Trust Matters: The Impact of Ingroup
and Outgroup Trust on Nativism
and Civicness∗
Markus M. L. Crepaz, The University of Georgia
Jonathan T. Polk, University of Gothenburg
Ryan S. Bakker, The University of Georgia
Shane P. Singh, The University of Georgia
Objectives. The objectives of this study are threefold: first, we separate trust into
a two-dimensional concept: ingroup trust and outgroup trust. Second, we apply
both types of trust to two dependent variables: nativism and civicness, hypothesizing that respondents with ingroup trust should display higher degrees of nativism
and lower degrees of civicness while the opposite should apply to respondents with
outgroup trust. Third, we control for the traditional trust question in order to determine whether there is any value added by separating trust into two dimensions.
Methods. After applying confirmatory factor analysis to a six-item measure in the
fifth (2005/2006) wave of the World Values Survey, we identify two kinds of trust—
“ingroup” and “outgroup.” We then use various regressions (linear, ordered logistic,
and binary logistic) to estimate their effects on different measures of nativism and
civicness. Results. Our results indicate that despite the existence of a moderately
strong positive correlation between the two trust measures, once applied to four
models of nativism and three models of civicness, these have statistically significant
and different relationships, even when controlling for traditional generalized trust.
Conclusions. Our conclusions suggest that outgroup trust reduces nativism while
ingroup trust tends to increase it, and, counter to expectations, we find that while
ingroup trust varies positively and significantly with civicness measures, outgroup
trust does not.
As modern societies become more diverse as a result of immigration, what
role do different types of trust play in affecting the relationship between natives
and newcomers? Similarly, what is the role of types of trust in determining
an individual’s willingness to contribute to the common good? Hardly any
policy area demarcates so starkly the difference between “us” and “them” as
∗
Direct correspondence to Markus M. L. Crepaz, The University of Georgia, Candler
Hall 308, Athens, GA 30602-1492 "[email protected]#. Shane P. Singh will host the data and
replication code at "www.shanepsingh.com#. Jonathan T. Polk and Markus M. L. Crepaz would
like to thank Gary Freeman who was discussant of an earlier version of this article presented
at a panel at the 2009 APSA Annual Meeting in Toronto, Canada, as well as the anonymous
reviewers of this article
SOCIAL SCIENCE QUARTERLY
$
C 2014 by the Southwestern Social Science Association
DOI: 10.1111/ssqu.12082
2
Social Science Quarterly
the relationship between natives and immigrants. Even among citizens, there
are differences in the degree to which some people are willing to contribute
to the welfare of others; to show solidarity with those less fortunate; to care
about the lives of neighbors, of people in different stations in life, of different
religions, gender, age, or strangers in the countryside or the far-off capital of
their nation.
We argue that orbits of trust—for example, the degree to which individuals
understand other people to be either “like them,” based on spatial, racial,
ethnic, religious, linguistic, class, and other proximities, or whether they perceive others to be different along these categories—can help us understand
relationships between natives and newcomers. The wider this circle of cooperative norms, the easier it is to resolve potential conflicts, compromise on
political and social issues, and provide public goods. We believe that this width
of the moral community matters in understanding interactions between different people, particularly as modern societies become more diverse. Soroka,
Banting, and Johnston (2002:2) claim that “[b]onds of community may be
more difficult to sustain as the population becomes more diverse” and that
“[t]rust is aided by identification with fellow citizens.” We suspect that forms
of trust, taken as predictors rather than dependent variables, should matter in
the readiness of people to embrace others who are different from themselves,
and help us understand the severity of the potential problem highlighted by
Soroka, Banting, and Johnston. For this reason, we focus our analysis on contemporary industrialized societies, which have become increasingly diverse in
recent years, potentially undermining the political attitudes necessary for these
societies to function efficiently.
This article answers calls for empirical research that explores the consequences of distinct forms of trust (e.g., Freitag and Traunmüller, 2009:798–
99).1 Across 11 Organization for Economic Co-operation and Development
(OECD) countries, our analysis reveals that the levels of “ingroup” and “outgroup” trust significantly affect people’s attitudes on a group of questions that
measure what we refer to as nativism and civicness.2 Yet we also find that
these two closely related trust concepts have markedly different impacts on
these attitudes, thereby uncovering something of a paradox for contemporary
1
Freitag and Traunmüller (2009), Gibson (2001), Herreros (2004), Uslaner (2002), Welch,
Sikkink, and Loveland (2007), Newton and Zmerli (2011), Zmerli and Newton (2008), and
Zmerli and Hooghe (2008) represent some of the empirical contributions to research on
formation of different types of trust. As Freitag and Traunmüller point out, “[t]he empirical
component has been neglected, due mainly to lack of adequate data addressing various aspects
of different forms of trust” (2009:783). Further, most empirical work in this field focuses on
single-country case studies: Freitag and Traunmüller tested whether there are “spheres of trust”
using data from a German panel, Gibson’s study focused on Russia, while both Uslaner and
Welch, Sikkink, and Loveland’s research was based on U.S. data.
2
Throughout this article, we use the terms ingroup and outgroup trust interchangeably with
particularized and generalized trust, respectively. We prefer the use of ingroup and outgroup
trust rather than another terminology both because the terms are simple and straightforward,
and because our conception of trust closely resembles that of Delhey, Newton, and Welzel
(2011), who use these terms of trust.
Trust Matters
3
diverse societies. Outgroup trust reduces nativist resentment while ingroup trust tends to increase it. However, ingroup trust varies positively
and significantly with our civicness measures, but outgroup trust does not.
The divergent effects of these two types of trust on two important sets of
political attitudes highlight the value of differentiating between types of trust,
and provide important insights for current debates about the appropriateness
of the traditional trust question (TTQ) as a measure of what is at root a
multidimensional concept (Delhey, Newton, and Welzel, 2011; Reeskens and
Hooghe, 2008; Sturgis and Smith, 2010). Our findings support the assertion
that “one cannot recommend measuring trust with just a single item as is often
done in comparative research” (Reeskens and Hooghe, 2008:530).
The next section reviews the debate about the conceptual meaning, dimensionality, and measurement of trust. Relying on a six-item question in the fifth
wave of the World Values Survey (WVS), we then construct a two-dimensional
measure that we use to examine the impact of forms of trust on nativism and
civicness while controlling for the TTQ. The final section concludes with
implications of the findings and suggestions for further research.
The Thrust of Trust: Toward a Multidimensional Concept of Trust
A society in which generalized trust—trusting individuals whom we do
not know personally—is abundant is said to be able to resolve a variety of
coordination problems. The presence of generalized trust within a society reduces transaction costs associated with principal-agent problems, encourages
more cooperative behavior in prisoner’s dilemma situations, and helps avoid
issues of asymmetrical information. Higher levels of generalized or outgroup
trust are also associated with increased confidence in political institutions and
satisfaction with democracy (Zmerli and Newton, 2008; Zmerli and Hooghe,
2011), as well as less corrupt and more trustworthy and generous institutions (Uslaner, 2002). Research also suggests that societies with high amounts
of outgroup trust experience higher growth and investment rates (Knack
and Keefer, 1997; Zak and Knack, 2001), higher tax compliance (Scholz and
Pinney, 1995; Scholz, 1998; Scholz and Lubell, 1998), higher income and
education rates (Alesina and Ferrara, 2002; Slemrod and Katuscak, 2005),
and increases in the efficiency of organizations and reduction of moral hazard
(Miller, 2001), to name a few.
