Você está na página 1de 27

Explicating Social Capital: Trust and Participation in the Civil Space

Author(s): Gerry Veenstra


Source: The Canadian Journal of Sociology / Cahiers canadiens de sociologie, Vol. 27, No. 4
(Autumn, 2002), pp. 547-572
Published by: Canadian Journal of Sociology
Stable URL: http://www.jstor.org/stable/3341590
Accessed: 29/12/2009 04:59

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/action/showPublisher?publisherCode=cjs.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.

Canadian Journal of Sociology is collaborating with JSTOR to digitize, preserve and extend access to The
Canadian Journal of Sociology / Cahiers canadiens de sociologie.

http://www.jstor.org
Explicating Social Capital:Trustand
Participationin the Civil Space*

Gerry Veenstra

Abstract:Accordingto some commentators,social capital,i.e., the natureof trustand participation


in the civil space, may influencepopulationhealth,politicalperformanceand/oreconomic growth.
In the social capitalliterature,however, the relationshipbetween trustand participationin social
networksin the civil spaceis underdevelopedtheoreticallyandempirically.Respondentsto a survey
in Saskatchewan,Canadatrustedpeople from nearbycommunitiesthe most, experts and profes-
sionals less stronglyandgovernmentsleastof all, suggestinga multifacetednotion of trust,although
those who trusted one referenttended to express trust for another,suggesting continuity to its
characterin the self. Religious affiliationproved one of the most salient predictorsof trust,sug-
gesting an intriguinglinkbetween faithandtrust.Respondentswho participatedin many secondary
associationsin the civil space were relativelymoretrusting,as were those who participatedin co-
operativeassociationsand in groupswith an interestin furtheringsome common good.

Resume:Selon certainscommentateurs,le capital social, c.-a-d. la mesurede la confiance et de la


participationa la vie civique, pourraitinfluencerla santedes populations,la performancepolitique
et/ou la croissance economique. Dans la litteraturedu capitalsocial, toutefois, la relationentre la
confiance et la participationaux reseaux sociaux du milieu civique est sous-developpee, tant du
point de vue theoriquequ'empirique.Les sujetsinterrog6slorsd'unsondagemene en Saskatchewan
(Canada)accordaientleur plus grande confiance aux gens des collectivit6s avoisinantes, mais
avaientmoins confianceaux expertset aux professionnels,et moins encore aux gouverements, ce
qui suggere une notion polyvalente de la confiance. Par ailleurs,les personnesqui disaient avoir

* Datacollection was facilitatedby the HEALNet-RegionalHealthPlanningTheme, Saskatoon,


Saskatchewan.The Social Sciences and HumanitiesResearchCouncil of Canada (SSHRC)
providedthe authorwith a doctoral fellowship for 1996-98, and SSHRCand the Centrefor
HealthServices andPolicy Researchat UBC providedthe authorwith postdoctoralfellowships
for 1998-2000. BrianWilson andthe reviewersmade helpful suggestions regardingan earlier
draftof this article.

CanadianJournalof Sociology/Cahierscanadiensde sociologie 27(4) 2(X)2 547


548 Canadian Journal of Sociology

confiancea l'unedes categoriesmentionn6estendaienta exprimerleurconfianceenversune autre,


ce qui suggereune continuit6de son caractereau sein du moi. L'appartenance religieuses'estavere
etrel'unedes meilleuresvariablespr6dictivesde la confiance,ce qui suggereun lien fascinantentre
foi et confiance.Les sujetsinterrog6squi participaienta de nombreusesassociationssecondairesau
sein dumilieucivique6taientrelativementplusdisposesa faireconfiance,toutcommel'etaientceux
qui participaienta des associationscooperativeset a des groupesvou6s a la poursuited'unbien
commun.

Introduction
The notion of social capitalhas attainedsome prominencein ratherdisparate
arenasof discourse.Researchersarguethatcertaincharacteristicsof the civil
space, i.e., trustandparticipationin social networks,mayconstitutea valuable
resourcefor social groups,communitiesor societies. In the world of public
health research,for example, social capital, or social cohesion, a closely re-
lated concept, is thoughtto constitutean importantdeterminantof the health
of populations(e.g. Wilkinson,1996;Kawachietal., 1997;LynchandKaplan,
1997; Coburn,2000; Hawe and Shiell, 2000; Kawachiand Berkman,2000;
Lynch et al., 2000; Putnam,2000; Veenstra,2000, 2001, 2002). Kawachiand
Berkman(2000) speculatethatinterpersonaltrustandparticipationin networks
may influence the health of individualsdirectly, e.g., by providing social
supportforpeople, butalso indirectly,e.g., interpersonaltrustmay accompany
egalitarianpatternsof participationthatinfluencehealth-relevantstatepolicies
pertainingto education,transportationor the distributionof wealth.
In like manner,social capital is thoughtto promote the performanceand
characterof political institutions(e.g. Putnamet al., 1993;Rice andSumberg,
1997;VeenstraandLomas, 1999;Coburn,2000) andeconomicgrowthandde-
velopment (e.g. Helliwell and Putnam, 1995; Knack and Keefer, 1997;
Temple, 1998; Woolcock, 1998, 2001). Usually stimulatedby the portraitsof
social capitalpresentedby Coleman(1988) andPutnamet al. (1993), empirical
measuresof said capital generally seek to measurequantitiesof social trust
and/orparticipationin secondaryassociationswithinsocieties or communities
(e.g. Putnamet al., 1993; Kawachi et al., 1997; Putnam,2000; Veenstra,
2002). Although the implicationsof participationand trustfor the shape of
social structureand thence outcomes such as populationhealth, the perfor-
mance of political institutionsor the natureof economic activity have been
empiricallyexploredat some length,relationshipsbetweentrustandparticipa-
tion in the civil space have not receivedthe detailedexplorationthey deserve.
The gap between theoreticallysophisticatedconceptualizationsof trustand
participationin the civil space on the one hand,andtheirempiricalapplication
to the appliedresearchthatutilizes these conceptson the other,is quitebroad.
Forms of voluntary or secondary organizations,in particular, "provide a
throwawayline withindiscussionsof trustor communityas desirablemodels
in theCivil Space 549
TrustandParticipation

of association, while remaining fairly unexamined on their own terms"


(Tonkiss and Passey, 1999:258). Measures of the 'weak ties' described by
Granovetter(1973) that are supposedly inherent in such organizations and
essential for communicating information and facilitating co-operation, for
instance, have rarelybeen assessed directly (Berkmanet al., 2000), let alone
studied in relation to a complex notion such as trust.
Using data drawn from a survey of adults in Saskatchewan,Canada,this
articletests hypothesesconcerningthe inter-relatednessof some characteristics
or types of trust and empirical relationshipsbetween these forms of trustand
one form of participationin the civil space, i.e., relationships among social,
political and expert trusts and participation in secondary or voluntary
associations. It seeks to elucidate the complexity of trustas well as the degree
to which it/they is/are fostered by participation in some kinds of secondary
associations more than in others. Are there dimensions to or types of trust?
Does the natureof an individual'sparticipation(e.g. breadthor depthof parti-
cipation) matterfor the developmentand expressionof trust?Do characteris-
tics of associations themselves influence the expression of trust by respon-
dents? If the answerto any one of these questions is 'yes' then we have good
reason to argue that empirical studies utilizing the notion of social capital
should conceptualize and measuremultidimensionalconceptions of trustand
participationin secondaryassociations;that certainforms of trustand certain
kinds of associations may be more relevant for the development of 'social
capital' than others. The resultspresentedhere will aid futurestudies seeking
to determine the precise characteristicsof lived public life that influence
health, the polity and the economy.

