Scott Sumner on the Princeton School of Macroeconomics and Overcoming Inflationary Fears

Scott Sumner is the Ralph G. Hawtrey Chair of Monetary Policy at the Mercatus Center and a returning guest to Macro Musings. He joins the podcast today to talk about his ongoing work on the Princeton School of Macroeconomics as well as his thoughts on monetary policy in 2021. Specifically, David and Scott discuss the economic contributions of various different Princeton economists as well as how the central bank can overcome inflationary fears and establish further institutional credibility.

Read the full episode transcript:

Note: While transcripts are lightly edited, they are not rigorously proofed for accuracy. If you notice an error, please reach out to [email protected].

David Beckworth:  Scott, welcome back to the show.

Scott Sumner:  Thank you for inviting me, David. Good to be here.

Beckworth:  Glad to get you back on the show, Scott. It's always a delight to discuss monetary policy with you. It's worth noting, since some of our listeners may not know this, but you, Scott, are a key reason we have this podcast. You started the Monetary Policy Program at the Mercatus Center, and this podcast is an outgrowth of that program. So, thanks to you, Scott, we have this podcast which has been going on four years now.

Beckworth:  Now, before we get going, I do want to briefly mention that we are recording this show on Wednesday, January 6, the day that Congress was supposed to certify the results of the presidential election as part of an ordinary transition of power between administrations, but as listeners know, today turned out to be anything but ordinary with the groundswell of Trump supporters charging the Capitol building. And, it was dramatic, some are calling it sedition. But hopefully, this will be the last of such antics, so that when we run this show, on Monday, January 11th, our country will be back on track for a normal transition of power between the Trump administration and the Biden administration. So, that this discussion, these topics today will be relevant.

Beckworth:  So, with that said, Scott, let's move into a project you're working on. This project centers around the Princeton School of Macroeconomics. Sounds very fascinating. So, why don't you tell us who the school is? Who are the key contributors? When were they around? And, what's their ultimate contribution to macroeconomic thought?

The Princeton School of Macroeconomics and Its Impact on Monetary Policy

Sumner:  Okay, so in the early 2000s, there were five people, at least, at Princeton, who, in my view, developed what we now view as the standard view of monetary policy in the 21st century, that is policy when interest rates are really low. And, these were, Paul Krugman, Ben Bernanke, Michael Woodford, Gauti Eggertsson, and Lars Svensson. And, they all did really interesting work that I think has, and I'll talk later about how it's shaped, both, how we view policy, and actually, perhaps, policy itself to some extent. So, I don't think people have quite connected the dots. And, that's what I hope to do with this project.

Beckworth:  Yeah, it is interesting to think that all these individuals came together, and they were all working on a very similar topic at the time. As we'll get into it, we'll talk more about this, but Japan was the primary driving force, as I understand it, and they all collaborated. And, I think, you make the argument that what we see now in terms of the average inflation target at the Fed, level targeting in general, the idea of more credible forward guidance through makeup policy. The fact that it's so widely understood, and maybe accepted, is because of their work. Is that fair?

Sumner:  Yeah, I think that's pretty accurate. And, for me, the key paper here is Paul Krugman's 1998 paper where he looked at the liquidity trap in Japan. And, what I'd like to do is spend a couple of minutes giving you an analogy for, I think, what Krugman did, because I believe the paper is underestimated and misunderstood to some extent. In fact, I view it as one of the most important macro papers of the last 40 years. So, the analogy I would use is, if you look back at Ronald Coase's paper on social costs way back in 1960, Coase basically dismantled the traditional Pigou theory of externalities, but he didn't really stop there, he rebuilt it on stronger foundations.

Sumner:  So, what Coase did, is he focused on transactions costs as being the key problem, not externalities, per se. And, what Krugman did is, he dismantled Keynesian liquidity trap theory and rebuilt the theory sort, as an expectations trap, if you will. So, the problem in Krugman's view, was not just that money and bonds are equivalent at zero interest rates. The real problem is sort of bearish expectations of future monetary policy. And, I think both papers have been somewhat misunderstood, because each paper can be interpreted in two different ways, what you might call a right wing interpretation and left wing interpretation.

Sumner:  So, in the case of Coase's paper, the more conservative interpretation is that you can have negotiated solutions to externality problems and you don't necessarily need government regulation. The more left wing interpretation is that, in the real world, there's a lot of transactions costs. So, you do need Pigouvian types of regulations to correct externalities. So, if we use that analogy, then I think the same thing can be said about Krugman's paper. One interpretation is that monetary policy still works at zero interest rates. The central bank just needs to commit to future inflation. So, you don't need fiscal stimulus. Krugman was actually surprisingly dismissive of fiscal stimulus in Japan during the late 90s.

Sumner:  The other interpretation, you might call it more progressive interpretation, is that central bank promises to inflate are not likely to be credible to inflate in the future. So, you need Keynesian fiscal policy, not so much because monetary policy doesn't work, but that a policy that would work wouldn't be credible, because central bankers are viewed as too conservative. So, in each case, with the Coase paper and the Krugman paper, there's two interpretations. And, they really reflect the fact that you had an original policy view that was conventional. And, each of these individuals showed that, that original policy view might be correct. But, you needed a special case for it to be correct.

Sumner:  In Coase's case, you needed there to be high transactions costs. And, in Krugman's case, what you needed is, that central banks couldn't credibly commit to future higher inflation. And, only then did you get the traditional policy assumptions being correct. Does that make sense?

