No, it's not. It's very dangerous to draw correlations like these(i.e. see all the "crime going down"/"crime going up" explanations) because of the huge potential for noise; it's not like we only take one societal decision per year to be able to measure its actual effects.