Hystersis is something that concerns me deeply. When we look at the relationship between unemployment and inflation its hard not come to the conclusion that past monetary episodes influence the current relationship between the two.

For me the 80s has always been a particarly important period because we have central banks officially adopting a set of policies designed to bring down inflation. Those policies ranged from slam-bam Volkerism to explicit inflation targets.

Its hard not to believe that this was money pushing on the economy and not real effects masquerading as nominal ones.

In any case, I wanted to look at the series Krugman uses to produce evidence of hysteresis because it doesn’t look consistent with the data to me.

Here is what Krugman offers

You can see that there was a mini-version of the current decline in manufacturing capacity after the 2001 recession: capacity basically stopped growing in the face of a protracted weak economy. But this time around, with manufacturers operating way below capacity with little prospect of needing more capacity any time soon, they’re both scrapping equipment and failing to expand. The result is that when we finally do have a real recovery, we’ll run up against capacity constraints much sooner than we would have if there had been no Lesser Depression.

The story here is that capacity is shrinking because machines are wearing and not being replaced. Well, we have data on investment in industrial equipment.


What we see either in the 2000s or in this period doesn’t look out of the norm overall, just more volatile this time around. For example, we have the biggest percentage collapse in investment followed by the biggest percentage surge.

It’s a volatile time but not one in which investment is completely out to lunch.

I am not sure what Krugman is picking up but given the explosion in the 90s and the drop recently I think it is likely related to the real terms of trade. Capacity went up because manufacturers benefits from changes in the terms of trade. Those effects peaked in the early part of the recession.

Here for example is my extremely rough attempt to compare capacity in US manufacturing to the real cost of Chinese imports.

FRED Graph

I am using the Chinese consumer inflation rate here when I should be using producer prices but its what I got. And, it suggests to me that there could be something there.