Wednesday, August 25, 2010

How to Get Worked Up Over Nothing

Have at look at this:

1. Krugman.
2. Rowe
3.Harless
4. Thoma

Is there something in the water, or did these guys sit through some funny macro classes? What they are objecting to in Kocherlakota's speech is one of the most innocuous things he said. Here's the simplest example I know. Suppose a cash-in-advance model with a representative consumer, period utility u(c), discount factor b, constant aggregate endowment y. c is consumption. The consumer needs cash to buy c each period. Suppose y is a fixed quantity of output received by a firm, which is sold for cash within the period, and then the cash is paid as a dividend to the consumer at the end of the period. Have the money stock grow at a constant rate m. The real interest rate is constant at 1/b -1. The nominal interest rate is (1+m)/b - 1, and the inflation rate is m. Constant m implies a constant nominal interest rate and a constant inflation rate. If m < 0, there is deflation, and the nominal interest rate is sufficiently low to support the deflation. I can think of the instrument the central bank sets as either the money growth rate or the nominal interest rate - that part is irrelevant. This type of result holds in virtually all monetary models, though of course sometimes the real rate may depend on the inflation rate. That's not a big deal though. What's the problem?

44 comments:

  1. Could it be the case that they have a different interpretation of the term "long run"? Maybe they are not thinking about a non-stochastic steady state equilibrium? I don't know.

    ReplyDelete
  2. Steve: "I can think of the instrument the central bank sets as either the money growth rate or the nominal interest rate - that part is irrelevant."

    No. That is the one crucial part I disagree with. What we take as exogenous/endogenous determines the central bank's out-of-equilibrium (off-the-Nash-equilibrium-path) play, and people's expectations of that play, and that in turn will either create or not create a process that leads to that equilibrium.

    Must go to bed. More later.

    I am so glad we are getting to the crux of this issue. This is related to a point I have been hammering away at for so long. Mostly against people who cannot think of monetary policy except as setting nominal interest rates. Now I get to argue with the opposite side!

    ReplyDelete
  3. I think you're assuming the consequent when you say "Have the money stock grow at a constant rate m." The four people you've cited are arguing that if the Fed takes Kocherlakota's advice, and shifts the fed funds rate up before inflation increases, the money stock _will not_ grow at a constant rate: instead it will grow at a decreasing rate or begin to shrink.

    In order for you and Kocherlakota to be right, you'd have to demonstrate that inflation must be constant.

    Now, as Nick says, if the Fed set the money growth rate directly, we wouldn't have any of these problems, but for now they set the nominal interest rate instead, which leaves the money growth rate free to change over time.

    ReplyDelete
  4. @ Anonymous: It's to do with path dependence and knife edge equilibrium. Williamson is describing an equilibrium, but it is an knife-edge equilibrium. If central bank policy is a simple fixed nominal interest rate rule, there is an equilibrium at i = r + inflation, but any perturbation of inflation upwards will cause inflation to follow an explosive upwards path. In the case of an economy in a recessionary liquidity trap with nominal interest rates set near zero, a recovery that causes positive inflation will almost certainly start an inflationary spiral *if* the central bank holds interest rates fixed near zero indefinitely.

    ReplyDelete
  5. I, Jeffrey Yasskin, and lockhoo, are on the same page. We are all making the same point, in different ways.

    Take a standard supply and demand diagram. Start in equilibrium. If the demand curve for apples shifts right, we know the equilibrium price of apples increases. But is there any reason to believe we will get to that new equilibrium price? Yes. At the original price of apples, there is now excess demand for apples when the demand curve shifts. And that excess demand will cause individual sellers to want to raise apple prices above the existing price, as a profit-maximising response.

    Suppose the Fed sets M. Start in equilibrium. Now the Fed doubles M. We know that equilibrium P doubles. But is there any reason to believe we will get to that new equilibrium P? Yes. At the existing price vector, and given existing expectations of future prices, there is now an excess supply of M. People try to get rid of the extra M by spending it, or lending it. This creates an excess demand for apples and bananas. So sellers of apples and bananas raise prices.

