A Markov Switching Model of GNP Growth With Duration Dependence

Abstract

We use a regime-switching model of real GNP growth to examine the duration dependence of business cycles. The model extends Hamilton (1989) and Durland and McCurdy (1994) and is estimated using both the postwar NIPA data and the secular data constructed by Balke-Gordon. We find that an expansion is more likely to end at a young age, that a contraction is more likely to end at an old age, that output growth slows over the course of an expansion, that a decline in output is mild at the beginning of a contraction, and that long expansions are followed by long contractions. This evidence taken together provides no support for the clustering of the whole-cycle around seven-to-ten year durations.

---