Read my lips: The case against election promises

We study a model in which politicians differ in their ability to im- plement some policy. In an election, candidates make binding promises regarding the plans they will implement. These serve as a signal of true ability. In equilibrium, candidates make overambitious promises. The candidate with the highest ability wins. Yet, the electorate may be better off having a random candidate implement her best plan, rather than seeing the election winner implementing an overambitious plan. This is more likely if the distribution of abilities is skewed toward high values, in the case of private benefits from being elected, or if parties select candidates.