bs_bs_banner

P O L I T I C A L S T U D I E S : 2 0 1 4 VO L 6 2 , 5 7 3 – 5 9 5
doi: 10.1111/1467-9248.12068

Does Diversity Erode Social Cohesion?

Conceptual and Methodological Issues

Gal Ariely
Ben-Gurion University of the Negev

Robert Putnam’s ‘hunkering down’ thesis regarding the negative effect of ethnic diversity on trust and willingness to
participate in collective life launched an ongoing debate concerning the ramifications of diversity for social cohesion.
Findings regarding the way in which diversity affects social cohesion are discrepant, some scholars arguing that diversity
has negative effects on social cohesion and others indicating insignificant or even positive effects.This study claims that
these conflicting conclusions are explained by the vagueness of social cohesion – a multidimensional concept.
Analyzing cross-national survey data from 42 European countries, it demonstrates how diversity is variably related to
the diverse dimensions and operationalization of social cohesion. While diversity is not associated with the most
commonly adduced dimension of social cohesion – namely, interpersonal trust – it does possess a negative relation to
two other dimensions of social cohesion: belonging and social solidarity. Even these negative relations are not
consistent across different operationalizations of belonging and social solidarity, however. In the face of increasing
concerns regarding the implications of diversity for social cohesion, these findings demonstrate that caution must be
exercised when examining the relations between these two phenomena.

Keywords: social cohesion; diversity

John Stuart Mill’s adage that ‘Free institutions are next to impossible in a country made up
of different nationalities.Among a people without fellow-feeling, especially if they read and
speak different languages, the united public opinion necessary to the working of represen-
tative government can not exist’ has become very popular recently (Mill, 1958 [1861] , pp.
232–3). This argument for the necessity of ‘fellow feeling’ and ‘united public opinion’ is
reflected in modern concern that the increasing diversity caused by immigration reduces
the conditions necessary for ‘social cohesion’ – a fashionable buzzword in the discussion of
immigration and integration policy (Holtug and Mason, 2010).
Similar apprehensions have also generated greater research into the links between
diversity and social cohesion. Robert Putnam’s (2007) ‘hunkering down’ thesis concerning
the negative effects of ethnic diversity on trust and the willingness to participate in
collective life launched an ongoing debate regarding the implications of diversity for social
cohesion.To date, none of these studies has produced robust results.While some have found
diversity to have negative effects in disparate countries (e.g. Agirdag et al., 2011; Fieldhouse
and Cutts, 2010; Laurence, 2011) or across countries (Anderson and Paskeviciute, 2006;
Delhey and Newton, 2005), others have not (e.g. Gesthuizen et al., 2009; Hooghe et al.,
2009; Savelkoul et al., 2011; Sturgis et al., 2011).
In light of these contrary results, several scholars have pointed out conceptual and
methodological flaws in the study of diversity and social cohesion. Marc Hooghe (2007), for
example, has argued that, in much of the literature, generalized trust is used to measure the
concepts of social capital and social cohesion – an operationalization that does not fully
account for the comprehensive meaning of these phenomena. Similarly, Allison Harell and
© 2013 The Author. Political Studies © 2013 Political Studies Association
574 G A L A R I E LY

Dietlind Stolle (2010) claim that other dimensions of social cohesion – such as norms and
solidarity – must be addressed, while Eric Uslaner (2004) has demonstrated the ways in
which general trust differs from social capital, the variant conceptualizations directly
affecting the results regarding the effect of diversity. In a recent review of the ‘hunkering
down’ thesis, Alejandro Portes and Erik Vickstrom (2011) assert that the key for under-
standing the inconsistent empirical findings regarding the relations between diversity and
social cohesion are the conceptual and methodological gaps in the ways social cohesion is
defined and measured.
This article does not attempt to decipher the link between diversity and social cohesion
or to offer a new conceptualization and operationalization for social cohesion – or even
claim to solve the methodological shortcomings of previous studies. Its principal aim is to
demonstrate how the relations between diversity and social cohesion vary according to the
dimensions of social cohesion examined and the indexes employed for their measurement.
I demonstrate that the relationship between diversity and social cohesion derived from a
single data set depends on the specific definition of social cohesion adopted.Thus while one
dimension of social cohesion may not be related to diversity, other aspects may exhibit
either negative or positive relations.
In the first section, I review Putnam’s ‘hunkering down’ thesis and the empirical studies
that have examined the relations between diversity and social cohesion on the community
and national level. This survey illustrates that the inconsistent relations between diversity
and social cohesion appear to be a function of the dimensions of social cohesion selected
by each individual study. I then discuss, in brief, the multidimensional nature of social
cohesion, with the empirical illustration based on international survey data to allow for the
measurement of diversity at the country level. Using data from the EuropeanValues Survey
2008 across 42 countries, I demonstrate how some forms of diversity are unrelated to
measures of general trust while other dimensions of social cohesion do evince a link with
diversity. Thus, for example, while ethnic and linguistic fractionalization at the country level
is negatively related to solidarity measured as concern for disadvantaged groups in one’s
country, it is unrelated to other measures of solidarity. Similarly, ethnic and linguistic
fractionalization at the country level was found to be negatively related to one measure of
belonging but not to the other. In the conclusion, I analyze these findings in order to
critique the ongoing discourse regarding diversity and social cohesion.

Putnam’s ‘Hunkering Down’ Hypothesis

Robert Putnam’s studies during the 1990s and 2000s accentuated the importance of
social capital – and in particular social trust – for maintaining prosperous societies,
proposing that ‘social networks and the associated norms of reciprocity and trustworthi-
ness’ (Putnam, 2007, p. 137) constitute facilitators of peaceful collective action, inclusive-
ness, tolerance, confidence in institutions and political participation (Putnam, 2000). Social
capital research has paid especial attention to interpersonal trust, which, it is argued,
comprises an important dimension of social capital. As Uslaner (2002, p. 1) asserts, trust is
the ‘chicken soup of moral life’. In light of this circumstance, numerous studies have
examined the factors shaping social capital, and social trust in particular (for a review, see
Nannestad, 2008; Newton, 2007).
© 2013 The Author. Political Studies © 2013 Political Studies Association
POLITICAL STUDIES: 2014, 62(3)
D O E S D I V E R S I T Y E RO D E S O C I A L C O H E S I O N ? 575

Among these, Putnam (2007) has suggested that diversity is strongly related to lower
levels of social capital.1 His ‘hunkering down’ hypothesis suggests that diversity has different
effects from those propounded in contact and conflict theories. In the short term, diversity
does not lead to greater inter-group contact between people from different backgrounds,
nor does it reinforce the solidarity of the in-group faced with the growing presence of
out-groups. Rather, Putnam argued, higher levels of diversity reduce both in-group and
out-group solidarity – a process related to changes in the sense of collective identity.
According to Putnam, one of the significant factors that shape social capital is related to
social identity. This, in turn, influences social distance – people’s sense of who they are.
Diversity impacts on the dynamic of social identity for both majority and minority groups
because it affects the sense and meaning of ‘we’ for both groups. While Putnam (2007, p.
159) acknowledges that ‘the linkage between identity and social capital is only beginning
to be explored’, he suggests that the ‘hunkering down’ effect derives from shifts in the sense
of social identity. Changes induced by increased immigration directly affect the social
identity of the receiving communities; in the short term diversity reduces the sense of
community.Thus, diversity ‘seems to trigger not in-group/out-group division, but anomie
or social isolation. In colloquial language, people living in ethnically diverse settings appear
to “hunker down” – that is, to pull in like a turtle’ (Putnam, 2007, p. 149, emphasis in
original). Evidence provided by the Social Capital Community Benchmark Survey sup-
ports this thesis. Inhabitants of diverse communities tend to withdraw from collective life,
distrust their neighbors and engage less in public life. They are also less likely to regard
themselves as possessing political efficacy, place less trust in their local leaders and volunteer/
give to charity less.
Further studies conducted in the US also corroborate Putnam’s thesis, with negative
relations appearing to obtain between diversity and various aspects of social cohesion. Using
longitudinal data at the state level, Daniel Hawes and Rene Rocha have found that, by and
large, a positive relation obtained between increasing minority diversity and lower levels of
social trust between 1985 and 2000 (Hawes and Rocha, 2011; see also Dincer, 2011). These
findings are consistent with the results of Rodney Hero’s (2007) study, which found that
racial and ethnic diversity in the United States is negatively related to social capital and civic
culture.
While Putnam’s work focuses on the concept and measures of social capital rather than
directly on social cohesion, social capital – and interpersonal trust in particular – are
frequently regarded as an important dimension of social cohesion, these two concepts at
times being considered synonymous (Harell and Stolle, 2010). As I address below, such a
narrow concentration on generalized trust ignores the various dimensions of social cohe-
sion and predetermines a specific set of conclusions regarding the relations between
diversity and social cohesion.

