Inefficient Policies and Incumbency Advantage

We study incumbency advantage in a dynamic game with incomplete information between an incumbent and a voter. The incumbent knows the true state of the world, e.g., the severity of an economic recession or the level of criminal activities, and can choose the quality of his policy. This quality and the state of the world determine the policy outcome, i.e., the economic growth rate or the number of crimes committed. The voter only observes the policy outcome and then decides whether to reelect the incumbent or not. Her preferences are such that she would reelect the incumbent under full information if and only if the state of the world is above a given threshold level. In equilibrium, the incumbent is reelected in more states of the world than he would be under full information. In particular, he chooses ine±cient policies and generates mediocre policy outcomes whenever the voter's induced belief distribution will be such that her expected utility of reelecting the incumbent exceeds her expected utility of electing the opposition candidate. Hence, there is an incumbency advantage through ineficient policies. We provide empirical evidence consistent with the prediction that reelection concerns may induce incumbents to generate mediocre outcomes.