Sunday, February 16, 2014

Email Conversation with Binyamin Appelbaum

In connection with this post, and this New York Times article by Binyamin Appelbaum, a question came up in the comments section from Noah Smith, as to how Prescott's comments may have been taken out of context. Prescott and I were both, I think, interviewed in the same way for Appelbaum's piece, i.e. by email. Here's how I was quoted in Appelbaum's article:
“Suffice to say that I am very surprised by his current policy views and how he articulates them,” Stephen Williamson, a professor of economics at Washington University in Saint Louis, said in an email. “I don’t think his views have changed. I think this is just a side of Narayana we didn’t know was there.”
It's useful to compare that to my email conversation with Appelbaum. I left out one question, which I thought was too personal, and I did not answer the question. Here's the rest of it:
BA: More than four years later, how would you assess his performance as president of the Minneapolis Fed?

SW: I'm guessing that if you asked the participants at FOMC meetings what they thought, they would be very enthusiastic.

BA: Narayana has argued that stronger forward guidance could help to increase employment, and that the Fed has room to try because inflation remains low. Why do you regard this view as mistaken?

SW: I think that the Fed's experiments with forward guidance have been unsuccessful. They have only succeeded in confusing people. A useful exercise is to count the number of words in the FOMC statement, and compare pre-financial crisis to post-financial crisis. I can parse that and understand what the FOMC is trying to say, but I think the average financial market participant has trouble with it, and the average person on the street would be completely lost. Further, the forward guidance has not been consistent over time. The FOMC first specified an "extended period" of low interest rates, then calendar dates, then a threshold specified in terms of the unemployment rate. Now they are back to extended period language - roughly - along with words about what happens if the inflation rate is low for a long period of time. It's hard to see why more language in there - for example about the unemployment rate threshold - could make any difference.

Further, I think that the reason inflation is low now is because short-term interest rates have been low for a long time. The longer they stay low, the longer inflation will remain low. That sounds counter-intuitive, but I think it's right. One of the most astute things that Kocherlakota said in public is in this speech:

That essentially makes the same point. The idea is that, the way the Fed thinks about the problem they're facing traps them in a low-inflation environment.

BA: You wrote in September, "Fed officials like Kocherlakota seem to want to argue that the failure of policy to 'cure the problem' is just a license for doing more." Kocherlakota, I think, would say that the Fed has a legal obligation to keep trying so long as inflation remains low. What is the downside to trying?

SW: The "legal obligation" i.e. the "dual mandate" is quite vague. For example, I think if you look at the FOMC statements during the Greenspan era (though granted that was fairly tranquil) you'll find little or no mention of anything to do with the "real" economy - e.g. the unemployment rate - in the policy statement. Greenspan was deliberately vague, and Congress never gave him a hard time about it. The danger of speaking too much to the real part of the mandate (i.e. "maximum employment") is that the Fed's ability to influence real activity is subject to considerable uncertainty, and no one thinks they can have an effect on real activity for long. If Fed officials continually make statements to the effect that they can do things which they cannot do, that is a problem.

BA: You wrote in 2010, "The Narayana I knew would have thought the worldview represented in standard Keynesian economics was hopelessly naive." How would you describe Kocherlakota's views on monetary policy before he joined the Minneapolis Fed? Are those views apparent in his work as an academic?

SW: As an academic, Narayana was a very sharp researcher, interested in a wide array of issues, few of which actually touched on monetary policy. So maybe I was just making inferences about what his views on policy would be, and how he might go about applying his knowledge in a policy setting. Suffice to say that I am very surprised by his current policy views and how he articulates them.

BA: Do you think his views have changed at the Minneapolis Fed, or that they were previously misunderstood?

SW: Actually, I don't think his views have changed. I think this is just a side of Narayana we didn't know was there.

BA: How do you think the Minneapolis Fed has performed as a research institution under Kocherlakota? Do you see evidence of a shift in the subjects under study, or in its perspective on those subjects? What is your understanding of the reasons for the termination of Pat Kehoe and the separation with Ellen McGrattan?

SW: To date, I don't think that the way Kocherlakota thinks about policy has mattered for how research is done at the Minneapolis Fed. Whatever is going on internally actually has nothing to do with economic ideas, as far as I can tell.


  1. With regards to Q3, I think your answer points to a common misconception (which is inherent in the question): Not taking an action is also an action. You could simply be waiting for more and/or better information, and there might well be downsides to acting prematurely.

  2. QE can bring down long rates, take risk out of private and onto a public balance sheet and it can raise inflation expectations (but as far as I see it it has not).
    The only disadvantage I see is the typical one of any form of expansionary monetary policy: inflation. The long-term inflationary potential of a massively blown up balance sheet of the Fed is hard to judge.

    So I'd say we can judge QE fairly well on a qualitative basis. Just because we don't have clear quantitative results yet doesn't mean that the Fed should not have engaged in QE.

    1. "QE can bring down long rates, take risk out of private and onto a public balance sheet and it can raise inflation expectations..."

      How do you know this?

    2. Epic fail. I wrote CAN reduce long rates and CAN raise inflation expectations.
      The only thing I actually know (because it is mere accounting) is the "temporary risk socialization". Iif Ricardian equivalence would hold this would obviously have no effect but the world is not Ricardianas ample of consumers are liquidity constrained and as credit markets are far from perfect. So while the CB has to sell the assets it bought one day in the meantime private players are happy that some risk has been taken off their balance sheets.

      As I already wrote, how significant these effects are is unknown. At least I am not aware of any papers which tried to analyze QE quantitatively, if somebody knows some I am interested.

    3. Imperfect credit markets is not sufficient to give the result. It's not clear how central bank offloads maturity risk from the private sector. In the process, the central bank has to take on more maturity risk. Which means that, in the event that short rates go up, the central bank turns over less revenue to the fiscal authority, which has to make it up somehow - i.e. from taxpayers, who are the people who supposedly are bearing less risk. I have never seen a model of how this works.

  3. This is an important point about the interplay between QE and fiscal policy. But the world is not Ricardian, consumers do not fully internalize the public budget and QE should have some effect upon demand via helping private players to deleverage.

    If the world were Ricardian public demand management would not be possible. But the world is not Ricardian so the public sector is the only one who can sail against the wind.

    I share your scepticism though, just because there should be positive effects in theory does not mean that there are in practice. Yet suggesting doing nothing because of this is wrong. Inflation is low and budget issues do not exist while interest rates on government debt are low. Unemployment on the other hand is still high.

    1. All that a non-Ricardian economy guarantees is that the debt matters. But we need more to make the case that the maturity structure of the debt matters, and more still to make the case that, under current circumstances, changes in the composition of the debt brought about by the central bank matter.

    2. I doubt that the buying of government debt had much effect in the US. But the purchase of all those crappy asset-backed securities has a clear effect. When risky private assets are substituted by safe assets the risk-bearing capacity of the respective private players is increased. In case of a bank this can increase lending, in case of a non-financial firm this can increase investment and in case of a household it can increase consumption.

      It is clear that this does not suffice (why should firms invest when demand is low so why should banks make new loans) which is why the obvious solution was deficit spending.
      No need to respong by the way, I know very well that you will ignore basic economics and deny that expansionary fiscal policy is expansionary.

    3. Go away John D.