Presidential elections and real GDP growth in U.S.A.: 1961-1969

Abstract

Political influences on macroeconomic policies and outcomes are again widely studied phenomena. The two major theories of such influence are, (1) competing parties with differing policy preferences and (2) the political business cycle (PBC). Most recent empirical work concludes that the PBC does not affect output or unemployment, though it may affect policy variables. In this paper, I present significant evidence of a PBC in u.s. output growth. The estimated cycle is significant during both Democratic and Republican administrations and has stable coefficients across the 1961 - 1996 sample period. I also develop a new test to discriminate between first and second generation PBC theories, and find support for the new theory based on rational, forward looking voters. Finally, my results also confirm the importance of temporary, post-election rational partisan effects as developed by Chappell & Keech (1986), Alesina (1987) and Alesina & Sachs (I988). Because of the wide-spread belief that the PBC does not systematically influence output, I begin the paper with my empirical evidence showing a significant PBC, then work backward to discussing theories of the PBC. Section II summarizes recent results and techniques used in testing the PBC hypothesis. Section III contains my initial tests and results. Section IV discusses the difficulty of obtaining precise empirical predictions from current game-theoretic PBC models, and presents a new test designed to help distinguish between traditional and new PBC theories. Section V contains a discussion of possible new directions to take in exploring the coexistence of rational voters and PBC's and Section VI provides a brief summary and conclusion.