The ‘Hunkering Down’ Hypothesis and the Link between Diversity and
Social Cohesion: Contradictory Findings
While evidence from the US has provided support for the ‘hunkering down’ thesis, findings
from other countries have been far from unanimous. Single-country analyses which have
focused on diversity at the local level have evinced contrary results not only between
© 2013 The Author. Political Studies © 2013 Political Studies Association
POLITICAL STUDIES: 2014, 62(3)
576 G A L A R I E LY

countries but also between studies conducted within a single country. One study in the
Netherlands, for example, found that the effects of ethnic diversity on social cohesion vary
according to the dimensions used to measure social cohesion.While at the municipal level
diversity is negatively related to the propensity to volunteer, in ethnically heterogeneous
neighborhoods it is positively related to tolerance of different-race neighbors. At this level,
ethnic heterogeneity is also positively rather than negatively related to generalized trust
among residents with above-average educational levels (Tolsma et al., 2009). In contrast,
another Dutch study indicated that ethnically heterogeneous neighborhoods exert a nega-
tive effect on the quality of contact with neighbors. Significantly, ethnic diversity was not
related to other dimensions of trust in this study.As Bram Lancee and Jaap Dronkers (2011)
have demonstrated, the criterion of religious diversity in neighborhoods must also be
adduced, being negatively related among native Dutch residents to the quality of contact
with their neighbors, trust within the neighborhood and inter-ethnic trust.
While in Canada one study found that the more diverse the city, the higher the levels of
trust (Kazemipur, 2006), another indicates that heterogeneity is also positively related to a
greater sense of belonging (Wu et al., 2011). A comparative study between the US and
Canada, on the other hand, confirms that the negative effect of diversity on trust is similar
in both Canada and the US (Stolle et al., 2008). Although a study conducted among
Flemish municipalities found negative relations between different measures of social capital
and the number of nationalities present, in this case, too, these relations are not necessarily
related to ethno-cultural differences but can also be explained by other social and economic
components (Coffé and Geys, 2006). Another study in Belgium focused on schools,
examining the link between diversity and the pupils’ national (Belgian) and sub-national
(Flemish) identification (Agirdag et al., 2011). Its results demonstrate that ethnic hetero-
geneity within the school is negatively associated with the pupils’ national and sub-national
identification. Conversely, however, ethnic diversity at the school level in Denmark is not
negatively related to social cohesion but rather exerts a positive impact on trust among
native Danes (Dinesen, 2011).
In Britain, while some studies found no relation between diversity and social capital
(Sturgis et al., 2011; Taylor et al., 2010), one indicated that community heterogeneity is
related to a single aspect of social capital alone – namely, neighborhood attitudes – and
unrelated to other dimensions such as formal and informal interactions (Letki, 2008).The
latter study also suggested that social capital is affected by neighborhood status rather than
diversity.While another study in Britain employed expansive definitions of social cohesion
and indexes similar to Putnam’s social capital, it also examined inter-ethnic relations
(Laurence, 2011), its findings demonstrating that, although greater diversity does have a
negative impact on social capital, it also has a positive impact on inter-ethnic relations. In
fact, a higher level of inter-ethnic relations may reduce the negative impact of diversity on
social capital. A comparative US/British study suggested that the proper factor to be
adduced is segregation rather than heterogeneity, with residents of integrated and diverse
communities with varied social networks more likely to trust others (Uslaner, 2012; see also
Fieldhouse and Cutts, 2010).
The contrary results from the single-case studies conducted at the local level also occur
in studies examining diversity at the country level. Unlike those which investigated other
© 2013 The Author. Political Studies © 2013 Political Studies Association
POLITICAL STUDIES: 2014, 62(3)
D O E S D I V E R S I T Y E RO D E S O C I A L C O H E S I O N ? 577

dimensions of social cohesion, studies focusing primarily on social trust found no evidence
that heterogeneity reduces trust. While Hooghe et al.’s (2009) multi-level analysis of data
from the European Social Survey (ESS) across twenty European countries examined the
effect of 26 different measurements of diversity on interpersonal trust, their overall analysis
found no indication that the two phenomena are related. Another study (Gerritsen and
Lubbers, 2010) employing the European Social Survey data focused specifically on trust in
the European Union and candidate populations, together with trust in members of their
own country. While heterogeneity was not related in any way to trust in their country
members, a positive relationship did obtain between ethnic fractionalization and trust in the
European Union. Using Eurobarometer as their source of data, Maurice Gesthuizen et al.
(2009) analyzed both social trust and other dimensions such as social ties and volunteering
across 28 European countries. Here too, however, diversity appears to be unrelated to these
dimensions of social cohesion.
In line with the latter three studies, on the basis of the European Social Survey Henrik
Lolle and Lars Torpe (2011) found that, while ethnic diversity at the country level is
unrelated to trust, a negative relation between diversity and trust does obtain at the local
level. Michael Savelkoul et al. (2011) used the same data set to examine regional and
country-level effects of diversity, focusing on various aspects of social capital such as
informal social capital and the density, strength and extensiveness of social networks.
The results of this study indicate that while a direct positive effect obtains between ethnic
diversity and informal aid at the country level, at the regional level the effect of
ethnic diversity is indirect. Another study across European countries has found that
immigrant diversity is not related to generalized trust, association membership or voluntary
involvement (Reeskens and Wright, 2013). Instead, their results suggest that ethnic identity
reduces social cohesion.While these studies used a multi-level design to assess the linkage
between attitudes of individuals and the country or regional level of diversity in Europe,
Ming-Chang Tsai et al. (2010) confined their analysis to the country level. Using various
cross-national surveys, Tsai et al. employed aggregate data regarding the levels of inter-
personal trust. Using the question: ‘In general, do you think that most people can be
trusted?’, they found no relation between diversity and trust across 98 countries (see also
Bjørnskov, 2007). Analyzing 80 countries, Jong-sung You (2012) also found that diversity
is not significantly related to social trust when fairness variables – lack of corruption,
income equality, mature democracy – are included in the analysis.
While these cross-national studies demonstrate limited support for the negative effect of
diversity, the findings from other studies differ. Using the World Values Survey to analyze
sources of trust across 60 countries, Jan Delhey and Kenneth Newton (2005) found that
ethnic heterogeneity is negatively related to trust, while Birte Gundelach (2013) used the
later version of the World Values Survey and finds that diversity is positively related to
out-group trust.Another study (Anderson and Paskeviciute, 2006) employed the same data
set to observe social trust, as well as political and civic engagement, their results demon-
strating a more complex relation in which, while ethnic heterogeneity decreases levels of
trust in established democracies, in less democratic countries the principal cause is linguistic
rather than ethnic heterogeneity. Analyzing trust across 115 countries, another study found
that, while no general relationship exists between ethnic fractionalization and trust, negative
© 2013 The Author. Political Studies © 2013 Political Studies Association
POLITICAL STUDIES: 2014, 62(3)
578 G A L A R I E LY

