Does religiosity have an impact on fatalities from violent behavior? Is this impact stronger than the effect of religiosity on fatalities from causes other than violent behavior? In this
paper, I have employed three instrumental variables to find the answers.