1. Introduction

In June 2005, a heart failure medication called “BiDil” became the first drug ever to be approved by the Federal Drug Adminstration (FDA) for treatment of members of a particular race (African Americans). Since that time, the medical community has shown increasing interest in pursuing research aimed at discovering other forms of treatments that can be useful for treating specific races. Given both the success that BiDil has had in treating African Americans and the medically relevant biological distinctions that are generally recognized by health care practitioners as existing among groups traditionally categorized as constituting separate races (e.g., susceptibility to sickle cell anemia among blacks), one might think that research into so-called race-based medicine would be an avenue that would be unanimously welcomed by both medical researchers and bioethicists alike. As it turns out, however, there has been significant opposition to the pursuit of race-based medicine. A significant amount of controversy surrounding race-based medicine has arisen around the question of whether the concept of race holds any legitimacy within a medical or any other context. The backlash against the movement toward race-based medicine—and, specifically, against the notion of race that underlies it—is captured in the following statement by Sharona Hoffman: “‘Race-based’ medicine is an inappropriate and perilous approach … the concept of ‘race’ is elusive and has no reliable definition in medical science, the social sciences, and the law. … ‘Race’, consequentially, does not constitute a valid and sensible foundation for research or therapeutic decision-making” (Reference Hoffman2005, 397).

In this essay I argue that there are strong reasons for preserving the concept of race in both medical and sociological contexts. While there are important reasons to conceive of race as picking out distinctions among populations that are both legitimate and important, the notion of race that I advocate in this essay differs in fundamental ways from traditional folk notions of race. As a result, I believe that the folk understanding of race needs to be either revised or eliminated altogether.

2. Race as a Problematic Biological Concept

In attempting to provide a defensible account of race, one is confronted by two key obstacles. On the one hand, one is compelled to address the myriad of dark episodes in human history that have been caused, to some extent, by the perceived racial distinctions among certain groups. Atrocities of this sort are often cited as a reason for why we should forego making any racial distinctions. The other obstacle—and the one that is perhaps most salient for the purposes of this essay—pertains to the biological difficulties of establishing meaningful racial distinctions among groups. The traditional concept of racial distinctions that many of us have—black, white, Asian, and so forth—roughly corresponds to what Pigliucci and Kaplan have termed “folk racial categories.” The problem with these categories is, as Kaplan (Reference Kaplan2010) puts it, that “there remains a broad consensus that current folk racial categories … do not correspond to meaningful biological categories” (283). This conclusion follows from several important biological findings, including how genetic variation within groups is significantly larger than it is among groups and how the genes of all human beings are 99.9% similar.

Despite an overwhelming number of evolutionary biologists and anthropologists denying that there are any biologically meaningful human races, recent efforts have been made to establish racial distinctions along lines that correspond roughly with folk racial categories. Risch et al. (Reference Risch, Buchard, Ziv and Tang2002), Rosenberg et al. (Reference Rosenberg, Pritchard, Weber, Cann, Kidd, Zhivotovsky and Feldman2002), and others have pointed out that it is possible to identify races by using genetic markers—in this case, nonfunctional DNA—that divide populations according to genetically traced geographical ancestry. In this way, individuals are assigned to racial categories that roughly correspond to folk notions of race. Notwithstanding the claims of some that this procedure has demonstrated the biological reality of races, there are several reasons for being skeptical about the importance of what these studies have shown. As anthropologists Kenneth Weiss and Stephanie Fullerton point out (Reference Weiss and Fullerton2005), these studies are highly arbitrary in that they set out to place individuals into groups (African, Caucasian, etc., i.e., the folk racial categories) that were picked out as constituting the primary racial categories. According to Weiss and Fullerton (Reference Weiss and Fullerton2005, 167), the experiments mentioned by Rosenberg et al. and others could just as easily have divided up populations according to a completely different selection of geographical ancestry—for instance, “Icelanders, New Zealand Maoris, and Mayans.” While the latter grouping may seem much less natural from our perspective, Weiss and Fullerton suggest that such a grouping is based not on “inherent features of the data, but choices investigators make for a host of reasons” (167).