relations obtain between low levels of ethno-religious cross-cutting and cross-

fractionalization types of diversity (Finseraas and Jakobsson, 2012).
In contrast to these two cross-sectional design studies, Christel Kesler and Irene
Bloemraad (2010) – who also used the World Values Survey – employed a cross-sectional
time-series design in order to follow changes in nineteen countries between 1981 and
2000. Looking not only at social trust but also at political and civic engagement, their
findings indicate that increasing diversity as a result of immigration decreases trust, civic
engagement and political participation in some – but not all – advanced democracies. In
those countries with a lower level of income inequality and multicultural policies, the
negative effects of diversity disappear (see also Kumlin and Rothstein, 2010, for similar
findings in Sweden). Another study examining the effect of integration policy on trust
across Europe reached a different conclusion, however, implying that the weak negative
effect of diversity on trust is unaffected by more open integration policies (Reeskens, 2010).
In summary, the macro-level studies of the relationship between levels of heterogeneity
and social cohesion conducted to date have failed to yield robust conclusions.The question
I address here is: what explains this discrepancy in the empirical literature? Different studies
employ disparate definitions and measures for outcomes generally considered to constitute
indicators of ‘social cohesion’ or ‘social capital’, some focusing solely on interpersonal trust;
others include such aspects as civic and political engagement, attitudes toward neighbors,
inter-ethnic relations or the sense of belonging. In light of the fact that a multitude of
definitions of social cohesion exist, it is surely no wonder that contrary conclusions have
been drawn regarding the impact of diversity. Any attempt to examine the effects of
heterogeneity on social cohesion must therefore recognize the multiple dimensions of social
cohesion.

Social Cohesion: A Multidimensional Concept

While social cohesion currently possesses a high profile in scholarly research and policy, it
remains a vague concept. Despite numerous attempts to define and measure the phenom-
enon, little consensus exists regarding its nature and the best way to measure it (Botterman
et al., 2012; Chan et al., 2006; Dickes et al., 2010; Hulse and Stone, 2007; Jenson, 2010).
Harell and Stolle (2010, p. 247), for example, assert that it ‘generally refers to belonging to a
political community, both in terms of socio-economic inclusion as well as attachment and
involvement within the larger society and its values’. Joseph Chan and his colleagues, on the
other hand, define social cohesion as ‘a state of affairs concerning both the vertical and the
horizontal interactions among members of society as characterized by a set of attitudes and
norms that includes trust, a sense of belonging and the willingness to participate and help, as
well as their behavioral manifestations’ (Chan et al., 2006, p. 290). A review of the various
definitions proposed reveals that the approach to social cohesion differs not only between
different scholars and disciplines but also according to context.Thus while in Canada, for
example, the sense of belonging to the community is accentuated, in the European context
social cohesion is understood as a process directed at social inclusion (Vergolini, 2011).To
date, the sole consensual recognition is that social cohesion is a multidimensional concept.
Various aspects of social cohesion have been propounded. In a recent study, Jene Jenson
(2010) adduces five components for measuring social cohesion: (a) the legitimacy of public
© 2013 The Author. Political Studies © 2013 Political Studies Association
POLITICAL STUDIES: 2014, 62(3)
D O E S D I V E R S I T Y E RO D E S O C I A L C O H E S I O N ? 579

and private institutions; (b) inclusion in the economic sphere; (c) involvement in political
and civic spheres; (d) tolerance of diversity; and (e) the feeling of belonging to the same
community. Other scholars have suggested such factors as shared norms, common bonding,
‘togetherness’, social order and social solidarity. Likewise, while social capital is sometimes
understood to be synonymous with social cohesion, some writers regard it as simply one
facet of social cohesion (Laurence, 2011). Frequently, the indexes used to measure social
capital – such as interpersonal trust and social networks – are also employed to measure
social cohesion (for reviews, see Berger-Schmitt, 2000; Chan et al., 2006).
Although attempts have been made to measure social cohesion, the proper method for
measuring it remains under dispute.While some scholars argue that, being an attribute of
a group, social cohesion should be measured at the group level, others maintain that it is best
measured at the individual level (Vergolini, 2011). Disagreement also exists regarding the
indexes appropriate for its measurement. Paul Dickes et al. (2010), for example, developed
a single index of social cohesion at the country level based on aggregate survey data from
the European Values Survey.Although this includes scales that relate to social capital – such
as voluntary participation – interpersonal trust was not included, while elements such as
political engagement, confidence in institutions, satisfaction with democracy and solidarity
with fellow country members were incorporated. A similar study was conducted by Loris
Vergolini (2011), who employed the European Social Survey to measure social cohesion
and included confidence in institutions, interpersonal trust and political and civic engage-
ment but not solidarity. Other research has emphasized further aspects of social cohesion,
such as the reciprocity and recognition of diversity itself or the existence of harmonized
inter-ethnic relations, without summarizing these in a single index (Laurence, 2011). Sarah
Botterman et al. (2012) even argue that the multidimensionality of social cohesion under-
mines the ability to offer a single index.
In this article I do not presume to suggest a comprehensive definition or measure of
social cohesion, discuss its affinities and differences from social capital (see Portes and
Vickstrom, 2011) or even attempt to determine which dimensions form part of social
cohesion and which do not. My concern is rather to demonstrate how measuring variant
aspects of social cohesion leads to divergent results regarding the effects of diversity. I focus
on three elements frequently adduced as elements of social cohesion: interpersonal trust,
belonging and solidarity.

Data
The survey data were drawn from the fourth wave of the European Values Study 2008
(EVS).2 The EVS 2008’s primary advantage is that it contains items that enable the
measurement of various dimensions of social cohesion and covers a relatively large number
of European countries, thereby giving variation at the state level and facilitating a multi-level
analysis.The 42 countries from the EVS 2008 included in this analysis are:Albania,Armenia,
Austria, Azerbaijan, Belarus, Belgium, Bosnia-Herzegovina, Bulgaria, Croatia, Cyprus,
Czech Republic, Denmark, Estonia, Finland, France, Georgia, Germany, Great Britain,
Greece, Hungary, Iceland, Ireland, Italy, Latvia, Lithuania, Luxembourg, Malta, Macedonia,
Republic of Moldova, the Netherlands, Norway, Poland, Portugal, Romania, Russian
Federation, Slovak Republic, Slovenia, Spain, Sweden, Switzerland,Turkey and Ukraine.3
© 2013 The Author. Political Studies © 2013 Political Studies Association
POLITICAL STUDIES: 2014, 62(3)
580 G A L A R I E LY

Individual-Level Measures
Interpersonal Trust
Interpersonal trust was measured by two separate items.The first is formulated in the classic
question: ‘Generally speaking, would you say that most people can be trusted or that you
can’t be too careful in dealing with people?’ While recent studies have indicated that the
meaning of ‘most people’ varies across countries, this remains the most common way to
measure interpersonal trust in cross-national surveys (Delhey et al., 2011).The second was
measured by the question:‘Most people try to take advantage of you or try to be fair’.The
scales are summarized in Table 1.

