Histories of Conquest,Diversity,and Social Cohesion in Former Colonial Europe

Abstract

Research has overlooked the influence of countries’ colonizing histories on how present-day diversity shapes forms of social cohesion. Ethnic boundary theories suggest that legacies of colonizing external territories could strengthen the boundaries that separate natives from foreign migrants. Alternatively, colonization could simply give countries greater experience with foreign populations over time, thus diffusing boundaries through sustained integration processes. This article investigates how history shapes diversity’s influence on horizontal and vertical forms of social cohesion: support for welfare to reduce hierarchical class differences and trust of the generalized other. Using 2002 to 2014 European Social Survey and country history data, I find that while diversity often directly reduces both welfare support and trust, histories of conquest moderate this relationship. Specifically, diversity’s negative influence on social cohesion outcomes gradually diminishes with the occupation of foreign territories, in contrast to their colonization. While prior research emphasizes diversity as a straightforward negative force, current findings show that it is shaped by historical episodes of symbolic boundary making.

Keywords

social trust welfare support immigration diversity colonialism

Introduction

Increased diversity in Europe has sparked debates about the boundaries of national belonging, the future of welfare state institutions, and the rights and mutual obligations associated with citizenship. Sociopolitical organization in many countries relies on interconnectedness and cooperation as the public foundation for social cohesion. Social cohesion can be defined in terms of at least two features of Western liberal democracy. Protection from social exclusion represents a vertical dimension of social cohesion, linking individuals to the state and social institutions, whereas the horizontal dimension of social cohesion is indicated by the quality of social interaction among members of society (Vergolini 2011). Scholars have linked both horizontal and vertical dimensions of cohesion to diversity by investigating whether increasing homogeneity erodes welfare state support (Alesina and Glaeser 2004; Eger 2010; Mau and Burkhardt 2009) and generalized social trust among the public (Hooghe et al. 2009; Kesler and Bloemraad 2010; Putnam 2007), both a linchpin of Europe’s democratic systems.

One aspect of demographic change that scholars have largely overlooked is the connection between diversity and Europe’s history of colonizing external territories. Many of the immigrant communities in Europe originate in former colonies (Beaman 2016; Killian and Manohar 2016; van Amersfoort and van Niekerk 2006). While a handful of studies acknowledge that history and cultural variables matter for social cohesion (Delhey and Newton 2005; Gesthuizen, van der Meer, and Scheepers 2009), none looks closely at indicators of a country’s past, for example, historical relationships of external colonization and occupation. With this study, I develop and test a conceptual framework for analyzing destination’s historical contexts that brings together insights from theories on race and ethnicity, symbolic boundary making, and Europe’s colonizing history. First, I test a broad range of diversity indicators on welfare support and social trust, while controlling for the individual characteristics known to shape cohesion. Next, I conduct a series of analyses on the joint role of ethnic heterogeneity and histories of conquest to determine whether diversity’s purported pernicious effects are dependent on a country’s historical legacy of boundary making. I end by discussing findings within the context of sociodemographic change across the continent.

This study examines how European powers’ past external conquests structure the way present-day diversity is experienced through its effect on social cohesion. I argue that ignoring episodes of conquest obscures the continuing relevance of past symbolic boundary making for contemporary demographic change. Relatedly, this study develops an innovative, generalizable framework by which to incorporate analyses of historical trajectories, which can be used in future research to link past structural forces to more recent social change. This is the first large-scale cross-national quantitative study to directly compare the dynamics underlying two important social cohesion processes in the same analyses: the maintenance of welfare support and preservation of social trust. Is welfare support characterized by path dependency, as prior research would suggest, compared to generalized social trust, which is possibly more fleeting? Do contexts of diversity influence both outcomes similarly? Finally, the study answers calls for a shift in focus from universal reactions to context-specific responses (Kesler and Bloemraad 2010). Findings show that ethnic diversity may not elicit the same reactions everywhere. Instead, diversity effects are channeled through prior social and institutional configurations.

Theoretical Background

Social cohesion provides the conditions for and shapes the consequences of public civic life in many liberal democracies (van der Meer and Tolsma 2014). In the current study, social cohesion is conceptualized in terms of its civic and public importance as the interpersonal and institutional connections that foster a sense of solidarity and shared bonds between members of society. While the concept of social cohesion has been treated extensively as a multidimensional concept arranged along different axes and embedded in the structure of networks (Moody and White 2003; Vergolini 2011), I maintain a focus on the horizontal and vertical dimensions of public social cohesion, which could form the basis for group solidarities at various scales and temporalities (Rusu and Gheorghita 2014). The vertical dimension represents connections between an individual and others through social institutions of protection. This dimension emphasizes shared fate and implies an interest in safeguarding a cohesive core to promote solidarity and prevent social exclusion. For the vertical dimension, I focus on respondents’ beliefs about minimizing differences—and thus strengthening the social core—between those at the top and bottom of the class structure. The horizontal dimension of social cohesion represents the frequency and quality of individuals’ interactions with others. This feature highlights individuals’ openness and willingness to engage with other people, with positive engagement implying a greater sense of social solidarity and the creation of bonds. For the horizontal dimension, I gauge respondents’ views and trust in other members of society more generally, regardless of their social position. The horizontal and vertical dimensions of cohesion correspond theoretically to the public’s welfare support and generalized social trust, respectively, as elaborated below.

Social cohesion is often disrupted by the perception of group boundaries, which shifts people’s focus from shared commonalities to the differences that distinguish and divide social groups. While symbolic boundaries can take different shapes depending on the social differences that define them, ethnic differences often constitute very powerful boundaries (Wimmer 2013). Ethnic boundary theories conceive of social cohesion as the product of perceived identity defined against other racial and ethnic groups (Barth 1969). The consistent regeneration of ethnic boundaries can result in fluid and mutable distinctions between ethnic groups, but can also harden under conditions where boundaries correspond to cultural differences, where there are high degrees of social closure, or where political networks are aligned along already existing ethnic boundaries (Wimmer 2008, 2013). Salient ethnic boundaries may strengthen intragroup cohesion, but simultaneously weaken intergroup cohesion.

Ethnic boundaries may be circumscribed along symbolic as well as geographic and contextual parameters. Nations as symbolic boundaries, and respective national modes of belonging, can be resistant to considerable change and are found to be path dependent, although the durability of these boundaries also varies between nations (Alba 2005; Bail 2008; Brubaker 1992; Zolberg and Woon 1999). An ethnic boundary approach to national cultures suggests that nations develop boundary-making strategies that permeate a number of social fields, thereby solidifying the way publics and immigrant populations view each other and themselves (Wimmer 2008:992).

Thus, the salience of ethnic boundaries could serve as the basic mechanism for the formation of natives’ views about who belongs in cohesive national polities and who does not, and whether the arrival of diverse populations threatens this sense of cohesion. The heightened salience of ethnic boundaries produces solidarity groups, which, when perceived to be distinct and in opposition to each other, undermine intergroup cohesion. Ethnic diversity has been theorized to negatively impact horizontal and vertical forms of social cohesion. For horizontal social cohesion, it is possible that welfare support is depressed when highly salient ethnic boundaries prompt the public to view diverse outsiders as non-contributors to the social pact, a drain on welfare systems, and less deserving than natives (van Oorschot 2006). Vertical social cohesion in the form of social trust could diminish when salient ethnic boundaries cause natives to avoid intergroup contact and gradually isolate themselves from the larger public (Putnam 2007). Figure 1 presents a conceptual model of the relations between ethnic diversity, histories of conquest, the salience of ethnic boundaries, and social cohesion outcomes. Ethnic diversity is expected to shape the salience of ethnic boundaries through type of immigration (total, non-Western, Muslim foreign-born). Histories of conquest are also expected to influence the salience of ethnic boundaries depending on the type and duration of countries’ foreign exploits. Below, I discuss the baseline ethnic heterogeneity and constrict hypotheses related to diversity and then link ethnic boundary making to colonization histories, which are expected to moderate the salience of ethnic boundaries.

Figure 1.

Conceptual model of relationship between ethnic diversity, histories of conquest, the salience of ethnic boundaries, and social cohesion outcomes.

