One of the most interesting things I heard at the APSA was an remark made by my fellow-panelist Mark Tessler of the University of Michigan. In response to a question about Arab attitudes towards the United States, he cited a paper he recently completed for the National Defense University (to be published soon) surveying the findings of ten different data sets collected over a fifteen year period in Egypt, Jordan, Palestine, Algeria, Morocco, and Turkey.
One set of findings wasn't surprising, but bears noting: the data consistently show that there is no relationship between levels of personal religiosity and attitudes towards democracy. Such findings should eventually begin to wear down the view that Islam is somehow incompatible with procedural democracy.
But the really interesting part was this: he ran regressions on nine different families of variables - everything from education to gender to religiosity to income - to find out what best explained expressions of support for terrorism. None of these shows any stasticially significant relationship. The only independent variables significantly correlated with support for terrorism were "negative views of U.S. foreign policy" and - and here's the interesting part - "negative views of one's own political system."
Those Jordanians and Algerians who expressed dissatisfaction with their own government were significantly more likely to approve of terrorism against the United States. Discontent with their own government was vastly more significant than religion, education, culture, class, or anything else for explaining support for terrorism against the U.S. Which offers some really interesting support for the idea that there is a clear American interest in promoting reform in such countries - even if Greg Gause is right that doing so will not itself end terrorism.