There is growing agreement that trust comes in two forms. Various terminologies have been proposed such as “strong” and “weak” ties (Granovetter,
1973), “bonding” and “bridging” trust (Putnam, 2002), “particularized” and
“generalized” trust (Uslaner, 2002), all basically claiming that there are different radii of trust that ripple out in concentric circles from the individual:
particularized or thick trust is extended to people whom we know, to family,
friends, to the locality, to kith and kin. However, these ripples begin to flatten
out the larger the radius from the individual becomes. Such trusters have a
4
Social Science Quarterly
very strong sense of “us” versus “them.” They are ready and willing to cooperate with others, as long as the others are like themselves (Goodhart, 2004).
The biggest challenge, however, is properly measuring these different types of
trust.
Most survey research on trust relies on the TTQ, which has become a
staple question in the WVS and the European Social Survey (ESS), among
others.3 The TTQ is worded as follows: “Generally speaking, would you say
that most people can be trusted or that you can’t be too careful in dealing with
people?” According to Delhey, Newton, and Welzel (2011:787), this question
was meant to measure “trust in a wide and unfamiliar circle of others, rather
than in a small circle of close and familiar others,” and Uslaner (2002) also
reports that people indeed think of trust in these general terms. Reeskens and
Hooghe, however, find that “almost every word [in the TTQ] is problematic”
(2008:516), and Sturgis and Smith (2010) conclude that when respondents in
their study of Great Britain (2007/2008) answered “most people” in the TTQ,
they had thought about people who were known to them personally not in
terms of “unknown others,” potentially seriously impairing the interpretation
of empirical results. This leads some to the conclusion that “relying on a single
item [the TTQ] measurement amounts to skating on thin ice” (Reeskens and
Hooghe, 2008:530).
The fifth wave of the WVS includes a battery of six questions that allow
researchers to more directly examine different types of trust. The six questions
(V125–V130) are as follows: “I’d like to ask you how much you trust people
from various groups. Could you tell me for each whether you trust people from
this group completely, somewhat, not very much, or not at all?” The groups are:
“your family,” “your neighborhood,” “people you know personally,” “people
you meet for the first time,” “people of another religion,” and “people of
another nationality.” The answer options were as follows: “trust completely,”
“trust somewhat,” “do not trust very much,” and “do not trust at all.”
In a detailed analysis of the TTQ, using these new questions that measure the radius of ingroup and outgroup trust, Delhey, Newton, and Welzel
find that “the question seems to work well in Western and affluent nations,
but less well in other parts of the world . . . ” (Delhey, Newton, and Welzel,
2011:800). More recently, Wollebaek, Lundåsen, and Trägårdh (2012) identified three types of trust within 33 Swedish municipalities: particularized,
generalized, and what they termed “community” trust. We take this variation
in the applicability of the TTQ and growing interest in multidimensional
conceptions of trust as the point of departure for our analysis.
In the next section, we further explore the two-dimensional concept of
trust, which we follow Delhey, Newton, and Welzel (2011) in referring to
as “ingroup” and “outgroup” trust. However, in addition to examining the
3
Unfortunately, the answer categories are not the same across the WVS, GSS, and ESS. While
the WVS forces the respondents into a dichotomy (can be trusted/cannot be too careful), the
GSS adds a third option,"depends,” and the ESS has an 11-point scale (0–10).
Trust Matters
5
impact of the two-dimensional trust measures on nativism and civicness,
we also include the TTQ as a control, allowing us to observe whether it
has a unique impact above and beyond the impacts of “ingroup” and “outgroup” trust. We restrict our sample to the modern industrialized countries for
which the six trust items were available (Australia, Finland, France, Germany,
Italy, the Netherlands, Norway, Sweden, Switzerland, the United Kingdom,
and the United States). Given Delhey, Newton, and Welzel’s (2011) finding
that the TTQ works well in Western affluent nations, it is of particular interest
to see whether the TTQ aligns either with the “ingroup” or “outgroup” form
of trust, or whether it taps a different dimension altogether.
Constructing Ingroup and Outgroup Trust
Social identity theory is animated by the dichotomy of “us” versus “them”
based on categorical differences such as race, gender, religion, “native,” or
“stranger,” and any other conceivable difference. Social psychologists have
examined these differences in most detail (Allport; 1954; Sherif et al., 1961;
Tajfel, 1981, 1982, among many others), but such issues have also garnered the
attention of ethologists (van de Berghe, 1981; Butovskaya et al., 2000; Salter,
2004) and, more recently, of political scientists and economists (McCarty,
1993; Hero and Tolbert, 1996; Alesina and Glaeser, 2004; Lindert, 2004,
among others) as well as of philosophers (Walzer, 1983; Wrong, 1994; Miller,
2000, among others). Generally, social identity theory posits that the selfesteem of ingroup members rises when the category to which they belong is
positively evaluated. A similar effect is observed when members of an outgroup
are negatively evaluated and, additionally, ingroup bias is also strengthened
when members perceive a threat from an outgroup.
In order to sharpen the specification of generalized trust, we will use the
six-item trust measures from the fifth wave of the WVS to produce a twodimensional map of ingroup and outgroup trust. A two-dimensional map of
forms of trust has three distinct advantages: first, the six items in the fifth
WVS wave allow us to explicitly specify both types of trust. Second, it allows
us to measure the intensity of trust on each dimension. In other words, it is
possible to derive ingroup trust (high/low) as well as outgroup trust (high/low),
suggesting that ingroup trusters do trust, perhaps to an even higher degree
than outgroup trusters, but that their trust orbit is limited to people they know
or with whom they share categorical similarities such as race, region, language,
or religion. Third, it will allow us to measure the relationship between the two,
which is important because the exact relationship between these two forms
of trust is unclear. Our analysis will therefore help us understand if ingroup
and outgroup trust should be thought of as orthogonal to each other (i.e.,
no relationship between the two) (Newton, 2001), or a zero-sum, negative
relationship as others have argued (Banfield, 1958; Fukuyama, 1995; Uslaner,
2000, 2002; Yamagishi and Yamagishi, 1994; Newton, 1999; Stolle, 2001,
6
Social Science Quarterly
2002). Our analysis may also be in line with those that find ingroup and
outgroup trust are positively related to each other (Whiteley, 1999; Bahry
et al., 2005; Yosano and Hyashi, 2005; Glanville and Paxton, 2007; Putnam,
1993; Freitag and Traunmüller, 2009), or even that “it [particular trust] seems
to be a necessary but not a sufficient cause of more general forms of . . . social
trust” (Newton and Zmerli, 2011:194). Finally, if there is a distinction between
forms of trust, we suspect that they should have predictably different impacts
when it comes to questions of nativism and civicness, which we explain more
fully in the following section.