The Nature of Trust


The debateregardingtrusthas in partdisplacedthe long-standingdebateon the
natureof civil society itself (Tonkiss and Passey, 1999). Social theoristssuch
as Luhmann (1980), Giddens (1984; 1987; 1990), Fukuyama (1995) and
Miszthal (1996) have craftedintricatetheoreticalportraitsof trust in moder-
nity, enablingus to distinguishtrustexpressedby individualsfrom trustwithin
social relationships and trust within social systems, trust from risk and
contingency, and dangerfromfortuna, for example. Does trusthave variants
and hues that ought to be explored and accountedfor? Should we investigate
the natureof the complex social contexts in which trustis supposedlyfostered?
It does and we should, starting with the seemingly self-evident notion of
'trust.'
Giddensdefines trustas "confidencein the reliabilityof a personor system,
regardinga given set of outcomes or events, where that confidence expresses
a faith in the probity or love of another, or in the correctness of abstract
Journalof Sociology
550 Canadian

principles"(1990:34). It involves confidenceand faithin certainattributesof


anotherperson,systemorprinciple.Trusthas been linkedto valuesof honesty
and fairness (Tonkiss and Passey, 1999), and can rest on the degree of
predictability,faith or co-operationthatprevailsin a given situation,on ex-
pectationsthatthe otherwill act in ways thatarecompetentandcaring,andon
perceptionsof commitmentto a goal and the fulfilmentof obligations(Kas-
personet al., 1992). To Simmel, trustis belief in someone, a belief that they
possess some consistency upon which one can rely, based on reasonsbut not
explained by them. Thus, according to Luhmann(1980), trust recognises
alternatives- the otherneed not follow predictablepathsof action- but, to
Giddens (1990), smoothes away the constantanxiety or dreadthat can grow
should extreme focus upon such alternatives(dangers)become the norm.As
such trustdoes involve a measureof rationalassessment,but it is more than
simply the recognition of contingencies because it has a rathercontinuous
character,more so thanmight be predictedby exclusive focus on the assess-
ment of risks (Giddens, 1990). It's continuity in the self has psychological
consequences,as a 'moralhostageto fortune'(Giddens,1990)is given. Govier
notes that"mostof us maintainconsiderabletrustmost of the time"(1997:3).
Trust invokes concepts such as 'perception,' 'confidence,' 'faith' and
'belief.' These are located in the individual mind, wherein each trusting
relation can be thoughtto have a (trusting)subject and a (trusted)referent.
Miszthal(1996) distinguishesi) trustas a psychologicalattributefromii) trust
as a propertyof social relationshipsand iii) trust as a propertyof social
systems, however,therebyadditionallylocatingdimensionsof trustbeyondor
outsideindividualmindandperception.Someindividualsmightbe morelikely
thanothersto bringa trustingattitudeto all social interactionsand towardall
institutions,possessing a degree of 'personaltrust.' On the otherhand, some
social relationshipsmight be deemed trusting relationships,regardless of
whetheror not the participantsin the interactiontend to bring trustto other
interactions.To assess this 'interpersonaltrust'we could ask individualsabout
trust pertaining to the relationship,but would do better to question both
membersof the relationand additionallyassess whetherthe relationexhibits
trustingactions. Thus trustcan have an inter-subjectivereality, an emergent
quality beyond perceptionlocated in the individual mind, althoughit is not
clear the degree to which an innate propensityto bring trust to a relation
influences the developmentof interpersonaltrustwithin relations, and vice
versa. Expandingfurtherinto social structure,some social systems might be
predicated upon trustingrelationshipsspanning time and space, patterned
relationslocated in space, with a historyand a future,whatmight be referred
to as 'systemtrust.'These patternsof thoughttrustandtrustingactions likely
have a more lastingontologicalrealitythandoes a sometimesephemeraland
contextually-specificinterpersonaltrust.
TrustandParticipation
in theCivil Space 551

Miszthal (1996) suggests that high interpersonaltrust may be positively


relatedto high societal (i.e. system) trust.It seems sensible to supposethat the
constituentpartsof a social system (actorsand social relationships)influence
and reflect characteristics of the system. It also makes sense to wonder,
however, whethercertainformsof personalor interpersonaltrustinfluence and
reflect system trustwhen othersdo not. It may be that strongpersonaltrust-
trust in people in general, from the standpointof the individual - forms a
basis of system trust, but that trust directed toward or confined to sub-com-
munities (i.e. interpersonaltrust) does not. Forms of interpersonal trusts,
expressed personally via their respective referents,might dispose themselves
differentlyamong geographicallydefined communities(e.g. neighbourhoods,
cities or societies) and among communities not necessarily located in space
(e.g. class, religious or ethnic groups),formingvicious (i.e. highly mistrustful)
or virtuous (i.e. highly trustful) cycles of trust, (mis-) trust breeding more
(mis-) trustin a formof self-fulfilling prophesy.This in-communitytrustmight
then affect trustrelations elsewhere in positive or negative ways. Yamagishi
et al. (1998) proposethatintense groupties prevent trustfrom developing be-
yond groupconfines. Stolle (1998) questionswhethertrustdirectedtowardor
confined within sub-communitiesnecessarilytranslatesinto trustin people in
general, as the 'continuousnatureof trust'thesis wouldpredict,or whetherthe
opposite relationshipis in fact the norm,such thatstrongin-grouptrustneces-
sitates mistrustdirectedtowardlargercommunities.
Communitiesof trustcan come in many forms. Giddens (1984) describes
ever-wideningzones, extendingfromthe nuclearfamily andmoving outwards
through the workplace, neighbourhood,city, nation-state and world-wide
system, within which individuals are positioned. Day-to-day physical inter-
action with othersmay producepatternsof interpersonaltrustthat differ from
those expressions of trustin others more spatiallyremoved from the subject.
Trust in neighboursmay differ substantiallyfrom trustin people from one's
community, with many of whom the individualhas not interactedpersonally.
Trust in people in general may differ from trust in people from one's
community, those with whom one might interact on a given day. Giddens
(1990) suggests thatthe pre-modernenvironmentsof trustare being gradually
replaced by newer versions. According to him, the pre-modernenvironments
were kinship relations, the local community (rooted in place), religious
cosmologies, and tradition(the routineswithin which beliefs are organized).
Supposedly, these are losing groundin the face of the newer communities of
trust,those of relationsof friendshipand sexual intimacy and trustin abstract
systems. The primacy of place is (being) slowly destroyed, and as place be-
comes less centralto social relationships,trustincreasinglybecomes relatedto
absence in time and space and the incomplete informationthat accompanies
such absence (Giddens, 1990). Increasinglyin modernity,then, interpersonal
Journalof Sociology
552 Canadian

social trust (very much local) is potentiallybecoming replacedby a trustin


abstractandexpertsystems(verymuchglobal)thatmaynow formtheprimary
basis of social organization.
As such trust need not necessarily have an exclusively social referent.
Expert systems are "systems of technical accomplishmentor professional
expertisethatorganiselarge areasof the materialand social environmentsin
which we live today" (Giddens, 1990:27). Political institutions might be
consideredto be one such entity, andtrustin politiciansand government(i.e.
political trust)might reflect the trustworthinessof theirestablishedexpertise.
Is interpersonalsocial trustbeing replacedby trustin such systems, and, if so,
is it necessarilytrue,as this interpretation
of Giddens'argumentsuggests,that
these forms of trust be negatively correlatedwith one another'?Contrarily,
Miszthal (1996) and Tyler (1998) suggest that social trustcan be positively
affectedby the credibilityof government.Kaspersonet al. (1992) also suggest
that political trustdependsheavily on the performanceof social institutions.
If trustis moreor otherthana rationalcalculationof risksanddangers,instead
possessing a continuous state in individuals and social relationships,then
perhapsits continuityextends from social to political trust,and vice versa. If
trust is fostered in social relationships,then perhapsall kinds of trust are
fostered in the same set of social relationships.