    Suppose the Fed sets the nominal interest rate i. Start in equilibrium. Now the Fed increases i. We know the new equilibrium has higher inflation. But is there any reason to believe we will get to that new equilibrium? No. At the existing price vector, and given existing expectations of future prices, the real interest rate is now higher, so demand for apples and bananas falls, so there is excess supply of apples and bananas. So sellers cut prices. We get deflation, and never get to the new equilibrium.

    Using interest rate control to target inflation is like balancing a long pole upright on the palm of your hand. It's unstable. Your hand is the nominal interest rate. The top of the pole is inflation. If you want the top of the pole to move right, you must first move your hand left. Then when the pole is leaning right, and the top moves right, you can move your hand right too.

    If prices and expectations are slow to adjust, it's like a very long pole. If they are quick to adjust, it's like a short pole. With a short pole, your hand and the top of the pole are always closer together on the left/right dimension. But it's harder to keep the pole upright. In the limit, with perfectly flexible prices and expectations, the pole is very short, has zero inertia, and it's impossible to keep the pole upright, because you can't move your hand quickly enough. You can't have the Fed set nominal interest rates with perfectly flexible prices and expectations. The monetary system explodes into inflation or implodes into deflation immediately.

    Monetary control is the opposite. It's like letting the pole hang down. It's stable.

    ReplyDelete
  6. And right now, continuing with the pole analogy, there is a wall on the Fed's left side, and the Fed's hand is up against the wall. The Fed wants the top of the pole to move right, but it can't move its hand any further left. And the pole is leaning further and further left.

    The solution is to grab the top of the pole, and move it right. Abandon the unstable control variable of nominal interest rates. Switch to something like M targeting, or targeting nominal asset prices. Just grab hold of anything that moves in the same direction as the top of the pole in the short run and in the long run. And announce in advance where you are planning to move that thing.

    At one level Steve, you are right. Interest rate targeting is not what the Fed does, it's how the Fed frames (communicates) what it does. But framing/communication are important, because they affect what can be communicated about out-of-Nash-equilibrium-off-path play by the Fed. Interest rate targeting is a social construction of reality. It's not really real. But it's an unfortunate social construction of reality, because it creates the zero bound problem. It means that the Fed cannot communicate that it is easing monetary policy, by increasing the growth rate of some nominal variable (with $ in the units) like M or the S&P. If it could communicate this, the Fed could loosen monetary policy, and watch nominal interest rates rise at the same time, as a *consequence* of its commitment to easing.

    At one level, this is all in Wicksell. At another, deeper level, it ain't.

    ReplyDelete
  7. Let's find something to agree on here. Central banking is intermediation. It is always about controlling some quantities - the quantities on the central bank's balance sheet. Some people like to argue as if the central bank operates by "setting" an overnight market nominal interest rate, which of course it can't in general. The only nominal interest rates a central bank can "set" are the interest rates on its deposits (reserves) and its loans. It turns out, however, for reasons having to do with short run fluctuations in the demand for outside money, that pegging a nominal interest rate over a short horizon seems to work well. Further, Friedman's prescription of constant money growth as a policy rule, over any horizon, has been a practical failure, as the "demand for money" is not stable over any horizon. What are you left with? Some monetary quantity is always what really matters, but it's impossible to figure out what that quantity is at any point in time. Central bank decisions therefore typically involve questions of how you move the nominal interest rate target in response to what you see. Currently, in the US, things are much more interesting than that, with questions involving the maturity structure of the central bank's assets and purchases of private assets. It's not like the standard operating procedure (nominal interest rate targeting) does not work. The central banks of the world have been doing this for a long time, and things don't come unhinged as a result.