The Dependent Variables: Nativism and Civicness
As we indicated above, different forms of trust animate many outcomes,
which explains the burgeoning literature on the consequences of trust. We
have chosen nativism and civicness as dependent variables because the distinction between “us” and “them” looms large and differences in types of
trust should systematically affect the perception of respondents in these two
areas of attitudes. If it is indeed true that “ingroup” and “outgroup” trust
capture different radii of cooperative norms, then this should manifest itself
in natives’ attitudes toward immigrants. Similarly, even among native citizens,
wider radii of trust should predict more civic behavior than shorter radii of
trust. What nativism and civicness both have in common is an emphasis on
community: nativists see their own culture and ways of life threatened by
foreigners (see, e.g., Barry, 2001; Huntington, 2004) while others argue that
a decline in civic behavior will undermine the very foundations and working
of democracy (Pharr and Putnam, 2000; Putnam, 2000, among others). Nativism is centered on the “us first” idea, a fear that strangers will undermine
the traditional way of life, and is xenophobic and exclusionary. If it is indeed
true that outgroup trust engenders a broader sense of community, mutual aid,
inclusiveness, reciprocity, tolerance, and solidarity, we hypothesize a negative
relationship between that form of trust and our various measures of nativism.
Civicness, on the other hand, refers to participatory, solidaristic behavior
that is communitarian in character and centered on the common good. It
suggests a responsibility for the public good, the idea that obeying the laws,
paying one’s fair share in taxes, not taking unfair advantage of social benefits,
and that participating in voting is a duty that will make government work
better and furthers the interest of all members of the community (Orviska and
Hudson, 2002). Letki (2006:306) argues that “such sense of civic responsibility
and respect for the norms and rules of the community enhance the willingness
to comply with them, even when one is unlikely to be caught and if the threat of
punishment is minimal.” Outgroup trusters should contribute to the provision
of public goods because they trust others too, independently of whether
they are kith and kin, of a different religion, race, or characterized by other
differences. Hence, in terms of our civicness measures we expect outgroup trust
Trust Matters
7
to buoy civicness. On the other hand, ingroup trusters should be less capable of
involving themselves in their society as their circles of identity are drawn more
tightly. For this reason, we hypothesize a positive relationship between ingroup
trust and nativism and a negative relationship between ingroup trust and
civicness.
Alternatively, if we conceive ingroup and outgroup trust as a continuum
from a primordial, traditional, and authoritarian pole to one that is more
cosmopolitan, open, and accommodating of difference, ingroup trust may
be associated with more conservative attitudes on questions pertaining to
civicness. In this scenario, it would be the individuals with the narrower,
more restrictive form of trust that would find it more unacceptable to cheat
on taxes, avoid payment for public services, or abstain from participating in
parliamentary elections. Thinking of this sort undergirds the belief of many
scholars that the welfare state project is viable only on the basis of some kind
of primordial sentiments (see, e.g., Marshall, 1950; Walzer, 1983; Alesina and
Glaeser, 2004; Goodhart, 2004; Miller, 2000, and others). If this were the
case, we would actually see a positive relationship between ingroup trust and
civicness.
In more general terms, we are interested in how different forms of trust affect
the width of the moral community of individuals. To what extent do they see
themselves as their “brother’s keeper”; how tightly drawn are their circles of
identity and belonging? Outgroup trusters should have a higher capacity to
contribute their “fair share” to society as they see themselves as part and parcel
of a community; their attitudes should be consistent with an embracing of
the “other” and with higher levels of public spiritedness. Primordial attitudes,
on the other hand, should be more centered on ingroup benefits, skepticism
toward newcomers, nativism, and perhaps a reduced willingness to contribute
toward the common good.
For each dependent variable, we run two models: one with ingroup and
outgroup trust as the central predictors (in addition to a series of control
variables) and a second model that adds the TTQ to the other two forms
of trust. This allows us to explore whether there is a difference between
ingroup and outgroup trust, and the degree to which the TTQ either has
significant/insignificant independent effects and whether it aligns itself with
one or the other form of trust. The sample for our analysis consists of all
industrialized countries queried by the fifth wave of the WVS for which data
were available for all the models applied (with the exception of France, the
United Kingdom, and the Netherlands, in which two of the civicness items
were not asked).
The reason for constraining our sample to industrialized societies is threefold: first, according to Delhey, Newton, and Welzel (2011:800), the TTQ
“seems to work well in Western and affluent nations,” meaning that equivalence issues are greatly reduced. Second, we want to keep as many potentially
confounding effects, such as levels of democracy or development, at bay.
Third, all of the countries in this sample have seen significant increases of
8
Social Science Quarterly
immigration over a number of decades, which animates nativism (Mudde,
2007), and these are also the “trilateral democracies” (sans Japan) that are said
to have been afflicted by public apathy and disaffection with their governments, leading to widespread “civic malaise” (Pharr and Putnam, 2000). As a
result, they should prove fertile ground for examining whether different forms
of trust have systematically different effects on civicness and nativism. The
next section derives the empirical measures for ingroup and outgroup trust,
maps them against our set of industrialized societies, examines the relationship
between them, and finally estimates their effects on a series of nativism and
civicness models.
Methods and Findings
Wave 5 of the WVS includes several questions measuring respondents’
levels of trust toward a variety of different “types” of people. For example,
respondents are asked to record their level of trust toward their family members
and neighbors, as well as people of different religions or nationalities. Given
our theoretical definitions of ingroup and outgroup trust, we expect some
subsets of these survey responses to be causally connected to each of the two
different types of trust. Because we have theoretical expectations about the
components of the two forms of trust, we estimate confirmatory factor models
(CFAs).
We estimate a two-factor CFA allowing the two latent variables to be correlated to one another. Each factor is estimated using three observed indicators.
The primordial trust factor is causally connected to the questions asking
about family, neighbor, and personal acquaintances (V125–V127), whereas
the cosmopolitan trust factor is connected to the questions regarding people of
different religions, nationalities, and new acquaintances (V128–V130). These
survey questions are all measured on four-point scales, making the standard
CFA assumption of multivariate normality among the observed indicators
unrealistic. As a result, our model takes the categorical measurement of the
observed indicators into account using polychoric, rather than Pearson, correlations. The data cover 11 countries with a total of 13,051 respondents. Note
that results are replicated in country-by-country factor models.
Following the notation of Kolenikov (2009), the CFA models take the
following general form:
Yi j = µ j +
!