The question to be asked here is that, given the possibility of categorizing human beings according to whatever geographical criteria you please, why should such a categorization be deemed important by anyone? From a medical perspective, is the possibility of dividing populations along arbitrary geographical ancestral lines something that is likely to further the aims of medicine? To see why there is reason to doubt this, consider international studies conducted on rates of hypertension. Numerous studies have shown that North American blacks have significantly higher rates of hypertension than North American whites. Now were we persuaded by the studies cited by Rosenberg et al. to categorize individuals along folk racial lines, we might reasonably expect blacks worldwide to have higher rates of hypertension than whites. But this conclusion would be wrong. As Hoffman points out, blacks on average were found to have less hypertension than whites worldwide (Reference Hoffman2005, 403). Hence, if one were to rely upon either folk racial categories or the recent attempts to establish racial categories through genetic ancestry (which lead to roughly the same conclusions according to Rosenberg et al.) in making generalizations about hypertension rates in blacks, one would be led into error.

Now it is true that tracing one's ancestry can provide some indication of whether one is more susceptible to certain types of genetic maladies (e.g., Tay-Sachs Disease, cystic fibrosis). But even with regard to such ailments, drawing conclusions based on broad ancestral information can lead to mistakes. As Root (Reference Root2003) recognizes, it is not the fact that one can trace one's recent ancestral history to Africa that puts one at risk for sickle cell anemia. The question is whether one can trace one's ancestry to a specific area in which malaria was rife. But since, as Kaplan (Reference Kaplan2010, 286) points out, carriers of the HbS (“sickle cell”) allele can trace their ancestry to malarial regions of both Africa and Europe, using race as an indicator of whether one carries the allele in question will lead to overpredicting black risk and underpredicting white risk. So it turns out that even for those ailments for which traditional notions of race would seem best equipped to provide information (e.g., sickle cell anemia), folk racial categories and genetic tests for ancestral geography are of limited use. Assuming, then, that any conception of race that is based exclusively on either folk racial categories or geographic ancestry (or some combination thereof) is unlikely to serve any useful purpose to medicine, should we dispense with discussing race altogether in biomedical contexts? In section 4, I suggest that it may be possible to develop a notion of race that can play an important role in informing medicine.

3. A Paradigmatic Racial Distinction

While folk racial categories seem ill-suited to the task of helping us distinguish among populations on the basis of relevant biological differences, there is little doubt that such categories can act as a potent social force that has important biological ramifications. It is in virtue of being categorized as being of a certain race and being exposed to distinct environmental forces over the course of numerous generations on the basis of this racial identity that a group can develop its own unique biological characteristics. This phenomenon is clearly at work in the case of American blacks. In building his case that folk notions of race have had a profound effect on the health of this group, Kaplan mentions how “by almost every measure, Black Americans have, on average, worse health and health-outcomes than do White Americans” (Reference Kaplan2010, 287). black Americans have a 6-year shorter average life expectancy than white Americans, as well as worse self-reported health than white Americans (Kaplan Reference Kaplan2010).

Kaplan maintains that the health discrepancies between black and white Americans are not attributable to inherent biological factors like genetic ancestry. In support of this view, he points out how several studies (e.g., David and Collins Reference David and Collins1997; Sankar et al. Reference Sankar, Cho, Condit, Hunt, Koenig, Marshall, Lee and Spicer2004) have shown that the general health deficiencies shared by American blacks are not shared by non-American blacks or recent African immigrants to America. He believes that the most likely explanation for why black Americans tend to be less healthy than their white counterparts has to do with the substantial differences between these two groups in terms of important social barometers such as wealth, education, and location—all fronts on which blacks are worse off than whites. And, as Kaplan recognizes, the difference between the prospects afforded to blacks and whites in America is directly attributable to the role that folk racial categories play in American society.

But there is more to this picture. It turns out that the health challenges faced by black Americans persist even when factors like socioeconomic status and education are controlled for. Furthermore, there is evidence that many of the health problems black Americans face persist for generations even when the environmental factors that seem to have given rise to such problems (e.g., poor nutrition) no longer appear to be present. All of this is evidence that the adverse social circumstances in which black Americans have tended to find themselves have brought about changes on the biological level. Thus, it turns out that while there is very little in the way of a legitimate biological basis for the folk racial categories, these categories do have significant biological/medical repercussions. This leads Kaplan to conclude that “the causal arrow points not from biology to race, but rather from race to biology” (Reference Kaplan2010, 291).

