Advances in Applied Sociology
2012. Vol.2, No.4, 253-259
Published Online December 2012 in SciRes (http://www.SciRP.org/journal/aasoci) http://dx.doi.org/10.4236/aasoci.2012.24033
How Do Network Size, Voluntary Association, and Trust Affect
Civic Engagement? Evidence from the Asian Barometer Survey
Harris H. Kim
Department of Sociology, Ewha Womans University, Seoul, South Korea
Received August 2nd, 2012; revised September 5th, 2012; accepted September 17th, 2012
This study examines the role of social capital in shaping the individual likelihood of “civic engagement”
defined specifically as informal and formal political participation. Based on a subset of the Asian Ba-
rometer Survey of Democracy, Governance and Development (2006), a representative cross-national
dataset, it examines how and to what extent network size, voluntary association, generalized trust, and
particularized trust differentially influence the political behaviors of the survey participants in Korea. The
dependent variable is measured in terms of first, discussing political topics in an informal social context
and, second, getting together with others in order to raise a political issue or sign a petition. Four inde-
pendent variables are measured: 1) the size of egocentric network; 2) the membership in voluntary or-
ganizations and formal groups; 3) the degree of trust placed in generalized others (i.e., strangers); and 4)
the extent to which survey respondents place their trust in particularized others (those with whom one has
a personal relationship). Quantitative analyses show that, ceteris paribus, network size and voluntary as-
sociation have a strong causal impact on both outcome variables. Generalized trust is found to be a
non-significant factor, however, while particularized trust has a contingent effect. Along with the inter-
pretation of statistical results, their broad theoretical implications are also discussed.
Keywords: Civic Engagement; Network Size; Social Trust; Voluntary Organization; Asia
Introduction
Social Capital as Generalized Trust & Voluntary
Association
What are the social structural conditions under which people
make political decisions and take political actions? Like any
other type of social action, individual political behavior, both
formal and informal types of civic participation, does not take
place in a social vacuum. Whether it’s going to the voting booth,
participating in a protest, or simply engaging in a casual politi-
cal discussion, people’s behaviors are shaped by a number of
relevant situational factors, i.e., those in which individual deci-
sion-making and actions are socially embedded (cf. Granovetter,
1985). When it comes to this particular empirical question,
researchers have made much use of the concept social capital.
According to Putnam’s (1993) much-cited work, social capital-
broadly cast in terms of interpersonal networks, shared organ-
izational membership, mutual trust, and norms of cooperation
and reciprocity, serves to increase political participation and
can improve democratic governance and accountability. There
is a voluminous and growing literature that shows a causal link
between social capital and various political outcomes (see e.g.,
Braithwaite & Levi, 1998; Dekker & Uslaner, 2001; Jamal &
Noorudin, 2010; Mishler & Rose, 2005; Paxton, 2002; Roth-
stein & Uslaner, 2005; Uslaner, 2002; Newton & Zimerli, 2011;
Zmerli & Newton, 2008).
The vast majority of the extant scholarship focuses on the
role of generalized trust and participation in voluntary associa-
tion in facilitating democratic values, support and involvement
on the part of individuals (Marsh, 2005; Tavits, 2006; Zmerli &
Newton, 2008). Concerning the causal relationship between the
propensity to trust others and the likelihood of civic engage-
ment, Rothstein and Uslaner (2005: p. 41) write that “at the
individual level, people who believe that in general most other
people in their society can be trusted are also more inclined to
have a positive view of their democratic institutions, to partici-
pate more in politics, and to be more active in civic organiza-
tions.” Researchers have found a similar causal link connecting
voluntary organizational membership and a certain set of values
and orientations that are conducive to individual democratic
responsibility (Paxton, 2002; Stolle, 2001; Van Egmond et al.,
1998). As Putnam (1993) explains, “associations instill in their
members habits of cooperation, solidarity, and public-spirited-
ness” and that “participation in civic organizations inculcates
skills of cooperation as well as a sense of shared responsibility
for collective endeavors” (90). Using the World Values Survey
data across 43 countries, Inglehart (1997) provides empirical
support for this view: that the stability of democracy among
different nations is indeed highly correlated with the individual
membership in voluntary associations as well as the aggregate
levels of social trust. In a case study of Sweden, Teorell (2003)
offers further evidence by showing that “organizational in-
volvement provides bridging social capital by connecting the
individual to a wide range of people” (49).
