More than 200 years into the modern experiment with democratic forms of rule, democratic aspirations continue to founder on the rocks of racial and ethnic hierarchies, and other patterns of domination constructed on social categorizations of difference. In the case of the U.S., African American disadvantage continues to be entrenched, fifty years after the Civil Rights Act of 1964 transformed the legal landscape. In Europe, we are witnessing the resurgence of the far-right, in response to dramatic demographic diversification, occurring simultaneously with economic instability. Civil war related to ethnic violence has devastated many African countries in recent years. India has the world’s largest affirmative action program and yet to cross caste and religious lines in marriage is to open oneself and one’s family to abuse and often murder by the locally dominant.
Persuasive arguments have been made that liberal theories that arose alongside and in partnership with systems of social domination are ill-equipped to provide guidance for the pursuit of democratic political forms that eschew such domination. Older liberal theories sought to address the problem of racial hierarchy by identifying the issue as a matter of “exclusion” to be addressed by “inclusion.” This formulation is problematic because it assumes a healthy socio-political whole into which the marginalized need only to be incorporated. Instead, the problem of “exclusion” emerged from the constitution itself of the mainstream socio-political order and the forms of domination practiced in that order. Moreover, the notion that rectification or transformation might be achieved by shifting from “exclusion” to “inclusion” analytically assigns all agency to the majority group, which can either exclude or include, and thereby reinforces the patterns that led to marginalization in the first place.
In contrast to the older work in liberal theory, recent work in sociology and political philosophy has underscored that the problem of hierarchies built on social difference is better understood as one of domination. The challenge, then, is to understand how to build socio-political systems on a framework of “non-domination.” In short, the question of how to achieve fair and just forms of democratic life in conditions of significant demographic diversity must be tackled afresh, from the ground up.
Importantly, knowing how to pursue the sort of world we desire involves both normative difficulties and challenges for positive analyses. First, for the normative. Precisely what ideal should govern our understanding of what counts as just social relations in contexts of demographic diversity? Over the course of the 20th century, candidates have included: “social cohesion,” “assimilation,” “integration” and “multiculturalism.” Each of these has come under critical attack. None currently provides a broadly accepted point of orientation. What theories of identity and of the human need for recognition should we defend? How do we incorporate those theories in our accounts of justice? How do we need to revise our theories of justice, or core social ideals, if we are to do better at achieving fair and just forms of democratic life in conditions of significant demographic diversity? Can we develop concepts that are applicable both in societies with a small number of social groups and those with a multitude of categories in a muddled hierarchy? In other words, can our concepts be independent of our many contexts and yet be useful in charting directions toward fair and just democracies?
As for positive social scientific analyses of social difference and inequality, conundrums abound as scholars seek to understand the behavior of agents within socially stratified societies. In the first instance, most contemporary social science research on discrimination and inequality does a good job of describing the problematic social equilibria in which we find ourselves, but a very poor job of describing the concrete relationships between the circumstances in which we find ourselves and the circumstances in which we would like to find ourselves. The problems are both theoretical and empirical. As an example of the theoretical difficulties, take Thomas Schelling’s insight about the disjunction between micro motives and macro effects. For instance, the literature on neighborhood sorting in economics has shown that even if all households prefer some racial diversity, uncoordinated actions by them can generate segregated neighborhoods that stubbornly persist in changing economic environments. This theoretical insight then has important implications how we should interpret empirical estimates of discrimination. Empirical studies that reveal lower interview callback rates for black households using identical fictitious resumes tell us about employer behavior in these existing segregated equilibria. We learn nothing about how these estimates would change if neighborhoods and schools were more diverse even though these alternative configurations are also feasible and rational within the same theoretical framework. In other words, the question of what we should analyze, empirically, is at present too closely tethered to an acceptance of status quo equilibria.
Similarly, the analytical categories that we use in our research flow too directly from the social hierarchies that are the problem we seek to overcome. For instance, in research on race and education in the U.S., white students are considered the baseline against which all other groups of students are measured, both those groups that on average perform less well than whites (Latinos and African-Americans) and those groups that perform on average better than whites (Asian-Americans). But why should white students be a baseline, so that other groups are constantly being compared to “whites” and never “whites” to anyone else? Would it not be better to use an external standard, for instance, median (or 75th or 90th percentile) global performance on PISA tests, to assess the differential distribution of achievement within a given national population? In a completely different context, most empirical research on caste inequality in India uses the categories of Scheduled Castes and Scheduled Tribes that were created to define eligibility to affirmative action programs in the 1950s. These categories have been politically contested and their membership has changed as some of the wealthier castes have tried to enter them to corner benefits of affirmative action in a resource- scarce economy, yet there is little intellectual discussion of alternative classifications.
Problems of this kind permeate the social sciences, and a fresh approach to issues of race, ethnicity, social difference, and social hierarchy requires comprehensively re-visiting our approaches to measurement and comparison, as well as to the basic enterprise of the construction of research questions. Such efforts necessarily also entail engagement with the normative or ethical aspects of these issues.
Diversity, Justice, and Democracy programming is funded in part by a grant from the Ford Foundation.