President Obama was born the child of a “black man from Kenya and a white mother from Kansas.”1 The dilemma for health disparities researchers is this: Should his birth have been recorded under the white or black racial category?
The surprising answer is that it would have depended on when and where his birth took place. Many (but not all) states follow the U.S. Census Bureau guidelines for “assigning” a child of mixed-race parents to a racial category, but those guidelines have changed considerably over time. For example, according to the 1960 guidelines, the race assigned to the child was that of the parent providing the child’s information, so Obama would have been white if his mother had provided information about him, but he would have been black if his father had volunteered his information. The 1970 guidelines assigned the race of the father to the child, so Obama would have been black, but the 1980 guidelines assigned the race of the mother, so Obama would have been white. However, as mentioned above, not all states follow these guidelines. Instead, some states use the “hypodescent” rule, where the race of the most nonwhite parent is assigned to the child. Still other states assign the race that the child “is regarded [as] in the community”—in other words, the local racial narrative determines a child’s race.2
The arbitrariness of these assignments is masked by the use of “standard” racial or ethnic (R/E) categories in health disparities research. Thus, we describe the growing contradiction of their use to predict health outcomes in an increasingly racially ambiguous population.
R/E Categories in Health Research
Health outcomes (mortality, life expectancy, infant mortality, etc.) are not randomly distributed across the United States; some populations and areas consistently seem to have better outcomes than the national norm, while others consistently have worse ones. Researchers attempting to identify and track health disparities and inequities generally use five R/E categories—four racial groups (white, black, Asian/Pacific Islander, and American Indian) and one ethnic group (Hispanic)—to analyze and predict variations in health outcomes in the overall U.S. population. These categories are used as if they were permanent, naturally occurring, internally homogeneous, and discrete.
In the United States, 14% of babies are already racially ambiguous (i.e., multiracial and/or multiethnic).3 If they are withheld from analysis because they do not fit into the standard R/E groups, or if they are arbitrarily assigned into one or more of the R/E groups, the validity of health disparities predictions may be seriously compromised. As the percentage of racially ambiguous babies continues to grow in the United States, the validity of predictions based on the standard R/E categories may become even more compromised. Researchers need to appreciate the socially constructed nature of the R/E categories commonly employed; one way to do this is to examine their historical use, which reveals an evolving, artifactual heterogeneity.
The Social Construction of Race
In 1958, a circuit court judge in Virginia proclaimed that “Almighty God created the races white, black, yellow, malay, and red,” placing them on separate continents so they would never mix. 4(p450) He was expressing a racial narrative, long-held in the United States, that an individual’s race/ethnicity was a fundamental and immutable characteristic of that individual (i.e., created by “Almighty God”).4(p450)
The racial narrative in the United States has been constructed on the premise of biological determinism—that is, that a person’s race, transmitted vertically by that person’s parents, determines that individual’s intelligence, behavior, and susceptibility to disease. Early theological explanations for racial differences were replaced in the late 19th century by “scientific racism,” which distorted early quantitative methods to justify racial categories. This expanded into “eugenics science” in the early 20th century, aimed at identifying and suppressing, if not outright eliminating, the reproduction of “less suitable” populations.5 Over the years, researchers influenced by these racial narratives have sought to explain variations in behavior, intelligence, and disease susceptibility in U.S. populations by endowing R/E categories with explanatory and predictive power.
Racial structuring of society
In 1790, the Naturalization Act incorporated contemporary U.S. racial narrative into law by restricting citizenship to “free white persons.” During the 19th and early 20th centuries, federal and state governments posited a binary racially structured society of white and nonwhite and accordingly provided full access to civil rights, privileges, and protections to white persons only. Nonwhite persons—then generally meaning Africans, American Indians, and their descendants—were barred from full citizenship. During the 19th century, the legal definition of nonwhite persons rested on a hard-edged division known as the one-drop rule—having one ancestor identified as black sufficed to place a person in the black racial category. Indian was somewhat more nuanced, largely defined by a tribe’s treaty with the U.S. government; each treaty nevertheless specified a hard-edged division between white and Indian. During the 19th and early 20th centuries, Asian populations were added to the nonwhite category.6 (The term Asian itself was used to encompass a range of countries and languages from the approximate geographic areas of East Asia and Southeast Asia.) This hard-edged, binary racial divide between white and nonwhite, which was enforced by the power of the state, intruded into daily lives—determining where one could buy a home or attend school, when one could swim in the local pool, or whom one could marry.
Racial Ambiguity: The Growing Contradiction
Notwithstanding the origins (and perhaps inappropriateness) of R/E categories, a growing problem with continuing to use them is that the U.S. population is becoming increasingly racially ambiguous as more and more individuals self-identify as biracial, multiracial, or “some other race.” Two demographic trends drive this growing ambiguity in the U.S. population:
- A growing Latino (the terms Hispanic and Latino are used interchangeably in this Invited Commentary) population, nearly half of whom do not identify as one of the four racial categories and often describe themselves as some other race, and
- A growing population of racially ambiguous babies, whose parents belong to different R/E groups.
Latinos and U.S. racial categories
Since 1980, Latinos have been asked to categorize themselves first under the new ethnic category (Hispanic) in the U.S. Census, and then under one of the standard racial categories (white, black, Asian/Pacific Islander, or American Indian). However, just under half of Latinos consistently do not see themselves as fitting into any of the four single-race categories. That is, in 1980, 42% did not identify with any of these single-race categories, choosing instead to describe themselves as some other race. Thirty years later, in the 2010 census, the percentage of Latino respondents who described themselves as some other race or two or more races was still 42%.
