Terror per Capita

Studies on the correlates of terrorism usually analyze total numbers of attacks or victims per country. However, what we may ultimately care about in terms of policy recommendations is the likelihood of any individual being subject to the respective phenomenon. Thus, we propose and explore a simple alternative measure of terrorism: terror per capita. Studying terror per capita across 162 countries from 1970–2015, the associated correlates differ substantially in terms of sign, levels of statistical significance, and magnitude from those when analyzing total terror. We illustrate two cases in point, serving as proof of concept. First, democracy, often associated with more total terror, emerges as a marginally negative predictor of terror per capita. Second, a larger share of Muslims in society is associated with a positive and statistically significant link to total terror, but emerges as a negative predictor of terror per capita. We find similar changes in sign and statistical relevance for GDP per capita and language fractionalization as correlates of terrorism. Depending on the policy question, studying terror per capita can greatly enhance our understanding of terrorism drivers, especially when analyzing data across countries with vastly differing population sizes.