Beckworth:  Yeah, it does. That's interesting. And, just for our listeners, the title of this 1998 Krugman paper is, *It's Baaack: Japan's Slump and the Return of the Liquidity Trap.* It was a Brookings paper. I stretched out back with, ‘it's back’, because he literally spells the word back with... He had three A's. I think it was from Poltergeist, the movie... Some movie where someone says that when the scary monsters come back. So, the concern was deflation was returning. And, again, this is motivated by Japan's experience. I like the way you summarize it. There's two ways you can look at it. One is, a central bank can work at a zero lower bound, it just has to commit to future inflation or a future permanent increase in the money supply, monetary base. Or, the alternative take is that, a central bank can't credibly commit for whatever reasons, and it needs fiscal policy to help it makes its policies credible moving forward. Where do you think Paul Krugman comes down on that divide?

Sumner:  Well, that's a very interesting question. And, one reason, by the way, I think the paper’s under-appreciated is that many people only see one of the two interpretations. And, I think what Krugman... The beauty of it is that, he saw that there are actually two possibilities. And, over the years, I think his view shifted somewhat maybe towards what you might call the more left wing view, more pro-fiscal policy. And, that was probably partly because Krugman became disappointed in the fact that central banks seemed kind of stubborn, and unwilling to do what he called promising to be irresponsible. That is promising to push inflation higher than the normal level once you've exited the liquidity trap.

Sumner:  Now, ironically, the Fed is just doing that now. And, last year, in 2020, they did issue a promise to make up for an undershoot of inflation with higher inflation in future years. So, I think with this, it's important to keep in mind that there really are two cases here. One paper, I would point to with Krugman is, a paper he wrote 20 years, later looking back on it. I think he's pretty proud of this paper. And, I think deservedly so. I'll talk about some of the things that are under-appreciated in a moment in the paper. But, one of the things I noticed is that, he really presents both cases, in a sense, in this look back paper from 2018.

Sumner:  The first was, in the early 2000s, where the Japanese Central bank, after injecting money in the economy, pulled a lot of it back out in 2006. And, this is exactly what Krugman warned against that if you just have a temporary currency injection, you inject a lot of money, but it's not permanent, it won't have any significant inflationary effect at the zero bound. So, QE is relatively ineffective unless you can somehow commit to making it permanent, at least to some extent. So, that's supported his pessimistic view about monetary policy. But, in the same paper, he later talks about what happened after Abe was elected and the central bank President in Japan, Kuroda, instituted a higher inflation target, and did a lot of other changes to boost inflation expectations, set a higher inflation target and move very aggressively.

This is exactly what Krugman warned against that if you just have a temporary currency injection, you inject a lot of money, but it's not permanent, it won't have any significant inflationary effect at the zero bound. So, QE is relatively ineffective unless you can somehow commit to making it permanent, at least to some extent.

Sumner:  And, Krugman pointed out that, that was sort of a regime change. And, it did actually improve the Japanese economy. They didn't quite hit their 2% inflation target. But, nominal GDP growth definitely picked up fairly dramatically after 2013. And, Krugman attributes that to the monetary stimulus. So, that's a little bit more of a conservative interpretation, because Krugman, himself, points out that Abe actually had a relatively conservative fiscal policy, they didn't do fiscal stimulus. They relied almost entirely on monetary stimulus. And, so essentially, under Abe's government, the Japanese government did exactly the opposite of what MMTers would recommend. They use monetary policy rather than fiscal policy. And, Krugman points out, it was successful to some extent.

Sumner:  So, even in this paper 20 years later, that looks back and it's written at a time where I think Paul Krugman is definitely more Keynesian and more skeptical of the efficacy of monetary policy, he could still point to both types of cases where monetary policy didn't work for the reasons he outlined, and where it did work for the reasons he outlined in the earlier paper.

Beckworth:  Okay, let's go back and talk about that evolution of his thought, and you attribute it to what he saw in practice with central banks, right? They tended to be more timid. And, even in the case of the Bank of Japan, like you said, they never really got that far above, or even close to 2%. On, at least, in a persistent basis, they never hit 2%. And, all across the advanced economies, the ECB, the Fed, the Bank of Japan, the Bank of Switzerland, the central bank there, they all have persistently undershot... And, you can tell stories as to why they did. But, this comes back in my mind to this question of being able to do what Paul Krugman prescribed, and that is credibly commit to being irresponsible.

Beckworth:  And, it seems none of the central banks are willing to do that. And, so this is a question we've discussed before, Scott, but it's worth fleshing out here on the podcast one more time. How do you make a central bank credible? And, again, it's not just one central bank that seems to have a problem with this. It's many of them, right? It's almost all the advanced economies, central banks. And, maybe this is more of a political economy question. Maybe, they don't feel they have the authority. Or, maybe, there needs to be a new regime. Paul Krugman said a new regime was needed. And, Christine Romer also had a paper on, Japan started a new regime. But, one could argue it really wasn't that much of a different regime, in retrospect, maybe a little bit better. So, how do you get a central bank to be truly credible, so that it can permanently increase or reflate the economy?

Establishing Central Bank Credibility

Sumner:  Okay, well, I think, first of all, I'll say that I'm a little bit older than you, and maybe a little more optimistic, because I can remember the reverse problem, that when nobody believed the central banks were able to control inflation. And, I've lived through that transition, where we first figured out how to do it academically, and eventually central bankers came around and implemented effective things like the Taylor Rule approach. So, I just think this takes time. And, you have to remember that the Fed’s decision last year to adopt average inflation targeting, is really the first step in that direction. And, obviously, it's too soon to know whether they're going to carry through with it.

