home if they undertake the move.
Individual and plant relocation differ in several respects of which only
those that are relevant to voluntary turnover are mentioned here. In the
62
Moshe Krausz
63
64
Moshe Krausz
Large and versatile organizations may be an arena for withinorganization lateral mobility. Employees may move between teams, units,
departments, functions, branches, and geographical location and still be
members of the same overall organization. Such moves may indicate personal as well as organizational flexibility. Taken from the organization's
standpoint, it is a way of coping with fluctuating work loads and with
structural and functional changes by utilization of the current human
resources thereby reducing the need to embark on expensive processes of
external recruitment and selection. For the employees, self-initiated internal mobility may be a safeguard against foreseeable decline in work volume in one's current unit thus strengthening the individual's job security.
Furthermore, internal mobility provides a route for voluntary change of a
current job or unit without the need to cross organizational boundaries.
Note that change of unit within the same organization may involve a
65
radical change of the culture of one's work environment. In addition, selfsought internal change is a way of enriching one's own job or just a way
to accomplish work and social variety.
Seen from another angle, however, within-organization moves may
be regarded as turnover. While the person making the change remains in
the same organization, the unit that she or he left still needs to find
a replacement, just as in the case of external turnover. Similarly, there are
at least two parties who may regard to move as a loss: colleagues of the
person who moves and her or his supervisor.
For co-workers, in addition to the loss of a role partner, it also has the
socioemotional implications of losing a friend (Riordan & Griffeth, 1995;
Weiss, 1990). In some respect, employees who leave their unit for another
one in the same organization may be perceived as rejecting their teammates even more than one who leaves the entire organization. Attributions
of the causes for departure may be easier and more comforting for the latter than for the former. Similarly, internal moves are a setback for the head
of the unit. The effect on the supervisor may be even more harmful: Loss
of employees to another organization can be rationalized by attributing its
reasons to factors beyond the control of the supervisor or the team, such as
higher payor improved promotion chances. Conversely, loss of employees
to departments in the same organization may be seen as rejection of the
supervisor and the work team as leaders, co-workers, and friends.
Organizations are, therefore, caught in a conflict between the desire to
retain employees by offering and allowing internal mobility to revitalize
work, and the negative emotions that such moves may evoke among peers
and superiors. This may be the reason that many organizations do not
encourage internal mobility or even actively block employees' desires
to change.
Authors have differed in outlook on this phenomenon. Some authors
offered various approaches for understanding external and withinorganization turnover. Whereas Jackofsky and Peters (1983) asserted that
within- and between-organization turnover should be combined and
treated as a single criterion, Krausz, Koslowsky, et al. (1995), based on
longitudinal data on nurses' turnover intention, supported the initial
assumption that within- and between-organization turnover merit
separate treatment. The latter findings favor the recommendation made
by Dalton and Todor (1993) that one way to reduce turnover from an
organization is by offering job alternatives within its boundaries. It has
to be acknowledged that organizational policies to encourage internal
mobility are not necessarily appreciated by all employees; willingness and
capability for change differ between individuals. Personality
constructs and behavioral styles such as openness for change, neuroticism,
66
Moshe Krausz
This chapter points out that the March and Simon model or its later
developments are no longer the sole explanation or antecedents of
turnover. Although for many employees, dissatisfaction is a major
antecedent and many of them will not leave an organization without prior
searching (and finding) an alternative job, in many cases, job dissatisfaction
and search for an alternative are not primary antecedents of turnover. There
seem to be growing numbers of career starters as well as more experienced
67
employees who prefer to be their own career masters who may preplan the
points and stages for career transition and change.
This chapter highlights the fact that turnover is not a unitClry concept.
Its traditional definition as termination of formal relations "w',ith an organization is limited in scope; within-organization job changes are a clear
departure from that definition. Moreover, within-organization changes
highlight the need to adopt a multilevel outlook on turnover; whereas
from the organization's perspective this is definitely not turnover,
although it is perceived as such by employees (supervisors or colleagues)
or entire units within that organization. In another situation that has been
illustrated, namely, employee relocation, different work or non work
parties differ in their views about the move; some may regard it as
turnover, based on its impact on them, whereas others, mainly the individual making the move and the entire organization, will conceive it
entirely differently. It has also been shown that in light of the heterogeneity of employment contracts that prevails today, the entire essence of
turnover may need to be turned over. The growing use of contingent
workers can itself be considered as an indication of a whole change in the
meaning of turnover.
The above also highlights the need for a multilevel outlook on
turnover. In some way, today's organizations are caught in an insolvable
conflict. Whereas Hall and Mirvis (1995:335) adamantly proclaim that lIan
organization should not be in the business of career planning," it is clear
that employees do expect training and personal development and would
turn down job offers from companies that are known to provide little of
that. Thus, while companies, mainly in hi-tech, are highly concerned with
issues of recruitment and preservation of high-quality employees, provision of challenging jobs accompanied by training and personal development entails a risk that some employees will seek further levels of
professional upgrading and career promotion elsewhere.
Social and economic welfare of communities and nations may be
improved by high rates of employee voluntary turnover since this may be
an indication of adaptation to the dynamic changes required by global
competition. Conversely, high numbers of employees sticking to a job
for very long periods should be a sign of rigidity and difficulties in timely
coping with required changes. The different facets and faces of turnover
that have been presented demonstrate the need for further expansion
of multilevel models and empirical research designs. This need has already
received attention: The 1999 volume of Research in Organizational
Behavior consists entirely of papers that present theory and propositions
that make linkages between different levels of analysis (Sutton & Staw,
1995). The topic of voluntary turnover clearly deserves a similar treatment.
Moshe Krausz
68
REFERENCES
Adkins, C. L. (1995). Previous work experiences and organizational socialization.
Bakker, A. B., Le Blanc, P. M., & Schaufeli, E. W. B. (1998). Burnout contagion: Are
leU nurses at risk for infection with the burnout virus? Unpublished manuscript,
Psychological Department, Utrecht University.
Barrick, M. R, & Mount, M. K. (1991). The big five personality dimensions and job
performance: A meta-analysis. Personnel Psychology, 44,1-26.
Belous, R S. (1989). The contingent economy: The growth of the temporary,
part-time and subcontracted workforce. Washington, DC: National Planning
Associates.
Bowlby, J. (1988). A secure base: Parent-child attachment and healthy human development. New York: Basic Books.
Carsten, J. M., & Spector, P. E. (1987). Unemployment, job satisfaction, and
employee turnover: A meta-analytic test of the Muchinsky model. Journal of
Fox, S., & Krausz, M. (1987). Correlates of relocation intention and emotional
responses to an Israeli plant relocation. Journal of Occupational Behavior, 8, 325-338.
Golembiewski, R. T., Munzenrider, R E, & Stevenson, J. G. (1986). Stress in organizations: Toward a phase model of burnout. New York: Praeger.
Hall, D. T. (1996). The career is dead-Long live the career. San Francisco: Jossey-Bass.
Hall, D. T., & Mirvis, P. H. (1995). Careers as lifelong learning. In A. Howard (Ed.),
The changing nature of work. San Francisco: Jossey-Bass.
Hatfield, E., Caciopo, J. T., & Rapson, R L. (1994). Emotional contagion. London:
Cambridge University Press.
Hazan, c., & Shaver, P. S. (1990). Love and work: An alternative theoretical
prospective. Journal of Personality and Social Psychology, 59, 270-280.
Herzberg, E, Mausner, B., Peterson, RD., & Capwell, D. E (1957). Job attitudes:
Review of research and opinion. Pittsburgh: Psychological Services of Pittsburgh.
Hom, P. w., Caranikas-Walker, E, Prussia, G. E., & Griffeth, R W. (1992). A metaanalytic structural equations analysis of a model of employee turnover.
69
Hom, P. W, & Griffeth, R W (1995). Employee turnover. Cincinnati, OH: SouthWestern Publishing.
Hulin, C. L. (1991). Adaptation, persistence, and commitment in organizations. In
M. D. Dunnette & L. Houghs (Eds.), Handbook of industrial and organizational psychology (2nd ed., Vol. 2, pp. 445-506). Palo Alto, CA: Consulting Psychologists
Press.
Hulin, C. L., Roznowski, M., & Hachiya, D. (1985). Alternative opportunities and
turnover decisions: Empirical and theoretical discrepancies and an integration. Psychological Bulletin, 97, 233-250.
Huselid, M. A. (1995). The impact of human resource practices on turnover,
productivity, and corporate financial performance. Academy of Management
Journal, 38, 635-672.
Jackofsky, E. E, & Peters, L. H. (1983). Job turnover versus company turnover:
Reassessment of the March and Simon participation hypothesis. Journal of
Applied Psychology, 68, 490-495.
Kilgore, J. E., & Shorrock, M. B. (1991). International relocation: The future of
spouse relocation assistance. Journal of Career Development, 17, 271-284.
Krackhardt, D., & Porter, L. W. (1986). The snowball effect: Turnover embedded in
communication networks. Journal of Applied Psychology, 71, 50-55.
Krausz, M., Bizman, A., Yaakobovitz, N., & Caspi, T. (1999). Stayers' intention to
withdraw following turnover of a coworker. Journal of Business and Psychology,
14,96-107.
Krausz, M., Brandwein, T., & Fox, S. (1995). Work attitudes and emotional responses
of permanent, voluntary, and involuntary temporary-help employees: An
exploratory study. Applied Psychology: An International Review, 44, 217-232.
Krausz, M., Koslowsky, M., & Eiser, A. (1998). Distal and proximal influences on
turnover intentions and satisfaction: Support for a turnover progression theory. Journal of Vocational Behavior, 52, 59-71.
Krausz, M., Koslowsky, M., Shalom, N., & Elyakim, N. (1995). Predictors of the
intention to leave the ward, the hospital, and the nursing profession: A longitudinal study. Journal of Organizational Behavior, 16, 277-288.
Landy, E J. (1989). Psychology of work behavior. Pacific Grove, CA: Brooks/Cole.
Lee, T. W, & Mitchell, T. R (1994). An alternative approach: The unfolding model
of voluntary employee turnover. Academy of Management Review, 19, 51-89.
Lee, T. W., Mitchell, T. R, Wise, L., & Fireman, S. (1996). An unfolding model of
voluntary employee turnover. Academy of Management Journal, 39, 5-36.
McLean Parks, J., Kidder, D. L., & Gallagher, D. G. (1998). Fitting square pegs into
round holes: Mapping the domain of contingent work arrangements onto the
psychological contract. Journal of Organizational Behavior, 19, 697-730.
Mael, E A., & Ashforth, B. E. (1995). Loyal from day one: Biodata, organizational
identification, and turnover among newcomers. Personnel Psychology, 48,
309-333.
March, J. G., & Simon, H. A. (1958). Organizations. New York: Wiley.
Mathieu, J. E., & Zajac, D. M. (1990). A review and meta-analysis of the
antecedents, correlates, and consequences of organizational commitment.
70
Moshe Krausz
Meyer, J. P., & Allen, N. J. (1991). A three component conceptualization of organizational commitment. Human Resource Management Review, I, 61-89.
Mikulincer, M. (1995). Attachment style and the mental representation of the self.
Journal of Personality and Social Psychology, 69, 1203-1215.
Mikulincer, M. (1997). Adult attachment style and information processing:
Individual differences in curiosity and cognitive closure. Journal of Personality
and Social Psychology, 72, 1217-1230.
Mobley, W. H. (1977). Intermediate linkages in the relationship between job satisfaction and employee turnover. Journal of Applied Psychology, 62, 237-240.
Munton, A. G., Forster, N., Altman, Y., & Greenbury, L. (1993). Job relocation, managing people on the move. New York: Wiley.
Noe, R A., Steffy, B. D., & Barber, A. E. (1988). An investigation of the factors
influencing employees' willingness to accept mobility opportunities.
Personnel Psychology, 41, 559-580.
Peters, L. H., Jackofsky, E. E, & Salter, J. R (1981). Predicting turnover: A comparison
of part-time and full-time employees. Journal of Occupational Behavior, 2, 89-98.
Reichers, A. E. (1985). A review and conceptualization of organizational commitment. Academy of Management Review, 10, 465--476.
Riordan, C. M., & Griffeth, R W. (1995). The opportunity for friendship in the
workplace: An underexplored construct. Journal of Business and Psychology, 10,
141-155.
Robinson, S. L. (1996). Trust and breach of the psychological contract.
Administrative Science Quarterly, 41, 574-599.
Rosse, J. G., & Miller, H. E. (1984). Relationships between absenteeism and
other employee behaviors. In P. Goodman & R Atkin (Eds.), Absenteeism
(pp. 194-228). San Francisco: Jossey-Bass.
Rousseau, D. M. (1995). Psychological contracts in organizations. Beverly Hills: Sage.
Rousseau, D. M. (1998). Why workers still identify with organizations. Journal of
Organizational Behavior, 19, 217-233.
Sagie, A., Krausz, M., & Weinstein, Y. (2001). To move or not to move: Factors
affecting employees' actual relocation when an entire plant moves. Journal of
Occupational and Organizational Psychology, 74,343-358.
Schaufeli, w., & Enzmann, D. (1998). The burnout companion to study and practice.
London: Taylor & Francis.
Sutton, R J., & Staw, B. M. (1995). What theory is not. Administrative Science
Quarterly, 40, 371-384.
Wegmann, RG. (1991). From job to job. Journal of Employment Counseling,2S, 8--12.
Weiss, R S. (1990). Losses associated with mobility. In S. Fisher & c. L. Cooper
(Eds.), On the move: The psychology of change and transition. New York: Wiley.
Wolpin, J., Burke, R. J., Krausz, M., & Freibach, N. (1988). Lateness and absenteeism: An examination of the progression hypothesis. Canadian Journal of
Administrative Sciences, 5, 49-54.
Zuckerman, M. (1994). Behavioral expressions and biosocial bases of sensation seeking.
London: Cambridge University Press.
4
Employee Withdrawal Behavior
Role of the Performance Management Process
71
72
Employee Powerlessness
Reduced Employee's
-Job involvement
-Organizationa
commitmentl
Strategy to
reduce
employee witbdrawal behaviors
Perfonnance Management
Process
Employee Empowennent
Increased Employee's
-Job involvement
-Organizational
commitment
,
Increased Employee
Withdrawal Behaviors
-Membership
-Performance
-Extra-role
Reduced Employee
Withdrawal Behaviors
-Membership
-Performance
-Extra-role
Figure 1. An overview of the chapter: conditions with the potential to increase, and strategy
to reduce employee withdrawal behaviors.
73
Adaptation/withdrawal behaviors
74
As existing theories of motivation would suggest, the withdrawal behaviors seek to either increase job outcomes or decrease work output in order
to equitably compensate for the perceived work inputs (Adams, 1965;
Kanungo & Mendonca, 1997).
If we focus exclusively on employee withdrawal behaviors, including those suggested by Hulin (1991), then we can group them into three
broad categories:
75
76
of job alienation and reduces the employees' self-efficacy belief, decreasing their motivation to invest in the psychological and physical resources
needed to accomplish the job tasks. As a consequence, they have been
found to engage in withdrawal behaviors such as tardiness, absenteeism,
turnover, and decrease in job effort and performance (Brown, 1996).
Organizational Commitment and Withdrawal Behavior
Employee empowerment also fosters strong organizational commitment because the empowered employees now perceive greater competence and control over the organizational resources and environment
(Menon, 1999). What is organizational commitment? Organizational commitment has been defined in different ways. In essence, it is a psychological state that characterizes the employee's bond or relationship with the
organization. Meyer and Allen (1997:11) describe organizational commitment in terms of three components: affective commitment, continuance
commitment, and normative commitment. "Affective commitment refers
to the employee's emotional attachment to, identification with, and
involvement in the organization .... Continuance commitment refers to
an awareness of the costs associated with leaving the organization ....
Normative commitment reflects a feeling of obligation to continue
employment."
Employees high on all three types of commitment have low turnover.
Employees high on affective commitment have better attendance records,
work harder, and perform better. Employees high on affective commitment and normative commitment are more inclined to engage in extrarole
behaviors that contribute to organizational effectiveness-such as helping
co-workers and protecting the organization (Bishop & Scott, 1997;
Brown, 1996; Mathieu & Zajac, 1990; Meyer & Allen, 1997; Somers, 1995).
Compared to employees with a low organizational commitment, employees with a high organizational commitment have internalized the organization's vision that creates a stronger bond with the organization and,
therefore, are more likely to engage in behaviors that contribute to its
effectiveness (Dessler, 1999).
On the other hand, the feeling of powerlessness weakens the employees' organizational commitment because they do not experience the psychological state of bonding with the organization and, therefore, engage in
withdrawal behavior (Dessler, 1999). Employees with weak affective, continuance, and normative commitment have high turnover. Employees
with weak affective commitment demonstrate a higher rate of voluntary
absenteeism, low effort, and poor performance. Employees with weak
affective and normative commitments avoid extrarole behaviors
77
(Bishop & Scott, 1997; Brown, 1996; Mathieu & Zajac, 1990; Meyer & Allen,
1997; Somers, 1995).
From the discussion in this section, it is clear that employee feeling of
powerlessness is the root of the malaise of job alienation and weak
organizational commitment, with the resulting withdrawal behaviors.
The next section examines the notion and practice of empowerment
in order to lay a foundation for the development of an appropriate
strategy to empower employees, and implement a commitment-oriented
management practice that seeks to effectively address the withdrawal
behavior.
THE NOTION OF EMPOWERMENT IN MANAGEMENT
PRACTICE
In order to overcome the adverse effects of powerlessness in workers,
organizational researchers have suggested "empowerment" as the key
dealienation strategy (Block, 1987; Kanter, 1979). "The buzzword of the
1990s is 'empowerment' and it's one of the hottest topics in management"
(Stern, 1991:B4). What is empowerment? What is the specific process by
which it overcomes the debilitating effect of worker alienation?
Conger and Kanungo (1988) proposed that empowerment can be
viewed in two different ways: as a relational and as a motivational construct (see Fig. 2). Most theorists in industrial sociology and organizational behavior have tended to view empowerment as a relational
construct. Derived from the literature of social exchange theory, the relational dynamics revolves around the relative power status of the parties
in the exchange situation that might be one of dependence and/ or interdependence (Blau, 1964; Emerson, 1962; Homans, 1974; Thibaut & Kelley,
1959). In the organizational context, power is interpreted to mean the possession of formal authority or control over organizational resources.
Social exchange theories suggest that subordinates are more likely to
experience alienation because of their dependence on their managers who
have the power to determine whether they will achieve their desired outcomes. By sharing power, managers seek to overcome the alienation
brought about by subordination and control (Schacht, 1970). Viewed in
terms of the relational dynamics, empowerment has come to mean the
process (flow # 1, Fig. 2) of delegation or decentralization of decisionmaking power in which management gives away or shares power with
workers (Burke, 1986; Kanter, 1983). This connotation of empowerment is
also reflected in the Merriam Webster's Dictionary meaning of empower as
"to authorize or delegate or give legal power to someone."
78
;--
Organizational Conditions:
Delegation/Power Sharing
(Stems out of Relational Dynamics in
Social Exchange Theories)
*2
I-
79
decision-making, and resource sharing? As discussed later, these techniques might well be a means of empowerment only to the extent that
they lead to enhancing the individual's self-efficacy belief (flow # 3,
Fig. 2). But, the mere act of sharing or delegating power, in the relationalbased approach, is inadequate to fully understand the nature and experience of empowerment because it ignores the basic motivational processes
involved in it. As will be seen in the ensuing discussion, these processes
are better and more fully captured when empowerment is interpreted
as a motivational construct.
The social psychological literature provides the basis for interpreting
empowerment as a motivational construct. Here empowerment is used as
a motivational and/or expectancy belief state internal to individuals. For
instance, individuals are assumed to have a need for power (McClelland,
1975) where power connotes an internal urge to influence and control
other people. A related but more inclusive disposition to control and
cope with life events has also been proposed by several psychologists
while dealing with the issues of primary/secondary control (Rothbaum,
Weisz, & Snyder, 1982), internal/external focus of control (Rotter, 1966),
and learned helplessness (Abramson, Garber, & Seligman, 1980).
Individuals feel empowered when they perceive that they can adequately
cope with events, situations, and/or the people they confront. On the
other hand, individuals feel powerless when they believe that they are
unable to cope with the physical and social demands of the environment.
Empowerment in this motivational sense also refers to a belief in
self-determination (Deci, 1975) or a belief in personal self-efficacy
(Bandura, 1986, 1997) as the individual copes with environmental
demands. Any organizational strategy or technique that strengthens this
self-determination or self-efficacy belief of workers will tend to make
them feel empowered at work (flow # 2, Fig. 2) and, consequently,
dealienated. Conversely, any strategy that weakens their self-determination or efficacy belief will increase their sense of personal powerlessness
or alienation in relation to the work context. The Oxford English dictionary meaning of empower as "to enable" also reflects the motivational sense
of empowerment. In contrast to the earlier definition of empowerment as
"delegation" (of authority and resource sharing), the connotation of
"enabling" implies motivation through the enhancement of one's personal efficacy and ability to cope with environmental demands.
The motivation-based approach to empowerment requires two sets
of actions:
1. To identify organizational conditions or contextual factors that
cause employees to feel powerless and self-estranged-resulting
in withdrawal behavior
80
2. To modify such conditions and provide employees with information that enhances personal efficacy, and increase job involvement
and organizational commitment in order to reduce the withdrawal
behaviors
We now deal with the first set of actions. The next section discusses
the Performance Management Process-a potent strategy to modify the
contextual conditions in order to empower employees and thereby reduce
withdrawal behavior.