Belonging
The EVS offers two sets of questions out of various possibilities relating to the measurement
of belonging and attachment. The first item measures the sense of belonging by asking
respondents how proud they are of their country. It is reasonable to wonder to what extent
national pride reflects the sense of belonging to one’s country, rather than nationalism.
Unfortunately, while items that explicitly ask respondents about their level of belonging to
their country are better indicators, none such features in the EVS 2008. It is also plausible
to assume that people with a strong sense of belonging will also tend to be proud of their
country. Regarded as the expression of an enduring bond with the political community, this
item is commonly used in cross-national surveys to measure the most basic level of system
support (Dalton, 2004; Norris, 2011).
The second item asked respondents:‘Which of these geographical groups would you say
you belong to first of all?’, with respondents required to choose from five options: locality
or town, region of country, country, Europe or the world. The answer was recoded as a
dummy variable in order to mark country attachment. It is important to emphasize that
while these two measures only relate to part of the broader categories of attachment and
belonging (Sinnott, 2006), these are the items available in the EVS.

Solidarity
The EVS contains a set of items asking respondents about their concern for the living
conditions of people in their community and for disadvantaged groups in their country,
such as the sick or unemployed, with these items commonly used as indicators of welfare
attitudes in Europe (Van Oorschot et al., 2005). An exploratory factor analysis leads to
the same two-factor solutions across virtually all of the 42 countries. The first factor
reflects care for the neighborhood, the region and one’s fellow countrymen, and the
second reflects concern for the disadvantaged – the two thus signifying solidarity with
the community as a whole and with the neediest groups, respectively. A confirmatory
factor analysis indicates that intensive cross-loadings exist between the factors and that
measurement equivalence cannot be established across all 42 countries. Nevertheless, the
results of the exploratory factor analysis demonstrate that the scales should not be sum-
marized in a single index.4 In addition to the dimensions of social cohesion the analysis
also controlled for socio-demographic variables: age, gender, education, income and
religiosity.5
© 2013 The Author. Political Studies © 2013 Political Studies Association
POLITICAL STUDIES: 2014, 62(3)
D O E S D I V E R S I T Y E RO D E S O C I A L C O H E S I O N ? 581

Table 1: Measuring Dimensions of Social Cohesion

Mean
Variables Items Scale (SD)

Interpersonal Trust dummy Generally speaking, would 1 = Most people 0.31

trust you say that most can be trusted (0.46)
people can be trusted
or that you can’t be too
careful in dealing with
people? (V62)
Trust scale Most people try to take 1–10 scale 10 = 5.44
advantage of you or try Try to be fair (2.53)
to be fair (V63)
Belonging Patriotism How proud are you to be 1–4 scale 3.32
a [nationality]? (V256) 4 = Very proud (0.76)
Country attachment Which of these Dummy variable 0.35
geographical groups 1 = Country (0.47)
would you say you
belong to first of all?
(V253)
Solidarity General concern Concern for the living 1–5 scale 3.04
conditions of: People 5 = Very much (0.94)
from your concern
neighborhood (V285),
People from your
region (V286), Fellow
countrymen (V287)
Concern for the Elderly people in your 3.64
disadvantaged country (V290), Sick (0.89)
and disabled people in
your country (V291),
Unemployed people in
country (V292)

Source: EVS 2008.

It is important to emphasize that this operationalization for social cohesion cannot avoid
the vagueness of the concept itself.While interpersonal trust is commonly used as a measure
of social cohesion, the measures for belonging and solidarity are less clearly related to social
cohesion. While national pride, for example, may be viewed as reflecting a sense of
belonging to one’s country it might also reflect other dimensions of attachment, such as
nationalism (De Figueiredo and Elkins, 2003).The same claim may be made with regard to
the solidarity scale – which might reflect economic ideology rather than aspects of social
cohesion. Such conceptual vagueness is evident in previous studies employing the EVS or
ESS for measuring social cohesion. Dickes et al. (2010), for example, used, inter alia, the
© 2013 The Author. Political Studies © 2013 Political Studies Association
POLITICAL STUDIES: 2014, 62(3)
582 G A L A R I E LY

items for solidarity but not for social trust, while Vergolini (2011) included interpersonal
trust but not solidarity. The use of secondary data limits the capacity of any study –
including the current one – to address this problematic.

Country-Level Measures
In this study, rather than attempting to assess all the possible forms of diversity, I followed
common practice in measuring for static diversity and changes in diversity. The static
measures are based on the prevalent indexes of ethnic and linguistic fractionalization
(Alesina et al., 2003), with fractionalization understood as the probability that two randomly
selected individuals from a single population belong to different groups.The fractionaliza-
tion index reflects diversity due not only to immigration but also to the country’s historical
ethnic and linguistic background, Switzerland being an excellent case in point. Data
concerning ethnic and linguistic fractionalization were obtained from Jan Teorell et al.
(2011). In addition I also examined the relative number of foreign-born residents in each
country (in relation to the total population) for the year 2010 using data obtained from the
World Bank. While this measure reflects diversity deriving from immigration (such as
immigrants in Germany and France) it also reflects other sources of foreign-born popula-
tions (such as residents of Russian origin in Estonia). Based on the World Bank data, I
constructed an index for changes in diversity examining shifts in the percentage of
foreign-born residents between 1995 and 2010. I also used an alternative time lag (2000 to
2010) in the analysis.
Alongside these direct measures of diversity I also controlled for alternative explanations
of social cohesion. One of the most common factors adduced for social cohesion – and
interpersonal trust in particular – is income inequality, higher levels of which are custom-
arily regarded as linked to lower levels of trust (Rothstein and Uslaner, 2005). Since the
Standardized World Income Inequality Database (Solt, 2009) maximizes the comparability
of income inequality data, I used these data to measure net income inequality (GINI index
net). Another explanation for social cohesion frequently appealed to in order to explain
trust is economic development, with wealthier societies considered to exhibit higher levels
of trust (Delhey and Newton, 2005). In this case, I used the GDP per capita from the World
Bank Data to measure economic development (average for 2000–5).As several studies have
indicated that some aspects of social cohesion – such as interpersonal trust – are lower in
post-communist countries (e.g., Dowley and Silver, 2002), a dummy variable to control for
their impact was constructed.
The correlation between the country-level variables suggests that some overlap obtains
between the variables. First, although ethnic and linguistic fractionalization are highly
correlated with one another (r = 0.82), Christopher Anderson and Aida Paskeviciute’s
(2006) findings regarding their different effects led us to analyze them separately. Second,
the more ethnically diverse countries are also those not faced with an increase in the
numbers of foreign-born residents between 1995 and 2010.The strong negative correlation
(r = −0.63) between immigration increase and post-communist countries illustrates that
Western Europe constituted the primary immigrant destination. Third, Table 2 indicates
that no overlap exists between ethnic/linguistic fractionalization and the relative number of
foreign-born residents; thus these two factors appear to comprise different measures of
© 2013 The Author. Political Studies © 2013 Political Studies Association
POLITICAL STUDIES: 2014, 62(3)
 Table 2: Means and Correlation of Country-Level Variables

POLITICAL STUDIES: 2014, 62(3)

Mean Ethnic Linguistic Foreign Foreign GINI GDP
(SD) fractionalization fractionalization born change born (net) per capita

1. Ethnic fractionalization 0.26

(0.18)

© 2013 The Author. Political Studies © 2013 Political Studies Association

2. Linguistic fractionalization 0.26 0.82***
(0.2)
3. Foreign born change 0.05 −0.45** −0.39**
(0.54)
4. Foreign born 0.09 0.14 0.26 0.24
(0.06)
D O E S D I V E R S I T Y E RO D E S O C I A L C O H E S I O N ?

5. GINI (net) 30.63 0.29 0.14 −0.23 −0.06

(5.77)
6. GDP per capita 22,562.93 −0.28 −0.08 0.52*** 0.65** −0.41**
(13,868.61)
7. Post-communist 0.49** 0.29 −0.63*** −0.36* 0.19 −0.73**

Notes: *p < 0.05; ** p < 0.01; ***<0.001 level. N = 42.

