Reconstruction of Pre-war U.S. Business Cycle Dates Using Markov Regime-Switching Model

Abstract

This article attempts to (re)assess the post-war stabilization hypothesis for U.S. economy. Unlike most works that look at the issue in terms of volatility and/or amplitude aspect of business cycles, I approach the stabilization issue from the duration perspective, in line with Diebold and Rudebusch (1992) and Watson (1994). One of the distinguishing features of the article is, in identifying the pre-war boom and recession periods, to employ the Markov regime-switching model and construct alternative pre-war business cycle reference dates. These newly constructed business cycle dates from the regime-switching model provide useful and significant implications for pre-war business cycle fluctuations. Finally, based on the newly created dates, I test the post-war stabilization hypothesis. The empirical results largely support the hypothesis.