λ j k ξik + δi j , j = 1, . . . , p,
where i indexes observation, j indexes the observed indicator, k indexes the
latent factor, and pk is the number of observed indicators associated with
each of the latent factors. The yij- s represent observation i’s response to survey
question j. λjk is the factor loading relating indicator j to factor k and the µj- s
Trust Matters
9
TABLE 1
Confirmatory Factor Analysis of Trust
Loading on Factor 1:
Ingroup Trust
Immediate family
Neighbors
Personal acquaintances
People you first meet
People of another religion
People of another nationality
n
χ2
CFI
RMSEA
AIC
BIC
0.52
0.70
0.76
.
.
.
Loading on Factor 2:
Outgroup Trust
13,051
3,456.39
0.95
0.09
123,194.35
123,381.26
.
.
.
0.68
0.92
0.95
NOTE: Countries included are Australia, Finland, France, Great Britain, Germany, Italy, the
Netherlands, Norway, Sweden, Switzerland, and the United States. Results from a confirmatory factor analysis of a polychoric correlation matrix. In order to obtain the CFI and RMSEA
fit statistics, we calculated an additional factor analysis of a Pearson’s correlation matrix.
are the intercepts. The ξ ik s are the latent factors and the δ ij s are measurement,
or unique, errors.
Before estimation, we must assess the identification of our model. First, we
set the mean and variance of the latent variables to zero and one, respectively.
Next we compare the number of estimated parameters in our model to the
degrees of freedom. In order for the model to be identified, the degrees of
freedom must equal or exceed the number of free parameters. In our model,
we have six observed indicators that correspond to 18 free parameters plus one
additional parameter to model the correlation between the two latent factors.
The degrees of freedom equal the number of nonredundant elements in the
variance/covariance matrix of the observed indicators, which is given by
df =
p ( p + 1)
,
2
where p is the number of observed indicators. Given six observed indicators,
this yields 21 degrees of freedom, which exceeds the 19 free parameters to be
estimated in the model. Finally, given that each factor is related to at least three
observed indicators, the residual covariances of these indicators are assumed
to be zero, and each observed indicator loads on one and only one factor, the
model is identified under the three indicator rule (Bollen, 1989). We estimated
the model via full information maximum likelihood using Mplus (Muthén
and Muthén, 2008).
10
Social Science Quarterly
FIGURE 1
italy
germany
netherlands
switzerland
australia
usa
britain
finland
france
norway
sweden
−1
−.5
mean of in−group trust
0
.5
1
mean of out−group trust
Table 1 presents factor loadings for each of the six observed indicators of
trust. As is apparent from the table, there are strong loadings between trust
in immediate family, trust in neighbors, and trust in personal acquaintances
on the first factor, what we call ingroup trust. In addition, trust in people
one first meets, trust in individuals of another religion, and trust in individuals of another nationality load highly on a separate factor, which we
call outgroup trust. The two factors correlate at r = 0.67, indicating that
that there is a moderately strong relationship between the two types of trust.
It appears that people who elicit ingroup trust also have the capacity for
outgroup trust. In other words, we find, unlike Newton (2001), that the relationship between the two dimensions is not orthogonal; rather, there is a
positive relationship. However, a correlation of 0.67 also suggests that there is
a fair proportion of variation in each factor that is unique to each theoretical
concept.
Using the results of the CFAs, we generate factor scores for each individual
in our sample for both types of trust. We prefer this method to additive indices,
which do not account for variation in the loadings of each constitutive variable
on the latent factors.4 We display the mean value of each type of trust across
4
We also generated the trust variables using additive indices, and results are substantively
the same.
Trust Matters
11
the countries in our sample in Figure 1. Unsurprisingly, Sweden displays the
highest levels of ingroup and outgroup trust and fellow Nordics of Norway
and Finland also display high levels of trust. It is equally unsurprising, based
on previous research (Banfield, 1958; Putnam, 2007), to find Italy with the
lowest amount of trust given the history of low trust in the southern part of the
country in particular. The figure further demonstrates that the two dimensions
of trust are positively related, a finding consistent with other results (Bahry
et al., 2005; Glanville and Paxton, 2007; Freitag and Traunmüller, 2009;
Newton and Zmerli, 2011; Zmerli and Hooghe, 2011). Interestingly, however,
although the two types of trust correlate positively with one another, later in
this article we will show that they have a markedly different impact on attitudes
toward nativism and civicness.
We identify four survey questions in the WVS that capture nativism and
three that capture civicness, which are listed below:
Nativism:
1. “Turning to the question of ethnic diversity, with which one of the
following views do you agree? Ethnic diversity erodes a country’s unity
(coded 1); Ethnic diversity enriches life (coded 10).” (V221)
2. “How do you feel about people from other countries coming here to
work? Which one of the following do you think the government should
do?” (V124)
!
!
!
!
“Let anyone come who wants to” (assigned a 4)
“Let people come as long as there are jobs available” (3)
“Place strict limits on the number of foreigners who can come here” (2)
“Prohibit people coming here from other countries” (1)
3. (V45) asked respondents to indicate whether they agreed with a statement about employers and the hiring of nationals over immigrants. We
code respondents who felt that nationals should be hired over immigrants a 0, and those who felt that nationals should not be favored over
immigrants as 1.
4. “Which of the following groups of people would you not like to have
as neighbors?” (V37). If the respondent checked the immigrants/foreign
workers box, we assign a 1. If not, we assign a 0.
Civicness:
“For each of the following actions, please indicate whether you think it can
always be justified, never be justified, or something in between.”
1. Cheating on taxes if you have a chance (always justified coded 1, never
justified coded 10; V200)
2. Avoiding a fare on public transport (always justified coded 1, never
justified coded 10; V199)
3. Did you vote in recent parliamentary elections? (coded 1 if yes, coded 0
if no; V234)
12
Social Science Quarterly
We estimate a series of models with the ingroup and outgroup trust measures
taken as predictors of each nativism and civicness variable. Each of the seven
models includes a number of variables that previous research suggests have
an impact on attitudes toward nativist and civic attitudes. Scholarship has
found that females and respondents with more education display significantly
less nativist resentment and more civic attitudes than males and those with
less education. Additionally, older people as well as those who are poorer
are more likely to hold negative attitudes toward immigrants and cultural
diversity than younger respondents or those who are in a less precarious
economic position (Crepaz, 2008:75–77). We expect similar trends to hold
in our analysis as well. Others have also found life satisfaction to be positively
related to interpersonal trust, with similar effects on democratic institutions
(Inglehart, 1990:33/41). We therefore expect respondents who display higher
levels of life satisfaction to be less likely to hold nativist attitudes and more
likely to hold attitudes consistent with civicness. Although older respondents
are expected to display more nativist attitudes, we do anticipate that they will
be more likely to vote and less likely to find it acceptable to cheat on taxes
or avoid fares (Crepaz, 2008). Finally, a number of studies have found that
left/right orientation has an effect on attitudes toward welfare state principles
and politics (Arts and Gelissen, 2001; Bean and Papadakis, 1998; Edlund,
2003; Feldman and Zaller, 1992; Gelissen, 2000; Jacoby, 1994), and we
accordingly include the respondents’ self-placement on the left/right scale.
