Monday, February 3, 2014

The Low-Inflation Policy Trap

Some central bankers in the world seem surprised that they are seeing low rates of inflation.
The chart shows headline CPI inflation for Canada, Germany, Italy, Japan, the UK, and the US. Except for Japan, all of these countries experienced the most recent peak in the inflation rate about mid-2011, and their inflation rates today are lower than they were then. All of the central banks responsible for monetary policy in these countries have an inflation target of 2%. But, except for the UK, which is now hitting the 2% target, all have fallen short, and two (Canada and Italy) have inflation below 1%.

As I have stated before, (here and here), for the U.S., the current policy stance is a trap. The Fed clearly intends to maintain its policy interest rate target at essentially zero, possibly well into 2016. But, particularly as the economy continues to strengthen, most forces are pushing short-term real rates of return up. With short-term nominal rates of return pegged at zero by the central bank, the inflation rate has nowhere to go but down. But central bankers appear only to understand the short-term liquidity effects of central bank actions. They seem to think that a low short-term nominal interest rate must mean high inflation, even if the nominal interest rate is persistently low. Though real rates of interest are certainly not constant, and there can be persistence in deviations of real rates of interest from long-run averages, if the short term nominal interest rate is low for a long period of time, then the inflation rate is guaranteed to be low.

The most recent FOMC statement contains exactly that confusion between the short-run and long-run effects of monetary policy:
The Committee continues to anticipate, based on its assessment of these factors, that it likely will be appropriate to maintain the current target range for the federal funds rate well past the time that the unemployment rate declines below 6-1/2 percent, especially if projected inflation continues to run below the Committee's 2 percent longer-run goal.
So, if inflation continues to be low, the policy interest rate will continue to be low, which implies a low inflation rate...

The ECB President seems to suffer from the same confusion. See this story on how he has pledged to keep nominal interest rates low in the face of low inflation:
Draghi reiterated his commitment to keeping borrowing costs low “for an extended period of time” as policy makers continue their deliberations over whether they have done enough to prevent deflation...
This is not exactly a quote from Mark Carney, the Governor of the Bank of England, but he seems to think the same thing:
Low inflation also meant that raising interest rates became less urgent even though the economy is growing.
With Carney, we'll have to see what he says if the inflation rate falls below his 2% target. My guess is that he'll fall into line with Yellen and Draghi.

Finally, this comes from a speech by Steve Poloz, the Governor of the Bank of Canada, who discussed current risks in Canadian central banking, including the possibility of deflation:
Let’s switch gears and look at the risk of outright deflation. Like out-of-control inflation, deflation can become a spiral, but a downward one. Expectations become unanchored on the downside, and people put off their purchases because they expect things to be less expensive later. Demand declines with prices, while the weight of debt on the economy grows.

In the Great Depression, consumer prices in Canada fell 25 per cent, and national output dropped by almost a third. The human cost was staggering, with unemployment reaching 20 per cent. A milder form of deflationary trap has nagged Japan for the past 20 years.

What I am describing is an economy-wide process of deflation, which is quite different from individual prices falling because of improved competitiveness in an economy that is still growing strongly.

Today, the concern is that even though policy-makers were successful in avoiding global deflation in the wake of the 2008 crisis, there is still a risk that inflation could creep down into deflationary territory as the aftershocks of this crisis persist. It is, at least in part, to counter that risk that central banks in a number of countries have kept interest rates very low and used unconventional monetary policies, such as quantitative easing, to provide additional stimulus to their economies.

History has taught us that deflation usually comes in the wake of a financial crisis. This was true of the Great Depression, and of the Japanese deflation of the 1990s. Perhaps the most important lesson of the crisis, then, is that a stable financial system is necessary to keep inflation low, stable and predictable - and limit the risk of falling into a deflationary trap.
The first part is a standard story one hears about deflation. As the story goes, the economy can enter a "deflationary trap" in which expenditure is indefinitely postponed, and things get generally awful. As far as I know, there is no sound theory that actually delivers such a phenomenon. We certainly have multiple equilibrium models in which the economy can be stuck forever in a Pareto-dominated equlibrium, but I don't know of a model like that in which a bad equilibrium is associated with deflation (maybe you do?). Indeed, in a wide class of models, deflation can be associated with Pareto optima - that's the logic of the Friedman rule (not that I'm endorsing that).

Poloz goes on to discuss the empirical relationship between deflation and poor performance on the real side of the economy. But it's quite possible that we associate low inflation, or deflation, with bad outcomes on the real side because those outcomes make the central bank respond with zero short-term nominal interest rates at zero for long periods of time, which in turn produces low inflation. But that wouldn't entirely explain the behavior of a Poloz central bank, as he also thinks good insurance against low inflation is a low policy interest rate for the central bank - sustained, apparently.

We can find a related type of argument in the minutes of the December FOMC meeting:
A few participants raised the possibility that recent declines in inflation might suggest that the economic recovery was not as strong as some thought.
We could call this the "hidden output gap argument." These FOMC participants are hardcore Phillips curve adherents. In the face of what appears to be a declining output gap, by any measure, they insist that the output gap must be rising, because inflation is falling.

