Migration, Diversity, and the Welfare State
KeywordsWelfare State Ethnic Diversity Public Attitude Social Program Welfare Reform
One of the most compelling challenges facing western democracies is how to maintain and strengthen the bonds of community in ethnically diverse societies. How can we reconcile growing levels of migration and multicultural diversity with the sense of a common identity and mutual support which underpins generous welfare states? In recent decades, a growing number of analysts argue that migration and growing ethnic diversity erode trust and a sense of community among citizens and that contemporary democracies face a trade-off between the accommodation of ethnic diversity on the one hand and support for redistribution on the other. This concern has been labeled the “progressive’s dilemma” (Goodhart 2004), which holds that reformist forces face a tough choice between pursuing greater redistribution or supporting a more diverse, multicultural society.
If these trade-offs are real, the welfare state is in considerable trouble. Migration is a reality in virtually all western democracies; and there is no reason to assume that ethnic minorities will stop pressing for recognition and accommodation of their differences. But are these tensions real? More importantly, are they universal? Or do the cases of tension that we do observe reflect particular contexts and circumstances?
This essay examines the evolution of research and scholarly debate on these questions. It begins by setting the debate in the larger context of research on the welfare state and then summarizes the trends in recent literature. A concluding section then draws together the general patterns and suggests some implications for future research.
Origins of the Debate
Students of social policy have long argued that the welfare state was built on and can only be sustained by a strong sense of community and associated feelings of trust, reciprocity, and mutual obligation. An early expositor of this view was T.H. Marshall, who argued that the expansion of social rights reflected the emergence of a national consciousness in Britain. “Citizenship,” he argued in an oft-quoted passage, “requires a bond of a different kind, a direct sense of community membership based on loyalty to a civilisation that is a common possession” (Marshall 1950/1992, p. 8). In recent years, however, analysts have increasingly argued that ethnic/racial diversity will erode the sense of a common community. They fear that the public may withdraw support from social programs that redistribute resources to people they regard as “strangers,” or “outsiders” who are not part of “us.”
Traditionally, researchers interested in the welfare state paid relatively little attention to such factors (Swank 2002; Huber and Stevens 2001; Esping-Andersen 1990). However, the issues surfaced in other fields. For example, development economists increasingly pointed to ethnic and tribal diversity in attempting to explain the poor economic performance and limited provision of public goods such as education in a number of developing countries, especially in Africa (Easterly 2001).
More importantly for current purposes, analysts began to extrapolate from US experience. The crippling effects of race run through the history of social policy in that country, from Civil War pensions to the Social Security Act in 1935 to the Great Society programs in the 1960s to welfare reform in the 1990s (Schram et al. 2003). Traditionally, the politics of race was seen as contributing to American exceptionalism (Lipset and Marks 2000). Increasingly, however, the racialized dimension of US welfare politics is seen as evidence of a normal, even inevitable, reaction to ethnic/racial heterogeneity and a warning to other countries being transformed by migration. Gary Freeman described migration as a “disaster” for the welfare state and predicted that it would lead to “the Americanization of European welfare politics” (Freeman 1986: 62). Alesina and Glaeser worried that “As Europe has become more diverse, Europeans have increasingly been susceptible to exactly the same form of racist, anti-welfare demagoguery that worked so well in the United States. We shall see whether the generous welfare state can really survive in a heterogeneous society” (Alesina and Glaeser 2004, pp. 180–181).
Since the early 2000s, research on these issues has exploded. Scholars in more and more countries are asking whether such warnings might be right. They also ask what governments can do to mitigate any tensions between migration and social solidarity. Not surprisingly, perhaps, the rapidly growing literature has generated a diverse range of findings and uncovered a terrain of considerable complexity. Nevertheless, some general patterns are beginning to emerge. We explore them here in several categories: public attitudes, political processes, the design of public policy, and social policy outcomes.
Does migration and ethnic diversity weaken support for redistribution among the general public? In the US, Robert Putnam famously concluded that trust and community engagement, which he argues is critical for collective action, is weakened by ethnic diversity; individuals living in more homogeneous parts of the country are in effect “hunkering down” in social isolation (Putnam 2007). Putnam did not extend his analysis directly to support for welfare spending. But Luttmer concludes that support for welfare spending is lower in diverse neighborhoods (Luttmer 2001); Gilens has demonstrated how the interaction between racial attitudes and media-driven images of poor blacks as lazy explains “why Americans hate welfare” (Gilens 1999); and Fox (2004) has found that a greater proportion of Latinos in a state is correlated with lower public support for welfare spending.