Sumner:  But, it is certainly a step in the right direction, even better would have been, price level targeting. And, even better than that, of course, nominal GDP level targeting. But, even average inflation targeting should help address this problem if the Fed is serious about it. And, I think the fact that they've publicly committed to average inflation targeting, does put some pressure on the Fed to actually carry through. In other words, if that regime had been in effect, back in the 2010s, the Fed would not have raised interest rates frequently between 2015 and 2018. So, you can actually see how the policy would have been different during that stretch of time under average inflation targeting.

Even average inflation targeting should help address this problem if the Fed is serious about it. And, I think the fact that they've publicly committed to average inflation targeting, does put some pressure on the Fed to actually carry through. In other words, if that regime had been in effect, back in the 2010s, the Fed would not have raised interest rates frequently between 2015 and 2018.

Sumner:  So, I'm a little bit more optimistic than some on that, although I can certainly understand that, especially what's going on in Europe and Japan, would give people reason to be pessimistic as well. But, I think we'll just have to see how this plays out in the US over the next decade. We should know within a few years, how serious the Fed is about actually implementing its new regime. But, I'd like to point out that this is, a mild or modest version of exactly what Paul Krugman is recommending. Make up inflation for the undershoot you get in the liquidity trap itself.

Beckworth:  Yeah, absolutely. And, we'll come back and talk about the average inflation target of the Federal Reserve a little bit later in the show when we talk about the outlook for US monetary policy in 2021. But, I do want to come back to Krugman’s paper. It's something I've joked about before, because Krugman loves to say, “It’s all in my '98 paper.” He loves that reference and it is. It's a great paper. And, and again, it I think it's the beginning of some other papers from Princeton that we'll get to in a minute.

Beckworth:  But, the one thing I want to flesh out from his paper that I think is underappreciated, and maybe this is one of the points you're going to make, is this idea of a permanent increase in government liabilities, or permanent increase in the monetary base precisely. And, he goes on to say, that's what this commitment to being irresponsible is. And, he really stresses that you've got to commit and create the expectation that you'll stick to it, that you need to increase the monetary base, and never reverse it. I would qualify that a step further. And, I think it's implicit in his work and others work. It's not just a permanent increase in the monetary base, it's a permanent increase in the monetary base, above and beyond what the market demands, and is expected to be demanded.

Beckworth:  So, you want to actually increase the monetary base beyond what banks want to hold, beyond the currency holdings of the public, such that there is this excess supply of monetary base. And, what's interesting, if you take this argument and you think through it, it's a quantity theory of money argument underneath the hood. Is that fair?

Sumner:  Well, I guess, here's how I would put it, I think that Krugman, himself, doesn't think in terms of Monetarist terms in terms of the transmission mechanism. So, I believe what he was actually doing in that paper was saying, look, if you're a monetarist, and you're claiming that printing money is the solution in the liquidity trap, here's the problem you run into. If you print a lot of money, and there's no commitment for future inflation, people are going to expect that the monetary injections are not permanent. So, it wasn't so much that Krugman was actually arguing for targeting the monetary base or doing something specific with it, he was using it as a rebuttal to monetarists who claimed that an easy way out of a liquidity trap is simply to inject a lot of money through QE.

Sumner:  So, that's, I think the way that Krugman would probably frame the claim. But, he certainly did talk about the monetary base again, when the Japanese reduced the base in 2006. He cited that as an example, I think, correctly, as showing what he was talking about in the paper, exactly the problem that he was worried about. But, his basic view of the transmission mechanism of monetary policy is more Keynesian. He focuses on the need to get real interest rates down to a low level through setting a higher inflation target. He talked about 4% inflation over 15 years in Japan, for instance, as a possible goal based on his estimates of the output gap there. So, he does have a more Keynesian view in terms of his own policy preferences. But yeah, he does definitely have a point of view on why base injections won't work unless they are view as permanent.

Beckworth:  Yeah, and I want to be clear, I'm not claiming he's invoking old Monetarism by any means. He definitely has a traditional New Keynesian perspective on getting the real interest rate down to its equilibrium or natural rate level, by raising inflation expectations, and thus, doing it appropriately long enough, you'll get the output gap clear, and you'll be back to a healthy recovery. However, he shows that it's equivalent to him. It's a better way of saying that. It's equivalent to a permanent increase in the monetary base.

Beckworth:  Let me quote here from his 1998 paper. He says, "A monetary expansion that the market expects to be sustained, that is matched by equal proportional expansions in all future periods, will always work. Whatever structural problems the economy might have, if monetary expansion does not work, if there's a liquidity trap, it must be because the public does not expect it to be sustained." So, it's got to be permanent... And, again, I think it's important to stress, it's got to be permanent and greater than what's needed by the economy. Needed by banks in terms of bank reserves and more than the public wants to hold in terms of currency in circulation. So, it's got to be this almost exogenous increase in the monetary base.

Beckworth:  Now, that's hard to do. And, how would you do that, it's a whole different discussion. You could think of, maybe, a helicopter drop as being an example of that, or some kind of deficit, that's money financed. Some way of getting that permanent increase. There's a commitment not to reverse it in the future. So, you're not going to have tax increases. One could argue, you're going to lower your primary surpluses. You're going to do whatever it is to keep that permanently in circulation.