Looking at the impressive evidence of health differences between blacks and whites in America, one might be tempted to conclude that it would be justifiable to distinguish between blacks and whites along lines that correspond to our folk notions of race. As I pointed out earlier, however, the adverse health conditions facing American blacks do not generally extend to non-American blacks. A reasonable conclusion to draw, therefore, is that black Americans are biologically different from other populations in important ways. But it is not merely that black Americans are different from groups such as whites, Asians, and nonblack Hispanics. Black Americans appear to be different from non-American blacks as well. The upshot is that there does appear to be a legitimate basis (i.e., a biological basis) for viewing black Americans as constituting a distinct race. Of course, in order to determine whether it is appropriate to categorize a population as a race, some analysis of the term “race” is called for. However, settling on an acceptable notion of race is complicated by the fact that—as Pigliucci and Kaplan (Reference Pigliucci and Kaplan2003) point out—the term “race” is defined differently throughout the biological literature. Nonetheless, a reasonable definition of “race” is offered by King and Stansfield in the Dictionary of Genetics (Reference King and Stansfield1990): “[A race is a] phenotypically and/or geographically distinctive subspecified group, composed of individuals inhabiting a defined geographical and/or ecological region, and possessing characteristic phenotypic and gene frequencies that distinguish it from other such groups.”

This definition of “race” seems acceptable in that while it approximates the folk understanding of “race”—which any acceptable definition of “race” should do to some degree—it rules out a significant degree of arbitrariness in the designation of specific races by requiring that groups designated as races share some distinctive genetic and phenotypic similarities as well as geographical or ecological proximity. That this definition would allow black Americans to constitute a race follows from how this group is designated by a distinct geographical locale (the United States), and shares various distinctive phenotypic and genetic features.

During their discussion of whether it is appropriate to recognize specific races among the human species, Pigliucci and Kaplan maintain that “the question is not whether biological ‘races’ exist; rather, it is which biological race concepts can be most usefully [my emphasis] applied to human populations” (Reference Pigliucci and Kaplan2003, 1164). I agree with Pigliucci and Kaplan that usefulness is the key concept when it comes to deciding both whether we should distinguish among races and, if so, how we ought to go about doing so. As the example of BiDil highlights, racial distinctions can be useful to medicine. Had researchers not been viewing race as a variable in the experiments involving BiDil, this drug that has helped numerous heart failure patients might never have been available to the public.Footnote ¹ The key point is that there is substantial empirical evidence indicating that black Americans possess distinctive biological traits that are likely to influence the particular illnesses that they are susceptible to as well as how likely they are to respond positively to a given treatment. Turning a blind eye to these biological facts is likely to impede the progress of medical knowledge and would likely result in worse health care for black Americans. Hence, there are important biomedical reasons for viewing black Americans as constituting a distinct race.

4. Self-Identified Race as a Reliable Indicator of Racial Distributions within Local Populations

Given that one accepts the reasons I have offered in favor of categorizing black Americans as a distinct race, what are the implications for how we might go about drawing other racial distinctions? To find out, it may prove useful to understand how black Americans came to constitute a specific race in the first place. I have already explained how folk concepts of race—and a particular form of social organization that has been constructed upon them—acted as the catalyst for bringing about the biological features that distinguish black Americans from other groups. Contra Kaplan (Reference Kaplan2010), there is compelling evidence that at least some of these biological features have a genetic basis. For instance, in 2005 a gene variant was discovered that increases the risk of heart attack in black Americans by 250%, though it is fairly harmless in white Americans (Wade Reference Wade2005).Footnote ² Additional research has indicated that there may be a genetic basis for other conditions afflicting disproportional numbers of black Americans, including preterm birth (Wang et al. Reference Wang, Feng, Zhang, Zhou, Jiang, Niu, Wei, Xu, Xu and Wang2006).

In addition to biological differences between black Americans and other groups that have a genetic basis, there is a substantial amount of empirical evidence indicating that some of the heritable conditions plaguing blacks in America have a nongenetic basis. Although the notion of inheritance of acquired traits has received much scorn over the years, there has of late been scientific evidence accumulating in its favor. In their Reference Drake and Walker2004 paper, Drake and Walker cite several studies indicating that nongenetic factors can bring about inherited characteristics across multiple generations. As an example, the authors discuss how dietary reductions for rats resulted in lower birth weights across subsequent generations even once the rats’ dietary allotments returned to normal (Drake and Walker Reference Drake and Walker2004, 5). It is notable in these studies both that a genetic basis for the observed changes appears to be lacking and that the observed changes persevere even in the absence of a continuing environmental stimulus. In regard to how the ability to inherit acquired traits might affect human beings, Drake and Walker suggest that “socio-economic factors may have a role to play in intergenerational effects. Lifelong minority status and disadvantage amongst black women in the US may have played a key part in perpetuating poor intrauterine growth across a number of generations” (7).