583
584 G A L A R I E LY

diversity. Fourth, some overlap obtains between the control variables: the post-communist
countries have lower levels of economic development, those with higher levels of economic
development also possessing lower levels of income inequality.

Results
Using the pooled sample of the 42 countries, multi-level regression models for the various
components of social cohesion were constructed. Such an approach enabled an examination
of the direct effects of country-level diversity indexes on the outcome variables (at the
individual level). All the models controlled for basic socio-demographic variables – age,
gender, education, income and religiosity – relevant to the outcome variables. This facili-
tated a more refined analysis of the effects of the country-level variables. In addition to
controlling the individual-level variables, the investigations also controlled for alternative
explanations of social cohesion: income inequality, economic development and post-
communist countries.While multi-level models are able to account for country-level effects
on the outcome variables at the individual level, the correlations between the country-level
variables should be considered. As Table 2 indicates, strong correlations exist among the
country-level variables, despite the sample only including 42 countries. The effects of the
country-level variables were thus examined using different models.While sufficient varia-
tion for the country-level variables exists, the analysis should also take into account the
possibility of bias within specific countries. The analysis was therefore replicated after
excluding outliers.
Table 3 presents the results regarding interpersonal trust. The trust scale measured with
the dummy variable was analyzed via logistic regression, the alternative trust scale via linear
regression. For simplicity, only the results for the country-level variables are presented.The
results from eight models in Table 3 indicate that, at the country level, no significant
relations obtain between general trust and the four diversity measures.The following sets of
models examined two indexes regarding belonging. The first set addressed people’s iden-
tification with their country – described in Table 4 and analyzed via logistic regression. For
‘belonging to the country’, three indexes of diversity were found to be insignificant (models
4.1, 4.3 and 4.4). Although linguistic fractionalization (model 4.2) was found to be
positively related, the level of significance was marginal ( p = 0.078), deriving from an outlier
(Georgia). If Georgia is excluded from the analysis, the relations become insignificant.
In contrast, the second measure of belonging was found to be related to diversity in the
linear multi-level regression.6 Models 5.1 and 5.2 (see Table 5a) indicate that ethnic and
linguistic fractionalization exert a negative effect on the level of national pride.These effects
are significant even when alternative country-level variables are controlled for (models
5.5–5.10).7 The finding of significant relations does not give us very much information,
however, concerning the contribution of ethnic and linguistic fractionalization to explaining
national pride.This should take the explained variance into consideration. Further support
for the effects of ethnic and linguistic fractionalization is evident when their relative
contribution to the reduction in the country-level unexplained variance vis-à-vis other
factors is examined. Overall, the analysis suggests that ethnic and linguistic fractionalization
contribute to a greater lowering of the levels of unexplained variance (at the country level)
than do economic development, rates of inequality or post-communist legacy.
© 2013 The Author. Political Studies © 2013 Political Studies Association
POLITICAL STUDIES: 2014, 62(3)
POLITICAL STUDIES: 2014, 62(3)
Table 3: Multi-level Analysis for Interpersonal Trust

Trust dummya Trust scale

© 2013 The Author. Political Studies © 2013 Political Studies Association

Models 3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.8

Ethnic fractionalization −0.34 (0.47) −0.10 (0.09)

Linguistic fractionalization 0.31 (0.40) 0.03 (0.08)
Foreign born −0.03 (1.39) −0.12 (0.24)
Foreign born change −0.22 (0.14) −0.03 (0.02)
D O E S D I V E R S I T Y E RO D E S O C I A L C O H E S I O N ?

Notes: The models include individual-level variables: Age, Gender, Income, Education and Religiosity. For variable scaling, see Table 1. Robust standard errors in parentheses. aDummy variable
outcome was analyzed as logistic regression with logit link function (unit-specific model). N = 42,265, countries = 42.
Source: EVS 2008.
585
586 G A L A R I E LY

Table 4: Multi-level Analysis for Belonging to the Country

Models 4.1 4.2 4.3 4.4

Ethnic fractionalization 0.38 (0.41)

Linguistic fractionalization 0.57 (0.31)∧
Foreign born −0.62 (10.11)
Foreign born change −0.17 (0.14)

Notes: The models include individual-level variables: Age, Gender, Income, Education and Religiosity. For variable scaling, see
Table 1. Dummy variable outcome was analyzed as logistic regression with logit link function (unit-specific model). ∧p < 0.1. N =
42,265, countries = 42.
Source: EVS 2008.

Figure 1 illustrates these relations at the aggregate level.The more diverse countries are
also those with low levels of an aggregate sense of national pride, this pattern cutting across
regional differences and historical factors. Thus some homogenized Eastern European
countries – such as Poland and Slovenia – have high levels of national pride while such
diverse Western European countries as Switzerland and Belgium exhibit low levels of
national pride. Figure 1 also implies that although Bosnia-Herzegovina might be an outlier,
replication of the analysis after its exclusion from the analysis leads to similar results.
While ethnic and linguistic fractionalization are negatively related to national pride, a
higher level of foreign-born residents is positively related to national pride (model 5.4).
These relations are positive and significant even after the other three country-level expla-
nations have been controlled for (models 5.11–5.13) and the outlier (Bosnia-Herzegovina)
excluded from the analysis. Furthermore, additional analysis (not reported in Table 5)
indicated that when ethnic and linguistic fractionalization are both included in the model,
these positive relations hold while negative relations obtain with the fractionalization
indexes. This pattern holds true even when the alternative country-level explanations are
included in the models. As the relative contribution of an increased level of foreign-born
residents to the reduction in the country-level unexplained variance is higher than the
other three country-level variables, one can conclude that, on average, in countries where
the percentage of foreign-born residents rose between 1995 and 2010, higher levels of
national pride also existed.
The final section of the multi-level regression analysis – based on linear models –focused
on the relationship between diversity and solidarity. The first four models in Table 6
(models 6.1–6.4) indicate that diversity is not related to general concern regarding the
living conditions of people from one’s neighborhood, region or country. With respect to
concern for the disadvantaged, however, model 6.5 demonstrates that ethnic fractionaliza-
tion is negatively related at a merely marginal level of significance, with linguistic fraction-
alization negatively related (model 6.6).These relations hold for linguistic fractionalization
even after the other three country-level variables are controlled for (models 6.9–6.11).
Thus in similar fashion to national pride, the country-level variables reduce the explained
variance to a greater degree than the control variables.
© 2013 The Author. Political Studies © 2013 Political Studies Association
POLITICAL STUDIES: 2014, 62(3)
POLITICAL STUDIES: 2014, 62(3)
Table 5a: Multi-level Analysis for National Pride

Models 5.1 5.2 5.3 5.4 5.5 5.6 5.7

Ethnic fractionalization −0.59 (0.16)*** −0.66 (0.16)*** −0.56 (0.14)*** −0.47 (0.12)**

© 2013 The Author. Political Studies © 2013 Political Studies Association

Linguistic fractionalization −0.56 (0.15)**
Foreign born 0.49 (0.46)
Foreign born change 0.23 (0.05)***
GINI index −0.11 (0.00)**
GDP per capita 0.04 (0.02)∧
Post-communist countries −0.18 (0.05)**
D O E S D I V E R S I T Y E RO D E S O C I A L C O H E S I O N ?

