Racial gaps on (insert outcome here): have 3 parts- (1) structural racism, (2) individual factors, and (3) discrimination/implicit bias. Generally researchers are taught to investigate (2) using covariates to explain away whatever the gap is. This is flawed.

Keep that in mind as the economists (or whoever else) start to weigh in why black people are dying at the highest rate from COVID-19, they lack the ability, desire, vision,and analytical tools to properly discuss (1) and (3).

And most of the stuff that’s happening at the individual level has structural components. For example- weight and stress. Black people are stressed. Part of the reason black women have high mortality in birth is “weathering”.

That being said, early data re: race + COVID is distressing. Before you race to put out a working paper, make sure you’re not falling into the trap of throwing everything into a regression and “solving” the problem. The effects work through the covariates + independent of them.