Network Size as Social Capital: A Neglected Variable
The preceding discussion serves to highlight two major
strands of research in the extant literature: that is, how general-
ized trust and voluntary association can lead to greater political
interest and participation of individual citizens in democratic
societies. While fully recognizing the past research to be em-

H. H. KIM
pirically and theoretically important, the current study proposes
to examine two additional factors that have been relatively
overlooked, namely network size and particularized trust. That
the network variable, though causally relevant, has not been
investigated fully in the literature is surprising since social
capital is typically defined as trust, norms of reciprocity, and
social networks (Putnam, 1993; see also Rothstein & Stolle,
2008). Part of the reason for this selective scholarly attention
lies with the limited availability of data on interpersonal net-
works and political behavior.
This study has a twofold purpose. Given the relative scarcity
of research findings related to the causal role of social networks,
one of its main objectives is to fill this empirical gap in the
literature. Network size, though understudied, is indeed a sig-
nificant factor driving individual political interest and behavior.
According to the empirical investigation carried out by Lake
and Heckfeldt (1998), along with political interaction frequency
and network expertise, the size of one’s interpersonal network
has a powerful influence on the person’s political actions, while
controlling for a host of background factors. Theoretically,
larger networks are associated with multiple interpersonal
sources that can provide access to novel and timely information
(Burt, 1992). They can also serve as a communication bridge
(cf. Granovetter, 1973). In the case of political participation,
being situated in large networks opens up more opportunities to
be in touch with others who may be politically interested or
active (Huckfeldt et al., 1995).
Particularized vs. Generalized Trust
Along with network size, the concept of particularized trust
has also failed to attract sufficient scholarly attention. A con-
ceptual distinction is drawn between “generalized” and “par-
ticularized” trust in the literature. Simply put, the former is
about trust in unknown others or strangers. It is related to a
situation where “a community shares a set of moral values in
such a way as to create regular expectations of regular and
honest behavior” (Fukuyama, 1995: p. 153) among people who
do not know each other personally. This is very different from
particularized trust which is based on specific face-to-face in-
teractions between concrete individuals (see Bahry et al., 2005;
Hardin’s, 2002; Yamagishi & Yamagishi, 1994). The vast ma-
jority of the past studies attempting to link social capital and
political participation have relied on the concept of generalized,
not particularized, trust. A study by Uslaner and Conley (2003)
is one important exception, which focuses on the latter concept.
They categorize people into two groups: generalized trusters
and particularized trusters. According to their definition, the
former group tends to engage in broad social interactions and
helps to build general social capital that is conducive to the
proper workings of democracy. The latter group, on the other
hand, is prone to remain disconnected from the mainstream
society. The empirical analysis carried out by Uslaner and
Conley (2003) is based on a sample of Asian immigrants in the
U.S. Their findings reveal that particularized trusters, compared
with their counterparts, are indeed less likely to engage in
mainstream American politics.
As Newton and Zmerli (2011) show, however, the relation-
ship between generalized trust and particularized trust may be
more nuanced than as described above. The authors challenge
the conventional notion that particularized trust is “either
harmful or of little importance in modern democracies” (2011:
p. 169) and that only generalized trust is instrumental in creat-
ing social capital that promotes democratic processes. They
contend that depending on the situation the two forms of trust
may be mutually exclusive or supportive. In their analysis of
the World Values Survey (1995-2007), the authors also demon-
strate that both trust types are positively associated with peo-
ple’s attitude toward (level of confidence in) political institu-
tions across 22 countries. In other words, contrary to the impli-
cations of the findings in Uslaner and Conley (2003), they re-
port that those who score higher on the particularized trust scale
are more likely to hold a positive view of their political leaders
and institutions. Based on the above literature review, a fol-
lowing set of hypotheses can be formulated concerning the
concepts of generalized trust, particularized trust, voluntary
associational memberships, and network size, as each relates to
the probability of individual political participation.
