Abstract
Empirical voting/popularity research examines parameters that relate economic performance to voting and popularity, not the interaction of these parameters with movements in economic performance, which is what actually influences voting and popularity. This paper examines that interaction for the USA from 1953:1–1990:4 by decomposing the response of popularity to economic performance by both political party and the quarter of the electoral term. Averaged over the electoral term, economic performance increases the popularity of Democrats but decreases it for Republicans, whereas near the election the opposite occurs. Overall, the results appear consitent with a mix of partisan and electoral influences, so that no single approach seems fully adequate.