Beckworth:  And, the reason I want to stress this, Scott, and flesh this out is, how do we take that idea and make sense of the fact that the Federal Reserve, the ECB, Bank of Japan, Swiss National Bank, they all have expanded their balance sheets dramatically since 2008. And, one could look at that and say, it looks like a permanent increase in the central bank balance sheets, which implies a permanent increase in the monetary base. Yet, we haven't seen the story Krugman has talked about. So, how would you reconcile what appears to be a ratcheting up of central bank balance sheets. It's been over 10 years, and, yet, you don't see this big recovery that Krugman says is possible in his paper.

Sumner:  Right, so I think at that time, you have to remember this is 1998, when the Japanese liquidity trap was viewed as a quirky oddity. And, two things have changed since then, one is, we have countries that seem to be permanently stuck at zero interest rates. And, we also have the payment of interest on bank reserves, which has greatly increased the demand for base money. So, both of those things have led to permanently enlarged central bank balance sheets. And, that's made things more difficult for central banks than before.

Sumner:  Now, I still think that monetary injections that are permanent, do have an expansionary effect. But, I think in that kind of environment, it's a mistake to think in terms of, oh, the monetary base needs to be at this or that level. What you really want to do is, have a price level or a nominal GDP level target. And, then basically, what you're doing is you're committing to inject as much base money as necessary to achieve that target.

What you really want to do is, have a price level or a nominal GDP level target. And, then basically, what you're doing is you're committing to inject as much base money as necessary to achieve that target.

Sumner:  Right. So, you'd have to have sort of, commitment to do whatever it takes to hit a price level target in the future. It's a little bit easier in the United States than in Europe and Japan, because there's more of a perception here that we would perhaps occasionally exit the liquidity trap and interest rates would rise above zero. But, even there, if they're really committed enough, I think they could probably do that. Now, Krugman might disagree with that point, given the way that things have changed recently.

Sumner:  But, the other thing I'd like to point out that... It's certainly true that this notion that monetary injections need to be permanent, has been around for a while. You can find this in earlier Monetarists’ writings, as well. They would talk about thought experiments permanently doubling the money supply. But, what Krugman did is, he took this basic idea and saw how it was essential to understanding the liquidity trap. And, to just give you a few examples of things that were, I think, well ahead of his time, when he wrote the paper in 1998. At that time, people thought this was a problem associated with the banking system in Japan, just like in the Great Depression in the US in the 30s, our banking system was in trouble and we were stuck at zero rates.

Sumner:  And, Krugman showed that you can have a liquidity trap, even if the banking system is doing fine. And, that's been proven to be the case in years since. He showed you can have a liquidity trap at full employment, and Japan was basically at full employment in 2019. And, they were in a liquidity trap. That was also a pretty bold claim when he made that in 1998. He pointed out that Europe could likely be next and it was, of course. He talked about the Great Depression in the US. And, he focused on Romer’s view that, what's really needed for recovery was negative real interest rates and the only way to get that was positive inflation expectations.

Sumner:  Now, one thing I think he missed there that would have been useful, is to discuss the role of dollar devaluation and that's where Gauti Eggertsson's 2008 paper really came in. He took the basic idea in Krugman's paper and saw that FDR's dollar devaluation In 1933, was a way to credibly commit to higher inflation over the next couple of decades. And, that's exactly what happened.

Beckworth:  Well, Scott, before we move on to Gauti Eggertsson's paper or Woodford's paper and the others, just to circle back and make sure we've covered it all. So, in Krugman paper... Again, I don't want to beat this horse too much. But, the two takeaways I get, at least, from this in terms of why this idea of permanency is so consequential, it has to be permanent in terms of being expected to be permanent going forward, it can't be reversed. But also, it has to be greater than the demand for monetary base.

Beckworth:  So, you mentioned is, when you're at the zero lower bound, that increases the demand for the base and also interests in excess reserves, increases the demand for the monetary base. So, you can reconcile Krugman views with what we've seen over the past decade. So, going back to my original question, the central bank balance sheets have expanded in part because the demand for their liabilities has expanded, given the zero lower bound environment, given the interest in excess reserves. I think one application of Krugman's argument is, that they need to do even more. They would have to expand their balance sheets even more above and beyond what banks want to hold in terms of bank reserves and what the public wants to hold in terms of currency.

Beckworth: Okay, Scott, so I think that's a good summary of Paul Krugman's 1998 paper. Again, the title was, *It's Baaack: Japan Slump and the Return of the Liquidity Trap.* Any final thoughts on that paper before we segue into some of the other Princeton macro papers from that period?

Sumner:  Well, I would just add that I think the issues you talked about recently with the greatly enlarged demand for base money, large central bank balance sheets, have made Krugman somewhat more pessimistic about monetary policy recently. And, that's, I think, why he's moved more towards fiscal stimulus as a solution. But again, I would recommend that people go back and look at this paper and also read the comments written by a number of prominent economists at the end of the paper in the Brookings volume. And, I think what you'll find is that, Krugman's interpretation holds up really well. And, in my view, much better than a lot of the commenters who criticize various aspects of what he was doing. So, I still view this as the key paper that underlies everything that came after that in terms of the Princeton School.

Beckworth:  Okay, one last thought, before we move on, on this paper, just to reiterate this point. Again, Krugman is not advocating a monetary base target by any means. In fact, I think, the way could look at this is, as he's advocating more of a price level target, or higher inflation, but that in turn implies a certain path for the monetary base. So, you can look at this through the lens of a level target, you can look at through the lens of some kind of permanent increase in the monetary base, either way, they're telling the same story, it just depends where you want to put your emphasis. All right, let's move on to the next paper and I believe that's Michael Woodford and Gauti Eggertsson's 2003 paper, is that right?