One feature that sticks out in the chart at the top of the page is the behavior of inflation in Japan - obviously it's very different from the inflation behavior in the other countries. Part of what's interesting about this is that the increase in the inflation rate in Japan, by about two percentage points, coincides with the Bank of Japan's "Qualitative and Quantitative Monetary Easing" policy, which commenced in April 2013. Details of this program are in the April 2013 policy statement of the Bank of Japan. Basically, the B of J intends to double the size of its balance sheet over a period of about two years, by purchasing longer-maturity Japanese government bonds and private assets (exchange-traded funds). A curious part of the policy is the following:
With a view to pursuing quantitative monetary easing, the main operating target for money market operations is changed from the uncollateralized overnight call rate to the monetary base.
The B of J pays interest on excess reserves and, under this policy, the interest rate on reserves (IROR), is a key part of its policy. It's not clear whether the officials at the B of J understand this or not. Though rules for how the IROR should be set are not addressed in any policy statements of the B of J post-April 2013, as far as I can tell, arbitrage tells us that the overnight call rate is equal to the IROR. Here's the overnight rate:
Since April 2013, this overnight rate has been exactly .07%. So, the IROR is .07%. The key question is: How is the B of J going to set the IROR going forward?

My advice to the B of J would be the following. The QE policy is just foreplay. Get to the main event. The B of J cannot sustain 2% inflation if it keeps the IROR at .07% for the next two years. That policy will make it likely that the B of J will fall short of its 2% inflation target, just as is happening in the Euro zone, the U.S., Canada, and - shortly - the U.K.

Here's the same scatter plot I constructed in this post for the U.S., but in this case for Japan, using data from 1985-2013, plotting the overnight call rate vs. the CPI inflation rate.
That's roughly the same thing as what you see for the U.S. There's a clear Fisher relation in the data, with some tendency for the real rate to be low at low inflation rates.

If inflation in Europe, North America, Japan, and elsewhere stays low - or falls into negative territory, that's not a disaster, I think. I don't think there is any serious economics that supports the conventional deflationary-trap-as-black-hole idea. But the policy trap that exists at the zero lower bound is a genuine trap, and it even has some genuine theory behind it, as Jim Bullard pointed out. I'm interested in watching how central bankers dig themselves out.

79 comments:

  1. Given your chart of the Fisher relation, are you saying that in order to combat high inflation, Paul Volcker should have cut interest rates to 1 or 2% rather than increasing them to 20%?

    ReplyDelete
  2. You're getting the idea. Look at what Volcker actually did. He increased the fed funds rate for only a short time, then he reduced it, and inflation fell. See this:

    http://newmonetarism.blogspot.com/2013/12/volcker-and-bernanke.html

    ReplyDelete
    Replies
    1. And I guess you'd say that Hjalmar Schacht stopped the Reichsmark hyperinflation in 1923 by dropping interest rates? You'd need some pretty good arguments to convince me of that one.

      Delete
    2. Different story entirely. For that you have to really get into the details of fiscal policy.

      Delete
  3. It all hinges on whether the Fed controls real rates, or whether they move subject to forces beyond the Fed's control...

    ReplyDelete
    Replies
    1. Well, all the Fed can actually do is to conduct asset swaps and, what is very important currently, set the nominal interest rate on reserves. In principle, there are things the Fed can do that will affect rates, but it's also true that there are many other factors - as with all real variables - that are going to determine real rates of return. It's not either/or.

      Delete
    2. Suppose we focus on standard models. Take a mainstream monetary model - e.g. cash-in-advance with some kind of liquidity effect in it. Then, do an experiment where there is a one-time, permanent increase in the money growth rate. Then, the real rate falls in the short run, and there is no effect in the long run. In the short run, the nominal rate falls, and at some point the Fisher effect takes over. In the long run the nominal rate increases one-for-one with the increase in the money growth rate.

      So, that's mainstream monetary economics. I've written down models where you can get permanent effects on real rates of interest from monetary policy, by taking into account the role that all assets play in exchange and as collateral. I think that's interesting, and I think those things matter a lot in the financial crisis and after.

      Now, for the other things driving real rates of return. Here you have to worry about technology, risk, collateral, liquidity, etc. Again, all of that stuff comes into play in a big way in the financial crisis and after.

      Note, for example, that the Fed has been telling us that the way QE works is really just like conventional monetary policy. Inflation expectations are "anchored," and if the Fed buys long-maturity assets, then long bond yields fall. But notice that the Fed always announced QE in terms of quantity targets. They never announced a target for a long bond rate. Why? They know that the Fed would have a hard time hitting such a target, because of all the other factors that are driving long bond rates.

      Delete
    3. I'd be lying if I said this wasn't very persuasive...

      Delete
    4. Here's what I learned from Deirdre (then Don) McCloskey: Our job as economists is persuasion. So if we can persuade people, that makes us happy. Thus, today I am a happy man.

      Delete
  4. I enjoy your inflation posts because you make don't hedge your position.