    ReplyDelete
  8. Steve: we are getting somewhere.

    "It's not like the standard operating procedure (nominal interest rate targeting) does not work. The central banks of the world have been doing this for a long time, and things don't come unhinged as a result."

    We always knew it couldn't work in theory (the indeterminacy problem), but it seemed to work in practice, so eventually we kept quiet. But now it has stopped working, because there is no way the Fed can communicate an easing of monetary policy by saying something about nominal interest rates when everybody knows the Fed can't lower the nominal interest rate further. By talking of monetary policy as setting interest rates, the Fed has, in effect, gagged itself.

    It worked for as long as it did because people had enough confidence in the Fed hitting an implicit inflation target (this creates a nominal anchor resolving the indeterminacy problem), and because prices and expectations were slow to adjust, relative to the speed at which the Fed would change the nominal interest rate, so the Fed was able to keep the pole balanced more or less upright.

    But when a big shock hit quickly, and the Fed didn't move its hand even quicker, and when people lost confidence in the Fed's ability to hit its target, and that loss of confidence was exacerbated by the zero bound, given the way people framed monetary policy, it stopped working. It has certainly stopped working now. Things have come very unhinged.

    ReplyDelete
  9. Continued (my comment was too long to be allowed).


    "Further, Friedman's prescription of constant money growth as a policy rule, over any horizon, has been a practical failure, as the "demand for money" is not stable over any horizon."

    Agreed, unfortunately. When I speak of monetary targeting, I do so mainly to illustrate a theoretical point. But I would not rule out a return to flexible monetary targeting. Sure, the demand for money moves around. But then the (nominal equivalent of) the natural rate of interest moves around too. So M-targeting and i-targeting are equivalent in that regard. And M-targeting creates a stable equilibrium, while i-targeting does not. And M-targeting would let the Fed communicate a commitment to monetary easing, while i-targeting under present circumstances (while we are at the zero bound) does not. So M-targeting beats i-targeting under present circumstances.

    But yes, something else would probably be better. My own inclination is to target the nominal price of some real asset, rather than i or M. Maybe the S&P500? Dunno.

    "Some people like to argue as if the central bank operates by "setting" an overnight market nominal interest rate, which of course it can't in general."

    Yep. Setting a nominal interest rate is not something that central banks really really do. It's a way for economists to model what central banks do. Just like (at a longer horizon) setting CPI inflation can be an economist's modelling assumption about what central banks do. (Thank God, I've been trying to convince people of this point for such a long time). The only thing central banks really control is their own balance sheets.

    But, they do control their communication. And right now they communicate what they are doing in terms of the nominal interest rate. And this means that people, (and that includes most economists) cannot even think of monetary policy except as the central bank setting nominal interest rates. The Fed's way of framing what it does now appears as a concrete reality. That framing is what creates people's expectations about off-equilibrium-path behaviour.

    Two methods of framing can be observationally equivalent to an outside observer, who only sees the data. He can't tell just by looking what's exogenous and what's endogenous. It only matters at nodes that are off the equilibrium path, and we never observe those nodes. But people's expectations of what would happen counterfactually at those nodes nevertheless determine which path is the equilibrium path.

    So the Fed needs to switch to another framing. So it can communicate monetary easing, and at the same time allow i to rise above zero as an endogenous consequence. And this is very different from recommending an exogenous increase in i. Given the current framing, that could only be interpreted, by the populace, as monetary tightening. Which is how we got into this argument in the first place.

    ReplyDelete
  10. If consistently low interest rates lead to deflation, why didn't Volcker use lower interest rates to stem inflation instead of using higher interest rates?

    It seems like you are assuming things to be constant which can be changing - output and real interest rates. As a non-economist, what I found to be Kocherlakota's logical flaw is the assumption of 1-2% real risk-free rates. A little research showed that there have been periods where the risk-free rate (fed funds - inflation) is a couple points higher or lower than the 1-2% range. It makes the math simpler to assume things, but the conclusion drawn is then dependent on the quality of the assumption.

