In his empirics, he employs negative binomial regression on
a data set consisting of a panel of one hundred twenty seven countries for the years 1970 through 2007.
His control variables include income inequality, democracy, state failure, population, and a post-cold
war dummy. He runs regressions of overall terrorism, and the three different types of terrorism,
domestic terrorism, international terrorism, and suicide terrorism, on overall economic growth, and
alternatively, on the two components of economic growth, agricultural sector growth and industrial
sector growth.