Contextual Factors Lead to Powerlessness
Conger and Kanungo (1988:477) have identified four categories
of contextual factors that contribute to the lowering of employees' selfefficacy beliefs (See Table 2).
Table 2. Context Factors Leading to a Feeling of Powerlessness that
Contributes to Employee Withdrawal Behavior
Organizational factors
Significant organizational change/transitions
Start-up ventures
Competitive pressures
Impersonal bureaucratic climate
Poor communications/network-forming systems
Highly centralized organizational resources
Supervisory style
Authoritarian (high control)
Negativism (emphasis on failures)
Lack of reason for actions/ consequences
Reward system
Noncontingency (arbitrary reward allocations)
Low incentive value of rewards
Lack of competence-based rewards
Lack of innovation-based rewards
Job design
Lack of role clarity
Lack of training and technical support
Unrealistic goals
Lack of appropriate authority/discretion
Low task variety
Limited participation in programs, meetings, decisions that have a direct impact on job
performance
Lack of appropriate/necessary resources
Lack of network-forming opportunities
Highly established work routines
81
Table 2. (Continued)
Job design
High rule structure
Low advancement opportunities
Lack of meaningful goals/tasks
Limited contact with senior management
We draw on a few factors from each category to illustrate how each has
the potential to lower the employees' self-efficacy beliefs, increase their
feeling of powerlessness, and reduce their job involvement and organizational commitment that contribute to the withdrawal behaviors.
82
Supervisory Style. Bureaucratic organizations and directing members by rigid rules and procedures tend to create inequities in the distribution of organizational power. Both of these characteristics deprive
employees of control and discretion and, thereby, lower the employees'
self-efficacy belief. This is largely because such a style does not favor,
much less foster, open communications.
In such a closed and oppressive climate, it is inevitable that employees
engage in withdrawal behavior such as tardiness, social loafing, and
increased grievances, and avoid the whole set of extrarole behaviors.
Through tardiness and social loafing-provided no adverse consequence,
such as a pay deduction, is attached to it-employees exercise their hidden
power to counterbalance that of the supervisor. The grievances, in particular the baseless ones, are an overt demonstration of employee power, especially if the grievance system is sanctioned by the collective agreement with
the union. One might also view the grievances as a vehicle through which
the employees are demanding a two-way communication in the organization. By avoiding the extrarole behaviors, employees communicate that
they too can play by those same rigid rules; in effect, they are making
a statement: "the extrarole behaviors are not part of my job description."
Reward System. Employees' sense of powerlessness increases
when they do not receive valued rewards; and when the rewards are not
based on employee competence and persistence in innovative behavior
(Sims, 1977; Szilagyi, 1980).
Rewards provide employees with a message on the value of their job,
and their performance. In reality, however, it is a message of what the
organization believes its employees are worth. Rewards are one aspect of
the workplace that directly affects the individual's self-worth. When
rewards do not reflect the job value and their performance, employees
experience serious inequity that leads to pay dissatisfaction. To restore
equity, employees engage in withdrawal behaviors of absenteeism,
reduced performance, and turnover (Brown, 1996; Kanungo & Mendonca,
1997). Since rewards perceived to be unfair are a serious reflection of one's
self-worth, employees' commitment to the organization will also be
reduced which will cause employees to avoid the extra role behaviors.
83
"You get what you reinforce"; therefore, if the performance and extrarole
behaviors are not reinforced, then employees will refrain from engaging
in these behaviors (Luthans & Stajkovic, 1999).
Job Design. When the job involves role ambiguity and conflict, the
employees' self-efficacy beliefs are greatly diminished and their job commitment is considerably weakened.
The direct impact of role ambiguity and conflict is withdrawal from
performance behavior. The task will not be performed not because the
employee does not wish to, but because the employee does not know
what it is that has to be done. The job has two components: what is to be
done and how it is to be done. It is the manager's responsibility to communicate clearly, in unambiguous terms, what is to be done because managers are expected to know how the performance of the job tasks fit or
relate to the successful attainment of the work unit's objectives.
Employees have been hired after ensuring that they have the job relevant
knowledge, skills, and abilities. Therefore, they do not expect to be told
how the job is to be done. Furthermore, employees will also withdraw
from task performance when the job involves role conflict. The reason for
the low self-efficacy and the resulting withdrawal is not because employees do not know how the job is to be done, but because the role conflict
creates uncertainties due to issues like ambiguous overlapping jurisdictions with regard to job tasks, responsibilities, and resources.
The discussion thus far has demonstrated that the phenomenon of
employee withdrawal behavior has its origin in the organizational conditions that lower employees' personal efficacy belief and create a feeling of
powerlessness which, in turn, cause employees to reduce job involvement
and commitment to the organization. The discussion has also examined
the conceptual framework of the empowering experience that provides a
foundation for commitment-oriented management practices. Such practices "commit to people-first values ... clarify the mission and ideology ...
guarantee organizational justice ... create a sense of community ... support
employee development" (Dessler, 1999:66). The Performance Management
Process, described in the next section, incorporates the commitmentoriented management practices needed to reduce employee withdrawal
behaviors.
THE PERFORMANCE MANAGEMENT PROCESS-AN ANTIDOTE
TO EMPLOYEE WITHDRAWAL BEHAVIOR
As can be seen from Fig. 3, the Performance Management Process
outlines a set of essential preconditions (Lawler, Mohrman, & Resnick,
84
1984) that constitute the foundation and rationale to ensure the effective
implementation of the procedural steps that follow. As we describe the
process, we will discuss how the essential conditions and the procedural
steps enhance the employees' self-efficacy belief in the workplace. The
empowerment experienced by employees increases their job involvement
and organizational commitment that prevent withdrawal behavior or
alleviate its adverse effects.
ESSENTIAL PRECONDITIONS
Organizational Work Culture
Employees are a vital resource
Management Commitment
~Communication
~Rewarding
PROCESS
Defining the Job
Job objectives
~
Organizational Objectives
-.
'--
Monitoring Performance
Ongoing feedback
Coaching and mentoring
~Goals
85
Essential Preconditions
86
the job's responsibility to satisfy the specific needs of the customers. The
customer approach enables the employees to experience the significance
of their tasks and promotes, in varying degrees, identification with the
organization's objectives.
This step addresses the factors that lower the employee's self-efficacy
belief-factors such as role ambiguity, role conflict, and lack of meaningful job goals. It gives managers the opportunity to communicate the organization's vision-including its values, mission, and goals-and to clearly
explain how the employee's job tasks fit and contribute to the attainment
of the vision. When employees understand the job's relationship to the
organization's vision and its underlying higher purpose, they internalize
the vision and identify with it because they share the beliefs and values
represented by the vision, and it also enables them to feel a sense of community (Kanter, 1972). Such internalization transforms the employees'
values and beliefs and heightens their sense that their job is a particularly
significant endeavor (Collins & Porras, 1997; Conger & Kanungo, 1998).
The resulting internalization and identification with the organization
greatly enhance the employees' affective and normative commitment,
which increase their willingness to undertake the extrarole behaviors
(Meyer & Allen, 1997). Also facilitated are the employees' ability to understand and their willingness to cope with organizational changes even at
some personal inconvenience or cost (Kanungo & Mendonca, 1996).
Fundamental to this and the other steps as well, is the manager's participative approach that involves the employee in the process of job definition that leads to an agreed understanding of and commitment to the
appropriate job behaviors.
Step 2. The second step of the cycle is the setting of performance standards and the corresponding rewards. The first step spelled out what needs to
be done. The second step specifies how well the tasks must be done. It sets
the standards by which the job performance will be assessed. Goal-setting
theory (Locke & Bryan, 1968), abundantly supported by empirical evidence (Locke, Shaw, Saari, & Latham, 1981), stipulates that employee
performance is greatly enhanced when the assigned goals are difficult,
but attainable; and specific, but appropriate to the objectives of the organization. This step also establishes the performance period, and ensures
that the rewards are contingent on performance and are valued by the
employee (Kanungo & Mendonca, 1988).
The process of performance assessment, despite the most sophisticated attempts to introduce objective procedures, will always remain a
subjective process and, as such, highly vulnerable to fallible judgments
even with all the goodwill in the world. Therefore, the participation of
87
employees in the establishment of standards and measures of performance is also crucial. The knowledge and expertise of both the supervisor
and the employees contribute to performance standards and measures
that are reasonable, realistic, and appropriate. This is not to suggest that
subordinates would set unrealistically low standards. In fact, there is considerable evidence that subordinates often tend to set unrealistically high
goals (Lawler, 1977). The joint mutual-influence process would, in addition to injecting realism, ensure manager-subordinate agreement and
give the subordinate "ownership and control" which go a long way to
generating the trust, acceptance, and commitment so necessary to this
highly subjective system (Locke & Latham, 1984).
In addition to establishing the job objectives and measures of performance, the manager should also identify the needed resources for task
attainment and ensure that these are available to the employee. Managers
should also explore the possible environmental constraints that might
interfere with the employee's job performance, and work to removing
them. If all of the required resources are not available, or the constraints
cannot be removed, then the performance standards would need to be
appropriately adjusted to reflect the fact that job performance at such levels is beyond the control of the employee.
The activities in this step ensure role clarity; realistic and meaningful
goals; performance-contingent and valued rewards; participation in decision making; communications; allocation of the needed resources; and
removal of environmental constraints. As discussed previously, these
activities are most conducive to enhancing the employees' self-efficacy
belief.
Step 3. The third step of the performance management cycle is monitoring the performance. During this phase, the manager provides informal
ongoing feedback, which is to be viewed not in terms of faultfinding but
rather as on-the-job coaching. When managers function as coach, they
seek to help employees grow and reach desired performance levels.
Coaching involves knowing how well the employees are performing relative to performance standards in terms of specific, measurable behaviors, and discussing areas for improvement. The coach gives praise for
work well done and offers constructive criticism when appropriate. In the
latter case, the manager cites specific behaviors with specific suggestions
on how to correct the performance problems.
Performance standards can never be etched in stone; their validity
always assumes a relatively stable environment. The manager, who functions like a coach, will be sensitive to changing environmental factors, and
make suitable adjustments to the performance standards. Perhaps the
88
Step 4. The fourth step is the formal appraisal review at the end of the
predetermined performance period, during which managers record their
assessment of the individual's performance. This phase usually poses the
greatest problem for most managers as it demands that they play two
apparently conflicting roles-as coach and judge. The most frequently
recommended approach to this phase is termed the "problem-solving"
approach because its focus is on the removal of obstacles to good performance-obstacles such as inappropriate or obsolete work procedures,
lack of resources and certain skills, lack of a clear understanding of the job
role and requirements. This approach also encourages a joint discussion
between the manager and the subordinate.
The mutual exchange of information provides a much clearer picture of
the individual's job performance and the context in which it was carried out.
As a result, the subordinates' trust in the fairness of the process increases
because they are now certain that the manager does indeed have all of the
information needed for a reasonable assessment. Furthermore, it provides
an opportunity to discuss the short- and long-term career objectives of the
subordinates, as well as their training and development needs. The problem-solving approach creates a climate of mutual trust, is nonthreatening,
and, therefore, makes the appraisal review the ideal event to discuss and set
goals for the next performance period. In this approach, managers function
as both coach and judge, but the obvious emphasis is on their mentoring role
as they seek to nurture the subordinates' strengths and minimize the negative effects of their weaknesses, if these cannot be completely eliminated.
If the performance review is to aid in performance management, the
assessment recorded should provide information that facilitates equitable
compensation decisions and identifies training and development needs.
When managers ensure equity in compensation decisions, they unleash
the tremendous motivational power of the compensation program with
89
two positive consequences, among many, that inevitably follow. First, the
correct and desirable performance behaviors are reinforced which greatly
increases the probability of such behaviors being repeated in the future
(Luthans & Stajkovic, 1999). The employees get a clear, direct, and unambiguous message of the type and level of performance behaviors expected
from them. Second, the satisfaction that follows equitable compensation
increases the value employees place on the rewards they receive from the
organization. The empirical evidence is overwhelmingly conclusive that
employee motivation is high when rewards are contingent on performance behavior, and are valued by the employees (Kanungo & Mendonca,
1988). One of the fundamental principles underlying this process is procedural justice. Therefore, the proper use of this process throughout the
organization, requires that the individual's performance measures, and
corresponding rewards, cover all facets of the job in relation to the organization's objectives, and duly recognize the support and collaboration of
colleagues who contribute to the individual's successful performance
(Bloom, 1999).
Identifying training and development needs is the necessary, but
not sufficient, first step to improving performance. It must be followed
through with appropriate programs, which remedies specific performance deficiencies or provides opportunities for the acquisition or enhancement of certain skills and abilities. This activity is especially critical to
organizations that compete in a highly dynamic and rapidly growing
industry. Therefore, continuing commitment to employee development
and growth cannot be overemphasized. As an employee observed of the
company's development program, "Federal Express made me a man. It
gave me the confidence and self-esteem to become the person I had the
potential to become" (Dessler, 1999:63).
The Performance Management Process and
the New Work Environment
90
91
organizational and individual levels. It ensures that the contextual factors-organizational conditions, supervisory style, reward system, and
job design-provide employees with an empowering and a reinforcing
experience to increase their job involvement and organizational commitment that is essential to reducing the withdrawal behavior. Developed on
a sound conceptual foundation, the Performance Management Process
provides a systematic approach to remedy the malaise of employee withdrawal behavior-be it in the "ordinary workplace" location or away
from it in telecommuting programs. However, it is recognized that the
evidence, cited in the chapter, relates to specific aspects of withdrawal
behavior-such as the relationship of absenteeism to its antecedent conditions <e.g., Nicholson & Payne, 1987). It does not validate the relationship of withdrawal behaviors to the various stages of the Performance
Management Process suggested in this chapter. This remains the task of
future research.
REFERENCES
Abramson, L. Y., Garber, J., & Seligman, M. E. P. (1980). Learned helplessness in
humans: An attributional analysis. In J. Garber & M. E. P. Seligman (Eds.),
Human helplessness: Theory and applications (pp. 3-34). New York: Academic
Press.
Adams, J. S. (1965). Injustice in social exchange. In L. Berkowitz (Ed.), Advances in
experimental psychology (Vol. 2, pp. 267-299). New York: Academic Press.
Bandura, A. (1986). Social foundations of thought and action: A social-cognitive view.
Englewood, NJ: Prentice-Hall.
Bandura, A. (1997). Self-efficacy: The exercise of control. San Francisco: Freeman.
Baron, J. N., & Kreps, D. M. (1999). Strategic human resources: Frameworks for general
managers. New York: Wiley.
Bishop, J. W., & Scott, K. D. (1997, February). How commitment affects team performance. HRMagazine, pp. 107-111.
Blau, P. M. (1964). Exchange and power in social life. New York: Wiley.
Blauner, R. (1964). Alienation and freedom. Chicago: University of Chicago Press.
Block, P. (1987). The empowered manager. San Francisco: Jossey-Bass.
Bloom, M. (1999). The performance effects of pay dispersion on individuals and
organizations. Academy of Management Journal, 42, 25-40.
Brown, S. P. (1996). A meta-analysis and review of organizational research on job
involvement. Psychological Bulletin, 120,235-255.
Burke, W. (1986). Leadership as empowering others. In S. Srivastra (Ed.), Executive
power (pp. 51-77). San Francisco: Jossey-Bass.
Collins, J., & Porras, J. (1997). Built to last. New York: Harper Business.
Conger, J. A., & Kanungo, R. N. (1988a). The empowerment process: Integrating
theory and practice. The Academy of Management Review, 13(3), 471-482.
92
93
94
Telecommuting and virtual office: Issues and opportunities (pp. 163-185). Hershey,
PA: Idea Group Publishing.
Reichers, A. E. (1985). A review and reconceptualization of organizational commitment. Academy of Management Review, 10, 465-476.
Roethlisberger, F. J., & Dickson, W. J. (1939). Management and the worker.
Cambridge, MA: Harvard University Press.
Rothbaum, F. M., Weisz, J. R., & Snyder, S. S. (1982). Changing the world and
changing self: A two process model of perceived control. Journal of Personality
and Social Psychology, 42, 5-37.
Rotter, J. B. (1966). Generalized expectancies for internal versus external control of
reinforcement. Psychological Monographs, 1 (Whole No. 609).
Sashkin, M. (1984). Participative management is an ethical imperative.
Organizational Dynamics, Spring, 5-22.
Schacht, R. (1970). Alienation. New York: Doubleday.
Scott, K. D., & Taylor, G. S. (1985). An examination of conflicting findings on
the relationship between job satisfaction and absenteeism: A meta-analysis.
Academy of Management Journal, 28, 599-612.
Shepard, J. M. (1971). Automation and alienation. Cambridge, MA: MIT Press.
Sims, H. P. (1977). The leader as a manager of reinforcement contingencies.
In J. G. Hunt & L. L. Larson (Eds.), Leadership: The cutting edge (pp. 121-137).
Carbondale: Southern Illinois University Press.
Somers, M. J. (1995). Organizational commitment, turnover and absenteeism: An
examination of direct and interaction effects. Journal of Organizational Behavior,
16,49-58.
Sommer, S., Bae, S. H., & Luthans, F. (1996). Organizational commitment across
cultures: The impact of antecedents on Korean employees. Human Relations,
49, 977-993.
5
Meaning and Measurement of
Work Role Withdrawal
Current Controversies and Future Fallout from
Changing Information Technology
David A. Harrison
Despite a massive empirical effort, there has been no consensus on appropriate solutions to lateness, absence, or withdrawal problems (Johns,
1997). A vigorous debate about the meaning and measurement of these
behaviors has continued for decades (Blau, 1998; Hanisch & Hulin, 1990,
1991; Hanisch, Hulin, & Roznowski, 1998; Harrison & Martocchio, 1998;
Hulin, 1984; Johns, 1984, 1998; Johns & Nicholson, 1982; Martocchio &
Harrison, 1993; Mobley, 1982; Muchinsky, 1977; Rosse & Miller, 1984). The
purpose of this chapter is to evaluate some of the arguments in that
debate, and to explore the implications of those arguments under conditions of rapidly changing information technology.l begin with a review of
the constitutive definitions, or the stipulated meanings ascribed to singleand multiple-behavior types of withdrawal. I then examine the implications of those meanings for improved measurement. Finally, I predict how
those meanings are likely to evolve as the information revolution severs
the performance of many work tasks from time and place. This change in
work arrangements might necessitate a new look at the ways withdrawal
behaviors are defined and studied.
DAVID A. HARRISON Department of Management and Organization, Pennsylvania
State University, University Park, Pennsylvania 16802
95
96
David A. Harrison
CONSTITUTIVE DEFINITIONS
Withdrawal research, as with most topics of study in the behavioral
sciences, is currently dominated by the hypothetico-deductive approach
(Task Force on Statistical Inference, 1996). Initially, constructs are defined
and linkages between them are proposed before operationalizations are
formed and empirical relationships are estimated (Schwab, 1998). In the
postpositivistic paradigm under which most of this research is conducted,
constructs are not "real." Instead, they are abstractions meant to provide
a more parsimonious representation of the psychological and behavioral
world. Much of the current controversy surrounding the meaning and
measurement of withdrawal stems from disputes over the most appropriate level of such abstraction.
Constructs are initially defined constitutively. That is, the first step that
researchers are urged to take in the hypothetico-deductive approach is to
clearly state what a construct is-to stipulate its meaning (Runkel & McGrath,
1972). Constitutive definitions state the "what" of constructs; theories in
which they are embedded state the "why." Clear constitutive definitions or
stipulated meanings are critical for ensuring comparability of findings
across studies. As Hulin (1984) has noted, however, research on most
forms of withdrawal did not begin with stipulated meanings. Lateness,
absenteeism, and turnover began and stayed for many years as organizational facts or variables in search of underlying constructs and theories
that would help explain them (see Organ's 1988 work on organizational
citizenship behavior for a criterion that started in a different way).
In other words, scientific meanings were generally not a priori considerations in the earlier decades of exploratory studies that preceded
full-blown models of each form of withdrawal (Johns & Nicholson, 1982;
Martocchio & Harrison, 1993). The seriousness of that problem continues
to haunt current investigations, and should not be downplayed (e.g., the
prevalent but rather loose usage of the word "withdrawal" to refer to
anyone of the behaviors discussed, including its use in this chapter).
Constitutive definitions are the foundation for every stage of research that
follows. Lack of attention to constitutive definitions can lead to poor
measurement, and therefore uninterpretable empirical findings (StoneRomero, 1994).
However, withdrawal research is not alone in having a long history
of study before constitutive definitions are given serious consideration.
Most major streams of behavioral science research in general, and of
organizational science research in particular, started with descriptive
work on social phenomena using labels that had been part of the common
vernacular before they were the subject of rigorous investigations (see
97
Katzell & Austin, 1992). Induction from these descriptive findings led to
insights about possible underlying constructs and relationships. One of
the best-known examples of this process is the observation of positive correlations between any two tests of intellectual performance, and the ensuing century of controversy about the constitutive and operational
definitions of intelligence (Carroll, 1993). The construct of job performance has had a similar, although somewhat shorter, scientific history
(Campbell, McHenry, & Wise, 1990). More recently, legislative actions and
judicial rulings regarding sexual harassment have clearly preceded clear
scientific definitions and comprehensive investigations of it (Fitzgerald,
Drasgow, Hulin, Gelfand, & Magley, 1997), although the behavior has
arguably been going on for as long as women have worked with men.
As with any realm of science, constitutive definitions for single- and
multiple-behavior withdrawal constructs are, to a large degree, arbitrary.