    ReplyDelete
  11. Karl Smith responds here:
    http://modeledbehavior.com/2010/08/26/wonk-city-more-on-deflation-and-interest-rates/

    ReplyDelete
  12. Nick,

    You had me a little worried with the balancing poles and such. However, you have hit on something with the "framing" issue. My guess is that there are people on the FOMC, principally the non-academic types, who are having problems understanding current policy. If they have a problem, your average Joe Schmoe can't figure it out either. A regime where the decision is always a choice among: (i) up 1/4 point; (ii) down 1/4 point; (iii) unchanged; seems easy to understand (though we know there are complicated things going on there too, that aren't communicated). Talk of "quantitative easing," "draining reserves" through term deposits and reverse repos, etc., creates a lot of noise. In that context, seemingly inconsequential changes in policy, like in the last FOMC statement, can move long bond yields substantially. I'm skeptical that targeting some other asset price is the way to go (the nominal interest rate is just asset pricing of course) though.

    jimcaserta,

    This is a long-run proposition. I know it's confusing. Tighter monetary policy, in the short run, raises the nominal interest rates. Ultimately, the lower monetary growth as the result of tightening leads to lower inflation and lower nominal interest rates over a longer horizon. Look at what happened after the Volcker tightening. Nominal interest rates went up, then they came back down again as the inflation rate fell. In the short run, there is a "liquidity effect" whereby tight monetary policy tends to increase the nominal interest rate. In the long run, what dominates is the "Fisher effect" whereby an inflation premium gets built into the nominal interest rate.

    ReplyDelete
  13. It seems like the consensus here is that the trouble with Narayana's assertion is that it ignores how off-equilibrium strategies determine which equilibrium actually arises. If so, should the Fed start following the policy prescriptions of Atkeson, Chari and Kehoe's Sophisticated Monetary Policy? If the theoretical results hold in practice (and obviously that's a big if) that type of policy should take care of off-equilibrium path strategies as well, right?

    ReplyDelete
  14. So if raising rates for a period can lower inflation, can't lowering rates for a period increase inflation? That seems the basic argument.

    Also, the premise would be that, had Volcker kept rates high, inflation would have stayed high? So if we had a constant fed-funds rate of 12%, we'd have constant 10% inflation?

    ReplyDelete
  15. It's a little hard to understand exactly where people are coming from here. Nick says things about Nash equilibrium, but I'm not sure he's using the right words to express himself. I saw the Sophisticated Monetary Policy paper, but I can't remember exactly what is in it. I do know that it is straightforward to construct policy rules that kill undesirable equilibria, in a wide class of models where multiple equilibria might be a problem. It's easiest to do this in overlapping generations setups - you just back out the rule from the first order condition.

    ReplyDelete
  16. I'm not sure exactly what happened to the Sophisticated Monetary Policy paper. Chari presented it at Minnesota Macro last summer but I haven't heard anything about it since then.

    You have the basic idea of it, though. They propose "hybrid" policies that use interest rate rules on the equilibrium path, but switch to something else (like money) if anyone deviates.

    ReplyDelete
  17. Stephen,

    "Tighter monetary policy, in the short run, raises the nominal interest rates. Ultimately, the lower monetary growth as the result of tightening leads to lower inflation and lower nominal interest rates over a longer horizon."

    The reason Kocherlakota is wrong is that he's arguing that, at some point, looser monetary policy will raise interest rates in the short run and that we can therefore interpret an increase in interest rates as a loosening of monetary policy (which will thus prevent deflation). In some models, he's right about the long-run equilibrium, but the conclusion he draws from it is wrong in any model that makes the type of short-run/long-run distinction that you are making in the quoted passage above.

    ReplyDelete
  18. Steve and Nick,

    Let me try to reconcile things here (you tell me if I have succeeded).