The suggestion that at least some of the intergenerational health difficulties facing black Americans are due to nongenetic inherited biological dispositions has been argued forcefully by Kaplan (Reference Kaplan2010). According to him, much of the health discrepancies among black and white Americans can be attributable to the stresses that American blacks have suffered as a result of racism. Following the parlance of Mays, Cochran, and Barnes (Reference Mays, Cochran and Barnes2007), Kaplan uses the phrase “allostatic load” to refer to these types of stressors. The basic idea behind allostatic load is that generations of discrimination and other forms of racial oppression have brought about fundamental changes in the biology of black Americans in the same way that adverse environmental pressures brought about biological changes to the animals in the studies cited by Drake and Walker. In support of this view, Kaplan illustrates how allostatic load has been correlated with numerous adverse medical conditions, including higher propensities toward hypertension and preterm birth. The idea that these racial distinctions in terms of overall health are at least partially attributable to nongenetic heritable factors would explain why non-American blacks do not suffer these ailments to an extent similar to American blacks. This also helps explain why American blacks suffer high incidences of these ailments even once socioeconomic status and education are controlled for.

The preceding paragraphs offer reason to believe that the heritable biological changes by which black Americans have come (I have argued) to constitute a distinct race are attributable to genetic as well as to nongenetic factors. This fact provides some guidance in how to go about identifying other distinct races. The genetic record provides a strong case for why we ought to refrain from positing races strictly according to folk racial categories. Whereas folk notions of race favor grouping all blacks together, the biological record cautions against doing so since there is hardly any more genetic similarity among all blacks than there is among blacks and all other races. Furthermore, what little additional genetic similarity there is among all blacks is unlikely to yield any benefits, be they medical or otherwise. Likewise, genetic similarity by itself seems unable to provide a satisfactory means by which to make racial distinctions. I have already mentioned how identifying racial groups according to genetic ancestry is often arbitrary and fairly useless in terms of informing medical decisions. Perhaps a bigger problem with relying on genetic similarity to identify race is that doing so would ignore important nongenetic factors that figure into biological inheritance (e.g., allostatic load). Paying heed only to genetics is bound to make one overlook important groupings that can inform medicine.

As an initial effort to provide a useful method of racial classification, I suggest that the first step should be to ascertain what particular race an individual considers herself to belong to. In many ways, a person's self-identified race will correspond with folk racial categories. Self-identified race is important in that understanding a person's racial identity can inform us about the kinds of environmental and social forces that could influence the individual's biology (e.g., allostatic load). It can also cue us in on which group an individual is likely to share key genetic characteristics with. One drawback to self-identified race as the basis of racial classification is that, for instance, both a black American and a black African might identify their races as “black.” Since these individuals are likely to share little in the way of relevant biological traits, classifying these two individuals as being members of the same race would seem a mistake. To prevent this kind of error, I suggest that self-identified race should be used to identify one's racial category only among those in a localized population—say, a country like the United States. Furthermore, it seems reasonable to hold that self-identified race should only identify one's race among those who are at least second-generation citizens. The purpose of this qualifier is to rule out racial groupings among, say, multi-generational black Americans and recent African immigrants, who likely share little in the way of medically relevant biological distinctions.

Now it may turn out that drawing racial distinctions in the way I have sketched out would result in racial groupings (e.g., white American, Asian American) that are of little to no help when it comes to revealing the medically relevant biology of a particular group's constituents. Drawing from the example of black Americans, it may be that a distinct racial categorization should only be afforded to populations that have undergone serious oppression over the course of generations. The reason for this is that it may be the case that medically relevant races are only likely to come into existence by being subjected to a substantial amount of adverse environmental stimuli over an extended period. By this account, there may be no legitimate reason to distinguish between, say, self-identified white Americans and self-identified Asian Americans, though it would probably be appropriate to make racial distinctions for historically subjugated groups like Native American Indians as well as other traditionally oppressed groups around the globe.

