Social Cohesion and Immigration in Europe and North America

Mechanisms, Conditions, and Causality

Ruud Koopmans, Bram Lancee, Merlin Schaeffer

• Publisher: Routledge
• Year: 2014
• eBook ISBN:
  9781317689072
• Topic: Sciences sociales
• Subtopic: Sociologie

Concerns about immigration and the rising visibility of minorities have triggered a lively scholarly debate on the consequences of ethnic diversity for trust, cooperation, and other aspects of social cohesion. In this accessibly written volume, leading scholars explore where, when, and why ethnic diversity affects social cohesion by way of analyses covering the major European immigration countries, as well as the United States and Canada. They explore the merits of competing theoretical accounts and give rare insights into the underlying mechanisms through which diversity affects social cohesion. The volume offers a nuanced picture of the topic by explicitly exploring the conditions under which ethnic diversity affects the 'glue' that holds societies together. With its interdisciplinary perspective and contributions by sociologists, political scientists, social psychologists, as well as economists, the book offers the most comprehensive analysis of the link between ethnic diversity and social cohesion that is currently available.

1 Ethnic diversity in diverse societies

An introduction

Ruud Koopmans, Bram Lancee, and Merlin Schaeffer

Concerns about immigration and the rising visibility of ethnic and racial minorities have triggered a lively scholarly debate on the consequences of ethnic diversity for social cohesion. Economists suggest that ethnic diversity is one of the reasons for stagnation and corruption in the developing world (Easterly and Levine, 1997), and explains why the USA does not have a European-style welfare state (Alesina et al., 2001). In political science and sociology Putnam’s (2007) “hunkering down thesis” is a central focus of debate: in neighborhoods, cities, or regions that are more ethnically diverse, citizens withdraw from public social life and reciprocity and trust go down.

While the debate originates in the USA, it has been receiving growing attention from European scholars, who fear that the high levels of trust, civic engagement, and redistribution that characterize European countries might be threatened by increasing levels of ethnic diversity. In Sweden, higher levels of ethnic diversity have been shown to be associated with declining levels of support for welfare state spending (Eger, 2010), and in Germany with declining support for unemployment benefits (Stichnoth, 2012). In many other European countries, ethnic diversity is related to a decline in social trust and civic engagement (Delhey and Newton, 2005; Dinesen, 2011; Laurence, 2011).

However, the ethnic diversity hypothesis is also disputed. Critics argue that the role of socio-economic deprivation has been overlooked (e.g., Letki, 2008; Twigg et al., 2010) and that findings that are particular to one national context, especially the US, are generalized to other countries (e.g., Harell and Stolle, 2010; Holtug and Mason, 2010). Moreover, heterogeneous networks often lead to positive outcomes, such as more productive job searches (Granovetter, 1973), more creativity (Burt, 2000), and greater problem-solving capacities (Gurin et al., 2004).

Whatever their disagreements, both sides of the debate agree that under some circumstances ethnic diversity can lead to declines of social cohesion, but also that, in the long run, immigration is beneficial to innovation and economic prosperity (cf. Alesina et al., 1999; Page, 2008; Putnam, 2007). Yet, if ethnic diversification is associated with short and medium-term declines in social cohesion and public goods production, we need to understand why and under which conditions this is the case. However, until now, most research has focused on the question whether a negative relation between ethnic diversity and social cohesion exists at all. Moreover, almost all evidence in the one or the other direction has been correlational. The causal effect of ethnic diversity therefore remains an issue of further study. Incomplete knowledge about causality and underlying mechanisms makes it hard to propose policy solutions and to bring the research field further.

Ethnic diversity can be a consequence both of indigenous ethnic heterogeneity and of recent immigration. While many of the arguments discussed in this volume also apply to examples of “native” heterogeneity such as in the cases of Basks, Catalans, and Castilians in Spain or relations between whites, blacks, and Native Americans in the United States, in this volume we focus exclusively on ethnic diversity that derives from recent immigration waves. It is the connection between ethnic diversity and immigration that has been at the heart of recent controversies over alleged declines in social cohesion in Europe and North America.

