Before you criticize it, already know the flaws, but it's the only government sourced study. There's lots of independent studies for sociological correlations to crime rates, education levels, religion, climate, race, religion, substance abuse; pick a reason. Not saying any one is correct. The only single factor which stands out above and beyond any other factor, is gender, and that can be consistently applied on a global scale. That could definitely be sociological. If wanting to stick to the race topic and consider genetics, studies for testosterone levels have also been correlated to crime for both males and females.