Comment by lanternfish

They come to the conclusion that their crime numbers are reflecting underlying density differences, but then don't redo the analysis with a per-capita control? Seems like, without that, this isn't a very useful presentation.