Ethnic Diversity and the Welfare State

In Europe, sociodemographic shifts have prompted governments and policymakers to rethink fundamental institutions of equality and democracy, including the occasionally threatened but already entrenched welfare state. The fear is that the growth of diverse populations that are viewed as less “deserving” of social assistance, namely, immigrants (van Oorschot 2006), will undermine the welfare state’s legitimacy. The anxiety captured in this deservedness hypothesis is compounded by observations that immigrants are proportionally more likely to receive welfare benefits (Boeri, Hanson, and McCormich 2002).

An ethnic heterogeneity hypothesis maintains that greater diversity in a country leads to diminished support for welfare. Of note is the work of economists Alesina and Glaeser (2004), who formulate a theory of welfare state development and evolution. Comparing social policy development in the United States and Europe, the authors argue that ethnic diversity in the United States curtailed support for redistributive policies. They contrast this with the experience in Europe, where ethnic homogeneity and active labor movements promoted the development of robust welfare state institutions. Importantly, extrapolating the correlation between diversity and welfare, Alesina and Glaeser (2004) relate the historically anemic American welfare state to the potential for retrenchment in Europe. An ethnic heterogeneity hypothesis as applied to Europe thus suggests that higher levels of diversity in a country diminish public welfare support (H1).

European welfare states have evolved in specific historical contexts, characterized by class struggle, the power of left parties, a resulting decommodification of labor, and the leveling of class differences (Esping-Andersen 1990). Prior research suggests that support for state intervention and redistribution tends to be path dependent, with the highest support in the social democratic welfare states of Nordic Europe and lowest in the Anglo-Saxon liberal welfare states. However, beyond these general patterns, research fails to provide conclusive evidence for a distinct clustering of attitudes by regime type (see discussion in Svallfors 2012:8–10).

Research that tests the ethnic heterogeneity hypothesis provides partial evidence that more ethnically heterogeneous populations exert downward pressures on the public’s welfare support. Mau and Burkhardt (2009) find that the percentage of non-Western foreign-born in a country and migration inflows have a significant negative influence on welfare support, although this varies by other macro-level indicators such as GDP, welfare regime type, and unemployment levels. Considering Sweden over a series of years, Eger (2010) finds that recent immigration and the proportion foreign-born in regional län reduce support for the robust Swedish welfare state. Similarly, Burgoon (2014) shows that across Europe, a larger foreign-born population generally dampens public support for welfare, an effect that is exacerbated by migrants’ economic non-integration. In contrast, Brady and Finnigan (2014) analyze a larger sample of rich countries and show meager evidence for a generic ethnic heterogeneity hypothesis. Instead, certain forms of population change, such as net migration and migration inflows, have a positive effect on public welfare support. The authors attribute this to a compensation dynamic, where more acutely visible immigration triggers a sense of insecurity and competition for jobs.

Ethnic Heterogeneity and Social Trust

The vertical dimension of social cohesion can be examined as a function of people’s trust in others they do not personally know. Social trust is a crucial component of social capital (Putnam 1993), and scholars have linked trust to democratic processes and institutions (Paxton 2002) and low crime rates and individual well-being (Putnam 2000), among other indicators. Trust is also associated with pro-social behavior such as volunteering, donating, participating in civic organizations, and giving to the needy (Uslaner 2002). Furthermore, the solidaristic foundations of important social and political institutions are found to be dependent on relatively high levels of social trust (Crepaz 2008; Soroka, Helliwell, and Johnston 2007). Social trust varies by national context, with the universalist Nordic countries exhibiting high levels and Southern and Eastern European countries showing lower levels of trust (Delhey and Newton 2005; Hooghe et al. 2009).

Similar to the ethnic heterogeneity hypothesis, research has linked increasing diversity with diminished social trust. Perhaps best known from Robert Putnam’s (2007) exposition on diversity and social cohesion in the United States, the constrict hypothesis maintains that ethnically heterogeneous environments are detrimental to interpersonal trust and the formation and preservation of social connections. The constrict hypothesis predicts troubling change for most advanced democracies whose populations are rapidly diversifying. Anomie and social isolation increase with diversity, thus affecting both in-group and out-group relations (Putnam 2007). In Putnam’s (2007:149) words, people “pull in like a turtle” in the face of diversity. A constrict hypothesis suggests that higher levels of diversity in a country suppress the public’s social trust (H2).

A flurry of research on social trust has produced a “cacophony of findings” (van der Meer and Tolsma 2014:460). Studies focusing largely on the United States provide some support that increased ethnic heterogeneity weakens generalized social trust (Alesina and Ferrara 2002; Costa and Kahn 2003). Research that casts a comparative net produces evidence that diversity suppresses generalized social trust (Delhey and Newton 2005; Paxton 2007), other mixed evidence (Gesthuizen et al. 2009; Hooghe et al. 2009), and yet additional findings that no negative relationship exists (Kesler and Bloemraad 2010; Savelkoul, Gesthuizen, and Scheepers 2011). The constrict hypothesis finds most support in the American context (van der Meer and Tolsma 2014). Increasingly, however, analysts of social trust have begun to look to lower-level contextual environments in Europe, like residential neighborhoods, uncovering additional support for the constrict hypothesis (Dinesen and Sønderskov 2015).

Ethnic Boundaries and Colonialism

National histories could also influence the salience of ethnic boundaries. Studies of boundary making show that ethnic distinctions are often accompanied by hierarchies of power and, where institutionalized, can generate and maintain systems of subjugation (Wimmer 2013:92–95). Relationships of subjugation can partly be evaluated by considering a country’s colonial history. Theories in the postcolonialist tradition argue that colonialism and conquest cannot be adequately analyzed as mere episodes in a country’s distant past, but rather that the West’s global exploits have had a lasting effects on how modern institutions and structures operate today (Bhambra 2014; Go 2016). Furthermore, colonialism fundamentally shaped the epistemologies underlying modern knowledge and ways of thinking about the world, including ideologies surrounding race and ethnicity (Go 2016).

The classical period of European colonial expansion witnessed incorporation of vast parts of the world’s territory under European powers’ domain and subsequent projects to dominate foreign peoples. Since that period, countries that were once colonial powers have experienced considerable postcolonial migration from their former colonies in Africa, the Middle East, Asia, and Latin America (Castles, de Haas, and Miller 2014). The presence of foreign populations that were once subjugated under the yolk of colonialism could provide the conditions for salient ethnic boundaries to emerge in countries such as France, the United Kingdom, the Netherlands, and Spain when compared to countries without extensive colonial histories, such as the Nordic European countries. Indeed, comparative scholarship on racial boundaries points to the role of colonialism in defining race discourses throughout the colonial and postcolonial periods in countries of conquest (Bleich 2003; Lentin 2004; Triandafyllidou 2001; Zølner 2000). Thus, each colonial relationship provides occasion for strengthening the salience of boundaries in a country between natives and immigrants from foreign territories, and possibly by extension all immigrants. These salient boundaries could in turn shape the association between ethnic diversity and social cohesion. A closed ethnic boundary hypothesis suggests that with histories of colonialism and conquest, ethnic diversity will have an overall greater negative impact on social cohesion through the intensification of boundary salience (H3).

However, a history of colonial relationships could also provide a greater basis for sustained immigrant integration. Countries with longer periods of postcolonial migration could foster systemic immigrant incorporation over the course of generations, compared to countries that have experienced more recent, non-colonial immigration. Given the mutability of ethnic boundaries, sustained episodes of integration could weaken the salience of such boundaries by providing the opportunity for natives and immigrants to continuously renegotiate them (Jiménez 2008; Lamont, Morning, and Mooney 2002; Waters 1999). Although not all immigrant groups are integrated at the same rate or in the same manner, in the aggregate, recognition of colonial migrants as part of the national fabric, like some early South Asian immigrant groups in the United Kingdom, gradually redraws national boundaries toward inclusion. In contrast, countries without colonial experiences preserve national boundaries along more native ethnic lines in the face of newer waves of immigration. Thus, an open ethnic boundary hypothesis suggests that a history of colonialism moderates the role of ethnic diversity by making boundaries less salient, allowing an overall lesser negative impact on social cohesion (H4).

