Endogenous Business Cycles and the Dynamics of Output, Hours, and Consumption

This paper studies the business-cycle fluctuations predicted by a two-sector endogenousbusiness-cycle model with sector-specific external increasing returns to scale. It focuses on aspects of actual fluctuations that have been identified both as defining features of the business cycle and as ones that standard real-business-cycle models cannot explain: the autocorrelation function of output growth, the impulse response function of output to demand shocks, and the comovement of forecastable changes in output, hours, and consumption. For empirically realistic calibrations of the degree of returns to scale, the results suggest that endogenous fluctuations do not provide the dynamic element that is missing in existing real-business-cycle models. (JEL E32) The hypothesis that actual business-cycle fluctuations are—at least in part—endogenous has a long history in economics. But only recently were endogenous-business-cycle models developed whose empirical validity can be evaluated using methods that have been standard in modern business-cycle research since the pioneering contribution of Finn E. Kydland and Edward C. Prescott (1982). In this class of endogenous-business-cycle models, the possibility of aggregate fluctuations typically arises as a consequence of increasing returns (e.g., Roger E. A. Farmer and Jang-Ting Guo, 1994) or variable markups of prices over marginal costs (e.g., Jordi Gaĺı, 1994 and Stephanie Schmitt-Grohé, 1997). However, the models just cited have been criticized as empirically implausible because endogenous fluctuations arise only for increasing returns or markups ∗Department of Economics, Rutgers University, 75 Hamilton Street, New Brunswick NJ 08901, USA. I would like to thank Mart́ın Uribe, three anonymous referees, Jess Benhabib, Michael Woodford, and seminar participants at the Board of Governors, Duke University, the Federal Reserve Bank of New York, the Federal Reserve Bank of Richmond, Georgetown University, the 1998 NBER Summer Institute, Princeton University, Rutgers University, and the University of Wisconsin-Madison for helpful comments.