Do Politicians Reward their Supporters? Regression Discontinuity Evidence from Ghana

We use a unique dataset collected from a random sample of schools in Ghana to test the “Patronage” hypothesis, that central government funds are targeted to districts that support the ruling party. We find evidence of patronage targeting to administrative districts, the level of government at which the national budget is allocated. However, there are no significant patronage effects at the parliamentary seat level. We further use 1996 and 2000 parliamentary election results to estimate incumbency effects for members of parliament, and can rule out even small incumbency effects (of 4 percentage points). These findings suggest a partial re-evaluation of the “conventional wisdom” on new African democracies: while there is some evidence of patronage, certain institutional arrangements – such as fiscal decentralization – partially limit the manipulation of public funds for political gain. These safeguards also appear to neutralize any potential electoral advantage that incumbents might reap from their position of power. *Correspondence to: Department of Economics, University of California, 549 Evans Hall #3880, Berkeley, CA 94720-3880, phone: (510) 642-7162, fax: (510) 642-6615, email: [email protected]. Frances Fontanilla has provided excellent research assistance. We thank Xiao Ye for her assistance with the data, and David Lee for many helpful discussions.