Per Bak, Kan Chen, José Scheinkman, Michael Woodford

Paper #: 93-01-004

Explaining the observed instability of economic aggregates is a long-standing puzzle for economic theory. A number of possible reasons for variation in the pace of production are easily given, such as stochastic variation in the timing of households’ desired consumption of produced goods, or stochastic variation in the costs of production. But is it hard to see why there should be large variations in those factors that are synchronized across the entire economy--why most households should want to consume less at exactly the same time, or why most firms should find it an especially opportune moment to produce at the same time. Instead, it seems more likely to suppose that variations in demand or in production costs in different parts of the economy should be largely independent. Thus, one might ask, should one not expect these local variations to cancel out, for the most part, in their effects on the aggregate economy, due to the law of large numbers? Fluctuations in activity of macroeconomic significance, it might be thought, should occur only when many independent shocks happen by coincidence to have the same sign, and this should be an extremely unlikely event (with the probability of occurrence decreasing exponentially with the square of the size of the event, by the central limit theorem).