Comparative research, however, suggests that US story is not universal. Cross-national findings have been mixed at best, with contradictory results reflecting different outcome measures, different migrant populations, different control variables, and different time periods. Nonetheless, cross-national studies tend to find either no relationship or, at most, a weak negative relationship between migration and public support for the welfare state. Certainly, negative effects seem smaller than one might expect from simply extrapolating from US experience and tend to fade when a number of moderating factors are included in the analysis. Other factors are clearly more important in predicting welfare attitudes (See, e.g., Brady and Finnigan 2014; Senik et al. 2009; Finseraas 2008; Mau and Burkhardt 2009; Crepaz 2008).
Case studies of other individual countries are similarly varied. In Canada, one the most diverse countries among western democracies, Soroka et al. (2006) find that although interpersonal trust tends to be lower in racially diverse neighborhoods, echoing Putnam, trust in government institutions and support for the welfare state are not strongly related to racial diversity. A more recent study qualifies this rosy image of multicultural Canada: perceptions of heavy migrant use of social assistance are not associated with a general rejection of redistribution, but perceptions of welfare dependence among aboriginal people are (Banting et al. 2013). In Germany, Stichnoth(2012) found greater opposition when he asked whether the native population supported state help for the unemployed in regions where the proportion of foreigners among the unemployed is high. However, after adjusting for individual differences such as income and East German origins, the negative association between migration and public support was relatively weak.
Not surprisingly, perhaps, the effects of ethnic diversity are especially sensitive in Scandinavia. The Nordic model of the welfare state was constructed in the context of striking homogeneity, but migration is transforming these countries, most notably Sweden, where the proportion of the population born outside of the country has grown the fastest. Once again, the evidence is mixed. Gerdes and Wadensjö (2008) study the ways in which the distribution of refugees across Danish municipalities affect voting patterns, and find no clear indication of a decline in support for the welfare state. In contrast, two studies in Sweden are less reassuring. Eger (2010) finds a negative association between the proportion of migrants in different regions and support for social spending, although the relationship is much weaker for universal social programs such as health care; and Dahlberg et al. (2012) report a negative impact of migration from nonwestern countries on public support for social benefits.
In retrospect, the idea that the public would abandon social programs they value simply because of migration seems a bit implausible. Not surprisingly, several alternative hypotheses have emerged. Some studies have found support for a “compensation hypothesis,” which holds that anxiety about migration actually increases support for the welfare state. High levels of migration generate economic insecurity in the majority population, especially among low-skilled workers, increasing their support for social programs (Finseraas 2008; Brady and Finnigan 2014). In effect, antimigrant attitudes and proredistribution views can and often do go together.
Importantly, however, the compensation effect does not guarantee an inclusionary approach to social solidarity. In many countries, there is considerable support among the majority population for maintaining the welfare state but denying its benefits to newcomers. Migrants are consistently seen as the least “deserving” group of beneficiaries (van Oorshot 2006), and there is substantial support for what is, in effect, a two-tier welfare state, generous for natives but restricted for migrants. This exclusionary approach, often called “welfare chauvinism,” can take several forms: lower levels of support for social programs thought to disproportionately benefit migrants and support for lengthy residency requirements which delay access to general programs for newcomers (Sainsbury 2012; Reeskens and van Oorschot 2012; Koning 2013).
It is worth underscoring that the compensation effect and welfare chauvinism are entirely compatible with each other. Interestingly, Brady and Finnigan (2014) find support for both in their cross-national analysis. Moreover, as we shall see next, this combination of support for the welfare state but exclusion of newcomers from its benefits resonates in the political processes of many countries.
Scholars have asked whether diversity has weakened social movements and political parties which were traditionally the primary champions of the welfare state, indirectly undermining it over time. Once again, the classic case comes from the US. In the 1960s, racial politics swirled around Aid to Families with Dependent Children (AFDC) and the Great Society programs. As welfare rolls expanded, the profile of the poor became racially charged: black families represented close to half of the AFDC caseload, and Hispanic groups were also increasingly over-represented. Resentment against these programs helped fracture the New Deal coalition and the electoral base of the Democratic Party. White union members, white ethnics, and southerners deserted their traditional political home, especially in presidential elections, in part because of its image on race and welfare issues. The effect was so powerful that the Democratic Party sought to insulate itself in the 1990s by embracing hard-edged welfare reforms, which President Clinton signed in 1995. Clinton’s strategy may have helped his re-election in 1996, but it failed to alter Americans’ racialization of and opposition to welfare (Dyck et al. 2008).