I realize that my admittedly cursory proposal for developing racial categories has an element of arbitrariness about it and that it may be more suited to the United States than to other countries. Nonetheless, I believe that it at least provides a useful first step for developing a more robust system of racial classification by including elements that help it to avoid some of the difficulties that will be inevitably faced by other systems—such as those that rely primarily on folk racial categories (e.g., the current U.S. Census) or strictly upon genetic ancestry. As far as the claim that the system I have suggested is hopelessly arbitrary in that it seeks to make a neat distinction among biological entities that are incapable of being divided in such a clear-cut manner, I admit that there is an element of truth to this objection. However, I would respond that in the field of medicine people are often placed into categories (such as age group 40–50) that have a similar air of arbitrariness about them. But the fact is that while such groupings do not provide pinpoint guidance in terms of what medical courses to pursue, they do allow for approaches to health care that are generally beneficial for members of these groups. In the same way that your falling between the ages of 40 and 50 gives a doctor a general idea about your health (illnesses to look for, etc.), there is reason to think that being aware of a patient's race would provide a doctor with an advantage in making health care decisions about a patient. It should also be pointed out that, from a medical perspective, people are usually categorized into several groups, such as age, sex, whether one is a smoker, and so forth. By virtue of a patient identifying himself or herself as belonging to several distinct groups, doctors are able to make fairly accurate generalizations about the patient's health. Likewise, I allow for the possibility that one might be classified as belonging to more than one race. Being aware that a patient identifies with multiple races may enable a doctor to check a patient for maladies known to be especially frequent among members of the particular races in question.

To sum up my essay to this point, I have argued that there are strong biological reasons to consider black Americans as constituting a distinct race. The distinctiveness of this race is owed not merely to the fact that most U.S. citizens view it as distinct but rather to distinctive biological characteristics that members of this group share with one another. Even so, I have discussed how traditional notions of race have played a key role in bringing about these distinct biological properties. Finally, I have offered what I take to be a reasonable first step in developing a methodology for assigning individuals to racial categories. In section 5, I consider an argument against the propriety of using race as a basis for medical decisions.

5. An Objection against Race-Based Medicine Considered

Toward the beginning of this essay, I mentioned how Sharona Hoffman rejected race-based medicine as being a “perilous approach” to medicine that relies on a shaky concept (i.e., race) that should not serve as the foundation for decisions in health care (Reference Hoffman2005, 397). One concern Hoffman has with race-based medicine has to do with the possibility that identifying a certain race as being more prone to an illness may reinforce negative stereotypes about the race in question, perhaps to the point of stigmatizing the race as being inferior to other races. In response to this concern, I would answer that recognizing the origins of racial distinctions along the lines that I have suggested is likely, if anything, to eliminate many of the negative attitudes held toward certain races. The fact that any major biological distinctions among races are not primarily due to inherent genetic properties—but rather to oppressive social structures—should garner increased sympathy toward those races that are viewed as “disadvantaged.” Taking the case of black Americans, the fact that much of health problems they suffer are the result of nongenetic factors like allostatic load should bring about increased awareness that their health difficulties are the result not of some innate genetic inferiority but rather of the oppressive environment that they have been thrown into. I would think that recognizing this fact is likely to result in more empathy than antagonism for nonblack Americans.

6. Conclusion

The pursuit of race-based medicine is a good idea for several reasons. The plain fact is that there are medically relevant biological distinctions among certain groups. The fact that these distinctions often correspond roughly to traditional notions of race—and are, in fact, primarily brought about by these very notions—lends credence to the view that race is a relevant concept for the fields of biology and medicine. Furthermore, racial distinctions in medicine have already yielded positive results (e.g., BiDil). While preserving the concept of race may be important for the field of medicine, it seems even more important from a sociological standpoint. The biological distinctions that exist among certain groups are a testament to how social organization can affect the lives of individuals in profound ways. If we are persuaded by arguments to ignore these biological distinctions—which I have argued are the defining characteristics of racial distinctions—we will be unable to see one of the most powerful ways that our current means of social organization can adversely affect the lives of individuals.

The issue of race is a difficult one that is complicated by how distinctions along racial lines—which have almost always been carried out via poorly founded criteria—have led to some of the most heinous atrocities perpetrated by the human species. Despite the monumentally poor track record of historical efforts to divide individuals into racial categories, we should not make the mistake of failing to recognize important biological differences among individuals when paying heed to these differences can generate positive results. Should we be persuaded by past mistakes to ignore these biological differences, we are likely to cause unnecessary suffering to those very groups whose welfare we are interested in furthering.