Though it should be noted that the publicly-available statistic with the highest correlation with the drop in violent crimes is... drumroll, please... the drop in lead contamination.
Yes, America is safer primarily (presumptively) because it banned lead-containing antiknock molecules in gasoline and banned lead-containing pigments in paint. The dispersion of remaining lead contamination out of the urban environment over the years so closely tracks the drop in violence that it almost makes one believe the correlation is due to causation. Almost. Unfortunately, experimental confirmation would be grossly unethical.
This generalization gets bandied about too much. Crime trends and patterns are highly local. I'm sure violent crime is dropping in many places, but in other places it's spiraling out of control, for example, South Chicago, East Oakland, Detroit. I think it's a mistake to throw those places in the same basket as a small city in Ohio (unless it really has similar crime stats).