    The reference paper I have in mind is by Peter Howitt (JPE, 1992), "Interest Rate Control and Nonconvergence to Rational Expectations." Nick, I'm sure you've read this, or at least familiar with the arguments. Steve, even though Peter was a colleague of yours, I am doubting that you ever read it (you should, if you haven't, and I apologize if you have read it!)

    The point of his paper is this. Interest rate pegging is perfectly feasible, if one assumes rational expectations. But if we depart from the rational expectations solution concept, even just marginally, and (I think) in a perfectly reasonable way (he studies a class of learning rules with very weak restrictions), then interest rate pegging will lead to the familiar Wicksellian "cumulative process." Pegging it too high will lead to a deflationary spiral, pegging too low will lead to the opposite.

    Of course, this assumes a credible commitment to a fixed interest rate. If I recall, Howitt also shows how an interest rate policy rule (the "Taylor principle") can mitigate the cumulative process.

    And so, I now interpret Nick's original "shock" as a disbelief that Narayana is apparently unware of the Wicksellian cumulative process. I would recommend cutting NK some slack. He was speaking to a lay audience about a number of issues with limited time on hand. It would be interesting to see how he formalizes his argument (we'll be meeting with him next week at the SL Fed).

    So, as a few have already pointed out, I think that the divergent viewpoints ultimately relate to the question of "out of equilibrium" dynamics.

    A final thought. If pegging the nominal interest rate at a "low" rate for a prolonged period of time results in a Wicksellian inflationary spiral, why have we not seen this happening in Japan? Just wondering.

    ReplyDelete
  19. David,
    BoJ has raised rates a bit every time there was a recovery, this was a credible signal that they are ready to do more if an inflationary spiral is starting. There were also changes in QE and fiscal policy.

    ReplyDelete
  20. Money Demand Blog,

    Can you point me to a useful reference in support of your point? Thanks.

    ReplyDelete
  21. Andy,

    No, this is what he says:

    "To sum up, over the long run, a low fed funds rate must lead to consistent—but low—levels of deflation. The good news is that it is certainly possible to eliminate this eventuality through smart policy choices. Right now, the real safe return on short-term investments is negative because of various headwinds in the real economy. Again, using our simple arithmetic, this negative real return combined with the near-zero fed funds rate means that inflation must be positive. Eventually, the real economy will improve sufficiently that the real return to safe short-term investments will normalize at its more typical positive level. The FOMC has to be ready to increase its target rate soon thereafter."

    He's saying that ultimately, as the economy recovers, of its own accord, then the Fed will respond by raising the nominal interest rate target. My story was purely one about what happens in response to a change in monetary policy, holding everything else constant.

    ReplyDelete
  22. David @10.34. Yes, that does largely reconcile it (for me). I can't actually remember if I have read Peter's paper, but your description of it is very familiar to me.

    Start with sticky prices and expectations. Pegging a nominal interest rate gives us an unstable long run equilibrium. We move away from it. Now take the limit of the model as price and expectations flexibility increases. It stays unstable, but diverges even more quickly, explosively in the limit. But *at* the limit, I would say that anything can happen.

    "A final thought. If pegging the nominal interest rate at a "low" rate for a prolonged period of time results in a Wicksellian inflationary spiral, why have we not seen this happening in Japan? Just wondering."

    I'm wondering the exact opposite. Given the recession, and deflation, I would assume the natural rate is so very low that it's still below the low actual rate. So the question should be: why hasn't Japan imploded into ever-accelerating deflation? I don't know. Maybe prices/wages really are totally sticky downwards. Maybe there's something missing from the standard Neo-Wicksellian model. Dunno.

    Steve: "You had me a little worried with the balancing poles and such. However, you have hit on something with the "framing" issue."

    Thank God! Nearly everybody else hates my "framing issue" stuff. They think I'm totally out to lunch when I say that interest rate targeting is a social construction of reality, and a very bad one at that. (But most people like my pole analogy).

