Stratifying a Population by Race

Philosophers have written a number of interesting books and articles about race in recent years, but most are about what race is (about the nature or reality of race or racial categories) or about how race bears on issues of distributive or retributive justice. My interest is different: not in what race is but what race does and, in particular, how race is used in the biomedical sciences to describe or explain differences within the U.S. population in health status or outcomes. Biomedical scientists in the United States routinely stratify populations by race when describing or explaining variations in health risks and find that Blacks and Whites differ in their risk of a variety of diseases and disease-related deaths. These findings are important, since they shape public policy and guide efforts to improve public health. However, a number of objections have been raised against the use of race as a variable in biomedical research. I discuss three objections and explain how studies of racial differences in the United States can avoid them and how race can be better used in the biomedical sciences. Rather than treat race as a fixed characteristic of members of a population, the sciences, I argue, should treat race as a variable whose value can vary from one health risk to another. Race, on this view, is not a property of members of a population but a relation between members and the risk within the population of morbidity and mortality; while, at the present time, the biomedical sciences assume that a member’s race is fixed, they should assume that his race can vary and that a member’s race can be White in relation to his risk of one medical condition or outcome and Black in relation to another.