National colonial experiences vary in the duration, territorial breadth, and nature of contact between people during conquest. In an exhaustive study of naturalization regimes, Janoski (2010) makes an important distinction between colonizing and occupying countries. Here, I adopt Janoski’s distinction as one based on duration of involvement since the focus is on historic, path-dependent moderation.¹ Occupiers include European powers that gained short-lived access to foreign lands, perhaps with an unfulfilled ambition of becoming a full colonial power. Occupiers remain in the first stage of colonialism and have no long-term interaction with occupied populations (Janoski 2010).² In contrast, colonizers remain at least 50 years in colonized territories and pass through a number of stages that set the scene for greater immigration to the colonial metropole. Non-colonizers are neither colonizers nor occupiers.

In Janoski’s view, these distinctions have important repercussions for native-immigrant relations. He argues that not experiencing conquest during the great colonial period led to a sense of envy, or ressentiment (Greenfeld 1992), among many nations. As a means of compensating, non-colonizers and occupiers developed romantic forms of nationalism where positive national identities were forged to the exclusion of foreigners and outside ethnic groups in the homeland. From this perspective, colonial powers should be the most open to foreign populations, non-colonizing powers should be less open as they value nationalistic ideals, and occupiers should be the least open as they are beset by the dual frustration of real, but limited, exposure to foreigners and unrealized colonial aspirations. However, if boundaries are closed, colonization’s longer-term exposure could simply cause boundaries to atrophy over time compared to occupation, thus intensifying their effects.

Data and Measures

I use 2002 to 2014 European Social Survey (ESS) data merged with historical data. The breadth of countries and time periods offered by the ESS is particularly well-suited to a cross-national analysis of social cohesion in Europe. To capture between-country variation while also attending to comparability regarding countries’ historical pasts, I include countries that are similarly situated in Europe’s geopolitical history and that have recently experienced immigration of ethnically different populations. This delimits the sample by excluding Eastern European countries. Eastern Europe has experienced its own period of occupation as a series of Soviet satellites, and in the more distant past was the site of vast empires, for example, the Bulgarian empires of the eleventh through fourteenth centuries, prior to the formation of the modern World System. However, the region has only recently begun to attract global immigrant populations. For the Eastern European countries available in the ESS, the foreign-born population makes up a median of only 3.5 percent of the total population. For the non-Western foreign-born, which we might expect to activate stronger boundaries, this figure drops to 2.2 percent. Theoretically, histories of occupation and colonialism in Eastern Europe are also not expected to shape cohesion in the same way since these countries were recently occupied, and not the occupiers. Experiencing protracted and recent periods of occupation may be expected to play a more direct—and not moderating—role by influencing trust and state support through other institutional characteristics, such as levels of democracy (Letki and Evans 2005). Thus, I include data from 17 Western, Southern, and Northern Europe countries that span 14 years.³ Countries contribute to the ESS at differing rates. The full sample includes between two and seven waves of survey data for each country, producing a dataset with just over 170,000 respondents over 103 country-years.

Outcome Variables

Social cohesion is conceptualized in its vertical and horizontal dimensions and measured as respondents’ support for welfare and generalized social trust, respectively. The vertical dimension is captured by an indicator of welfare support, measured by a single item available in ESS waves asking respondents whether they believe “the government should take measures to reduce differences in income levels.” Scaled responses are from 1 “Agree strongly” to 5 “Disagree strongly.” A dichotomous variable is generated to represent respondents who either agree or agree strongly. The ESS unfortunately does not provide a regular battery of items related to the welfare state, thus studies analyzing welfare support have relied on this single item (Burgoon 2014; Mau and Burkhardt 2009).

The horizontal dimension of social cohesion is captured by an indicator of social trust, measured as a mean index combining three survey items gauging respondents’ trusting view of the generalized other. These items ask whether respondents believe “Most people can be trusted or you can’t be too careful,” “Most people try to take advantage of you or try to be fair,” and “Most of the time people are helpful or mostly looking out for themselves.” While it should be noted that these items could exhibit multigroup differences, the three items are found to be sufficiently equivalent across countries (Reeskens and Hooghe 2008) and have been used in similar studies of social trust (Dinesen and Sønderskov 2015; Hooghe et al. 2009; Ziller 2015). The social trust scale ranges from 0 (low) to 10 (high). Respondents answering at least two of the survey questions are retained for all analyses.

Predictor Variables

For diversity measures, I draw primarily from the Eurostat database, which measures the foreign-born stock by their countries of origin in each destination country across time. Occasionally, reporting countries are missing data for specific origin countries in between time-points. In these cases, I fill in missing values using linear interpolation and avoid extrapolating beyond first or last available time-points.⁴ All diversity variables are standardized to produce meaningful coefficient estimates and odds ratios.

First, I include a measure of the general foreign-born population as a percentage of a country’s total population at time t. For ethnic boundary hypotheses, I also include measures of the foreign-born population that indicate racial, cultural, and religious difference. The second predictor measures the non-Western foreign-born population as a percent of total population, where Western countries include the EU-15, EFTA countries, the European micro-states, the United States, Canada, Australia, and New Zealand. Similar measures are used in prior research to represent the non-Western population (Schneider 2008; Ziller 2015).

Third, I construct a measure of the Muslim foreign-born population to assess the role of cultural- and religious-based differences on social cohesion. The role of Islam in Europe and its compatibility with Western democratic ideals have been widely researched in recent years from a number of perspectives (Joppke and Torpey 2013; Klausen 2005; Modood, Triandafyllidou, and Zapata-Barrero 2006). Ethnic boundary theories identify Muslim cultural and religious differences as a bright ethnic boundary in the European context (Alba 2005). I measure the Muslim foreign-born population as those originating from predominately Muslim societies as a percentage of total population.⁵ While this measure may over-represent first-generation Muslims, particularly for origins where there are large non-Muslim populations, it is preferable to other means of estimating the Muslim population, for example, by relying on counts from aggregated survey data, since surveys regularly underrepresent smaller minority populations. Furthermore, the overall Muslim population in each country is likely undercounted since foreign-born statistics do not capture Muslims from the second and third generations. The propensities for the measure to over-count and undercount the Muslim population are assumed to approximately cancel each other out. This measure highlights “otherness” by representing newer migration populations in Europe perceived as culturally and religiously different.

Historical variables reflect each country’s colonial past, particularly during Europe’s classical period of expansion and conquest. I adopt two variables from Janoski (2010), which represent a country’s experience colonizing and occupying foreign territories. For colonization, an average score is generated as the number of colonized people in each country’s territories divided by the home country population from 1870 to 1955. The occupation measure is similarly calculated but only applies to foreign exploits of less than 50 years.⁶ While histories of conquest focus on domination overseas, this does not imply that contemporary struggles related to territorialization and power are not also currently taking place. Janoski’s measure is the only quantified measure constructed for cross-national research that also considers the qualitative differences between stages of civilizational contact. Colonization requires 50 years, which spans the amount of time required to move through all stages of contact (see Note 2), thus solidifying power differentials between colonizers and colonized. In contrast, occupiers may have more superficial engagement and engage in contact where power differentials have not yet calcified, which should matter for the hypothesized boundary-making processes.

Control Variables

Control variables are included to account for individual characteristics that are conventionally controlled for in similar studies (see Supplementary Table S1). Sociodemographic controls include binaries for female gender, married status, unemployed status, status as an unskilled worker (ISCO 9000 unskilled occupations), low income (bottom third of country-year income distribution), (far) right political ideology (7–10 on 11-point left-right scale), (sub)urban residence, current or past union membership for welfare support, and status as foreign-born. Age and education are measured continuously in years and are standardized. Attitudinal controls include satisfaction with government (7–10 on 11-point extremely dissatisfied-extremely satisfied scale) and anti-immigrant attitudes (0–3 on 11-point scale from “immigrants undermine cultural life” to “immigrants enrich cultural life”). The latter has typically not been analyzed in studies of social cohesion and is included here to determine whether the influence of individualized xenophobia is shaped by diversity contexts.