Country-level variance 0.042 0.042 0.047 0.036 0.040 0.043 0.043

Reduction in the country-level variance 8.6% 8.6% 0 21.7% 13% 6.5% 6.5%

Notes: The models include individual-level variables: Age, Gender, Income, Education and Religiosity. For variable scaling, see Table 1. Robust Standard errors in parentheses. ∧p < 0.1;
*p < 0.05; **p < 0.01; ***< 0.001. N = 42,265, countries = 42.
Source: EVS 2008.
587
 588

Table 5b: Multi-level Analysis for National Pride

Models 5.8 5.9 5.10 5.11 5.12 5.13

Linguistic fractionalization −0.60 (0.15)*** −0.58 (0.12)** −0.48 (0.14)**

Foreign born change 0.23 (0.05)*** 0.18 (0.07)* 0.25 (0.04)***
GINI index 0.01(0.00)* 0.00 (0.02)
GDP per capita 0.05 (0.02)* 11 (0.06)∧
Post-communist countries −00.20 (0.05)** 0.01 (0.00)
Country-level variance 0.041 0.043 0.042 0.035 0.037 0.038
Reduction in country-level variance 10.8% 6.4% 8.6% 23.9% 19.5% 17.3%

Notes: The models include individual-level variables: Age, Gender, Income, Education and Religiosity. For variable scaling, see Table 1. Robust Standard errors in parentheses. ∧p < 0.1;
*p < 0.05; **p < 0.01; *** < 0.001. N = 42,265, countries = 42.
Source: EVS 2008.
G A L A R I E LY

POLITICAL STUDIES: 2014, 62(3)

© 2013 The Author. Political Studies © 2013 Political Studies Association
D O E S D I V E R S I T Y E RO D E S O C I A L C O H E S I O N ? 589

Figure 1: Plot for Ethnic Fractionalization and Aggregate Level of National Pride

Figure 2 illustrates the fact that more linguistically diverse countries also exhibit a lower
level of aggregate concern for disadvantaged groups.While high levels of concern for the
disadvantaged occur in the homogenized Scandinavian countries these also appear in
homogenized post-communist states such as Armenia and Albania, while lower levels of
concern appear in such multilingual countries as Belgium and Luxembourg. While
Figure 2 also implies that Luxembourg, Latvia and Turkey might be outliers, replication of
the analysis after excluding these three countries leads to similar results. I can thus conclude
that, on average, countries with higher levels of linguistic fractionalization also display lower
levels of care for disadvantaged groups.

Conclusions and Implications

The fragmentation of societies in the face of increasing immigration, and the ethnic and
linguistic diversity to which the latter gives rise, are a source of concern for many scholars
and pundits.The view that such heterogeneity reduces social cohesion and leads people to
‘hunker down’ and withdraw from public life is disputed, the theoretical debate spilling over
into the arena of policy making and the public sphere, where the claim that diversity
© 2013 The Author. Political Studies © 2013 Political Studies Association
POLITICAL STUDIES: 2014, 62(3)
 590

Table 6: Multi-level Analysis for Solidarity

General concern Concern for the disadvantaged

Models 6.1 6.2 6.3 6.4 6.5 6.6 6.7 6.8 6.9 6.10 6.11

Ethnic fractionalization 0.18 (0.19) −0.32 (0.17)∧

Linguistic fractionalization 0.06 (0.16) −0.42 (0.13)** −0.48 (0.12)** −0.43 (0.14)** −0.36 (0.13)**
Foreign born −0.03 (0.50) −0.59 (0.40)
Foreign born change −0.03 (0.06) 0.10 (0.06)
GINI index −0.19 (0.00)*
GDP per capita 0.03 (0.02)
Post-communist countries 0.08 (0.06)
Country-level variance 0.102 0.101 0.101 0.100 0.078 0.074 0.081 0.075 0.066 0.076 0.072
Reduction in country-level 0 0 0 0 2.5% 7.5% 0 6.2% 17.5% 5% 10%
variance

Notes: The models include individual-level variables: Age, Gender, Income, Education and Religiosity. For variable scaling, see Table 1. Robust Standard errors in parentheses. ∧p < 0.1; *p < 0.05; ** p < 0.01; ***<0.001.
N = 42,265, countries = 42.
Source: EVS 2008.
G A L A R I E LY

POLITICAL STUDIES: 2014, 62(3)

© 2013 The Author. Political Studies © 2013 Political Studies Association
D O E S D I V E R S I T Y E RO D E S O C I A L C O H E S I O N ? 591

Figure 2: Plot for Linguistic Fractionalization and Aggregate Level of Concern for
the Disadvantaged

reduces social cohesion forms the justification on occasion for a restriction on immigration
(Holtug, 2010; Portes and Vickstrom, 2011). Despite the popularity of the concept of social
cohesion, the results from the numerous studies addressing this issue are inconsistent. Some
findings support the ‘hunkering down’ thesis while others do not. Making no boast to
provide empirical evidence in favor of or in dismissal of this theory, this contribution
endeavors to demonstrate how the relations between diversity and social cohesion depend
on which aspects of social cohesion are selected for analysis and which indexes are adopted
to measure them.
Analyzing cross-national survey data from 42 countries, I found that diversity exhibits
dissimilar relations with different dimensions and operationalizations of social cohesion.
How should these results be interpreted? Do they support the claim that diversity is indeed
related to lower levels of social cohesion? If the focus is placed on interpersonal trust, the
results match those of previous studies examining diversity at the state level across European
© 2013 The Author. Political Studies © 2013 Political Studies Association
POLITICAL STUDIES: 2014, 62(3)
592 G A L A R I E LY

countries (Gerritsen and Lubbers, 2010; Gesthuizen et al., 2009; Hooghe et al., 2009; Lolle
and Torpe, 2011). No link obtains between the four indexes of diversity and social trust.
Observation of other dimensions of social cohesion, however, suggests a more complex
picture. Here, the relations between belonging and diversity diverge when dissimilar ways
are employed to measure the sense of belonging, the relations between diversity and
solidarity also changing according to the manner in which solidarity is measured. If this
dimension is measured by items reflecting general concern with the living conditions of
other members of the community, no relation with diversity obtains. If solidarity is
measured by very similar items explicitly asking about the disadvantaged groups in one’s
country, negative relations obtain between linguistic diversity and solidarity. Despite the fact
that these two variables for solidarity closely correspond, using one of them alone might
lead to mistaken conclusions regarding the relationship between diversity and solidarity.
The findings also demonstrate that different indexes of diversity have a dissimilar impact.
While the more static indexes of ethnic and linguistic fractionalization are negatively related
to national pride, the more dynamic index – reflecting changes in the relative number of
foreign-born residents between 1995 and 2010 – is positively related to national pride.
In light of these results, one may speculate that further investigation of additional
dimensions of social cohesion – such as ethnic relations, civic norms or engagement in
public life – will similarly lead to inconsistent findings. It is important to note, however, that
even if one selects only a single dimension of social cohesion, inherent limitations exist
regarding the drawing of conclusions regarding the implications of diversity.
First, like most of the studies that address this issue at the community or country level of
analysis, this contribution constitutes a cross-sectional study and thus precludes asking
questions related to causal relations at the individual and macro levels.While the ‘hunkering
down’ thesis emphasizes the dynamic effect of diversity, this cannot be corroborated
without establishing casual relations.
Second, given that the debate over the relations between diversity and social cohesion
explicitly addresses the possibility that it is not diversity per se that reduces social capital but
rather segregation (Uslaner, 2012), the possibility of omitted variable bias should be taken
into account.While the analyses controlled for alternative explanations at the country level,
other factors affecting the outcomes might well exist (see, for example, You, 2012).
Third, an intrinsic gap exists between the individual-level measures employed and the
concept of social cohesion, as well as between the country-level indexes of diversity and the
concept of diversity itself. One can only assume that the survey questions regarding national
pride reflect a sense of belonging and that concern for disadvantaged groups reflects
solidarity across all 42 countries.Without further examination to validate this assumption,
one cannot be sure how narrow the gap is between the concept and the measures. This
problematic also applies to the measures of diversity.While I used the common indexes for
analyzing static ethnic and linguistic fractionalization, together with a more dynamic index
for immigration, previous studies conducted in single countries point to the limitation such
indexes face in fully reflecting the complex meaning of diversity (Laurence, 2011). The
validity and comparability of such scales therefore remains an open question.
Fourth, as Christian Bjørnskov (2007) has argued, the significance of ethnic diversity for
the issue of trust depends on which specific countries are included in the analysis. Can the
© 2013 The Author. Political Studies © 2013 Political Studies Association
POLITICAL STUDIES: 2014, 62(3)
D O E S D I V E R S I T Y E RO D E S O C I A L C O H E S I O N ? 593