Hypothesis 1: People who are more willing to trust general-
ized others are more likely to participate in political activities,
ceteris paribus.
Hypothesis 2: People who are more active in voluntary asso-
ciations are more likely to participate in political activities,
ceteris paribus.
Hypothesis 3: People who have larger social networks are
more likely to participate in political activities, ceteris paribus.
Hypothesis 4a: People who have greater particularized trust
are more likely to participate in political activities, ceteris
paribus.
Hypothesis 4b: People who have greater particularized trust
are less likely to participate in political activities, ceteris pari-
bus.
Hypothesis 4 has two opposing versions since, given the
contradictory views in the literature, it is debatable whether
particularized trust promotes or hinders the likelihood of infor-
mal and formal political engagement. The remainder of this
paper is devoted to evaluating the validity of these hypotheses
through empirical testing. The next section describes the data,
the variables, and the methods used for the quantitative analysis,
followed by the interpretations of the findings and the broad
implications of this study.
Data, Model & Variables
Data analyzed in this paper were collected by the Asian Ba-
rometer Project, which was co-directed by Professors Fu Hu
and Yun-han Chu and received major funding support from
Taiwan’s Ministry of Education, Academia Sinica and National
Taiwan University. This is the second wave of the Asian Ba-
rometer Survey (ABS) collected in 2006 covering major politi-
cal systems in Asia. This study focuses specifically on the Ko-
rean dataset. There have only been a limited number of nation-
ally representative studies on Korea, which previously pro-
duced conflicting results based on older data (see Kim, 2005;
Lee, 2008). The dataset available from the ABS provides more
recent information that can be used to shed novel light on the
role of social capital in individual political participation in Ko-
rea. The survey was conducted during the month of September
(2006) by Gallup Korea. The survey population consisted of all
citizens aged 19 and older residing in the territory of South
Korea at the time it was conducted. The survey, using a
multi-stage probability sampling method, was done through
face-to-face personal interviewing, which lasted on average 60
minutes. A total of 1212 interviews were completed, with the
254

H. H. KIM
response rate of 44.9%.
Dependent Variable
The dependent variable for this study is “political participa-
tion”. A number of questions were asked in the survey to tap
into the respondents’ levels of political interest and behavior. In
this study, the focus will be on one informal and one formal
type of political participation. The exact wording for these two
questions is as follows:
Q.1 When you get together with your family members or
friends, how often do you discuss political matters? (Frequently,
Occasionally, Never).
Q.2 Here is a list of actions that people sometimes take as
citizens. Tell me whether you, personally, have never, once, or
more than once done any of these things during the past three
years. For example, “Got together with others to raise an issue
or sign a petition”.
The answers for the first question was coded on a 3-point
scale (“Frequently” = 3; “Occasionally” = 2; “Never” = 1) to
create the first dependent variable (DISC_POL). From the sec-
ond question, a variable called SIGN_PETN is constructed by
assigning the value of “0” if the answer is never and “1” other-
wise. In the sample, 31% stated that they never talk about poli-
tics with their friends and families; about 59% said that they do
so occasionally, while 10% claimed to do so frequently. When
it comes to getting together with others to raise an issue or sign
a petition, about 12% gave an affirmative answer.
Independent Var iables
A number of independent variables are created to test the va-
lidity of the hypotheses stated above. The standard question in
the literature used to measure generalized trust, as found in the
American General Social Survey (GSS) or the World Values
Survey (WVS), is typically phrased as: General speaking,
would you say that “most people can be trusted” or that “you
must be very careful in dealing with people”? The same word-
ing was used in the ABS. In the sample, about 33% of the Ko-
rean respondents agreed with the question, a figure that is rela-
tively lower in comparison with the earlier studies of Europe
and North America (see Nannestad, 2008). A variable named
GEN_TRUST, coded “1” if answered “yes” to the above state-
ment, is used to test the validity of Hypothesis 1. Hypothesis 2
deals with the level of participation in voluntary groups or or-
ganizations. Under the subheading of “Social Capital,” the ABS
asks subjects to name up to “3 most important organizations or
formal groups you belong to”. Based on the individual re-
sponses, a variable called VOL_ASSOC is constructed.