As for the slight majority of Latinos (53%) who self-identified as white in 2010, describing themselves this way may not imply that they see this as their biological race. Rather, they may use the term to indicate that they see themselves as Americans, as U.S. citizens, as members of the middle class, or as members of other nonbiological identities.7
Racially ambiguous babies
A California Health Care Foundation (CHCF) report on 2014’s nearly half million births represented 97.7% of the state’s babies as belonging to a single race, with only 2.3% categorized as belonging to two or more races.8(p5) Thus, it might appear that nearly all of California’s babies fell neatly into mutually exclusive and internally homogeneous R/E categories. Presented in this fashion, the data hide a great deal of racial ambiguity for three reasons:
- Births are categorized only by the mother’s R/E group; the father’s R/E group is not mentioned,
- Latina is presented as if it were a single, homogenous R/E category with no further racial component, yet a footnote alerts the reader that Latina “includes any race group,” and
- This footnote also alerts readers that “unknown and other races are not shown.”8(p5)
We addressed the details not available in the CHCF report by using the American Community Survey to add the following details for California births in 2011–2015:
- The race or ethnicity of the father,
- Race categorization of both Hispanic mothers and fathers based on their responses this question, and
- “Unknown” and “other race” responses.9
Our results are shown in Figure 1. Less than half of the babies (47.4%) had a mother and father who were “racially congruent”—that is, they belonged to the same single-race group (non-Hispanic white mother and non-Hispanic white father, non-Hispanic black mother and non-Hispanic black father, etc.).
Over half of the babies (52.6%), however, were of considerable racial ambiguity and did not fall into the standard categories. In 25.0% of all births, the mothers and fathers were of noncongruent (different) single-race categories. In 16.0%, at least one parent chose the residual category of some other race to describe themselves. In 3.5%, at least one parent described themselves as being of two or more races, while the final 8.1% provided no racial information. Therefore, a majority of California’s babies cannot be adequately described as pertaining to one single hard-edged racial category or another.
Racial Narratives as Social Determinants of Health
From innate, individual trait to racial narrative
The biological determinism that has undergirded the U.S. racial narrative has led researchers to conclude that an individual’s R/E background is a major, if not the major factor in determining that individual’s health outcomes. Although most genomic researchers would argue that humans do not come in such neat, nonoverlapping groups, health researchers continue to use the same five R/E categories (see above) as if they did.
Over half of California’s babies are already racially ambiguous, and other states are seeing increasing numbers as well.3 How then will future researchers be able to identify and track health disparities and inequities as these categories become operationally obsolete in an increasingly racially ambiguous population?
We argue that researchers should measure the effects of the United States’ hard-edged, binary racial narrative; its social construction of race; and its racial structuring of society. Through laws, policies, and institutions, U.S. society has assigned certain individuals and groups to legally defined, socially constructed races. All of these groups (other than whites) have subsequently suffered impaired access to society’s rights, services, and protections. This impaired access has created differentials in income, education, voting behaviors, and the like. These different, racialized life courses have also led to different health outcomes.
For example, the discrepancy between an arbitrarily assigned racial category and the lived experience of the U.S. racial narrative can be illustrated by President Obama’s political life. If he had been born after the 1980 Census, Obama’s birth would have been recorded in the white category, and the race of his Kenyan father might not have been a factor in his political career. Yet, the U.S. racial narrative is so strong that he has lived his life as a black individual, not as a white one. The important point for population-based research is that differences in health outcomes are the result of the hard-edged, binary U.S. racial narrative and are not due to innate, individual racial traits. Thus, we suggest that the hard-edged, binary racial narrative itself is a social determinant of health.
Operationalizing the racial narrative
Having reconceptualized the U.S. racial narrative as a social determinant of health, how do we now measure it? Given our understanding that race is not biologically determined but, rather, socially constructed, our goal in health disparities research is to operationalize this binary narrative and its influence on the racial structuring of society and the subsequent racialization of the life courses of R/E groups.
Alternative racial narratives
Moreover, racialized groups can create their own racial narratives, usually in response to the United States’ hard-edged one. Latinos, in particular—as the result of over 500 years of population mixing among the Indigenous peoples of the Americas and populations from Africa, Europe, and Asia—have developed a narrative that embraces racial mixture (mestizaje in Spanish) as opposed to hard-edged distinctions. In epistemological terms, Latinos in California are heirs to a 500-year-old fuzzy-edged, multivalent (see below) racial narrative that has coexisted for 170 years alongside the United States’ hard-edged, binary one. In this narrative, fuzzy-edged differences in ancestry can shade into one another, so that boundaries between them may be blurred rather than hard-edged. Additionally, race and ethnicity, in the Latino narrative, can be multivalent, and thus constructed in terms of intersectionality (many elements present all at once); situationality (e.g., an individual’s urban, Spanish-speaking construct in California may differ from the construct he/she uses when visiting family in a Nahuatl-speaking, rural village); and intentionality (the intentions of both the individual and the interrogator are considered).10
Health disparities between groups currently categorized by race or ethnicity are not randomly distributed. We suggest that health disparities do not result from biological differences between different R/E groups but, instead, are partially the result of the United States’ hard-edged, binary racial narrative, which has fashioned racialized life courses for different groups, resulting in their different health outcomes.
Some implications of this for academic medicine are that:
- In trying to increase diversity among medical students, admissions committees may want to consider how to operationalize applicants’ relationships to racial narratives (rather than their putative biological backgrounds).
- While other R/E narratives exist internationally (and can be just as profound and impactful in other countries), an awareness of the ones specific to the United States would help inform research in population health.
We invite readers to collaborate in developing an epidemiology of diversity—a field of health disparities research that rejects assumptions of biological determinism and focuses instead on racial narratives as a source of R/E health disparities.
The authors gratefully acknowledge the support of Adventist Health White Memorial’s Center for Hispanic Health.