-
Chapter 3: Ethnicity and Ethnic Identifiability
- Russell Sage Foundation
- Chapter
- Additional Information
Chapter 3 Ethnicity and Ethnic Identifiability A ny claim that ethnic diversity is associated with better or worse public goods provision requires some notion of what ethnic diversity means. Yet the very idea of ethnicity or ethnic diversity is itself a contentious issue. As measured by most scholars, the concept of ethnic diversity appears simple at first: if two people are paired at random in a given community, what is the likelihood that they belong to the same ethnic group? If the likelihood is high, we call this a homogeneous community. If the likelihood is low, the community is said to be diverse. All that matters for determining this likelihood is data on the community’s demography. Scratching the surface of this apparently straightforward approach to measuring diversity, however, reveals a number of complications. What does it mean to say that an individual “belongs” to an ethnic group? How are we to know if two people are from the same “group”? Of the several groups in which individuals might claim membership, to which do we assign them? When we presented the scatter plot relating ethnic diversity to public goods provision in figure 1.1, we assigned community members to ethnic group categories defined by the Uganda Bureau of Statistics (and used in the national census), and we calculated each community’s diversity in terms of the relative sizes of the memberships of each of these groups. Implicitly, we assumed that the ethnic demography defined by these group categories matters for the people living in MulagoKyebando . We, like most scholars who use demographies based on ethnic categories officially defined by the government, assumed that people ’s actions are shaped by the way in which they (and others in their community) are classified by their country’s national statistics department .1 But is this a viable assumption? People living in real communities may not see themselves as fitting neatly into the categories utilized by census-takers. As we move toward a deeper understanding of how 36 ethnic diversity can undermine collective action—and how binary and small-group interactions are shaped by the match (or not) between the ethnic backgrounds of participating individuals—we need to develop a richer understanding of the way that individuals actually perceive their own, other people’s, and the broader landscape of ethnic backgrounds around them in their everyday interactions. We call this “subjective demography ,” and we contrast it with the “benchmark demography” that derives from census classifications. To make a distinction between “subjective” and “benchmark” demographies is to do more than simply point out that people may categorize themselves differently from how outsiders do. It is also to recognize that a person’s ethnic background cannot always be taken as an objective fact that is easily measured, as most formal and empirical work on ethnic politics usually assumes. Theorists generally assume that, at any point in time, individuals belong to some (or perhaps multiple) ethnic categories. They also assume that other individuals know these categories and have little or no problem working out the category to which any given individual belongs. Both assumptions present difficulties. The first assumption presupposes what some might call an “objective ” demography. We refer to it instead as a “benchmark” demography to reflect the fact that membership in a particular category is always subject to some form of assignment criterion—that is, relative to some necessarily subjective benchmark. We take the statement that individuals “truly” belong in a given category as nothing more than an application of some specified classification rule: given a well-defined identification rule, these are the categories to which each individual belongs.2 The second assumption—that people know and use these benchmark categories —rests on an empirical claim that community members share an understanding of the relevant social categories and that they can assign people to them without ambiguity. It assumes that people’s subjective demographies match a common benchmark demography. Yet, since we know so little about how people perceive the ethnic backgrounds of their fellow community members, this assumption is quite heroic. Our goal in this chapter is to find a way to speak about ethnic identities and to measure ethnic classifications and distributions that does not depend on the existence of objective identities. Instead, we rely on something more epistemologically defensible and more strategically important: what people actually believe about their own ethnic group membership and about the group memberships of others...