Best practices for using race in public health research

Download the best practices guide

Why do we talk about how we use race? There are a number of concerns with using race, so we should do so intentionally. Heading link

  • Race is a social construct

    Race is a social construct that was created to differentiate people based on physical features like skin color and was used to justify racism, eugenics, experimentation and other societal horrors.5

  • No biological basis

    Race does not have a biological basis,5,8 so it is conceptually questionable to equate race with biological outcomes.

  • Race is not a clear construct

    Race is not a clear construct and lacks consensus definitions or has changing definitions.4,7,8

  • Imprecise measurement

    Using race results in imprecise measurement.4,7,8

  • Confounding factors

    Race is confounded by social class, socioeconomic status , nationality, ethnicity – affecting its overall validity for even being a proxy for racism.3,5,7–10

Even if your research aligns with health equity principles, research that focuses on structural and social determinants of health generally has race as a demographic factor used to identify groups that have experienced historical disadvantage and continue to experience discrimination and systemic racism. The National Institutes of Health continue to require researchers to report and analyze race as a demographic characteristic of the sample.5,11 We must push to change this policy and remain reflexive and intentional in our use of race in health research.2,5

If we want our research to address racial equity, it must explicitly name racism as a factor and acknowledge the context in which racial inequities occur, including those that examine the role of other determinant s of health that interlock with racism, such as socioeconomic status, resource accessibility and place.2,7,10,12,13 While this is true, race is broadly used as a proxy for other constructs such as racism, skin color, discrimination, disadvantage, etc. Further, we must engage in antiracism praxis, where the knowledge gained from theory, research, personal experience and practice inform one another.14

Best practices to address concerns with using "race" Heading link

Thoughtful study design

Show that you have thoughtfully considered race, ethnicity, and disparities in your study design and interpretation.2,5

  • If you are using race as study variable, specify the reason why, along with the limitations of using race as a demographic variable.2,7,9,14
  • Indicate how race is measured in your data and its limitations as a valid measure. When doing original data collection, consider asking participants to self identify.8,9
  • Note the limitations with the typical Office of Management and Budget (OMB) and Census categories (listed on page 2 for reference).5,7,9,14,15

Let hypotheses drive the use of race

  • Do not assume race is relevant to your hypotheses.5 Do not use race and ethnicity when ther e is no biological, scientific, or sociological reason for doing so.9
  • Consider if you are using race as a proxy for ethnicity, culture, skin color, discrimination, a genetic marker, or racism, for which there are more valid measures.3,4,7–9,15,16 Include these measures in your study.
  • Consider the implication s of “controlling ” for race in statistical analyses.8

Pathologizing and medicalizing race

Avoid pathologizing or medicalizing race2,7,24 and supporting a narrative of race-based medicine, that there is a biological difference between racial groups.5,25,26

  • Do not use white as reference group. This replicate s the notion that non-white groups are “other”.24
  • Consider the implication s of excluding data from a specific subset of the sample.8
  • Do not use race as an explanatory variable.9

Acknowledge intersectional identities

  • Describe the heterogeneity of groups8,15,16,24 and examine data by at multiple intersections of social identity, such as race, gender, socioeconomic class, sexual orientation and citizenship status.
  • Examine models within racial groups.8

Racism as a fundamental cause of health inequality

Treat racism as a fundamental cause of health inequality when choosing methods and interpreting results.7,10

  • Consider using critical approaches that consider history and systems of oppression14 such as critical feminist theory,19,20 decolonial theory,21 critical Whiteness studies,22,23 or critical race theory, which elucidates contemporary racial phenomena, expands the vocabulary with which to discuss complex racial concepts and challenges racial hierarchies .17–19
  • Include measures related to racism and the social and structural determinants of health as relevant to your research (i.e. housing age and density, housing quality, distance to grocery stores, percentage of population criminal justice system involved, etc.).3 Much of the time, including race and socioeconomic status such as education and income are rudimentary and do not address the complex factors linked with racism.9
  • Utilize research approaches that allow for sharin g experiences of marginalized people.
  • In discussions that focus on differences between racial groups, emphasize that that race itself is not a risk factor; racism is.3,8
  • Define populations by the presence and impacts of structural racism rather than ethnically or racially. Draw attention to the large r social, historical and environmental context of groups.15,16

We are presenting these data by ‘race.’ We are using ‘race’ here as a proxy for racism. While ‘race’ is socially constructed and has no genetic basis, racism has real biological, physiological, political and economic consequences. These consequences are rooted in state-sanctioned historical and contemporary racial oppression. We acknowledge that racism is complex. We challenge you to critically engage with this material and question the use of ‘race’ without the context of racism. Public health and other health professionals have a responsibility to contextualize measures of inequity along the lines of ‘race’ with a discussion of the negative impacts of racism on health.27

Radical Public Health

This statement was developed by the 2016 Curricular Praxis Workgroup of Radical Public Health, a group of students, faculty, practitioners, and alumni connected to the University of Illinois Chicago School of Public Health who insist on addressing the roots of public health issues. It is intended for members of our school community to use when presenting data by race. Similar to the way that a map is expected to include a legend, compass rose, and scale for it to be usable, the authors contend that charts, graphs, maps, and text that describe inequities along “racial” categories cannot be accurately understood without the context of racism.27

If we acknowledge and name racism in our work, writing, research, and interactions with [students] and colleagues, we can advance understanding of the distinction between racial categorization and racism and clear the way for efforts to combat the latter.

Rachel Hardeman

OMB and US Census Race & Ethnicity Categories (for reference) Heading link

  • American Indian or Alaska Native

    A person having origins in any of the original peoples of North and South America (including Central America), and who maintains tribal affiliation or community attachment.

  • Hispanic or Latino

    A person of Cuban, Mexican, Puerto Rican, South or Central American or other Spanish culture or origin, regardless of race. The term “Spanish origin” can be used in addition to “Hispanic or Latino.”

  • Native Hawaiian or Other Pacific Islander

    A person having origins in any of the original peoples of Hawaii, Guam, Samoa or other Pacific Islands.

  • Asian

    A person having origins in any of the original peoples of the Far East, Southeast Asia or the Indian subcontinent including, for example, Cambodia, China, India, Japan, Korea, Malaysia, Pakistan, the Philippine Islands, Thailand and Vietnam.

  • Black or African American

    A person having origins in any of the Black racial groups of Africa.

  • White

    A person having origins in any of the original peoples of Europe, the Middle East or North Africa.