One cannot argue about their construct validity, or use data to support or
refute them. On the other hand, one can argue about how useful such definitions are, by referring to the amount of agreement scientists have regarding their stipulated meaning, and by evaluating the degree to which those
definitions support the logic, processes, and goals of scientific inquiry (e.g.,
analysis and synthesis; understanding, prediction, and control). Armed
with solid constitutive definitions of withdrawal behaviors, one can then
argue about the validity of particular operational definitions.
OPERATIONAL DEFINITIONS
Both stipulated and nomological meanings are necessary for knowledge to accrue about any phenomenon. In the second step of the idealized
research sequence mentioned above, investigators develop operational
definitions or measurements for their constitutively defined constructs.
Once developed, researchers can use a variety of data-based tools (e.g.,
multitrait multimethod matrices; Campbell & Fiske, 1959) to refine those
measurements so that they (1) fully span the construct space and (2) provide high-fidelity, low-error numerical representations of it (Hinkin,
1995). Measurements shown to be less susceptible to these dual problems
of deficiency and contamination should have higher construct validity
(Schwab, 1980). Only after operational definitions have been developed
can researchers empirically examine relationships of withdrawal behaviors to a proposed network of other constructs, and thereby come to
conclusions about their nomological meaning(s).
Deficiency and contamination issues are a strong source of contention
in withdrawal research. In current debates, proponents of a multibehavior
98
David A. Harrison
99
100
David A. Harrison
101
102
David A. Harrison
103
104
David A. Harrison
Constitutive Definition. Lateness is constitutively defined as arriving at work after the start, or leaving before the end, of a scheduled workday
(Koslowsky, Sagie, Krausz, & Singer, 1997). The latter part of that definition (leaving early) is probably the only portion that differs from what
might be found in an English dictionary, and what might generate disagreement among withdrawal scholars (e.g., Blau, 1985). Further distinctions are sometimes made about how much time must elapse before an
instance of lateness turns into an absence. For example, Price (1997) followed Chadwick-Jones, Nicholson, and Brown (1982) in suggesting a time
threshold of 4! hours. A different threshold of 2 hours might distinguish
lateness from a "partial absence." Exact time limits may be more a function of organizational rules about sick leave or wages than behavioral
substance, and for that reason might be better handled as an operational
rather than constitutive issue.
The considerable overlap between absence and lateness definitions is
obvious. Both are temporary separations from one's work role. Both lend
themselves readily to observation in work organizations. Both involve not
performing an expected action (working), on the same target (job-related
tasks), in the same context (workplace). As with absence, the lateness definition can be abstracted to lateness propensity or lateness proneness and can
be taken to include behaviors enacted over extended time intervals or
broad sets of social settings (Harrison & Price, 2003, in press). Indeed,
Blau (1994) defines different forms of lateness in terms of patterning over
time, although it is not clear whether the operational or constitutive definitions came first for each form.
105
unit norms for what constitutes arriving late or leaving early, would likely
contribute to variance in that kind of operational definition.
The fact that there are fewer available lateness (than absence) data
has weakened the unspoken mandate to use only records-based measures. In a comprehensive, meta-analytic summary of lateness correlates,
Koslowsky et al. (1997) did not distinguish self-reported versus recordsbased measures, although 20% of the original studies used self-reports.
Moreover, lateness researchers do not have the benefit of published studies devoted to the reliability, stability, or validity of lateness operationalizations (lateness data were included in Chadwick-Jones et aI., 1971, but
they were not the focus of that study). Aside from Blau's (1994) contention
that different forms of lateness represent different behavioral processes,
there are also no compelling arguments about what constitutes the most
valid records-based index of lateness. Most studies appear to use a frequency measure, although time lost measures are possible. In sum, if lateness is to remain a viable, single-behavior criterion in organizational
research, a good deal of psychometric work needs to be done.
Some of that work should compare the convergent validity of selfreports, social reports, and archival records of lateness. All of the suggestions given above for improving self- and social-report instruments for
absence propensity would apply to lateness propensity as well. As a
minor, tolerated, or perhaps even ignored event in some organizations, it
is improbable that respondents clearly perceive and store instances of
lateness behavior in their memories. Hence, recall and judgment errors
will be more likely. There may be a need for stronger memory cues or a
priori questions that begin to activate more effortful cognitive processing
about coming to work late or leaving early (Harrison & McLaughlin,
1996). Although it might generate contamination due to attributional content, it could be helpful to start a series of questions with: "Occasionally,
things happen to almost everyone that make them late for work. Can you
recall events that made you late for work with in the past 3 months? What
were they? For each one, how many times did it happen over the same
3-month period? Because of their relative mildness or even banality, it
may be necessary to use a shorter recall interval for lateness probes than
for questions about other forms of withdrawal. If lateness was a serious
or highly visible problem in a research context (e.g., for flight attendants
or city bus drivers), these suggestions would be less critical. However, it
would also be more likely in such situations that the organization had put
together its own system for monitoring lateness behavior. And, in all cases
for which lateness propensity is specified as a dependent construct, lateness measurement must have a clearly defined time interval that occurs
after the measurement of independent constructs.
106
David A. Harrison
Turnover
107
108
David A. Harrison
Roznowski, & Harrison, 1996). On the other hand, they require moving
out of the realm of linear thinking and correlation coefficients, even
though their general specification is similar to that of a regression model.
Since their introduction to withdrawal researchers in the late 1980s (e.g.,
Morita, Lee, & Mowday, 1989), nearly a dozen studies in the organizational literature have applied event history techniques to turnover, and
many have provided insights into its determinants that would not have
been possible with conventional methods (e.g., Sheridan, 1985). However,
it is disappointing to see that the use of these demonstrably superior
methods remains far outpaced by conventional and more fallible ways to
operationalize turnover propensity in withdrawal research (Somers &
Birnbaum, 1999).
109
deviant work behavior (e.g., Robinson & Bennett, 1995), which is, in turn,
a part of the more general construct of job performance (Van Scotter &
Motowidlo, 1996).
Yet, in each of these substantive domains, progress can be and has
been made on the molar construct without calls for abandoning the
molecular one, even though progress in the molecular approach has been
stunted at various times. The key appears to be keeping compatible levels
of abstraction within each research approach-to keep the same level of
specificity or generality between independent and dependent constructs
(Ajzen & Fishbein, 1977). For example, the demyelinization that occurs in
MS provided immunology researchers with clues about different classes
of T cells and the mechanisms by which they detect foreign tissue. New
knowledge about immune functioning led MS researchers to develop
tailored immune-suppression drugs to help slow the progress of that
disease. In the withdrawal domain, studies have shown that group
absence norms or cultures, which are shared beliefs about legitimate types
and amounts of absences, are consistently related to absence propensity
(Johns, 1997). From these investigations of more specific forms of withdrawal, researchers can learn about the transmission and reception of
normative information about work, or the development and maintenance
of within-group standards for behavior, and use them to build comprehensive models of response thresholds and temporal patterning in generalized forms of withdrawal.
Narrow Is Preferable to Broad
Looking at the controversy from the other perspective, lateness (e.g.,
Blau, 1998), absence (e.g., Martocchio & Harrison, 1993), and turnover
(e.g., Hom & Griffeth, 1995) investigators have been reticent to acknowledge the utility of a generalized withdrawal approach. This has occurred
despite the fact that most of those taking the specific approach have
adopted the "withdrawal" label for their research domain. Using that
descriptor implicitly endorses the oldest proposition-in-use (Johns &
Nicholson, 1982) for each of the single withdrawal behaviors: that they all
stem, partly, from a tendency to distance oneself from dissatisfying work.
Single-behavior advocates have also been slow to recognize the many
theoretical frameworks and empirical studies that posit or support some
form of generalized withdrawal, albeit using other labels. Table 1 shows
"neglect," "avoidance," "disengagement," "noncompliance," and "production deviance" are all labels for similar dimensions of individual work
behavior used in over a dozen different theories, taxonomies, or typologies. A thesaurus-driven literature search might find a dozen more. In
David A. Harrison
110
Construct label
Impression management
Withdrawal (- )
Not working to potential ( - )
Dimensions of job
performance
Responses to job
dissatisfaction
Neglect (-)
Exit (-)
Behavioral response to
(dis)satisfaction
Organizational
withdrawal
Organizational
delinquency
Absenteeism ( - )
Hunt (1996)
Attendance ( + )
Off-task behavior (-)
Kahn (1990)
Calibration of self-in-role
Personal disengagement ( - )
Prediction of job
performance
Organizational citizenship
behavior
Counterproductive
behavior (- )
Compliance ( + )
Production deviance ( - )
Avoidance ( - )
Contextual performance
Job dedication ( + )
groups studies, labels such as "shirking," "free-riding," and "socialloafing" cover similar conceptual territory (Kidwell & Bennett, 1993).
Withdrawal, by any other name, would seem to have the same conceptual
sweetness.
Different Goals
Another wedge driving the single- and multiple-behavior withdrawal camps apart seems to be differential emphasis on one of two different goals of science. Those who prefer the multiple-behavior approach
111
112
David A. Harrison
aspects of their specific work role or minimize time spent on their specific
work tasks while maintaining their current organizational and work role
membership." Job withdrawal is defined as "employees' efforts to remove
themselves from a specific organization and their work role" (Hanisch &
Hulin, 1991:111).
Even after being taken to task for problems with these definitions
(e.g., Johns, 1998; Martocchio & Harrison, 1993), no revision has been
offered (Hanisch et al., 1998). The definitions are not confined to a set of
behaviors. Instead, job attitudes and intentions are stipulated to be part of the
meaning of the withdrawal construct. Under these terms, work withdrawal
is meaningless unless it is accompanied by dissatisfaction. As the cause is
constitutively defined (rather than hypothesized) in the effect, this stipulated meaning does not support the logic of research that would investigate job satisfaction as an antecedent of work withdrawal. The definition
of job withdrawal does not carry a stipulation of its cause, but both definitions give rise to operational definitions that mix attitude, intention,
and behavior variance.
Operational Definition
113
Contamination. Relative to (other) constitutive definitions of withdrawal as behavior, these two measures appear to contain a large amount
of contamination. The desirability of an act is most closely related to attitude toward that act. Attitudes, intentions, and expectations are conceptually distinct from one another and from enacted behavior (Fishbein &
Ajzen, 1975; Fishbein & Stasson, 1990). The authors themselves point out
that many of the questions follow directly from operationalizations of
withdrawal cognitions (see Hom & Griffeth, 1995, for a recent summary),
which they note are hypothesized to be precursors of withdrawal
(Roznowski & Hanisch, 1990) rather than the behavior itself. At least
some portion of the fit of the authors' withdrawal models is likely to be
preordained by the overlap of attributional, and especially attitudinal,
content in these operational definitions (Schwab, 1980; Stone-Romero,
1994).
Common Method Variance. Proponents of the multiple-behavior
approach to withdrawal point to the positive relationships among the
work withdrawal or job withdrawal indicators and note in some cases
that these covariances produce "impressive" internal consistencies (e.g.,
a = .81 for job withdrawal in one sample; Hanisch et al., 1998:473).
Without positive covariation in these operational definitions, there would
be no empirical basis for a general withdrawal construct. Opponents (or
at least critics) of the multiple-behavior approach suggest that this consistency is biased upward by the use of overtly similar questions and
response formats, all asked of the same person at the same time (Blau,
1998; Johns, 1998). Counterarguments (Hanisch et al., 1998) cast the claim
of common method bias as a nebulous one (it is in most cases, but is
better specified in this one), and suggest that it would disallow any
content-based covariation, factor discrimination, or differential relationships between constructs. It does not.
114
David A. Harrison
115
often they think about it, how much they might like to do it, and how
strongly they expect to do it. Although the authors of these instruments
are unwilling to consider perception of covariation as evidence about the
structure of withdrawal (i.e., from multidimensional scaling; Hanisch
et al., 1998: 469), they are willing to consider covariation of perceptions as
unambiguous evidence of consistency within families of withdrawal
behaviors.
A more critical perspective on the data gathered with these measures
might regard them as a mix of content and common method variance.
Table 2 shows the expected reliability estimates one would obtain under
differing levels of (item-item) true score and error correlations, for scales
that have from 3 to 12 items. The range of true-score correlations, .15 ::::::
P'TT :::::: .35, was taken from the values of disattenuated meta-analytic correlations involving lateness, absence, and turnover (e.g., Koslowsky et al.,
1997; Mitra, Jenkins, & Gupta, 1992). The range of error correlations, .05::::::
PEE:::::: .20, was taken from the studies of common method variance listed
6-item scale
9-item scale
12-item scale
PIT = .15
PIT = .20
PIT = .25
PIT = .30
PIT = .35
PIT = .15
PTT = .20
PTT = .25
PTT= .30
PTT= .35
PIT =.15
PIT = .20
PTT= .25
PTT = .30
PIT = .35
PTT=15
PTT= .20
PTT = .25
PTT = .30
PTT = .35
.35
.43
.50
.56
.62
.51
.60
.67
.72
.76
.61
.69
.75
.79
.83
.68
.75
.80
.84
.87
PEE = .05
PEE = .10
PEE =.15
PEE = .20
.43
.50
.56
.62
.67
.60
.67
.50
.56
.62
.67
.71
.67
.56
.62
.67
.71
.75
.72
.72
.76
.80
.83
.75
.79
.83
.86
.88
.80
.84
.87
.89
.91
.76
.80
.83
.86
.79
.83
.86
.88
.90
.84
.87
.89
.91
.92
.62
.67
.71
.75
.79
.76
.80
.83
.86
.88
.83
.86
.88
.90
.91
.87
.89
.91
.92
.94
.76
.80
.69
.75
.79
.83
.86
.75
.80
.84
.87
.89
.72
116
David A. Harrison
To be fair to the authors of the multiple-behavior withdrawal instruments, they have always been candid in stating that their research and
their instruments are works in progress (Hanisch et al., 1998). And, their
constitutive definition of withdrawal as a psychological soup of attitudes,
intentions, and behaviors matches their operational definitions. However,
if the stipulated meaning of withdrawal was that of a process (a sequence
of relationships between constructs) instead of a construct per se, it is
likely there would be much less resistance to the multiple-behavior line of
reasoning and research. A more direct and distinct set of constitutive definitions for withdrawal behaviors could then be stipulated.
Unified Meaning. There may be more agreement between the singleand multiple-behavior approaches to withdrawal than is reflected in the
tenor of recent discussions. If the fairly thin distinction between absence
and lateness is removed, it becomes apparent that both approaches to this
research domain specify two sets of withdrawal behaviors, one temporary
and one permanent. Moreover, these two forms of withdrawal share a conceptual basis that is both clear and at a fairly high level of abstraction. Each
one can be defined as part of a propensity to reduce or withhold inputs from
one's work role. This definition is simple, straightforward, and does not stipulate causes or effects. Use of the word "inputs" helps to rope in some of
the well-articulated concepts and predictions of equity theory into the withdrawal area (e.g., Johns, 1997; Schwarzwald, Koslowsky, & Shalit, 1992). If
the word "contributions" were substituted for "inputs," this underlying
theme could be traced back to original propositions by March and Simon
(1958). A similar kind of conceptual integration has been achieved in the
groups literature, where Kidwell and Bennett (1993) tied "shirking," "freeriding," and "social loafing" together as slightly different manifestations of
the propensity to withhold effort during team performance.
117
Measurement Implications. What if this unifying constitutive definition were applied to the multiple-behavior perspective on withdrawal?
There would still be an opportunity to differentiate temporary and permanent withholding of inputs. A variety of actions in addition to lateness
and absence could be gainfully applied in the operational definition of
temporary withdrawal, including some behaviors that appear in current
theories and in withdrawal-like instruments (e.g., Roznowski, Rosse, &
Miller, 1992; Rusbult et aI., 1988): taking long breaks, missing scheduled
meetings, leaving tasks for someone else to do, or performing any number of "neglectful" off-task behaviors during work hours (Hunt, 1996).
Theories of common method variance would suggest dispersing
questions about those behaviors throughout a larger surveyor interview
instrument, using items that tap unrelated constructs as psychological
"filler" tasks (Harrison et aI., 1996). To the same end of reducing potentially false consistency, the withdrawal questions should have different
response formats, perhaps asking for open-ended percentages and frequencies, as well as conventional forced choice responses that involve
numerical ranges when appropriate (e.g., "once a month"), or adverbs
("never" ... "all the time"). When opportunities for the stimulus behavior
might be limited, or might vary widely across respondents, such as for
"missing scheduled meetings," it would be beneficial to first ask how
many times the behavior was scheduled before finding out how many
times it was missed. The final number to be standardized and added to
the work withdrawal composite would be a ratio of the two (Harrison &
Price, 2003, in press).
Measurement of the behavioral tendency toward permanent withdrawal would be more circumscribed, but might benefit from the same
guidelines. Questions could probe forms of job search behavior (Blau,
1998): "how many times have you contacted other companies about job
openings," " ... checked newspaper or Internet listings of other job opportunities?" Overt behaviors such as " ... asked for a transfer" or " ... told
a co-worker that you planned to leave" would also fit within the constitutive definition.
Time. Current work withdrawal and job withdrawal instruments
have self-report questions about actions, thoughts, attitudes, and intentions spread throughout the past, present, and future. If the measures
suggested above were developed, questions or items should be about the
occurrence of (near) past behavior. Just as all behaviors are defined by
time, all items should have clear time boundaries specified (Atkin &
Goodman, 1984). Finally, as with the casual order concerns raised about
measurements of lateness and absence, these behavioral reports would
118
David A. Harrison
119
Distributed Work
More than simply making communication easier or cheaper, information technology (IT) is allowing work tasks to he done without historically strong constraints on schedule and location (Cascio, 1998). That is,
IT is facilitating the coordination of employee inputs in ways that could
only be accomplished in the past through spatial proximity and face-toface interaction (Agres, Edberg, & Igbaria, 1998). Thus, IT enables a wide
array of flexible work arrangements. Various forms of these arrangements, such as satellite work centers, virtual offices, or telecommuting,
can be characterized as distributed work (Belanger & Collin, 1998).
Distributed work is a term describing a variety of practices that allow
employees and their tasks to be dispersed over settings outside a central
work location or times outside a synchronous work schedule (Lindstrom,
Moberg, & Rapp, 1997). Employees take part in distributed work to the
extent they work remotely from the physical and temporal locations of
their managers, co-workers, or subordinates, but on an interdependent
task. The most well-known and often-researched type of distributed work
is telecommuting, which is almost always done from an employee's home
(Feldman & Gainey, 1997).
Estimates of the number of full-time employees in the United States
working as telecommuters have ranged from 4 million people in 1990 to
11 million in 1997 (Shellenbarger, 1997). Annual growth in telecommuting
has averaged 15%. Conservative projections place the number of telecommuters at 15 million by the end of the year 2000, approaching 15% of the
U.S. work force. By 2010 that number may exceed 25% of full-time
employees (U.S. Department of Commerce, 1998), given expanding bandwidths and faster speeds for digital communication.
Change in the Constitutive Meaning of Withdrawal Behaviors
120
David A. Harrison
121
Two broad classes of jobs are easiest to accommodate under distributed work: those that involve a high division of labor and external control, and those that involve a low division of labor and internal control
(Ford & Butts, 1995). Although withholding of inputs is possible in any
work role, it is the former type of job (e.g., clerical tasks; customer service;
programming) for which distributed work is growing the fastest
(Dickisson, 1997). In these arrangements, the physical separation of
employees from supervisors and co-workers makes them less susceptible
to absence and lateness measures such as time sheets and punch cards
(Handy & Mokhtarian, 1995).
122
David A. Harrison
networked computer. These records can provide supervisors with information about "electronic attendance," work speed, work completed (keystrokes, pages, lines of code), completed forms, correspondence, call logs,
connect time, and errors-as soon as an employee works on a connected
telecommunications system (Ambrose & Alder, 2000). Other forms of IT
can even allow a firm to monitor physical presence in front of a computer
(e.g., a small video camera or "webcam").
Despite legislative and scientific disputes surrounding the intrusiveness of CPM (George, 1996), it is already commonplace in U.S. businesses.
In a survey conducted by the American Management Association, over
two-thirds of its members reported doing some form of electronic
employee surveillance. Over half reported using CPM (Lavelle, 1999).
Some of the critical issues in the disputes are the recording of taskrelevant versus irrelevant information, and disclosure to employees that
monitoring is occurring (Ambrose & Alder, 2000). If these procedural justice hurdles can be overcome, CPM-based records will become ubiquity
for employees both in and out of distributed work arrangements.
With employee permission, observations could be taken from these
records and modified by researchers to use as measures of generic inputs
to one's work role. The same IT that allows bidirectional communication
might also be used to administer short, time-sensitive measures of job attitudes and perceptions in a more-or-Iess continuous fashion (e.g., at the
beginning and end of each work session; Hackett et al., 1989). Such data
would permit unambiguous testing of some of the major within-person
hypotheses regarding withdrawal (Harrison & Martocchio, 1998). A final
benefit of this technology is its ability to detect out-of-bounds responses
and other random errors at the point of measurement, and to immediately
record and structure data files.
Outcome-Based Measures. The other approach to withdrawal
measurement under distributed work might take advantage of the elaborate processes by which a supervisor and employee in such an arrangement are urged to agree on what constitutes good performance, and how
such performance will be evaluated. For example, employees who work
remotely might be asked to sign a "telecommuter agreement." This collaborative contract states exactly what is expected of both employee and
supervisor, covering such items as the regularity of communication, what
constitutes completion of particular objectives, the deliverable quality
and quantity of task products, deadline parameters, the format in which
results will be reported and assessed, and specification of the frequency
and timing of conventional attendance at a central workplace (Cooper,
1996). Although these outcome-based measures would not contain the
123
same level of detail as process-based measures, development of the agreement described above usually requires more careful thought from the
organization about what should be measured for the telecommuter than
for employees in conventional work arrangements (e.g., Dickisson, 1997;
McNerney, 1995).