    "It's a little hard to understand exactly where people are coming from here. Nick says things about Nash equilibrium, but I'm not sure he's using the right words to express himself."

    I'm pretty sure I'm not using the right words. I'm desperately grabbing for every bit of vocabulary I can find to try to express my point.

    If I wanted to keep it simple, I would just stick to talking about what's exogenous vs endogenous under i and M targeting, and stability of the long-run equilibrium.

    But I think there's something deeper going on here too. And from a policy perspective, right now, that something deeper could let the Fed escape the lower bound. So it really matters.

    ReplyDelete
  23. David,

    Things are confusing enough without "disequilibrium" and ad-hoc expectations, and Wicksellian cumulative processes coming into the mix. You're free to do that, if you want, but it doesn't help me understand anything.

    ReplyDelete
  24. Stephen:

    This discussion started me thinking about the postbellum period of deflation (1867-1897) and the fisher identity during this time. This was a 30-year period of real growth that average almost 4% a year while deflation averaged about 2% a year. Though there was no central bank, it would make for a great case study of interest rates in a deflationary environment. Unfortunately, this period is problematic because it is plagued by the Gibson Paradox,that observation that nominal interest rates followed the price level rather than the inflation rate. Still, there has to be some insights from this time for this discussion...

    ReplyDelete
  25. David Beckworth:

    I imagine that the data is pretty lousy for that time period, but I would love to see it anyway. Can you share any good references that study that episode?

    DA

    ReplyDelete
  26. Yes, Champ, Smith and I (Champ-Smith-Williamson, CJE 1996) looked at issues related to banking panics in Canada/US and currency elasticity in the National banking era. We used some call loan interest rate data - not the greatest.

    ReplyDelete
  27. David Andolfatto,
    this is an excellent starting point:
    http://www.boj.or.jp/en/type/release/zuiji_new/k060714.pdf

    ReplyDelete
  28. Not really very useful to look at 1867-97 - completely different monetary regime. Not comparable with today's fiat currency/ central bank. Expectations totally different.


    pe

    ReplyDelete
  29. Steve:

    I posted a response to your reply above, but it appears not to have appeared. Let me try again.

    If I understand Nick correctly, he is worried about the "stability" properties of RE equilibria. I recommend that you read Howitt's paper in order to understand the issues that concern some people. You may not find it personally useful as way to organize your own thinking, but at least you'd be better aware of where your critics are coming from. And, ultimately, it would allow you to communicate more effectively why you feel the way you do.

    ReplyDelete
  30. Stephen, you write, "He [Kocherlakota] is saying that ultimately, as the economy recovers, of its own accord, then the Fed will respond by raising the nominal interest rate target."

    That's absolutely true, but then it's the higher inflation rate leading to a higher nominal interest rate target. I don't think that's what Kocherlakota said: he said the higher nominal rate would lead to higher inflation.

    Now, in my understanding, the Fed sets a higher nominal rate (in the short run) by decreasing the money supply, which would decrease inflation and, in the long run, decrease the nominal rate. If you think Kocherlakota meant the Fed to increase the money supply (which would increase the nominal rate in the long run, but leave it the same, at 0, it in the short run), why do you think he didn't mention the money supply?

    ReplyDelete
  31. Jeffrey:

    Narayana says this: "Eventually, the real economy will improve sufficiently that the real return to safe short-term investments will normalize at its more typical positive level. The FOMC has to be ready to increase its target rate soon thereafter."

