Monday, August 15, 2011

Commitment, the State of the World, and Dissent

Here are some more thoughts with regard to my previous three posts (this one, this one, and this one.)

As academics, our policy arguments take place in the following way. We agree on what the state of the world is, but we may disagree on what model we should use to guide policy. Typically our arguments are in terms of the merits of the models at hand. Which model is better? Once we agree on that, then we can determine an operating rule for policy that tells the policymaker what action to take given any possible state of the world.

Policymaking on the FOMC seems not to work like that. Some of the non-economists on the committee are not aware of all the models available and how they work. The trained economists may have very different views concerning the merits of the models at hand. Over time, what seems to make the decision-making process work is to have the argument over the state of the world rather than over the theory. Pre-financial crisis, the question would be framed as follows. The FOMC members took as given that they did the right thing at the last FOMC meeting. The question for the current meeting was whether the state of the world was worse, better, or about the same relative to the last meeting. If things were worse, they would move the target for the fed funds rate down 1/4 point; if things were better, they would move up 1/4 point; if things were about the same they wouldn't do anything. Easy.

Post-financial crisis, things are not quite so easy, as now we supposedly have some more tools (I say supposedly because this is Bernanke's claim; I actually don't think so). Primarily, quantitative easing is on the table, and so is "extended period" language. But the policy decision is still framed in the same way. Is the state better or worse? Based on the answer to that question, the FOMC either tightens or eases. Then the question is how to do the tightening or easing.

Now, go to Bernanke's explanation for why the FOMC voted for the QE2 program in November 2010, i.e. this piece in the Washington Post. He discusses the dual mandate and some broader issues associated with the state of the world. Here is the important part:
The FOMC decided this week that, with unemployment high and inflation very low, further support to the economy is needed.
What Bernanke is telling you is that the FOMC has a decision rule, and for purposes of communicating with you, that decision rule takes account of inflation and unemployment. Unemployment is bad, and more inflation is good in this case. That's not the unemployment or inflation that he or someone else is forecasting, that's the unemployment and inflation we are actually observing. Then, he is telling you that the state of the world dictates that the FOMC should ease, following which he lays out the QE2 plan:
With short-term interest rates already about as low as they can go, the FOMC agreed to deliver that support by purchasing additional longer-term securities, as it did in 2008 and 2009. The FOMC intends to buy an additional $600 billion of longer-term Treasury securities by mid-2011 and will continue to reinvest repayments of principal on its holdings of securities, as it has been doing since August.

So, it's possible that the FOMC might now think that QE2 did not work, in which case maybe they want to try something else. However, Bernanke seems to think it worked, at least that's what he said in this speech in February. As of the end of the QE2 program, Jim Bullard certainly thought the program worked, as did other Fed officials.

Thus, in terms of Bernanke's own publicly-stated criteria for what directs easing and tightening, Kocherlakota tells us the state is worse. In November 2010, the FOMC agreed that the state of the world directed them to ease. They eased, and according to them the easing worked. The state of the world is now better, so why ease further?

I think macroeconomists agree broadly on issues of commitment. The decision rules of policymakers need to be simple enough for the public to understand, and policymakers need to behave in ways that are consistent with their publicly-stated policy rules. If there is a change in a decision rule, that needs to be clearly-communicated.

In the case of last week's FOMC decision, there was certainly no statement that the FOMC was thinking about the world in a different way. Thus, presumably the decision rule has not changed. If that is true, than the FOMC's decisions should be consistent with what it has been doing. But Kocherlakota argues that is not the case, and I think he is right. A decision that may appear to make the Fed's behavior more predictable actually makes it less so. Instead of asking why the dissenters on the FOMC voted the way they did, we should be asking why the other people on the committee voted the way they did. And we should not have to ask that. We are now more confused, and that is not good.


  1. Steve,
    if the public really believes that in a better state
    than last meeting the fed raises rates and in
    a worse state it lowers than they might as well
    believe that in a desperate state the fed announces
    something more aggressive, discounts bernanke appropriately
    and assumes that in 2012:2 if the unemployment rate
    went down and inflation up we return to the old game discarding announcements in the desperate state
    asked differently what model do you have in mind
    that describes the believes on what the public opinion has
    in mind on the behavior of the fed at the lower bound
    to my knowdledge the taylor rule cant be it without adjustment

  2. If I understand you (I'm just looking at the last bit you wrote), I agree that it's not appropriate to be discussing the Taylor rule at the lower bound. Some people in the Fed system do that, unfortunately. To Bernanke's credit, he seems to recognize that it's going to be difficult to communicate what the FOMC is up to in the current context, and takes pains to get the message across that QE2 and such are just business as usual. There's a sense in which it is in fact business as usual, but I don't think in the way he is communicating it.

  3. Isn't the truth that QE2 didn't work as well as they hoped so they want more easing, but that if they say QE2 didn't work then people will oppose more of the same? This would explain why the public statements and the actions don't quite match up.

  4. That could be, but I think they are sticking to their guns, for the most part. There may have been a few public doubts expressed about QE2 by Fed officials - Plosser, for example, I think, but you would have to look through his speeches. However, the overwhelming message coming from the Fed system is that QE2 worked, with empirical evidence to back it up. Bullard's slides that I linked to above were from a St. Louis Fed conference. The thrust of the empirical work at that conference - from Jim Hamilton and others - was that QE2 worked. The problem is that the theory is either weak or non-existent, and there is therefore no structural empirical work on the matter.

  5. Stephen, I agree with your post. I think the FOMC said: we had some bad data revisions and a couple of bad days on the stock market, so let's put aside the inflation and unemployment numbers, let's put aside Taylor rules, let's put aside what we did in the past and how we explained our previous actions, and let's show the public that we care by doing something. Keeping rates at zero till 2013 is something, so let's do that.

  6. Seems to me that there is some experimentation going on. Bernanke et al's conviction could easily be that 1) QE2 worked but 2) It didn't work as well as they hoped so 3) If you double down you could get there.

    Where things gets muddied is with communication and politics. For political purposes they may want to de-emphasize 2) and emphasize 3). This is relatively easy as a communication strategy because the data has been so bad and they don't need to go out there and say that they are doing a lot of experimentation.


  7. Yes, you could say this is a response to financial instability. You have to do something, so let's try this. On the other hand, you could interpret it as responding to panic with panic.

  8. Wouldn't you agree that growth prospects are much lower now than they were at the beginning of this year? Many were predicting 4% growth this year (after all the slump we had in the recent past), but now forecasts seem to be about half that. Unemployment as well is only very marginally better than it was at the beginning of the year.

    As Hamilton and many others have shown in their research, QE did have an impact on long term rates, but that impact was not big. And, besides that, other things happened in between. Growth prospects in Europe are lower (even Germany is now trending down), the high energy prices etc. To say that QE worked, but economic conditions haven't improved a lot, is not necessarily contradictory, right?


  9. "To say that QE worked, but economic conditions haven't improved a lot, is not necessarily contradictory, right?"

    Right. But to evaluate whether QE worked, we need a structural model, otherwise we can't evaluate the effectiveness of the policy. Hamilton and other researchers who have attempted to determine the effects of QE are serious economists, but the problem is a difficult one. I have yet to see a theory that meets "industry standards" and also tells us that QE works the way Bernanke claims it does.