That is, under telecommuting and other forms of distributed work,
the attendance measurement system (Gardiner, 1992) is likely to be
expanded, or perhaps be aligned completely with the firm's performance
measurement system (Naylor et aI., 1980). Such a change would also be
consistent with a change in the constitutive definition of absenteeism.
Obviously, researchers would need to work closely with firms and
employees to make these weekly or monthly types of records into useful
assessments of withdrawal behavior. Current withdrawal models could
still be tested with powerful within-subject designs, if attitudinal and
perceptual data could be collected at regularly spaced intervals (e.g.,
Westman & Eden, 1997).
Other Sources of Data. It is possible that organizational records created from IT-enabled work will be unavailable or unconnected to
researchers' constitutive definitions of withdrawal. In that case, tacit prescriptions against the use of other sources of data will need to be relaxed
(Johns, 1998). And, given the nature of such work arrangements, social
reports of withdrawal behaviors from co-workers or supervisors would
have little construct validity. These circumstances underscore the importance of developing sound self-report measures of everyday forms of
withdrawal (behaviors, not feelings) using the guidelines suggested
above for self-reports of absence, lateness, or work-role separation.
If withdrawal is redefined as the reduction of inputs to one's work
role, it may be necessary to develop sets of job-specific (tailored to the
context) and generic (portable across contexts) items. Reverse-scored
measures of discretionary (OCB-like) behaviors could fall into either category depending on the jobs being studied, and would be especially sensitive to changes in levels of purported independent variables (they
would also be a source of deficiency in measures based solely on organizational records). Under distributed work, reductions of these kinds of
prosocial inputs to one's work role are the easiest to enact, involve the
lowest cost, and entail the least risk. The potency of interpersonal consequences is reduced by the limited bandwidth of communication offered
by most information technologies. Limited face-to-face interaction, and a
more permeable membrane between work and family demands is also
likely to change the nomological meaning of withdrawal behavior
(Harrison et a1., 2000).
124
David A. Harrison
CONCLUSION
REFERENCES
Agres, c., Edberg, D., & Igbaria, M. (1998). Transformation to virtual societies:
Forces and issues. The Information Society, 14, 71-82.
Ajzen, 1., & Fishbein, M. (1977). Attitude-behavior relations: A theoretical analysis
and review of empirical research. Psychological Bulletin, 84, 888-918.
125
Ajzen, I., & Fishbein, M. (1980). Understanding attitudes and predicting social behavior. Englewood Cliffs, NJ: Prentice-Hall.
Ambrose, M. L., & Alder, G. S. (2000). Designing, implementing, and utilizing
computerized performance monitoring: Enhancing organizational justice. In
G. Ferris (Ed.), Research in personnel and human resources management (Vol. 18,
pp. 187-219). Greenwich, CT: JAI Press.
Atkin, R. 5., & Goodman, P. S. (1984). Methods of defining and measuring absenteeism. In P. S. Goodman & R. S. Atkin (Eds.), Absenteeism: New approaches to
understanding, measuring, and managing employee absence (pp. 47-109). San
Francisco: Jossey-Bass.
Becker, T., & Martin, S. (1995). Trying to look bad at work: Methods and motives
for managing poor impressions in organizations. Academy of Management
Journal, 38,174-199.
Belanger, E, & Collins, R. W. (1998). Distributed work arrangements: A research
framework. The Information Society, 14, 137-152.
Blau, G. J. (1985). Relationship of extrinsic, intrinsic, and demographic predictors
to various types of withdrawal behaviors. Journal of Applied Psychology, 70,
442-450.
Blau, G. J. (1994). Developing and testing a taxonomy of lateness behavior. Journal
of Applied Psychology, 79, 959-970.
Blau, G. J. (1998). On the aggregation of individual withdrawal behaviors into
larger multi-item constructs. Journal of Organizational Behavior, 19, 437-451.
Boye, M. W., & Wasserman, A R. (1996). Predicting counterproductive behavior
among drug store applicants. Journal of Business and Psychology, 10, 337-349.
Campbell, D. T., & Fiske, D. W. (1959). Convergent and discriminant validation by
the multitrait-multimethod matrix. Psychological Bulletin, 56,81-105.
Campbell, J. P., McHenry, J. J., & Wise, L. L. (1990). The modeling of performance
in a population of jobs. Personnel Psychology, 43,313-333.
Campion, M. A (1991). Meaning and measurement of turnover: Comparison of
alternative measures and recommendations for research. Journal of Applied
Psychology, 76,199-212.
Carroll, J. B. (1993). Human cognitive abilities: A survey of factor-analytic studies.
London: Cambridge University Press.
Cascio, W. E (1998). The virtual workplace: A reality now. The IndustrialOrganizational Psychologist, 37, 32-36.
Chadwick-Jones, J. K., Brown, C. A, Nicholson, N., & Sheppard, C. (1971).
Absence measures: Their reliability and stability in an industrial setting.
Personnel Psychology, 24, 463-470.
Chadwick-Jones, J. K., Nicholson, N., & Brown, C. A (1982). The social psychology
of absenteeism. New York: Praeger.
Cheney, M.S., & Harrison, D. A (1999). Uncertainty revisited: Nonlinear, chaotic
changes on the journey toward decision resolution. Paper presented at the national
meeting of the Academy of Management, Chicago.
Clegg, C. W. (1983). Psychology of employee lateness, absence, and turnover:
A methodological critique and an empirical study. Journal of Applied
Psychology, 68, 88-101.
126
David A. Harrison
Cooper, R. C. (1996). Telecommuting: The good, the bad and the particulars.
Supervision, 57, 10-12.
Crampton, S. M., & Wagner, J. A. (1994). Percept-percept inflation in microorganizational research: An investigation of prevalence and effect. Journal of
Applied Psychology, 79, 67-76.
Dalton, D. R., & Todor, W. D. (1993). Turnover, transfer, absenteeism: An interdependent perspective. Journal of Management, 19, 193-219.
Dickisson, K. H. (1997). Telecommuting: Got your homework done? CMA
Magazine, 3-14.
Dickter, D., Roznowski, M., & Harrison, D. A. (1996). Temporal tempering: An
event history analysis of the process of voluntary turnover. Journal of Applied
Psychology, 81, 705-716.
Doty, D. H., & Glick, W. H. (1998). Common methods bias: Does common methods
variance really bias results? Organizational Research Methods, 1(4), 374-406.
Epstein, S., & O'Brien, E. J. (1985). The person-situation debate in historical and
current perspective. Psychological Bulletin, 98, 513-537.
Farrell, D. (1983). Exit, voice, loyalty, and neglect as responses to job dissatisfaction: A multidimensional scaling study. Academy of Management Journal, 26,
596--607.
Feldman, D. c., & Gainey, T. W. (1997). Patterns of telecommuting and their consequences: Framing the research agenda. Human Resource Management Review,
7,369-388.
Fichman, M. (1988). Motivational consequences of absence and attendance:
Proportional hazard estimation of a dynamic motivation model. Journal of
Applied Psychology, 73, 119-134.
Fichman, M. (1989). Attendance makes the heart grow fonder: A hazard rate
approach to modeling attendance. Journal of Applied Psychology, 74,325-335.
Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention, and behavior: An introduction to theory and research. Reading, MA: Addison-Wesley.
Fishbein, M., & Stasson, M. E (1990). The role of desires, self-predictions, and perceived control in the prediction of training session attendance. Journal of
Applied Social Psychology, 20,173-198.
Fisher, C. D., & Locke, E. A. (1992). The new look in job satisfaction research and
theory. In C. J. Cranny, P. C. Smith, & E. E Stone (Eds.), Job satisfaction: How people feel about their jobs and how it affects their performance. New York: Lexington,
165-194.
Fitzgerald, L. E, Drasgow, E, Hulin, C. L., Gelfand, M. J., & Magley, V. J. (1997).
Antecedents and consequences of sexual harassment in organizations: A test
of an integrated model. Journal of Applied Psychology, 82, 578-589.
Ford, R. c., & Butts, M. A. (1995). Questions and answers about telecommuting
programs. Business Horizons, 38, 66-72.
Gardiner, R. C. (1992). Tracking and controlling absenteeism. Public Productivity &
Management Review, 15, 289-307.
George, J. E (1996). Computer-based monitoring: Common perceptions and
empirical results. Management Information Systems Quarterly, 20, 459-480.
127
128
David A. Harrison
Henne, D., & Locke, E. A. (1985). Job dissatisfaction: What are the consequences?
Intemational Journal of Psychology, 20, 221-240.
Hinkin, T. (1995). A review of scale development practices in the study of organizations. Journal of Management, 21, 967-988.
Hogan, J., & Hogan, R. (1989). How to measure employee reliability. Journal of
Applied Psychology, 74, 273-279.
Hom, P. w., & Griffeth, R. W. (1995). Employee tumover. Cincinnati, OH: SouthWestern.
Hulin, C. L. (1984). Suggested directions for defining, measuring, and controlling
absenteeism. In P. S. Goodman & R. S. Atkin (Eds.), Absenteeism: New
approaches to understanding, measuring, and managing employee absence (pp. 391420). San Francisco: Jossey-Bass.
Hulin, C. L. (1991). Adaptation, persistence, and commitment in organizations. In
M. D. Dunnette (Ed.), Handbook of industrial and organizational psychology (2nd
ed., pp. 445-505). New York: Wiley.
Hunt, S. T. (1996). Generic work behavior: An investigation into the dimensions of
entry-level, hourly job performance. Personnel Psychology, 49, 51-83.
Johns, G. (1984). Suggested directions for defining, measuring, and controlling
absenteeism. In P. S. Goodman & R. S. Atkin (Eds.), Absenteeism: New approaches
to understanding, measuring, and managing employee absence (pp. 360-389).
San Francisco: Jossey-Bass.
Johns, G. (1994a). Absenteeism estimates by employees and managers: Divergent
perspectives and self-serving perceptions. Journal of Applied Psychology, 79,
229-239.
Johns, G. (1994b). How often were you absent? A review of the use of self-reported
absence data. Journal of Applied Psychology, 79, 574-59l.
Johns, G. (1997). Contemporary research on absence from work: Correlates,
causes, and consequences. International Review of Industrial and Organizational
Psychology, 12, 115-173.
Johns, G. (1998). Aggregation or aggravation? The relative merits of a broad withdrawal construct. Journal of Organizational Behavior, 19,453--462.
Johns, G., & Nicholson, N. (1982). The meanings of absence: New strategies for
theory and research. In B. M. Staw & L. L. Cummings (Eds.), Research in organizational behavior (Vol. 4, pp. 127-172). Greenwich, CT: JAI Press.
Judge, T. A., & Watanabe, S. (1995). Is the past prologue?: A test of Ghiselli's Hobo
Syndrome. Journal of Management, 21, 211-229.
Kahn, W. A. (1990). Psychological conditions of personal engagement and disengagement at work. Academy of Management Journal, 33, 692-724.
Katzell, R. A., & Austin, J. T. (1992). From then to now: The development of industrial-organizational psychology in the United States. Journal of Applied
Psychology, 77, 803-835.
Kidwell, R. E., Jr., & Bennett, N. (1993). Employee propensity to withhold effort: A
conceptual model to intersect three avenues of research. Academy of
Management Review, 18,429-456.
Koslowsky, M., Sagie, A., Krausz, M., & Singer, A. D. (1997). Correlates of
employee lateness: Some theoretical considerations. Journal of Applied
Psychology, 82, 79-88.
129
Lavelle, L. (1999, April 16). Survey shows increase in firms that monitor employees' communications. The Record. Hackensack, NJ: Knight-Ridder Tribune
Business News.
Leonard, c., Dolan, S. L., & Arsenault, A. (1990). Longitudinal examination of the
stability and variability of two common measures of absence. Journal of
Occupational Psychology, 63, 309-316.
Lindstrom, J., Moberg, A., & Rapp, B. (1997). On the classification of telework.
European Journal of Information Systems, 6, 243-255.
Mabe, P. A., III, & West, S. G. (1982). Validity of self-evaluation of ability: A review
and meta-analysis. Journal of Applied Psychology, 67, 280-296.
Mathieu J. E. & Zajac D. M. (1990). A review and meta-analysis of the antecedents,
correlates, and consequences of organizational commitment. Psychological
Bulletin, 108, 171-194.
McCloskey, D. W., & Igbaria, M. (1998). A review of the empirical research on
telecommuting and directions for future research. In M. Igbaria & M. Tan
(Eds.), The virtual workplace (pp. 338-358). Hershey, PA: Idea Group Publishing.
McNerney, D. J. (1995). Telecommuting: An idea whose time has come. HR Focus,
72(11), 1-4.
Mangione, T., & Quinn, R. (1975). Job satisfaction, counterproductive behavior,
and drug use at work. Journal of Applied Psychology, 60,114-116.
March, J. G., & Simon, H. A. (1958). Organizations. New York: Wiley.
Martocchio, J. J., & Harrison, D. A. (1993). To be there or not to be there?
Questions, theories and methods in absenteeism research. In K. Rowland &
G. Ferris (Eds.), Research in personnel and human resources management (Vol. 11,
pp. 259-329. Greenwich, CT: JAI Press.
Miner, M. G. (1977). Job absence and turnover: A new source of data. Monthly
Labor Review, 100, 24-31.
Mitra, A., Jenkins, G. D., & Gupta, N. (1992). A meta-analytic review of the relationship between absence and turnover. Journal of Applied Psychology, 77, 879-889.
Mobley, W. H. (1982). Some unanswered questions in turnover and withdrawal
research. Academy of Management Review, 7, 111-116.
Morita, J. G., Lee, T. W., & Mowday, R. T. (1989). Introducing survival analysis to
organizational researchers: A selected application to turnover research.
Journal of Applied Psychology, 74, 280-292.
Muchinsky, P. M. (1977). Employee absenteeism: A review of the literature. Journal
of Vocational Behavior, 10, 316-340.
Naylor, J. c., Pritchard, R. T., & Ilgen, D. R. (1980). A theory of behavior in organizations. New York: Academic Press.
Nicholson, N., & Payne, R. (1987). Absence from work: Explanations and attributions. Applied Psychology: An International Review, 36(2), 121-132.
Organ, D. W. (1988). Organizational citizenship behavior: The good soldier syndrome.
Lexington, MA: Heath.
Ostroff, c., & Harrison, D. A. (1999). Meta-analysis, level of analysis, and best estimates of population correlations: Cautions for interpreting meta-analytic
results in organizational behavior. Journal of Applied Psychology, 84, 260-270.
Podsakoff, P. M., & Organ, D. W. (1986). Self-reports in organizational research:
Problems and prospects. Journal of Management, 12, 531-544.
130
David A. Harrison
of Manpower, 18,303-558.
Puffer, S. M. (1987). Prosocial behavior, noncompliant behavior, and work performance among commission salespeople. Journal of Applied Psychology, 72,
615-62l.
Robinson, S. L., & Bennett, R. J. (1995). A typology of deviant workplace behaviors: A multidimensional scaling study. Academy of Management Journal, 38,
555-572.
Roznowski, M., & Hulin, C. L. (1992). The scientific merit of valid measures of
general constructs with special reference to job satisfaction and job withdrawal. In C. J. Cranny, P. C. Smith, & E. F. Stone (Eds.), Job satisfaction:
Advances in theory and research. New York: Free Press.
Roznowski, M., Rosse, J. G., & Miller, H. E. (1992). On the utility of broad-band measures of employee behavior: The case for employee adaptation and citizenship. Paper
presented at the annual meeting of the Academy of Management, Las Vegas.
Runkel, P. J., & McGrath, J. E. (1972). Research on human behavior: A systematic guide
to method. New York: Holt, Rinehart & Winston.
Rusbult, C. E., Farrell, D., Rogers, G., & Mainous, A. G. (1988). Impact of exchange
variables on exit voice, loyalty, and neglect: An integrative model of responses
to declining job dissatisfaction. Academy of Management Journal, 31, 599-627.
Rushton, J. P., Brainerd, C. J., & Pressley, M. (1983). Behavioral development and
construct validity: The principle of aggregation. Psychological Bulletin, 94, 18-38.
Schwab, D. P. (1980). Construct validity in organizational behavior. In
L. L. Cummings & B. M. Staw (Eds.), Research in organizational behavior (Vol. 2,
pp. 3-43). Greenwich, CT: JAI Press.
Schwab, D. P. (1998). Research methods for the organizational sciences. Hillsdale, Nl:
Erlbaum.
Schwarz, N. (1999). Self-reports: How the questions shape the answers. American
Schwarzwald, J., Koslowsky, M., & Shalit, B. (1992). A field study of employees'
attitudes and behaviors after promotion decisions. Journal of Applied
131
Shellenbarger, S. (1997, August 20). Madison Avenue may need to alter image of
'90s telecommuter. The Wall Street Journal, p. B1.
Sheridan, J. E. (1985). A catastrophe model of employee withdrawal leading to low
job performance, high absenteeism, and job turnover during the first year of
employment. Academy of Management Journal, 28, 88-109.
Sickness absence monitoring and control: A survey of practice. (1994, September).
6
Developing and Testing a
Taxonomy of Lateness Behavior
Gary Blau
133
134
Gary Blau
Relationship to
other behaviors
Antecedents of such
lateness
Increasing
chronic
Nonrandom,
increasing,
and increasing
Stable
periodic
Nonrandom, stable,
and stable
Unavoidable
Random
Positive to voluntary
absence, voluntary
turnover, leaving
work early
Independent of voluntary
absence, voluntary
turnover, leaving work
early
Positive to nonwork
lateness
Category
Transportation
Bad weather
Personal illness or
accident
135
136
Gary Blau
137
138
Gary Blau
METHOD
Hospital Sample and Procedure
A questionnaire was constructed and administered in October 1990
on a voluntary basis to all nonexempt hospital employees for whom daily
attendance records were kept. This hospital, located in a large Eastern
city, offers patient care in many medical specialties. Sample respondents'
jobs included secretary, clerk, receptionist, lab technician, maintenance
worker, housekeeper, bill processor, and food service worker. Hospital
management was concerned about employee lateness and supported the
study. Union approval was also obtained. Subjects were given time at
work to fill out the survey. The hospital's human resource department
assisted with the survey administration across all departments and shifts.
Complete confidentiality of subject responses was maintained and participants mailed their survey directly to the investigator in a preaddressed
business reply envelope.
Of the 641 surveys distributed, 509 (79%) were voluntarily completed
and mailed to the investigator. Two follow-up letters helped to increase
subject participation. Of the 509 subjects, 483 (95%) were willing to give
the last four digits of their social security number so that subsequent lateness and other behavioral data could be collected. Eighteen months after
the survey (April 1992), record-based data (i.e., lateness, absence,
turnover, and leaving work early) were collected. Of these 483 subjects,
102 had no recorded lateness incidents over the 18-month period.
A demographic breakdown of the 381 subjects with lateness incidents
indicated that: the average age was 34 years; 71 % were women; 55% were
married; 47% were black, 39% were white, and 14% were either Hispanic
or Asian; 70% had at least a high school degree; 64% worked on the day
shift; average organizational tenure was about 11 years; and 42% had one
or more children living at horne. A comparison of these demographics to
the 102 subjects with no lateness incidents revealed that significantly less
"never late" participants (28%) had children living at home.
Bank Sample and Procedure
A questionnaire was constructed and administered in September
1990 on a voluntary basis to 844 nonexempt bank employees for whom
daily attendance records were kept. This nonunionized bank is headquartered in a large Eastern city. Sample respondents' jobs included teller,
data entry clerk, secretary, receptionist, mail clerk, check processor, and
maintenance worker. Bank management was concerned about employee
139
lateness and supported the study. Subjects were given time at work to fill
out the survey. Supervisory personnel assisted with the survey administration at each work site (e.g., branch or department). All 66 branches participated in this study. Complete confidentiality of subject responses was
maintained, and participants mailed their survey directly to the investigator in a preaddressed business reply envelope.
Of the 844 surveys distributed, 720 (85%) were voluntarily completed
and mailed to the investigator. Two follow-up letters helped to increase
subject participation. Of the 720 subjects, 619 (86%) were willing to give
the last four digits of their social security number so that subsequent lateness and other behavioral data could be collected. Eighteen months after
the survey (March 1992), record-based data (i.e., lateness, absence,
turnover, and leaving work early) were collected. Of these 619 subjects,
171 had no recorded lateness incidents over the 18-month period.
A demographic breakdown of the 448 subjects with lateness incidents
indicated that: the average age was 37 years; 63% were women; 61 % were
married; 58% were white, 25% were black, and 17% were either Hispanic or
Asian; 81 % had at least a high school degree; average organizational tenure
was about 9 years; and 40% had one or more children living at home.
A comparison of these demographics to the 171 subjects with no lateness
incidents indicated that the "never late" sample had significantly higher
organizational tenure (12 years) and less children (25%) living at home.
Following the same survey and record-based data collection procedures
across the hospital and bank samples will allow for stronger comparison.
Measures
140
Gary Blau
incidents. Part of the reason for both organizations approving this study
was to increase supervisors' awareness about employee lateness, including the application of organizational policy for dealing with lateness.
Rosse and Hulin (1985) found a similar negligent supervisory attitude in
recording employee lateness incidents and applying company policy.
To help overcome this lack of diligence in recording lateness and to
allow a stronger application of the pattern, frequency, and duration decision rules for determining the type of lateness behavior an employee
exhibited, bank and hospital personnel agreed that a longer (i.e., 18
month) time frame should be used. For example, after 12 months when
employee records were wiped clean, both organizations wanted to know
if (and by how much) employee lateness behavior increased. Analysis of
12-month versus 18-month lateness behavior for 100 randomly selected
participants across both organizations indicated a significant increase in
overall lateness behavior. This suggests that some employees may "work
the system."