For social trust, I also include binaries for religious attendance (attend at least once a week) and potential for political engagement (very/quite interested in politics) since people who attend religious services and are politically effective show higher levels of social trust (Hooghe et al. 2009).⁷

Analytic Technique

Multilevel models estimate the influence of predictors at different levels of analysis based on the grouping of observational units. Ignoring the hierarchical nature of the data can violate the assumption of independent errors and lead to underestimation of standard errors (Snijders and Bosker 2012). Multilevel techniques have advantages over other methods, such as ordinary least squares (OLS), since they allow the simultaneous estimation of individual- and upper-level effects and variance components. The data include individuals (ij) nested within 17 countries (j). To account for this nested structure, hierarchical models are estimated with random intercepts for countries.⁸

For welfare support, the log odds of holding supportive attitudes for each individual ij is calculated as:

\begin{array}{l} \ln (\frac{p_{i j}}{1 - p_{i j}}) = γ_{00} + {x^{'}}_{i j} β_{i j} + Z_{j}^{B^{'}} γ_{j}^{B} + Z_{j}^{W^{'}} γ_{j}^{W} \\ + P E + ϵ_{j} + ϵ_{i j} \end{array}

The level of social trust for each individual ij on the social trust index is calculated as:

\begin{array}{l} Y_{i j} = γ_{00} + {x^{'}}_{i j} β_{i j} + Z_{j}^{B^{'}} γ_{j}^{B} + Z_{j}^{W^{'}} γ_{j}^{W} + P E \\ + ϵ_{j} + ϵ_{i j} \end{array}

where $x_{i j}$ is the vector of individual-level control variables, $Z_{j}^{B}$ is the vector of country means and time-invariant country-level variables, and $Z_{j}^{W}$ is the vector of time-varying variables for country j centered around the country mean $(Z_{j}^{W} = Z_{j} - {\bar{Z}}_{j})$ . $γ_{j}^{B}$ captures between-country effects, and $γ_{j}^{W}$ captures within-country variance. PE are year-specific period effects that affect all countries, such as external economic shocks. $ϵ_{j}$ represents a country-specific intercept, and $ϵ_{i j}$ is an idiosyncratic error.

The data’s cross-sectional nature allows an analysis of both between-country differences and within-country changes in diversity. I estimate hybrid models to decompose difference and change effects. Between-country estimates provide results consistent with cross-sectional models, whereas within-country estimates are equivalent to those from a standard fixed-effects model. I estimate hybrid models as a first step to determine which component of diversity is significantly associated with the social cohesion outcomes and then continue diversity-history interaction models using the most relevant decomposed measure of diversity.

Results

Multivariate Analyses

Is greater heterogeneity associated with lower levels of welfare support? Panel 1 of Table 1 presents welfare support results from hybrid models estimating both within-country changes and between-country differences in diversity.⁹ Results show that while between-country differences exert no significant influence, within-country changes in diversity work to depress welfare support. This pattern holds for each form of diversity, with the most culturally distinct change, increases in the Muslim population, showing the greatest odds decrease. A 1 percentage point increase in the total foreign-born population is associated with a decrease in the odds of supporting welfare by a factor of 0.94, holding other variables constant. The same increase in the Muslim foreign-born population is associated with a greater decrease in the odds by a factor of 0.88. Overall, Panel 1 provides evidence for the ethnic heterogeneity hypothesis (H1). Furthermore, distinguishing the foreign-born population by the bright ethnic boundary of Muslim otherness (Alba 2005) uncovers evidence for greater opposition to welfare.

Table 1.

Multilevel Models for Welfare Support and Interactions between Diversity and Historical Contexts.

	Hybrid models (within and between)			Colonization models			Occupation models
	FB	NW FB	Muslim FB	FB	NW FB	Muslim FB	FB	NW FB	Muslim FB
Predictors	1	2	3	4	5	6	7	8	9
Diversity contexts
Foreign-born (within)	0.94***			0.92***			0.94***
	(−13.63)			(−11.12)			(−13.02)
Foreign-born (between)	0.99
	(−0.74)
Non-Western FB (within)		0.93***			0.92***			0.93***
		(−12.18)			(−10.54)			(−11.85)
Non-Western FB (between)		0.96
		(−0.75)
Muslim FB (within)			0.88***			0.88***			0.88***
			(−8.85)			(−6.72)			(−8.76)
Muslim FB (between)			0.91
			(−0.89)
Historical context
Colonization^a				0.87	0.87	0.87
				(−0.95)	(−0.93)	(−0.94)
Occupation^a							0.90	0.90	0.90
							(−0.78)	(−0.78)	(−0.77)
Interactions
FB (within) × colonization				0.97*
				(−2.56)
NW FB (within) × colonization					0.98*
					(−2.16)
Muslim FB (within) × colonization						1.01
						(0.24)
FB (within) × occupation							1.01*
							(2.57)
NW FB (within) × occupation								1.01**
								(2.94)
Muslim FB (within) × occupation									1.03***
									(3.35)
Constant	2.33**	2.62*	2.57***	1.91***	1.94***	2.13***	1.95***	1.98***	2.13***
	(3.18)	(2.48)	(3.94)	(4.85)	(4.95)	(5.63)	(4.95)	(5.08)	(5.58)
Period effects	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Variance: countries (N = 17)	0.29	0.30	0.29	0.29	0.29	0.29	0.29	0.29	0.30
	(0.54)	(0.54)	(0.54)	(0.54)	(0.54)	(0.54)	(0.54)	(0.54)	(0.55)
Intraclass correlation (ρ)	.08	.08	.08	.08	.08	.08	.08	.08	.08
Proportion variance explained	0.06	0.06	0.07	0.08	0.08	0.07	0.06	0.06	0.05
N (individuals)	182,514	182,514	182,514	182,514	182,514	182,514	182,514	182,514	182,514
Log-likelihood	−101,820.90	−101,839.80	−101,874.30	−101,817.40	−101,837.40	−101,874.20	−101,817.50	−101,835.50	−101,868.80
AIC	203,705.80	203,743.60	203,812.60	203,700.90	203,740.70	203,814.50	203,701.10	203,737.00	203,803.60

Note. Z-statistics in parentheses, except for variance components where standard deviations are shown. All individual-level controls from Supplemental Table S3 included but not shown. Proportion variance explained calculated against null model. FB = foreign-born; NW = non-Western; AIC = Akaike information criterion.

Variable standardized for ease of interpretation.

†

p < .10. *p < .05. **p < .01. ***p < .001 (two-tailed test).

Individual controls are presented in Supplementary Table S3. Results conform to expectations from prior research (Brady and Finnigan 2014; Burgoon 2014). Respondents who are female, older, unemployed, engaged in unskilled work, or union members are more supportive of the welfare state, all else equal. In contrast, those who are married, more highly educated, or who identify with the political right are less supportive of welfare, net of other factors. Furthermore, respondents who are more satisfied with the government are significantly less likely to support redistributive policy, as shown by prior research (Burgoon 2014). Interestingly, those who believe that immigrants are a cultural threat are only marginally less likely to support welfare than those who believe immigrants enrich the national culture. The caustic effect of xenophobia on welfare support thus does not appear to work directly through individual anti-immigrant attitudes but instead through other proxies, such as individual social position and political affiliation. This finding qualifies previous research on welfare chauvinism, which largely has not considered individual-level anti-immigrant sentiment (Mewes and Mau 2012; Reeskens and van Oorschot 2012).

Do histories of conquest moderate the relationship of diversity to welfare support? Panels 2 and 3 of Table 1 show that history matters for how diversity contexts shape support, with divergent patterns for colonization and occupation. Caution should be exercised when interpreting interaction coefficients and p-values in non-linear models since common residual variance cannot be assumed. Histories of colonization involve possession of a foreign territory for at least 50 years, and thus represent more prolonged engagement with ethnically different populations. The United Kingdom has an exceptionally high colonization score from its vast and populous empire, followed by the Netherlands with former colonies in the densely populated Indonesian archipelago, and then more distantly by Austria and Portugal (see Supplementary Table S2). Many countries have very low colonization scores (Belgium, Italy, and Sweden), or zero scores. In supplementary analyses, the United Kingdom is excluded from colonization models to ensure results’ robustness.