Trustand Participationin SecondaryAssociations

Psychologists such as ErikErikson(1950) have exploredthe developmentof


personaltrustduringchildhood,locating the formationand maintenanceof a
lasting personaltrustin the dynamicsof family relationships.More recently,
however, researchershave begun to explore the developmentand expression
of trust among people within the civil space, within relationshipsoutside
family and kinship groupsand distinctfrom those networkslocated in work
and economy. These civil relationsinclude so-called 'bonding'relationships
with friends and neighbours, 'bridging'relations with distant friends, col-
leaguesandassociates,andparticipationin secondaryorvoluntaryassociations
not directlyfinancedor maintainedby the state.Thusfar,severalstudieshave
demonstratedsignificantrelationshipsbetweenrudimentarymeasuresof par-
ticipationin secondaryassociationsand trust(e.g. Putnam,1996; Brehm and
Rahn, 1997; Uslaner, 1998; Shah, 1998; Stolle, 1998).
Most such studieshave been conductedin the UnitedStates,at the level of
the individual.Utilizing a pooledsamplefromthe GeneralSocial Survey,from
1972-94, with annualresponseratesrangingfrom72%to 79%and an overall
sample size greaterthan 9,000, Brehm and Rahn (1997) posit a reciprocal
relationshipbetween interpersonaltrustand civic engagement(measuredby
the number of membershipsin voluntaryassociations) and, via structural
TrustandParticipation
in theCivil Space 553

equation modelling that included a multitudeof exogenous and endogenous


factors,found a strongempiricaleffect from such engagementto interpersonal
trust. Probit analysis by Uslaner (1998) among responses on the same GSS
(N=1520), but for one time period only, also found an effect for such
membershipon trust.Shah (1998) used the DDB Needham life cycle survey,
with a response rate of 72% and a sample size greaterthan 3,000, to find that
a similar measure of civic engagement was strongly related to a similar
measureof interpersonaltrustin two-stage least squareregressions,also with
the inclusion of multiple factors. These are compelling results drawn from
nationallyrepresentativesamples:in the United States at least, participationin
secondary associations seems to be ratherstronglyrelated to expressed trust.
Perhapsless compelling statistically, but as provocative theoretically, Stolle
(1998) also determinedan effect for associational membershipon trust in a
pooled sample of Germansand Swedes, drawingupon a sample (N=484, 95%
responserate)of membersof 30 associationalgroups,notinga 'trustboost,' an
increase in social trust,uponjoining a new association.
Which comes first, participationin secondaryassociations within the civil
space (action) or interpersonaltrust(perception,feeling)? Trustcan be viewed
as the basis for voluntary association: "agents do not come together in
voluntaryassociationson the back of a contractualrelationship,but insteaddo
so on the basis of trust"(Tonkiss and Passey, 1999:261). Reciprocally, "the
more that citizens participatein theircommunities,the more thatthey learnto
trustothers"(BrehmandRahn, 1997:1001).Thatis, "membershipin voluntary
associations should increase face-to-face interactions between people and
create a setting for the developmentof trust"(Stolle, 1998:500). The issue of
causal priority is not easily answered: complex research designs aside, a
researcher'stake on the issue will rest upon her/his theoreticalorientationto
the agency-structureand mind-action dichotomies. This article adopts the
perspective taken by Woolcock (2001) and Putnam(2001), wherein trust is
seen to be a productof social interactionand social networks,resultingfrom
social capital (a by-productof relationships)ratherthanforminga constituent
part of social capital (a cause of certain kinds of relationships),an approach
that privileges (participationin) social networksover trust,the latter serving
as the dependentvariablein the multivariateanalysisshowcasedin this article.
There is some empirical supportfor contendingthat interpersonaltrustflows
from social interactionmore readily than it provides a necessary stimulantfor
social interaction. The structuralequations described by Brehm and Rahn
(1997) showed a strongereffect for civic engagementon trustthanthe reverse,
as did the structuralmodel provided by Shah (1998). On the other hand,
Uslaner's (1998) three-stageleast squaresestimation led him to deduce that
trust influences membership in associations ratherthan the reverse. In his
sample of Germansand Swedes, Stolle (1998) showed a self-selection bias:
554 Canadian
Journalof Sociology

even given a 'trustboost' uponjoining a new association,trustingpeople are


more likely to join associations,and more trustingpeople are more likely to
join more diverse associations.The most sensible conclusionto adoptis that
trustandparticipationareindeedreciprocallylinked,at least at the level of the
individual,and that a focus on one directionof causality alone can only tell
partof the story.
It would be unwise to assertthatinterpersonaltrustof all kindsis relatedto
all aspectsof participationin all kindsof secondaryassociationslocated in the
civil space. We should explore trustin its variedmanifestations,the various
types of participationand types of secondaryassociations,and then relation-
shipsamongthesemanifestations.Beyondthe diversewaysof conceptualizing
dimensions of trust,discussed above, "[w]e do not know whether trust and
cooperativeattitudes[..] are a functionof a particulartype of involvementor
a special type of group"(Stolle, 1998:499).It may be that the way in which
people participatematters for trust, or that participationin some kinds of
networksof associationwithin the civil space mattersmost for the develop-
ment of interpersonalandpersonaltrust.Thusparticipationin a wide breadth
of groupsmight foster generalizedtrustbut intenseparticipationin only one
groupmight decreaseit: e.g. we trustone anotherto the deathbut mistrustall
others.Group-specifictrustmay be fosteredin demographicallyhomogeneous
groupswhile generalizedor personaltrustis not:e.g. we distrustanyone who
is not like us. Political mistrustmight be particularlyfosteredin politically
orientednetworks:e.g. we workto achieve the goals thatgovernmentclearly
cannotattainon its own.

Research Questions
This articleempiricallytestshypothesesconcerningvarious'hues' of trustand
their relationshipswith participationof a certainkind in the civil space via
analysis conductedat the level of the individual.This exercise will point to
some of the types of secondary associations in the civil space that may
constitutethe social resourcethatis describedin the social capitaldiscourse.
First,the articleassesses the degree of geographically-specificsocial trust
professed by respondents for their neighbours, people from their own
communityandpeople fromtheirpartof the province(i.e. region). These are
comparedwith social trustexpressedtowardmembersof two (possibly)non-
spatialcommunities,membersof respondents'religiousandethniccommuni-
ties, and with trustin people in general (an avowal of personal trust by the
individual).Two formsof trustin abstractor expertsystemsarealso compared
with these: trustin experts and professionals (i.e. trustin representativesof
scientific knowledge)andtrustin governments(i.e. politicaltrust).Hypothesis
one: following Giddens(1990), the social formsof trustare weakerthantrust
TrustandParticipation
in theCivil Space 555

in political and expert systems, suggesting that the 'primacy of place' has
indeed been supplantedby 'placeless' forms of trust.
Second, the article assesses the degree to which these dimensions or ex-
pressionsof trustaremutuallyinclusive (i.e. positively correlated)or exclusive
(i.e. negatively correlated), leading to hypothesis two: trust is not especially
fostered within certain communities, necessitating mistrust toward those
belonging to othercommunities,butinsteadhas a continuousnaturein the self.
Third, the article assesses whether these forms of trust are concentrated
within specific geographic communities and demographic categories. In
particular,given that the survey was administeredin eight health districts of
Saskatchewan,Canada,it determineswhethersome healthdistrictshave more
trustingresidents than do others. It also assesses whether these measures of
trust accrue differently within social classes and demographic categories.
Hypothesis three:higherstatuspeople aremore trusting,following otherwork
in the United States (e.g. Putnam, 1996; Brehm and Rahn, 1997; Uslaner,
1998; Shah, 1998), as are older people (e.g. Putnam, 1996; Uslaner, 1998;
Shah, 1998) and those who live in ruralsettings (e.g. Putnam, 1996).
Fourth, the article determines whether the forms of trust are fostered by
strong or weak ties among people participatingin one kind of network, the
'social club' or secondaryassociation."[S]ocialcapitalis built on the effect of
various contacts and regular interactionsof groups of people who stand in
ratherloose contactto each other[..] Priorresearchimplies thatstrongergroup
bondsareusuallydevelopedin interactionswith morelike-mindedandsocially
similar people, whereas weaker bonds are indicated by group diversity or
greatersocial distancebetween groupmembers"(Stolle, 1998:501-504). Hy-
pothesis four: a preponderanceof weak ties correspondswith higher levels of
trustexpressed by respondents.
Fifth, it determineswhetherthe structureof secondaryassociationsre. lines
of power (as assessed by respondents)is related to the expression of trust.
"[R]elationshipswithin verticalnetworks,because of theirasymmetry,are not
able to create similar experiences of mutuality and reciprocity to the same
extent as relationships in horizontal networks could" (Stolle, 1998:502),
leading to hypothesisfive: participationin informal(versusformal)secondary
associations is positively relatedto the expression of trust.
Sixth, is some attentionto a common good, versus exclusive attentionthe
needs of the secondary association itself, related to the expression of trust?
According to Stolle (1998: 501), the social capital discourse tends to devalue
"membershipsin newly developed self-help groups,where a strong 'I' orien-
tation (as opposed to a 'we' orientation) might prevent the experience of
successful cooperation,"leading to hypothesissix: participationin a groupthat
strivesto meet the needs of some largercommunitymitigatesan 'I' orientation
and thusly fosters social trust.
556 Canadian
Journalof Sociology

Seventh and finally, the article determineswhetherthe forms of trustare


strongerfor people who claim to belong to secondaryassociationsthatare co-
operativelyoriented.The relationshipbetween collective action and trustis
integralto the social capitaldiscourse,leadingto hypothesisseven: participa-
tion in groups that are deemed to be co-operativeby participantsfoster the
expressionof trust.

