Menzie Chinn makes the point that there is no sharp distinction between cyclical unemployment and structural. That is, the longer a recession goes on the more and more people will become semi-permanently unemployed. Ryan Avent co-signs.
I think the data on this phenomenon, known in economics circles as hysteresis, is compelling. Though, we don’t know exactly why it occurs. Stories about lost skills and disaffected workers are easy to come by. Hard data is not.
However, hysteresis should be properly interpreted as saying “Monetary policy has long lasting real effects” not that recessions are purely the result of real shocks or mismatch.
Hysteresis increases the urgency for higher inflation targets and indeed encourages us to pursue higher permanent inflation. If we believe hysteresis is true then that suggests that there is a semi-permanent tradeoff between inflation and unemployment.
This is the exact opposite of the matching / micro-imbalance story. I have not emphasized hysteresis because I think it is controversial and I don’t think we need to focus on it to come to an agreement that the Fed needs more aggressive monetary policy.
However, the take away is this: to the extent hysteresis is true this episode of disinflation and recession has dragged up the NAIRU, or natural level of unemployment. Yet, the same theory suggests that a period of rising inflation levels could drag down the NAIRU.
Hysteresis is not a reason not to act but a further imperative to action.
As far as I am aware the only reasons not to act are that either
1) We think 3 – 4% inflation is worse than 10% unemployment.
2) We believe that there are purely real effects at work and monetary policy cannot affect those.
The first is, of course, a value judgment though I consider it wildly off base. Still, I am willing to give more time to discussing the long run consequences of a little inflation.
The second is an empirical question with a right answer and a wrong one. I suggest that the right answer is that monetary policy can affect our situation and I will continue to post that evidence here.