By Keith Banting, Richard Johnston, Will Kymlicka and Stuart Soroka
During the second half of the 20th century, the hope of progressives in Canada and elsewhere was to build a society that was both caring and diverse. To nurture a caring society based on social solidarity, they built a welfare state designed to protect citizens from the risks and insecurities inherent in a market economy. To embrace diversity, they adopted a more expansive immigration program as well as policies that gave greater public recognition to difference through Aboriginal rights, official language policies and multiculturalism.
During the first decade of the twenty-first century, critics have increasingly deemed this progressive agenda doomed. Growing diversity, it is claimed, weakens the willingness of majorities to support the weak, since the newcomers are seen as “other,” not part of “us.” The policy implications of such conclusions are troubling. The choice is between limiting immigration and reducing public recognition of ethnic and linguistic difference, on one hand, and accepting erosion of the welfare state on the other.
Over a number of years we have conducted research focused on testing the feasibility of the progressive agenda. What forms of diversity and multiculturalism, if any, erode which forms of solidarity? Under what conditions? What steps can governments take to mitigate this effect? While ongoing, our research to date suggests that the original progressive agenda of combining solidarity and diversity is still viable, and that the tradeoffs between diversity and solidarity are both more localized and more manageable politically than widely believed. Further, our results suggest that the diversity-solidarity tradeoff is less salient in Canada than elsewhere.
To show this, we first set our approach in the context of the international debate surrounding the progressive agenda, distinguishing between three key forms of social solidarity – democratic, civic and redistributive. Then we highlight our own findings, starting from a cross-national portrait across the advanced countries that make up the Organisation for Economic Co-operation and Development (OECD) and then moving the focus to the Canadian experience.
The solidarity/diversity debate
A new pessimism pervades international debates about the consequences of ethnic diversity. There is a widespread perception that ethnic and religious diversity erodes various forms of solidarity, with some commentators portraying the tensions as universal, inherent in the nature of human relations. Starting in the early 1990s, a number of large-scale cross-national studies suggested that countries with high levels of ethnic diversity were susceptible to a wide range of pathologies: they were more prone to violent conflicts, were less likely to develop into democracies, were less likely to enact redistributive welfare policies, displayed lower levels of trust and so on.
These findings are being challenged. More recent research has questioned the idea of a universal or inevitable tension between diversity and solidarity. Instead, the relationship between diversity and solidarity turns out to be contingent on the nature of the diversity, the larger socioeconomic context and the political structures which contain and manage diversity. These newer studies have challenged the idea that ethnically diverse societies are inherently more prone to civil wars,1 less likely to be democracies,2 less likely to respect human rights or less able to sustain social redistribution.3
Nevertheless, a vague pessimism lingers. Moreover, there remains an undeniable potential for deep-seated tensions between ethnic groups, leaving us with a compelling research agenda. We need to understand the diversity-solidarity relationship better and explore the factors that may mediate between ethnic diversity and solidarity. What are the intervening factors that either generate or mitigate tensions between the two? What factors contribute to solidarity amid diversity?
In considering the relationship between diversity and the welfare state, we cannot rule out the arguments advanced by pessimists as intrinsically implausible. They worry that ethnic and racial diversity will erode the sense of community and common identity and weaken feelings of trust in fellow citizens, with potentially debilitating consequences for social policy. They worry that growing diversity will fragment the historic coalitions that supported the welfare state and split emerging groups that might otherwise coalesce around redistributive agendas. They worry that members of the majority public might withdraw electoral support from social programs that redistribute resources to people they regard as “strangers” or “outsiders” who are not part of “us.” They cite apparently increasing support in Europe for conservative or radical right parties that oppose immigration.
Evidence to support these arguments has emerged from very different settings. First, development economists increasingly point to ethnic and tribal diversity in attempting to explain the poor economic and social performance of a number of developing countries, especially in Africa. Initially, the focus was on the impact of heterogeneity on economic growth, but subsequent research has extended the analysis to the impact of heterogeneity on the provision of public goods such as education.4
Second, studies of the politics of social policy in the United States provide substantial evidence of racial diversity weakening redistribution. For example, in a 2001 study, Alberto Alesina, Reza Baquir and William Easterly demonstrated that public spending tends to be lower in cities and states with higher levels of racial heterogeneity, even when other relevant factors are held constant,5 and their results have been replicated by others.6 Alesina and Edward Glaeser have extended this approach to cross-national differences.7 They conclude that almost half of the difference in social spending between the United States and European countries can be explained by differences in the level of racial diversity.
