Evidence from Public Opinion Polls in Pakistan

Abstract

Reducing terrorism is one desired outcome of increased development aid to Pakistan. Targeting of this aid toward education rests on the conventional wisdom that the more educated are less likely to become terrorists and to support terrorism. Recent evidence, however, shows that education levels are not negatively related to terrorism in various contexts. In this paper, I use data from public opinion polls in Pakistan (PIPA 2009 and 2007) and econometric analysis to show the following novel results: 1) in the cross-section, more educated women are less likely to support militancy and terrorism relative to educated men, whereas uneducated women are more likely to support militancy and terrorism relative to uneducated men, controlling for demographics and region; 2) this result continues to hold when the years of data are combined in a pseudo-panel, controlling for time effects; 3) it also holds after accounting for actual terrorist events in the district using data from the Global Terrorism Database; 4) the result reverses when it comes to views of the United States, so that more educated women have more negative views of the US relative to educated men. Terrorist events in the district also directly reduce support for terrorism. The resulting policy prescription is clear: to successfully counter support for terrorism, we must focus on increasing the education of girls in Pakistan.