By GARY D. HANSEN AND EDWARD C. PRESCOTT *
Real-business-cycle theory has been used to determine the statistical properties of aggregate fluctuations induced by technology shocks. The finding is that technology shocks have been an important contributor to fluctuations in the U. S. economy. For example, Finn E. Kydland and Prescott (1991) estimate that if the only impulses were technology shocks, the U. S. economy would have been 70 percent as volatile as it has been over the postwar period. In this paper we employ the theory to answer the question “Did technology shocks cause the 1990-1991 recession?”
Answering this question requires the determination of the effects of technology shocks on the path of the economy. The procedure that we use to make this determination is as follows. First, we construct a model economy and calibrate it so that its steady-state matches actual 1987 values. Data for other years are used to determine the realizations of the shocks and the nature of the stochastic processes generating these shocks. Given these stochastic processes, the equilibrium decision rules for our calibrated economy are computed. We then construct the path for the economy implied by the model for the period 1984:1-1992:3. We set 1984:1 model values of the state variables equal to actual 1984:1 values and use the decision rules and the realized shocks to construct this path. Finally, we examine whether the model economy experiences a recession in 1991 as did the U. S. economy.
Business cycles are variations in output per adult that are in large part accounted for by variation in the per-adult time allocated to market production. Figure 1 plots the time path of the per-adult labor input. During the 1983:1-1989:1 period there was a remarkable 12-percent increase in the labor input per adult. Beginning in 1990:2 there has been a decline of nearly 6 percent in this per-adult labor input. An unusual feature of the recovery subsequent to the 1990-1991 recession is that this labor-input variable has continued to decline well into the recovery. This unusual behavior leads to a related question, “Did technology shocks cause the slow recovery?”
We found it necessary to modify the standard real-business-cycle model in four ways in order to answer the posed question. First, given the enormous changes in the relative prices of durables as depicted in Figure 2, we could not treat technology change as being neutral with respect to different types of final goods. Following John B. Long and Charles Plosser (1983) and Jeremy Greenwood et al. (1992) we consider multiple production sectors with technology change differing across sectors. Second, we assume that there is a technology employed within the household sector that produces a consumption flow from the stock of consumer durables. Third, we consider land to be a factor of production in addition to labor and capital. Our fourth modification is to introduce population growth.
By definition, technology shocks are changes in the production functions or, more generally, the production possibility sets of the profit centers. In a growing economy we observe positive technology change over
* Hansen: University of California-Los Angeles and Department of Economics, University of Pennsylvania, Philadelphia, PA 19104-6297; Prescott: Research Department, Federal Reserve Bank of Minneapolis, Minneapolis, MN 55401, and University of Minnesota. We thank Fernando Alvarez and Terry Fitzgerald for their more than able assistance and the NSF for financial support. The views expressed herein are those of the authors and not necessarily those of the Federal Reserve Bank of Minneapolis or the Federal Reserve System.