Social cohesion can include a broad array of phenomena. In this volume, we reserve it in the first place for a community’s capacity for collective action in pursuit of public goods, and the attitudes and expectations of trust that undergird this capacity. In a wider sense, the absence of prejudices and hostility among groups, the density and quality of intergroup social contacts, shared norms, values, and identities, and the ability to communicate through a common language are often also counted as aspects of social cohesion. We will argue, however, that the latter are better seen as mechanisms and conditions through which social cohesion—in the sense of cooperative capacity and trust—may be negatively affected by ethnic diversity.

In order to better understand how and why ethnic diversity affects social cohesion in Europe and North America, we move beyond the omnipresent research that analyzes whether or not there is a relationship between diversity and social cohesion. Instead, the chapters in this volume engage explicitly with three core issues that we believe to be most pressing in the current debate: 1) the causal status of ethnic diversity effects; 2) competing explanations and mechanisms; and 3) conditional effects. While the chapters differ in many respects, such as the countries of analysis and the measures of social cohesion that are investigated, the explicit discussion of explanations and mechanisms, conditions and causality, is what unites the chapters into a coherent volume. An additional innovative feature of this volume is that, in addition to the more traditional focus on the diversity of spatial areas such as neighborhoods, the contributions in Part III expand the focus to the analysis of diversity effects in a key functional context: schools. Within the school setting, chapters focus on causality, mechanisms, and conditional effects.

Spurious correlation or causal effect?

The first contribution of this book is to better establish whether the relation between ethnic diversity and social cohesion is causal. The question of causality is very salient in the debate on the consequences of ethnic diversity. Is the observed effect of ethnic diversity spurious, due to unobserved contextual heterogeneity? For example, some scholars argue that studies do not fully take into account socio-economic deprivation (Letki, 2008). In this view, the effect of ethnic diversity is significant only because of its correlation with the socio-economic status of neighborhoods or cities. Similarly because of data limitations or because the amount of variables that can be included on the contextual level is limited, it could be that previous work suffers from further sources of unobserved heterogeneity. Any such omitted variable bias would render observed ethnic diversity effects at least partly spurious, implying that correlational studies overestimate the effect of ethnic diversity.

Another important issue in establishing a causal relationship between diversity and social cohesion is individual’s self-selection. The interpretation of correlations between diversity and social cohesion is complicated by the potential influence of self-selection, for instance if better-situated, high-trusting people move to other areas when the ethnic composition changes (e.g., Alesina and La Ferrara, 2002; Growder et al., 2011), leaving behind the deprived, low-trusting, and disengaged. In this case, self-selection would also cause an overestimation of negative diversity effects. Alternatively, it may also be the case that more trustful people are more comfortable with diversity and stay in, or move to diverse areas, whereas low trusters avoid or move away from diversity. In this scenario, self-selection would lead to an underestimation of negative diversity effects. Either way, as most previous studies do not control for selection, existing findings may provide biased estimates of the negative effects of diversity.

Experimental and longitudinal designs are the best options to investigate causality issues, and this volume contains several examples of such designs. In order to address the bias arising from endogenous residential sorting, Camille Hémet presents in Chapter 2 a quasi-experimental study on individuals living in the French public housing sector, in which housing is allocated independent of households’ ethnic origins or their preferences for diversity. The absence of choice in this section of the housing market eliminates the problem of self-selection. Nevertheless, Hémet finds that ethnic diversity is associated with increases of various indicators of degradation (e.g., vandalism, graffitti, trash lying around) in common areas of housing blocks. In addition, Hémet finds negative effects of ethnic diversity on individuals’ chances of finding employment, suggesting that communication barriers and ethnically clustered networks limit the circulation of information in diverse areas. These effects are strongest when analyzed on the level of small neighborhoods rather than larger geographical areas.