    Whether it's March of 2011:

    "1. All of the necessary conditions are there for a substantial inflation. I'm thinking of something on the order of 5% to 10%, and once it gets going it will be costly to stop it. The total stock of reserves will rise to about $1.6 trillion by the end of June, and that represents an accident waiting to happen. Right now, as Bernanke points out, financial markets are taking the Fed at its word. The margin between TIPS yields and nominal Treasury yields is not very large, reflecting modest anticipated inflation. But if people start to anticipate that the Fed will not be reducing the size of its balance sheet any time soon, and start to anticipate higher inflation, then nominal rates of return on assets will rise, making reserves undesirable to hold. The Fed can make reserves more desirable to hold by increasing the interest rate on reserves, thus cutting off the incipient inflation, but it will have a hard time doing that."

    or March of 2012:

    "We'll just have to wait and see. I think I've been consistent in saying that the Fed can control inflation if it wants to, but I've had my doubts about its resolve. At this point, the FOMC has boxed itself in, and I can't see any other future scenario than more inflation in the future. How much? I don't know. 5% or 6% by the end of the year?"

    I believe you'll be wrong about inflation this year as well.

    Regardless of the Feds low interest rate/tapering QE stance, employment and wages will improve more this year than any other during the recovery. Federal Employees have received the first across the board pay raise in 3 years. California, Texas, Florida and others are making up for years of layoffs, furloughs and/or salary freezes by raising wages and will likely begin hiring.

    Private sector employment has been carrying the recovery until now but there has been no wage pressure. That will change now that the public sector will no longer be providing a stream of newly unemployed workers.

    ReplyDelete
    Replies
    1. Thanks for neatly pointing out how precise Williamson's forecasts have been.

      He must hate Krugman who has all along predicated low inflation hate even more. If he were a proper scientist he would change his theory ... but as usual it is all about ego or ideology, not about facts.

      Delete
    2. Maybe I just enjoy confusing you. Come back in a year, and we'll see how it turned out.

      Delete
    3. Steve,

      Honestly, it's pretty hard buying anything you say with your forecasting record. You got the sign wrong and you're anywhere from 3 to 8 percentage points off on the actual level. And you continually double down and lose again. It's hard to imagine a worse performance. And, er, I don't think anyone is confused. Just the opposite.

      Delete
    4. I am just throwing ideas out there. I don't pretend to make forecasts. If I were doing it for a living (which I actually have in the past), obviously I wouldn't be offering my forecasts for free in a blog post. So stop being silly.

      Delete
    5. No no no you don't get it, Kyle. Empirical verifiability as a measure of epistemic progress is so passe in serious macroeconomics; you are just not sophisticated enough to understand, please...

      Delete
    6. "All of the necessary conditions are there for a substantial inflation. I'm thinking of something on the order of 5% to 10%, and once it gets going it will be costly to stop it."

      and

      "At this point, the FOMC has boxed itself in, and I can't see any other future scenario than more inflation in the future. How much? I don't know. 5% or 6% by the end of the year?"

      These are NOT forecasts!??

      If you "don't pretend to make forecast" stop writing what you think inflation will be at some future date. People tend to think of reflections concerning future outcomes as forecasts. And I don't think that's silly.

      Delete
    7. Let me put it this way. There are people like you, who really have no interest in the issues I want to discuss, or in exploring what is going on in the world. But for some reason - even though you know little about me or what I do - you have decided you don't like me. So you're looking for whatever you can quote out of context that you think will make me look like an idiot.

      If I cared about what you thought, I would be very careful about saying anything about the future, which of course none of us can see. But, in fact, I don't care what you think, and I'm not staking my career on whether those things come to pass. It's actually more interesting when the future doesn't unfold as I expected, as then I have something to think about.

      Here's another way to look at this:

      http://newmonetarism.blogspot.com/2013/12/noahs-complaint_2.html

      Delete
    8. The worst part is that he/she keeps repeating the same wrong things over and over. For example, empirical evaluation of theories relies on conditional forecasts! Being right or wrong in an unconditional forecast says nothing about the validity of the scientist's theory. What matters is if the forecast was right or wrong for the right reasons. Watching someone who cannot understand that simple notion trying to portray themselves as an expert of epistemology is amusing the first few times. But, frankly, it gets old pretty quickly.

      Delete
  5. "If inflation in Europe, North America, Japan, and elsewhere stays low - or falls into negative territory, that's not a disaster, I think. I don't think there is any serious economics that supports the conventional deflationary-trap-as-black-hole idea."

    The guy upon whose definition of real interest rates your funky theory relies did. Time for you to re-read "The Debt-Deflation Theory of Great Depressions".

    Let me tell you why low inflation / deflation is bad besides increasing the real burden of debt and slowing down deleveraging.
    Second, low inflation increases real rates which makes monetary policy less effectual.
    Third, low inflation makes real wages drop more slowly over time which together with wage rigidity leads to a slower decline of unemployment.

    ReplyDelete
    Replies
    1. I didn't say it wasn't bad. I'll rephrase what I think you're trying to say. The Fed says it is committed to 2% inflation, and currently people seem to believe that, given how we measure inflation expectations from asset prices. So, unanticipated deviations from 2% inflation are not good. Such deviations redistribute wealth between debtors and creditors in unanticipated ways, and a similar phenomenon applies to labor contracts. Predictable inflation is a good thing. We don't want the Fed to lose credibility. That said, I'm saying that maybe in this instance it's not a big deal. And maybe the Fed will catch on.

