When eminent economists agree to disagree

One thing the internet has achieved is to open up to the public the discussions between academic experts that previously took place mostly in face-to-face gatherings or scholarly journal articles. Neither medium was particularly accessible to the bulk of the interested audience.

This is particularly true in the economic blogosphere, where research results and policy thinking are communicated, often eloquently and accessibly, by many of the fields’ greatest minds to the costless benefit of anyone who cares to follow.

A recent exchange between Larry Summers, Brad DeLong and Paul Krugman is a case in point. Before Christmas, Summers posted a characteristically succinct statement on why he disagreed with the Federal Reserve’s decision to begin tightening monetary policy. His analysis is well worth reading in full, but the trigger of the ensuing debate was his explanation for why the Fed thinks differently: “I suspect it is because of an excessive commitment to existing models and modes of thought. Usually it takes disaster to shatter orthodoxy.”

DeLong and Krugman both responded on their blogs, in an exchange that is fascinating in that all three agree that the Fed should not have raised rates, but disagree on what its intellectual error was. That disagreement throws light on some deep questions about how economic theorising is and should be used to inform policymaking.

DeLong expressed doubts that the Fed’s analysis was indeed compatible with existing models; Krugman asserted that conventional models straightforwardly showed the Fed to be in the wrong, and that the central bank’s policy was driven not by a commitment to existing models but by “a conviction that you and your colleagues know more than is in the textbooks”.

DeLong then added an admission that conventional textbook analysis was indeed sufficient to prove the Fed wrong, provided one at least adds the newer modelling of why increased fiscal stimulus does not increase public debt burdens in today’s monetary conditions — a contribution that was made by DeLong himself and Summers in 2012.

Summers then responded with a comment and a postscript which showed a fascinating divergence between him and two other great economists, all three of whom agree on the right policy conclusion. Whereas DeLong and Krugman think the Fed erred by ignoring the models they cite, Summers thinks the Fed erred by ignoring things that such models do not capture.

In this sense, he is someone much more in awe of the importance of policymakers’ ignorance than are his interlocutors. Note that it is informed ignorance we are talking about here: what we should admit we do not know beyond the understanding the models we have, rather than ignorance of the models themselves.

Summers is also is much more comfortable with the notion that policymakers should aim to underpin market confidence.

That notion has often been derided by Krugman in the past six years of debate on fiscal austerity, especially in the US and the UK, on the grounds that in models of economies with their own floating currency with interest rates near zero, no market panic should follow from large deficits, and if one does, it will not affect the government’s borrowing costs but only the currency.

So a panic will at most cause a boost to exports and therefore be positive. That was the key claim of Krugman’s Mundell-Fleming lecture at the IMF in 2013, a very important paper whose “simple model-based thinking” Summers thinks is “dangerously wrong”.

Two quotes rather nicely capture the methodological disagreement here. Summers writes: “I think maintaining confidence is an important part of the art of policy ... Paul is certainly correct in his model but I doubt that he is in fact.”

DeLong responds: “Larry, however, says: We know things that are not in the model. Those things make Paul’s claim wrong. My problem with Larry is that I am not sure what those things are.”

Krugman himself says that his point was that “simple models don’t seem to have room for the confidence crises policymakers fear”. That, as Summers points out, is no longer true: Olivier Blanchard has written down a model in which the capital outflows from a crisis of confidence make an economy contract rather than expand through an export boom.

But even without that formal response, which proves that what policymakers fear in reality can indeed happen in a model, Krugman’s lecture acknowledged that a sovereign debt crisis could trigger a banking crisis: “Private-sector foreign-currency debt can effectively remove monetary autonomy and leave nations vulnerable to severe damage from sudden stops.

“This is an important caveat to my earlier analysis. It is not, however, relevant to countries that, even though they may have substantial external debt, borrow in their own currencies — that is, to the US and the UK.”

The US, perhaps, but the UK? UK banks have huge non-sterling liabilities (they have similarly large non-sterling assets, but these being banks, they presumably borrow short-term and lend long-term). On Krugman’s own argument, then, those worried about confidence seem to have a case when it comes to the UK.

But Summers’ broader point is that a crisis of confidence just in the sovereign can depress demand domestically, even if only through “animal spirits” and with no good reason. What is a policymaker to do if she thinks this is the case in reality, even if no extant model captures it?

Surely not wait for 30 years in the hope that new mathematical techniques enable economists to model that reality. In his willingness to listen to those who may have an untheoretical “feel” for the market, and in his intellectual respect for the limits of his own knowledge, Summers comes across as the most Keynesian of these three.

(Surprising, perhaps, for someone who left derivatives unregulated on the grounds, as models predicted, that this would be safe and beneficial — but then Keynes himself is often quoted in support of changing one’s mind when one is wrong.)

— Financial Times