Interesting. From Driving While Black:
Like most people, I suppose, I had thought that there was some pretty firm statistical evidence that racial profiling was going on in regard to highway stops — i.e., that black drivers are a lot more likely, all other things being equal, to be pulled over by cops than white drivers. According to an article by Stephen Michelson in the most recent Jurimetrics, while racial profiling of this kind may be occurring, none of the studies purporting to demonstrates its existence actually does so.
The argument, in more detail:
But Michelson points out that, first, none of the studies measures the extent to which cops can actually perceive the race of drivers as they whiz by at 75 mph. If cops were completely unable to tell which cars are driven by blacks and which by whites, of course, whatever was accounting for this disproportionality could not, of course, be “racial profiling.”Second, and somewhat more complicatedly, the relevant baseline for comparison with the proportion of stopped cars with black drivers is not “the proportion of black drivers going over the speed limit,” but rather “the proportion of black drivers going so far over the speed limit that they are likely to be stopped.” That is: suppose (as is the case) one of the studies shows that 25% of blacks go above 55 mph in a 55 mph zone, but that 45% of the stops there are of black drivers. Looks suspicious. But nobody gets stopped for driving 56 mph, or even 62 mph, in a 55 mph zone. The real question is: what proportion of black drivers are exceeding the speed limit by a significant enough amount to be likely to be pulled over — i.e., what proportion of black drivers is going, say, 75 or higher. If 45% of the drivers going over 75 mph were black, the data’s no longer so suspicious. And it turns out that all of the data collected thus far are consistent with the idea that blacks are significantly over-represented in this category — enough to account for the higher proportion of blacks among the stopped cars.