On the causal interpretation of race in regressions adjusting for confounding and mediating variables

On the causal interpretation of race in regressions adjusting for confounding and mediating variables