We expect respondents further to the right of the left/right scale to display
less support for immigrants, but are uncertain what impact a respondents’
left/right self-placement will have on our measures of civicness.
The number of observations varies across tables, as response rates fluctuated
by question. The ethnic diversity and immigration policy questions were
not asked in Britain, France, and the Netherlands, meaning these countries
are not included in the models corresponding to these dependent variables.
Age is measured in years, gender is measured dichotomously, with females
coded as 1 and males 0, and income is sorted into 10 strata within each
country. Education is measured on a nine-point ordinal scale, ranging from
no formal education, coded as 1, to completed university education, coded 9.
Left–right placement is measured with a question that asks individuals to
report their ideology, with 1 indicating “left” and 10 indicating “right.” We
also include a control for life satisfaction, which ranges from 1 to 10, with lower
values indicating dissatisfaction with one’s life and higher values indicating
satisfaction. To assure that ingroup and outgroup trust have predictive power
beyond their association with generalized trust, we also control for generalized
trust using the WVS question that asks respondents if they agree that most
people can be trusted. Those who agree are coded 1, and those who do not
are coded 0 (V23).
Regarding nativism, results in Table 2 provide strong evidence that ingroup
and outgroup trusters display very different attitudes in regard to the benefits
of immigration policy, ethnic diversity, the hiring of nationals, and having
Outgroup trust
−4.08
−0.73
2.26
7,873
0.06
−7,559
0.44∗∗∗
(0.08)
−0.28∗∗
(0.10)
0.363∗∗∗
(0.08)
0.041∗∗∗
(0.02)
−0.01∗∗∗
(0.003)
−0.013
(0.06)
0.08∗∗∗
(0.02)
−0.040
(0.08)
−0.159∗∗∗
(0.02)
Immigration Policy
0.47∗∗∗
(0.08)
−0.24∗∗
(0.11)
0.05∗∗∗
(0.02)
−0.01∗∗∗
(0.003)
−0.02
(0.05)
0.08∗∗∗
(0.01)
−0.03
(0.08)
−0.16∗∗∗
(0.02)
−4.232
−0.887
2.104
8,074
0.06
−7,761
TABLE 2
Trust and Nativism
0.11∗∗∗
(0.03)
−0.01∗∗∗
(0.002)
0.27∗∗
(0.10)
0.08∗∗∗
(0.02)
−0.16
(0.10)
−0.20∗∗∗
(0.03)
7.04∗∗∗
(0.34)
7,907
0.62∗∗∗
(0.03)
−0.43∗∗∗
(0.07)
0.43∗∗∗
(0.05)
0.09∗∗∗
(0.02)
−0.01∗∗∗
(0.002)
0.28∗∗
(0.02)
0.08∗∗∗
(0.02)
−0.17
(0.10)
−0.20∗∗∗
(0.03)
7.889∗∗∗
(0.32)
Ethnic Diversity: Good
8,112
0.18
0.64∗∗∗
(0.03)
−0.37∗∗∗
(0.001)
0.17
0.45∗∗∗
(0.08)
−0.44∗∗∗
(0.100)
0.68∗∗
(0.29)
0.06∗∗
(0.03)
−0.01∗∗∗
(0.003)
0.05
(0.06)
0.07∗
(0.03)
0.04
(0.16)
−0.10∗∗∗
(0.03)
1.90
(0.68)
Immigrant Neighbors
0.08∗∗
(0.03)
−0.01∗∗∗
(0.003)
0.04
(0.06)
0.08∗
(0.04)
0.05
(0.16)
−0.11∗∗∗
(0.03)
2.07
(0.70)
10,004
0.07
−3,386
Hiring Immigrants
0.44∗∗∗
(0.05)
−0.27∗∗∗
(0.06)
0.45∗∗∗
(0.09)
0.04∗∗
(0.02)
−0.01∗∗∗
(0.003)
−0.03
(0.03)
0.07∗∗∗
(0.02)
0.03
(0.09)
−0.14∗∗∗
(0.02)
0.55∗∗∗
(0.20)
10,249
0.06
−3,495
0.49∗∗∗
(0.08)
−0.38∗∗∗
(0.09)
0.06∗∗∗
(0.02)
−0.01∗∗∗
(0.003)
0.04∗
(0.02)
0.07∗∗∗
(0.02)
0.03
(0.09)
−0.14∗∗∗
(0.02)
0.66∗∗∗
(0.18)
10,069
0.09
−6,341
0.46∗∗∗
(0.05)
−0.23∗∗∗
(0.06)
10,318
0.08
−6,546
Trust Matters
Ingroup trust
Generalized trust
Life satisfaction
Age
Female
Income
Education
Left–right ideology
Constant
τ1
τ2
τ3
n
Pseudo-R2
Log-likelihood
R2
NOTE: Cluster robust standard errors in parentheses. Immigration policy model estimated with ordered logistic regression. Diversity model estimated
with linear regression. Hiring and neighbors models estimated with binary logistic regression.
∗ ∗∗∗ p < 0.01, ∗∗ p < 0.05, ∗ p < 0.1.
13
14
Social Science Quarterly
FIGURE 2
.8
.6
Predicted probability
Predicted probability
.6
.4
.2
.4
.2
0
0
−4
−2
0
In−group trust
2
prohibit from coming
if jobs available
−4
−2
0
Out−group trust
2
4
strict limits
let anyone come
immigrants as neighbors. Across each of the tables, the coefficient on the
outgroup trust variable is positive and significantly related to the dependent
variable. This positive association remains even when controlling for generalized trust. Alternatively, across each of the nativism models ingroup trust
is negatively related to the dependent variable, meaning it is associated with
higher levels of nativism, whereas outgroup trust is associated with lower levels of nativism; the opposite is true for ingroup trust. Again, the negative
association between ingroup trust and the dependent variables remains even
when controlling for generalized trust.
For the models that we estimated with ordered or binary logit, we display the
relationships between ingroup and outgroup trust and the dependent variables
graphically. Figure 2 displays the relationships between the trust variables and
attitudes toward immigration policy. As shown in the right-hand panel of
the figure, as outgroup trust increases, individuals are more likely to feel that
all immigrants should be allowed, or that immigrants should be allowed if
jobs are available. Conversely, as outgroup trust increases, individuals are less
likely to feel that there should be strict limits on immigration or that no
one should be allowed to immigrate to the country. As shown in the lefthand panel of the figure, opposite patterns are found in regard to ingroup
trust.
