Skip Navigation

Race / Ethnicity


Scholars (e.g., Harris, 2000; Hirschman, et al., 2000; Perez & Hirschman, 2009; Saenz & Morales, 2005) and federal agencies, including the Census Bureau, view both race and ethnicity as social constructs. Federal agencies must measure and report race and ethnicity in accordance with standards in a 1997 revision of the Office of Management and Budget's (OMB) Directive 15. These standards "reflect a social definition of race recognized in this country… and do not conform to any biological, anthropological or genetic criteria" (Office of Information and Regulatory Affairs, 1997b). (The minimum racial categories required by the OMB are listed under Suggested Measure, below.)

The Census Bureau defines ethnicity or origin as "the heritage, nationality group, lineage, or country of birth of the person or the person's parents or ancestors before their arrival in the United States" (U.S. Census Bureau, 2001).

The OMB standards include the ethnic categories of Hispanic or Latino and Not Hispanic or Latino, with racial categories as separate. Census respondents are to respond in both ethnicity and race categories, but "the popular assumption is that virtually everyone can be fit into one, and only one, of these categories" (Hirschman, Alba, & Farley, 2000). For example, research (Rodriguez, 1992) has found that "the racially diverse Hispanic population regards their 'Hispanic' identity as a 'racial' one" (Office of Information and Regulatory Affairs, 1997a). Indeed, there has been "official ambivalence" at the Census Bureau, itself, as to "whether Spanish-speaking groups should be considered a separate race, or not" (Bates, de la Puente, DeMaio, & Martin, 1994).

Historically, race has transitioned from "a biological concept to a social category" (Hirschman, p. 403), including in biomedical research (Winkler, 2004). In biomedical research, this transition is further reinforced by the Human Genome Project which found that the human population has 99.9% of its DNA in common (Landers, Linton, Birren, et al., 2007); that greater genetic differences exist "within such groups as 'black' or 'white' than between groups"; and that "[t]here is no biological or scientific basis for the term 'race' much less the categories commonly used to describe it" (Rivara & Finberg, 2001, p. 119). The editor of the New England Journal of Medicine stated the same in terms of medical practice, maintaining that "attributing differences in a biologic end point to race is not only imprecise but also of no proven value in treating an individual patient," concluding that "[r]ace is a social construct not a scientific classification" (Schwartz, 2001, p. 1392).

Biomedicine, like social science, must explain health differences in terms of different cultural, social, and environmental influences. Moreover, both sciences are cautioned to avoid using race as a proxy measure for these influences (Hirschman, 2004; Winkler, 2004), and instead measure such underlying explanatory variables directly (Winkler, 2004). Furthermore, caution must be taken in how underlying variables of interest are measured and interpreted. Studies which use oversampled data of targeted racial or ethnic groups should be cognizant that such oversampling has been associated with a large increase in variances for the total population but only a marginal increase for the targeted population (Waksberg, Judkins, & Massey, 1997). Researchers using targeted oversampled data from a specific geographic area must be careful not to make statistical inferences about the targeted groups that, instead, should be made about the specific geographic area (Denton & Deane, 2010). Lastly, in using Census or other secondary data, an understanding of how missing data from less responsive groups are imputed is important. Skerry (2000) argues that Census Bureau imputation techniques have overcounted Hispanics in decennial Census results.

Some believe that "the term ethnicity may supplant race" (Hirschman, et al., 2000); others believe that race is a subset of ethnicity (Perlmann & Waters, 2002). Even a past Director of the U.S. Bureau of the Census testified before Congress that the "Bureau recognizes that the concepts are not mutually exclusive…" (Statement of Harry A. Scarr, 1994). However, notwithstanding its progressive blurring together with ethnicity, "race" is still used and its inclusion as a variable is required in studies funded by agencies important to health communication, such as the National Institutes of Health (National Institutes of Health, 2001).

There are obvious historical and social reasons why accounting for race in the census, other governmental initiatives, and research is important. A clear example is in support of governmental civil rights enforcement. Unlike assignment of ethnicity in daily interactions, described as "optional" (Waters, 1990) in the sense of being done with less certainty, people assign race based on physical appearance, namely skin color. Racism includes snap judgments made on appearance. Racism is a topic of study across the social sciences; an example in health services research includes the role of racism in health disparities (Thomas, 2001).

Communication studies, broadly, have been more concerned with culture as a variable (Jackson & Garner, 1998), which is more closely related to heritage and nationality included in definitions of ethnicity. Strength of ethnic identification varies across ethnic groups and is weak for many Americans (Hout & Goldstein, 1994). Ethnic culture can affect health behavior in, for example, diet (James, 2004) and care-seeking (Frisby, 2006).