Migration has also had a dramatic impact on the party systems in Europe. A potent public backlash against migration has contributed to electoral volatility, voter realignment, and the rise of radical right parties. This backlash has placed immense pressure on mainstream parties, especially social democratic parties which traditionally have supported both the welfare state and cultural diversity. The voters most attracted to radical anti-immigrant parties have been white, low-skilled male voters who traditionally represented an important part of the base of left parties, which might suggest that US commentators about Europe’s future were right. However, as anti-immigrant parties became more dependent on working-class votes, they have moved left, increasingly defending the welfare state for the majority population and pledging to protect it from the “burden” of supporting migrants (Bay et al. 2013; Oesch 2008; Lefkofridi and Michel forthcoming). Welfare chauvinism is now the dominant position among these parties.
The Mediating Role of Policy Regimes
Some researchers have asked whether there is anything states can do to mitigate the progressive’s dilemma. In particular, they have investigated whether the structure of the welfare state itself or the nature of multiculturalism policies matter to the compatibility of migration and solidarity.
Since Marshall (1950), the welfare state has been seen as an instrument of social integration, which can strengthen the sense of cohesion and solidarity in divided societies. Contemporary analysts, however, insist that such integrative potential depends heavily on the actual design of the welfare state. Some argue that heavy reliance on means-tested benefits lock societies into an unending conversation about the legitimacy of particular groups of recipients, such as migrants, while universal programs dampen discussion of which recipients are deserving and which are not. Others caution that easy access to the welfare state for newcomers reduces their incentive to integrate socially and economically and courts majority backlash as newcomers become associated with welfare dependence (Koopmans 2010). To date, the empirical evidence suggests that the public in countries with highly selective welfare states is more inclined to welfare chauvinism and that egalitarian policies and institutions can help in fighting such sentiments (van der Waal et al. 2013; Larsen 2006).
Others argue that support for redistribution depends more on trust in government than interpersonal trust and that such trust is sustained by the quality of government, especially fairness and effectiveness in the administration of government programs (Rothstein 2011). Indeed, Rothstein (forthcoming) argues that a high quality of government institutions generates high levels of social trust even in societies with high levels of ethnic diversity, dampening down welfare chauvinism.
Similar debates arise over the state’s own approach to diversity policy, especially the role of multiculturalism policies. Some countries have responded to growing ethnic diversity with the adoption of multiculturalism policies that recognize distinctive rights or entitlements for ethnic and religious groups. (Despite political rhetoric about a “retreat” from multiculturalism policies, there is strong evidence of their persistence, including in Europe (Banting and Kymlicka 2013; Koopmans et al. 2012).) However, multiculturalism policies have been controversial, and some countries have steadfastly resisted that approach. Not surprisingly, much of the debate has centered on their implications for social solidarity. Critics insist that multiculturalism policies exacerbate the tension between diversity and redistribution, by encouraging identity politics which crowds out redistributive issues from the policy agenda, corrodes trust among vulnerable groups who would otherwise coalesce in a proredistribution lobby, or misdiagnoses the real problems facing minorities, leading them to believe that their problems reflect their culture rather than economic barriers they confront (Barry 2001). Defenders of multiculturalism policies reply that such policies do not create distrust among groups but that such policies can ease intercommunal tensions over time and strengthen the sense of mutual respect, trust, and support for redistribution.
Given the intensity of the debates, there has been surprisingly little empirical research on the impact of multiculturalism policies on support for redistribution. (Schaeffer 2014 notes the virtual absence of such studies. For analyses of the impact of multiculturalism policies on other aspects of integration, see Bloemraad and Wright 2014; Koopmans 2013.) Nevertheless, the studies that are available provide little evidence that multiculturalism policies weaken public support for the welfare state. Countries that have adopted such programs did not experience an erosion of their welfare states or even slower growth in social spending than countries that have resisted such programs (Banting et al. 2006). This finding has been replicated with updated data (available on request) and has been confirmed by Brady and Finnigan 2014.
Finally, research has analyzed the impact of migration, not on public attitudes or the political process but on actual social spending. Once again, there are two possibilities. The first is that migration increases social spending, especially in generous welfare states, because low-income migrants are drawn to countries with expansive social programs and rely on benefits when they arrive (Borjas 1999). The second possibility is that migration decreases social spending, because of the political mechanisms discussed earlier: migration tends to reduce support for social spending generally and/or increase support for excluding migrants from benefits.