Methods

Eight of 33 healthdistrictsin Saskatchewanwere selected for administration


of a survey of randomly selected citizens. The natureof the largerproject,
which exploredrelationshipsamongsocial capital,the political effectiveness
of regional health authorities,community wealth, income inequality and
mortality rates (Veenstraand Lomas, 1999; Veenstra, 2000, 2002), led to
inclusionof the two urbandistricts,two of the fourmid-sizeddistrictsandtwo
of the 27 ruraldistricts.A CDROMfromthe companyPro-CD,containingthe
phone numbers, names, addresses and postal codes of the households in
Saskatchewanwith a listed telephone,providedeight samplingframes.From
each a randomsampleof households,220 fromeach of the largerdistrictsand
180 from each of the smaller ones, was chosen. A four-partstrategy was
pursued,mailing the questionnairetwice and reminderpostcardstwice in the
summer of 1997. The recipient was asked to give the questionnaireto a
member of the household,selected randomlyby listing the membersof the
household, 18 years of age or older, in orderof birthdaywithin the year, and
choosing the person whose birthdayfalls earliest in the year. 534 question-
naireswere returnedfor an overallresponserateof 40.3%, which, becauseof
the natureof bulk mailing (where non-deliveredmail is not necessarily re-
turnedto the sender)is probablyan underestimate.To gauge non-representa-
tiveness, aggregated age, gender and income characteristicsof the samples
were comparedto knowncharacteristicsof the districts.Most of the district
samples were slightly biased toward higher income, female and older
respondents.
To measurepersonaltrust,respondentswereaskedto agreeordisagreewith
the statement:'Mostpeople can be trusted.'The referentsfor items assessing
formsof interpersonalsocial trustwere respondents'neighbours,membersof
theircommunitiesandregionsof theprovinceandmembersof theirethnicand
religiouscommunities.Otheritems assessingparticularaspectsof trustworthi-
ness exploredhonour(i.e. othersupholdtheirobligations,they are honest and
committed to fairness), safety (i.e. the risk of being hurt by others is low,
communitymembersaregenerallytrustworthy)andwillingnessto help others
(i.e. others are caring, their intentionsare honourable).To measurepolitical
trust, respondentswere presenteditems assessing respondents'trustin four
TrustandParticipation
in theCivil Space 557

levels of governmentin Saskatchewan,namely, the local municipal, District


Health Board, provincial, and federal governments.They were asked to rate
the performanceof these governmentsat problem solving (i.e. confidence or
faith, perceptionof competence) and asked to assess the degree to which they
trust governmentaldecisions and whether governmenthas the public's best
interests at heart (i.e. perception of honourablemotives, commitment to a
goal). Finally, to measureexperttrust,respondentswere askedwhetherexperts
and professionals can aid in solving communityproblems.
To measurebreadthof participationin secondaryassociations,respondents
listed the clubs or secondary associations they currentlyparticipatein, to a
maximum of ten, producinga positively skewed variablewith a mean of 2.4
and standarddeviation of 2.1. Three hundred and seventy respondentsaddi-
tionally answeredspecific questions about the characteristicsof at least one
and up to three secondaryassociation(s) with which they are 'most involved
with.' With respect to these groups, questions were asked pertaining to the
depth, breadth and nature of participationin the group by the respondent,
followed by questions assessing organizational formality, strength of ties,
densityof relations,socio-demographicheterogeneity,altruistictendenciesand
degree of co-operation.The survey items are delineated in the appendix.
In bivariateanalysis, when the two variableswere categorical,SPSS 10.0.7
was used to calculate the chi-squarefor the test of significance and Cramer's
V for the measure of association. When one variablewas categorical and the
other quantitative,the ONEWAY ANOVA provided the test of significance
and eta the measureof association.When both variableswere continuousand
the relationship was linear, Pearson'sr provided the measure of association,
but when a variable was ordinalratherthan interval or ratio in nature,or the
assumption of linearity was violated, Kendall's tau_b, a non-parametric
measure,was used instead.When constructingindices, PrincipalComponents
Analysis was used to identify clusters of variables and then scale reliability
analysis was used to demonstratethe reliability and coherence of resultant
indices. In multivariateanalysis, multiple linear regressionwas the technique
of choice.

Results

The Nature of Trust


Table 1 describes aggregatedresponses to selected items pertaining to trust,
where ratherstarkdifferences in strengthof trustwere found. Trustin people
from spatially-definedcommunitiesand personaltrustwere distinctly strong-
er thantrustin experts andprofessionals,which in turnwas strongerthantrust
in governments. Of the spatially defined communities, the spatially-near
558 Canadian Journal of Sociology

Table 1. Aggregate scores for indicators of trust

Mean response %agree


(n, sd) at least %agree
Questionnaireitem (95%CI) a little strongly

My communityis a prettysafe place. 2.07 (530, 1.33) 89.4 38.9


(1.96 .. 2.19)
Most people in my communityare 2.12 (529, 1.21) 88.8 34.0
willing to help if you requireassistance. (2.02 .. 2.22)
Most people in my neighbourhoodcan 2.14(532, 1.21) 88.5 33.1
be trusted. (2.03 .. 2.24)
Most people in my religious/spiritual 2.34 (519, 1.33) 77.3 31.0
communitycan be trusted. (2.23 .. 2.46)
Most people in this communitycan be 2.44 (531, 1.34) 81.7 23.7
trusted. (2.32 .. 2.55)
Most of the people who live in my part 2.45 (530, 1.14) 83.8 16.8
of Saskatchewanare honourable. (2.35 .. 2.55)
Most people can be trusted. 2.60 (528, 1.39) 80.9 19.1
(2.48 .. 2.72)
Most people in my ethnicgroupcan 2.61 (513, 1.36) 70.8 22.2
be trusted. (2.50 .. 2.73)
When it comes down to it, you can
always trustthe people in my part 2.67 (529, 1.27) 80.9 11.7
of Saskatchewan. (2.56 .. 2.78)
Expertsand otherprofessionalscan help 3.05 (516, 1.40) 72.5 10.7
solve problemsin my community. (2.93 .. 3.17)
AlthoughI may have some complaints
aboutsome decisions the DistrictHealth
Boardmakes,I trustit to makegood 3.68 (520, 1.60) 52.1 4.2
decisions. (3.54 .. 3.82)
The local governmentwill tell the public
whatthey need to know aboutrelevant 3.80 (517, 1.53) 50.3 3.9
issues in my community. (3.67 .. 3.93)
The provincialgovernmenthas the 3.86 (520, 1.76) 52.5 5.4
public's best interestsat heart. (3.70 .. 4.01)
AlthoughI may have some complaints
aboutsome decisions the federal
governmentmakes,I trustit to make 3.87 (521, 1.71) 53.4 3.8
good decisions. (3.72 .. 4.01)