Before leaving the description of race and ethnicity, the impact of social construction and self-identification on the interpretation and measurement of these variables should be mentioned. The social constructions of race and ethnicity evolve and vary through time and across context (Saenz & Morales, 2005). Cohen (1978) argues that subjective self-identification is the most conceptually valid measure of ethnicity because it represents the respondent's beliefs and thus the salience of their ethnic affiliation or identification. As a social construction, the same would apply to race. The Journal of the American Medical Association prefers, in studies it publishes, self-identification to identification by others (e.g., parent). To the point, its Deputy Editor, supported by recommendations from researchers in public health (Mays, Ponce, Washington, & Cochran, 2003), argues that "individuals should self-designate race to ensure that the designation most closely matches what they believe reflects their personal and cultural background" (Winkler, 2004, p. 1613).

James (2001) observes that, when used as a control variable, race has been treated as a fixed characteristic. Race, it is argued, should not be viewed as fixed in the way age, gender, and education are viewed. "[I]t is not race per se that is the causal factor, but rather how race has led some groups to be treated in society" (Denton & Deane, 2010, p. 69).

Suggested Measure

The measure suggested is to use the minimum categories of the OMB's revision of Directive 15 (Office of Information and Regulatory Affairs, 1997b). These are, for race: American Indian or Alaska Native, Asian, Black or African American, Native Hawaiian or Other Pacific Islander, and White. And for ethnicity: Hispanic or Latino and Not Hispanic or Latino. It is further suggested that, consistent with this directive, respondents be allowed to select more than one race.

The 2010 Census used additional, more specific, choices for the Asian and Native Hawaiian or Other Pacific Islander race categories (U.S. Census Bureau, 2010). The Asian category was specified into six racial choices plus "Other Asian" with a blank field to be filled. The Native Hawaiian or Other Pacific Islander category was specified into three racial choices plus "Other Pacific Islander" with a blank field to be filled. Hispanic ethnicity was labeled Hispanic, Latino, or Spanish origin and further specified into three racial categories and a blank to be filled for "another Hispanic, Latino, or Spanish origin, in addition to the choice of, "No," not of this origin.


The minimum categories and allowed multi-race designation of the OMB's revised directive are suggested mainly for reasons of comparability across research, since it is mandated for federal agencies that fund research in health communication. For example, the Health Information National Trends Survey (HINTS) follows this (Health Information National Trends Survey, 2008). Federal data and data from federally funded studies are also used as background or rationale for research, and consistency with data in the research could be important.

Ultimately, of course, whether race/ethnicity is included and, if included, what specificity is required depends on the research questions asked (and whether the funding source requires this data, regardless). Additionally, if used, the researcher must decide "which measure of race represents the best match with theory [used]" (Harris, 2000).

Additional Commentary

Consistency in self-identification is made more relevant by: 1) OMB's revision of Directive 15 that directs, beginning with the 2000 Census, the Census Bureau (and other federal agencies) to allow respondents the option of selecting more than one racial category; and 2) the growing number of Americans who are of more than one of the races that have historically been measured (Perez & Hirschman, 2009). Some research sheds light on consistency of racial self-identification across surveys. Furthermore, the NIH has called for demographic studies of race and ethnicity in the United States, which would include questions of measurement.

Analyzing patterns of racial classification in the National Longitudinal Study of Adolescent Health, Harris and Sims (2002) found that about 12% of youth provided different responses to almost identical questions at school and at home. Examining biracial groups of white/American Indian, white/black, and white/Asian, they found reported identity of white/American Indian to be more unstable across contexts than the other two groups. Hahn, Truman, and Barker (1995) analyzed patterns of racial classification in the First National Health and Nutrition Examination Survey and Epidemiologic Follow-up. Respondents who initially reported four different ethnic backgrounds were, ten years later, 3.4 times as likely to change their identity as those who reported one ethnic background. The greatest inconsistency between responses were among those who initially identified themselves as American or American Indian, the latter finding being consistent with that of Harris and Sims.

Hirschman, Alba, and Farley (2000) conducted a survey in anticipation of the 2000 Census that allowed multiple responses to the race question as written for the Census. They concluded that the planned change would have minimal impact on existing measured racial composition of the population: that "[a] small fraction of the largest groups, perhaps 1 or 2% of whites and blacks, will identify with more than one race" (p. 390). The other main finding was a dramatic reduction of non-responses in, of all questionnaire formats tested, the one with a combined race/Hispanic origin question. Consistent with earlier description, this suggests possible confusion in the inclusion of a separate Hispanic origin or ethnicity question but no Hispanic choice under "race."

The OMB has recommended Bridge Tabulation Methods for reassigning multi-racial counts into single race categories (Office of Management and Budget, 2001). These methods vary, including deterministic rules and probability-based techniques. An alternative bridging method has been proposed that shifts the unit of analysis from respondents to exhaustive racial identifications. Post hoc statistical corrections are then made to "allocate the impact of the multiple identification individuals" (Denton & Deane, 2010, p. 84).


