| Abstract: | This paper presents two tests of the rational partisan theory (RPT) of business cycles. First, I develop and test an RPT model in which wage contracts are staggered and overlapping. Bureau of Labor Statistics (BLS) data on the density and duration of wage contracts and estimates of the incumbent party's election-win probability are employed to calibrate partisan intervention variables entered in output growth regressions. Next, I perform a more “flexible” test of the RPT by comparing partisan output growth differences after elections in which the outcome was relatively surprising with partisan output growth differences after elections in which the outcome was widely anticipated. The two approaches produce the same conclusion: The RPT is not supported by the data. |