results from this study be expected to hold true if other countries are included in the
analysis? Given the contrary findings of previous studies, this probability is not high.
Fifth, while the ‘hunkering down’ thesis was developed and tested primarily in within-
country studies, the present article only observes people clusters in states, not cities or
regions. Its findings thus cannot be used to validate or reject results from within-country
studies.
Together with the contrary results, these methodological shortcomings should alert us to
the dangers of drawing any conclusions regarding the relationship between diversity and
social cohesion.These findings highlight the concerns raised by Hooghe (2007), Harell and
Stolle (2010) and Portes and Vickstrom (2011) concerning the integration of the issue of
social cohesion into future research. In light of the increasing sense of foreboding regarding
the implications of diversity for social cohesion, I hope that this study demonstrates the
caution that should be exercised when the multifaceted concept of social cohesion is
employed in discussions of immigration and integration.
(Accepted: 1 November 2012)
(Published online: 13 September 2013)

About the Author

Gal Ariely is Assistant Professor in the Department of Politics and Government at Ben-Gurion University of the
Negev. He studies issues of citizenship and national identity in Israel and across countries. Gal Ariely, Department of
Politics and Government, Ben-Gurion University of the Negev, POB 653, Beer-Sheva, 84105 Israel; email: galariel
@bgu.ac.il

Notes
1 Putnam’s theory of the effect of diversity on social capital was in fact pre-empted in the economic literature, which drew attention
to the effects of ethnic diversity on social trust and economic growth (e.g. Zak and Knack, 2001).
2 European Values Study 2008: Integrated Dataset (EVS 2008) ZA 4800.
3 For further details, see: http://www.europeanvaluesstudy.eu/. Other countries included in the EVS 2008 were excluded from this
analysis due to the absence of country-level data.
4 In fact, not only does the Akaike Information Criterion (AIC) comparison between the one-factor model and two-factor model
indicate that the two-factor model fits the data better, but the global fits of the one-factor model are unacceptable.
5 Age was measured with V303, gender with V302, education with V336, income with V353 and religiosity with V109.
6 As the outcome variable ‘national pride’ is a 4-point scale skewed to ‘very proud’, the analysis was replicated with logistic regression
for an outcome variable recoded as a dummy variable. The results were similar.
7 Due to the strong correlations between the country-level control variables they were modeled separately.

References
Agirdag, O., Van Houtte, M. and Van Avermaet, P. (2011) ‘Ethnic School Context and the National and Sub-national
Identifications of Pupils’, Ethnic and Racial Studies, 34 (2), 357–78.
Alesina, A., Devleeschauwer, A., Easterly, W., Kurlat, S. and Wacziarg, R. (2003) ‘Fractionalization’, Journal of Economic Growth,
8 (2), 155–94.
Anderson, C. J. and Paskeviciute, A. (2006) ‘How Ethnic and Linguistic Heterogeneity Influence the Prospects for Civil
Society: A Comparative Study of Citizenship Behavior’, Journal of Politics, 68 (4), 783–802.
Berger-Schmitt, R. (2000) ‘Social Cohesion as an Aspect of the Quality of Societies: Concept and Measurement’. EuReporting
Working Paper No. 14, Centre for Survey Research and Methodology (ZUMA).
Bjørnskov, C. (2007) ‘Determinants of Generalized Trust: A Cross-Country Comparison’, Public Choice, 130 (1),
1–21.
Botterman, S., Hooghe, M. and Reeskens, T. (2012) ‘ “One Size Fits All”? An Empirical Study into the Multidimensionality
of Social Cohesion Indicators in Belgian Local Communities’, Urban Studies, 49 (1), 185–202.
Chan, J., To, H. P. and Chan, E. (2006) ‘Reconsidering Social Cohesion: Developing a Definition and Analytical Framework
for Empirical Research’, Social Indicators Research, 75 (2), 273–302.
© 2013 The Author. Political Studies © 2013 Political Studies Association
POLITICAL STUDIES: 2014, 62(3)
594 G A L A R I E LY

Coffé, H. and Geys, B. (2006) ‘Community Heterogeneity: A Burden for the Creation of Social Capital?’, Social Science
Quarterly, 87 (5), 1053–72.
Dalton, R. J. (2004) Democratic Challenges, Democratic Choices:The Erosion of Political Support in Advanced Industrial Democracies.
New York: Oxford University Press.
De Figueiredo, J. R. J. P. and Elkins, Z. (2003) ‘Are Patriots Bigots? An Inquiry into the Vices of In-Group Pride’, American
Journal of Political Science, 47 (1), 171–88.
Delhey, J. and Newton, K. (2005) ‘Predicting Cross-National Levels of Social Trust: Global Pattern or Nordic
Exceptionalism?’, European Sociological Review, 21 (4), 311–27.
Delhey, J., Newton, K. and Welzel, C. (2011) ‘How General is Trust in “Most People”? Solving the Radius of Trust Problem’,
American Sociological Review, 76 (5), 786–807.
Dickes, P., Valentova, M. and Borsenberger, M. (2010) ‘Construct Validation and Application of a Common Measure of
Social Cohesion in 33 European Countries’, Social Indicators Research, 98 (3), 451–73.
Dincer, O. C. (2011) ‘Ethnic Diversity and Trust’, Contemporary Economic Policy, 29 (2), 284–93.
Dinesen, P.T. (2011) ‘Me and Jasmina Down by the Schoolyard: An Analysis of the Impact of Ethnic Diversity in School on
the Trust of Schoolchildren’, Social Science Research, 40 (2), 572–85.
Dowley, K. M. and Silver, B. D. (2002) ‘Social Capital, Ethnicity and Support for Democracy in the Post-communist States’,
Europe-Asia Studies, 54 (4), 505–27.
Fieldhouse, E. and Cutts, D. (2010) ‘Does Diversity Damage Social Capital? A Comparative Study of Neighborhood Diversity
and Social Capital in the US and Britain’, Canadian Journal of Political Science, 43 (2), 289–318.
Finseraas, H. and Jakobsson, N. (2012) ‘Trust and Ethnic Fractionalization: The Importance of Religion as a Cross-Cutting
Dimension’, Kyklos, 65 (3), 327–39.
Gerritsen, D. and Lubbers, M. (2010) ‘Unknown is Unloved? Diversity and Inter-population Trust in Europe’, European Union
Politics, 11 (2), 267–87.
Gesthuizen, M., Van Der Meer, T. and Scheepers, P. (2009) ‘Ethnic Diversity and Social Capital in Europe:Tests of Putnam’s
Thesis in European Countries’, Scandinavian Political Studies, 32 (2), 121–42.
Gundelach, B. (2013) ‘In Diversity We Trust: The Positive Effect of Ethnic Diversity on Outgroup Trust’, Political Behavior,
Online First.
Harell, A. and Stolle, D. (2010) ‘Diversity and Democratic Politics:An Introduction’, Canadian Journal of Political Science, 43 (2),
235–56.
Hawes, D. P. and Rocha, R. R. (2011) ‘Social Capital, Racial Diversity, and Equity: Evaluating the Determinants of Equity
in the United States’, Political Research Quarterly, 64 (4), 924–37.
Hero, R. E. (2007) Racial Diversity and Social Capital. New York: Cambridge University Press.
Holtug, N. (2010) ‘Immigration and the Politics of Social Cohesion’, Ethnicities, 10 (4), 435–51.
Holtug, N. and Mason, A. (2010) ‘Introduction: Immigration, Diversity and Social Cohesion’, Ethnicities, 10 (4), 407–
14.
Hooghe, M. (2007) ‘Social Capital and Diversity Generalized Trust, Social Cohesion and Regimes of Diversity’, Canadian
Journal of Political Science, 40 (3), 709–32.
Hooghe, M., Reeskens, T., Stolle, D. and Trappers, A. (2009) ‘Ethnic Diversity and Generalized Trust in Europe’, Comparative
Political Studies, 42 (2), 198–223.
Hulse, D. K. and Stone, W. (2007) ‘Social Cohesion, Social Capital and Social Exclusion’, Policy Studies, 28 (2), 109–28.
Jenson, J. (2010) Defining and Measuring Social Cohesion. London: Commonwealth Secretariat.
Kazemipur, A. (2006) ‘A Canadian Exceptionalism? Trust and Diversity in Canadian Cities’, Journal of International Migration
and Integration, 7 (2), 219–40.
Kesler, C. and Bloemraad, I. (2010) ‘Does Immigration Erode Social Capital? The Conditional Effects of Immigration-
Generated Diversity on Trust, Membership, and Participation across 19 Countries, 1981–2000’, Canadian Journal of Political
Science, 43 (2), 319–47.
Kumlin, S. and Rothstein, B. (2010) ‘Questioning the New Liberal Dilemma: Immigrants, Social Networks, and Institutional
Fairness’, Comparative Politics, 43 (1), 63–80.
Lancee, B. and Dronkers, J. (2011) ‘Ethnic, Religious and Economic Diversity in Dutch Neighborhoods: Explaining Quality
of Contact with Neighbors, Trust in the Neighborhood and Inter-ethnic Trust’, Journal of Ethnic and Migration Studies,
37 (4), 597–618.
Laurence, J. (2011) ‘The Effect of Ethnic Diversity and Community Disadvantage on Social Cohesion:A Multi-level Analysis
of Social Capital and Interethnic Relations in UK Communities’, European Sociological Review, 27 (1), 70–89.
Letki, N. (2008) ‘Does Diversity Erode Social Cohesion? Social Capital and Race in British Neighborhoods’, Political Studies,
56 (1), 99–126.
Lolle, H. and Torpe, L. (2011) ‘Growing Ethnic Diversity and Social Trust in European Societies’, Comparative European Politics,
9 (2), 191–216.
Mill, J. S. (1958 [1861]) Considerations on Representative Government. New York: Liberal Arts Press.
Nannestad, P. (2008) ‘What Have We Learned about Generalized Trust, if Anything?’, Annual Review of Political Science, 11,
413–36.
Newton, K. (2007) ‘Social and Political Trust’, in J. R. Dalton and H. Klingemann (eds), Oxford Handbook of Political Behavior.
New York: Oxford University Press, pp. 342–61.
Norris, P. (2011) Democratic Deficit. New York: Cambridge University Press.