The Work of Woodford and Eggertsson and Its Implications

Sumner:  Right. And, I haven't had as much time to really focus on that. So, I won't say much on that, other than that, it takes some of the ideas in Krugman's paper and develops it into a more fleshed out New Keynesian model with multiperiod and all the bells and whistles. But, it still ends up with moving towards a recommendation of something like price level targeting or something where you're making up for past undershoots of inflation. In fact, maybe even stronger than a price level target where you move the price level above trend once you exit the liquidity trap.

Sumner:  So, I think it was basically continuing what Krugman was doing. I know a little bit more about Gauti Eggertsson's work in 2008, looking at the Great Depression, because my own research was also somewhat on that topic. And, one of the interesting things about that is that, Krugman was recommending 4% inflation over 15 years. So, if you compound that, that's about an 80% increase in the price level for Japan. And, in some ways, that's actually easier to do under a gold standard than under a fiat money system.

Sumner:  Because, in a fiat money system, you've got a real credibility problem, a commitment problem, if you will. With a gold standard, you have this sort of, exogenous device, you can just arbitrarily change the price of gold. So, what FDR did is, he raised the price of gold 69%, just between '33 and '34. And, what that did is, it created expectations of a much higher future price level.

Sumner:  So, it was kind of, an interesting commitment device. And, sure enough, prices did start rising, and over the next 15 years, rose even more than 69%. So, what that effectively did because interest rates were stuck, basically, close to zero is, they probably depressed real interest rates. They created inflation expectations. And, that stimulated the recovery, or the recovery began, at least, in 1933, right about the time of the devaluation. So, that's, I thought, a kind of, an interesting application of that basic approach.

Beckworth:  Yeah, so the work done by Michael Woodford, Gauti Eggertsson, you mentioned his paper on the Great Depression as an application of this work. Extension, kind of, a key part of this Princeton School of Economics or School of Macroeconomics. And, if you read their subsequent work... So, they had other papers. They had their paper together at the Brookings paper in 2003, I believe, but a number of other papers they put out since then, it's a similar vein. And, I'll just mention here, Michael Woodford's, big paper in 2012, the Jackson Hole meeting, where he reiterates these comments. Again, he makes the case in terms of a level target, makeup, or what the New Keynesian might say, truly credible forward guidance that comes from a level target.

Beckworth:  But, he also says, this is equivalent to a permanent increase in the monetary base. And, that 2012 paper, he also mentions, Scott, the case of Japan, like you did, the 2001, 2006 QE, where they reversed it. And, he says, it's as if the Bank of Japan kind of, confirm the expectations of the public that this wouldn't be a permanent increase. Now, I want to mention, we've had Gauti Eggertsson on the show, so listeners can go out and check out his episode, we'll provide a link to it. Also, worth mentioning, though, Scott, Gauti was a graduate student during part of this time, correct. He studied under Michael Woodford. Studied under all of them. And, I want to just highlight that because he has a paper from 2016, where he recounts an experience from the department during this time. I want to get your take on this.

Beckworth:  But, the paper’s title is, *Bernanke's No-arbitrage Argument Revisited: Can Open Market Operations and Real Assets Eliminate the Liquidity Trap.* And, let me just read a couple of paragraphs from this paper. I won't spend too much time on this. But, he gets into an interesting discussion as a graduate student with Ben Bernanke. He was the chair of the department back then. So, Gauti Eggertsson says this, "From the perspective of the author of this paper, however, the most pertinent statement about the academic consensus at the turn of the century, came up in a conversation with Ben Bernanke, then chairman, not of the Federal Reserve, but the Princeton economics department and editor of the American Economic Review. When proposing the liquidity trap as a PhD dissertation"... This is Gauti talking... "Bernanke issued the following warning, ‘I have to warn you, I do not believe in the liquidity trap.’”

Beckworth:  And, then he goes on to tell why Bernanke doesn't believe in this. He quotes from a speech where he brings up this idea that the central bank could buy up every last real asset on the planet and get rid of a liquidity trap. So, the argument is, look, if you really think the path of the price level and monetary policy are separate, then there's a free lunch to be had. Right? You can just go up and start buying up assets. Anyways, Gauti Eggertsson responds to that argument in this paper. But, I'm interested to hear your thoughts on that argument.

Sumner:  I think Bernanke is correct, but in a way that... I would emphasize a slightly different point. Because, I think critics of Bernanke might say, well, if you buy up everything that's really fiscal policy, which I don't think is quite correct, but I can see why people would say that. But, I think that thought experiment is, the implications are not exactly what it seems at first glance. The actual implication is, if you committed to do whatever it takes, to buy up as much as necessary, you might not have to buy very much at all, in fact, you might end up selling.

Sumner:  So, let's do a different thought experiment to drive this point home. Suppose, the Bank of Japan were to raise their inflation target to 10%, not 2%, 10%, and, commit to do whatever it takes to hit that target. Now, at first glance, you might think, well, wow, just imagine all the heavy lifting they'd have to do. They're not even able to hit 2%. Imagine how much harder to hit 10%. But, if you would kind of, reverse it, and ask yourself, well, what would success look like? If they really were willing to do whatever it takes, and the commitment was credible, and people expected 10% inflation, how much base money would people in Japan want to hold? And, the answer is going to be, something like less than 10% of GDP would be held as cash and bank reserves. Now, it's over 100% of GDP.

