Economic Conditions and National Elections, Post-Sample Forecasts of the Kramer Equations

Abstract

There has long been a division of opinion among economists concerning the effects of economic developments on voting behavior. In his well-known study, Kramer (1971a) concluded that the fluctuations in economic activity have a systematic influence on U.S. congressional elections. His findings were challenged by various economists, including Stigler (1973).The purpose of this paper is to extend the analysis of Kramer's results by examining the post-sample forecast performance of his models. As noted by Christ (1966, pp. 546–49), the acid test of a time series model is its ability to perform well outside the sample period. The results discussed below demonstrate that Kramer's equations have relatively good post-sample predictive ability. The findings thus support the view that short-term economic fluctuations have important effects on national elections. However, the results also suggest that economic factors alone are not sufficient to account for all variations of voter behavior.