Galton’s Problem and Contagion in International Terrorism along Civilizational Lines

Abstract

If terror attacks from groups of one country are followed by similar attacks on the same target from groups of other similar countries, then this could be the consequence of contagion. However, just because one terror incident follows another does not necessarily imply that one is caused by the other or, in other words, that terror attacks are what is called spatially dependent. Rather, both incidents could have been triggered by the same underlying cause. This is known as Galton’s problem. One area where this problem is particularly prevalent is international terrorism. According to Huntington, international terrorism is contagious because of civilizational rallying effects. If radical groups from one country attack targets from a country of another civilization, then groups from other countries of the same civilization as the initial terrorist groups will become more likely to also attack this target. Any test of this hypothesis has to solve Galton’s problem and thus to disentangle spatial dependence from spatial clustering of attacks and common shocks and trends, which affect similar groups from different countries similarly. Accounting for such potentially confounding effects, we nevertheless find evidence for spatial dependence in international terrorism along civilizational lines in the post-Cold War period and particularly so for specific inter-civilizational combinations. However, while contagion consistent with Huntington’s predictions exists, spatial dependence seems to have a substantively small effect on patterns of international terrorism.