Each item was a 7-point Likert scale ranging from 1 (agree strongly) to 7 (disagree strongly)
TrustandParticipation
in theCivil Space 559

communities of neighbourhoodand community garneredthe strongest ex-


pressions of trustoverall. Of the variantsof political trust,there was little to
choose amongtrustin differentlevels and types of governments.1Thus, on the
positive side, approximately89%of respondentsexpressedat least some trust
for people from theirneighbourhoodsandcommunities,while on the negative
side, only 53%of respondentsexpressedat least some trustfor the federalgov-
ernment.Given the magnitudeof this difference it is by no means clear that
trusttakes on a single formfor a given respondent,that 'trust'is a single some-
thing, thereby implying that various forms of trust may create differently
textured social spaces with subsequentimplications for health, politics and
economic activity. It could still be the case, however, that those who trust
membersof the communityare still more likely thanothersto trustthe federal
government,even if the absolutemagnitudeof trustof the one is less thanthe
otherin the aggregate:thatis, these expressionsof interpersonaltrustmay still
manifest or reflect either or both a personal trust or a province-wide system
trust.
PrincipalComponentsAnalysis was conductedon twenty items pertaining
to trustin most people, neighbours,membersof the community,region, ethnic
and religious communities,experts and professionalsand four levels of gov-
ernment, utilizing the Varimax rotation method to identify three factors
accountingfor 58%of the total variability(Table 2). The first factorrepresents
the nine items pertainingto social trust,the second factorrepresentstrustin the
District Health Board and local municipal government, and the third factor
representstrustin the federalandprovincialgovernments.Trustin expertsand
professionals did not load highly on any of the factors. This analysis demon-
stratesa cleardemarcationamongsocial trust,trustin largergovernments,trust
in smaller governments,and trustin experts and professionals.
The PCA results serve to guide the creationof appropriateindices, in this
case by illuminatingthe coherentnatureof items pertainingto a trustin other
people (i.e. social trust), the coherence of items pertainingto trust in gov-
ernmentsand the uniquenatureof trustin expertsandprofessionals.Social and
political trust indices were created to provide measures of reliability and to
simplify subsequentinvestigationinto the concentrationof trustwithinvarious
geographic and non-spatialcommunitiesand relationshipswith participation
in secondaryassociationswithinthe civil space. As delineatedin the appendix,
those items pertaining to trust in community members and others from the
region were collected togetherto weight equally the geographicreferents.The

1. This finding supportsJennings' (1998) work in the 1990s, where local and state governments
in the United Statesgarneredlevels of trustsimilarto thatawardedthe federalgovernment.He
argues that the decline in federal trustfrom the 1960s and onwardprecipitateda movement
towardthe devolutionof decision-makingin thatcountry.
560 CanadianJournalof Sociology

Table 2. Rotated component matrix resulting from analysis of 20 trust items

Factor
1. 2. 3.
Social Trust Trust
trust small gov't large gov't

1. Most people can be trusted. .838


2. Most people in my neighbourhoodcan be trusted. .815
3. Most people in this communitycan be trusted. .808
4. Most of the people who live in my partof
Saskatchewanare honourable. .782
5. Most people in my ethnicgroupcan be trusted. .755
6. Most people in my communityare willing to help if
you need assistance. .741
7. Whenit comes down to it, you can always trustthe
people in my partof Saskatchewan. .740
8. Most people in my religious/spiritualgroupcan
be trusted. .718
9. My communityis a prettysafe place. .624
10. How would you ratethe performanceof the current
federalgovernmentin solving problemsin Canada? .782
11. AlthoughI have may have some complaintsabout
some decisions the federalgovernmentmakes,I trust
it to make good decisions. .754
12. How would you ratethe performanceof the current
provincialgovernmentin solving problemsin
Saskatchewan? .405 .680
13. The provincialgovernmenthas the public'sbest
interestsat heart. .462 .678
14. How would you ratethe performanceof yourDistrict
HealthBoardin solving problemsin yourdistrict? .805
15. AlthoughI may have some complaintsaboutsome
decisionsthe DistrictHealthBoardmakes,I trustit
to make good decisions. .746 .328
16. When the DistrictHealthBoarddecides whatpolicies
to adopt,how muchattentiondo you thinkthey pay
to whatthe generalpublicthink? .727
17. How would you ratethe performanceof yourlocal
governmentin solving problemsin your community? .671
18. The local governmentwill tell the publicwhatthey
need to know aboutrelevantissues in my community. .642
19. In general,how muchof the taxpayers'money, if
any, is wastedby yourlocal government? .504
20. Expertsand otherprofessionalscan help solve
problemsin my community.
Eigenvalue 7.199 3.180 1.217
Percentof explainedvariance 36.0 15.9 6.1
Cumulativepercent of explainedvariance 36.0 51.9 58.0
Rotation method: Varimax with Kaiser Normalization
Only loadings exceeding 0.30 in magnitude are shown, N = 397
Trust and Participationin the Civil Space 561

social trustindex formed an internallycoherentand reliable index with Cron-


bach's alpha= 0.906 (Appendix),whereinall items assessing formsof personal
andinterpersonalsocial trustwere positively, stronglyandsignificantlyrelated
to one another(resultsnot shown). Given the clear distinctionin the literature
between social and political trust, all items pertainingto political trust were
collected into one index, with an alphaof 0.871 (Appendix).2The two indices
were positively and significantly related to one another(r=0.347, p<0.001),
corroboratingfindings in the United States(e.g. BrehmandRahn, 1997; Shah,
1998), and both social and political trust were positively related to trust in
experts andprofessionals(r=0.231, p<0.001; r=0.330, p<0.001, respectively).
Results from this exploration into the nature of expressed personal and
interpersonal social trust, where most inter-relationshipswere strong and
positive, may mean that individuals either participatein (non) trustingrela-
tionships in general, or that they have instead, or additionally, developed
strongpersonalorientationsregardingthe (perhapsindiscriminate)expression
of trust. Respondents who trusted one referent tended to trust another,
supportingthe notion of the continuousnatureof trust.They additionallygive
reason to suspect that social and political trust,if they do indeed congregate
within communities,congregatewithin precisely the same communities.As it
turns out, in this case they did not.
The two trust indices dispersed themselves differently among various
geographic and socio-demographic categories3.Social trust was strongerin
smaller,ruralcommunitiesthanit was in larger,more urbanones (eta=0.302,
p=0.001). Protestantswere more trustingthan Roman Catholics, who in turn
were more trusting than those with no religious affiliation (eta=0.237,
p<0.001). Older respondentswere more trustingalong both social (tau_b=0.
214, p<0.001) andpoliticallines (tau_b=0.147, p<0.001), andrespondentswho
were unemployed professed less political trust, on average (eta=0.127,
p=0.004). Widowed, married and common-law respondentsprofessed more
social trust than did divorced, separatedand single respondents(eta=0.177,
p=0.005). Home owners were more trusting, on average (with social trust,

2. Inthisinstance,socialtrustencompassesperceivedwillingnessto help andperceptionsof safety


and honour,butis best capturedby the simplepersonaltrustitem. The two items assessing trust
in the provincialgovernmentwere stronglyrelated,suggesting that perceptionof honourable
motives andperceivedcompetenceof governmentarerelateddimensionsof politicaltrust.The
inter-relationships amongthe threeitemsassessingtrustin the DistrictHealthBoardsuggestthat
perceivedaccountabilityformsanotherrelateddimensionof politicaltrust.Perceptionthatthe
local government wastes taxpayers' money was most weakly related to the other forms of
political trust,suggestingthat perceptionof fiscal ineptitudemay form a distinctdimension of
political trust distinguishablefrom the dimensions pertaining to honour, competence and
accountability.
3. Trust in experts and professionalswas relatedto maritalstatusonly, p=0.018 and eta=0.163,
such that divorced and separatedrespondentswere less trustingon averagethan others.
562 CanadianJournalof Sociology

eta=0.147, p=0.001), as were those respondentswith more children (with


social trust,tau_b=0.143, p<0.001). Lengthof time in the neighbourhoodwas
significantlyrelatedto both social (tau_b=0.230,p<0.001) and political trust
(tau_b=O.115, p<0.001). Gender,income andeducationalattainmentwere not
significantly related to the expression of trust. A multivariatemodel was
created from the socio-demographicvariablesusing stepwise regressionto
predictthe (transformedfor normality)social trustindex,eventuallyincluding
age, numberof yearslived in the neighbourhoodandreligiousaffiliationas the
representativeindependentvariables(Table 3). A similarprocess resultedin
a regressionmodel with age and unemploymentstatusas predictorsof the (al-
readynormallydistributed)politicaltrustindex (Table4). These models serve
as controlswhen investigatingrelationshipsbetweentrustandparticipationin
the public space, removing socio-demographiccandidatesfor spuriousness
from relationshipsamong trustand participationvariables.

Table 3. Multiple regression on transformedsocial trust index

beta sig.

