Anyone pondering systemic racism for a while will quickly realize that it is very hard to measure.

E.g. you might think one can look at police statistics ins a straightforward manner. However you cannot, because the ones making the statistical recordings are the ones we want to study the bias of.

Thus we need natural experiments. The kind of tactics used here is widely used in other areas. Have you read Freakonomics? It has an abundance of examples where studies are conducted in this manner.

You look for special cases that form natural experiments, that give you an ability to study the phenomenon. A benefit of a lot of these studies is that they are conducted in such a manner that it is very hard for the people with the racist biast to cover up what they are doing.

E.g. take something like fixing a boxing match. You cannot prove an individual case. But when match fixing happens on a regular basis you can see it in the statistics. It will be skewed.

It is the same here with systemic racism. It is next to impossible to prove in idividual cases. However a systemic bias will easily show up in the statistics.

E.g. individual prosecutors/police who lie on how much a suspect had of drugs on them, know that they are unlikely to get caught. But in the study I referred to you see that the statistics catch their lying, beause it skewes the numbers over time in a very predictable pattern.

I honestly don't see a problem with the studies done. I think they are very cleverly conducted. And honestly there is nothing surprising here at all. Anyone who has lisitened to the stories African-Americans tell, read American history or just observed life in America themselves can see the patterns clear enough. But I suspect many Americans are so invested in the ideals of America that they cannot see what is plain before them.