But there are other factors that mitigate the tension between diversity and solidarity. A number of candidates have been put forth. The impact of immigration on solidarity may depend, for example, on whether immigrants are seen as posing an economic threat, on whether there are shared associational and political forums for native-born and immigrants to meet and learn about each other and on whether a shared national identity is developed that encompasses both immigrants and native-born. Hence, where policies and practices are put in place that reduce perceptions of economic threat, reduce social isolation, promote political participation and build inclusive shared identities, then the potentially corrosive effect of diversity may be mitigated or even reversed.
Yet these suggestions remain largely speculative. We have little firm evidence about the extent to which these different factors do mediate the impact of diversity on solidarity. And perhaps because mediating factors have been understudied, we have conflicting evidence regarding the basic diversity-solidarity relationship. The aim of our research has been to expand what we know.
Three versions of solidarity
In order to make progress on these issues we need first to be precise about what we mean by “social solidarity.” Our approach distinguishes three different dimensions of this broad notion:
Democratic solidarity is characterized by support for basic human rights and equalities and an inclusive approach to democratic decision-making, including acceptance of: equal participation of citizens from all backgrounds, tolerance for the political expression of diverse cultural views consistent with basic rights and equalities, and acceptance of compromises among legitimate contending interests.
Civic solidarity is characterized by an openness to newcomers, mutual tolerance and acceptance of people of diverse ethnicities, languages and religions as legitimate members of the community – as part of “us.”
Redistributive solidarity is characterized by support for redistribution toward the poor and vulnerable groups; support for the full access of people of all backgrounds, including newcomers, to core social programs; and support for programs that recognize and accommodate the distinctive needs and identities of different ethnic groups.
We have focused on these three because they are, in our view, inherently valuable features of a society, and essential if a society is to be minimally decent and just. And yet none can be taken for granted. All three may require individuals to act against their initial inclinations and self-interest, or at least to exercise self-restraint in the pursuit of those interests and beliefs. All three forms of solidarity, therefore, must continually be nurtured.
It is important to note that our definition of solidarity is not as wide as that in certain other discussions. For example, some go beyond civic, political and redistributive solidarity to include levels of interpersonal trust, strength of national identity and levels of civil society participation.8 On these broader accounts, solidarity is not just about civic tolerance, commitment to a pluralistic democracy and redistribution, but also about trusting your neighbour, feeling a sense of pride and belonging in your country and being an active participant in social and political life.
In our view, these additional dimensions of trust, identity and participation are certainly important, and may indeed be essential to ensuring a decent and just society. It may be impossible to achieve civic, political and redistributive solidarity without the right sorts of trust, identity and participation. For example, interpersonal trust and trust in government may contribute to tolerance, effective democratic governance or support for redistribution; and a shared sense of national identity may increase social inclusiveness, enhance democratic governance or reinforce support for transfers to the poor.
But these relations are, we believe, conditional and contingent. Societies that exhibit high levels of trust, national pride or civic participation may not in fact be particularly tolerant of minorities, or particularly solicitous of the poor, or even particularly democratic. For this reason, we consider these as “intervening variables” or “mediating variables,” in order to study their effects on the more foundational values of civic, political and redistributive solidarity. We thus treat trust, identity and participation as potential sources of social solidarity, rather than as elements of solidarity itself. And in this article we focus on the third form, redistributive solidarity, where our work is more advanced.
The cross-national view: Diversity, multiculturalism and redistribution
We start at 30,000 feet, surveying the broad relationships between immigration, multiculturalism policies and social spending across OECD countries. We seek to answer two deceptively simple questions. First, are higher levels of immigration associated with lower levels of social spending? Second, does the government’s adoption of a multiculturalism policy affect that relationship? Answering these questions is a complex task, as one needs to take into account all of the other factors that influence social spending – the proportion of the population over age 64, the percentage of women in the labour force, the strength of left parties and the like.9
The picture that emerges is mixed. On the one hand, there is no clear relationship between the proportion of the population born outside the country and growth in social spending over the last three decades of the 20th century. On the other hand, the pace of change does seem to matter. Figures 1 and 2 display the basic data: the first shows the relationship between levels of diversity and spending, and the latter shows the link between changes in these variables. In the first case, there is a slight negative relationship, but it is only slight and statistically insignificant. The relationship is more striking in the second case, where greater changes in diversity are clearly linked to smaller increases in spending. In the final analysis, it may be the pace of social change rather than the existence of difference that has the potential to unsettle political life.