Constant 5.949 <0.001


Age -0.120 -2.280 0.023
Numberof years in neighbourhood -0.027 -5.155 <0.001
Religiousaffiliation
RomanCatholic -0.133 - 1 970 0.050
Protestant -0.225 -3.232 0.001
Orthodox -0.009 -0.197 0.844
Other 0.030 0.651 0.516
(versusnone)

R-squared = 0.174, adjusted R-squared = 0.162


F = 14.580 (df = 6,414), p<0.001

Table 4. Multiple regression on political trust index

beta t sig.

Constant 26.648 <0.001


Age-0.208 -0.208 -4.793 <0.001
Unemployed 0.116 2.664 0.008
, ,, , ! ,, ,,,, ,,,,~~~~~~~~~~

R-squared = 0.060, adjusted R-squared = 0.057


F = 16.087 (df = 2,501), p<0.001
Trust and Participationin the Civil Space 563

Trustand Participation in SecondaryAssociations

This section investigates the generationof social, political and expert trusts
within the civil space throughattentionto attributesof participationitself in
secondary associations from the standpointof the respondentand to analysis
of characteristicsof the associationsthemselves. In both instancesthe analysis
remains at the level of the individual:in the latter instance respondentsare
deemed to be qualifiedinformantsconcerningthe groupsto which they belong.
Specifically, it will seek to determinewhetherparticipationin those associa-
tions built upon weak versus strong ties (i.e. bridging versus bonding social
capital) correspondsmore closely with individually-expressedtrust;whether
trustappearsto manifest itself differentlyin formal organizationsin compari-
son with informal ones (i.e. vertical versus horizontal lines of power); the
degree to which groups with altruistic goals and activities may explain
variability in trust;and whether trust appearsto be specially fostered in co-
operative groups.

Strengthof ties: The strengthof ties was measuredalong several lines. First,
it is assumed that participationin a breadth of secondary associations by a
respondentpresumesweakerties betweenthatpersonandthe membersof each
group. Second, it is assumed that depth of participationalso reflects the
strength of ties: respondentswho spend much time with a given group will
have craftedstrongerties than will those who spend less time therein, as will
those who have served in a position of leadershipin the group and those who
feel they have personalinfluence in the group.Third,the size of the groupwas
measured, assuming that smaller groups will tend to have strongerties than
will largerones. Fourth,the degree to which a groupis thoughtto be open to
newcomers was measured,this time assumingthat exclusiveness corresponds
with strongin-groupties. Fifth,it was determinedwhetherthe groupis thought
by the respondentto be organizedalong the lines of a commoninterest,in this
instanceassumingthatassociationsformedarounda single interest(i.e. to play
soccer, to play chess) are more engaged in bridgingpractices than they are in
bonding ones, as are those groups where members meet one anotherin that
settingalone. Finally, socio-demographicheterogeneity,alongthe lines of age,
communityof residence, ethnicity and religion, was measured,whereinmore
heterogeneityis assumed to correspondwith weaker ties.
Only a few of these measures proved meaningful when it came to
explaining variability in trust. Social trust was positively and significantly
related to the number of secondary associations within which respondents
currentlyparticipate,both before and after controlling for the socio-demo-
graphic model (r=0.163, p<0.001; p=0.015 with controls);political trust was
564 Canadian
Journalof Sociology

only relatedto the numberof groupsbeforesuch controlswere added(tau_b=


0.077, p=0.016). The empowermentimplied by personalinfluence was also
positively related to social trust after adding socio-demographiccontrols
(eta=0.188, p<0.001; p=0.016 with controls), but not to political trust
(eta=0.108, p=0.038). Of the measuresof socio-demographicdiversity only
similarityaroundreligion was relatedto social (eta=O.198, p<0.001; p=0.005
with controls)or political trust(eta=O.146, p=0.006; p=0.032 with controls)
aftercontrollingfor socio-demographiccharacteristics,suchthathomogeneity
along religious lines correspondedwith higher levels of trust.In sum, little
evidence was foundin this dataset suggestingthatstrongor weakties between
respondents and others, and weak ties within the secondary associations
themselves, plays a particularlypotentrole in the generationof trust.

Organizationalformality: Having an executive and formaldecision-making


rulesareindicatorsof organizationalformalityandserveto reflecta distinction
between leader-directed(i.e. vertical) and member-directed(i.e. horizontal)
forms of association. The level of organizationalformalitymeasuredin this
way did not play a statisticallysignificantrole in explainingvariabilityin trust
among these respondents,however.

Altruisticgoals: In this instance,political trust(eta=O.165, p=0.001; p=0.006


with controls) was strongerfor respondentsparticipatingin groups that do
things for others outside the group.That is, groupsthat direct their attention
beyond theirown needs had more (politically)trustingmembersthandid the
more 'egoistic' groups.This findingcould be explainedby religion. As noted
earlier,respondentswho claimed no religious affiliationwere markedlyless
trusting than that majority of respondentswho claimed affiliation with
Protestantor Roman Catholic churches. Respondentswere also asked how
often they attendreligious services: this variablewas positively and signifi-
cantly relatedto both social (tau_b=0.198,p<0.001; p<0.001 with controls)
and political trust (tau_b=0.083, p=0.016; p=0.080 with controls). The role
played by religious organizationsfor the generationof trustis not surprising,
given thatreligious communitiesform one of Giddens' (1990) environments
of trust, although social capital researchershave not yet given religious
institutionsthe attentionthey deserve.Wisdom(1953) claims thata personof
faith has a differentattitudetowardthe world thandoes the unbeliever,that
events are interpreteddifferently:theseresultssuggestan intriguingsimilarity
between religious faith and trust.

Co-operation:The supposed relationshipbetween trustand co-operationor


collective actionformsthe bedrockof social capitaltheory.In thisinstance,all
of social (eta=0.174, p=0.001; p=0.015 with controls), expert (eta=0.155,
TrustandParticipation
in theCivil Space 565

p=0.003; p=0.004 with controls) and political trust (eta=0.163, p=0.002;


p=0.007 with controls) were strongerfor respondentsparticipatingin groups
that are perceived by respondentsto co-operatewell.
In summary, respondents trusted people from nearby spatially-defined
communities the most and governmentsthe least, althoughthose who trusted
one referentdid tend to be more trustingtowardanother.There were distinct
sub-dimensionsto trust, a clear demarcationamong social trust (i.e. trust in
people with whom one might interactin the course of daily life), political trust
(i.e. trustin large and small governments)and trustin experts and profession-
als. Olderpeople were more trustingthan youngerpeople, suggesting that the
cohort effect arguedfor in the United States (Putnam,1996, 2000) may apply
to this context as well. Gender,educationandincome were not relevantfor the
expressionof trust.Among these respondents,religious affiliation proved one
of the most salient predictorsof (especially social) trust, suggesting a link
between religious faith and trust.Above and beyond the socio-demographic
characteristicsof respondents,some aspects of participationin the civil space
appearedto aid in understandingvariabilityin expressed trust:those respon-
dents who belonged to many secondary associations were more trusting,on
average, as were those who participatedmeaningfullyin co-operativegroups
and in groups with an interestin furtheringthe common good. The weak-ties
thesis, whereinweak ties arethoughtto especially fostertrust,was only weakly
supportedby this data set, while the organizationalformality thesis, wherein
informalityis thoughtto foster trust,was not supportedat all.

Discussion
The advantagesof this data set are the breadthand scope of items pertaining
to bothtrustandparticipationin secondaryassociations.The disadvantagesare
its cross-sectional nature, making claims about causality no more than
speculative in nature;the small sample size (N=534 overall; 370 respondents
answered questions pertaining to a group within which they currently
participate),making it difficult to discernweak butreal relationships;and the
low responserate(40%),raisingimportantissues pertainingto generalizability.
The latteris particularlyrelevantgiven the abstractnatureof the analysis that
purportsto addressgeneralizeable 'laws' or 'tendencies.'