Sumner:  So, in fact, if Japan were actually able to credibly commit to 10% inflation, and commitment to do whatever it takes, they would have to do reverse QE, dramatically reduce the money supply... And, this gets into the whole Neo-Fisherian thing we can talk about later, they would have to raise nominal interest rates, not cut them to implement that kind of expansionary monetary policy. Now, of course, 10% is kind of far-fetched. But, even for smaller increases in inflation, I think sometimes people forget, when contemplating what central banks would have to do, that if the policies actually credible, and you do get somewhat higher inflation and higher nominal GDP growth, then the so-called natural rate of interest will be higher. And, you actually won't have to do as much as it might appear. And, importantly, you won't have to do as much as central banks in places like Japan and Switzerland, do as purely defensive maneuvers. Countries that have failed to credibly commit to higher inflation. And, as a result, there's an enormous demand for their currency, right. And, they're accommodating that demand with a lot of QE to avoid outright depression and deflation.

I think sometimes people forget, when contemplating what central banks would have to do, that if the policies actually credible, and you do get somewhat higher inflation and higher nominal GDP growth, then the so-called natural rate of interest will be higher. And, you actually won't have to do as much as it might appear. And, importantly, you won't have to do as much as central banks in places like Japan and Switzerland, do as purely defensive maneuvers.

Sumner:  So, I think people kind of, mix up those two cases. One type of QE is basically defensive. And, you see very large central bank balance sheets in countries with the lowest inflation rates in the world. Another type of QE is inflationary. And, that's what we've seen in Latin America and places like that, where you have rapid increase in money supply and high inflation over time. So, all of these things need to be looked at in context, and it's very dangerous to estimate how much you'd have to do to hit a particular target, without thinking about whether your commitment to hit that target is credible, because the answer, in terms of how much you have to do, very much depends on how the public views the credibility of your action.

Beckworth:  That brings us back to credibility, right? I think that's the big hang up. That is the big hang up in our conversation. What does it mean to be credible, right? Why is it that central banks seem to have a hard time with credibility? Can they commit on their own to being credible? Again, Scott, it's hard not to look around the world and see central banks be averse to having inflation overshoot a little bit. Even the Fed, this year, with average inflation targeting... Getting ahead of myself here, but if you look at forecasts like breakevens and such, you don't see massive... At least, the market yet doesn't think there's going to be a huge amount of overshooting taking place.

All of these things need to be looked at in context, and it's very dangerous to estimate how much you'd have to do to hit a particular target, without thinking about whether your commitment to hit that target is credible, because the answer, in terms of how much you have to do, very much depends on how the public views the credibility of your action.

Beckworth:  I think back to the Great Depression as another example. In the mid-1930s, right? So, they've gone through the worst part of the great contraction, there's still a huge hole in the economy. In 1935, '36, they begin to worry about inflation, right? The price level's fallen, I think, 30%, at the bottom. So, the Fed gets worried about inflation going a little bit above normal. And, they completely missed the big decline in the price level relative to its pre-crisis trend. I think, in terms of growth rates, in terms of levels... And, maybe, that's in our blood, maybe that's just some kind of cognitive bias we have, but what makes you hopeful, Scott, that they can overcome this?

Overcoming Inflationary Fears

Sumner:  I don't think so, because, first of all, we've had several cycles here. We had the Great Depression, we had the Great Inflation, and, for several decades, everybody believed or almost everybody believed that central banks couldn't stop the inflation because they didn't have the self-control or discipline, or politically, it was too difficult or whatever. And, then they basically got inflation down to around 2% in the 90s, and kept it there. Recently, we've been undershooting, and there's a lot of the pessimism that you describe.

Sumner:  But, I would say two things about credibility. There's really two questions that get raised here. One is, can we get central banks to sincerely want more to hit their inflation target?  The second question is, if central banks are sincere, will the public believe them? The time inconsistency problem. I don't think the time inconsistency problem is the actual problem. I think the actual problem is, central banks just have to want to do the right thing, basically. And, if you look at all the failures we've seen over the last two decades, none of them can possibly be attributed to central banks being well-meaning but unable to convince the financial markets.

If you look at all the failures we've seen over the last two decades, none of them can possibly be attributed to central banks being well-meaning but unable to convince the financial markets.

Sumner:  If we look at Japan, they raised interest rates in 2000, they raised interest rates in 2006. This is not a central bank trying to inflate, and there was no inflation at the time. If we look at the European Central Bank, all the way through the Great Recession from 2008 to 2013, interest rates in Europe were above the zero bound, often well above the zero bound. The ECB was not out of ammunition, they just didn't want a more stimulative policy.

Sumner:  Then, you come to the United States, the Fed raised interest rates nine times between 2015 and 2018. And, during that period, you could pick up any newspaper, a financial paper, and see articles claiming that the Fed was powerless to get inflation up to 2%. A period of time when they raised rates nine times. The analogy I would use is like someone buying a car and saying I can't get this car up to 60 miles an hour, I keep tapping the brake, but it won't go up to 60 miles an hour. All nine of those interest rate increases were designed to lower inflation. That's why the Fed increases its target.

Sumner:  So, by the end of that period, I was pulling my hair out reading these articles, the Fed just can't hit its 2% inflation target when they were raising interest rates, nine different occasions. So, there has to be the commitment. One reason I'm a little bit optimistic is, I sense from some of Jay Powell statements that he sees the mistake that was made during the 2010s, and he's more reluctant now to increase interest rates until there's a sign of hitting inflation on sort of, an average basis. Not just at the moment, like they did briefly in 2018, but over an extended period of time. And, that's the idea behind the average inflation target.

Sumner:  So, if the Fed actually does what it says it's going to do with that average inflation target, I think the conduct of monetary policy should improve. We should get a little bit closer to what Krugman was recommending back in 1998. But, we'll see.

