Using term limits to estimate incumbency advantages when office holders retire strategically

Empirical study of U.S. elections over the last 50 years has documented a strong electoral advantage to incumbency in state and federal elections. Recently, however, critics have argued that traditional estimates of the incumbency advantage may overstate the advantage by as much as 100% because the estimates fail to consider strategic retirements. This article directly examines whether or not strategic retirement biases conventional regression estimates of incumbency advantages. We use term limits in state executive and legislative elections as instrumental variables to correct for strategic retirement. We find that, as an empirical matter, strategic retirement is not substantively important. Estimates of incumbency advantages that take account of strategic retirement actually are marginally larger than estimates that do not.