Trust Matters
15
FIGURE 3
P (Accept as neighbor)
P (Not favor nationals)
1
.8
.6
.4
.2
−4
−2
0
Trust
2
4
−4
−2
0
Trust
2
4
0
Trust
2
4
1
.9
.8
.7
.6
.5
P (Voted)
.9
.85
.8
.75
−4
−2
Out−group Trust
In−group Trust
The top panel of Figure 3 displays the relationships between the trust
variables and attitudes toward the hiring of immigrants. As outgroup trust
increases, individuals are more likely to feel that preference should not be
given to nationals, while as ingroup trust increases, individuals are more
likely to feel that employers should give preference to nationals. The patterns
in the middle panel of Figure 3 are similar, with outgroup trusters more
16
Social Science Quarterly
TABLE 3
Trust and Civicness
Cheat
on Taxes
Outgroup trust
Ingroup trust
Generalized trust
0.01
(0.04)
0.19∗∗∗
(0.06)
Life satisfaction
0.06∗∗∗
(0.02)
Age
0.01∗∗∗
(0.002)
Female
0.42∗∗∗
(0.06)
Income
−0.01
(0.01)
Education
−0.03
(0.06)
Left–right ideology −0.02
(0.02)
Constant
7.59∗∗∗
(0.33)
n
Psuedo-R2
Log-likelihood
R2
Fares
on Public
Transport
Voted in
Parliamentary
Elections
0.003 −0.07∗∗ −0.07∗
0.01
−0.01
(0.05)
(0.03)
(0.03)
(0.03)
(0.03)
0.20∗∗
0.14∗∗∗
0.15∗∗∗
0.29∗∗
0.16∗
(0.06)
(0.04)
(0.04)
(0.10)
(0.09)
−0.03
−0.05
0.04∗∗∗
(0.11)
(0.08)
(0.10)
0.07∗∗∗
0.06∗∗
0.06∗∗
0.06∗
0.04
(0.02)
(0.02)
(0.02)
(0.03)
(0.03)
0.01∗∗∗
0.03∗∗∗
0.03∗∗∗
0.04∗∗∗
0.04∗∗∗
(0.002) (0.002) (0.002) (0.003) (0.003)
0.43∗∗∗
0.13∗∗
0.13∗∗
0.04
0.03
(0.07)
(0.05)
(0.05)
(0.09)
(0.10)
−0.01
−0.001 −0.0004
0.12∗∗∗
0.12∗∗∗
(0.01)
(0.02)
(0.02)
(0.03)
(0.03)
−0.03
0.02
0.01
−0.04
−0.05
(0.06)
(0.09)
(0.09)
(0.10)
(0.10)
0.04∗∗
0.01
0.02
−0.02
0.04∗∗
(0.02)
(0.02)
(0.02)
(0.02)
(0.02)
7.59∗∗∗
6.29∗∗∗
6.35∗∗∗ −1.13∗∗∗ −1.21∗∗∗
(0.37)
(0.37)
(0.39)
(0.41)
(0.39)
10,415
10,157
10,420
10,164
0.04
0.04
0.09
0.09
10,253
0.08
−4,302
10,009
0.08
−4,187
NOTE: Cluster robust standard errors in parentheses. Taxes and transport models estimated
with linear regression. Voting model estimated with binary logistic regression.
∗∗∗∗ p < 0.01, ∗∗ p < 0.05, ∗ p < 0.1.
likely to accept immigrants as neighbors and ingroup trusters less likely to
do so.
We next turn to our civicness measures. Here, the relationships between
the dependent variables and the trust measures are different. Across each
of the models, displayed in Table 3, ingroup trust is positively related to
civicness, while outgroup trust is negatively related or unrelated to civicness.
These patterns hold when controlling for generalized trust. We display the
relationships between the trust measures and the likelihood of voting in the
bottom panel of Figure 3, as the turnout models in Table 3 were estimated
with binary logistic regression. As is clear from the figure, turnout propensity
is essentially unchanged as outgroup trust rises, but increases as ingroup trust
rises.
Trust Matters
17
Our results demonstrate remarkable differences in the attitudes and behavior of ingroup and outgroup trusters. In terms of nativism, whereas outgroup
trusters are more likely to see value in ethnic diversity and to exhibit attitudes that are open to immigration and immigrants, the opposite is true of
ingroup trusters. In terms of civicness, however, ingroup trusters are more
likely to “accept the system” and participate in it. This is not true of outgroup
trusters.
Interestingly, generalized trust tends to run in the same direction as outgroup trust in the models with the “nativist” dependent variables, whereas in
the “civicness” models, it either has the opposite effect of outgroup trust or is
insignificant. More specifically, it is insignificantly related to the “tax cheating” and “fare jumping” variables and positively and significantly related to
the likelihood of voting. These diverging results indicate that the more highly
specified ingroup and outgroup trust measures capture something quite different from the unidimensional trust question. However, scholars using the
TTQ only have found no relationship between generalized trust and civic
morality (Letki, 2006) or between generalized trust and legal permissiveness
(Marien and Hooghe, 2011). While the emphasis of our research is on the
difference between ingroup and outgroup trust measured on two dimensions,
in the tables that use the unidimensional TTQ as a control, our findings converge with previous scholarship demonstrating no effect of generalized trust on
civicness. This serves to highlight the importance of using a two-dimensional
map of ingroup and outgroup trust as it reveals relationships that otherwise
would remain obscured. Hence, we agree with Delhey, Newton, and Welzel
(2011:801) that “comparative research should discontinue using unspecified
trust [TTQ], at face value, as a measure of general trust . . . ”
Regarding the remaining control variables, patterns are also quite different
across the nativism and civicness models. In the nativist models, age and rightward political ideology tend to relate to more nativist attitudes (or negatively
with the dependent variables), while life satisfaction and education tend to be
associated with less nativism. Results on the control variables are less consistent
across the civicness models, though life satisfaction and age are consistently
positively related to “civic” attitudes and turnout propensity.
Conclusions and Implications
The case for differentiating between types of trust has been made in previous theoretical studies and our findings add to a growing consensus that
differentiating between trusts is conceptually accurate and of practical significance in furthering our understanding of the impact of trust in diverse
societies. While our empirical analysis provides further support for the existence of two separate dimensions of trust, ingroup and outgroup, we have
also shown that these two forms of trust are not orthogonal to each other.
18
Social Science Quarterly
Consistent with the expectations of Putnam (1993:144) and the studies of
Bahry et al. (2005), Glanville and Paxton (2007), Freitag and Traunmüller
(2009), Newton and Zmerli (2011), Zmerli and Hooghe (2011), and others,
we find a positive relationship between these two types of interpersonal trust
in our sample of West European countries, the United States, and Australia.
Importantly, however, despite the positive relationship, it is striking to observe
the systematically different impact of the two dimensions when applied to a
set of political attitudes in the fifth wave of the WVS measuring attitudes
toward nativism and civicness.
In all of the nativism regressions, ingroup trusters display more nativist
attitudes than outgroup trusters even when controlling for the TTQ, which
we added to explore how ingroup and outgroup trust would be affected by
the unidimensional predictor variable for trust. When asked whether “ethnic
diversity enriches life,” outgroup trusters strongly agree (together with generalized trusters) while ingroup trusters strongly believe that it tends to erode a
country’s unity. Outgroup trusters, as well as generalized trusters, do not think
that employers should give preference to nationals when jobs are scarce while
ingroup trusters strongly believe they should. While outgroup trusters believe
that immigration should be liberalized, together with generalized trusters,
the coefficient for ingroup trust points in the opposite direction. Similarly,
while outgroup trusters feel comfortable (and traditional trusters also) with
immigrant neighbors, the ingroup trusters are uncomfortable living among
immigrants.