Bates, N., de la Puente, M., DeMaio, T., & Martin, E. (1994).

Research on Race and Ethnicity: Results from Questionnaire Design Tests.
Paper presented at the Census Bureau's Annual Research Conference.

Denton, N., & Deane, G. (2010).

Researching Race and Ethnicity: Methodological Issues. In P. Hill-Collins & J. Solonos (Eds.),
The SAGE Handbook of Race and Ethnic Studies. London: SAGE Publications, Ltd.

Frisby, C. (2006).

Messages of Hope: Heatlh Communication Strategies that Address Barriers Preventing Black
Women from Screening for Breast Cancer. Journal of Black Studies, 37, 103-126.

Harris, D. R. (2000).

Demography's Race Problem.
Retrieved June 2, 2010, from

Health Information National Trends Survey. (2008).

Health Information National Trends Survey 2007.
Retrieved June 15, 2010, from

Hirschman, C. (2004).

The Origins and Demise of the Concept of Race.
Population and Development Review, 30(3), 385-415.

Hirschman, C., Alba, R., & Farley, R. (2000).

The Meaning and Measurement of Race in the U.S. Census: Glimpses Into the Future.
Demography, 37(3), 381-393. 

Hout, M., & Goldstein, J. R. (1994).

How 4.5 Million Irish Immigrants Became 40 Million Irish Americans: Demographic and Subjective Aspects of the Ethnic Composition of White Americans. [Article].
American Sociological Review, 59(1), 64-82.

Jackson, R. L., & Garner, T. (1998).

Tracing the Evolution of "Race," "Ethnicity," and "Culture" in Communication Studies.
The Howard Journal of Communication, 9, 41-55.

James, D. (2004).

Factors Influencing Food Choices, Dietary Intake, and Nutrition-Related Attitudes Among African Americans: Application of a Culturally Sensitive Model.
Ethnicity & Health, 9(4), 349-367.

Landers, E., Linton, L., Birren, B., et al. (2007).

Initial Sequencing and Analysis of the Human Genome.
Nature, 409, 860-921.

Mays, V., Ponce, N., Washington, D., & Cochran, S. (2003).

Classification of Race and Ethnicity: Implications for Public Health.
Annual Review of Public Health, 24, 83-110.

National Institutes of Health. (2001).

NIH Policy On Reporting Race and Ethnicity Data: Subjects in Clinical Research.
Retrieved June 14, 2010, from

Office of Information and Regulatory Affairs. (1997b).

Revisions to the Standards for the Classification of Federal Data on Race and Ethnicity.
Federal Register, October 30, 1997. Retrieved May 19, 2010, from

Office of Management and Budget (2001).

Provisional Guidance on the Implementation of the 1997 Standards for Federal Data on Race and Ethnicity [OMB Federal Register 2001].
Washington, DC: Retrieved September 29, 2011 from

Perlmann, J., & Waters, M. (2002).

The New Race Question: How the Census Counts Multiracial Individuals
New York: Russell Sage Foundation.

Perez, A. D., & Hirschman, C. (2009).

The Changing Racial and Ethnic Composition of the US Population: Emerging American Identities.
Population and Development Review, 35(1), 1-51.

Rivara, F. P., & Finberg, L. (2001).

Use of the Terms Race and Ethnicity.
Archives of Pediatrics and Adolescent Medicine, 155.

Saenz, R., & Morales, M. C. (2005).

Demography of Race and Ethnicity. In D. L. Poston & M. Micklin (Eds.), Handbook of Population (pp. 169-208).
National Institutes of Health, Bethesda, MD: Springer.

Schwartz, R. S. (2001).

Racial Profiling in Medical Research.
The New England Journal of Medicine, 344(18). Statement of Harry A. Scarr, Acting Director, Bureau of the Census, House of Representatives, 103 Congress, first Session Sess. 3-17 (1994).

Thomas, S. B. (2001).

The Color Line: Race Matters in the Elimination of Health Disparities.
American Journal of Public Health, 91(7), 1046-1048.

U.S. Census Bureau. (2010).

The Questions on the Form.
Retrieved May 20, from

U.S. Census Bureau. (2001, January 25, 2002).

Questions and Answers for Census 2000 Data on Race.
Retrieved May 29, 2010, from

Waksberg, J., Judkins, T., & Massey, J. (1997).

Georgraphic-Based Oversampling in Demographic Surveys of the United States.
Survey Methodology, 23(1), 61-71.

Waters, M. (1990).

Ethnic Options: Choosing Identities in America.
Berkeley: University of California Press.

Winkler, M.A. (2004).

Measuring Race and Ethnicity: Why and How? [Editorial].
Journal of the American Medical Association, 292(No. 13), 1612-1614.


PDF documents can be viewed with the free Adobe® Reader™