You're completing missing my point...those two variables (crime data with encryption vs. crime data without encryption) are not measurable as there are many factors outside that can directly impact crime data alone... increase/decrease in the amount of officers on the street, directed patrols in certain areas resulting in decreased crimes, number of crimes being reported, increase/decrease in population/immigrant surge in NYC, etc etc etc.
So while you want to stand on your internet soapbox and exclaim "provide your documentation," i'm saying you shouldn't even be standing on that soapbox using Nat as a measurement, so get down before I keep knocking you off.