If the Fed actually does what it says it's going to do with that average inflation target, I think the conduct of monetary policy should improve. We should get a little bit closer to what Krugman was recommending back in 1998.

Beckworth:  Yeah, we will. I agree with a lot of that. I completely agree, the Fed could have done a lot more over the past decade. It made a choice to raise interest rates nine times, as you mentioned, and I agree completely, inflation would have been higher, had they not done that. I guess the question I'm asking is, well, why did they do that? And, why does the ECB do what it does? Why does the Bank of Japan? Why is there this tendency, not just in the US, but across the world... Is it a generational thing? Is it the, fighting the last war of the 1970s? Why have we got to this point? If the Fed, and other central banks, have the ability and they're not using it... Maybe that's the point of us talking today, this podcast, our work at the Mercatus Center, is to get them over that bridge of troubled waters where they're worried about inflation taking off excessively.

Sumner:  It is partly a generational thing. Each generation makes the reverse mistakes of the previous one, going all the way back to the Great Depression. And, I think Krugman used the term focal point or something for the 2% inflation target. In one of his papers, he talked about how, when we first thought about what should the inflation target be, we thought about the costs and benefits of inflation. And, one of the benefits of a little bit of inflation was, you want it high enough to avoid a liquidity trap. Okay. So, we thought 2% was enough to avoid a liquidity trap and not enough to cause major menu costs, etc.

Sumner:  And, then when we found that there was a secular decline in the equilibrium real interest rate, logically, we should have reevaluated what would be the appropriate inflation target with that new information, right? That would be the logical thing to do. But, central bankers had focused so much on 2% inflation. I think Krugman used the analogy, the number became like the gold standard, something they couldn't break away from. And, I think that's right, the commitment was so strong, that psychologically, it's tough to kind of reorient everything you've been focused on in your academic career, in the early part of your career as a central banker and so on, and see that the actual problem is something different. But, enough time goes by a new generation comes along, and I'm pretty sure a lot of the younger people you know, probably, roll their eyes at older economists that worry that inflation is just around the corner, right? Well, eventually, they'll be in power and they'll be able to implement… Maybe, they'll push things so far the other way, we get the opposite and say, who knows?

Sumner:  Nothing is permanent. But, these changes take place over a timescale of decades. One reason I'm optimistic is, I think, I'm starting to see some evidence that things are kind of moving a little bit in that direction. I happen to think that if we hadn't been hit by COVID, we were actually making a lot of improvements in the monetary policy, and that we were on track for a fairly long expansion. So, I'm pretty optimistic that when we come out of this, we might be in pretty good shape for the rest of the decade. But, that may be famous last words too.

I happen to think that if we hadn't been hit by COVID, we were actually making a lot of improvements in the monetary policy, and that we were on track for a fairly long expansion. So, I'm pretty optimistic that when we come out of this, we might be in pretty good shape for the rest of the decade.

Beckworth:  To the extent that does happen, that is, to the extent that the Fed does do some kind of, makeup policy, implicitly some kind of quasi-level targeting, through its new average inflation targeting, we can thank this Princeton School of Macro, and we need to get back to that, Scott, and finish our conversation. There's two other individuals in that group. We touched on one of them, it's Ben Bernanke. But, I know that he has some papers that you want to speak to directly, so why don't we talk about him now and his contribution to this school of thought?

Ben Bernanke’s Contributions to the Princeton School of Thought

Sumner:  The Bernanke paper I'd like to talk about is a 1999 working paper he did on the situation in Japan, I think he called it, *A Case of Self-Induced Paralysis,* where he diagnosed the problem in terms of how the Bank of Japan was thinking about the problem in the wrong way, and was misdiagnosing the problem and had convinced itself that there was nothing they could do. And, one of the things that's really interesting about this paper is that, it's also an excellent description of what went wrong in the United States and Europe in the Great Recession, which occurred a decade later than the papers.

Sumner:  So, he was criticizing the Bank of Japan, and yet, the United States, and even more so Europe, made some of the same mistakes. So, now the cynics will, of course, pile on Bernanke and say, well see... It was easy to criticize others, but when you're put in charge of the Federal Reserve, you find out that things aren't exactly as you thought. But, I think those critics missed two points. So, I'll make two points in defense of Bernanke here. One point is that, the problems in the United States were not so much technical as political. So, to the best of my knowledge, when Bernanke got on board at the Fed, he did discuss ideas like price level targeting, but there was a lot of institutional resistance of the Fed to that.

Sumner:  And, there's also a lot of worry about taking on too much risk on the central bank's balance sheet. And, he had sort of dismissed that problem with Japan correctly, because it's actually not risky for the Fed to buy treasury bonds. The Fed is part of the federal government as a whole, it's all one consolidated balance sheet in a sense. So, any losses to the Fed are offset by gains to the Treasury. But, the Fed doesn't look at things that way. So, the Fed is actually, as an institution, much more conservative than what Bernanke recommended to the Japanese. So, one problem Bernanke ran into at the Fed, is just institutional resistance to being too aggressive in dealing with the zero bound problem during the Great Recession.

One problem Bernanke ran into at the Fed, is just institutional resistance to being too aggressive in dealing with the zero bound problem during the Great Recession.

Sumner:  The second point I'd make in defense of Bernanke is that, the Fed was relatively aggressive compared to either Europe or Japan, and the US did much better than Europe, especially, even though the so-called trigger for the great recession was in the United States with the subprime crash, the banking crisis, the real estate crisis here, the recession ended up being much worse than Europe, which people did not anticipate in 2008, people thought it was going to be worse in the United States.