For the hospital, lateness was defined as when an employee reported
to work at least 8 min past the beginning of his or her shift. Over a I-year
period, a certain number of lateness incidents led to the following penalties: 3 incidents-general counseling; 5 incidents-verbal warning
with documentation; 9 incidents-written warning; 13 incidents-3
working days' pay suspension; 15 incidents-5 working days' pay suspension; and 16 incidents-job termination. The hospital's policy also
stated that employees would be docked wages in IS-min increments
beginning 8 min after start time as follows: 8 to 15 min late-IS min
docked; 16 to 30min late-30min docked; 31 to 45min late-45min
docked; and so on.
For the bank, lateness was defined as when an individual reported
to work at least 5 min past the beginning of his or her shift. The bank
used the following incident to penalty schedule over a 12-month period:
4 incidents-general counseling; 6 incidents-verbal warning with
documentation; 8 incidents-written warning; 11 incidents-5 working
days' pay suspension; and 14 incidents-job termination. The bank's
policy also stated that employees would be docked wages in 30-min
increments as follows: 5 to 30 min late-3D min docked; 31 to 60 min late60min docked; and so on. Thus, the bank's lateness policy was more
punitive.
Employees were initially classified as having either a nonrandom or a
random lateness pattern. Nonrandom was defined as when at least twothirds (>66%) of an individual's lateness incidents occurred at the beginning of a workweek (usually Monday) or after a legal holiday. Random
was defined as when less than one-third 33%) of an individual's
141
142
Gary Blau
top-third versus bottom-third subjects on a personality measure) to represent "truer" personality distinctions (Blass, 1977).
This approach also acknowledges the dynamic nature of lateness for
individuals (i.e., that employees can display more than one category of
lateness in a given period). The middle-third group of subjects on pattern,
frequency, and duration should best represent employees with varying
types of lateness. This dynamic lateness category group is defined by onethird to two-thirds of the lateness incidents occurring at the beginning of
a workweek or after a legal holiday and where frequency and duration
increased 4 to 6 of the 18 months. This dynamic lateness group will be
compared to the three previously defined lateness category groups:
increasing chronic, stable periodic, and unavoidable. Because the
dynamic lateness category represents a mix of the other three lateness
behaviors (increasing chronic, stable periodic, and unavoidable), it is not
possible to a priori hypothesize relationships of dynamic lateness to other
work behaviors or antecedents. Therefore, dynamic lateness was not
included in Table 1.
Using these decision rules, five raters coded the lateness behaviors of
the study participants in each sample. For the hospital sample, 353 out of
381 (93%) subjects clearly fell into one of the four categories as follows:
increasing chronic lateness (n = 59); stable periodic lateness (n = 70);
unavoidable lateness (n = 104); and dynamic lateness (n = 120). The
remaining 28 subjects met only one of the frequency or duration cutoffs
and were excluded from further analysis. For the bank sample, 402 out of
448 (90%) subjects clearly fell into one of the four categories as follows:
increasing chronic lateness (n = 66); stable periodic work lateness (n = 79);
unavoidable lateness (n = 125); and dynamic lateness (n = 132). The
remaining 46 subjects met only one of the frequency or duration cutoffs
and were excluded from further analysis.
To check rater reliability, all five raters coded a common group of 50
randomly selected participants into the four latensss categories. Interrater
reliability was assessed using coefficient kappa, which estimates the
agreement between pairs of raters while correcting for chance agreement
(Cohen, 1960). With five raters there are 10 between-rater comparisons.
The average kappa coefficient for these comparisons was .87, and the
range was .97 to .81. Given the four types of lateness, three dummy coded
variables were created for later canonical analysis.
143
Turnover. From hospital and bank records, frequency counts for the
number of subjects who voluntarily left their organization after 18 months
were recorded (1 = stay and 2 = leave). For the hospital, there were 247
(70%) stayers and 106 (30%) leavers. Of these 106leavers, exit interviews
indicated that 23 said that they left for "unavoidable" (Abelson, 1987)
non-job-related reasons (e.g., family related or medical). For the bank,
there were 288 (72%) stayers and 114 (28%) leavers. Of these 114 leavers,
exit interviews revealed 29 "unavoidable" leavers. On the basis of
Abelson's (1987) research, these unavoidable leavers were deleted from
each sample when analyzing the turnover data.
Leaving Work Early. Because daily attendance records were kept
for nonexempt employees at the bank and hospital, the number of times
a subject left work early without permission was tabulated.
Nonwork Lateness. In the survey, subjects were asked to retrospectively indicate the number of times they were late over the last 18 months
in each of the following nonwork categories: (a) religious (e.g., mass or
wedding); (b) volunteer related (e.g., PTA or charity work); (c)
family-social (e.g., picnic or party); and (d) community related (e.g., town
meeting, athletic event, garden club, or political event). The following
response scale was used: 1 = never (0 times); 2 = rarely (1 or 2 times);
3 = occasionally (3 to 5 times); 4 = frequently (6 to 9 times); and 5 = very frequently (at least 10 times). A frequency range is given within each verbal
anchor under the assumption that subjects are more likely to correctly
recall an approximate (versus exact) number of times they were late to an
event category. Across both samples 40 spouses were willing to answer
(after getting subject permission) this three-item nonwork lateness measure about their participating spouse. The resulting correlation of .69
between subjects and spouses suggests that the participants' retrospective
nonwork lateness self-reports are reliable.
144
Gary Blau
145
frame. Mean levels of unavoidable lateness were compared for the badweather versus other months.
Canonical analysis was initially done to examine the overall relationship of the dummy-coded lateness behavior categories to the antecedents.
Table 2 presents the correlations of study variables with canonical variates
broken down by sample. With the Bartlett residual test (Barcikowski &
Stevens, 1975), three significant canonical correlations (R)-representing
uncorrelated antecedent variable-lateness behavior composites-were
found for each sample. The pattern of canonical variate-variable correlations indicates that lower job satisfaction, job involvement, and organizational commitment lead to increasing chronic lateness; higher
leisure-income trade-off and work-family conflict are related to more stable periodic lateness; and transportation concerns, and to a lesser extent
personal illness or accidents, lead to unavoidable lateness. Overall, these
results support the proposed lateness taxonomy. Given the N to K (subjects to variables) ratio for both samples, these canonical correlations
should be stable (Barcikowski & Stevens, 1975).
Means, standard deviations, reliabilities, and correlations among
study variables are shown for the hospital sample and the bank sample
(see Table 3). As noted earlier, 353 of the 381 hospital employees and 402
of the 448 bank employees could be coded into one of the four lateness
categories. The lower variable means on the work attitudes for the bank
sample suggest that these employees are somewhat less committed to
their work than the hospital employees. The self-report scales, in general,
146
Gary Blau
Hospital sample
First
Bank sample
Second
Third
First
Second
Third
-.51
-.46
-.43
-.21
-.13
-.18
-.14
-.11
-.15
-.53
-.44
-.45
-.17
-.20
-.22
-.05
.18
.20
.34
.37
.15
.08
.08
.19
.39
.22
.27
.10
.12
.17
.06
.05
.40
.38
.13
-.01
.04
.16
.21
.32
.23
.35
.15
.21
.47
.49
.27
.23
.52
.18
.19
.17
.50
.11
-.04
.09
.45
.24
.08
R! = .43**
.10
.26
.40
.13
R2 = .35*
R3 = .30*
.11
-.11
-.09
147
Hypothesis 1 (increasing chronic lateness will have a stronger positive relationship to voluntary absence, voluntary turnover, and leaving
work early than stable periodic lateness) was supported across both samples for voluntary absence and leaving early but not for turnover. As
shown in Table 3, significant differences in correlations (McNemar, 1969)
were found (p < .05, one-tailed) for these relationships.
Hypothesis 2 (a significant positive relationship will be found
between unavoidable lateness and nonwork lateness across both samples)
was supported. Hypothesis 3 (significant negative relationships will be
found between job satisfaction, job involvement, and organizational commitment to increasing chronic lateness across both samples) was supported. Hypothesis 4 (significant positive relationships will be found
between leisure-income trade-off and work-family conflict to stable periodic lateness for both samples) was supported. Hypothesis 5 (a significant
positive relationship between transportation concerns and unavoidable
lateness will be found) was supported for both samples. Personal accident
or illness was positively related to unavoidable lateness for the bank sample but not the hospital sample. Correlation results with only the first 12
months of lateness data were consistent with the results cited above.
To test the relationship between bad weather and unavoidable lateness, a t test was done between winter months (i.e., November through
mid-March) versus the other months to see if unavoidable lateness was
higher during the winter months. The results did not support this
hypothesis. The mean minutes level of unavoidable lateness during winter months was 219.4 versus 217.7 for the other months (t = 1.18, P > .05)
for the hospital sample and 219.1 versus 215.5 (t = 1.47, P > .05) for the
bank sample.
Tables 4 and 5 show the mean levels of study variables for both samples, broken down by lateness category. Consistent with prior lateness
research, the never late group is also shown. An analysis of variance was
used to test for significant overall mean differences on each variable; this
was followed by the Duncan multiple range test to examine the differences between pairs of means at the .01 level of significance. The Duncan
multiple range test is favored over the least significant difference test
(Nie, Hull, Jenkins, Steinbrenner, & Bent, 1975) because it uses a different
148
Gary Blau
Table 3. Means, Standard Deviations, Reliabilities, and Correlations
Hospital sample
Bank sample
M
SD
.86
.85
73.8
28.7
11.9
6.8
.87
.88
7.5
3.1
.84
.73
33.6
10.8
7.7
3.3
.77
.75
.38-.20"
.41"
-.24"
-.16-
24.3
4.6
.81
25.7
4.9
.80
-.25"
-.17"
-.19"
.14"
9.6
2.4
.70
8.3
2.1
.69
-.07
-.03
0.2
.11"
9.1
2.2
.61
8.2
2.0
.65
.04
.01
-.04
.05
255.3'
68.5
NA
231.8'
64.2
NA
-.39"
-.41-
-.38"
.15
217.4'
71.7
NA
212.1'
63.4
NA
-.15
-.17
-.19
.35-
218.6'
73.3
NA
217.3'
67.1
NA
-.11
-.12
-.10
.16
230.4'
85.8
NA
222.5'
88.3
NA
-.19-
-.22-
-.18-
.17
8.3
3.2
NA
7.1
2.5
NA
-.24-
-.20-
-.17"
.19"
1.3'
13.5
12.1
0.6
4.6
4.3
NA
NA
.75
1.3'
12.1
11.6
0.6
3.8
4.2
NA
NA
.72
-.22-.25-.06
-.23-.21-.05
-.14"
-.16"
.02
.10
.15"
.12-
Variable
SD
1. Job satisfaction
2. Organizational
commitment
3. Job involvement
4. Leisure-income
trade-off
5. Work-family
conflict
6. Transportation
concerns
7. Personal illness
or accident
8. Increase chronic
lateness'
9. Stable periodic
lateness'
10. Unavoidable
lateness
11. Dynamic
lateness'
12. Voluntary
absence
13. Turnover
14. Leaving early
15. Nonwork
lateness
75.6
30.9
12.4
7.2
35.1
12.2
'"
'"
.41"
.39-
3
.36"
.39"
4
-.18"
-.16-.13"
Note . Correlations for the hospital sample are below the diagonal, and correlations for the bank sample
Reported in minutes based on ~(Frequency " Duration).
'N = 59 for increase chronic lateness, N = 70 for stable periodic lateness, N = 104 for unavoidable
'N = 66 for increase chronic lateness, N = 79 for stable periodic lateness, N = 125 for unavoidable
dNA = not applicable because there were separate groups of subjects within each lateness category.
e Significant difference in correlations, p < .05, one-tailed test.
" p<.05.
range value for different size subgroups. The results in Tables 4 and 5 provide further support for the study hypotheses. Increasing chronic late
subjects exhibited significantly higher voluntary absence and leaving
early behavior and had significantly lower job satisfaction, organizational
commitment, and job involvement than subjects in the other lateness
categories. Stable periodic late subjects had significantly higher
leisure-income trade-off and work-family conflict, whereas unavoidably
late subjects had a significantly higher level of nonwork late behavior and
transportation concerns (hospital sample only).
Perhaps most interesting is the comparison of mean level variables
between the never late versus lateness categories. The results across both
149
10
11
12
-.17
-.18
-.08
-.13
-.19*
-.18*
-.14
.32*
-.10
.16
.10
.31*
.07
.11
-.14*
-.15*
-.05
-.07
-.06
-.02
-.45*
-.38*
-.11*
.10*
-.01
.08
-.05
.01
-.35*
-.14
.12
.09
.14
.09
13
14
-.23*
-.17*
-.26*
-.20*
-.24*
-.16*
.03
.04
-.20*
.15
-.15*
.13*
-.17*
.07
-.19*
.18*
.02
.09
.12
.19*
.11*
.14*
.15*
-.03
.13
.27*
.08
.04
-.03
.11*
.05
.07
.29*
.11
.09
.05
.10*
.12*
NA d
NA d
NAd
.42*'
.19
.38*'
.01
NA d
NAd
.16'
.13
.14'
.04
NAd
.12
.08
.06
.34*
.20*
.17
.16
.13
.31 *
.29*
.10*
.07
.15*
.13
.08
.09
.28*
.12
.07
NAd
.14
.25*
.18
NAd
NAd
.20*
.12
.09
NA d
NA d
NA d
.18*
.08
.11*
.39*'
.12e
.10
.20*
.11 *
.24*
.09
.03
.13*
.06
-.02
.10*
.07
.20
.43*'
.03
.14
.13e
.08
.04
-.01
.32*
.10
.18*
.15*
.30*
.26*
.14*
.15*
.10*
.07
15
.02
-.11*
-.09
samples show that the never late participants had significantly higher job
satisfaction, organizational commitment, and job involvement and significantly lower leisure-income trade-off, work-family conflict, personal
illness or accident, voluntary absence, and leaving work early. These
patterns of results are supportive of the lateness taxonomy and reinforce
the importance of antecedent and work behavior variables for distinguishing late versus never late behavior.
The demographic difference of never late subjects being less likely to
have dependent children at horne would help to explain some of these
results, including the lower levels of work-family conflict, leisure-income
trade-off, and leaving work early. Both samples in this study consisted
Gary Blau
150
SO
Organizational
commitment
M
SO
Job involvement
M
SO
Leisure-income
trade-off
M
SO
Work-family
conflict
M
SO
Transportation
concerns
M
SO
Personal illness or
accident
M
SD
Voluntary
absence
M
SO
Turnover
M
SO
Leaving early
M
SO
Nonwork late
M
SO
Chronic
Periodic
(N=59)
(N= 70)
Unavoidable
(N= 104)
Dynamic
Never late
(N= 120)
(N=102)
67.3_
11.9
78.4 e
12.9
76.1 e
12.4
81.2d
13.1
5.23**
12.3
25.4_
6.6
30.2b
7.2
32.3b
7.1
30.7b
7.3
35.1e
7.4
3.86*
29.8_
7.1
34.~
35.7b
7.6
36.2b
7.8
41.6e
7.5
4.11*
11.5_
3.1
14.~
12.3_
3.0
11.9_
2.8
7.4e
3.1
3.22*
3.3
22.7_
4.8
29.2b
4.9
24.1_
4.7
23.8_
4.4
20.1e
4.3
4.37**
9.1_
2.2
8.9_
2.3
11.5b
2.6
9.4.
2.5
8.3.
2.4
2.15*
8.8.
2.3
8.7.
2.2
9.3.
2.1
9.1.
2.2
3.~
4.68**
2.3
11.5.
3.6
8.1b
3.0
7.9b
3.2
5.1 e
2.0
6.02**
3.1
1.7
0.6
1.4
0.5
1.2
0.5
1.1
0.4
1.37
0.4
18.5.
4.9
15.9b
4.8
12.5e
4.4
12.3e
4.5
7.3d
3.1
9.22**
10.3.
4.5
11.6.
4.2
13.9b
4.4
12.0_
4.3
9.5_
2.7
3.53*
73.~
7.2
1.1
7.~
Note. Across each row, cell means that do not share the same subscript are significantly
different at the .01 level by Duncan's multiple range test.
p< .05. p<.01.
151
Chronic
(N= 66)
Periodic
(N= 79)
Unavoidable
(N = 125)
Dynamic
(N= 132)
Never late
(N = 171)
67.1 a
12.0
72.9 b
12.1
74.1b
11.8
74.3 b
11.9
80.3 c
4.77**
25.9a
6.5
29.2b
6.7
30.6b
7.0
28.4b
6.9
34.0 c
30.3a
7.9
33.1b
7.8
34.5 b
7.7
33.9 b
7.5
37.7c
9.8a
3.6
12.6b
3.5
9.9a
3.2
10.6a
3.3
6.7c
25.7a
5.0
28.3 b
5.1
24.2 c
24.5 c
20.2c
4.6
4.8
4.4
8.0a
2.2
7.8a
2.1
9.2a
2.2
8.1 a
2.0
6.8b
1.9
2.29*
8.1 a
2.0
7.3a
1.9
8.2a
2.1
8.4 a
1.9
3.8b
1.4
4.46*
9.5a
2.7
7.0b
2.6
6.6 b
2.4
6.7b
2.5
4.9 c
6.35**
1.6
0.6
1.3
0.5
1.1
0.4
1.2
0.5
1.1
0.4
1.09
15.5a
4.0
12.3b
3.9
11.2b
3.7
11.5b
3.8
6.9c
8.97**
1O.5a
4.1
11.4a
4.2
13. 7b
4.3
10.9a
4.2
9.8a
3.7
Job satisfaction
M
SD
10.8
Organizational
commitment
M
SD
4.06**
7.1
Job involvement
M
SD
4.50**
8.0
Leisure-income
trade-off
M
SD
5.51**
2.6
Work-family
conflict
M
SD
5.13**
Transportation
concern
M
SD
Personal illness or
accident
M
SD
Voluntary absence
M
SD
1.8
Turnover
M
SD
Leaving early
M
SD
2.6
Nonwork late
M
SD
3.39*
Note. Across each row, cell means that do not share the same subscript are significantly
different at the .01 level by Duncan's multiple range test.
p< .05. p<.OJ.
152
Gary Blau
The results of this study empirically demonstrate, across two samples, that three distinct categories of lateness behavior can be operationalized by pattern, frequency, and duration. By isolating such lateness
categories, it was argued earlier that stronger antecedent-lateness relationships could be found than in previous research that did not break
down general lateness behavior into specific categories. For example, job
involvement accounted for 14% (hospital) and 12% (bank) of increasing
chronic lateness behavior versus other studies (e.g., Beehr & Gupta, 1978;
Cummings & Manring, 1977) where 1% to 8% of lateness has been
accounted for Organizational commitment accounted for 17% (hospital)
and 14% (bank) of increasing chronic lateness behavior versus other studies (e.g., Blau, 1986; Clegg, 1983) where 4% to 7% of lateness has been
accounted for Transportation concerns accounted for approximately 7%
of unavoidable lateness behavior in both samples versus 2% to 5% of lateness in other studies (e.g., Farrell & Robb, 1980; Gupta & Jenkins, 1983;
Leigh & Lust, 1988).
Furthermore, although the antecedents used were generally related
to specific lateness behavior categories as hypothesized, such antecedents
were generally not significantly related to other categories of lateness.
This lack of statistically significant relationships when crossing
antecedents to other lateness categories provides additional support for
the suggested lateness behavior taxonomy. Also, a dynamic lateness
behavior category was created that represented subjects with a mixture of
increasing chronic, stable periodic, and unavoidable lateness. The correlational results found between antecedent variables and work behaviors to
dynamic lateness are quite similar in magnitude to the results found by
153
previous research (e.g., Blau, 1986; Clegg, 1983; Cummings & Manring,
1977; Gupta & Jenkins, 1983). This suggests that previous lateness studies
may have measured lateness with some combination of different types of
lateness behaviors.
Various limitations of this study need to be noted. The use of retrospective nonwork late and personal illness or accident measures, especially over a long time period, is problematic. For example, there is a
longer time frame (i.e., 3 years) between the retrospective non work lateness and personal illness measures to the behavioral data versus the 18month time frame between the attitudinal and perceptual measures to the
behavioral data. Having a greater time frame between variables would be
expected to weaken their relationships (e.g., Mobley et al., 1979). Also,
asking subjects to remember if their arrival to work was delayed by an
unusual event over a retrospective 18-month time frame could pick up
antecedents (e.g., work-family conflict), which may spill over into different lateness categories. Impression management (Paulhus, 1986), or trying to look good to others, could further distort subject responses (e.g.,
more likely to report feeling ill than having a slow alarm clock).
Future researchers could consider collecting such self-report data
over shorter time periods or having subjects use a diary or log to help
them more accurately remember incidents. Similarly, finding the selfreport measure of transportation concerns, but not the objective bad
weather measure, to be significantly related to unavoidable lateness reinforces the need to take precautions against retrospective justification bias
affecting results. Certainly other methodological concerns (e.g., range
restriction or lower measure reliability) may also affect the amount of
lateness category variance explained.
It is also important to acknowledge that, despite the inductive support within both organizations studied for developing random versus
nonrandom patterns and stable versus increasing frequency and duration
classifications, the decision rules and cutoff values applied were organization specific. In other organizations, the decision rules and cutoff values
to create such lateness behavior categories may well be different.
However, the key issue seems to be that some type of decision rule-cutoff
value scheme consistently applied to frequency and duration lateness
data can result in identifying meaningful and useful categories of
employee lateness behavior.
