race and ethnicity. For example, there is a large literature on health differences across income, wealth, and educational groups, showing that the poor typically have far worse health than the rich. (Concern about such inequalities is often greater in other countries than in the United States [Gwatkin, 2000; Independent Inquiry into Inequalities in Health, 1998]).
It is often argued that health inequities—including those across racial and ethnic groups—are important because they identify lives that can be saved. If there is no “biological” reason why poor people should be sicker than rich people or why blacks or American Indians and Alaska Natives should be sicker than whites, Hispanics, or Asians, then Americans who are poor or black or from native populations are obvious targets for health interventions. Yet it is not clear whether such targeting is the most effective way, from a medical standpoint, of improving population health.
One issue that naturally arises when group differences in health are discussed is the question of within-group heterogeneity in health. The need for health care is best identified by health status, not by membership in a socioeconomic or racial or ethnic group. Many poor people are not sick, and many rich people are. If the objective is to improve health, one should simply select people by their need for care, not by their socioeconomic status or ethnicity. In many cases, though, selecting on health status may be too late to do much good. If the goal is to prevent a disease or health-threatening condition, one needs to select people at risk, and although race or ethnicity is never the sole relevant risk factor, it is often a major risk factor. So, for example, a strategy for reducing hypertension or diabetes may target racial or ethnic groups that are at high risk for those conditions once other risk factors have been taken into account.
It is also worth asking why health differences matter from a research perspective. For example, it costs money to collect data on health differences, especially for small, hard-to-sample ethnic groups, and policy makers must decide whether or when public support for such efforts is justified. Policy priorities also depend to some extent on what causes health differences, so that it becomes an important goal of research to uncover mechanisms. For example, some people argue that health differences that are “chosen,” through differences in behaviors, are a less important issue for public policy than are differences that are involuntary or that are imposed by other people’s behavior. In particular, if research shows that black-white health differences are in large part attributable to differences in health care, the policy agenda would be much clearer than it currently is. One good reason for caring about black-white health differences is the possibility that this might be the case, that there are more “preventable” deaths among minorities, and that some relatively straightforward policies could bring the health of blacks closer to that of whites. The extent to which health care