      Delete
  6. Steve,

    as you know I am receptive to your viewpoint. However, I am not sure that there is only one way to read the data. Specifically, I am not sure which the endogenous variable is. For example, suppose that to increase inflation a central bank lowers its targeted interest rate, thereby stimulating economic activity through the liquidity effect. As inflation rises towards or even above what is considered acceptable, the CB increases its policy rate to slow down and eventually end the ascend. Therefore, the economy ends up with higher inflation and a higher nominal rate. The reverse would happen if the CB wanted to lower inflation. I believe this is the New Keynesian story with endogenous monetary policy. In this case the Fisher relation will hold in the data. However, it is inflation that is driving interest rates rather than the other way round.

    To move on to current events, suppose that PK is correct, that monetary policy has run its course, and that inflation remains low despite the nominal rate being near zero. Again, the data will be consistent with a Fisher relation. However, this does not mean that the low nominal rate is causing the low inflation. Nor does it mean that a higher nominal rate is the remedy. It is possible that other factors that are keeping economic activity subdued, and that unless they are addressed raising the policy rate will only increase the output gap and push inflation further down.

    ReplyDelete
    Replies
    1. Yes, exactly. What is endogenous depends on the policy rule. And it's well known that, for example, the Taylor rule can have bad properties. This paper is precise about what that means:

      http://www.columbia.edu/~mu2166/perils.pdf

      Roughly, there are a lot of paths that lead you to a liquidity trap, where the economy gets stuck. And it's worries like this that keep it stuck there:

      "It is possible that other factors that are keeping economic activity subdued, and that unless they are addressed raising the policy rate will only increase the output gap and push inflation further down."

      You're assuming that the output gap drives inflation.

      I'm not sure what you mean by saying "suppose that PK is correct." I thought his argument was that more government spending solves everything. More G, more Y, and the Phillips curve gives us more inflation. Or, the NK argument is that more G gives us higher inflation, which gives us a lower real rate, which makes Y go up.

      Delete
    2. "You're assuming that the output gap drives inflation."

      Yes. I think this is what the typical New-Keynesian Phillips curve assumes. About PK, I had in mind the Krugman-Eggertsson deleverage story.

      But I think I am not clear about which interest I had in mind. I accept that issuing more Treasuries until their interest rate rises is a good idea. This is because most models support such a policy, even if they do so for different reasons. But I am not convinced that the same holds about the IROR.

      Thanks for the link! I will check the paper out once I find some time.

      Delete
    3. On the Phillips curve, I wasn't sure if you were talking about reality or a model.

      Eggertsson/Krugman is a bit funny. Basically, Krugman called in Eggertsson, who knows these models, to help him construct a model that would have all the working parts to justify his policy conclusions. The model says that if the borrowing constraint of the impatient people tightens, then the real interest rate will fall, much like in any incomplete markets model. But with sticky prices, the zero lower bound gets in the way, and so it's optimal for the nominal rate to be at the zero lower bound, but the real rate is too high. That's the context that justifies more government spending. But given the way the model is set up, there is a better policy, which actually goes after the credit market friction. That's a tax/transfer scheme. Conclusion: I think people are overly focused on sticky price frictions. If the problems are financial, that's what you want to address.

      Delete
  7. "all of these countries experienced the most recent peak in the inflation rate about mid-2011, and their inflation rates today are lower than they were then."
    At that time the ECB tried your "medicine" and increased interest rate twice! Seems it didn´t work!

    ReplyDelete
    Replies
    1. " increased interest rate twice! Seems it didn´t work!"

      Not much, and they didn't sustain it.

      Delete
    2. They didn't sustain it because raising interest rates when inflation is low and the economy is faltering is bad for economic activity, as they experienced. That is why.

      Delete
    3. So, if the ECB had wanted to increase inflation in 2011, permanently. how would they have done it? They would have reduced the nominal short rate for a short time, then increased it, and kept the nominal rate high forever. Basically Volcker in reverse. But what happens when we are at the zero lower bound and the central bank wants to increase inflation permanently? Obviously the the short rate can't go down, so there is nowhere to go but up. Simple.

      Delete
    4. What do you mean, reduce the nominal rate for a short period of time? How do you define short here? Is it time-contingent or is it state-contingent? If that is time-contingent, all it would happen is a interest rate increase, which we know is contractionary and more so under the European conditions. Well, come to think of it, if the ECB had followed your advice, we may well have seen some inflation... In the new peseta, new lira, new escudos...

      Delete
    5. "How do you define short here?"

      overnight.

      "If that is time-contingent, all it would happen is a interest rate increase..."

      You'll have to expand on that. I don't understand what you mean.

      Delete
    6. You said: "they would have reduced the 2011, permanently. how would they have done it? They would have reduced the nominal short rate for a short time, then increased it..."

      I asked what you mean reducing for a short period of time? Could you define this? I can think of two interpretations, one that makes sense and the other that does not. If "short period of time" means just enough time to affect inflation expectations, then that is state-contingent and looks like how a policy maker would think. On the other hand, if you mean "short period of time" as time in the chronological sense (one week, three months...), then that is what I would call time-contingent, which does not make a lot of sense and if anything needs to be better defined or else someone may read you and end up with the erroneous impression that you did not really think hard about it.

      Delete
    7. I'm not sure how hard you want me to think. You can understand I'm not going to sit all day and think about your comment, I think. If we were to write it down formally then, yes, we would need to write down a contingent rule for monetary policy, take proper account of monetary nonneutralities in the model, etc. "Short time" might apply to calendar time in a deterministic setting.