Panel 2 of Table 1 shows that colonization histories moderately intensify diversity’s negative influence on welfare support. Within-country increases in the foreign-born and non-Western foreign-born populations are expected to exert significantly greater downward pressures on welfare support with more extensive colonization histories. Increases in the Muslim foreign-born population appear to work directly on welfare support and are not moderated by colonization histories. Results provide evidence for a closed ethnic boundary hypothesis (H3) since greater experience with colonization appears to make boundaries more salient in their influence on welfare support.

Occupation represents more short-lived territorial possession and engagement with foreign-populations and is more evenly distributed between countries (see Supplementary Table S2). Several countries held no occupied territories and two countries show very low scores: Italy and Spain. Austria, Germany, and France show moderate occupation scores. The Netherlands and the United Kingdom show higher scores, whereas Belgium has the highest score for its short-lived but intense occupation of the Congo and, later, Rwanda and Burundi.

Panel 3 of Table 1 shows that histories of occupation have largely the opposite moderating effect. With greater occupation, the negative influence of within-country changes in diversity significantly diminishes. The greatest and most significant moderation occurs with increases in the Muslim foreign-born population. Thus, histories of occupation appear to partially temper diversity’s negative influence on welfare support, particularly for the most culturally salient changes. Results provide evidence for an open ethnic boundary hypothesis (H4) as boundaries become less salient with increasing occupation.

To illustrate interactions, Figure 2 (Panels A and B) represents diversity’s influence on welfare support for countries with different conquest profiles. Panel A shows that for all countries, regardless of colonization history, diversity exerts a similar negative influence on welfare support at very low levels of diversity. However, this negative influence differs in its slope for countries with different colonization profiles. For high colonizers, increases in the within-country foreign-born population result in more dramatic decreases in welfare support compared to low colonizers. Thus, colonization tends to exacerbate diversity’s negative influence on welfare support, particularly for high colonizers.

Figure 2.

Interactive effects of historical and diversity contexts on welfare support and social trust.

Panel B of Figure 2 represents the influence of increases in the Muslim foreign-born population on welfare support for countries with different occupation profiles, as presented in Model 9. A different pattern emerges for histories of occupation, with two notable characteristics. First, the influence of low levels of diversity on welfare support appears to be more stratified by occupation profile than the influence of high levels of diversity. Second, while low levels of diversity result in a higher probability of welfare support for low occupiers compared to high occupiers, increases in diversity result in a more dramatic drop in welfare support for low occupiers. For high occupiers, the slope of diversity’s negative influence is rather flat across levels of diversity. Thus, it appears that occupation significantly moderates diversity’s negative influence on welfare support.

Does diversity reduce social trust, as the constrict hypothesis suggests (H2)? Panel 1 of Table 2 presents social trust results from hybrid models estimating both within-country changes and between-country differences in diversity. Results show that while between-country differences exert no significant influence, within-country changes in diversity work to reduce social trust. This pattern holds for each form of diversity, although culturally distinct changes—increases in the Muslim population—are associated with the least significant effects. A 1 percentage point increase in the total foreign-born population is associated with a 0.02 point decrease in the social trust scale, net of other variables. Similar effects are found with increases in the non-Western and Muslim foreign-born populations. Overall, Panel 1 provides evidence for the constrict hypothesis (H2) that increases in diversity lead to social isolation and, ultimately, a reduction in social trust.

Table 2.

Multilevel Models for Social Trust and Interactions between Diversity and Historical Contexts.

	Hybrid models (within and between)			Colonization models			Occupation models
	FB	NW FB	Muslim FB	FB	NW FB	Muslim FB	FB	NW FB	Muslim FB
Predictors	10	11	12	13	14	15	16	17	18
Diversity contexts
Foreign-born (within)	−0.02***			0.003			−0.01***
	(−4.97)			(0.65)			(−3.85)
Foreign-born (between)	−0.005
	(−0.21)
Non-Western FB (within)		−0.02***			0.002			−0.01***
		(−4.26)			(0.41)			(−3.64)
Non-Western FB (between)		0.02
		(0.33)
Muslim FB (within)			−0.02*			0.01			−0.02*
			(−2.20)			(0.76)			(−2.05)
Muslim FB (between)			−0.03
			(−0.26)
Historical context
Colonization^a				0.01	0.01	0.01
				(0.04)	(0.04)	(0.04)
Occupation^a							−0.10	−0.10	−0.10
							(−0.63)	(−0.63)	(−0.63)
Interactions
FB (within) × colonization				0.04***
				(5.27)
NW FB (within) × colonization					0.03***
					(5.24)
Muslim FB (within) × colonization						0.08***
						(3.72)
FB (within) × occupation							0.01***
							(6.29)
NW FB (within) × occupation								0.01***
								(6.52)
Muslim FB (within) × occupation									0.04***
									(5.28)
Constant	5.47***	5.28***	5.50***	5.44***	5.44***	5.45***	5.41***	5.41***	5.43***
	(18.24)	(11.84)	(20.23)	(35.92)	(35.95)	(36.02)	(36.14)	(36.16)	(36.33)
Period effects	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Variance: countries (N = 17)	0.38	0.38	0.38	0.38	0.38	0.38	0.37	0.37	0.37
	(0.62)	(0.61)	(0.62)	(0.62)	(0.62)	(0.62)	(0.61)	(0.61)	(0.61)
Intraclass correlation (ρ)	.13	.13	.13	.13	.13	.13	.13	.13	.13
Proportion variance explained	0.30	0.30	0.30	0.30	0.30	0.30	0.31	0.31	0.31
N (individuals)	184,231	184,231	184,231	184,231	184,231	184,231	184,231	184,231	184,231
Log-likelihood	−352,370.80	−352,374.10	−352,380.80	−352,357.00	−352,360.40	−352,373.90	−352,350.90	−352,352.70	−352,366.70
AIC	704,811.70	704,818.20	704,831.60	704,785.90	704,792.90	704,819.80	704,773.80	704,777.40	704,805.30

Note. T-statistics in parentheses, except for variance components where standard deviations are shown. All individual-level controls from Supplemental Table S4 included but not shown. Proportion variance explained calculated against null model. FB = foreign-born; NW = non-Western; AIC = Akaike information criterion.

Variable standardized for ease of interpretation.

†

p < .10. *p < .05. **p < .01. ***p < .001 (two-tailed test).

Individual controls are presented in Supplementary Table S4. At the individual level, controls exhibit patterns that are consistent with prior research (Hooghe et al. 2009; Ziller 2015). Respondents who are female, older, more educated, more interested in politics, or who attend religious services show higher levels of social trust, all else equal. In contrast, respondents who have lower incomes, are unemployed or engaged in unskilled work, those who identify with the (far) right, those living in urban centers, and foreign-born respondents are less likely to trust others, net of additional factors. For subjective measures, respondents who show higher levels of government satisfaction are overall more trusting than the unsatisfied. Furthermore, respondents who view immigrants as a threat are significantly less trusting than those who believe immigrants enrich culture. Thus, the constricting effect of xenophobia on social trust works also through the individual level, in contrast to welfare support. Results could indicate a stronger social psychological basis for ingroup-outgroup distinctions with respect to social trust, perhaps also motivated by realistic group conflict and ethnic competition as identified by the literature on anti-immigrant attitudes (Scheepers, Gijsberts, and Coenders 2002). To test whether the influence of individual xenophobia is influenced by diversity contexts, I estimate cross-level interactions in supplementary analyses.

Do histories of conquest moderate the relationship of diversity to social trust? Panels 2 and 3 of Table 1 show that history matters for how diversity contexts shape trust, with similar interaction patterns for colonization and occupation, yet differentiated results for main effects. Developed histories of colonization appear to render diversity’s negative effect on social trust positive. For every form of diversity, main effects disappear while positive interactions emerge between colonization and diversity, suggesting that diversity works principally through its interaction with countries’ histories of colonization. Results support an open ethnic boundaries hypothesis (H3), as the salience of boundaries decline with greater experience as a colonizer.

