>>105632639I think the point being made is that stats on their own don't have that much value.
In the scientific community, corollary data is considered only slightly better than case stories - that is, anecdotes. Corollary data doesn't really prove anything, it's merely a decent place to pose a hypothesis and get to actual testing.
In terms of crime data, it's pretty simple to come up with a non-racist hypothesis. Black people have traditionally been pushed into poor neighborhood, zip code weighs a lot on success in the US so poverty begets poverty, the police know there's crime in these in impoverished areas so they spend a lot of additional time shaking down those neighborhoods. If you also assume there's profiling behind who gets pulled over, it's not surprising the arrest stats would be skewed like that.
I mean, when they were looking for Ted Bundy in Aspen, during traffic stops they wound up collecting hundreds of pounds of marijuana. From fucking Aspen. They just normally didn't shake down Aspen.