      Delete
  8. I can understand why a banker would be attracted to the idea of a 2% interest rate floor, since that is traditionally the average spread between the Fed Funds rate and bank accounts. When rates are low, banks respond by accepting lower profits rather than lowering customer rates below 0%. Don't ask me why. For some companies, higher rates fall directly to the bottom line. No wonder we see all sorts of crazy arguments for higher rates.

    ReplyDelete
  9. Steve: I very strongly disagree, of course. But I do give you credit for not fudging your views.

    As I have said before, if Steve Poloz (and all the other central bankers) has got the sign wrong, it would be an amazing fluke if the Bank of Canada had managed to keep inflation close to the 2% target for 20 years, if they had been moving the nominal interest rate the wrong way.

    "As far as I know, there is no sound theory that actually delivers such a phenomenon."

    John Cochrane said he liked my (sketch of a) theory, and he has *some* sympathy for your views on this question. So maybe you would like it too?

    http://worthwhile.typepad.com/worthwhile_canadian_initi/2013/12/getting-the-right-sign-on-the-nominal-interest-rate-signal.html

    ReplyDelete
    Replies
    1. Remember that Steve Poloz hasn't been Governor for the last 20 years, or even working for the Bank, so we can't say he was responsible for the success of inflation targeting. In "conventional" times, the Bank of Canada's inflation targeting policy (in conjunction with some kind of Taylor rule, presumably) seemed to work fine (in spite of Benhabib et al., which seems to say the Taylor rule sucks you into the liquidity trap forever). But now we're in a different world apparently. I didn't quite believe Bullard's reasoning when he was talking about this a few years back. Surely when faced with low inflation, central bankers would understand what is going on and change the policy rule. But they don't, and Poloz is a good example.

      Delete
    2. The central bank setting a higher interest rate is a signal. How will people interpret that signal? Will they say "the CB has increased its inflation target, so I expect higher inflation"? Or will they say "The CB thinks that inflation is above its target, so I should expect lower inflation"?

      Steve Poloz is new. But the idea that you need to raise the nominal interest rate if you think that inflation will be too high otherwise, and want to signal your belief, is old.

      Yes, it's a very crappy way to signal, especially at the ZLB, but this is the language that central banks use, and this is how people are used to interpreting that language.

      Delete
    3. Sorry to hijack a thread, but I find this comment interesting:

      "Surely when faced with low inflation, central bankers would understand what is going on and change the policy rule."

      So people have been wasting time arguing over whether or not QE is deflationary, when in reality pretty much everyone seems to agree that the target/policy rule needs to be changed to effect regime change and escape the current situation?

      Or have I misunderstood?

      Delete
    4. There are features of the Taylor rule that cause the economy to get trapped at the zero lower bound. If that is undesirable, i.e. the central bank wants higher inflation, the policy rule has to change.

      Delete
    5. OK, thanks for the response.

      So "policy rule" in this case would be "communication + action" or is it more of a question of appropriate instruments vs. targets? Is there any guidance in economic theory on how the central bank should choose a policy rule to effect regime change? I suppose it depends on what is the policy-maker's goal.

      I read Svenson's paper on a "Foolproof Method", but it's been a while, and honestly, I'm quite naive on this and need to learn more to really participate in this (or else I have to give up and stop reading these blogs) - I have a good quant background (but applied purely to finance) and I've only taken monetary econ at the undergrad level. I suppose I should read your undergraduate textbook? :)

      Delete
    6. So, forget about quantitative easing, as that complicates things. The relevant policy instrument for the Fed now is the interest rate it sets on reserves (IROR). There are a few caveats, but roughly that will determine short-term nominal interest rates. So, we can think of a policy rule for the Fed as how the interest rate on reserves depends on what the Fed is interested in. Suppose the Fed is just interested in the inflation rate and the unemployment rate. The Taylor rule says basically that the Fed should increase the IROR if the unemployment rate goes down, and should decrease it if the inflation rate goes down. You can show that, under some conditions, this will give you a long-run outcome where the Fed hits its inflation target. But it's also a long-run equilibrium for the IROR to be zero, and inflation to be low or negative. So the Fed sees low inflation, and would like to lower the IROR, but it can't. Perhaps it sees a lower unemployment rate, but it's convinced that there is much more slack in the economy than this is capturing, which, again, makes it want to reduce the IROR. So, you get stuck with IROR=0 and inflation below where the Fed wants it to be. To solve the problem, the Fed has to do something different than is dictated by the simple Taylor rule. You might be able to figure out Bullard's paper, that I link to near the end of the post.

      Delete
    7. Ok, I'll try the Bullard paper. Thanks.

      Delete
    8. Ok, I read the Bullard paper quickly, and I found myself in agreement with much of it, even if I don't understand the underlying models of the Benhabib et al paper. I can't say there's much there in Bullard's exposition to disagree with, actually.