Panel 3 of Table 2 shows that histories of occupation moderate diversity’s negative influence on social trust. In occupation models, main negative effects remain but are moderated by occupation histories, indicated by positive interaction coefficients. Interaction coefficients for Muslim foreign-born are somewhat higher than other forms of diversity. As with welfare state models, occupation results provide further evidence for an open ethnic boundaries hypothesis (H3).

To illustrate these interactions, Figure 2 (Panels C and D) represents diversity’s influence on social trust for countries with different conquest profiles. Panel C shows that countries with little colonial history have somewhat higher levels of social trust at low levels of diversity compared to high colonizers. This pattern changes, however, across levels of diversity change. For high colonizers in particular, increases in diversity result in a rather dramatic increase in social trust. In contrast, for low colonizers, increases in diversity lead to subtle decreases in social trust. Panel C suggests that greater historical exposure to foreign populations weakens the negative impact of diversity on social trust, and for those with the greatest exposure the relationship is positive.

Panel D of Figure 2 represents the influence of increases in the foreign-born population on social trust for countries with different occupation profiles. The influence of diversity on social trust is stratified by occupation profile at the low end of diversity. Furthermore, while low levels of diversity result in higher social trust for low occupiers compared to high occupiers, increases in diversity result in a drop in social trust for low occupiers. In contrast, for high occupiers, increases in diversity result in higher social trust. Thus, it appears that occupation significantly moderates diversity’s negative influence on social trust, and for the highest occupiers greater diversity is associated with high levels of trust.

Conditional effects analyses

All interactions imply that the influence of diversity varies in its magnitude, and perhaps its significance, depending on historical context. To test whether diversity effects are consistently significant under different historical conditions, I conduct a series of conditional effects analyses. I plot the conditional odds ratios or coefficients for diversity at different levels of colonization and occupation, with 95 percent confidence intervals.

Panels A and B of Figure 3 show conditional odds ratios for diversity’s effect on welfare support, and Panels C and D show conditional coefficient estimates for diversity’s effect on social trust. Panel A shows that within-country changes in the foreign-born population result in decreases in welfare support at all levels of colonization. Panel B shows that the negative influence of increases in the Muslim foreign-born population consistently weakens with greater occupation, except for the highest occupiers, where the diversity effect becomes non-significant. Conditional effects analyses show that diversity’s negative effect on welfare support is consistently intensified across all levels of colonization and, in contrast, moderated by occupation for all but the highest occupiers.

Figure 3.

Effects of diversity contexts conditional on historical contexts.

Panel C shows that for the lowest colonizers, changes in the foreign-born population result in significantly lower levels of social trust. However, this pattern reverses with even moderate levels of colonization. For most colonizers, increases in the foreign-born population are associated with significantly higher levels of social trust. Panel D shows that for low occupiers, occupation results in a weakened negative diversity effect on social trust. In contrast, for high occupiers, increases in diversity lead to greater social trust with great occupation. For countries with middling occupation profiles, increases in diversity appear to have no relationship with social trust.

Conditional effects analyses provide additional evidence for an open ethnic boundaries hypothesis (H3), particularly for occupation’s moderating effect. Histories of colonization, however, have a moderating effect on social trust, yet intensify diversity’s negative effect on welfare support, which suggests closed ethnic boundaries.

In sum, analyses suggest that we cannot attribute a mechanistic, negative effect to diversity, since the latter is dependent in many cases on historical trajectory.

Supplementary analyses qualify presented findings (see full description in Supplementary Materials pp. 14–15). First, for welfare support, results suggest that the diversity-intensifying effect associated with histories of colonization (Models 4 and 5) could be eclipsed when accounting for other country-level characteristics such as inequality, development, and economic globalization. This challenges the only evidence for a closed ethnic boundary hypothesis, although caution should be exercised when interpreting results with multiple higher-level variables and few higher-level units. In contrast, social trust results appear stable when controlling for other country-level characteristics. Second, the United Kingdom appears to drive the colonization results for social trust (Models 13 to 15). Third, for social trust, diversity appears to intensify individualized xenophobia’s negative effect, particularly with culturally different outgroups.

Discussion and Conclusion

Migration scholars acknowledge the link between Europe’s colonial conquests and current migration trends (Castles et al. 2014). Yet, in studies of social cohesion, analyses of migration-led diversity often overlook histories of conquest. Two common hypotheses suggest that diversity directly influences social cohesion. Ethnic heterogeneity and constrict hypotheses claim that higher levels of diversity in a country diminish public welfare support and social trust, respectively. Ethnic boundary theories qualify these hypotheses and suggest two divergent possibilities. Histories of colonialism could produce closed ethnic boundaries, as boundaries between natives and colonial migrants become salient and solidified over time. Alternatively, histories of colonialism could create the opportunity for open ethnic boundaries, where colonial powers gain more sustained experience with integrating foreign populations compared to non-colonizers.

Does the classical period of European colonialism have any bearing on the way diversity is experienced today? Results point to a qualified “yes.” Findings suggest that welfare support and social trust are in part shaped by a society’s diversity. Furthermore, this negative diversity effect is in most cases moderated by countries’ histories of conquest.

Within-country increases in ethnic heterogeneity directly diminish support for the welfare state and reduce social trust, providing plain evidence for ethnic heterogeneity and constrict hypotheses, and consistent with previous research (Burgoon 2014; Delhey and Newton 2005; Mau and Burkhardt 2009; Paxton 2007). Importantly, however, histories of occupation appear to ease diversity’s negative influence, even leading to positive diversity effects for very high occupiers. Overall, the balance of evidence points to historical trajectories providing the conditions for open ethnic boundaries (H4).

A closed ethnic boundary hypothesis finds some support, but only under certain conditions. With more developed colonization histories, within-country increases in diversity lead to greater decreases in the probability of supporting welfare. Interestingly, and in contrast, occupation histories moderate this negative effect. Some have posited that colonization and occupation differ not only in relation to their length and terms of engagement but also in the qualities of resulting symbolic boundaries (Janoski 2010). However, contrary to the expectation that occupiers become frustrated aspiring colonizers, current results suggest that occupation provides the most consistently moderating context for diversity, thus weakening the salience of symbolic boundaries. This could be due to occupation’s status as a sort of middle ground between non-colonizers who have little experience with diversity and colonizers whose longer-term engagement could cause ethnic boundaries to become more salient and even harden over time. It is also possible that occupation presents the opportunity to increasingly question old conceptions of diversity, such as the growing debate surrounding the Zwarte Piet practice in Netherlands and Belgium, without threatening foundational notions of a grand imperialist past. Findings suggest that some, but not too much, historical engagement with diversity creates possibilities for moderating present-day boundaries between natives and ethnic outgroups.

This balance that histories of occupation represent may be more important for welfare support than for social trust. Findings suggest that the closed boundaries associated with colonization do not apply to social trust. Histories of colonization instead offer some level of moderation of diversity’s negative influence on trust, although this is likely driven by high colonizer the United Kingdom. It appears, then, that the potential for boundary hardening is greatest for vertical forms of social cohesion, that is, cohesion that is connected to institutionalized forms of solidarity like the welfare state. This makes some sense considering scholars have often focused on the welfares state’s institutional path dependency, that is, the tendency of welfare states, and their support, to withstand external pressures linked to globalization (Pierson 1994). In this context, ethnic boundaries and their associations, such as stereotypes about outgroups and work ethic or racial cleavages, could just as likely harden in relation to the appraisal of cherished welfare institutions. Horizontal forms of social cohesion, such as generalized trust, appear to be more resistant to the salientization of boundaries. Because social trust is more interpersonal, historical experience with ethnic outgroups, in whatever form it takes, could weaken boundaries through consistent integration and exposure compared to those contexts with little historical engagement with diversity.