As discussed earlier, there is substantial evidence from the US that racial diversity weakens redistribution. Alesina and his colleagues demonstrate that public spending tends to be lower in cities and states with higher levels of racial heterogeneity (Alesina et al. 1999; Alesina and Glaeser 2004). However, a growing body of cross-national research suggests that negative effects on social spending are limited or are offset by other factors. Lipsmeyer and Zhu’s analysis of EU countries (Lipsmeyer and Zhu 2011) suggests that increased migration may increase social spending if left parties are strong or union density is high (also Taylor-Gooby 2005). Other studies suggest that the design of the welfare state is important, not only to public attitudes as discussed earlier but also to the impact of migration on spending levels (Mau and Burkhardt 2009). In short, the relationship between migration and welfare spending is complex and mediated by a variety of factors.
One distinction that deserves more attention is that between levels of migration and change in migration. Consistent with much existing cross-national work, Soroka et al. (forthcoming) find no relationship between the proportion of the population born outside the country and the level of social spending. However, they find a negative relationship between the rate of increase in migration and the rate of increase in social spending over time. Countries with large increases in the proportion of their population born outside the country tend to have smaller increases in social spending. This suggests that public reactions take time to develop. The public needs to notice the change and react, and their reactions need to work through political processes. In addition, Soroka et al. find it useful to disaggregate social spending into subdomains, such as pensions, employment benefits, and so on. In most domains, there is only weak evidence of a relationship between changes in migration and changes in spending, but negative effects do emerge in the case of unemployment benefits and labor market programs.
What can we conclude from the diverse and sometimes contradictory findings that populate this terrain? Most obviously, one needs to be cautious about sweeping statements about a toxic relationship between migration and social solidarity. The evidence is mixed at best. In many studies, there is no impact of migration on public attitudes about social programs or levels of public spending. In other studies, there is evidence of a relationship, but the association is often weak and the direction varies. In general, estimates of the impact of migration, positive or negative, tend to be small, and other factors emerge as more important in explaining the views of the public or the spending decisions of governments. (For a similar conclusion, see Stichnoth and Van der Straeten 2013.) (Similarly, in a related debate not discussed here, estimates of the net impact of migration on the fiscal position of government tend to be small (Rowthorn 2008).
Interestingly, this pattern of mixed findings is consistent with research on the implications of ethnic diversity for social capital or social cohesion more generally. A survey of 464 findings found that “there are nearly as many studies rejecting the negative effects of diversity as arguing for them” (Schaeffer 2014, p. 4 and Table 2.1). A meta-analysis of 90 articles testing Putnam’s conclusions found that 26 articles tend to support his findings, 25 articles reject them, and 39 studies provide mixed or neutral evidence (Van der Meer and Tolsma 2014). In the words of two leading scholars “the debate about the consequences of ethnic diversity on social cohesion has reached a stalemate” (Stolle and Allison 2015, p. 117).
Yet the progressive’s dilemma has not been laid to rest. Serious tension between migration, ethnic diversity, and the welfare state is clearly possible. The key challenge in future research is to identify more systematically the conditions which mitigate or exacerbate such tensions. To advance the debate, we need more differentiated understandings of both of our core concepts, migration and the welfare state. Which types of migrants are seen as “other”? Much of the literature assumes simply being foreign born is what matters. But other studies focus on low-skilled migrants, illegal migrants, racialized minorities, or religious minorities. A systematic comparison of the implications of different categories of migrants in different contexts would help. Similarly, we need more a systematic analysis of the dimensions of the welfare state at stake. Is it total social spending, the generosity of benefits, inclusive access to benefits, or the redistributive impact of the state?
Relatedly, we need a finer-grained understanding of the politics of migration. How does the public decide who is "us" and who is a "stranger”? We need a fuller understanding of “othering” and of how to overcome it. In particular, more attention needs to be paid to the role of the media and political parties in framing migration and newcomers. Much of the “flash” potential of migration as a political issue undoubtedly flows from such processes. Future research should also pay more attention to the political agency of migrants themselves and their attitude to redistribution. As the size of the migrant electorate grows, their views and political behavior will increasingly shape the future (Koopmans et al. 2012). There are likely to be tipping points in this process. When the size of the migrant electorate reaches a critical mass, political parties will find it harder to win power by attacking migrants. Countries such as Canada are already at that stage. The Republican Party in the US is struggling with the electoral calculus. This too will be part of Europe’s future.
Context clearly matters in all of this. Part of that context is the structure of the welfare state itself. The evolution of the welfare state reflects considerable path dependence, and welfare states at different stages of development respond differently to common social pressures. We therefore need to be cautious about simple extrapolations from the US to other contexts. There is considerable difference between the American experience of racial diversity constraining the development of a welfare state from its very beginning and European countries coming to terms with new forms of diversity in the context of mature welfare states which are well embedded in national cultures and voters’ expectations (Crepaz 2008). In the European context, the compensation hypothesis and welfare chauvinism clearly deserve primary attention.