The unsettling effects of change pose important questions about how states should manage periods of demographic transition. What is the role of public policy? Does the policy response of government to ethnic diversity matter when it comes to the resiliency of the redistributive state?
Historically, Western states tended to view immigrant ethnic identities with indifference or suspicion, and sought to assimilate newcomers into a common national culture. During the last decades of the 20th century, however, many states increasingly accepted some obligation to accommodate such identities, adopting what have become known as “multiculturalism policies.” The essence of multiculturalism policies is that they go beyond the protection of basic civil and political rights guaranteed to all individuals in a liberal-democratic state to extend some level of public recognition and support to ethnocultural minorities to maintain and express their distinct identities and practices.10 This trend sparked lively debate about the nature of citizenship and rights in diverse societies, and countries vary in the extent to which they adopted this approach. Moreover, several countries, including the Netherlands and Britain, have seen some retrenchment of the multicultural approach in recent years.
Some theorists insist that multiculturalism policies exacerbate the tensions underlying tradeoffs between diversity and redistribution. They argue that such policies trigger a politics that crowds redistributive issues out of the policy agenda, corrodes trust among vulnerable groups which would otherwise coalesce into a pro-redistribution lobby, or misdiagnoses the real problems facing minorities, leading them to believe that their problems emerge from their culture rather than economic barriers that they confront along with vulnerable members of many other cultural groups.11 Defenders of multiculturalism reply that distrust is the historical legacy of indifference to or repression of ethnic differences by earlier generations, and that rather than creating distrust among groups, multiculturalism policies can ease intercommunal tensions over time, and strengthen the sense of mutual respect and trust which sustains redistribution.12
In our research we have sought to subject these often anecdotally based assertions to more systematic empirical evidence. We asked whether the welfare states of countries that have adopted strong multicultural policies over the last two decades have been more fragile than those of countries that have resisted such policies. The first step, following work by Keith Banting and Will Kymlicka,13 was to classify OECD countries in terms of the relative strength of eight policies most commonly associated with this approach:
- constitutional, legislative or parliamentary affirmation of multiculturalism;
- explanation/celebration of multiculturalism in the school curriculum;
- inclusion of ethnic representation/sensitivity in the mandate of public media or media licensing;
- exemptions from dress codes, Sunday-closing legislation, etc.;
- acceptance of dual citizenship;
- funding of ethnic group organizations or activities;
- funding of bilingual education or mother-tongue instruction;
- affirmative action for disadvantaged immigrant groups.
The first three policies celebrate multiculturalism, the middle two reduce legal constraints on diversity and the final three represent forms of active support for minority communities and individuals. Countries that adopted six or more of these policies were classified as “strong,” countries that adopted no more than two were classified as “weak” and those falling in between were categorized as “modest.”
The second step was to examine how the welfare state in the three categories fared between 1980 and 2000. Did countries that adopted strong multiculturalism policies fare worse than countries that resisted such approaches in maintaining and enhancing their welfare states over the two decades? We found no evidence that this was the case. Countries that adopted such programs did not experience an erosion of their welfare states or even slower growth in social spending than countries that resisted them. Indeed, social spending and the redistributive impact of taxes and transfers on child poverty and inequality increased more in countries with the most developed multiculturalism policies, providing a hint that perhaps multiculturalism policies may actually ease the tension between diversity and redistribution.14
This lack of a systematic tension between recognition of diversity and redistribution is confirmed by subsequent analysis.15 Our evidence to date thus provides no support for the bald claim that “a politics of multiculturalism undermines a politics of redistribution.”16 This leaves open the question of whether the expected corrosive effects of multiculturalism policies only emerge over time, as suggested by Philippe Van Parijs in response to our findings.17 However, the 20-year period was long enough for other political factors, such as the role of left-wing parties, to emerge strongly in the multivariate analysis, as did – more tellingly perhaps – the effects of change in the proportion of the population born outside the country.
While much remains to be explored about this relationship, we can conclude from our vantage point at 30,000 feet that it is simply not true that multiculturalism inevitably corrodes solidarity.