Notably, distinguishing between types of diversity uncovers patterns related to racial-ethnic distinctiveness. Typically, the foreign-born and non-Western foreign-born populations exert similar significant effects on social cohesion outcomes. For welfare support, however, increases in the Muslim foreign-born population are associated with somewhat greater decreases in support. This effect also finds more significant mediation by histories of occupation. This suggests that the bright ethnic boundary of Muslim otherness (Alba 2005) is more directly relevant for vertical, institutionalized forms of social cohesion, perhaps due to hardened, racialized stereotypes that play into growing welfare chauvinism (Reeskens and van Oorschot 2012). An implication of welfare chauvinism is that people are increasingly reluctant to contribute to a system of risk pooling when they think benefits will accrue to people unlike themselves. Welfare chauvinism could explain why typically path-dependent processes and institutions, like the welfare state, are perceived to be increasingly susceptible to changing demographics. In contrast, social trust does not appear to be directly influenced by different forms of diversity. However, culturally different diversity appears to be associated with greater moderation in interaction with histories of colonization and occupation, which produces positive effects at high levels of diversity and conquest. This could be due to greater proportions of Muslim foreign-born populations between countries with and without histories of conquest. On the one hand, findings related to cultural difference are promising considering Europe’s increasing racial and religious diversity, since it challenges common perceptions of immigrant cultures as incompatible with Western ways of life, a view that is increasingly articulated by right-wing nativist parties. From a less optimistic perspective, findings only confirm that while diversity’s negative effects may be moderated in many cases, they are still largely negative.

Individualized xenophobia is also shaped by diversity contexts. Supplementary findings show that in countries with higher levels of diversity, individual anti-immigrant attitudes have a larger negative effect on social trust (see Supplementary Materials, p. 15). Results extend prior studies of social trust (Hooghe et al. 2009; Ziller 2015), which have not examined this relationship, and suggest that diversity works partially through micro-level processes as well as through macro-level interactions, as studies of anti-immigrant attitudes have found (Schneider 2008).

What do findings mean for the future of ethnic relations on the continent? To some extent, the way diversity influences social cohesion is already historically circumscribed. Some countries have more open ethnic boundaries as a result of conquest experiences. Yet, it is important to note that diversity effects are still largely negative, and thus histories of conquest merely moderate already negative trends. History places occupiers at a slight advantage in dealing with the challenges that increasing diversity poses. To the extent that history and collective memory serve as a source of public meaning, occupiers could highlight their intercultural histories while focusing on the present as a means of fulfilling history’s trend toward cultural diversity and debating outdated ways of thinking about others. Certainly, this will take concerted political effort at a national level, since nativist political movements have gained considerable ground in framing diversity for the public. Non-occupiers who are at a disadvantage, like the Nordic countries, will have to find other ways to adopt and maintain positive frames for diversity and institutional means of sustaining equality. Innovative solutions that emphasize membership in a global community could help broaden conceptions of social cohesion, provided local conditions do not deteriorate. For the Nordic countries in particular, incorporating a positive diversity focus into already entrenched welfare state structures might prove the most feasible solution.

Despite its findings, the study leaves open several questions for future research and debate. One of the study’s distinct limitations is that it reduces historical contingencies to quantified values. While this is necessary for testing historical hypotheses with a large cross-national sample, quantification simplifies historical complexities. Findings from this study should be taken as a point of departure for more in-depth historical studies of the ethnic boundaries, perhaps in a smaller sample of differently positioned occupiers or colonizers. Second, because the study uses multilevel statistical modeling, the number of macro-level variables included in models must be kept to a minimum to avoid exhausting sufficient degrees of freedom for meaningful inference. While supplementary analyses show that the inclusion of important macro-level controls does not negate findings, a comprehensive analysis of additional controls is beyond methodological capabilities. As methods improve in the cross-national comparative field, more extensive tests should be conducted to confirm diversity findings. Relatedly, future research should test findings in different conquest contexts. The current study is limited to a sample of countries active during Europe’s classical colonial period. As diversity increases in neighboring regions, such as Eastern Europe, systems of domination should be examined to determine whether historical trajectories shape diversity effects. Finally, this study brings to the forefront questions regarding native populations’ adjustment to immigrant populations. Social scientists should revisit assumptions that diversity exerts a mechanistic, negative effect on cohesion if integration does not happen naturally. Native populations should also be aware of their histories of conquest, how such histories are perceived by immigrant populations, and the possibilities and challenges associated with drawing on historical experience while avoiding nationalist revisionism. As the demographic tables turn, future research should continue to illuminate the meaning diversity holds for now majority native groups as migrant groups become a larger part of rich destinations’ histories.

Supplemental Material

SocCoh_post_RR_2_supplementary_materials – Supplemental material for Histories of Conquest, Diversity, and Social Cohesion in Former Colonial Europe

Supplemental material, SocCoh_post_RR_2_supplementary_materials for Histories of Conquest, Diversity, and Social Cohesion in Former Colonial Europe by Aaron Ponce in Social Currents

Footnotes

Acknowledgements

The author thanks the journal’s editors and anonymous reviewers, Patricia McManus, Kody Steffy, Clem Brooks, and Art Alderson for their very thoughtful comments and valuable suggestions. All mistakes are the author’s own.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the National Science Foundation under grant number SES-1519228. Support was also provided by the German Universities Excellence Initiative through the program on migrants’ vocational integration hosted by Technische Universität Dresden.

Notes

Supplemental Material

Supplemental material for this article is available online.

References

Alba

Richard

. 2005. “Bright vs. Blurred Boundaries: Second-generation Assimilation and Exclusion in France, Germany, and the United States.” Ethnic and Racial Studies 28(1):20–49.

Alesina

Alberto

La Ferrara

Eliana

. 2002. “Who Trusts Others?” Journal of Public Economics 85(2):207–34.

Alesina

Alberto

Glaeser

Edward

. 2004. Fighting Poverty in the U.S. and Europe: A World of Difference. New York: Oxford University Press.

Bail

Christopher A.

2008. “The Configuration of Symbolic Boundaries against Immigrants in Europe.” American Sociological Review 73(1):37–59.

Barth

Fredrik

. 1969. Ethnic Groups and Boundaries: The Social Organization of Cultural Difference. Boston, MA: Little, Brown.

Beaman

Jean

. 2016. “As French as Anyone Else: Islam and the North African Second-generation in France.” International Migration Review 50(1):41–69.

Bhambra

Gurminder K.

2014. Connected Sociologies. London, England: Bloomsbury Academic.

Bleich

Erik

. 2003. Race Politics in Britain and France. Cambridge, England: Cambridge University Press.

Boeri

Tito

Hanson

Gordon

McCormick

Barry

. 2002. Immigration Policy and the Welfare System. New York: Oxford University Press.

10.

Brady

David

Finnigan

Ryan

. 2014. “Does Immigration Undermine Public Support for Social Policy?” American Sociological Review 79(1):17–42.

11.

Brubaker

William Rogers

. 1992. Citizenship and Nationhood in France and Germany. New York: Columbia University Press.

12.

Burgoon

Brian

. 2014. “Immigration, Integration, and Support for Redistribution in Europe.” World Politics 66:365–405.

13.

Castles

Stephen

de Haas

Hein

Miller

Mark J

. 2014. The Age of Migration: International Population Movements in the Modern World. New York: Guilford Press.

14.

Costa

Dora L.

Kahn

Matthew E

. 2003. “Civic Engagement and Community Heterogeneity: An Economist’s Perspective.” Perspectives on Politics 1:103–11.

15.

Crepaz

Markus M. L.

2008. Trust beyond Borders: Immigration, the Welfare State, and Identity in Modern Societies. Ann Arbor, MI: University of Michigan Press.

16.

Delhey

Jan

Newton

Kenneth

. 2005. “Predicting Cross-national Levels of Social Trust: Global Pattern or Nordic Exceptionalism?” European Sociological Review 21(4):311–27.

17.

Dinesen

Peter Thisted

Sønderskov

Kim Mannemar

. 2015. “Ethnic Diversity and Social Trust: Evidence from the Micro-context.” American Sociological Review 80(3):550–73.

18.

Eger

Maureen A.

2010. “Even in Sweden: The Effect of Immigration on Support for Welfare State Spending.” European Sociological Review 26(2):203–17.

19.

Esping-Andersen

Gøsta

. 1990. The Three Worlds of Welfare Capitalism. Princeton, NJ: Princeton University Press.

20.

Gesthuizen

Maurice

van der Meer

Tom

Scheepers

Peer

. 2009. “Ethnic Diversity and Social Capital in Europe: Tests of Putnam’s Thesis in European Countries.” Scandinavian Political Studies 32(2):121–42.