    My interpretation of this is that: (i) The Fed intends to be holding the total supply of outside money constant for a while. (ii) The large quantity of outside money outstanding is not causing high inflation because the banks are holding most of it as reserves - essentially the stuff looks like T-bills. The quantity of currency outstanding is actually increasing at about 4-5% per year. (iii) The inflation starts to happen when the reserves start to "run off," i.e. lending becomes more profitable, reserves become less attractive, and in the process they turn into currency, increasing the price level. To make the reserves more attractive to hold and hold down inflation, the Fed has to increase the interest rate on reserves. Nowhere in this scenario did the Fed do anything to the stock of outside money. Open market operations to change the money supply (in order to target the fed funds rate) is what goes on in normal times - when the stock of excess reserves is essentially zero.

    ReplyDelete
  32. "I can think of the instrument the central bank sets as either the money growth rate or the nominal interest rate - that part is irrelevant."

    This was the problem for me as well, it's a matter of cause and effect. Which is which isn't irrelevant. If either you or Kocherlakota had suggested the Fed needed to raise long term inflation expectations, in order to prevent a long period of deflation and low interest rates, that would have made sense to me.

    But phrasing it the other way around, that a low fed funds rate can "lead to" deflation seems to be putting the cart before the horse. It may be that Kocherlakota mispoke and "lead to" was a poor choice of phrasing.

    But I think this is where the reaction is coming from. I don't think there's really disagreement as to what the long term equilibrium would look like. But as to how you get there, it seems to me that changes in m cause the changes in i. And holding the fed funds rate low for an extended period, even as the economy recovered, would lead to an increase in m.

    ReplyDelete
  33. No, let me correct that, it's not possible he misspoke:

    "Long-run monetary neutrality is an uncontroversial, simple, but nonetheless profound proposition. In particular, it implies that if the FOMC maintains the fed funds rate at its current level of 0-25 basis points for too long, both anticipated and actual inflation have to become negative."

    He is clearly saying that holding the fed funds rate low would lower anticipated inflation.

    Now I happen to think that monetary neutrality is not exactly correct either, as money does impact inflation, which like all forms of taxation does have some real effects.

    But monetary neutrality, even if you accept it, does not suggest that inflation expectations respond to interest rate targets. How would you even maintain a low target for "too long" without this resulting in an increase in m? What is the mechanism here?

    ReplyDelete
  34. Stephen: I'm beginning to think we agree about everything except what Kocherlakota really meant by his speech. You seem to believe that:
    1) The Fed should keep the money supply high and growing for now, even if that leads to prolonged low interest rates.
    2) When the economy begins to recover, inflation will begin to head up, and at that point the Fed needs to reduce money supply growth so that interest rates go up so that inflation stays in its normal range.

    I agree with all that, and I think Nick Rowe, et. al. do too. We've read Kocherlakota differently, and I'm still more persuaded by acerimusdux interpretation, but I'm satisfied as long as we agree on what Mr. K. should have meant.

    ReplyDelete
  35. I urge everyone who has not done so to read the paper by Peter Howitt mentioned by David Andolfatto. Nick and Andy's positions correspond exactly to that of Howitt.

    Here's what Narayana said:

    "over the long run, a low fed funds rate must lead to consistent—but low—levels of deflation."

    Imagine, instead, that he had said the following:

    "over the long run, an equilibrium with a low fed funds rate must also involve consistent—but low—levels of deflation"

    This latter statement would have been completely uncontroversial to Nick and Andy, and we would be arguing only over its practical relevance. This question of relevance is exactly what the Howitt paper addresses.

    Howitt claims that the older adaptive expectations models (which would predict accelerating inflation if the nominal rate were held too low for too long) generate more robust insights in this environment than RE models do. That is, the equilibrium that NK has in mind is not dynamically stable under a broad class of sophisticated learning mechanisms including Marcet/Sargent least squares learning and various Bayesian variants.

    It's impossible to appreciate this (very important) point without stepping outside the bounds of equilibrium models for a moment.

    ReplyDelete
  36. David: "If I understand Nick correctly, he is worried about the "stability" properties of RE equilibria."

    You understand me correctly.

    Jeffrey: "I agree with all that, and I think Nick Rowe, et. al. do too."

    Yes, I do too.

