Scott Sumner had a post a while back on “why no short recessions” and I can’t remember whether I responded to it or not.

However, a point that I want to keep making is that from the top it looks like recessions really are different. Or, to put it in terms I prefer: a recession is a general phenomenon, not a set of phenomena in general.

Here are both Unemployment and Excess Industrial Capacity in the US economy.

FRED Graph

First, especially in the unemployment rate, we can play “some of these periods are not like the others.” Specifically the time horizon over which unemployment rises is extremely short and the increases are really fast. The time horizon over which unemployment is declining are long and the declines typically fairly slow.

Second, for the most part these two phenomena move together. That is to say there are workers with out plants and the same time there are plants without workers, and that’s not a coincidence.

A recession doesn’t seem to be when capital and labor both happen to be underutilized but when something happens that underutilizes both capital and labor.

The one of the few really weird periods is the late 90s when excess capacity started to rise but unemployment continued to fall. That really was about a big structural change in the economy where the US was losing manufacturing jobs at the fastest pace on record but was gaining so many tech and service jobs that unemployment continued to fall.