21.

Julian

. 2016. Postcolonial Thought and Social Theory. New York: Oxford University Press.

22.

Greenfeld

Liah

. 1992. Nationalism: Five Roads to Modernity. Cambridge, MA: Harvard University Press.

23.

Hooghe

Marc

Reeskens

Tim

Stolle

Dietlind

Trappers

Ann

. 2009. “Ethnic Diversity and Generalized Trust in Europe: A Cross-national Multilevel Study.” Comparative Political Studies 42(2):198–223.

24.

Janoski

Thomas

. 2010. The Ironies of Citizenship: Naturalization and Integration in Industrialized Countries. Cambridge, England: Cambridge University Press.

25.

Jiménez

Tomás R.

2008. “Mexican Immigrant Replenishment and the Continuing Significance of Ethnicity and Race.” American Journal of Sociology 113(6):1527–67.

26.

Joppke

Christian

Torpey

John

. 2013. Legal Integration of Islam: A Transatlantic Comparison. Cambridge, MA: Harvard University Press.

27.

Kesler

Christel

Bloemraad

Irene

. 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 43:319–47.

28.

Killian

Caitlin

Manohar

Namita N

. 2016. “Highly Skilled Immigrant Women’s Labor Market Access.” Social Currents 3(2):138–59.

29.

Klausen

Jytte

. 2005. The Islamic Challenge: Politics and Religion in Western Europe. New York: Oxford University Press.

30.

Lamont

Michèle

Morning

Ann

Mooney

Margarita

. 2002. “Particular Universalisms: North African Immigrants Respond to French Racism.” Ethnic and Racial Studies 25(3):390–414.

31.

Lentin

Alana

. 2004. Racism and Anti-racism in Europe. Ann Arbor, MI: University of Michigan Press.

32.

Letki

Natalia

Evans

Geoffrey

. 2005. “Endogenizing Social Trust: Democratization in East-Central Europe.” British Journal of Political Science 35(3):515–29.

33.

Mau

Steffen

Burkhardt

Christoph

. 2009. “Migration and Welfare State Solidarity in Western Europe.” Journal of European Social Policy 19(3):213–29.

34.

Mewes

Jan

Mau

Steffen

. 2012. “Unraveling Working class Welfare Chauvinism.” Pp. 119–57 in Contested Welfare States: Welfare Attitudes in Europe and beyond, edited by Svallfors

. Stanford, CA: Stanford University Press.

35.

Modood

Tariq

Triandafyllidou

Anna

Zapata-Barrero

Ricard

, eds. 2006. Multiculturalism, Muslims and Citizenship: A European Approach. London, England: Routledge.

36.

Moody

James

White

Douglas R

. 2003. “Structural Cohesion and Embeddedness: A Hierarchical Concept of Social Groups.” American Sociological Review 68(1):103–27.

37.

Paxton

Pamela

. 2002. “Social Capital and Democracy: An Interdependent Relationship.” American Sociological Review 67(2):254–77.

38.

Paxton

Pamela

. 2007. “Association Memberships and Generalized Trust: A Multilevel Model across 31 Countries.” Social Forces 86(1):47–76.

39.

Pew Research Center. 2011. “The Future of the Global Muslim Population: Projections for 2010–2013.” Washington, DC: Pew Forum on Religion & Public Life.

40.

Pierson

Paul

. 1994. Dismantling the Welfare State? Cambridge, England: Cambridge University Press.

41.

Putnam

Robert D.

1993. Making Democracy Work: Civic Traditions in Modern Italy. Princeton, NJ: Princeton University Press.

42.

Putnam

Robert D.

2000. Bowling Alone: The Collapse and Revival of American Community. New York: Simon & Schuster.

43.

Putnam

Robert 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.

44.

Reeskens

Tim

Hooghe

Marc

. 2008. “Cross-Cultural Measurement Equivalence of Generalized Trust: Evidence from the European Social Survey (2002 and 2004).” Social Indicators Research 85(3):515–32.

45.

Reeskens

Tim

van Oorschot

Wim

. 2012. “Disentangling the ‘New Liberal Dilemma’: On the Relation between General Welfare Redistribution Preferences and Welfare Chauvinism.” International Journal of Comparative Sociology 53(2):120–39.

46.

Rusu

Horaţiu

Gheorghita

Andrei

. 2014. “Transnational Solidarity and Public Support for the EU Enlargement.” Sociologia 46(3):261–82.

47.

Savelkoul

Michael

Gesthuizen

Maurice

Scheepers

Peer

. 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):1091–1107.

48.

Scheepers

Peer

Gijsberts

Mérove

Coenders

Marcel

. 2002. “Ethnic Exclusionism in European Countries: Public Opposition to Civil Rights for Legal Migrants as a Response to Perceived Ethnic Threat.” European Sociological Review 18(1):17–34.

49.

Schmidt-Catran

Alexander W.

Fairbrother

Malcolm

. 2016. “The Random Effects in Multilevel Models: Getting Them Wrong and Getting Them Right.” European Sociological Review 32(1):23–38.

50.

Schneider

Silke L.

2008. “Anti-immigrant Attitudes in Europe: Outgroup Size and Perceived Ethnic Threat.” European Sociological Review 24(1):53–67.

51.

Snijders

Tom A. B.

Bosker

Roel J

. 2012. Multilevel Analysis: An Introduction to Basic and Advanced Multilevel Modeling. Thousand Oaks, CA: Sage Publications.

52.

Soroka

Stuart N.

Helliwell

John F.

Johnston

Richard

. 2007. “Measuring and Modelling Interpersonal Trust.” Pp. 95–132 in Social Capital, Diversity, and the Welfare State, edited by Kay

F. M.

Johnston

. Vancouver: University of British Columbia Press.

53.

Svallfors

Stefan

. 2012. “Welfare States and Welfare Attitudes.” Pp. 1–24 in Contested Welfare States: Welfare Attitudes in Europe and beyond, edited by Svallfors

. Stanford, CA: Stanford University Press.

54.

Triandafyllidou

Anna

. 2001. Immigrants and National Identity in Europe. New York: Routledge.

55.

Uslaner

Eric M.

2002. The Moral Foundations of Trust. New York: Cambridge University Press.

56.

van Amersfoort

Hans

van Niekerk

Mies

. 2006. “Immigration as a Colonial Inheritance: Post-colonial Immigrants in the Netherlands, 1945–2002.” Journal of Ethnic and Migration Studies 32(3):323–46.

57.

van der Meer

Tom

Tolsma

Jochem

. 2014. “Ethnic Diversity and Its Effects on Social Cohesion.” Annual Review of Sociology 40(1):459–78.

58.

van Oorschot

Wim

. 2006. “Making the Difference in Social Europe: Deservingness Perceptions among Citizens of European Welfare States.” Journal of European Social Policy 16(1):23–42.

59.

Vergolini

Loris

. 2011. “Social Cohesion in Europe: How Do the Different Dimensions of Inequality Affect Social Cohesion?” International Journal of Comparative Sociology 52(3):197–214.

60.

Waters

Mary C.

1999. Black Identities: West Indian Immigrant Dreams and American Realities. New York: Russell Sage Foundation.

61.

Wimmer

Andreas

. 2008. “The Making and Unmaking of Ethnic Boundaries: A Multilevel Process Theory.” American Journal of Sociology 113(4):970–1022.

62.

Wimmer

Andreas

. 2013. Ethnic Boundary Making: Institutions, Power, Networks. Oxford, England: Oxford University Press.

63.

Ziller

Conrad

. 2015. “Ethnic Diversity, Economic and Cultural Contexts, and Social Trust: Cross-Sectional and Longitudinal Evidence from European Regions, 2002–2010.” Social Forces 93(3):1211–40.

64.

Zolberg

Aristide R.

Woon

Long Litt

. 1999. “Why Islam Is like Spanish: Cultural Incorporation in Europe and the United States.” Politics & Society 27(1):5–38.

65.

Zølner

Mette

. 2000. Re-imagining the Nation: Debates on Immigrants, Identities and Memories. Brussels, Belgium: P.I.E.-Peter Lang.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.52 MB