The Economics of Identity and the Endogeneity of Race
National Bureau of Economic Research
Working Paper 9962
September 2003
Howard Bodenhorn, Professor of Economics
Clemson University, Clemson, South Carolina
Christopher S. Ruebeck, Associate Professor of Economics
Lafayette University, Easton, Pennsylvania
Economic and social theorists have modeled race and ethnicity as a form of personal identity produced in recognition of the costliness of adopting and maintaining a specific identity. These models of racial and ethnic identity recognize that race and ethnicity is potentially endogenous because racial and ethnic identities are fluid. We look at the free African-American population in the mid-nineteenth century to investigate the costs and benefits of adopting alternative racial identities. We model the choice as an extensive-form game, where whites choose to accept or reject a separate mulatto identity and mixed race individuals then choose whether or not to adopt that mulatto identity. Adopting a mulatto identity generates pecuniary gains, but imposes psychic costs. Our empirical results imply that race is contextual and that there was a large pecuniary benefit to adopting a mixed-race identity.
1. Introduction
Economic and social theorists have modeled race and ethnicity as a form of personal identity adopted in response to the costliness of maintaining a specific identity (Hechter, Friedman, and Appelbaum 1982; Stewart 1997; Mason 2001; Akerlof and Kranton 2000; Darity, Mason, and Stewart 2002). These models of racial and ethnic identity recognize that race and ethnicity is contextual because racial and ethnic identities are fluid (McElreath, Boyd, and Richerson, undated). Harris and Sim (2001) report recent evidence of this fluidity among contemporary mixed black-white youth. Although 75 percent of today’s mixed black-white children self-identify as black, 17 percent self-identify as white, and the remaining 8 percent prefer not to select a single racial designation. About 10 percent of mixed-race youth adopt one racial designation at school and a different one at home. It is evident that among modern mixed race youth racial identification is contextual.
Racial and ethnic self-identification have economic consequences because the choice of self-identity is likely to be entwined with the acceptance of and acculturation into dominant social norms. If race or ethnicity is endogenous in certain circumstances, a self-identity may or may not be selected to distance oneself from a subordinate group or to improve one’s standing with or acceptance into the dominant group. In a study of people of Mexican descent, Mason (2001) tests a model in which acculturation is a dominant strategy, and finds that light-complected people of Mexican descent may acculturate more easily. Murguia and Telles (1996) report different educational opportunities for Mexicans of light and dark complexion and argue that these may result from conscious choices. Phenotypic differences, they argue, influence individual strategies. Light-skinned people of Mexican descent learn early in life that by assimilating or acculturating they can defuse negative stereotypes and attain more than their dark-complected counterparts. Later in life, light-skinned Mexicans are able to increase their incomes by adopting a non-Hispanic white identity (Mason 2001). Yet there may also be situations in which members of the subordinate group decide to maintain identities separate from the dominant group.
Our study considers the choices and life chances of black and mixed black-white individuals residing in the urban U.S. South prior to the Civil War. The experience of mixed black-white individuals in this period is particularly germane to the study of the social and economic consequences of racial identification because the so-called one-drop rule was not yet firmly established. Most Upper South states legally adopted a one-fourth rule separating black from white. But the line was not as sharply drawn because the dominant white culture accepted mixed-race people as a separate class. As Williamson (1984, p. 13) notes for Virginia, “there were some people who were significantly black, visibly black, and known to be black, but by the law of the land and the rulings of the courts had the privileges of whites.” Lower South states generally adopted no formal definition of “whiteness,” and were even more accepting of a separate mixed-race or mulatto class. “Known and visible mulattoes could by behavior and reputation be ‘white’” (Williamson 1984, p. 19). Acculturation was an option for at least some mixed-race people living in the antebellum South.
We first model a mixed-race individual’s choice of self-identity. Acculturation brought a degree of acceptance from the dominant white community, which opened the door to a wider set of economic opportunities, but acculturation carried an implicit cost, namely that by adopting the norms of the dominant white culture (dress, language, mannerisms, religious affiliation, group membership, etc.), the individual alienated himself or herself from the black community. To the extent that the recognition of an individual’s heritage generates utility, the rejection of black culture was costly.
We then test the model empirically. We find that African Americans were more likely to identify as mulatto when there were already a substantial number of other mulattos who had formed social networks and established a community. Yet, the probability of declaring a mulatto identity declined with the size and extent of the African-American community. We interpret this to mean that if blacks ostracized mulattos for separating themselves socially and economically, then the larger the black community (holding the number of mulatto households constant) the more costly it was to be ostracized. Similarly, whites became less accepting of a mulatto’s distinctiveness as the city became increasingly African American and thus showed mulattos fewer preferences.
Once we demonstrate that the choice of a mulatto identity was associated with racial composition of the individual’s neighborhood and city, we then investigate the economic consequences of adopting a mulatto identity. We estimate differences in wealth between blacks and mulattoes and find that mixed-race householders, both male and female, accumulated more wealth than black householders. Regression decompositions suggest that a substantial portion of the wealth gap was due to racial identification and to community factors. Consistent with our model, we find that mixed-race people realized smaller advantages relative to blacks as the size of the African-American community increased both absolutely and relatively. Thus, mixed-race people benefited when they could form a distinct intermediate racial class, standing between the dominant white and subordinate black communities…
Read the entire paper here.