As I said in a previous blog, involvement in a religious community tends to reduce criminality among young men. But that depends on the religion. Here is an abstract (with my emphases) from a recent article:
Dirk Baier, “The Influence of Religiosity on Violent Behavior of Adolescents: A Comparison of Christian and Muslim Religiosity.” Journal of Interpersonal Violence 29, no. 1 (2014): 102-127.
“Different criminological theories assume that religiosity protects against violent behavior. Up to now, this assumption is tested empirically almost exclusively for the Christian religiosity. The study presented here questions whether such a relationship between religiosity and violent behavior could be found for Muslims, likewise. Using a German-wide representative school survey of 16,545 male students in the ninth grade, who belong either to a Christian or an Islamic denomination, it can be revealed that only for Christians a higher religiosity correlates with a lower rate of violent behavior. This influence of Christian religiosity can be explained by mainly control theory variables. For Muslims, there is no significant correlation between religiosity and violent behavior in a bivariate analysis. A multivariate analysis, however, reveals a suppression effect: Controlling for alcohol consumption, Muslim religiosity increases violent behavior. In addition, high religious Muslims agree more often to norms of masculinity and consume more often media violence, which are risk factors of violent behavior. Accordingly, it can be concluded that religiosity is not a violence-protecting factor in general; instead, a more differentiated view for separate religious groups is necessary.”
Yes, it depends on the religion. There are very few Methodist or Amish terrorists. Methodism is a church of older women, but the Amish have big families and many teenagers – and little crime.