Thursday, July 31, 2014

The Call Is Coming from Inside the House

Paul Krugman wonders why no one listens to academic economists. Almost all the economists in the IGM Survey agree that the 2009 stimulus bill successfully reduced unemployment and that its benefits outweighed its costs. So why are these questions still controversial?

One answer is that economists don’t listen to themselves. More precisely, liberal economists like Krugman who want the state to take a more active role in managing the economy, continue to teach  an economic theory that has no place for activist policy.

Let me give a concrete example.

One of Krugman’s bugaboos is the persistence of claims that expansionary monetary policy must lead to higher inflation. Even after 5-plus years of ultra-loose policy with no rising inflation in sight, we keep hearing that since so “much money has been created…, there should already be considerable inflation.” (That’s from exhibit A in DeLong’s roundup of inflationphobia.) As an empirical matter, of course, Krugman is right. But where could someone have gotten this idea that an increase in the money supply must always lead to higher inflation? Perhaps from an undergraduate economics class? Very possibly -- if that class used Krugman’s textbook.

Here’s what Krugman's International Economics says about money and inflation:
A permanent increase in the money supply causes a proportional increase in the price level’s long-run value. … we should expect the data to show a clear-cut positive association between money supplies and price levels. If real-world data did not provide strong evidence that money supplies and price levels move together in the long run, the usefulness of the theory of money demand we have developed would be in severe doubt. 
Sharp swings in inflation rates [are] accompanied by swings in growth rates of money supplies… On average, years with higher money growth also tend to be years with higher inflation. In addition, the data points cluster around the 45-degree line, along which money supplies and price levels increase in proportion. … the data confirm the strong long-run link between national money supplies and national price levels predicted by economic theory. 
Although the price levels appear to display short-run stickiness in many countries, a change in the money supply creates immediate demand and cost pressures that eventually lead to future increases in the price level. 
A permanent increase in the level of a country’s money supply ultimately results in a proportional rise in its price level but has no effect on the long-run values of the interest rate or real output. 

This last sentence is simply the claim that money is neutral in the long run, which Krugman continues to affirm on his blog. [1] The “long run” is not precisely defined here, but it is clearly not very long, since we are told that “Even year by year, there is a strong positive relation between average Latin American money supply growth and inflation.”

From the neutrality of money, a natural inference about policy is drawn:
Suppose the Fed wishes to stimulate the economy and therefore carries out an increase in the level of the U.S. money supply. … the U.S. price level is the sole variable changing in the long run along with the nominal exchange rate E$/€. … The only long-run effect of the U.S. money supply increase is to raise all dollar prices.
What is “the money supply”? In the US context, Krugman explicitly identifies it as M1, currency and checkable deposits, which (he says) is determined by the central bank. Since 2008, M1 has more than doubled in the US — an annual rate of increase of 11 percent, compared with an average of 2.5 percent over the preceding decade. Krugman’s textbook states, in  unambiguous terms, that such an acceleration of money growth will lead to a proportionate acceleration of inflation. He can hardly blame the inflation hawks for believing what he himself has taught a generation of economics students.

You might think these claims about money and inflation are unfortunate oversights, or asides from the main argument. They are not. The assumption that prices must eventually change in proportion to the central bank-determined money supply is central to the book’s four chapters on macroeconomic policy in an open economy. The entire discussion in these chapters is in terms of a version of the Dornbusch “overshooting” model. In this model, we assume that

1. Real exchange rates are fixed in the long run by purchasing power parity (PPP).
2. Interest rate differentials between countries are possible only if they are offset by expected changes in the nominal exchange rate.

Expansionary monetary policy means reducing interest rates here relative to the rest of the world. In a world of freely mobile capital, investors will hold our lower-return bonds only if they expect our nominal exchange rate to appreciate in the future. With the long-run real exchange rate pinned down by PPP, the expected future nominal exchange rate depends on expected inflation. So to determine what exchange rate today will make investors willing to holder our lower-interest bonds, we have to know how policy has changed their expectations of the future price level. Unless investors believe that changes in the money supply will translate reliably into changes in the price level, there is no way for monetary policy to operate in this model.

So  these are not throwaway lines. The more thoroughly a student understands the discussion in Krugman’s textbook, the stronger should be their belief that sustained expansionary monetary policy must be inflationary. Because if it is not, Krugman gives you no tools whatsoever to think about policy.

Let me anticipate a couple of objections:

Undergraduate textbooks don’t reflect the current state of economic theory. Sure, this is often true, for better or worse. (IS-LM has existed for decades only in the Hades of undergraduate instruction.) But it’s not much of a defense, is it? If Paul Krugman has been teaching his undergraduates economic theory that produces disastrous results when used as a guide for policy, you would think that would provoke some soul-searching on his part. But as far as I can tell, it hasn’t. But in this case I think the textbook does a good job summarizing the relevant scholarship. The textbook closely follows the model in Dornbusch’s Expectations and Exchange Rate Dynamics, which similarly depends on the assumption that the price level changes proportionately with the money supply. The Dornbusch article is among the most cited in open-economy macroeconomics and international finance, and continues to appear on international finance syllabuses in most top PhD programs.