The Nature of Trust

Accepting these limitations, the results presented here do contribute to


understandingthe natureof trustand where it is fostered.The notion of 'trust'
addressesthe very natureof social order.Low societal trust recalls Hobbes'
Leviathan, where the state is requiredto create and maintain order among
566 Canadian
Journalof Sociology

warring,self-interestedactors.Withhigh levels of trustin society, on the other


hand,social ordercomes fromthe bottomup, at least in part,createdanew in
interactionafter interaction.The discourseon trustseeks to explore the very
basis of social life: why do people enterinto interactionwithone anotherwhen
the risks are so great? People are rational, in part, and can calculate any
numberof adverseoutcomespotentiallyarisingfrom any numberof relation-
ships, yet they often overlooksuch risks,place themselvesin vulnerableposi-
tions, andrelatewithone another,interpersonallyandalso withrepresentatives
of abstract and expert systems. Assuming that risks and dangers manifest
themselvesdifferentiallyacrossandamongsocial scenarios,the cohesiveness
of trust as presentedhere suggests that it over-rideshomo economicus, the
rationalself-interestedhuman- trustis more thanthe rationalcalculationof
risk. Althoughonly a few referentsfor trustwere measured,no evidence was
found suggesting thatpeople direct theirwillingness to enterinto interaction
toward selected persons while necessarily mistrusting others. From the
perspective of the individual, no pockets of trust bracketed off from a
dangerousand risky social world were discernedherein.
Giddens(1990) hashypothesizedthattrustin modernsocieties haschanged.
It was formerly based mostly upon geographiccommunity, wherein those
people andentities with which one could interactarerootedin place, and also
within religious cosmologies and tradition which provided stability and
predictabilityto the universalorder.The strangeris enemy. God has a plan.
But in modernsociety, given the ephemeralnatureof interaction,not often
geographically rooted, and also the collapse of kinship ties and religious
meaning systems, people interactwith differentcommunitiesof people and
especially interactwith representativesof abstractsystems transcendingtime
and place. These survey respondents professed greatest trust for their
neighbours,andmoretrustfor people with whom they mightphysically inter-
act than for those spatiallyremoved.Trustin abstractentities was markedly
weakerthanwas trustin people. Accordingto these results,then,place is still
essentialwhen it comes to trust,as arepeopleandcommunities.Modernityhas
not yet entirelyreplacedthe local with the global or the abstract.

Trustand Participationin SecondaryAssociations

Many (if not most) social networksarestill local. This articlearguedthattrust


is generated,at least in part,by social networksin the civil space, andthatthe
relationshipbetween trustand such participationis not a simple one (noting
thatresearchutilizingthe notionof social capitalis often restrictedby the data
generatedby large-scalegovernmentsurveys).Bothtrustandparticipationcan
take on many forms, as can the natureof the networksof associationthem-
TrustandParticipation
in theCivil Space 567

selves, and certainkinds of trustwill be more stronglyrelatedwith participa-


tion in some groups than in others. Some of the infinite variety in patternsof
participationin the civil space was capturedby this small survey, i.e. parti-
cipation in secondaryassociations, and trustwas indeed found to be differen-
tially relatedto certaincharacteristicsof such participation.This suggests that
the 'social capital' that influences economic, political and population health
'outcomes' is not simply comprisedof communitiesor societies with trusting
people who belong to many clubs. Instead, when exploring and measuring
social capitalwe shouldbe especially sensitive to the natureof secondaryasso-
ciations and to the nature of trust itself. When it comes to the fostering of
variouskinds of trust,some networksof associationmay mattermore thando
others.
In particular,participation in a breadth of secondary associations was
significantlyrelatedto social andweakly relatedto political trust.Participation
in many networksof associationmaximizes face-to-face contacts with varied
interests, backgroundsand communication styles. People discover that so-
called 'foreigners' are indeed interestedin engaging in meaningful dialogue
and so develop an appreciationfor the trustworthinessof these people who are
'not really so different.' Even so, 'empowered' participationalso seems to
matterfor social trust.Superficialparticipationin networksof associationmay
not teach people lessons about trust: they must engage meaningfully in
dialogue with others for lessons regardingtrustworthinessto take. Adding
one's name to a membershiplist and paying yearly dues may not mean much
when it comes to learningthese kinds of lessons.
With respect to the structuralnatureof secondaryassociationsthemselves,
trust may be especially fostered in religiously oriented groups. Religious
groups remain one repositoryof certain values that facilitate trust:e.g. good
will toward others and a sense of responsibilityfor the common good. Reli-
gious communitiesemphasise the fundamentalimportanceof 'faith,' a corre-
late or component of trust. Thus a relationship between participation in
religious groups and trust may reflect a relationship between personal reli-
giosity and trust. Perhaps a belief in the benevolent nature of the universe
translatesinto a benevolent perspective of the intentionsof others. Perhapsa
belief that, in the end, everythingwill work out for the best translatesinto an
attitudeof forgiveness and trustfulnesstowardothers.Finally, trustappearsto
be especially fosteredin co-operatively-orientedsecondaryassociations.Trust
is a focus in social capital researchbecause it is thought to lubricateinterac-
tions among people, facilitatingcollective and collaborativeaction leading to
the achievement of common goals. The relationshipsbetween respondents'
perceptions of the collaborative context of their groups and both social and
political trustwere quite strong in this instance.
568 CanadianJournalof Sociology

References
Berkman,Lisa F., ThomasGlass, Ian Brissette,andTeresaE. Seeman
2000 "Fromsocial integrationto health:Durkheimin thenew millenium."Social Scienceand
Medicine51:843-857.
Brehm,John andWendy Rahn
1997 "Individual-levelevidenceforthe causesandconsequencesof socialcapital."American
Journalof Political Science 41(3):999-1023.
Coburn,David
2000 "Incomeinequality,socialcohesionandthehealthstatusof populations:Theroleof neo-
liberalism."Social Science and Medicine51:135-146.
Coleman,James
1988 "Social capital in the creation of human capital."AmericanJournal of Sociology
94:S95-S121.
Erikson,Erik
1950 Childhoodand Society.New York:Norton.
Fukuyama,Francis
1995 Trust: TheSocial Virtuesand the Creationof Prosperity.New York:Free Press.
Giddens,Anthony
1984 The Constitutionof Society.Berkeley:Universityof CaliforniaPress.
1987 Social Theoryand ModernSociology. Stanford:StanfordUniversityPress.
1990 The Consequencesof Modernity.Stanford:StanfordUniversityPress.
Govier, Trudy.
1997 Social Trust and Human Communities.Montrealand Kingston: McGill-Queen's
UniversityPress.
Granovetter,Mark
1973 "Thestrengthof weak ties."AmericanJournalof Sociology78(6):1360-1380.
Hawe, Penelopeand Alan Shiell
2000 "Social capital and health promotion:A review." Social Science and Medicine
51:871-885.
Helliwell, JohnF. and RobertD. Putnam
1995 "Economicgrowthandsocial capitalin Italy."EasternEconomicJournal21:295-307.
Jennings,M. Kent
1998 "Politicaltrustand the roots of devolution."In Braithwaite,Valerieand MargaretLevi
(Eds.) Trustand Governance.New York:Russell Sage Foundation.
Kasperson,RogerE., DominicGolding,and Seth Tuler
1992 "Social distrustas a factor in siting hazardousfacilities and communicatingrisks."
Journalof Social Issues48(4): 161-187.
Kawachi,Ichiro,BruceKennedy,KimberlyLochner,and DeborahProthrow-Stith
1997 "Socialcapital,income inequality,andmortality."AmericanJournalof Public Health
87(9):1491-1498.
Kawachi,Ichiroand Lisa Berkman
2000 "Socialcohesion,social capitalandhealth."InL. BerkmanandI. Kawachi(Eds.)Social
Epidemiology.New York:OxfordUniversityPress.
Knack,Stephenand Philip Keefer
1997 "Doessocialcapitalhaveaneconomicpayoff?A cross-countryinvestigation."Quarterly
Journalof Economics112(4):1251-1288.
Trust and Participationin the Civil Space 569