      I grabbed some data on Total CPI in Canada, the U.S., and Japan over the past 10 years ago, and also the Japanese Overnight Call Loan Rate. I also plotted a hypothetical price level for 2.5% inflation and 0% inflation. All series indexed to 100 at Dec 2001. You can see my graph here:

      http://imgur.com/9qqbyFG

      I think it's not controversial to say that Canada was targeting about 2% CPI over this time, and the U.S. was probably targeting something close to that or maybe a little higher, or whatever given the dual mandate. However, am I wrong to look at the Japanese case and conclude that they were just targeting 0% total CPI the whole time? Seems like there was a supply shock in 2007-2008 and total CPI rose above trend (just like in the other two countries) and they raised the overnight rate to tighten policy and bring CPI back down to the 0% target.

      I'm trying to understand how I make the leap from that interpretation to one that is consistent with my (perhaps incorrect, admittedly) interpretation of your view (or at least the view in the Bullard paper, which I think might be more like "The Bank of Japan saw inflation rising (in around 2007-2008) due to a supply shock, and took advantage of that to escape the unintended steady state and move up rates to the targeted steady state".

      Delete
    9. One last follow-up and then I'll stop - when I say I don't disagree with much of what Bullard says, I also can't say that I agree with it either - mainly I simply don't have the knowledge to come to any conclusion on this yet.

      The Benhabib et all idea seems interesting and I can understand (or at least partially understand) how that could work mechanically, where randomness or supply shocks or whatever can get you to the undesired equilibrium (without having read the paper or even knowing the slightest bit about these sort of models, I'm thinking of this like a first hitting time of a Brownian motion where when time horizon is long enough the probability of hitting the undesirable equilibrium could become very high if the drift rate is low - I have no idea if this is helping me understand better of it it's way off base, but that's what came to mind), but also it seems that a central bank can modify the FX value of it's currency or do something else to modify expectations even when in the undesirable equilibrium, and so there really can't be a liquidity trap in the sense of "there's nothing we can do, even if the central bank replaces all government liabilities with currency, we can't influence the nominal size of the economy".

      Even now, when I observe market reactions to Fed policy surprises, it seems that things "make sense" in that QE3 seemed to increase inflation expectations and raise asset prices (to which I guess you would respond with the "short run vs. long run" distinction).

      So really at this point I would summarize my understanding to be that QE (and low rates) are deflationary (even in the long run) only if the central bank is a strict Taylor Rule bank and the economy has already entered the undesired steady state. If the bank is not constrained by a strict Taylor rule, there should be no liquidity trap, there is no "QE = deflationary". Unconventional policy + policy rule changes (such as price level targeting with a catch up provision) accompanied by actions like large scale asset purchases could allow for exit from the unintended equilibrium. Is that a reasonable interpretation?

      Of course, if the bank isn't a Taylor Rule follower, then perhaps the economy never reaches that unintended equilibrium anyway. Which raises more questions.

      Delete
    10. "I think it's not controversial to say that Canada was targeting about 2% CPI over this time, and the U.S. was probably targeting something close to that or maybe a little higher, or whatever given the dual mandate. However, am I wrong to look at the Japanese case and conclude that they were just targeting 0% total CPI the whole time? "

      Canada has had explicit inflation targeting since the early 1990s. The target is 2%, with a tolerance of 1-3%. The Fed's target is now somewhat explicit, though they are now taking the "dual mandate" more seriously. In Japan, the inflation target is now explicitly 2%. I don't know enough to say what the Bank of Japan's explicit goals were before April 2013, but my guess is the goal was to be above 0%. The B of J I think has attempted to have more inflation, without success. Of course, my point is that they may have been doing the wrong thing. Now they think a large QE program has them on the right track. We'll see. QE is still poorly understood, that's for sure.

      Delete
    11. Maybe too early to tell, but I think I've actually learned something here, so thanks.

      Delete
  10. The NK model provides a clear explanation of why nominal rates need to be cut in order to return inflation back to target, or, as now, expected future rates. And BSG explains that same model has a costly FR like liquidity trap steady state. Are you saying that this is not a 'sound' model. My post at longandvariable elaborates on this reply. Tony Yates.

    ReplyDelete
    Replies
    1. This doesn't have much to do with whether NK is sound or not. It's certainly related to BSG, as it's Taylor rule reasoning that leads to the trap at the zero lower bound. But there are other things that can go wrong with the Taylor rule. The central bank can mismeasure or misperceive the output gap and the long-run real rate, for example.

      In the NK world, the Fisher effect is what determines the nominal interest rate in the long run. So, it's true in an NK model, as in any standard monetary model, that if the nominal interest rate goes up permanently, then the inflation rate goes up one-for-one with it, in the long run. In the NK model, presumably it's optimal to reduce the nominal rate temporarily before you increase it, if the inflation rate is to go up permanently, because of the sticky prices. Here's a conjecture. If there are menu costs instead of Calvo pricing, then if the central bankers wake up in the morning and determine that the inflation rate should rise by x%, then the way to do that is to increase the nominal rate by x% immediately, and permanently.

      Delete
  11. To everyone who thinks this post is crazy and they believe in the standard models-
    1) Aside from housing (which is pseudo fiscal policy), where do you think low interest rates generate real spending? What is the channel apart from wealth effect?
    There is NONE visible in the data.
    2) Zero (low) interest rates = contractionary fiscal policy as savers get zero (low) interest on savings and therefore income goes down.
    3) How do you explain Japan? Japan is the fly in the models. Your models are hopeless at explaining Japan.
    4) Bernanke, in trying to prevent a Japanese scenario in the US, has delivered us right into it - through a combination of bubble blowing and low rates for ever.

