Abstract
Contrary to the extant literature, this paper shows that the impact of terrorism on economic growth is insignificant in five regional samples. These surprising results follow when Nickell bias and cross-sectional dependence are taken into account. Previous studies have not properly adjusted for these biases. Our provocative findings are robust to alternative measures of terrorism (i.e., domestic, transnational, and total) and to alternative specifications of terrorism (e.g., level, first differenced, and attacks per capita). These findings are also robust to the inclusion of investment and population growth. Moreover, we find that the various forms of terrorism do not affect consumption, investment, and government expenditures. Our results have important policy implications.
