‘Population profiling’ and public health risk: When and how should we use race/ethnicity?

Abstract

The limited reliability of race/ethnicity and the sensitivity surrounding its use to stereotype, discriminate and rationalize difference suggest it should not be used in population profiling for public health risk. However, race/ethnicity is crucial for assessing the risk of discrimination along the lines of race/ethnicity, and it can be the best available proxy for important unmeasured variables. Nonetheless, generating any data disaggregated by race/ethnicity can fuel the use of biological reductionism or cultural essentialism to explain inequalities in health. To maximize the benefits and minimize the disadvantages of using race/ethnicity this commentary suggests that: (1) race/ethnicity should primarily be used to assess the risk and impact of discrimination; (2) race/ethnicity should only be used as a proxy for variables that cannot be measured and when it is the most reliable proxy available; and (3) the use of race/ethnicity may be desensitized by emphasizing how categorization and discrimination are ultimately responsible for the utility of race/ethnicity as a proxy for related variables. These suggestions accept that race/ethnicity will continue to be used in population profiling, but that its use could be refined and reframed. To this end, the commentary explores the dilemma faced by the South African National Blood Service (SANBS), which was criticized for using race/ethnicity to identify blood donors at increased risk of undetectable HIV infection. While race/ethnicity might not have been the only reliable proxy for undetectable HIV infection, its use might have been more palatable had the SANBS explained how the utility of race/ethnicity reflected the legacy of apartheid and its enduring impact on the differential spread of HIV.

Acknowledgements

The author is grateful to Jane Wills and Thea de Wet for helping to draw these arguments together. This commentary, and the responses which follow, are supported by grants from the Wellcome Trust's research programme in biomedical ethics: 057182 and 073524/2/03/2/AW/HH.

Notes

 Although ‘race’ and ‘ethnicity’ are often interpreted as measuring different (biological vs. social) aspects of group identity, both are social constructions and both are associated with a range of biological and social variables (Oppenheimer, 2001). For this reason they are often used interchangeably and debates about what each of them means distract attention from their common role in stereotyping social groups (through biological reductionism and cultural essentialism). For this reason, and given the focus of this commentary, I have adopted the term ‘race/ethnicity’ throughout to avoid drawing any unnecessary distinctions between the two.