Finally, we should ask why migration has emerged so dramatically in debates about the future of the welfare state. Clearly, the redistributive state is subject to a wide range of pressures, rooted in globalization, economic restructuring, dualization, neoliberalism, and on and on. If the effects of migration appear small at best, why does the issue command so much attention? Is migration really a proxy for wider fears about the future? Does migration policy rise to the surface because it appears to be one final domain in which the nation state, so constrained elsewhere, can actually have an impact?
- Banting K, Johnston R, Kymlicka W, Soroka S (2006) Do multiculturalism policies erode the welfare state? An empirical analysis. In: Keith B, Kymlicka W (eds) Do multiculturalism policies erode the welfare state? Oxford University Press, OxfordGoogle Scholar
- Banting K, Soroka S, Koning E, Edward K (2013) Multicultural diversity and redistribution: immigrants, aboriginal peoples and public support for social programs. In: Banting K, Myles J (eds) Inequality and the fading of redistributive politics. University of British Columbia Press, Vancouver, pp 165–186Google Scholar
- Barry B (2001) Culture and equality: an egalitarian critique of multiculturalism. Harvard University Press, CambridgeGoogle Scholar
- Bay A-H, Finseraas H, West A, Axel W (2013) Welfare dualism in two Scandinavian welfare states: public opinion and party politics. West Eur Polit 36(1):199–220Google Scholar
- Borjas G (1999) Heaven’s door: immigration policy and the american economy. Princeton University Press, PrincetonGoogle Scholar
- Crepaz M (2008) Trust beyond borders. University of Michigan Press, Ann ArborGoogle Scholar
- Easterly W (2001) The elusive quest for economic development. MIT Press, CambridgeGoogle Scholar
- Esping-Andersen G (1990) The three worlds of welfare capitalism. Princeton University Press, PrincetonGoogle Scholar
- Gerdes C, Wadensjö E (2008) The impact of immigration on election outcomes in Danish municipalities. In: Gerdes C (ed) Studying the interplay of immigration and welfare states. Swedish Institute for Social Research, SwedenGoogle Scholar
- Goodhart, David (2004) Too diverse? Prospect, February, pp 30–37Google Scholar
- Koning E (2013) Selective solidarity: the politics of immigrants’ social rights in Western welfare states. Ph.D. dissertation, Queen’s University, KingstonGoogle Scholar
- Larsen CA (2006) The institutional logic of welfare attitudes: how welfare regimes influence public support. Ashgate, AldershotGoogle Scholar
- Lefkofridi Z, Michel E (Forthcoming) The electoral politics of solidarity. In: Banting K, Kymlicka W (eds) The strains of commitment: the political sources of solidarity in diverse societies. Oxford University Press, OxfordGoogle Scholar
- Lipset SM, Marks G (2000) It didn’t happen here; why socialism failed in the United States. Norton, New YorkGoogle Scholar
- Marshall TH (1950/1992) Citizenship and social class. In: Marshall TH, Bottomore T (eds) Citizenship and social class. Pluto Press, LondonGoogle Scholar
- Rothstein Bo (Forthcoming) Solidarity, diversity and the quality of government. In: Banting K and Kymlicka W (eds) The strains of commitment: the political sources of solidarity in diverse societies. Oxford University Press, OxfordGoogle Scholar
- Schaeffer M (2014) Ethnic diversity and social cohesion: immigration, ethnic fractionalization and potentials for civic action. Ashgate, FarnhamGoogle Scholar
- Schram SF, Soss J, Fording RC (2003) Race and the politics of welfare reform. University of Michigan Press, Ann ArborGoogle Scholar
- Soroka S, Richard J, Keith B (2006) Ethnicity, trust and the welfare state. In: Fiona K, Richard J (eds) Social capital, diversity and the welfare state. University of British Columbia Press, Vancouver, pp 279–303Google Scholar
- Soroka S, Banting K, Johnston R, Kevins A, Kymlicka W (Forthcoming) Migration and welfare state spending. Eur Polit Sci Rev. First view online: 3 March 2015Google Scholar
- Stolle D, Harell A (2015) The consequences of ethnic diversity: advancing the debate. In: Koopmans R, Lancee B, Schaeffer M (eds) Social cohesion and immigrant in Europe and North America. Routledge, New York, pp 105–122Google Scholar
- van der Waal J, De Koster W, Van Oorschot W (2013) Three Worlds of welfare chauvinism? J Comp Policy Anal 15(2):164–181Google Scholar