That's a surprise.

Sometimes I end up being quite astonished by new data. This is an instance.

I have long assumed that some of the disparity in criminal action by race is a function of the relative availability of evidence, the low cost of that information between street crime and white collar crime.

Someone gets mugged or shot on the street (more common in black urban cores) and you often have witnesses, video, forensic evidence. Not saying that it is cheap or easy to prove the mugging or shooting. Just saying it is easier and cheaper than when dealing with a white collar crime.

The mugger gets away with a watch and $20. And is caught and convicted and jailed.

The white collar embezzler quietly defrauds another person of $20 a week for twenty years. No one sees. There often is not much of a paper trail. Frequently the distinction between a fraud and an innocent error is hard to make. The fraudster in some ways commits the more serious crime; not just for depriving another of more dollars but also for undermining societal trust upon which we are so dependent.

So if most street crime is committed by African-Americans and most white collar crime is committed by white Americans (both being crude generalizations), you are likely to see the former group being arrested and convicted at a much higher rate. Not necessarily because they are committing worse crimes (though physical violence does brings its own special considerations) but because the nature of their crimes is easier to see and convict. Nothing systemically racist because there is no intended outcome. It is just a function of costs and evidence.

As a consequence, I have long carried an unstated assumption that it is possible that white Americans may cause a greater crime burden than that which is acknowledged because of their possible propensity of committing more white collar crime than street crime.

Then I come across this from The Oxford Handbook of White-Collar Crime by Shanna Van Slyke. The data is somewhat dated (from 2005-2011). It is Federal convictions so the crime pattern is different from what state or municipal might look like.

Here is the essence of the data.

Click to enlarge.

African Americans and Hispanics are roughly 30% of the population and per these numbers committing a disproportionate share of crime. Nothing new there.

But look at the split between street crime and white collar crime. In both cases African Americans and Hispanics are committing a disproportionate volume of crime but also a slightly higher percentage of white collar crime as street crime.

That is totally unexpected. I have looked at how they define white collar and street crime and tried to assess whether the nature of federal crimes versus State and Local might account for this unexpected pattern.

I do not have a good answer. No obvious reason that I can see.

Is the data indicative of an underlying truth? No idea. I cannot assess.

But I now need to at least attach an asterisk to the lazy assumption that the relative ratio of street versus white collar crime commission might be different between groups. Looks like the rates of commission might be the same within a group, whether street or white collar.

Hmmm.