As discussed above, the variables for generalized trust and
voluntary associations are standard measures frequently used in
the literature. The two additional variables examined in this
study are network size and particularized trust. One of the new
questions asked in the ABS is the following, which was not
available in the first wave of the survey conducted in 2003: “On
average, about how many people do you have contact with in a
typical week day? Please include only people you know.” This
question is designed to directly tap into the size of the interper-
sonal network in which the individual subjects are embedded. It
resembles the question in the GSS, which asks for the number
of confidants with whom one discusses important matters. The
range of answers varies from 1 to 5, with 2.7 being the average
network size for the Korean sample. This mean value is higher
than that for the Americans, which decreased over time, from
2.94 in 1985 to 2.08 in 2004 (McPherson, Smith-Lovin, &
Brashears, 2006). For this study, a variable called NET_SIZE is
created to evaluate Hypothesis 3.
Lastly, in order to test the validity of Hypothesis 4a and 4b,
particularized trust is measured. In addition to generalized trust,
the ABS inquires about the extent to which the respondents
trust three groups of people whom they know personally,
namely “your relatives”, “your neighbors”, and “others you
interact with”. Based on the answers given, three separate vari-
ables are put together using a 4-point scale (“A great deal of
trust” = 1; “Quite a lot of trust” = 3; “Not very much trust” = 2;
“None at all” = 1), namely TRUST_REL, TRUST_NEIGHB,
TRUST_OTHERS. The mean value for trusting one’s relatives
is 3.14, that for trusting one’s neighbors is 2.8, and for trusting
others it is 2.4. According to an earlier study done by Gibson
(2001), 59.6% of the Russians feel that people they know per-
sonally can be “trustful”. In the Korean sample, 59.8% said that
they put “quite a lot of trust” in their relatives and 64.4% ex-
pressed the same feeling concerning their neighbors, while
40.3% gave the same answer for the third trust category (“oth-
ers you interact with”).
Control Variables
A number of relevant control variables are included in the
analysis. According to Newton and Zimerli (2011: p. 183), “a
reading of the already voluminous literature on trust suggests a
fairly short and consistent list of individual variables associated
with it”. In this study, such factors are controlled for—includ-
ing income, education, gender, marital status, age, and the level
of subjective interest in politics. The age variable (AGE) is a
continuous variable with the mean value of 42.7; the income
variable (INCOME) is measured as an ordinal one whose scale
is divided into quintiles; the level of education (EDUC) is
gauged on a 10-point scale (“No formal education” = 1;
“Post-graduate degree” = 10); the gender variable (FEMALE)
is coded 1 if “female”; the marital status variable (MARRIED)
is also dichotomously coded (“married” = 1); and the variable
that gauges the degree of interest in politics (INTEREST) is
coded using a 4-point scale (“Not at all interested” = 1; “Very
interested” = 4).
The descriptive information concerning all the variables is
summarized in Table 1. Table 2 contains the bivariate correla-
tions among the variables. The quantitative results are pre-
sented in two separate tables (Tables 3 and 4) since there are
two different outcome variables under investigation: discussing
political matters (DISC_POL) and signing a petition (SIGN_
PETN). Since the dependent variables are all categorically dis-
tributed, nominal logistic regression models are estimated. The
following section describes the statistical findings and their
interpretations.
Results and Discussion
For each of the two regression tables, two models are pre-
sented. Model 1 contains only the independent variables used to
test the hypotheses. Model 2 introduces the control variables,
offering a more conservative test of their validity. Table 3 re-
ports the findings from regressing DISC_POL on the selected
independent and control variables. According to Model 1, coef

H. H. KIM
256
Table 1.
Descriptive statistics.
Min Max Mean Std. Dev.
