Mark Kleiman, one of the smartest scholars on crime and punishment, offers this explanation.
Crime is linked to long-term concentrated poverty and deprivation; there’s no evidence that it’s linked to unemployment rates...
One explanation that appeals to me is that there’s a strong positive feedback built in to crime rates; the fewer crimes, the more cops and the more cells per crime, so crime declines and crime increases both tend to feed on themselves.
\n
Of course, white collar crime isn't on this chart. I'd be interested to see those numbers. And giving a bunch of worthless mortgages a AAA rating, for example, doesn't even count as a crime.