Everything changes at the zero lower bound. Defending the textbook on the ground that it's pre-ZLB effectively concedes that what economists were teaching before 2008 has become useless since then. (No wonder people don’t listen.) If orthodox theory as of 2007 has proved to be all wrong in the post-Lehmann world, shouldn’t that at least raise some doubts about whether it was all right pre-Lehmann? But again, that's irrelevant here, since I am looking at the 9th Edition, published in 2011. And it does talk about the liquidity trap — not, to be sure, in the main chapters on macroeconomic policy, but in a two-page section at the end. The conclusion of that section is that while temporary increases in the money supply will be ineffective at the zero lower bond, a permanent increase will have the same effects as always: “Suppose the central bank can credibly promise to raise the money supply permanently … output will therefore expand, and the currency will depreciate.” (The accompanying diagram shows how the economy returns to full employment.) The only way such a policy might fail is if there is reason to believe that the increase in the money supply will subsequently be reversed. Just to underline the point, the further reading suggested on policy at the zero lower bound is an article by Lars Svennson that calls a permanent expansion in the money supply “the foolproof way” to escape a liquidity trap. There’s no suggestion here that the relationship between monetary policy and inflation is any less reliable at the ZLB; the only difference is that the higher inflation that must inevitably result from monetary expansion is now desirable rather than costly. This might help if Krugman were a market monetarist, and wanted to blame the whole Great Recession and slow recovery on bad policy by the Fed; but (to his credit) he isn’t and doesn’t.

Liberal Keynesian economists made a deal with the devil decades ago, when they conceded the theoretical high ground. Paul Krugman the textbook author says authoritatively that money is neutral in the long run and that a permanent increase in the money supply can only lead to inflation. Why shouldn't people listen to him, and ignore Paul Krugman the blogger?

[1] That Krugman post also contains the following rather revealing explanation of his approach to textbook writing:
Why do AS-AD? First, you do want a quick introduction to the notion that supply shocks and demand shocks are different ... and AS-AD gets you to that notion in a quick and dirty, back of the envelope way. 
Second — and this plays a surprisingly big role in my own pedagogical thinking — we do want, somewhere along the way, to get across the notion of the self-correcting economy, the notion that in the long run, we may all be dead, but that we also have a tendency to return to full employment via price flexibility. Or to put it differently, you do want somehow to make clear the notion (which even fairly Keynesian guys like me share) that money is neutral in the long run. That’s a relatively easy case to make in AS-AD; it raises all kinds of expositional problems if you replace the AD curve with a Taylor rule, which is, as I said, essentially a model of Bernanke’s mind.
This is striking for several reasons. First, Krugman wants students to believe in the "self-correcting economy," even if this requires teaching them models that do not reflect the way professional economists think. Second, they should think that this self-correction happens through "price flexibility." In other words, what he wants his students to look at, say, falling wages in Greece, and think that the problem must be that they have not fallen enough. That's what "a return to full employment via price flexibility" means. Third, and most relevant for this post, this vision of self-correction-by-prices is directly linked to the idea that money is neutral in the long run -- in other words, that a sustained increase in the money supply must eventually result in a proportionate increase in prices. What Krugman is saying here, in other words, is that a "surprising big" part of his thinking on pedagogy is how to inculcate the exact errors that drive him crazy in policy settings. But that's what happens once you accept that your job as an educator is to produce ideological fables.


  1. Of course, there is another possibility: there was not actually a real increase in the money supply. That the fed put money out there as fast as it could, but it utterly failed to actually enter the economy and effect prices. Actually, I think that is the case. Weirdly, the pattern of inflation we're seeing is incredibly opportunistic and only occurring where demand is relatively inflexible. That is, prices have risen where they could do so; where they couldn't, they didn't. Now, that throws a bunch of monkey wrenches in the prevailing theories, mind you, but not because the Fed actually succeeded in increasing the real money supply.

  2. Right. In the textbook Krugman is very clear that "money" means M1 and is under the control of the central bank. But if you think of money as some broader aggregate that is determined by demand (and perhaps by conditions in the financial system), then things look different: the failure of expansionary policy to produce inflation is inked to its failure to increase the broader money supply.

    The problem with this argument -- as people like Nick Rowe have explained at length -- is that if money is endogenous it's not clear what it means to say that it is "neutral," and that knocks the foundation out from under the argument that the macroeconomy is ultimately self-correcting via price level adjustments.

    Incidentally, this is a very old argument. The claim that money is endogenous was first formulated as "the Law of Reflux" in the debates between the Banking School and the Currency School back in the 1830s. The context was pretty much exactly that of today's discussions: The Currency School believed that the role of the central bank was to set an appropriate quantity of money, while the Banking School believed that the money stock was determined endogenously and that the central bank should instead be trying to regulate the pace of lending.