Future research might well expand the number of antecedents used
in the proposed lateness taxonomy. For example, time management skills
or procrastination (Burka & Yuen, 1983) may be a useful individual difference variable to measure, particularly for subjects in the unavoidable
lateness category. By "cutting things too close" in planning to be on-time
154
Gary Blau
155
156
Gary Blau
absence may be stronger than the link between general lateness and general absence. Unfortunately, testing more specific relationships between
withdrawal behaviors was beyond the scope of this study, because only
limited measures of other withdrawal behaviors were gathered.
However, in future research a broader taxonomy of withdrawal
behaviors-including lateness, absence, and leaving early-could be
developed and tested with inductively defined thresholds to determine
pattern, frequency, and duration categories for all of the measured withdrawal behaviors. Ideally, the same threshold values would be applicable
across all of the withdrawal behaviors, but the investigator should be prepared to develop different cutoff values for each withdrawal behavior.
Such different behavioral thresholds would reflect the organization's
sensitivity to each withdrawal behavior. For example, with a sample
of nurses Blau (1985b) found that a hospital's control policy for lateness
was not as strict as it was for unexcused absence.
To conclude, despite the inductive approach used to test the proposed taxonomy of lateness categories, the results seem promising.
Certainly, additional work-operationalizing lateness-behavior categories
with other samples is needed. The dominance of blue-collar and shiftbased jobs in lateness research may be at least partially due to lateness
incidents being more carefully monitored for such jobs. It would be interesting to see the extent to which this taxonomy was supported with managerial or other jobs traditionally given more flexibility regarding
lateness. Finally, this study suggests that investigators must be more sensitive to how lateness is operationalized. More careful operationalization
can result in a greater understanding of the lateness construct.
ACKNOWLEDGMENTS. I gratefully acknowledge the input of Celia
Kamath, John Deckop, Frank Linnehan, and Diana Deadrick on earlier
versions of this article. This research project was funded in part by a
grant-in-aid-of-research from Temple University. Portions of this article
were presented at the 1993 National Academy of Management meeting in
Atlanta.
REFERENCES
Abelson, M. (1987). Examination of avoidable and unavoidable turnover. Journal
of Applied Psychology, 72, 382-386.
Adler, S., & Golan, J. (1981). Lateness as a withdrawal behavior. Journal of Applied
Psychology, 66, 544-554.
Angle, H., & Perry, J. (1981). An empirical assessment of organizational commitment
and organizational effectiveness. Administrative Science Quarterly, 26,1-14.
157
Barcikowski, R., & Stevens, J. (1975). A Monte Carlo study of the stability of
canonical correlations, canonical weights, and canonical variate-variable
correlations. Multivariate Behavioral Research, 10, 354-364.
Beehr, T., & Gupta, N. (1978). A note on the structure of employee withdrawal.
Organizational Behavior and Human Performance, 21, 73-79.
Blass, T. (1977). Personality variables in social behavior. Hillsdale, NJ: Erlbaum.
Blau, G. (1985a). A multiple study investigation of the dimensionality of job
involvement. Journal of Vocational Behavior, 27, 19-36.
Blau, G. (1985b). Relationship of extrinsic, intrinsic, and demographic predictors
to various types of withdrawal behaviors. Journal of Applied Psychology, 70,
442-450.
Blau, G. (1986). Job involvement and organizational commitment as interactive
predictors of tardiness and absenteeism. Journal of Management, 12, 577-584.
Blau, G., & Boal, K. (1987). Conceptualizing how job involvement and organizational commitment affect turnover and absenteeism. Academy of Management
Review, 12, 288-300.
Brooke, P. (1986). Beyond the Steers and Rhodes model of employee attendance.
Academy of Management Review, 11, 345-361.
Brooke, P., Russell, D., & Price, J. (1988). Discriminant validation of measures
of job satisfaction, job involvement and organizational commitment. Journal
of Applied Psychology, 73, 139-145.
Burka, J., & Yuen, L. (1983). Procrastination. Reading, MA: Addison-Wesley.
Campion, M. (1991). Meaning and measurement of turnover: Comparison of
alternative measures and recommendations for research. Journal of Applied
Psychology, 76, 199-212.
Cascio, W. (1987). Costing human resources: The financial impact of behavior in organizations (2nd ed.). Boston: Kent.
Clegg, C. (1983). The psychology of employee lateness, absence and turnover:
A methodological critique and empirical study. Journal of Applied Psychology,
68,88-101.
Cohen, J. (1960). A coefficient of agreement for nominal scales. Educational and
158
Gary Diau
Hanisch, K., & Hulin, C. (1991). General attitudes and organizational withdrawal:
An evaluation of a causal model. Journal of Vocational Behavior, 39,110-128.
Harrison, D., & Hulin, C. (1989). Investigations of absenteeism: Using event history models to study the absence-taking process. Journal of Applied Psychology,
74,300-316.
Hermann, J., deMontes, A., Dominquez, B., Montes, F., & Hopkins, B. (1973).
Effects of bonuses for productivity on the tardiness of industrial workers.
Herzberg, F., Mausner, B., Peterson, R., & Capwell, D. (1957). Job attitudes: Review
of research and opinion. Pittsburgh: Psychological Services.
Jamal, M. (1981). Shift work related to job attitudes, social participation and
withdrawal behavior: A study of nurses and industrial workers. Personnel
Jones, E., & Nisbett, R. (1972). The actor and the observer: Divergent perceptions
of the cause of behavior. In E. Jones, D. Kanouse, H. Kelley, R. Nisbett,
S. Valins, & B. Weiner (Eds.), Attribution: Perceiving the cause of behavior
(pp. 63-86). Morristown, NJ: General Learning Press.
Kanungo, R. (1982). Measurement of job and work involvement. Journal of Applied
Management,29,25-29.
Leigh, J., & Lust, J. (1988). Determinants of employee tardiness. Work and Occupations,
15,78-95.
Lobel, S. (1991). Allocation of investment in work and family roles: Alternative theories and implications of research. Academy of Management Review, 16, 507-521.
McNemar, Q. (1969). Psychological statistics. New York: Wiley.
Macy, B., & Mirvis, P. (1976). A methodology for assessment of quality of work life
and organizational effectiveness in behavior-economic terms. Administrative
159
Nicholson, N., & Goodge, P. (1976). The influence of social, organizational, and
biographical factors on female absence. Journal of Management Studies, 13,
234-254.
Nie, N., Hull, c., Jenkins, J., Steinbrenner, K., & Bent, D. (1975). SPSS: Statistical
package for the social sciences. New York: McGraw-Hill.
Nollen, S., & Martin, V. (1978). Alternative work schedules: Flextime. New York:
AMACOM.
Nunnally, J. (1978). Psychometric theory. New York: McGraw-Hill.
Paulhus, D. (1986). Self deception and impression management in test responses.
In A. Angleitner, & J. Wiggins (Eds.), Personality assessment via questionnaire
(pp. 143-165). Berlin: Springer-Verlag.
Payne, R, & Nicholson, N. (1987). Absence from work: Explanations and attributions. Applied Psychology: An International Review, 36,121-132.
Porter, L., Crampon, W., & Smith, F. (1976). Organizational commitment and
managerial turnover: A longitudinal test. Organizational Behavior and Human
Performance, IS, 87-98.
Richard, D., & Slane, S. (1990). Punctuality as a personality characteristic: Issues of
measurement. Journal of Psychology, 124, 397-402.
Rosse, J. (1988). Relations among lateness, absence and turnover: Is there
a progression of withdrawal? Human Relations, 41, 517-531.
Rosse, J., & Hulin, C. (1985). Adaptation to work: An analysis of employee health,
withdrawal and change. Organizational Behavior and Human Decision Processes,
36, 324-347.
Ruchti, W. (1967). Is there an answer for lateness and absenteeism? Supervisory
Management, 12, 21-22.
Schultz, J., & Henderson, C. (1985). Family satisfaction and job performance:
Implications for career development. Journal of Career Development, 12, 33-47.
Schwab, D. (1980). Construct validity in organizational behavior. In B. Staw, &
L. Cummings (Eds.), Research in organizational behavior (Vol. 2, pp. 3-43).
Greenwich, CT: JAI Press.
160
Gary Blau
Steers, R., & Mowday, R. (1981). Employee turnover and post-decision accommodation processes. In B. Staw, & L. Cummings (Eds.), Research in organizational
behavior (Vol. 3, pp. 235-281). Greenwich, CT: JAI Press.
Steers, R., & Rhodes, S. (1978). Major influences on employee attendance:
A process model. Journal of Applied Psychology, 63, 391-407.
Thompson, c., & Blau, G. (1993). Moving beyond traditional predictors of job
involvement: Exploring the impact of work-family conflict and overload.
Journal of Social Behavior and Personality, 8, 635-646.
Weiss, D., Dawis, R., England, G., & Lofquist, L. (1967). Manual for the Minnesota
Satisfaction Questionnaire. Minneapolis: Industrial Relations Center,
University of Minnesota.
Williams, K., 5uls, J., Alliger, G., Learner,S., & Wan, C. (1991). Multiple role juggling and daily mood states in working mothers: An experience sampling
study. Journal of Applied Psychology, 76, 664-674.
Youngblood,S. (1984). Work, nonwork and withdrawal. Journal of Applied
Psychology, 69, 106-117.
7
New Conceptualizations of
Lateness since Blau, 1994
Gary Blau
161
162
Gary Blau
Hepburn and Barling (1996) have more recently expanded the definition of employee lateness to include any unexcused time missed during
the workday, such as taking too long of a lunch or work break, or leaving
work early, in addition to a tardy arrival. Using Hepburn and Barling's
(1996) broadened definition of lateness allows it to be understood as a
form of "production deviance" within Robinson and Bennett's (1995)
model of deviant work behaviors.
Blau (1994) did find significant correlations between increasing
chronic lateness and leaving work early for both bank (r = .38) and hospital (r = .43) employee samples.
EMPIRICAL RESEARCH ON LATENESS SINCE BLAU, 1994
There has been limited research on employee lateness since 1994. In a
longitudinal study, Blau (1995) found that a measure of presurvey group
(department)-levellateness explained a significant amount of postsurvey
individual lateness, beyond that due to presurvey individual lateness, and
various surveyed work measures, i.e., job involvement, organizational
commitment, job satisfaction, leisure-income trade-off, work-family conflict, transportation concerns, and illness or accidents. Describing the influence of a construct (lateness) at one level (i.e., group) on that variable
measured at a different level (i.e., individual) necessitates a cross-level
research design, in which the mean aggregate (group-level) value for a
variable (i.e., group lateness) is assigned to all members within a given
group. Then the aggregate variable can be correlated with individual-level
variables across all individuals to assess cross-level influence (Rousseau,
1985).
In Blau's (1995) study, 66 bank branches and 71 hospital departments
were used to measure group-level lateness. Group-level lateness scores
were obtained by averaging occurrences (frequency times duration) by
bank branch for tellers and by department for hospital employees, and
then assigning the respective mean values back to individuals (Rousseau,
1985).
Since the group measure would include the individual measure, to
control for measure contamination it is necessary to separate their measurements, i.e., pre- and postsurvey.
Why would group-level measures of lateness affect individual lateness? Applying Nicholson and Johns's conceptual work (1985) on the
social dynamics of absence in work settings suggests that social norms for
absence are based on three key referent groups: co-workers, supervisors,
and family and friends. Furthermore, the salience of norms and beliefs
163
164
Gary Blau
size problem, which will prevent lateness behavior categories, i.e., pattern,
frequency, and duration of lateness behavior, from being operationalized
(Blau, 1994). Self-report lateness may be a necessary research strategy
option. Borrowing from Johns's (1994) discussion of using self-report
absence measures, one should consider: making the lateness accumulation
interval correspond to a salient and natural time unit (e.g., for teachers a
school year); lacking a salient time-frame, make the accumulation long
enough to enhance reliability but short enough to minimize memory loss
(e.g., 3-6 months); opt for a free-response format for obtaining time-lost or
frequency data (lower scale categories suggest to subjects that lower rates
are normal); assume that lateness questions may be threatening, so try to
ask longer questions which assume the behavior has been exhibited; and
to reduce self-generated validity and other artifacts, separate lateness
items within the survey.
Dalton and Mesch (1991) operationalized separate self-report measures of unavoidable (involuntary) and avoidable (voluntary) absence in a
creative manner, which could be extended to lateness behavior. By asking
employees "how many times were you late for work in the last 6 months
due to such factors as bad weather, getting stuck in traffic, having a car
problem or accident, unreliable public transportation, having a child or
other dependent care problem, or feeling sick?" this would measure
unavoidable (involuntary) lateness. Avoidable lateness would be measured as the difference between employees' response to this item and a
"total" lateness item, "how many times were you late to work for any reason in the last 6 months?" Using this approach, the avoidable lateness
measure would combine Blau's (1994) increasing chronic and stable periodic lateness categories, while the unavoidable lateness measure would
reflect Blau's (1994) same behavior category. Such a research design
should allow for a stronger understanding of lateness than using only a
general or overall lateness measure. It is hoped that this updated discussion of lateness will stimulate further research efforts.
REFERENCES
Apgar, M.1. (1998). The alternative workplace: Changing where and how people
work. Harvard Business Review, 76, May-June, 121-136.
Bercovits, M. D., & Koslowsky, M. (1996). Time urgency: A personality attribute for predicting lateness. Ramat Gan: Bar-Han University, Unpublished master's thesis.
Blau, G. (1994). Developing and testing a taxonomy of lateness behavior. Journal of
Applied Psychology, 79,959-970.
Blau, G. (1995). Influence of group lateness on individual lateness: A cross-level
examination. Academy of Management Journal, 38, 1483-1496.
165
Dalton, D. R., Johnson, J. L., & Daily, C. M. (1999). On the use of "intent to" variables in organizational research: An empirical and cautionary assessment.
Dalton, D. R., & Mesch, D. (1991). On the extent and reduction of avoidable absenteeism: An assessment of absence policy provisions. Journal of Applied
Igbaria, M., & Tan, M. (1998). The virtual workplace. Hershey, PA: Idea Group
Publishing.
Johns, G. (1994). How often were you absent? A review of the use of self-reported
absence data. Journal of Applied Psychology, 79, 574-591.
Koslowsky, M., Sagie, A., Krausz, M., & Singer, A. D. (1997). Correlates of employee
lateness: Some theoretical considerations. Journal of Applied Psychology,
82,79-88.
Nicholson, N., & Johns, G. (1985). The absence culture and the psychological
contract-who's in control of absence. Academy of Management Review, 10,
397-407.
8
Smoking and Absence from Work
A Quantitative Review
E. Kevin Kelloway, Julian Barling,
and Caroline Weber
167
168
rates and result in lessening the magnitude of the observed attitudeabsence correlation (Johns, 1991). In this chapter we turn our attention to
the latter suggestion. Specifically, we present data suggesting a link
between employee lifestyle behaviors (Le., cigarette smoking) and workplace absence.
One of the major social questions in recent years has been the effect
of cigarette smoking on individual health and well-being. Sufficient data
have now been accumulated to conclude that smoking does exert adverse
health consequences (Bush & Wooden, 1995), especially increased risks
for coronary heart disease and cancer. Moreover, smokers are more likely
than nonsmokers to experience less severe ailments such as upper respiratory tract infections (Bush & Wooden, 1995; Henningfield et al., 1994).
The association between smoking and ill health begs the question of
whether there are other "hidden" costs associated with cigarette smoking;
"hidden" in that the impact of smoking on outcomes other than physical
ailments has to date received comparatively less attention.
The purpose of our study was to investigate one such potential cost,
namely, whether tobacco smokers are disproportionately more likely to be
absent from work than are nonsmokers. The basis for this question lies
in the observation that smokers are less healthy than nonsmokers
(Henningfield et al., 1994) leading to the suggestion that smokers' attendance at work will be negatively affected. Certainly there are numerous
references in the medical and human resource literatures suggesting an
association between smoking and absenteeism. However, close inspection
of this literature suggests that there is considerably more discussion about
this possible association then there are data supporting such a link. In
other words, the literature is currently driven more by rhetoric than by
empirical data. In addition, studies that do contain empirical data vary
markedly in quality. Therefore, the primary purpose of our study was to
statistically summarize the empirical data linking smoking and absenteeism. In doing so, we empirically integrate findings from numerous
studies to provide an estimate of the effect of smoking on absenteeism.
Second, a cursory inspection of the empirical literature on smoking
and absenteeism suggests that this question has drawn the attention of
researchers from many countries (e.g., Germany, Sweden, Poland, China,
Israel, Australia, the United States, the United Kingdom, and New
Zealand). Many of these countries differ from one another in terms of
both their "smoking culture" (e.g., public policy, attitudes, and legislation
regarding cigarette smoking) and workplace structures (including workplace policies regarding absenteeism). Accordingly, we also address
whether the magnitude of the association between smoking and absenteeism varies across studies.
169
METHOD
Database
For a quantitative analysis to provide an accurate estimate of a relationship, it is critical that all available data be collected and analyzed. To
ensure that we collected as many studies on the relationship between
smoking and absenteeism as were obtainable, we conducted computerized searches of several databases, namely, Medline Medical Index,
American Business Index, Psychological Information Database, Social
Sciences Index, and Humanities Index. All studies purporting to provide
data on smoking and absenteeism were obtained. This included several
articles in languages other than English (e.g., German, Japanese,
Afrikaans) which were translated into English. Numerous articles that
purported to obtain data on the relationship of interest but did not contain any empirical data were also located. Because discussions on the link
between smoking and absenteeism frequently refer to the large database
available, a bibliography of these studies is available from the second
author on request. The articles that did produce usable data are presented
in Table 1.
Measures
In order to combine results across studies, it was necessary first to
develop common operationalizations of both cigarette smoking and
170
Data
year
smokers
Smoker
absence
Year
Anderson &
Malmgren
Athansou
Batenburg &
Reinkein
Bertera
Carmichael &
Cockcroft
Fawer et al.
Foerster et al.
Gabel & ColleyNiemeyer
Gallop
Green et al.
Hawker &
Holtby
Hayden et al.
Hendrix et al.
Hokomb&
Meiggs
Janzon et al.
Kark et al.
Kozak
Lowe
Manning et al.
Marti
McMillan
Qun&
Dobson
Ryan et al.
Smith et al.
Tsai et al.
1986
Sweden
1975
369
22.5
556
18.85
1979
1990
Australia
New Zealand
1978
1988
205
337
18.71
6.18
219
555
10.79
5.51
1990
1990
USA
UK
1988
1988
13,060
99
3.69
7
32,916
79
2.79
4
1982
1976
1990
UK
Germany
USA
1979
1975
1987
26
110
194
2.6
28.5
13.48
12
55
590
1.65
15.8
10.87
1991
1992
1988
New Zealand
Israel
UK
1988
1987
1987
20
2,067
40
10.3
10.99
4.63
33
3,759
121
7.9
10.3
3.39
1984
1991
1972
UK
USA
USA
1983
1986
1964
277
97
114
1981
1982
1987
1960
1989
1986
1981
1992
Sweden
USA
Czech
UK
USA
Germany
UK
China
1978
1981
1984
1957
1984
1984
1979
1987
517
186
373
2,284
96
299
1,962
1,479
13
1.36
25.9
6.57
5.74
3.57
5.58
0.39
230
163
250
1,057
206
3,660
929
377
4
0.91
28.1
5.49
3.95
2.1
4.11
0.27
1992
1981
1990
USA
Australia
Sweden
1989
1980
1986
825
226
2,388
12.96
2.94
7.79
1,712
479
4,609
9.84
2.11
5.12
28.5
6.4
6.9
nonsmokers
Nonsmoker
absence
Authors
239
366
29
15
5.8
4.3
171
Ideally one would also derive an estimate of sampling error and be able to correct for study
artifacts (e.g., restriction in range, measurement error). Unfortunately, the data available
to us did not allow for such corrections-there were insufficient data reported to allow us
to estimate these effects nor were sufficient data available to allow the estimation of
artifact distributions.
172
Smokers
Nonsmokers
9.53
19.32
14.43
9.15
7.10
5.69
6.58
15.33
9.32
5.61
5.69
5.29
173
Country
Data
year
Smoker N
N
Nonsmoker
smokers absence nonsmokers absence
USA
1987
194
13.48
210
11.39
20
10.30
29
9.10
1972
1988
New
Zealand
USA
1964
114
6.90
44
6.18
1989
1981
1989
1981
1990
USA
Sweden
USA
UK
Sweden
1986
35
1978 517
1984
96
1979 1962
1986 2388
Mean
41
13
5.74
5.58
7.79
7.84
35
290
96
686
866
30
7
3.44
4.21
7.19
6.87
Year
DISCUSSION
The primary purpose of our investigation was to investigate the existence and size of an association between cigarette smoking and absence
from work by cumulating results gleaned from almost 40 years of published empirical research on this question. Our results strongly support
the existence of such an association with smokers being absent 2.07 days
per year more than nonsmokers; an increase of 48%. These results point to
the substantial organizational costs of smoking.
Two potential moderators of the smoking absenteeism relationship
were also investigated. First, we found that the size of the effects reported
in the literature did not vary according to where the data were collected.
Consequently, any effects of tobacco smoking on absence from work may
be considered to be stable across countries despite national! cultural
differences in smoking and absenteeism patterns. Second, a limited
number of studies reported data on former smokers in addition to data
from smokers and nonsmokers. This allowed us to assess the potential
174
175
176
behavior. Indeed, in evaluating the effects of smoking cessation on absenteeism, Wooden and Bush (1995) observed that absenteeism declines as a
function of the number of years as an ex-smoker, little effect being
observed in the first year.