INCOME 1 5 3.06 1.22
EDUC 1 10 6.79 1.98
MARRIED 0 1 .72 .44
FEMALE 0 1 .50 .50
AGE 19 80 42.65 14.01
INTEREST 1 4 2.38 .80
NET_SIZE 1 5 2.71 1.17
VOL_ASSOC 0 3 .39 .79
GEN_TRUST 0 1 .32 .46
TRUST_REL 1 4 3.14 .62
TRUST_NEIGHB 1 4 2.80 .60
TRUST_OTHERS 1 4 2.44 .66
DISC_POL 1 3 1.77 .59
SIGN_PETN 0 1 .11 .32
ficients for two of the six variables are found to be statistically
significant (p < .001), namely NET_SIZE and VOL_ASSOC.
In other words, individuals with larger social networks are more
likely to participate in informal political discussions with their
friends and relatives. Also, those with more active organiza-
tional life or formal group affiliation are more prone to engage
in such informal political behavior, a finding that complements
the earlier research by Lee (2008) on Korea. Contrary to the
hypotheses stated earlier, however, neither generalized nor
particularized forms of trust are found to exert any significant
causal influence.
Model 2 tests the robustness of the coefficients in Model 1
by including the control variables. Among them, three have a
significant effect on the outcome variable. Specifically, indi-
viduals with higher educational attainment (p < .001), those
who are older (p < .05), and people who profess to have greater
interest in politics (p < .001) are more likely to converse with
their friends and relatives about political issues. More impor-
tantly, the coefficients for network size and associational
membership remain their significance at .05 and .01 level, re-
spectively, thus offering strong empirical evidence that con-
firms Hypothesis 2 (“People who are more active in voluntary
associations are more likely to participate in political activities,
ceteris paribus”) and Hypothesis 3 (“People who have larger
social networks are more likely to participate in political ac-
tivities, ceteris pa ribus”).
Table 2.
Correlation matrix.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
1. INCOME 1
2. EDUC .400** 1
3. MARRIED .020 −.114** 1
4. FEMALE −.016 −.229** .066* 1
5. AGE −.300** −.611**. .438** .050 1
6. INTEREST .061* .123** .074* −.215** .065*1
7. GEN_TRUST .021 .052 .045 −.048 .024 .095** 1
8. TRUST_REL −.007 .090** .010 −.061* .003 .093** .168** 1
9. TRUST_NEIGHB −.011 −.018 .092** −.037 .122** .100** .289** .499** 1
10. TRUST_OTHERS .023 .070* .014 −.034 −.033 .105** .335** .320** .501** 1
11. DISC_POL .121** .157** .115** −.177** .059*.435** −.005.040 .025 .005 1
12. SIGN_PETN .061* .138** .019 –.049 −.087** .098** .031 .048 −.015 −.021 .093** 1
13. CONTACT .008 .047 .064* −.107** .008 .082**.052 .022 .036 −.041 .108** .146** 1
14. NET_SIZE .143** .103** −.029 −.143** −.060*.105**.059*.049 .032 .060*.104** .087** .121** 1
15. VOL_ASSOC .092** .108** .137** −.190** .083** .170** .086** .050 .095** .045 .198** .128** .183** .121**1
Note: *<.05; **<.01; ***<.001.

H. H. KIM
Table 3.
Logistic coefficients from regressing DISC_POL on selected vari-
ables.
Model 1 Model 2
Estimate S.E. Estimate S.E.
INCOME .106# .062
EDUC .172*** .048
MARRIED .290 .178
FEMALE −.259# .146
AGE .017* .007
INTEREST 1.211*** .101
NET_SIZE .189*** .054 .122* .060
VOL_ASSOC .521*** .082 .287** .089
GEN_TRUST −.163 .140 −.253 .155
TRUST_REL .095 .114 −.032 .128
TRUST_NEIGHB .015 .135 .051 .150
TRUST_OTHERS −.074 .110 −.313# .122
INTERCEPT1 4.090*** .707
INTERCEPT2 8.015*** .757
−2LL 861.853 1463.182
N 1069 983
Note: # < 0.1; * < .05; ** <.01; *** <.001 (two-tailed tests).
Table 4.
Logistic coefficients from regressing SIGN_PETN on selected vari-
ables.