Luhmann,Niklas
1980 Trustand Power. New York:Wiley.
Lynch, John, PernilleDue, CariesMuntaner,and George Davey Smith
2000 "Social Capital - Is it a good investment strategy for public health?"Journal of
Epidemiology& CommunityHealth 54(6):404-408.
Lynch, John and George Kaplan
1997 "Understandinghow inequalityin the distributionof income affects health."Journal of
Health Psychology 2:297-314.
Misztal, BarbaraA.
1996 Trustin ModernSocieties. TheSearchfor the Bases of Social Order.Cambridge,MA:
Polity Press.
Putnam,RobertD., RobertLeonardi,and RaffaellaNanetti
1993 Making Democracy Work:Civic Traditionsin Modern Italy. New Jersey: Princeton
UniversityPress.
Putnam,RobertD.
1996 "Thestrangedisappearanceof civic America."TheAmericanProspect 24:34-48.
2000 BowlingAlone. Thecollapse and revivalofAmericancommunity.New York:Simonand
Schuster.
2001 "Social capital:Measurementand consequences."isuma: CanadianJournal of Policy
Research 2(1):41-51.
Rice, Tom W. and AlexanderF. Sumberg
1997 "Civic culture and government performance in the American states." Publius
27(1):99-114.
Shah, DhavanV.
1998 "Civic engagement, interpersonal trust, and television use: An individual-level
assessmentof social capital."Political Psychology 19:469-496.
Stolle, Dietlind
1998 "Bowling together,bowling alone: The development of generalized trustin voluntary
associations."Political Psychology 19:497-525.
Temple, Jonathan
1998 "Initialconditions,social capitaland growthin Africa."Journal of AfricanEconomies
7(3):309-347.
Tonkiss, Fran and AndrewPassey
1999 "Trust, confidence and voluntary organizations:Between values and institutions."
Sociology 33(2):257-274.
Tyler, Tom R.
1998 "Trustand democraticgovernance."In Braithwaite,Valerie and MargaretLevi (Eds.)
Trustand Governance.New York:Russell Sage Foundation.
Uslaner,Eric M.
1998 "Socialcapital,television, and the "meanworld": Trust,optimism,and civic participa-
tion."Political Psychology 19(3):441-467.
Veenstra,Gerryand JonathanLomas
1999 "Homeis wherethegoverningis: Social capitalandregionalhealthgovernance."Health
& Place 5:1-12.
Veenstra,Gerry
2000 "Social capital, SES and health: An individual-level analysis." Social Science and
Medicine 50(5):619-629.
570 CanadianJournalof Sociology

2001 "Socialcapitaland health."isuma: CanadianJournalof Policy Research2(1):72-81.


2002 "Social capital and health (plus wealth, income inequality and regional health
governance)."Social Science and Medicine54(6):849-868.
Wilkinson,RichardG.
1996 UnhealthySocieties: TheAfflictionsof lnequality.London:Routledge.
Wisdom,John.
1953 "Gods."In Philosophyand Psychoanalysis.Oxford:Basil Blackwell.
Woolcock, Michael
1998 "Socialcapitaland economic development:Towarda theoreticalsynthesisand policy
framework."Theoryand Society 27:151-208.
2001 "Theplace of social capitalin understandingsocial and economic outcomes."isuma:
CanadianJournalof Policy Research2(1):11-17.
Yamagishi,Toshio Cook KarenS. and WatabeMotoki
1998 "Uncertainty,trustandcommitmentformationintheUnitedStatesandJapan."American
Journal of Sociology 104(1):165-194.

Appendix
In the questionnaire 'yourregion' was definedto be 'thepart of the province in which you live,'
your community'was defined to be 'yourcity, town or ruralarea' and 'yourneighbours'were
definedto be 'thosepeople who live near you.' ThePearson's correlationbehindeach itemin an
index representsits correlationwith the remainderof the indexminusthat item.

Item: Personal trust


Most people can be trusted.<7 responsecategories,agreestrongly... disagreestrongly>

Item: Trustin neighbours


Most people in my neighbourhoodcan be trusted.<agreestrongly... disagreestrongly>

Index: Trustin people from respondents'communities


1. Mostpeople frommy communityare willing to help if you requireassistance.<agreestrongly
... disagreestrongly>(r=.683)
2. Most people in this communitycan be trusted.<agreestrongly... disagreestrongly>(r=.649)
3. My communityis a prettysafe place. <agreestrongly... disagreestrongly>(r=.589)
Themean of the inter-itemcorrelationswas r=.568 and Cronbach'salpha = .795.

Index: Trustin peoplefrom respondents'parts of Saskatchewan


1. Whenit comes down to it, you can alwaystrustthe people in my partof Saskatchewan.<agree
strongly... disagreestrongly>
2. Most of the people who live in my partof Saskatchewanare honourable.<agree strongly...
disagreestrongly>
The inter-itemcorrelationwas r=.695 and Cronbach'salpha = .817.

Index: Social trust


1. Most people can be trusted(r=.805). <agreestrongly... disagreestrongly>
2. Mostpeoplein my neighbourhoodcanbe trusted(r=.756).<agreestrongly... disagreestrongly>
3. Index:Trustin people from respondents'partsof Saskatchewan.(r=.750)
4. Mostpeople in my ethnicgroupcan be trusted.<agreestrongly... disagreestrongly>(r=.740)
Trust and Participationin the Civil Space 571

5. Most people in my religious/spiritualcommunitycan be trusted.<agree strongly ... disagree


strongly> (r=.715)
6. Index:Trustin people from respondents'communities.(r=.696)
The mean of the inter-itemcorrelationswas r=.620 and Cronbach'salpha = .906.

Index: Political trust


1. The provincialgovernmenthas the public's best interestsat heart.<agreestrongly ... disagree
strongly>(r=.712)
2. Although I may have some complaintsaboutsome decisions the DistrictHealthBoardmakes,
I trustit to make good decisions. <agree strongly... disagreestrongly>(r=.709)
3. How would you rate the performanceof the provincial governmentin solving problems in
Saskatchewan?<6 responsecategories,excellent ... very poor> (r=.645)
4. How would you ratethe performanceof yourDistrictHealthBoardin solving problemsin your
district?<excellent ... very poor> (r=.641)
5. Whenthe DistrictHealthBoarddecides whatpolicies to adopt,how muchattentiondo you think
they pay to whatthe generalpublicthink?<6 responsecategories,very much ... none> (r=.611)
6. The local governmentwill tell the public what they need to know aboutrelevantissues in my
community.<agree strongly ... disagreestrongly>(r=.581)
7. Although I may have complaintsaboutsome decisions the federalgovernmentmakes, I trust
it to make good decisions. <agree strongly... disagreestrongly>(r=.571)
8. How would you rate the performanceof your local governmentin solving problemsin your
community?<excellent ... very poor> (r=.550)
9. How wouldyou ratethe performanceof the federalgovernmentin solving problemsin Canada?
<excellent ... very poor> (r=.520)
10. In general, how much of the taxpayers' money, if any, is wasted by your local government?
<very much ... none> (r=.364)
The mean of the inter-itemcorrelationswas r=.402 and Cronbach'salpha = .871.

Item: Trust in experts and professionals


Experts and other professionals can help solve problemsin my community.<agree strongly ...
disagree strongly>

Item: Participationin secondaryassociations


Please list the groups,clubs or associationsthatyou currently participatein (if any):
For example, some types of groups are: church-related,school-service, sports, veterans,' card-
playing, recreational,youth, multicultural,hobby,political or civic,farm groups, etc., as well as
professional societies, literarysocieties, labourunions,services clubs,volunteerorganisationsand
seniors clubs.

Items: Characteristicsof secondary associations


For the three groups listed on the previouspage that you are most involved in right now, please
answerthe following:

Approximatelyhow many people are in the group?


Approximatelyhow many hours per monthdo you spend with this group?
Does the group have an executive? <yes, no>
Are you or have you been an executive memberin this group?<yes, no>
Are there formalrules for makingdecisions?(ex. a constitution,by-laws, voting) <yes, no>
Do you feel that you personallyinfluence decisions? <yes, no>
572 CanadianJournalof Sociology

Can anyonebecome a member?<yes, no>


Does the groupdo thingsfor people outsideits own membership?<yes, no>
Do the membersof the groupget togetherfor otherreasonsin othersettings?<yes, no>
Do groupmembersco-operateand workwell with one another?<yes, no>
Describethe people who belong to the group:
most are fromthe same church<yes, no>
most are fromthe same ethnicgroup<yes, no>
most sharea commoninterest<yes, no>
we have a wide varietyof ages represented<yes, no>
most live in the same community<yes, no>

Você também pode gostar