In conclusion, the available data support the conclusion that cigarette
smoking is associated with increased absenteeism from work. Smokers
reported approximately 48% more absenteeism from work than did nonsmokers. Moreover, the magnitude of this effect was stable across countries. In addition, some empirical data suggest that smoking cessation
may be associated with decreased absenteeism. Taken together these findings suggest that there are significant organizational costs associated with
cigarette smoking and that organizations may be able to avoid these costs
by supporting smoking cessation and similar lifestyle programs.
ACKNOWLEDGMENTS. This research was supported by grants from
Marion Merrell Dow to all three authors and by Social Sciences and
Humanities Research Council of Canada grants to the first and second
authors. The authors express their appreciation to Doug Grant for his
assistance throughout the research.
REFERENCES
Andersson, G., & Malmgren, S. (1986). Risk factors and reported sick leave among
employees of Saab-Scania, Linkoping Sweden between the ages of 50 and 59.
Scandinavian Journal of Social Medicine, 14, 25-30.
Athansou, J. A. (1979). Smoking and absenteeism. Medical Journal of Australia, 1,
234-236.
Ault, R. w., Ekelund, R. B., Jackson, J. D., Saba, R. 5., & Saurman, D. S. (1991).
Smoking and absenteeism. Applied Econometrics, 23, 743.
Batenburg, M., & Reinkein,]. A (1990). The relationship between sickness absence
from work and patterns of cigarette smoking. New Zealand Medical Journal,
103,11-13.
177
178
9
Job Involvement and Absence
The Role of Constraints as Moderators
Vishwanath V. Baba and Muhammad Jamal
179
180
Constraints as Moderators
181
182
--------+.
Absence
-Time lost
-Frequency of Absence
Constraints
-Psychological
-Behavioral
Figure 1. Model of job involvement and absence with constraints as moderators.
METHOD
Sample
The sample consisted of 193 nonunionized professional employees
from a largo aerospace facility located in a northeastern metropolis in
Canada. The average respondent was a university-educated male, about
Constraints as Moderators
183
40 years of age, married with one or two children and has been with the
organization between 2 and 5 years holding a permanent job.
Procedure
184
2.1
1.6
2.7
1.4
2.4
0.9
0.8
1-5
1-4
1-5
1-3
1-3
1-3
1-3
Range
0.77
0.32
1.1
0.53
0.63
0.69
0.50
SD
42
10
23
-15
-23
-01
Job
involvement
-28
23
-32
-12
-08
16
21
Time lost
(M = 4.0, SD = 5.4)
Correlations (r)
-14
-04
22
26
-34
11
-34
Frequency
(M = 2.5, SD = 2.2)
Table 1. Descriptive Statistics and Correlations among Independent, Dependent, and Moderator Variables
(J1
.....
ex>
;jl
0"
o
c..
~
....
I),>
<Il
I),>
<Il
<Il
;;r
::s
n
o
186
Table 2 provides the results of the moderator analysis using both the
OLS and Tobit techniques. The results indicate substantial support for the
model linking job involvement and absence with both psychological and
behavioral constraints as moderators. Job involvement was a significant
predictor of both time lost and frequency of absence explaining 7.5% of
the variance (OLS) and 7.4% (Tobit) of time lost and 11.4% (OLS) and
12.2% (Tobit) of the variance in frequency of absence.
Time Lost
When the time lost measure of absence was used as a criterion, felt
stress significantly constrained the negative relationship between job
involvement and time lost providing positive regression coefficients for
both OLS and Tobit models. Similarly, job challenge acted as a significant
negative constraint on absence when interacting with job involvement
enhancing the negative relationship between job involvement and
absence. Work group norms of attendance likewise acted as a negative
constraint on absence behavior when interacting with job involvement.
Similarly, organizational norm measured through supervisor's positive
impression and emphasis of attendance acted as a significant negative
constraint on absence as indicated by the negative regression coefficients
and significant B values for both OLS and Tobit models. Further, both
time pressure cognition and time pressure behavior significantly constrained the observed negative relationship between job involvement and
time lost, encouraging absence behavior even among the job-involved.
However, this moderator effect was supported only in the OLS model.
In addition to the moderator effects mentioned above, stress, challenge, and time pressure behavior showed significant main effects on time
lost though they were all weaker compared to their moderator effects
except in the case of stress.
Frequency of Absence
When frequency of absence was used as the criterion measure, the
empirical support for the model was not as strong as in the case of time
lost. Significant constraining effects on the relationship between job
involvement and frequency of absence were observed for job challenge,
work group attendance, and time pressure cognition for both the OLS and
Tobit models in the same direction as in time lost. Time pressure behavior
showed significant moderator effect on the relationship between job
involvement and frequency of absence only for the OLS model which is
consistent with the previous observation for this variable. Both stress and
R2
B
R2
B
SE
R2
B
B
SE
R2
0.80
0.84 13.4
2.3
9.7
11.4
1.5
1.6
0.71
9.1
9.7
0.98 0.57
0.76
2.7
2.0
2.2
1.7
1.8
2.2
1.5
1.3
-2.0 -1.8
1.2
1.5
1.0
1.0
1.0
13.4
10.5
0.88 11.6
9.3
2.7
0.65 0.34 16.7
1.7
1.9
2.7
0.85 0.32 14.9
2.4
0.44
1.1
0.75
0.67
1.0
2.7
2.1
2.3
11.8
2.2
1.7
1.8
0.88 0.71
-1.2
-2.1 -2.5 -1.7 1.03 10.1 -2.5 -0.97 0.34 18.4 -2.8 -1.3 -1.0 0.41 17.1 -3.1
-0.3 -0.12 -0.08 0.61 7.5 -0.2
0.14 0.26 11.6
0.5 -0.07
0.06 0.25 12.2 -0.3
Note. t-values 2: 1.9 are significant at .05 level. Brefers to the Tobit estimated coefficient. B refers to the regression coefficient in OLS and the
estimated partial derivative at the mean in Tobit.
Time pressure
behavior (TB)
JI X TB
Time pressure
cognition (TC)
JI X TC
-1.5
8.4
-0.95 0.75
Workgroup
attendance (WA)
JI X WA
Supervisor's
impression (SI)
JI X SI
0.64 7.4 -4.1 -0.97 0.21 11.4 -4.7 -1.2 -0.95 0.26 12.2 -4.7
2.9
1.5 12.8
0.76 0.50 12.6
1.5
0.88
0.69 0.64 13.5
1.4
2.0 15.2
2.0
1.0 0.62 13.9
1.7
0.91 0.82 14.8
1.4
1.2
1.3
1.4
0.53
SE
Tobit
-1.8 -0.93 -0.63 0.49 9.6 -1.9 -0.24 0.16 12.5 -1.5 -0.30 -0.24 0.20 13.5 -1.5
-3.1 -1.4 -0.97 0.54 14.0 -2.6 -0.50 0.18 16.3 -2.8 -0.51 -0.41 0.22 16.9 -2.3
-1.8
3.1
2.7
OLS
SE
Tobit
Frequency
Challenge (C)
JI X C
OLS
Time lost
(Xl
'-l
....
0'
"'
'""'
Q.
'"
'"~
0
::I
-['""'
Ii
188
supervisor's impression failed to act as significant constraints of the relationship between job involvement and absence frequency. In addition,
both time pressure cognition and time pressure behavior showed significant positive main effects on absence frequency.
When the OL5 and Tobit models are compared in terms of their overall performance, their relative merits appear to be marginal at best. When
the regression coefficients are compared, Tobit seems to provide stronger
estim~tes at first glance though they vanish when the partial derivatives
from B are used in comparison. Out of 13 possible predictions of time lost,
9 were significant with the OL5 model while Tobit produced 8 significant
predictions. In terms of explained variance in time lost, OL5 estimates
were higher than Tobit in 6 cases while Tobit estimates were higher than
OL5 in 6 other cases. In 1 case, they were both identical. Maximum
increase in explained variance in time lost (.1R2) remained the same at
7.8% for both OL5 and Tobit model.
In the case of absence frequency, out of a possible 13 predictions, the
OL5 model produced 7 significant predictions while Tobit produced 6.
However, in terms of explained variance, 11 out of 13 times, the Tobit
model yielded higher R2 compared to OL5. Maximum increase in
explained variance in absence frequency (aR2) was 7% for the OL5 model
and 6.4% for Tobit.
DISCUSSION
The results provided substantial support to the notion that psychological and behavioral constraints moderate the relationship between job
involvement and absence (Johns, 1991). Absence literature suggests that
time lost measures reflect involuntary absence due, for example, to sickness or other calamities where an individual may have been absent for a
long duration though such occurrences may not be frequent at all
(Chadwick-Jones, Nicholson, & Brown, 1982). In other words, the individual may not have much control over his or her absence in this case. On
the other hand, the impact of such incidents are minimal when one deals
with frequency of absence occurrences. The degree of control an individual has over his or her absence is higher in this case compared to the former. Therefore, frequency of absence has been used as a measure of
voluntary absence as the term is referred to in the literature (ChadwickJones et aI., 1982). The differences in the pattern of results obtained for
time lost and absence frequency seem to underscore the theoretical
importance of differentiating between involuntary and voluntary absence
(Chadwick-Jones et aI., 1982; Johns, 1997). Time lost measure of absence
Constraints as Moderators
189
yielded better predictions in terms of significance compared to the frequency measure though variances explained were higher for the frequency measure. If the time lost measure is viewed as an indication of
involuntary absence, the results seem to suggest that constraints playa
more significant role in bringing about involuntary behavior than voluntary behavior. In other words, when a job-involved person encounters a
constraint to his behavior, it appears to bring about some form of involuntary behavior more than voluntary behavior. Simply put, constraints
seem to weaken the control an individual has over his behavior. Removal
of such constraints may indeed facilitate the free impact of attitude on
behavior, which in this case refers to the negative impact of job involvement on absence, especially if the latter is involuntary. It appears that the
general tenets of the path goal theory of leadership (House & Mitchell,
1974) apply to this situation in terms of designing an appropriate intervention strategy.
Specific results obtained appear to support these general observations. For example, while stress seems to interact significantly with job
involvement in influencing time lost, it had no impact, either directly or
as a moderator, on frequency of absence. Thus, the conceptual distinction
between involuntary and voluntary absence becomes empirically meaningful. Similarly, supervisor's positive impression or emphasis of good
attendance has no effect on the employee's voluntary absence behavior
simply because the employee is in control of his absence behavior and
perhaps makes compensatory accommodations for such impressions
either prior or subsequent to the occurrence of absence. The employee is
in control of his absence behavior in this case including its repercussions.
As reported earlier, stress and time pressure show a direct effect on
absence in addition to their moderator effects. These findings are supported in the literature on stress-induced absence (Johns, 1997) as well as
in the Steers and Rhodes (1978) model of attendance. In other words, they
influence the absence behavior of even those who are not involved in their
job which is understandable, but parenthetical to the concerns of this
study.
This study made an explicit attempt to compare the OLS model with
Tobit in exploring the relationship between job involvement and absence
with certain constraints as moderators. Contrary to the findings reported
by Price and Mueller (1986), the number of significant predictors actually
diminished in this study when the Tobit model was used. Since the Price
and Mueller (1986) study did not report either the regression coefficients
or the variances explained in the two models, other systematic comparisons could not be made. Based on the results of this study, one cannot
argue in favor of one model over the other. However, it seems clear that
190
the choice of Tobit over OLS depends on the degree of truncation exhibited by the distribution and its measure of skewness. As reported earlier,
the degree of skewness for absence frequency (1.12) was roughly twice as
high as for time lost (0.65). The Tobit model seems to perform better in the
case of absence frequency where in 11 out of 13 cases, the variance
explained was higher for Tobit. On the other hand, both OLS and Tobit
models provided comparable performance for time lost. This observation
is in line with the reasons provided by Hammer and Landau (1981) favoring the use of Tobit when the distribution is highly skewed. Therefore, it
is suggested that the researcher examine the nature of distribution of
absence data and choose either OLS or Tobit for data analysis depending
on the degree of truncation and the measures of skewness obtained for
that distribution.
In conclusion, it is clear that a job-involved professional male will be
less absent from work. However, he is also sensitive to stress, job challenge,
attendance norms, and nonwork time pressure in terms of his absence
behavior. Specifically, a job-involved person under stress will tend to absent
himself more than the one with less stress. Similarly, a job-involved person
with a challenging job tends to absent himself less than one with a less challenging job. Further, where attendance norms are strong, a job-involved
person tends to absent himself less compared to a situation where attendance norms are weak. Finally, a job-involved person with less nonwork
pressures tends to absent himself less than one with more time pressure.
Thus, a low-stress environment, challenging job, strong attendance norm,
and supportive nonwork environment will enhance attendance among jobinvolved people. Future research must concentrate on how such interventions can be effected as well as explore the actual processes by which
constraints influence both attitude and behavior.
ACKNOWLEDGMENTS. Financial support for this study from FCAR
(OO-ER-0506) and SSHRC (410-99-1491) is gratefully acknowledged. The
authors are thankful for the critical comments on an earlier draft by Gary
Johns and Willi Wiesner.
REFERENCES
Baba, V. V. (1990). Methodological issues in modeling absence: A comparison of
least squares and tobit analysis. Journal of Applied Psychology, 75, 428-432.
Beehr, T. A., Walsh, J. T., & Taber, T. D. (1976). Relationship of stress to individually and organizationally valued states: Higher order needs as a moderator.
Journal of Applied Psychology, 61, 41-47.
Constraints as Moderators
191
192
Mathieu, J. E., & Kohler, S. S. (1990). A test of the interactive effects of organizational commitment and job involvement on various types of absence. Journal
of Vocational Behavior, 36, 33-44.
Nicholson, N. (1977). Absence behavior and attendance motivation: A conceptual
synthesis. Journal of Management Studies, 14, 231-252.
Price, J. L., & Mueller, C. W. (1986). Absenteeism and turnover of hospital employees.
Greenwich, CT: JAI Press.
Rabinowitz,S., & Hall, D. T. (1977). Organizational research on job involvement.
Psychological Bulletin, 84, 265-280.
Saal, F. E. (1978). Job involvement: A multivariate approach. Journal of Applied
Psychology, 63,53-61.
Scott, K. D., & Taylor, G. S. (1985). An examination of conflicting findings on the
relationship between job satisfaction and absenteeism: A meta-analysis.
Academy of Management Journal, 28, 599-612.
Steers, R. M., & Rhodes, S. R. (1978). Major influences on employee attendance:
A process model. Journal of Applied Psychology, 63, 391-407.
Steers, R. M., & Rhodes, S. R. (1984). Knowledge and speculation about absenteeism. In P .S. Goodman & R. S. Atkin (Eds.), Absenteeism (pp. 229-275).
San Francisco: Jossey-Bass.
Youngblood, S. A. (1984). Work, nonwork and withdrawal. Journal of Applied
Psychology, 69, 106-117.
10
The Timing of Thinking about
Quitting
The Effect on Job Attitudes and Behaviors
Kathy A. Hanisch
193
194
Kathy A. Hanisch
195
LITERATURE REVIEW
Although no articles or empirical studies were located that have
addressed the issue dealt with in this chapter, a significant amount of
research is available that 0) has evaluated the relations among attitude
measures, turnover, and other withdrawal behaviors, (2) proposes the
relations among withdrawal behaviors, and (3) evaluates the different
withdrawal models oftentimes comparing stayers and leavers on relevant
attitude and behavior variables. Each will be briefly examined below.
JOB ATTITUDES AND TURNOVER
The relationship between turnover and attitudinal variables has been
examined repeatedly since Taft and Mullins (946) included dissatisfaction in their reasons for turnover. Most researchers find a significant negative association between quitting and attitudinal measures generally
focusing on job satisfaction and organizational commitment. For example,
George and Jones (1996) found a significant correlation between job satisfaction and turnover of - .36. Job satisfaction and organizational commitment were both significant variables in a regression analysis predicting
intent to leave in a study by Rosin and Korabik (1995). In a sample of
nurses, Eberhadt, Pooyan, and Moser (1995) reported a correlation of - .50
between intent to quit and job satisfaction. It should be noted that
some studies (e.g., Good, Page, & Young, 1996) did not find significant
associations between intent to leave and job satisfaction. Sometimes other
alternative attitudinal measures (e.g., organizational commitment) are
related to intent to leave or quitting.
Researchers have also found notable decrements in job satisfaction and
commitment shortly before turnover occurred (e.g., Youngblood, Mobley, &
Meglino, 1983). Job satisfaction was found to be a stronger predictor of
withdrawal cognitions/intentions than organizational commitment (Tett &
Meyer, 1993). Mobley (1977) found high negative correlations between satisfaction and frequency of thinking of quitting. Atkinson and Lefferts (1972)
found that the frequency of thinking about quitting was significantly
related to actual quitting.
ORGANIZATIONAL WITHDRAWAL
Organizational withdrawal was defined by Hanisch (1995b:604) as a
construct denoting "various behaviors employees engage in to remove
196
Kathy A. Hanisch
197
have common antecedents, but they have different functions and consequences for the employees. March and Simon argued that dissatisfaction,
or motivation to withdraw, is "general and holds for both absences and
voluntary turnover" (p. 93). Absences are hypothesized to reduce negative job attitudes by removing employees temporarily from the stresses of
their day-to-day tasks. Quitting reduces stress from a specific job by
removing the individual permanently from the organization and required
work tasks. There are few empirical results that support this model
(Mowday & Spencer, 1981; Weiner, 1980) and it is generally not considered a likely candidate for the relationship among withdrawal behaviors
(Dalton & Todor, 1993).
Compensatory Behaviors Withdrawal Model. The Compensatory
Behaviors Withdrawal Model, first described by Hill and Trist (1955),
assumes that enacting one withdrawal behavior, in varying amounts (e.g.,
once, several times), in response to negative organizational attitudes or
work-related stress is sufficient to reduce job stress and dissatisfaction. All
withdrawal behaviors appear to have the same psychological function in
this model: Relieving job dissatisfaction and job stress. Any withdrawal
behavior, once enacted, compensates for other withdrawal behaviors
making them unnecessary. Hill and Trist (1955) observed that absences
were a characteristic of "stayers" rather than "leavers." Those who were
not absent were more likely to quit than those who had been absent in the
past. From this observation they concluded that absenteeism compensated for any job dissatisfaction that may have been experienced by the
"absence-prone" employees making it unnecessary for them to quit.
Dalton and Todor (1993) argue that one of the two prevailing views
regarding the relation between absenteeism and turnover is the adjustment model. This model suggests that absenteeism is a means to cope
with an unsatisfactory job and that it substitutes for turnover. Thus, being
absent makes one's job manageable and negates having to quit. This view
would predict the same negative relation between absenteeism and
turnover as the Compensatory Behaviors Model.
Progression of Withdrawal Model. Several articles contain elements of the Progression of Withdrawal Model. Baruch (1944) described
absences from work as "short-term quits." Herzberg, Mausner, Peterson,
and Capwell (1957) argued that the problems of absenteeism and
turnover should be considered jointly because "the small decision which
is taken when the worker absents himself is a miniature version of the
important decision he makes when he quits his job." Melbin (1961)
argued that leaving a job is the outcome of a chain of experiences leading
198
Kathy A. Hanisch
199
200
Kathy A. Hanisch
201
10 samples, all of the people who left had significantly higher absenteeism
than those who stayed. Porter, Crampon, and Smith (1976) using a
matched group of recently hired stayers and leavers found no significant
differences with the attitudes of leavers and stayers from 2 to 6 months
prior to the time the leavers left the organization. Significant differences
were found for organizational commitment in the last month of employment of leavers. Job satisfaction was found to be one of the most important discriminators between stayers and leavers in a study conducted by
Gaertner and Nollen (1992). Gupta and Jenkins (1982) found that overall,
leavers had higher absence levels than stayers. Finally, Wolpin and Burke
(1985) evaluating voluntary turnover and voluntary absenteeism found
them to be positively correlated suggesting that voluntary leavers had
higher rates of voluntary absenteeism.
Current Study Variables
METHOD
Subjects
202
Kathy A. Hanisch
employees had worked for the university on average 4.7 years with a
range of 3 months to 25 years. Eighteen (15%) participants had worked
less than 1 year with 8 (8.3%) of those working less than 6 months. The
average salary was approximately $35,000 and 97% were Caucasian. The
individuals in this study had been employed in professional and scientific
positions (e.g., research associate, program coordinator, associate scientist) at the university.
Procedure and Questionnaire
The names of respondents were obtained from the university human
resources office. A letter and an exit survey with a preaddressed, postagepaid return envelope were sent to each individual on the list who had a
complete mailing address. The letter described the study and encouraged
participation. The author only contacted individuals; information shared
with the human resources office was in aggregate form only. Relevant sections of the exit survey included the demographic items described earlier
and the five scales from the Job Descriptive Index 001; Smith, Kendall, &
Hulin, 1969) assessing work satisfaction (18 items; e.g., "fascinating"),
supervision satisfaction (18 items; e.g., "hard to please"), satisfaction with
promotion (9 items; e.g., "opportunities somewhat limited"), co-worker
satisfaction (18 items; e.g., "stimulating"), and pay satisfaction (9 items;
e.g., "barely live on income"); all items have response scales of Yes, No,
or ? (i.e., cannot decide). The JOI has been used numerous times in past
research with excellent reliabilities and validities (e.g., Hanisch & Hulin,
1990, 1991); the developmental, psychometric properties, and scoring of
the index have been studied and the validity is well documented
(Hanisch, 1992; Smith et a1., 1969).
