INTERNATIONAL ECONOMIC REVIEW
Vol. 30, No. 4, November 1989
TIME-TO-BUILD AND AGGREGATE FLUCTUATIONS: SOME NEW EVIDENCE*
BY SUMRU 1
This paper presents maximum likelihood estimates and tests of a model similar to one Kydland and Prescott (1982) suggested. For this purpose, it derives equilibrium laws of motion for a set of aggregate variables as functions of the model’s parameters and the innovation to the technology shock. The paper shows that a single unobservable index can explain the variability in the observed series, but identifying the single index with the innovation to the technology shock implies that per capita hours is not well explained. It also shows that time-separable preferences with respect to leisure are consistent with the data.
In a paper that has received much recent attention, Kydland and Prescott (1982) presented a competitive equilibrium model of cyclical fluctuations. Using the one-sector optimal growth model to construct a prototype competitive economy, they argued that postwar U. S. business cycles could be explained in terms of the dynamic response of the aggregate economy to persistent technology shocks. To provide a more complicated propagation mechanism for such shocks, they modified the optimal growth model by introducing a time-to-build feature in investment and by allowing leisure to be a durable good.
This paper presents maximum likelihood estimates and tests of a model that is similar to Kydland and Prescott’s, using postwar U. S. data for per capita hours and the growth rates of per capita output, per capita expenditures for the acquisitions of durable consumption goods and for the stocks of aggregate structures and equipment, respectively. It derives testable restrictions from the assumption that such time series are generated from the competitive equilibrium of an economy with the postulated preferences, technology, and stochastic environment. For this purpose, laws of motion describing the evolution of equilibrium quantities are calculated as functions of the underlying parameters of the model and the innovation to the technology shock.
* Manuscript received September 1986; revised October 1988.
1 I received helpful comments and suggestions from Lars Peter Hansen, Robert Litterman, Kenneth Singleton, and an anonymous referee. In addition, I thank Thomas J. Sargent for his encouragement at various times. An earlier version of this paper entitled “Gestation Lags and the Business Cycle: An Empirical Analysis, ” 1983 was presented at seminars at Carnegie-Mellon University; Massachusetts Institute of Technology; Northwestern University; and the Universities of California (Los Angeles), Chicago, Minnesota, Rochester, and Southern California; as well as at the 1984 National Bureau of Economic Research Conference on Business Fluctuations and the 1984 North American Summer Meetings of the Econometric Society. A major part of the calculations for the current draft were done on the Cray 2 computer, on a grant from the Minnesota Supercomputer Institute.