I leverage exogenous variation in the historical origins of chieftaincy to study the effects of traditional leaders on voters’ ability to extract state resources. Using original data on the history of traditional institutions in Northern Ghana combined with fine-grained census data, survey data, and polling station-level election results, I show that communities with chiefs from ethnic groups assigned to the colonial invention of chieftaincy in the late-19th century have less leverage to benefit from patronage exchanges with politicians today. I argue that this is because traditional institutions invented by colonial authorities are especially prone to elite capture, empowering electoral intermediaries who engage in rent-seeking. The paper demonstrates the contemporary importance of the historical origins of chieftaincy in Africa and identifies conditions under which voters benefit from brokered politics in clientelistic political systems.