When it comes to civicness, however, the relationship is reversed. Ingroup
trusters significantly believe that it is not justifiable to cheat on taxes while the
coefficients for outgroup trusters, together with generalized trusters, did not
reach statistical significance. Strikingly, when asked whether it is unacceptable
to avoid fares on public transport, ingroup trusters again tend to be more civic
and find it unacceptable, while outgroup trusters believe that it is acceptable
to avoid fares on public transport. Finally, when it comes to one of the
most civic acts of all, namely, voting, it turns out that ingroup trusters,
together with generalized trusters, are more likely to have said they voted in
the most recent parliamentary election while outgroup trusters did not. Given
the characteristics of outgroup trust, these findings are puzzling and call for
further research into these matters.
Our findings suggest that those with more ingroup attitudes are trusters
too, and that although their orbit of trust is more tightly circumscribed and
more limited in its reach, this type of trust is important in fostering civicallyoriented attitudes. In this our findings are somewhat consistent with research
suggesting that primordial sentiments facilitate ingroup trust, a necessary
component for the continued viability of modern welfare states (Marshall,
1950; Walzer, 1983; Freeman, 1986; Alesina and Glaeser, 2004; Goodhart,
2004; Miller, 2000; Huntington, 2004; Lindert, 2004). It is the outgroup
trusters, however, who are able to transcend the boundaries of otherness, and
for whom the traditional markers of identity are no deterrence to trust.
Trust Matters
19
Three things stand out from these findings: first, significant differences exist
in the attitudes of ingroup and outgroup trusters, even though the two types
of trust are correlated with one another. Second, there are significant differences between the two-dimensional concept of trust and the TTQ. In the
nativist battery of questions, the TTQ aligns with the outgroup dimension of
trust, while for the civicness questions, it either has no effect (for two of the
three questions) or it aligns with the ingroup dimension. These different outcomes demonstrate the value added of the multidimensional concepts of trust
compared to the unidimensional TTQ. Finally, despite the communitarian
outlook of outgroup trusters, the negative findings on the civicness questions
are somewhat puzzling and require further examination.
Modern societies are coping with dramatic increases in immigrationinduced diversity and the attendant political and social challenges. It appears
that when it comes to issues of nativism, outgroup trusters have an expanded
view of community, a belief in the shared fate of all that transcends ethnic,
racial, and religious differences. Paradoxically, the same attitudes do not engender more civicness on the part of outgroup trusters. It is people with a
wider orbit of trust who appear ready to embrace increasing diversity in our
modern societies, but people with a more narrow orbit of trust are also needed
to ensure the common good at home.
REFERENCES
Alesina, A., and E. L. Ferrara. 2002. “Who Trusts Others?” Journal of Public Economics
85(2):207–34.
Alesina, A., and E. L. Glaeser. 2004. Fighting Poverty in the US and Europe. Oxford: Oxford
University Press.
Allport, G. 1954. The Nature of Prejudice. Cambridge: Addison-Wesley.
Arts, W., and J. Gelissen. 2001. “Welfare States, Solidarity and Justice Principles: Does the
Type Really Matter?” Acta Sociologica 44:284–99.
Bahry, D. , M. Kosolapov, P. Kozyreva, R. K. Wilson, M. Butovskaya, F. Salter, I. Diakonov,
and A. Smirnov 2005. “Ethnicity and Trust: Evidence from Russia.” American Political Science
Review 99:521–32.
Banfield, E. 1958. The Moral Basis of a Backward Society. New York: Free Press.
Barry, B. 2001. Culture and Equality. Cambridge: Harvard University Press.
Bean, C., and E. Papadakis. 1998. “A Comparison of Mass Attitudes Towards the Welfare
State in Different Institutional Regimes, 1985–1990.” International Journal of Public Opinion
Research 10:211–36.
Bollen K. A. 1989. Structural Equations with Latent Variables. New York: John Wiley &
Sons.
Butovskaya, M. et. al. 2000. “Urban Begging and Ethnic Nepotism in Russia. An Ethological
Pilot Study.” Human Nature 11:157–82.
Crepaz, M. M. L. 2008. Trust Beyond Borders. Immigration, the Welfare State and Identity in
Modern Societies. Ann Arbor, MI: University of Michigan Press.
20
Social Science Quarterly
Delhey, J., K. Newton, and C. Welzel. 2011. “How General is Trust in ‘Most People’? Solving
the Radius of Trust Problem.” American Sociological Review 76:786–807.
Edlund, J. 2003. “The Influence of the Class Situations of Husband and Wives on Class
Identity, Party Preference and Attitudes Towards Redistribution: Sweden, Germany and the
United States.” Acta Sociologica 46:195–214.
Feldman, S., and J. Zaller. 1992. “The Political Culture of Ambivalence: Ideological Responses
to the Welfare State.” American Journal of Political Science 36:268–307.
Freeman, G. 1986. “Migration and the Political Economy of the Welfare State.” Annals of the
American Academy of Political and Social Science 485:51–63.
Freitag, M., and R. Traunmüller. 2009. “Spheres of Trust: An Empirical Analysis of the
Foundation of Particularised and Generalised Trust.” European Journal of Political Research
48:782–803.
Fukuyama, F. 1995. Trust: The Social Virtues and the Creation of Prosperity. London: Hamish
Hamilton.
Gelissen, J. 2000. “Popular Support for Institutionalized Solidarity: A Comparison Between
European Welfare States.” International Journal of Social Welfare 9:285–300.
Gibson, J. L. 2001. “Social Networks, Civil Society and the Prospects for Consolidating Russia’s
Democratic Transition.” American Journal of Political Science 45:51–68.
Glanville, J. L., and P. Paxton. 2007. “How Do We Learn to Trust? A Confirmatory Tetrad
Analysis of the Sources of Generalized Trust.” Social Psychology Quarterly 70:230–42.
Goodhart, D. 2004. “Too Diverse?” Prospect February:30–7.
Granovetter, M. S. 1973. “The Strength of Weak Ties.” American Sociological Review
78:1360–80.
Hero, R., and C. J. Tolbert. 1996. “A Racial/Ethnic Diversity Interpretation of Politics and
Policy in the States of the US.” American Journal of Political Science 40:851–71.
Herreros, F. 2004. The Problem of Forming Social Capital: Why Trust? New York: Palgrave
Macmillan.
Huntington, S. 2004. Who Are We? The Challenges to America’s National Identity. New Haven:
Yale University Press.
Inglehart, R. 1990. Culture Shift in Advanced Industrial Society. Princeton, NJ: Princeton
University Press.
Jacoby, W. G. 1994. “Public Attitudes Towards Government Spending.” American Journal of
Political Science 38:336–61.
