Despite the plethora of studies demonstrating that economic perceptions affect how a person votes, relatively little is known about how economic perceptions affect whether individuals will vote. Using the calculus of voting as our starting point, we develop a simple, but novel, hypothesis regarding the influence of sociotropic evaluations on voter turnout. We argue that this relationship will be curvilinear, with particularly negative and particularly positive evaluations of the economy increasing the likelihood of voting. Using an instrumental variables approach with individual-level data from eight recent U.S. presidential elections, we find that economic evaluations affect the decision to vote in the curvilinear manner hypothesized, but—counter to existing theory—only when there is not an incumbent president seeking reelection.