© 2013 The Author. Political Studies © 2013 Political Studies Association

POLITICAL STUDIES: 2014, 62(3)
D O E S D I V E R S I T Y E RO D E S O C I A L C O H E S I O N ? 595

Portes, A. and Vickstrom, E. (2011) ‘Diversity, Social Capital, and Cohesion’, Annual Review of Sociology, 37, 461–79.
Putnam, R. D. (2000) Bowling Alone:The Collapse and Revival of American Community. New York: Simon & Schuster.
Putnam, R. D. (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.
Reeskens, T. (2010) ‘Ethnic-Cultural Diversity, Migrant Integration Policies and Social Cohesion in Europe: Investigating the
Conditional Effect of Ethnic-Cultural Diversity on Generalized Trust’. Paper presented at the ‘Migration: A World in
Motion’ conference, Maastricht, 18–20 February.
Reeskens, T. and Wright, M. (2013) ‘Nationalism and the Cohesive Society: A Multi-level Analysis of the Interplay between
Diversity, National Identity, and Social Capital across 27 European Societies’, Comparative Political Studies, Online First.
Rothstein, B. and Uslaner, E. M. (2005) ‘All for All: Equality, Corruption, and Social Trust’, World Politics, 58 (1), 41–72.
Savelkoul, M., Gesthuizen, M. and Scheepers, P. (2011) ‘Explaining Relationships between Ethnic Diversity and Informal
Social Capital across European Countries and Regions: Tests of Constrict, Conflict and Contact Theory’, Social Science
Research, 40 (4), 1091–107.
Sinnott, R. (2006) ‘An Evaluation of the Measurement of National, Sub-national and Supranational Identity in Cross-
National Surveys’, International Journal of Public Opinion Research, 18 (2), 211–23.
Solt, F. (2009) ‘Standardizing the World Income Inequality Database’, Social Science Quarterly, 90 (2), 231–42.
Stolle, D., Soroka, S. and Johnston, R. (2008) ‘When Does Diversity Erode Trust? Neighborhood Diversity, Interpersonal Trust
and the Mediating Effect of Social Interactions’, Political Studies, 56 (1), 57–75.
Sturgis, P., Brunton-Smith, I., Read, S. and Allum, N. (2011) ‘Does Ethnic Diversity Erode Trust? Putnam’s “Hunkering
down” Thesis Reconsidered’, British Journal of Political Science, 41 (1), 57–82.
Taylor, J., Twigg, L. and Mohan, J. (2010) ‘Exploring the Links between Population Heterogeneity and Perceptions of Social
Cohesion in England’, in J. Stillwell and M. Ham Van (eds), Ethnicity and Integration. Berlin: Springer, pp. 269–87.
Teorell, J., Charron, N., Samanni, M. and Holmberg, S. R. B. (2011) The Quality of Government Dataset. Available from:
http://www.qog.pol.gu.se [Accessed 20 July 2012].
Tolsma, J., van der Meer, T. and Gesthuizen, M. (2009) ‘The Impact of Neighborhood and Municipality Characteristics on
Social Cohesion in the Netherlands’, Acta Politica, 44 (3), 286–313.
Tsai, M. C., Laczko, L. and Bjørnskov, C. (2011) ‘Social Diversity, Institutions and Trust: A Cross-National Analysis’, Social
Indicators Research, 101 (3), 305–22.
Uslaner, E. M. (2002) The Moral Foundations of Trust. Cambridge: Cambridge University Press.
Uslaner, E. M. (2004) ‘Trust and Social Bonds: Faith in Others and Policy Outcomes Reconsidered’, Political Research
Quarterly, 57 (3), 501–7.
Uslaner, E. M. (2012) Segregation and Mistrust. Cambridge MA: Cambridge University Press.
Van Oorschot, W., Arts, W. and Halman, L. (2005) ‘Welfare State Effects on Social Capital and Informal Solidarity in the
European Union: Evidence from the 1999/2000 European Values Study’, Policy & Politics, 33 (1), 33–54.
Vergolini, L. (2011) ‘Social Cohesion in Europe: How do the Different Dimensions of Inequality Affect Social Cohesion?’,
International Journal of Comparative Sociology, 52 (3), 197–214.
Wu, Z., Hou, F. and Schimmele, C. M. (2011) ‘Racial Diversity and Sense of Belonging in Urban Neighborhoods’, City &
Community, 10 (4), 373–92.
You, J.-S. (2012) ‘Social Trust: Fairness Matters More than Homogeneity’, Political Psychology, 33 (5), 701–21.
Zak, P. J. and Knack, S. (2001) ‘Trust and Growth’, Economic Journal, 111 (470), 295–321.

© 2013 The Author. Political Studies © 2013 Political Studies Association

POLITICAL STUDIES: 2014, 62(3)