Knack, S., and P. Keefer. 1997. “Does Social Capital Have an Economic Payoff? A CrossCountry Investigation.” Quarterly Journal of Economics 112:1251–88.
Kolenikov, S. 2009. “Confirmatory Factor Analysis Using Confa.” Stata Journal 9.3:329–73.
Letki, N. 2006. “Investigating the Roots of Civic Morality: Trust, Social Capital, and Institutional Performance.” Political Behavior 28:305–28.
Lindert, P. H. 2004. Growing Public: Social Spending and Economic Growth Since the Eighteenth
Century. Cambridge: Cambridge University Press.
Marien, S., and M. Hooghe. 2011. “Does Political Trust Matter? An Empirical Investigation
into the Relation Between Political Trust and Support for Law Compliance.” European Journal
of Political Research 50:267–91.
Marshall, T. H. 1950. Citizenship and Social Class and Other Essays. Cambridge, UK: Cambridge University Press.
Trust Matters
21
McCarty, T. 1993. “Democratic Diversity and the Size of the Public Sector.” Kyklos
46:225–40.
Miller, D. 2000. Citizenship and National Identity. Cambridge: Polity.
Miller, G. 2001. “Why is Trust Necessary in Organizations? The Moral Hazard of Profit
Maximization.” Pp. 307–31 in K. S. Cook, ed., Trust in Society. New York: Russell Sage
Foundation.
Mudde, C. 2007. Populist Radical Right Parties in Europe. Cambridge: Cambridge University
Press.
Muthén, L. K., and B. O. Muthén. 2008. “Mplus User’s Guide: Statistical Analysis with
Latent Variables." Available at <http://www.statmodel.com/download/usersguide/Mplus%
20Users%20Guide%20v6.pdf>. Accessed December 4, 2011.
Newton, K. 1999. “Social and Political Trust in Established Democracies.” Pp. 169–87 in P.
Norris, ed., Critical Citizens: Global Support for Democratic Government. Oxford, UK: Oxford
University Press.
—— 2001. “Social Trust and Political Dissatisfaction: Social Capital and Democracy." Paper
presented at the EURESCO Conference on Social Capital: Interdisciplinary Perspectives,
Exeter, UK.
Newton, K., and S. Zmerli. 2011. “Three Forms of Trust and Their Association.” European
Political Science Review 3:169–200.
Orviska, M., and J. Hudson. 2002. “Tax Evasion, Civic Duty and the Law Abiding Citizen.”
European Journal of Political Economy 19:83–102
Pharr, S. J., and Putnam, R. 2000. Disaffected Democracies: What’s Troubling the Trilateral
Countries? Princeton, NJ: Princeton University Press.
Putnam, R. D. 1993. Making Democracy Work. Civic Traditions in Modern Italy. Princeton,
NJ: Princeton University Press.
——. 2007. “E Pluribus Unum: Diversity and Community in the Twenty-First Century. The
2006 Johan Skytte Prize Lecture.” Scandinavian Political Studies 30(2):137–74.
——. 2000. Bowling Alone. The Collapse and Revival of American Community. New York:
Simon and Schuster.
———, ed. 2002. Democracies in Flux. Oxford: Oxford University Press.
Reeskens, T., and M. Hooghe. 2008. “Cross-Cultural Measurement Equivalence of Generalized
Trust: Evidence from the European Social Survey.” Social Indicators Research 85:515–32.
Salter, F. K. 2004. Welfare, Ethnicity, and Altruism. London: Frank Cass.
Scholz, J. T. 1998. “Trust, Taxes and Compliance.” Pp. 135–66 in V. Braithwaite and M. Levi,
eds., Trust and Governance. New York: Russell Sage Foundation.
Scholz, J. T., and M. Lubell. 1998. “Trust and Taxpaying: Testing the Heuristic Approach to
Collective Action.” American Journal of Political Science 42(2):398–417.
Scholz, J. T., and N. Pinney. 1995. “Duty, Fear, and Tax Compliance: The Heuristic Basis of
Citizenship Behavior.” American Journal of Political Science 39:490–512.
Sherif, M., O. J. Harvey, B. J. White, W. Hood, and C. W. Sherif. 1954/61. "Intergroup Conflict and Cooperation: The Robbers Cave Experiment.” In Classical Texts in Psychology. Toronto,
ON: York University. Available at: <http://psychcentral.com/classics/Sherif/index.htm>.
Slemrod, J., and P. Katuscak. 2005. Do Trust and Trustworthiness Pay Off? Journal of Human
Resources 40:621–46.
22
Social Science Quarterly
Soroka, S., K. Banting, and R. Johnston. 2002. “Ethnicity, Trust, and the Welfare State." PDF
paper. Available at <http://www.mcgill.ca/files/politicalscience/EthTrustWelf.pdf>.
Stolle, D. 2001. “Clubs and Congregations: The Benefits of Joining an Association.” Pp.
202-44 in K. S. Cook, ed., Trust in Society. New York: Russell Sage Foundation.
——. 2002. “Trusting Strangers: The Concept of Generalized Trust in Perspective.” Ö
Zeitschrift fur Politikwissenschaft 31:397–412.
Sturgis, P., and P. Smith. 2010. “Assessing the Validity of Generalized Trust Questions: What
Kind of Trust Are We Measuring?” International Journal of Public Opinion Research 22:74–
92.
Tajfel, H. 1981. Human Groups and Social Categories. Cambridge: Cambridge University Press.
———. 1982. Social Identity and Group Relations. Cambridge: Cambridge University Press.
Uslaner, E. M. 2000. “Producing and Consuming Trust.” Political Science Quarterly 115:569–
90.
———. 2002. The Moral Foundations of Trust. Cambridge, UK: Cambridge University Press.
van den Berghe, P. 1981. The Ethnic Phenomenon. New York: Elsevier.
Walzer, M. 1983. Spheres of Justice. New York: Basic Books.
Welch, M. R., Sikkink, D., and M. T. Loveland. 2007. “The Radius of Trust: Religion, Social
Embeddedness and Trust in Strangers”. Social Forces 86:23–46.
Whiteley, P. 1999. “Economic Growth and Social Capital.” Political Studies 48:443–66.
Wollebaek, D., S. Wallmann Lundåsen, and Lars Trägårdh. 2012. “Three Forms of Interpersonal Trust: Evidence from Swedish Municipalities.” Scandinavian Political Studies 35:319–46.
Wrong, D. 1994. The Problem of Order. What Unites and Divides Society. New York: Free Press.
Yamagishi, T., and M. Yamagishi. 1994. “Trust and Commitment in the United States and
Japan.” Motivation and Emotion 18:129–66.
Zak, P., and S. Knack. 2001. “Trust and Growth.” Economic Journal 111:295–321.
Zmerli, S., and M. Hooghe, eds. 2011. Political Trust. Why Context Matters. Essex: University
of Essex ECPR Press.
Zmerli, S., and K. Newton. 2008. “Social Trust and Attitudes Toward Democracy.” Public
Opinion Quarterly 72:706–24.