The Canadian view: Diversity, solidarity and mediating factors
To round out our perspective, we need to descend to the experience of specific countries. Our concentration is on Canada, which is one of the most multicultural countries in the world.18 Sustaining solidarity while accommodating diversity has been a central task in Canada since the emergence of a single state in northern North America, and a variety of strategies have been employed. Clearly, Canada provides an important test case for identifying which forms of solidarity are affected by diversity, and which factors might mediate that relationship.
In thinking about mediating factors, it is helpful to distinguish between factors that might explain why diversity erodes solidarity and factors that might sustain solidarity in the face of diversity. In the first category, the most influential work, by far, has focused on interpersonal trust and civic engagement. Robert Putnam has argued, on the basis of research in the United States, that ethnic and racial diversity erodes trust, leading people to “hunker down” in social isolation and lose faith in government solutions to social problems.19 We therefore test two related propositions: that ethnic diversity erodes feelings of trust in one’s neighbours, and that this weakens support for social redistribution.20
To test the first proposition – that ethnic diversity erodes interpersonal trust – our survey adopted an innovative measure of interpersonal trust known as the “wallet question.” Respondents were asked, “Say you lost a wallet or purse with $100 in it. How likely is it that the wallet or purse will be returned with the money in it if it was found by a neighbour?” They were then asked the question three more times, with “a police officer”, “a clerk at the local grocery store” and “a stranger” substituting for the neighbour in turn. This measure does reveal a tension between the ethnic diversity of the neighbourhoods in which Canadians live and the level of trust they have in their neighbours. The larger the presence of visible minorities in the neighbourhood, the less trusting is the majority even when one controls for other factors that influence trust levels, such as economic well-being, education, gender and age.
Members of racial minorities, in contrast, are much less trusting where the majority is very dominant, and their trust levels rise with the ethnic diversity of their neighbourhood. In this particular analysis, the two lines crossed when the racial minority percentage was about 60 per cent. Above that level, the average racial minority respondent was more interpersonally trusting than his or her majority counterpart. So, sustaining trust across racial differences does appear to be a challenge even in the most multicultural of countries, a pattern that parallels the United States.
Even here, however, the tension between diversity and interpersonal trust is hardly universal. Important social factors limit its reach. For example, among white majority respondents, the effects are strongest for those who have lived in a neighbourhood for some time – they have watched diversity grow around them, and are more likely to be concerned about it.21 Those who have moved more recently have “selected in” to that diverse neighbourhood, and are much less negatively affected by diversity.
Figure 3 shows predicted levels of trust for white “majority” respondents (using the combined wallets questions, rescaled from 0 to 1), across different levels of neighbourhood diversity.22 One line shows the average effect of diversity on respondents who have not moved recently; the other shows the effect for those who have moved. Note first that the impact of diversity is relatively small. A change from a neighbourhood that is 0 per cent visible minority to one that is 70 per cent visible minority shifts this version of trust downward from 0.7 to 0.55 – and only for those who have not moved. For the often sizable proportion of respondents who have moved into the neighbourhood, the effects of diversity are negligible.
Relatedly, in a comparative study of the United States and Canada, Dietlind Stolle, Stuart Soroka and Richard Johnston found that diversity affects trust only for those who have little interaction with their neighbours.23 Those who live in diverse settings but interact with their neighbours more regularly are not adversely affected by diversity. Supplementary tests suggest that this is not a case of reverse causation – it is not the case that high trustees speak to their neighbours more. Interactions with diverse others makes diversity less threatening.
More important for current purposes is our test of the second proposition: that neighbourhood diversity also weakens support for the welfare state. Many analysts stop before this step, and simply assume that diminished interpersonal trust necessarily weakens support for redistribution. This turns out not to be true, at least not in any direct way.
In addition to measuring trust levels, we explored respondents’ support for the welfare state through a battery of questions about specific social programs. Analysis of the data revealed virtually no relationship between respondents’ own ethnicity or the ethnic complexion of their neighbourhoods on one hand, and their support for social programs on the other, which is not the case for income, gender and age, all of which do influence support for social spending. Moreover, to the extent that there are even hints of a relationship, it is the minorities, not the majorities, that are less supportive of redistribution. There is no evidence of majorities turning away from redistribution because some of the beneficiaries are “strangers.” While the evidence is preliminary and the research project remains underway, there is little evidence to date to support a tradeoff between diversity and redistribution.