    ReplyDelete
    Replies
    1. Looks like somebody did not pass his undergraduate exams.

      1) If low rates spread throughout the banking sector and actually lead to low loan rates (Stiglitz pointed out long ago the importance of credit rationing, i.e. when we take a look at banking aka loan markets we cannot just look at the price (loan rate) but also have to look at quantity) the costs of investment falls and investment rises. All kind of investment, including stuff like tertiary education which is investment into human capital, not just housing,

      2) No idea how you connect the dots from low rates to contractionary fiscal policy and how you connect the dots from low returns on savings to a reducing income. Given that you already did not understand that low rates increase investment I doubt that actually have any mechanism or model in mind.

      3) Nope. There is the General Theory, there is "It's Baaack: Japan's Slump and the
      Return of the Liquidity Trap", some Woodford papers about the liquidity trap and so on. Mainstream macro has no problem explaining why monetary policy loses its traction at the ZLB and why inflation doesn't rise .... unless you actually try to get out of deep shit via fiscal policy.

      4) If Bernanke had actually raised rates it would have been worse, we would have had something like the Great Depression. Read your Friedman.

      This isn't rocket science, it is just Econ 101. Just because you did not learn or understand it doesn't mean it is wrong.

      Delete
  12. Right. It is econ 101. That is why millions are out of a job and countless others stuck in low paying part time jobs. Stop regurgitating dumb theories that do not work and look at what the real world is telling you.
    Investment will not rise until demand catches up and excess capacity disappears. Does not matter what interest rate the Fed engineers. When government debt is paying zero or near zero returns, the govt is NOT paying interest (equivalent to contractionary fiscal policy).

    Start thinking for a change.

    ReplyDelete
    Replies
    1. "...demand catches up and excess capacity disappears."

      Demand catches up to what?
      Excess capacity measured how?

      Delete
    2. "Demand catches up to what?
      Excess capacity measured how?"

      Typical neoclassical demand denial.

      Delete
    3. Does your tinfoil hat ever get too hot in the sun?

      Delete
    4. I'm not denying anything, I'm asking how these things are being measured. Make the case.

      Delete
    5. Do yourself a favour and take a look at the unemployment stats from time to time.

      Delete
    6. Seems you're avoiding answering the question.

      Delete
    7. You asked for a definition of the natural rate of interest (when you meet a monetary economist who has not read his Wicksell you know that you can stop taking him seriously) and now you ask for how we can measure cyclical unemployment or the output gap.

      Obviously you are not asking seriously, you are just indirectly showing that you do not think that either concept makes much sense.
      This is totally natural in an ideology which dictates that economies are always supply- and never demand constrained. It is also totally wrong.

      So tell us about why those millions of people are unemployed: skill mismatch, high taxes, luxurious unemployment benefits?

      If you had any connection with the real world you would know that millions of people are involuntarily unemployed (just ask a worker) and that there is ample of excess capacity (just ask a business owner) in all the Western economies right now.

      But I apologize for this intrusion, I won't ever bother you or any other Lucas fanboy with what happens in the real world again.

      Delete
    8. Well, that was a pretty empty rant. Give me some substance.

      Delete
    9. Steve, unless you say that millions of people are unemployed because the government wont hire them dont expect any civilized response. At least he hasnt called you scum of the earth yet...I retain the title, lol.

      Delete
    10. "Well, that was a pretty empty rant. Give me some substance."

      As I said, it is not about measuring anything, it is about you being in demand denial. When some folks have unlearned stuff that is more than seventy years old (or even longer given that you have not read your Wicksell) we are indeed in a dark age of macro.

      Delete
  13. It's interesting that natural rate proponents think they "know" that real rates should be negative, not positive. Therefore they won't believe your model, as it assumes that the real rate is persistently positive. They are taking a theoretical relationship and treating it as an observed fact. You are taking an observed fact (the tendency for positive real rates) and translating it into a theoretical relationship. I prefer your approach.

    ReplyDelete
    Replies
    1. Nope, we are OBSERVING that tens of millions are unemployed so the Wicksellian natural rate of interest is MOST LIKELY negative.
      You guys are on the other just playing the three monkey game and ignoring that there is unemployment at all. What cannot be according to the gospel of Lucas, a shortfall of demand aka an excess supply on the output and labour market, is simply ignored if it actually occurs.

      Delete
    2. I wonder what you think you are accomplishing here. Perhaps your personal unemployment, likely due to some form of mental deficiency based on my observations of your behavior, has left you too much free time. Perhaps we should ask the government to "employ" you digging holes in the ground, preferably in Guantanomo, which apparently would be good for everyone. Of course, in this case it actually would be good for us, since it would prevent the negative externality your stupidity imposes on us.

      Delete
    3. The issue on the real rate is the following. The NK position seems to be that there was a financial shock that made the real rate low. Sometimes that's represented simply as a preference shock, i.e. people got much more patient. Sometimes it's represented as a tighter borrowing constraint. But then there are two inefficiencies. There's the fundamental inefficiency - think of that as a credit friction. Then there's the inefficiency associated with the zero lower bound. Given the greater credit friction, the real rate falls, but it can't go low enough. So the real rate is low because of greater credit frictions. So the solution to the problem is that the real rate should rise - you need to think about financial market solutions that relax credit constraints. I think the NK reasoning is focused on the wrong problem. The real rate is too low, not too high.

