Nicholas Lemann's survey of some of the recent academic literature on terrorism is one of the more interesting New Yorker pieces I've read in a while. Here's his final paragraph:
Long ago, great powers that had vital interests far away simply set up colonies. That wound up being one of the leading causes of terrorism. Then, as an alternative to colonialism, great powers supported dictatorial client states. That, too, often led to terrorism. During the Bush Administration, creating democracies (by force if necessary) in the Middle East was supposed to serve American interests, but, once again, the result was to increase terrorism. Even if all terrorism turns out to be local, effective, long-running counterterrorism has to be national. States still matter most. And finding trustworthy partner states in the region of the world where suicide bombers are killing Americans is so hard that it makes fighting terrorism look easy.