In addition to the satisfaction scales, the participants also reported
the frequency with which they engaged in 17 withdrawal behaviors (e.g.,
late to work, surfing the Web on work time) in their last year of employment at the university. Many of the behaviors used in this study were first
generated and reported in a study by Hanisch and Hulin (1990) and have
been used successfully in later research. Employees rated each behavior
using an 8-point scale (1 = Never in the past year, 8 = Once or more a day
in the past year) to indicate how often they had engaged in each behavior
in the past year.
RESULTS
203
(1) Gender"
(2) Age
(3) Salary
(4) Months employed
(5) Months thinking
about quitting
(6) Work satisfaction
(7) Pay satisfaction
(8) Co-worker
satisfaction
(9) Supervision
satisfaction
(10) Promotion
satisfaction
(11) Withdrawal
behaviors
.08
.08
.15
.01 -.08
.07 -.05
.05 -.02 -.25
-.02 -.17 -.18 -.08 -.19
.90
.19
.28
.86
.05
.91
.39
.21
.44
.88
.26
0.26
.33
.19
.06
-.22
-.04
-.13
.37
.31
.21
.38
.28
.48
.07 -.10
-.01
-.02 -.04
.09
10
11
.81
Note. Correlations statistically significant (p < .05) are bold. Coefficient alpha underlined
and inserted in the main diagonal.
a Male = 1, female = 2.
the main diagonal and range from .79 to .91 indicating homogeneous and
consistent scales.
Many of the correlations shown in Table 1 are consistent with expectations. For example, gender is significantly correlated with salary indicating that men have higher salaries than women. Age, as expected, is
positively correlated with salary and months employed. It is also positively correlated with the number of months individuals thought about
resigning and negatively correlated with promotion satisfaction indicating that older individuals report being less satisfied with their promotion
opportunities and spend more time (in months) thinking about quitting
than younger individuals. Total months employed and months thinking
about resigning are both positively correlated with salary; an expected
finding. In addition, those who spent more time thinking about quitting
are also employed longer in this study.
Number of months thinking about quitting is negatively correlated
with pay satisfaction, supervision satisfaction, and co-worker satisfaction
indicating that these dissatisfactions were related to their thinking of quitting; its correlation with promotion satisfaction approached statistical significance (p < .10). Several of the attitude scales are intercorrelated (e.g.,
work satisfaction is positively correlated with all other satisfaction scales).
Work satisfaction and withdrawal behaviors are negatively correlated
Kathy A. Hanisch
204
(- .23) and supervision satisfaction is negatively correlated with withdrawal behaviors ( - .24). These results suggest that those who are less satisfied with their work and co-workers are more likely to engage in
withdrawal behaviors replicating the work of others discussed earlier
(e.g., George & Jones, 1996; Waters & Roach, 1971) and when absenteeism,
intent to quit, or quitting was used singularly as the withdrawal behavior.
Discriminant Analysis
The job attitudes and behaviors that might differentiate those who
had thought about their decision to quit longer than those whose cognitive processes about quitting were more recent were examined. Decision
time was determined based on each employee's response to the question,
"What date did you start thinking you might like to resign from your job
(e.g., such as looked at advertisements, sent out a resume, made phone
calls to friends at other organizations inquiring about jobs)?" The median
length of time reported was 6 months in this study; individuals with
values above the median were classified as longer decision, those with
scores below the median were classified as shorter decision. Discriminant
function analyses were completed using the five JDI job attitude scales
and the withdrawal behavior scale as the variables to classify the individuals into their decision time lag group. The discriminant function results
are shown in Table 2.
The attitudes and behaviors successfully discriminated between
those who had made their decision 6 months or more and less than
6 months prior to quitting as shown in the main diagonal of Table 2; individuals were classified correctly into their respective decision time lag
Table 2. Attitudes and Behaviors by Decision Time
Classification Results
Predicted group membership
Actual group membership
Less than 6 months
No. classified
% classified
Six months or greater
No. classified
% classified
Prior probability
Total
Less than
6 months
6 months
or greater
38
71.7
15
28.3
53
16
26.5
.51
54
35
73.5
.49
50
51
Total
104
205
Variable means
-.71
.08
-.09
34.42
41.63
.61
.49
.21
-.22
7.17
4.84
.55
.41
.19
-.19
42.36
37.57
.35
-.15
.12
-.12
33.36
29.96
.33
.25
.17
-.28
.11
-.28
-.11
.29
14.15
40.64
12.43
38.45
206
Kathy A. Hanisch
The opposite was true for the withdrawal behaviors; the mean was higher
for those with a longer rather than shorter time lag indicating they reported
engaging in more withdrawal behaviors prior to quitting.
DISCUSSION
207
response was no and several said that they worked harder in an effort to
make the transition for their replacement easier. This was true regardless
of when they had their first thoughts of leaving.
Given this is the first study to address this issue, additional research
on this topic is needed. Specifically, evaluating individuals' thoughts of
leaving with their attitudes and behaviors over time in a longitudinal
evaluation would be useful to more fully explicate the process of leaving.
The average length of time employees in this study had thought about
leaving was about 12 months; it was serendipity that the question about
frequency of engaging in the withdrawal behaviors was framed with a
time reference of the last year of employment. However, several individuals had thought about leaving for longer than 12 months so the question
of their attitudes and behaviors since their first thoughts of quitting
would add to our understanding of the withdrawal process.
Additional research is needed on the models of withdrawal particularly to begin to discern which model is the most accurate description
under specific circumstances. As Wolpin and Burke (1985) have noted,
research supports at least three of the models with regard to the relation
between absenteeism and turnover. They posit this may be due to a
number of reasons: (1) measurement error, (2) third variables causing the
relationship, (3) population type, and (4) conceptual and operational definitions. All of these issues need to be addressed and carefully evaluated
in order to give us a better understanding of the withdrawal process and
the relations among the various withdrawal behaviors. The research area
becomes difficult to pursue once withdrawal behaviors are expanded to
include more than just the typically evaluated absenteeism and turnover.
However, as noted by Thurstone (1931) over 70 years ago, it is possible
and probable that two individuals with the same attitude may enact quite
different behaviors. This, in effect, argues for studying the multiple withdrawal behaviors employees have available to them and understanding
the relations and parameters that define the relations among the withdrawal behaviors. One model may work well for one subset of employees
while another may be better suited for another group. First thoughts of
quitting may be an important parameter to consider when conducting
further tests of the withdrawal models.
This study focused only on voluntary turnover decisions and as others have noted, it is important to distinguish between voluntary turnover
and voluntary forms of withdrawal versus involuntary forms of withdrawal (Benson & Pond, 1987; Wolpin & Burke, 1985). An employee who
is absent due to factors such as jury duty or a funeral would not be
viewed as withdrawing from the organization. An interesting extension
of this research would be to examine the timing of layoff decisions and
208
Kathy A. Hanisch
REFERENCES
Atkinson, T. J., & Lefferts, E. A. (1972). The prediction of turnover using
Herzberg's job satisfaction technique. Personnel Psychology, 25, 53-64.
Baruch, D. W. (1944). Why they terminate. Journal of Consulting Psychology,B,35-46.
Beehr, T. A., & Gupta, N. (1978). A note on the structure of employee withdrawal.
209
Fisher, C. D., & Locke, E. A (1992). The new look in job satisfaction research and theory. In C. J. Cranny, P. C. Smith, & E. F. Stone (Eds.), Job satisfaction (pp. 165-194).
New York: Lexington.
Fitzgibbons, D., & Moch, M. (1980). Employee absenteeism: A multiple analysis
with replication. Organizational Behavior and Human Performance, 26, 349-372.
Gaertner, K. N., & Nollen, S. D. (1992). Turnover intentions and desire among
executives. Human Relations, 45, 447-465.
George, J. M., & Jones, G. R. (1996). The experience of work and turnover intentions: Interactive effects of value attainment, job satisfaction, and positive
mood. Journal of Applied Psychology, 81, 318-325.
Good, L. K., Page, T. J., & Young, C. E. (1996). Assessing hierarchical differences in
job-related attitudes and turnover among retail managers. Journal of Academy
of Marketing Science, 24, 148-156.
Gupta, N., & Jenkins, G. D., Jr. (1982). Absenteeism and turnover: Is there a progression? Journal of Management Studies, 19, 395-412.
Hanisch, K. A. (1992). The Job Descriptive Index revisited: Questions about the
question mark. Journal of Applied Psychology, 77(3), 377-382.
Hanisch, K. A. (1995a). Behavioral families and multiple causes: Matching the
complexity of responses to the complexity of antecedents. Current Directions
in Psychological Science, 4, 156-162.
Hanisch, K. A (1995b). Organizational withdrawal. In N. Nicholson (Ed.), The
Blackwell dictionary of organizational behavior (p. 604). Oxford: Blackwell.
Hanisch, K. A, & Hulin, C. L. (1990). Job attitudes and organizational withdrawal:
An examination of retirement and other voluntary withdrawal behaviors.
Journal of Vocational Behavior, 37, 60-78.
Hanisch, K. A., & Hulin, C. L. (1991). General attitudes and organizational
withdrawal: An evaluation of a causal model. Journal of Vocational Behavior, 39,
110-128.
Hanisch, K. A, Hulin, C. L., & Roznowski, M. (1998). The importance of individuals' repertoires of behaviors: The scientific appropriateness of studying multiple
behaviors and general attitudes. Journal of Organizational Behavior, 19,463-480.
Hanisch, K. A., Hulin, C. L., & Seitz, S. T. (1996). Mathematical/computational
modeling of organizational withdrawal processes: Benefits, methods, and
results. In G. R. Ferris (Ed.), Research in personnel and human resources management (Vo1.14, pp. 91-142). Greenwich, CT: JAI Press.
Herzberg, F., Mausner, B., Peterson, B., & Capwell, D. (1957). Job attitudes: Review
of research and opinion. Pittsburgh: Psychological Services of Pittsburgh.
Hill, J. M., & Trist, E. L. (1955). Changes in accidents and other absences with
length of service: A further study of their incidence and relation to each other
in an iron and steel works. Human Relations, 8,121-152.
Katzell, R. A (1994). Contemporary meta-trends in industrial and organizational
psychology. In H. C. Triandis, M. D. Dunnette, & L. M. Hough (Eds.),
Handbook of industrial and organizational psychology (2nd ed., Vol. 4, pp. 1-89).
Palo Alto, CA: Consulting Psychologists Press.
Keller, R. T. (1984). The role of performance and absenteeism in the prediction of
turnover. Academy of Management Journal, 27,176-183.
Kathy A. Hanisch
210
Lee, T. w., & Mitchell, T. R (1994). An alternative approach: The unfolding model
of employee turnover. Academy of Management Review, 19, 51-89.
Lee, T. w., Mitchell, T. R, Wise, L., & Fireman, S. (1996). An unfolding model of
voluntary employee turnover. Academy of Management Journal, 39, 5-36.
Lyons, T. F. (1972). Turnover and absenteeism: A review of relationships and
shared correlates. Personnel Psychology, 25, 271-281.
March, J. G., & Simon, H. A. (1958). Organizations. New York: Wiley.
Melbin, M. (1961). Organizational practice and individual behavior: Absenteeism
among psychiatric aids. American Sociological Review, 26, 14-23.
Mitra, A., Jenkins, G., Douglas, & Gupta, N. (1992). A meta-analytic review of the
relationship between absence and turnover. Journal of Applied Psychology, 77,
879-889.
Mobley, W. H. (1977). Intermediate linkages in the relationship between job satisfaction and employee turnover. Journal of Applied Psychology, 62, 237-240.
Mobley, W. H., Horner, S. 0., & Hollingsworth, A. T. (1978). An evaluation of precursors of hospital employee turnover. Journal of Applied Psycholof;>Y, 63,408-414.
Mowday, R T., & Spencer, D. G. (1981). The influence of task and personality characteristics on employee turnover and absenteeism incidents. Academy of
Management Journal, 24(3), 634-642.
Porter, L. W., Crampon, W. J., & Smith, F. J. (1976). Organizational commitment
and turnover: A longitudinal study. Organizational Behavior and Human
Performallce, 15, 87-98.
Porter, L. W., & Steers, R M (1973). Organizational work and personal factors in
employee turnover and absenteeism. Psychological Bulletin, 80,151-176.
Rice, A. K, & Trist, E. L. (1952). Institutional and subinstitutional determinants of
change in labor turnover. Human Relations,S, 347-372.
Rosin, H., & Korabik, K (1995). Organizational experiences and propensity to
leave: A multivariate investigation of men and women managers. Journal of
Vocational Behavior, 46,1-16.
Smith, P. c., Kendall, L. M., & Hulin, C. L. (1969). The measurement of satisfaction in
work and retirement: A strategy for the study of attitudes. Skokie, IL: Rand-McNally.
Somers, M. J. (1996). Modelling employee withdrawal behaviour over time: A
study of turnover using survival analysis. Journal of Occupational and
Organizational Psychology, 69, 315-326.
Taft, R, & Mullins, A. (1946). Who quits and why. Personnel Journal, 24, 300-307.
Tett, R P., & Meyer, J. P. (1993). Job satisfaction, organizational commitment,
turnover intention, and turnover: Path analyses based on meta-analytic findings. Personnel Psychology, 46, 259-293.
Thurstone, L. L. (1931). The measurement of social attitudes. Journal of Abnormal
and Social Psychology, 26, 249-269.
Waters, L. K, & Roach, D. (1971). Relationship between job attitudes and two
forms of withdrawal from the work situation. Journal of Applied Psychology, 55,
92-94.
211
Williams, A., Livy, B., Silverstone, R., & Adams, P. (1979). Factors associated with
labour turnover among ancillary staff in two London hospitals. Journal of
Occupational Psychology, 52(1), 1-16.
Wolpin, J., & Burke, R. J. (1985). Relationships between absenteeism and turnover:
A function of the measures. Personnel Psychology, 38, 57-74.
Youngblood, S. A., Mobley, W. H., & Meglino, B. M. (1983). A longitudinal analysis of the turnover process. Journal of Applied Psychology, 68, 507-516.
Index
Absence
definitions
constitutive, 98-100
integrative, 120
operational, 100-104
frequency, 100, 186, 188
perceptions of one's, 102
Absence behaviors, recall of, 102-103
Absence legitimacy, at cultural and
individual levels, 24-25
Absenteeism, 22, 167-168; see also Smoking
voluntary, 147, 148
measures of, 142-143
Absenteeism questions, encoding of, 102
Accidents: see Illness-accident, personal
Alternate Forms of Withdrawal Model,
199-200
Appraisal review, formal, 88-89
Attendance behaviors for superiors and
subordinates, protocol of, 5-6
Attendance rules, 5-6
Attitudes, job, 144, 195, 203, 205
Demographics
of composition of workforce, 36
and lateness, 149, 152
and withdrawal behavior, 203
Discipline, 9, 199
Distributed work, 119
213
214
Flexible organization-home boundaries,
companies with, 17
Flexplace, 119-120
Flextime, 30-31, 120
Gender differences: see Demographics
Gender role differentiation, 35-38
Group cohesiveness, 9
Group-level lateness behavior, 8-10
Hazard rate, 101
Identification
in-group, 63
with organization, 86
Illness-accident, personal
and lateness, 145, 147-154
Illnesses, 39-40
perceived legitimacy of various, 24-25
Independent Forms of Withdrawal Model
196-197
'
Individualism, 12, 23
Information technology (IT), 119; see also
under Withdrawal behaviors,
measurement of
"Jarring events," 59
Job alienation, 75-77
Job challenge (C), 182, 185-189
Job design, 80-81, 83
Job involvement (JI), 26
and lateness, 147-154
Job involvement and absence, model of
with constraints as moderators, 180, 182,
185-190
Job satisfaction, 144, 147-151, 203-206; see
also Turnover
and job alternatives
in the context of progression models,
56-57
differential importance of, 57
Lateness
causes identified by managers, 13-15
costs, 1-2
definitions
constitutive, 104
integrative, 120
operational, 104-105
Index
Lateness (cont.)
empirical research on, since Blau, 1994,
162-163
increasing chronic, 135-136,147-156
new conceptualizations of, since Blau,
1994, 161-164
nonwork, 143, 147-154
sanctions on, 199
stable periodic, 136-137, 147-156
unavoidable/involuntary, 137, 154, 164
Lateness behavior, 133
measures, 139-142
multiple level antecedents, 15-18
extraorganizationallevel, 4, 10-13
group and organizational levels, 4, 810
individual level, 2-8
multiple-level model of, 2-3
taxonomy of, 134-135, 145-156
Lateness proneness, 104
Lateness propensity, 104
Lateness records, 120
Layoffs, 12, 106
Leaving work early, 143, 147-154
Leisure-income-trade-off, 136, 144
Locus of control, internal-external, 34-35
Mobility: see also Employment prospects
perceived ease of, viii
within-organization, 64-66
Modified Organizational Adaptation/
Withdrawal model, 73
Monitoring performance, 87, 121-122; see
also Performance Management
Process
Moves, internal/within-organization: see
Transfers
National culture, 11-12
effects on absenteeism, 21-24, 42-43
future theory and research, 40-42
work centrality and, 25-28
Native Americans, 23
"New Economy," vii
Organization-home boundaries, companies
with flexible, 17
Organizational flexibility, 64-66
Organizational records, 100-101, 120
Index
Pace of life, 33
Pay satisfaction, 205
Performance Management Process, 90-91
as antidote to withdrawal behavior, 8391
essential preconditions for, 85
and new work environment, 89-90
procedural steps, 84-89
Performance monitoring, computerized,
121-122
Performance reviews, 88-89
Performance standards, setting, 86
Personality traits and factors
and attendance behavior, 4
and turnover, 58
Political climate, 13
Power distance, 12
Powerlessness
contextual factors leading to, 80-83
organizational factors, 81-82
job alienation and, 75-77
Problem-solving approach, 88
Progression of Withdrawal Model, 197-198,
206
Promotion satisfaction, 205
Punctuality, culture and, 11-12
Quitting, 106; see also Turnover, voluntary;
Withdrawal behaviors, models of
timing of thinking about, 193-195, 203208
Relocation, 61-62
Resonance, 60
Retirement, 107
Reward systems, 9, 80, 82-83
Rewards, setting, 86
Satisfaction: see Job satisfaction
Saudi Arabia, 27
Schedules, work, 9-10, 29-32
Self-determination, 79
Self-efficacy, 79, 88
Sick pay, 39-40
Sickness: see Illnesses
Sickness insurance, 39-40
Smoking and absenteeism, 168-169, 171-176
international studies, of 172-173
Smoking cessation and absenteeism, 172173
215
Social support systems, 38-40
Societies; see also National culture
sequential vs. synchronic, 12
Spillover Model of Withdrawal, 198-199,
206
Stimulus event, 59
Stress (S), 182, 185-190
Student absenteeism, academic
achievement and, 27-28
Supervision satisfaction, 205
Supervisor's impression (SI), 185-190
Supervisory style, 80, 82
Teams, 14
Telecommuting, 122-123
Temporary help agency workers, 55
Termination, 106
Time loss, 100, 186
Time orientation, 28-34
Time pressure behavior (TB), 185-190
Time pressure cognition (TC), 185-190
Time urgency, 4-5, 33
Transfers, within-organization, 65, 66, 106
Transportation concerns, 144
Turnover, employee, 53-54; see also Job
satisfaction
changing meaning of, 54-56
constitutive definition, 106-107
difficulty disentangling voluntary from
involuntary, 61-62
without dissatisfaction, 58-62
dissatisfaction without, 58
impulsive / non -sa tisfa ction-dependen t,
58-59
incited by others' behavior, 59-61
and initial intention to leave, 63-64
intra- vs. inter-organizational, 65
job attitudes and, 195, 203, 205
material and psychological aspects, 5456
measures of, ]43
model(s) of, 194
unfolding, 59, 62-63, 194
need for multilevel outlook on, 66-67
operational definition, 107-108
organizational flexibility vs., 64-66
without prior search for alternatives, 6263
socioemotional implications, 65
voluntary, ]47-154; see also Quitting
216
Turnover, employee (cont.)
imposed, 61-62
Turnover propensity, short- and long-term,
106
Type A personality, and attendance
behavior, 4
Unemployment, 12
Vietnamese culture, 23
Weather, inclement, 144-145, 147-154
Withdrawal
as behavior, 113
as construct vs. process, 116
constructs overlapping with, 109-110
job, 23, 112, 113; see also Quitting
organizational,196-200
defined, 195-196
Withdrawal behaviors, 57, 71-72, 84, 205;
see also Empowerment; Performance
Management Process; Powerlessness
categories of, 73-74, 116
changes in constitutive meaning, 119-121
job involvement, job alienation, and, 7476
lateness, 1
measurement of, 95, 113-117, 123-124
changes in information technology
and, 118-124
outcome-based measures, 122-123
process-based measures, 121-122
self- and social-report measures, 115
Index
Withdrawal behaviors (cont.)
measurement of (cont.)
and time, 117-118
models of, 196-200, 206
empirical research relevant to, 200-201
multiple, 112-116
broad constructs as preferable to
narrow, 108-109
constitutive definition, 111-112
different goals, 110-111
narrow constructs as preferable to
broad, 109-110
operational definitions of, 112-116
changes in, 121-123
organizational commitment and, 76-77
single, 98, ]20; see also Absence;
Lateness; Turnover
temporary vs. permanent, 116
unified meaning, 116
Withdrawal research
constitutive definitions in, 96-97, 119-121
operational definitions in, 97-98
Work and nonwork time and activities,
28-33
flexibility of boundaries between, 33
Work attitudes, 144, 195, 203, 205
Work centrality, 25-28
Work-family conflict, 136,144,147-154
Work group attendance (WGA), 185-190
Work role, reducing/withholding inputs
from, 116
Work role withdrawal: see Withdrawal
Work satisfaction: see Job satisfaction