Field interviews conducted by the author in Afghanistan suggest current theories linking conflict to development do not adequately account for ideological drivers of resistance. We present a model demonstrating how reconstruction/development led by a foreign occupier can exacerbate violence through popular discontent, if projects are ideologically controversial. We test the model using detailed data on military-led reconstruction and public opinion from NATO, and a US-Government violence log covering Afghanistan from 2005 until 2009. We find projects in the health sector successfully alleviate violence, whereas those in the education sector actually provoke conflict. The destabilizing effects of education projects are strongest in conservative areas, where public opinion polls suggest education projects breed antipathy towards international forces. Further underscoring the role of local perceptions, project-driven violence appears to be homegrown, rather than sourced externally. Our findings do not support competing theories; are not driven by reverse causation; and are robust when considering many sources of endogeneity.