In the latest issue of Capitalism and Society. I wrote one of the papers, and then Peter Howitt wrote a comment. I recommend both of them, although personally I find Howitt’s paper more interesting, because I already knew what I was saying. I will use the rest of this post (below the fold) to comment on Howitt’s paper.Howitt argues that we ought to think of unemployment as a coordination failure. He points out that even Keynes can be read that way. That is, savers want more future output, but entrepreneurs are not sure which type of output they want, and the result is a coordination failure.

Howitt argues that this coordination failure paradigm focuses attention on organizations that provide coordination services. Call these coordinators. Let me give two examples of entities that were coordinators. Borders Books brought together suppliers of books with readers of books. AIG Financial Products brought together people who wanted to earn a low-risk return on saving with companies that needed credit default swaps on mortgage-backed securities.

As you know, both Borders and AIG FP are no longer with us. We could look at this in three ways.

1. Their demise is an autonomous, causal factor in economic fluctuations.
2. Their demise is a symptom of lower aggregate demand.
3. Their demise is a symptom of economic displacement, as their patterns of specialization and trade are no longer sustainable. In the case of Borders, the problem is technology. In the case of AIG FP, it was the combination of mortgage lending to unqualified borrowers and falling house prices that changed the value of the options that they had written.

It strikes me (and this may be uncharitable) that Howitt is leaning toward (1) as a way to describe coordination failure. You can think of the Bernanke model (this is Bernanke the academic, not Bernanke the Fed chairman, although the first clearly influenced the second) of bank failures destroying the “relationship capital” that allows banks to intermediate between savers and users of savings. So, in Howitt’s model, one can get a wave of failures among coordinators, and this means that there will be less coordination in the economy, so that there is more unemployment.

If the failures of coordinators are autonomous, then one does not need a PSST story. I think that is why Howitt believes that PSST does little to advance the coordination story. Instead, I think it does, although I am aware of its many shortcomings.