That seems exceptionally high, and I'm assuming the study was mostly (all?) urban subjects, which could skew the results heavily. I can't think of more than 2 people I know that have an arrest history, and I know a LOT of (non-urban dwelling) people.