In Chapter 3, Bram Lancee and Merlin Schaeffer employ a longitudinal panel design to analyze causal effects of ethnic diversity in Germany. There is hardly any empirical research on diversity effects that relies on longitudinal data. This is problematic because as Hopkins (2010: 160) states: “To understand how diversity influences public good provision, we should look to those towns that are diversifying, not those towns that are diverse.” Put differently, the guiding question in the debate is how changes in ethnic diversity affect changes in trust and social cohesion. Applying a difference-in-difference design, Lancee and Schaeffer investigate how the event of moving to a more diverse neighborhood affects people’s opinions about immigration. They show that moving to diversity indeed results in more concerns about immigration. Moreover, the effect lasts over time: even three years after moving, people who moved to a more diverse neighborhood remain significantly more xenophobic.

Another way of dealing with selection problems is to make use of natural experiments. When diversity arrives as an exogenous shock, one can observe potential changes in social capital among a “treatment group” experiencing an increase of diversity it did not choose, compared to a similar “control group” whose level of diversity has not changed. In Chapter 4, Abigail Fisher Williamson reviews the benefits and challenges of using natural experiments, drawing on the example of Lewiston, Maine, which experienced an unanticipated arrival of Somali refugees. With qualitative interviews and survey data she investigates changes in local social capital in Lewiston after the Somalis’ arrival and carefully compares these changes to developments in otherwise similar localities. The findings are mixed. On the one hand, on the county and municipal levels of analysis there is no evidence that organizational involvement, trust, and interracial accord have disporportionately declined in Lewiston. However, on the neighborhood level she does find marked declines for various indicators of social cohesion comparing the neighborhoods where diversity had increased most strongly to otherwise similar neighborhoods that had not experienced strong increases in diversity. For instance, agreement to the statement that neighbors are likely to cooperate declined by as much as 41 percent in the neighborhoods that experienced the greatest increases in diversity.

Causality problems are not completely avoided in field-experimental studies, because ethnic diversity is used as an explanatory contextual variable, but is not itself subject to experimental control (see also Falk and Zehnder, 2013; Koopmans and Veit, 2014). Designs that fully randomize experimental and control conditions can establish causal claims with greater confidence. However, existing laboratory experiments that have taken into account ethnicity were usually conducted with interethnic pairs of participants instead of ethnically diverse groups, leading to mixed findings reflecting trust, cooperation, and discrimination among specific ethnic groups rather than general effects of ethnic diversity (e.g., Chuah et al., 2013; Fershtman and Gneezy 2001). Two exceptions are the studies of Alexander and Christia (2011) and Koopmans and Rebers (2009), which both use public goods games to study cooperation in culturally homogeneous and diverse groups with random assignment and find lower cooperation levels in the latter. While these studies provide a stronger basis for causality claims, they raise—as laboratory experiments generally do—the question of external validity.

In Chapter 9, Susanne Veit presents the results of two survey-experimental studies among neighborhood residents across Germany and parents at primary schools in Berlin. In both experiments, respondents were randomly allocated to experimental primes that heightened the salience of ethnic diversity in their neighborhoods, respectively schools, without, however, suggesting any negative or positive evaluation of such diversity. As control conditions, other respondents were randomly primed either with another form of neighborhood or school diversity or with no particular form of diversity at all. The results show that people who are cogiiitively exposed to ethnic diversity primes are significantly more pessimistic about the trustworthiness of neighbors and parental cooperation at school than those who are exposed to age or income diversity primes, or to no particular form of diversity at all. The findings for the treatment in which people’s attention was directed to income differences are particularly interesting. This study is, as far as we are aware, the first that experimentally tests ethnic diversity and socioeconomic inequality explanations for declines in trust against each other. The results show that raising the cognitive salience of ethnic diversity does reduce trust in neighbors, whereas pointing towards economic inequalities does not.

By eliminating potential biases due to self-selection and unobserved heterogeneity, these four studies provide strong evidence of the causal nature of negative diversity effect. In addition, they provide further evidence for the recurrent research finding (see the literature reviews of Schaeffer, 2014 and Van de...

Table of contents

• Cover
• Half Title
• Title Page
• Copyright Page
• Table of Contents
• List of figures
• List of tables
• List of contributors
• 1 Ethnic diversity in diverse societies: an introduction
• Part I The causal nature of diversity effects
• Part II The moderating role of interethnic contacts, identities, and policies
• Part III Ethnic diversity in schools
• Index