Model 1 Model 2
Estimate S.E. Estimate S.E.
INCOME –.012 .093
EDUC .161* .079
MARRIED .341 .281
FEMALE .207 .221
AGE –.015 .011
INTEREST .352** .136
NET_SIZE .222** .083 .229* .088
VOL_ASSOC .398*** .100 .309** .109
GEN_TRUST .290 .217 .273 .226
TRUST_REL .442* .174 .389* .187
TRUST_NEIGHB –.196 .207 –.125 .221
TRUST_OTHERS –.323# .174 –.450* .181
INTERCEPT1 3.024*** 4.396***
–2LL 417.631 668.580
N 1069 981
Note: # < 0.1; * < .05; ** < .01; *** < .001 (two-tailed tests).
Table 4 contains regression results from examining the sec-
ond outcome variable (SIGN_PETN), which has to do with a
more formal way of political engagement, i.e., getting together
with other citizens in order to raise an issue or sign a petition.
Looking at Model 2, the full model with the control variables
included, a new set of findings that is different from the one in
Table 1 emerges. As was the case with Table 1, people sur-
rounded by a larger number of interpersonal contacts (p < .05)
and those with greater organizational and group ties (p < .01)
exhibit a higher propensity toward political behavior, while
controlling for other covariates. In addition, two new variables
are found to be significant (at p < .05), namely TRUST_REL
and TRUST_OTHERS. Interestingly, however, the causal
flows are in opposite directions. As the negative sign for
TRUST_OTHERS suggests, individuals who shows greater
trust in those whom they know personally are less likely to
engage in formal political behavior, in accordance with Hy-
pothesis 4b. This finding thus lends support to the view pro-
posed by Uslaner and Conley (2003), as mentioned above. To
the contrary, the positive coefficient sign for TRUST_REL
offers a completely different picture, i.e., that people who score
higher on particularized trust are more prone to get involved
politically, which confirms Hypothesis 4a. This regression re-
sult challenges Uslaner and Conley (2003) and instead supports
Newton and Zmerli (2011), thereby adding further to the con-
troversy surrounding the relationship between particularized
trust and individual political behavior.
In sum, the quantitative analyses provide consistent and rela-
tively strong empirical support for Hypothesis 2 and Hypothe-
sis 3, as illustrated by the fact that the coefficients for network
size and organizational or group affiliation are robust across all
the models. Contrary to Hypothesis 1, however, generalized
trust is not related to political participation. In fact, the coeffi-
cients for this variable (GEN_TRUST) are insignificant in all of
the estimated models, a finding that deviates from much of the
previous research that confirms the function of generalized trust
in facilitating individual democratic values, responsibilities and
activities. The two contradictory hypotheses (Hypothesis 4a &
Hypothesis 4b) concerning the causal link between particular-
ized trust and individual political behavior receive partial sup-
port. As shown by the mixed statistical results, particularized
trust is not causally related in any way to informal political
participation, as measured by the frequency of holding political
discussions in a social setting with friends and relatives. None
of the coefficients for the three measures (TRUST_REL,
TRUST_NEIGHB, TRUST_OTHERS) reach the conventional
significance level. But when it comes to being mobilized to
sign a petition or to raise a political issue, trusting in relatives
and trusting in others do have significant, albeit opposite, ef-
fects, though trusting in neighbors is found to be causally in-
significant.
Conclusion
The purpose of this study is to examine the role of social
capital in predicting the level of individual civic engagement, at
both informal and formal levels. It tests four distinct arguments
related to the causal influences of generalized trust, particular-
ized trust, social network, and voluntary associational mem-
bership. Data analysis produces results that are expected as well
as surprising. First, when it comes to the role of voluntary asso-
ciation, measured in terms of membership in an organization or

H. H. KIM
a formal group, this study finds evidence that bolsters one of
the oldest views in the social capital theory—that healthy asso-
ciational life fosters citizen interests and activities in domestic
politics—an argument that dates back to Tocqueville’s (1969
[1865]) original observation about the workings of democracy
in America. Second, generalized trust seems to have no impact
on the individual probability of getting politically involved.
