Modeling Political Change with a Regime-Switching Model

In: European journal of political economy, Band 16, Heft 4, S. 739-762

Abstract

The purpose of this paper is to investigate whether legislators vote in two distinct regimes as suggested by the US political system & whether these voting patterns can be explained by economic variables. The paper exploits US Roll-Call votes from a rich unbalanced panel data set of 540 legislative positions over 104 congressional sessions & presidents since the Korean War. The paper employs a variant of the Hamilton Regime-Switching Model (1989) to uncover three major patterns in the data. There appear to be two distinct regimes closely related though not identical to the two-party system. Shifts in these regimes are most frequent in presidential elections & are less frequent in the Senate & the House. Finally, since WWII, economic variables do not seem to influence the shifts implying that these political shocks are exogenous. However, over longer horizons, recessions do tend to force political change in the legislature. 7 Tables, 5 Figures, 33 References. Adapted from the source document.