Dr. Christopher Williams' Open Letter to Dr. Chandra ford and Dr. Rae Anne M Martinez, following commentaries in the American Journal of Epidemiology

Dear Dr. Ford and Dr. Martinez, 

I have been following your recent commentaries in the American Journal of Epidemiology. I am the principal investigator for the Critical Race Framework Study. This study published the first critical appraisal tool concerning the use of race in public health research. I was hoping that your commentaries would have more fully discussed the statistical threats to scientific quality due to the use of race, as was laid out in the CR Framework Study. Insistence upon refining race, even within race-conscious methods or better defined meanings of race and ethnicity, is apt to considerable effects on research quality. Dr. Ford highlights a reliability issue, "Similarly, an emigrant from North Africa may have learned to “self-report” White race even if they actually thought of themselves as African American." Except, a validity evaluation would question the significance or scientific value of, say, a MENA designation. I do not support the insistence that a "Black race", or any other globalized notion of race, has scientific value - a lingering legacy of slavery and scientific racism. Dr. Martinez and colleagues' position -- "Black as a racial group (of which one may be any ethnicity)" - is an ideological position without any scientific or evidence-based support. Even so, the relevance of race is only important in epidemiology to the extent that there are shared attributes, exposures, or other attributes that hold weight for scientific advancement. 

Then, there is the question of statistical versus practical significance. Even if studies align with refinement of race and race-consciousness, those studies do not necessarily provide for 1) an inclusive public health research because a) inequity is not just a manifestation of racism (see Public Health Liberation theory) and b) racism risks marginalizing populations for which structural racism does not traditionally apply and 2) does not enhance practical significance because of a) the differences of within-group experiences across various public health economies and b) statistical hacking. The position of the Critical Race Framework Study is clear, "Other disciplinary theories or approaches about race in research are only relevant to subsequent stages of the Critical Race Framework study if they can contribute to reliability or validity and successfully defend internal and external validity in current practices." The phenomenon of statistical hacking perpetuates research practices that do not hold meaning for public health translation - policies, interventions, resource redistribution. Data analysis can become so mired in statistical analysis that it lacks insight for accelerating health equity and reform. In other words, statistical hacking can miss the forest for the trees. There is also the risk of race/ethnicity refinement assuming a monolith, as in the treatment of Black MSM, "Critical Examination of Race in HIV-Related Systematic Reviews of Black Men Who Have Sex with Men in the United States: A Methodological Review". Two articles are also instructive: Indices of neighborhood disadvantage and individual cancer control behaviors among African American adults: An Article Critique and Dr. Christopher Williams, Principal Investigator of the Critical Race Framework Study, Raises Alarm over “Statistical Hacking”.

I also am the lead author of Public Health Liberation theory. It views explanations of racism as underdeveloped or over-deterministic as explanatory models. Every barrier to health inequity for minoritized communities cannot plausibly be explained by racism because there are other dynamics at play to include the lack of horizontal and vertical integration, interest-driven action (i.e., public health realism), academic estrangement, legal theories and practices (e.g., requirements for standing, statutory obligations of public attorneys to defend governments, including bad ones), lax regulation, governance coalitions and politics, etc. Another reason why racism is not the highest quality analytical framework is because the field still struggles to form a consensus on structural racism conceptualization and operationalization, part of the dynamics attributable to the fragmentation of the public health economy that permits ever-shifting, agent-led agenda-setting. The performance of macro and micro public health economies varies considerably spatially and temporally. An African American (ADOS) born in DC in the 1960s would have stark life outcomes than an African American (ADOS) born in Central Virginia in the same period, as would with an African American (ADOS) born in DC in 1999, given the severe underperformance of DC's public health economy until the beginning of the 21st century.  A recent study highlights such differences, "All-Cause Mortality and Life Expectancy by Birth Cohort Across US States".

Sincerely, 

Christopher Williams, PhD