This finding is actually not very surprising in light of what
some scholars have pointed out in the past: namely that “there
are patchy and weak associations between social and political
trust” (Newton, 2001: p. 202) and that “trust has small if any
independent effect on support for the current regime” (Mishler
& Rose, 2005: p. 14). In fact, despite the voluminous literature,
there is still an ongoing controversy concerning the exact link-
age between social/generalized trust and political/institutional
trust, as well as various dimensions of democratic political
engagement (see, e.g., Delhey & Newton, 2003; Mishler &
Rose, 2005; Newton & Norris, 2000; Rothstein, 2002). The
main issue stems from disagreements concerning the definition
and the measurement of this frequently used, yet thorny, con-
cept. Broadly speaking, there are two opposing conceptions of
generalized trust found in the literature: one that is character-
ized by strategic rationality, on the one hand, and one that is
norm-driven, on the other (Nannestad, 2008). Many studies do
not make an explicit differentiation between them but conflate
the two approaches, which calls for future research that better
theorizes about and gauges generalized trust in understanding
its multifaceted causal role.
Third, network size as a relatively understudied concept is
found to be of major significance. In his investigation of Russia,
Gibson (2001) writes that “weak social networks” are critical
for the building of civil society in transitional societies such as
Russia. By “weak,” he means networks that are open and thus
cut across multiple social groups (cf. Granovetter, 1973). This
particular characteristic of network is also related to size (see
Burt, 1992). To the extent that a network is large, there is a
greater possibility of cross-cutting that would enable actors to
be connected to groups that transcend their narrow circles of
contacts based on kinship or other characteristics of similarity.
The argument is that networks that are large, disconnected, or
weak create bridging, rather than bonding, social capital (Put-
nam, 2000: p. 22) that encourages individual political aware-
ness, interest and engagement, which is ultimately beneficial
for the entire society. One shortcoming of this research is that it
relies on network size as a proxy variable and, due to data un-
availability, does not deal with the direct measures of network
openness or density. More nuanced network information is in
order to probe into the complex relationship between interper-
sonal networks and individual political engagement.
Lastly, the quantitative results in this study further add to the
debate concerning the value of particularized trust in promoting
political behavior. Does particularized trust deter people from
engaging in mainstream politics, as Uslaner and Conley (2003)
contend? Or does it serve as a foundation on which generalized
trust and political confidence in government institutions could
be built, as Newton and Zmerli (2011) insist? Perhaps the right
answer is that the role of particularized trust is contingent, that
is, “different types of social networks ··· lead to dissimilar
types of civic engagement” (Uslaner & Conely, 2003: p. 355).
It is the task of researchers to figure out the conditions under
which social networks of one kind result in one form of politi-
cal behavior, while another kind leads to a different form. So-
cial capital theory has provided a great deal of analytical mile-
age in explaining, for example, why people support different
types of political regimes, what democratic values they espouse,
when they are likely to participate in collective action, or how
they decide to trust government institutions. Regretfully, as is
the case with other popular theoretical innovations, the concept
of social capital has run into many conceptual, definitional, and
methodological problems. Greater attention is thus required in
correctly framing future inquiries about the political conse-
quences of social capital so as to maximize its heuristic value
and minimize the pitfalls associated with it.
REFERENCES
Bahry, D., Kosolapov, M., Kozyreva, P., & Wilson, R. (2005). Ethnic-
ity and trust. American Political S cience Review, 99, 521-532.
doi:10.1017/S0003055405051853
Braithwaite, V., & Levi, M. (2002). Trust and governance. New York:
Russell Sage Foundation.
Burt, R. (1992). Structural holes: The social structure of competition.
Cambridge: Harvard University Press.
Dekker, P., & Uslaner, E. (2001). Social capital and participation in
everyday life. London: Routledge.
Delhey, J., & Newton, K. (2003). Who trusts: The origins of social trust
in seven societies. European Societies, 5, 93-137.
doi:10.1080/1461669032000072256
Fukuyama, F. (1995). Trust: The social virtues and creation of pros-
perity. New York: Basic Books.
