Render unto Caesar: Taxes, charity, and political Islam

Using an original, nationally representative survey of 600 Tunisians, we show that support for the Islamic party in the first post-Arab Spring election came from wealthier districts and individuals. We demonstrate that standard public finance arguments explain this voting pattern better than other available explanations. Our model predicts that a voter's probability of voting for a religious party: (i) increases with income for the poorest voters, but possibly decreases with income for the richest; (ii) is greater for voters in richer districts; and (iii) increases with the voter's religiosity. Our empirical results align with our predictions and suggest that individual and district wealth were key drivers of support for the Islamic party. We test for other possible factors affecting voting, such as economic disgruntlement, migration, access to media, or attitudes towards gender parity or towards the West. Finally, we document similar patterns in other key elections in the Muslim world. (C) 2019 Elsevier B.V. All rights reserved.