Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
The authors study the movements in output, consumption, and hours that are forecastable from a vector autoregression and analyze how they differ from those predicted by standard real-business-cycle models. They show that actual forecastable movements in output have a variance about one hundred times larger than those predicted by the model. The authors also find that forecastable changes in the three series are strongly positively correlated with each other. On the other hand, for parameters whose implications are plausible in other respects, the model implies that output, consumption, and hours should not all be expected to move in the same direction. Copyright 1996 by American Economic Association.