It is dogmatic to throw out all studies because they may have data dredging. It is prudent to carefully consider studies individually and look for any flaws in their methodology.
it's different in every area and sometimes those laws are stupid. So you're not really measuring crime rate, you're measuring *obedience rate*.