If diversity does not weaken the welfare state by corroding trust, are there mediating factors which strengthen support for redistribution in the context of diversity? One argument24 supported by our data is that it is institutional trust rather than interpersonal trust that is critical to social solidarity, and that institutional trust is less sensitive to diversity.25
But perhaps most intriguing are our findings concerning national identity.26 Liberal nationalist theorists contend that strong national identities both mitigate opposition to redistribution among high-income earners and reduce the corrosive effects of immigration.27 Our findings support this contention, rather than that of those who view nationalism as chauvinism and, as such, hostile to redistributive solidarity, all the more so if redistribution is to groups vulnerable to being cast as outsiders. Identification with Canada increases tolerance for immigrants and support for the welfare state, especially among the more affluent among us. Figure 4 shows the effects of economic status on support for redistribution and health care for those with high versus low levels of national identity.28 In both cases, higher incomes are associated with decreasing support for social welfare policy, but the negative effect of income is smaller for those who identify strongly with Canada.
These results lend support to claims that national identity may help to protect the welfare state from toxic effects of cultural suspicion. But we also find that identity most strongly supports only some parts of the welfare state in Canada. In figure 4, there is a much wider gap in support for health care between low- and high-identity respondents than there is in support for redistribution. This implies, we believe, that the relationship is highly contingent, reflecting distinctive features of the history and national narratives of each country. National identity may not have any general tendency to strengthen support for redistribution, but it may do so for those aspects of the welfare state seen as having played a particularly important role in building the nation or in enabling it to overcome particular challenges or crises.
The progressive agenda: Feasible but fragile
Is the progressive agenda feasible? Or is any attempt to combine diversity and solidarity doomed to self-destruct? Our evidence should dispel such fatalism. Cross-national analysis reveals no evidence that countries with high immigration have greater difficulty maintaining their welfare states. Nevertheless, the pace of change seems to matter, and countries with sudden surges in the size of their immigrant population do face challenges. But there is no evidence that multicultural policies exacerbate tensions. Hence, while tensions can and do emerge among societies combining diversity and multiculturalism policies, they are not condemned to erode solidarity. They are contingent on the role played by mediating factors.
Canadian experience is revealing about this role, and tells a distinctive story. In contrast to Putnam’s findings about our neighbours to the south, Canadians in diverse neighbourhoods do not hunker down. There is very little evidence that diversity weakens individual Canadians’ support for solidarity, redistributive or otherwise. There is evidence of indirect and mediated relationships, to be sure, but not such as to lead us to jump to the conclusion that diversity and solidarity are fundamentally opposed.
Why does Canada suggest a different narrative? Various explanations compete here. In many countries, the native population sees immigrants as an economic threat. Canadians, however, are unusual, even compared to the population of other settler societies, in the extent to which they see immigration as a benefit to the economy.29 Undoubtedly, this optimism is partly a reflection of the traditional success of immigrants in integrating economically, and of the success of their children – the second generation – in educational and social terms. As a result, in contrast to many European countries, immigrants and members of racial minority communities have not been especially dependent on social benefits in Canada.
Danger signals are beginning to flash at this point. Immigrants arriving in the last two decades have not enjoyed the same economic success as previous cohorts. If the engine of economic integration remains stalled and immigrants increasingly depend on social assistance and other benefits, the nexus between diversity and redistribution might well change.
But economic factors do not provide a sufficient explanation. Again in contrast to many other countries, immigrants have also not been seen as a cultural threat in Canada. In part, this is because of the effect of multiculturalism policies.30 For younger generations raised in the multicultural era, the drop in trust levels in diverse neighbourhoods disappears. Our data also point to a distinctive role for national identity. In most countries, individuals with a strong sense of national identity tend to be more strongly opposed to immigration. In contrast, a sense of being Canadian helps strengthen support for both diversity and solidarity, in part because, at least in English Canada, both health care and multicultural diversity are key elements of our national identity.
This suggests that governments have tools to manage potential conflicts between diversity and solidarity. While rapid change (at the neighbourhood or the national level) can be corrosive, governments can manage these by promoting multiculturalism, by nurturing institutional trust and by building ideas of diversity and solidarity into a national narrative. This is by no means a recipe available to all countries, and at the extreme it might be unique to Canada, but it does provide an important counternarrative to the United States. Realizing the progressive agenda of diversity combined with solidarity is feasible, if inevitably fragile.
Continue reading “Are Diversity and Solidarity Incompatible? Canada in Comparative Context”