On the Causal Interpretation of Race

To the Editor: VanderWeele and Robinson’s article “On the causal interpretation of race in regressions adjusting for confounding and mediating variables”1 is surprisingly uninformed regarding work on health inequities and injustice more broadly. Consider the following statements, all appearing with nary a supporting citation: “The notion of an effect of race is ambiguous: the effects may vary depending on what is meant by race. It may include skin color and its perceptions by others, parental skin color, and its perceptions by others, cultural context, or genetic background – all considered separately or jointly.”1 (p. 474) “However, race is not something we can intervene on, and the associated counterfactual queries generally strike researchers as meaningless. The question of what would a black person’s health outcome have been had they been white seems like a strange one to pose.”1 (p. 474) “Our discussion has focused on differences in outcomes across racial groups … A similar approach might also be used with other nonmanipulable exposures such as sex.”1 (p. 480) Really? On what scholarship are these assertions based? And who, by the way, is this “we” who decides what is “nonmanipulable” or feasible?—or designs or enacts interventions? One would never know, from this article or its citations, that an enormous body of work in health and social sciences, biology, philosophy, and the humanities has already addressed the article’s counterfactual confusions and has torn apart—conceptually, methodologically, and empirically—the article’s unsubstantiated claim that “race” is an a priori innate biological (ie, genetic) property of individuals2—and also has rebutted its deeply erroneous approach to conceptualizing “sex” with zero mention of “gender” (a word entirely absent from the article and its commentaries; forget even thinking about transgender!).2 Indeed, per Rawls, the question of whether whites would fare as badly as blacks, or men as women, if subjected to the same types of social injustice, has been repeatedly posed, since at least the early 1800s,2 by many working, in Kaufman’s concise causal terms, “to redraw the DAG, such that we would live in a society in which there is no longer an arrow going from race to SES at all.”3(p. 487) Perhaps it is a fitting irony that the empirical study published alongside this article and its commentaries and rejoinder focused on Jim Crow and health inequities.4 Generations of hard-fought action infused by democratic values have changed US race relations, race classifications, and racial/ethnic health inequities. “Non-manipulable”—really? Nancy Krieger Department of Social and Behavioral Sciences Harvard School of Public Health Boston, MA [email protected]