    1. Also, one request: Please don't comment as "anonymous," use a handle of some kind.

  3. The liquidity trap is a useless concept: if you are outside it you cannot predict that you are going to fall into it, and once you are in it, there is no telling when it will end. So saying that the laws of monetarism work outside of the liquidity trap is equivalent to saying: "these laws work, until they don't, and then they stop working, until they start working again". Very scientific.

  4. To be fair the book was written during 2010 (see the acknowledgements). After blogging throughout 2010-14, he might bring textbook Krugman more in line with blogger policy Krugman.

    Go through the reviewers on the acknowledgments page. Are they all New Keynesians?

    "Liberal Keynesian economists made a deal with the devil decades ago, when they conceded the theoretical high ground."

    And became New Keynesians. What would be the school that didn't concede? The heterodox? Paleo-Keynesians? Disciples of Minsky and Tobin?

    1. After blogging throughout 2010-14, he might bring textbook Krugman more in line with blogger policy Krugman.

      I hope he will. But to be clear, this wouldn't just be a matter of adding "except in a liquidity trap." The whole section on open-economy macro policy would have to be rewritten from the ground up, probably in terms of IS-LM or Mundell-Fleming. On the other hand, blog-Krugman is on record preferring David Romer's version of Mundell-Fleming to the "intertemporal maximization with whipped cream" that he himself serves. So...

      What would be the school that didn't concede? The heterodox? Paleo-Keynesians? Disciples of Minsky and Tobin?

      Well, this is the problem. There are lots of dissents from orthodoxy, but there is no well-defined alternative school.

  5. I liked this post very much, but I liked the footnote more than I can say. Could I ask you to expand it into a post at some point?
    I was a grad student at a Bay Area political science department, and I remember a similar kind of halo existed about free trade and how it had to be promoted by any means necessary. It's not that I questioned it--the reverse, actually, which is why I'm now curious about the ways in which this kind of odd, counter-intuitive orthodoxy is transmitted.
    Or was Larkin all along?
    ". . . Where do these
    Innate assumptions come from? Not from what
    We think truest, or most want to do:
    Those warp tight-shut, like doors. They’re more a style
    Our lives bring with them: habit for a while,
    Suddenly they harden into all we’ve got "

    1. At one point I knew a disturbingly large number of Larkin poems by heart. But I'd forgotten that one. Lovely.

      The question of where orthodoxy comes from and how it reproduces itself -- hard questions. One thing that's striking is how old and how stable this orthodoxy is. I've just been reading Arie Arnon's Monetary Theory and Policy from Hume and Smith -- which I HIGHLY recommend -- and it's remarkable how much today's monetary policy debates are reenacting the debates over Britain's suspension and resumption of gold convertibility during the Napoleonic wars.

  6. Nick Rowe has replied to you:

    Even after 5-plus years of ultra-loose policy with no rising inflation in sight But it hasn't been. Low interest rates are generally a sign that money has been tight, not loose. Monetary policy that will clearly be tightened the moment inflation gets anywhere near going over 2%p.a. is not "ultra-loose".

    1. My post isn't strictly a reply to this post by JW. More of an aside. (I couldn't think of the right word).

      I would reply:

      1. The recent increase in M in the US is definitely not permanent, and the QT applies to permanent increases in M.

      2. Other things are not equal. There is a reason they increased Ms. The QT tells us what happens to P when you increase M, *relative to what would have happened otherwise*.

      But JW's last paragraph is getting at important stuff. Whether economies are "self-correcting" or not, (and whether and how price flexibility works to do the correction), and what we even mean by 'the long run" depends on what it is the central bank is targeting. JW and me are on the same page here.

  7. Hi Nick.

    Thanks for the comments, and my apologies for the slow reply. My excuse is that the I've been taking the kiddo to visit the grandparents on the other coast, and also I'm defending my dissertation in two weeks.

    I take your points to be the following:

    (1) It doesn't matter whether we think the central bank is fixing the money stock or some other nominal variable; the important point is that whatever nominal variable it sets, all other nominal variables should eventually adjust in proportion so that real variables converge to the same level in the long run. So while the classical dichotomy was originally stated in terms of a change in the exogenous money stock, it can apply just as easily to a world where money is endogenous. And

    (2) How quickly and reliably this adjustment takes place, depends which nominal variable the central bank targets. In particular, price flexibility is stabilizing with some targets and destabilizing with others.

    Within the framework of the kind of models we're talking about, I don't have any problem with this. If Krugman-the-columnist in fact believed that an appropriate NGDP target would eliminate the liquidity trap, the textbook would be less objectionable. The problem as I see it is that he doesn't believe that, but his textbook doesn't give any basis for the beliefs he does hold (nor does his scholarly work).

  8. Also, I'm curious why you say "The recent increase in M in the US is definitely not permanent." Can you elaborate?

  9. There's a broader question of whether it's reasonable to summarize monetary policy as setting a nominal variable. An alternative view is that the type and scale of economic activity always depends on the terms on which credit is available, and there is no way to separate out "real" variables from the credit system. But I don't want to get into that here.