    ReplyDelete
  37. "He's saying that ultimately, as the economy recovers, of its own accord, then the Fed will respond by raising the nominal interest rate target."

    From what I've read, the idea is that as the economy recovers, keeping the fed-funds rate low would cause deflation? I think the problem is the assumption that the real risk-free return would trend back to its 'norm'. If you had a growing economy with 0.25 fed funds rate, you'd have people searching for yield in all the wrong places (didn't this happen). Did this cause prices to fall? If you had businesses being more attractive investments, and less buyers of treasuries, the fed would have to buy more bonds to keep the rate low.

    The whole equation rests on the assumption that the real risk-free return goes back to its norm. If that doesn't happen, then the equation will tell you something else. Is there some axiom about 1-2% risk free real return on your money?

    Why wouldn't the real risk-free return be zero or negative? What is the benefit to rewarding taking no risk?

    ReplyDelete
  38. What a difference a word makes. I agree with Rajiv that this controversy seems to be due to a poor choice of words by Narayana. He must have been talking about steady state equilibria consistent with the Fisher equation. For that matter even Bullard's comments on the seven perils involve a lot of implicit verbal theorizing on the dynamics around the Fisher equation and another relatively non-controversial steady state relationship, an interest rate policy rule conditioning on inflation and subject to a lower bound of zero. Even if Narayana is not aware of Howitt or Wicksell he must know Sargent and Wallace's paper on the optimal monetary instrument. So he is unlikely to argue for a non-contingent interest rate policy, and in his speech he does talk about raising the rate once the economy improves, whatever that means. Of course, given current fiscal policy he might be influenced by Mike Woodford's pathbreaking work on the fiscal theory of the price level.

    ReplyDelete
  39. Anonymous,

    Yes, he certainly knows all that stuff well. I found the fiscal theory of the price level hard to understand, until I read this:

    http://www.minneapolisfed.org/publications_papers/pub_display.cfm?id=792

    ReplyDelete
  40. He probably knows these models, but I think he still got confused in this instance and jumbled together a short term model with a long term model in a way that doesn't fit.

    If you are assuming monetary neutrality, I would think that would occur in two ways. One would be that private actors can adjust to an increase in the money supply by raising prices, leading to inflation, but affecting no real variables. The other would be that they could offset the Fed action in ways that decrease the money supply, which not only would prevent inflation from occurring, but would offset any move in the FFR as well.

    In order to have a model which assumes monetary neutrality, and in which inflation does not occur in response to an attempt to increase the money supply, I think you would have to do away with the assumption that the FFR is exogenous.

    And in the long run, it may well be that private markets, moreso than the Fed, will be responsible for pushing the FFR higher as recovery proceeds. But should the Fed try to sustain lower rates against market forces in this case, that policy would still be inflationary, not deflationary.

    ReplyDelete
  41. All I can say is that I read the speech, and I was not confused.

    ReplyDelete
  42. "I can think of the instrument the central bank sets as either the money growth rate or the nominal interest rate - that part is irrelevant."

    Is that "money growth" actually debt growth?

    ReplyDelete
  43. Stephen Williamson said: "jimcaserta,

    This is a long-run proposition. I know it's confusing. Tighter monetary policy, in the short run, raises the nominal interest rates. Ultimately, the lower monetary growth as the result of tightening leads to lower inflation and lower nominal interest rates over a longer horizon."

    What if supply falls?

    ReplyDelete
  44. Steve: to make yourself understood it may be useful to cite Milton Friedman,

    "Low interest rates are generally a sign that money has been tight, as in Japan; high interest rates, that money has been easy... After the U.S. experience during the Great Depression, and after inflation and rising interest rates in the 1970s and disinflation and falling interest rates in the 1980s, I thought the fallacy of identifying tight money with high interest rates and easy money with low interest rates was dead. Apparently, old fallacies never die."

    ReplyDelete