      Delete
  14. I think what you are saying is that the rate is too low because of a capital mobility problem. There is too much capital locked in legacy uses, too little in higher return projects. This is consistent with the observed low rate of new firm formation. At the broadest level, this lack of mobility shows up as a lower threat of competition from new entrants or expansion projects. This, in turn, forces up corporate profits to gdp; normally this is a sign of health, but in this case a sign of structural problems. These are some of the observed facts that the natural rate argument has no explanation for (high profits, high commercial real estate valuations, etc).

    The solution to capital immmobility is not to subside legacy assets with negative real rates, but to allow positive real rates to achieve market-clearing prices for those assets. In most quarters, this is known as "liquidationism", and should never be spoken of.

    ReplyDelete
    Replies
    1. Put another way, government debt, which serves as collateral and is exchanged in financial markets, has a low real rate of return, which can be consistent with the marginal product of capital being high - or at least higher.

      Delete
  15. Steve,

    Thanks for the debate - this whole back and forth between you and others has given me a nice window into the mathematical models behind modern macro. It's been fun!

    I have to say though that I struggle with causation in your theory - and I think one has to be careful to distinguish backward looking inflation with inflation expectations. I read your theory as saying that low interest rates cause low expectations for inflation; low expectations for inflation then cause low actual inflation. And to be honest, I struggle with both steps. But a thought experiment - in your theory, why do stock prices tend to rise when the Fed lowers interest rates, and fall when the Fed raises them?

    ReplyDelete
    Replies
    1. "But a thought experiment - in your theory, why do stock prices tend to rise when the Fed lowers interest rates, and fall when the Fed raises them?"

      That's very tricky, and I'm not sure we can even say that's a characterization of the data. In general, stock prices move in response to new information, though there are theories which could explain some stock price movements as pure coordination phenomena. But let's say the latter don't matter. Then, if stock prices move in response to announcements about changes in the fed's policy rate, it's because market participants are surprised. So, the market could be surprised because the policy rate went up 1/4 point and they didn't expect it, or the market could be surprised because the policy rate went up 1/4 point, and they expected a half point. So in one case the market moves one way, and in the other case the market moves the other way, but it's the same increase in the policy rate that happened. So, it's hard to use asset price movements and the timing of policy announcements (i.e. event studies) to discriminate among different theories.

      Delete
    2. IR futures markets are commonly used to measure market expectations of Fed policy changes. If we can use futures markets as a proxy for market expectations before a fed announcement, and we can observe actual policy changes directly, then where does the difficulty lie in conducting event studies?

      Delete
    3. Stephen,
      Nicely put on event studies. Papers that use this technique unquestioningly seem to sail through peer review. Then they are cited over and over, including by people as intelligent as Ben Bernanke. How this all happens is one of the most curious things about the economics profession.

      Delete
    4. JSR,

      See the comment below. What do market participants know, and when do they know it? Our ability to discern that from market prices is limited. For example what you can pick up from TIPS yields and nominal bond yields about anticipated inflation is far from perfect. Cleveland Fed people construct an anticipated inflation measure from asset pricing data. I've looked at that, and I don't trust it. An event study definitely can tell you something like: the Fed said something and asset prices subsequently moved in a particular direction. Beyond that, it's pretty dubious science.

      Delete
  16. "So in one case the market moves one way, and in the other case the market moves the other way, but it's the same increase in the policy rate that happened."

    All one has to do is to control for ex-ante market expectations. Those can be obtained to a reasonable approximation from market prices or, if you will, surveys.

    ReplyDelete
  17. Sure, you can get some information, for example, there is a market in fed funds futures. There isn't a complete market in all the contingent claims we might be interested in, of course, so the information is limited, and in some circumstances could be quite poor, given what we want to know. We can't look inside the heads of market participants.

    ReplyDelete
  18. Your comment only implies that we should be cautious about interpreting the empirical results, which is true of inference of any sort, econometric or not.

    For instance, if I am measuring up Lebron James, I can tell he is taller than me even if I cannot tell his height but for some 3 inches error margin. Besides, the fact that I am biased and think that I am 2 inches taller than what I actually am does not invalidate my inference that Lebron James is taller than me. It would actually not invalidate any inference about my height relative to any professional basketball player that I know of.

    Where am I going with this? Some of the evidence about the effects of monetary policy is just like my height comparison with Lebron James. There may be biases and some uncertainty about the exact estimates, but there is a core of truth about which we can be fairly confident.

    ReplyDelete
    Replies
    1. "Your comment only implies that we should be cautious about interpreting the empirical results..."

      Yes, and sometimes we want to be cautious, to the point of ignoring some empirical results altogether.

      Delete
  19. "ignoring some empirical results altogether."

    This neatly sums up the the central problem of your school of "economics": it ignores facts and is thus unscientific.

    ReplyDelete
    Replies
    1. I'm not sure what my "school of economics" is supposed to be. I'm curious, though. What facts am I ignoring?

      Delete
  20. "the inflation rate has nowhere to go but down"

    Wrong. Dec-Jan was an historic inflection point for inflation. What lies ahead is stagflation.

    ReplyDelete