Gibson, J. (2001). Social networks, civil society, and prospects for
consolidating Russia’s democratic transition. American Journal of
Political Science, 45, 51-68. doi:10.2307/2669359
Granovetter, M. (1973). The strength of weak ties. American Journal of
Sociology, 78, 1360-1380. doi:10.1086/225469
Hardin, R. (2002). Trust and trustworthiness. New York: Russell Sage.
Ikeda, K., & Richey, S. (2005). Japanese network capital: The impact
of social network on Japanese political participation. Political Be-
havior, 27, 239-260. doi:10.1007/s11109-005-5512-0
Inglehart, R. (1997). Modernization and postmodernization: Cultural,
economic, and political changes in 43 societies. Princeton, NJ:
Princeton University Press.
Jamal, A., & Nooruddin, I. (2010). The democratic utility of trust: A
cross-national analysis. Journal of Politics, 71, 45-59.
doi:10.1017/S0022381609990466
Kim, J. Y. (2005). Bowling together isn’t a cure-all: The relationship
between social capital and political trust in South Korea. Interna-
tional Political Science Review, 26, 193-213.
doi:10.1177/0192512105050381
Lake, R., & Huckfeldt, R. (1998). Social networks, social capital, and
political participation. Political Psychology, 19, 567-584.
doi:10.1111/0162-895X.00118
Lee, J. (2008). Path toward democracy in South Korea: Social capital &
democracy embedded in the citizens. Asian Survey, 48, 580-602.
doi:10.1525/as.2008.48.4.580
Marsh, J. (2005). Social capital and democracy in a new democracy.
The Sociological Quarterly, 46, 593-615.
doi:10.1111/j.1533-8525.2005.00028.x
Mishler,W., & Rose, R. (2005). What are the political consequences of
trust? A test of cultural and institutional theories in Russia. Com-
parative Political Studies, 38, 1050-1078.
doi:10.1177/0010414005278419
Nannestead, P. (2008). What have we learned about generalized trust, if
anything? American Review of Political Scienc e , 11, 413-436.
doi:10.1146/annurev.polisci.11.060606.135412
Newton, K. (2001). Trust, social capital, civil society, and democracy.
International Political Science Review, 22, 201-214.
doi:10.1177/0192512101222004
Newton, K., & Norris, P. (2000). Confidence in public institutions:
Faith, culture, or performance? In S. Pharr, & R. Putnam (Eds.), Dis-
258

H. H. KIM
affected democracies: What’s troubling the trilateral countries?
Princeton, NJ: Princeton University Press.
Newton, K., & Zimerli, S. (2011). Three forms of trust and their asso-
ciation. European Political Sc i e n c e Review, 3 , 169-200.
doi:10.1017/S1755773910000330
Paxton, P. (2002). Social capital and democracy: An interdependent
relationship. American Sociological Review, 67, 254-277.
doi:10.2307/3088895
Putnam, R. (1993). Making democracy work: Civic traditions in mod-
ern Italy. Princeton, NJ: Princeton University Press.
Putnam, R. (2000). Bowling alone: The collapse and revival of Ameri-
can community. New York: Simon and Schuster.
Rothstein, B. (2002). Social capital in the social democratic state. In R.
Putnam (Ed.), Democracies in flux. The evolution of social capital in
contemporary society. Oxford: Oxford University Press.
Rothstein, B., & Uslaner, E. (2005). All for all: Equality, corruption,
and social trust. World Politics, 58, 41-72.
doi:10.1353/wp.2006.0022
Rothstein, B., & Stolle, D. (2008). The state and social capital. Com-
parative Politics, 40, 441-459.
doi:10.5129/001041508X12911362383354
Stolle, D. (2001). Clubs and congregations: The benefits of joining an
association. In K. Cook (Ed.), Trust in society. New York: Russell
Sage Foundation.
Tavits, M. (2006). Making democracy work more? Exploring the link-
age between social capital and government performance. Political
Science Quarterly, 59, 211-225.
De Tocqueville, A. (1969 [1865]). Democracy in America. New York:
Harper Perennial.