Sumner:  And, you can very clearly see why it was worse in Europe. The central bank in Europe, the ECB, was much more restrictive in its monetary policy. They kept interest rates well above zero during that period. They didn't go in, initially, for the QE. And, then only when they had the double dip recession around 2011 to '13, did they realize they'd made a mistake, and then they started doing things that looked very aggressive, but were essentially defensive moves, the QE, the negative interest rates, all of that.

Sumner:  And, the irony here is that the ECB had sort of, I think, some ECB officials had thought Bernanke was being too aggressive with the Fed in the early parts of the Great Recession, being too stimulative. And, in the end, it got so bad in Europe that they had to do things that were even more radical than what Bernanke would have liked the negative interest rates and so on. So, that shows you the long run, short run thing. A monetary policy that looks highly expansionary, like Europe in recent years, can actually be the long run response to an earlier contractionary policy that drove nominal GDP growth to really low levels. So, although Bernanke wasn't able to do maybe exactly what he would have liked, and the Fed certainly made some of the same mistakes as Japan made in the 90s, I think that, in a relative sense, we were probably still better off having him at the Fed then someone else in that situation, certainly, compared to the situation in Europe.

Although Bernanke wasn't able to do maybe exactly what he would have liked, and the Fed certainly made some of the same mistakes as Japan made in the 90s, I think that, in a relative sense, we were probably still better off having him at the Fed then someone else in that situation, certainly, compared to the situation in Europe.

Sumner:  So, I think that's a very interesting paper from the point of view... It complements Krugman's paper in the sense that he's really looking at how the Japanese were thinking about the problem the wrong way, in all sorts of dimensions that are really quite interesting even today to look at.

Beckworth:  Yeah, that's a great summary of his paper. And, there were other papers that followed. I'll just mention, Scott, that in many of those papers, he, too, alludes to this idea of permanence. When you increase the monetary base, it's got to be permanent. And, in fact, in 2015, after he has left the Federal Reserve, he's now outside, he was interviewed in the Financial Times. And, I want to read a quote from this interview that underscores this idea of permanence. And, again, to be clear, he doesn't focus, like Krugman, on some kind of monetary-based target, but he focuses on the implication of it, that is the price level target or some kind of makeup policy's important, but the implication is this permanent increase.

Beckworth:  So, in 2015, he said in this interview, his words, "The theoretical result is that if the Fed increases the money supply in a period of deflation risk, what's important is that the central bank persuasively raised the money supply on a permanent basis rather than a temporary basis." So, he stresses, he'd have to persuade the public and it's got to be viewed as permanent. So, he's understood this point all the way through. Again, he would have viewed it probably more from a price level target.

Beckworth:  Okay, we're running near the end of our time, but we have one more economist to add our discussion. That's Lars Svensson. So, talk us through his contribution to the Princeton School of Macro thought.

Lars Svensson’s Contributions to Macro Thought

Sumner:  Alright, so Lars Svensson has two interesting papers from 2003, both of which are, I find, quite interesting. One has a discussion of what Svensson calls, targeting the forecast. And, that means you set monetary policy in sort of a forward-looking way. You set it at a level where you expect the goal variable to be equal to the target. Let me make that more concrete. Let's say you have a 2% inflation target, you set policies such that you're also forecasting 2% inflation. Okay. And, from one point of view, this is just common sense, why wouldn't you set policy at a level where you expected the policy to be successful. But, you'll find in the real world, central banks will often set policy at a level where they forecast the outcome to be somewhat off target. Maybe, they forecast inflation at 1.5%, but their goal is 2% inflation, and so on.

Sumner:  Svensson actually served on the board at the Swedish Central Bank for a while in the 2010s. And, he ran into some resistance there to this idea. So, he was complaining that his colleagues were setting policy in a position where inflation was likely to come in below target. And, in his view, and my view as well, that didn't make any sense. And, one of the things I would point to about this idea of targeting the forecast or sort of, a forward-looking approach to monetary policy, is that, I think, somewhat meshes with the direction that policy is actually evolving in the real world. So, there's some signs that Svensson's ideas, I think, are having some effect. The alternative approach is often to rely on say a Taylor Rule formula that's kind of, backward-looking. But, if you look at central banks like the Fed, in recent years, they've been often moving interest rates in a way that I don't think would be justified under a Taylor Rule approach. Like in 2019, even before the recent COVID situation, the Fed cut rates three times, even though unemployment was only 3.5%.

Sumner:  So, if you're using a Taylor Rule approach, I don't think you'd have those rate cuts. That's not a situation where, in the past, the Fed would have done so. But in my view, those three rate cuts actually staved off a recession that otherwise would have occurred in 2019. We're about a decade into the expansion, that's often when expansions peter out. And, the past expansions often ended because of the so-called natural rate of interest fell, and the Fed was behind the curve in reducing its policy rate. So, the policy rate was getting to be above the natural rate of interest. The Fed was more proactive in 2019. It was relying on forward-looking forecasts, and it saw signs that the economy might be slowing. And, some of these were in the financial markets and responded to those signals by moving interest rates aggressively lower.

Sumner:  And, then by early 2020, actually, it looked like we'd avoided a recession. I think we probably would have if COVID had not hit, of course, in March of 2020. So, that's one paper that I think is very interesting. And, even more interesting is the paper he wrote the same year on, I think, he called it a foolproof way of escaping from a liquidity trap. And, so, like the other Princeton economist, this was very much on his mind, what can monetary policy do at the zero bound. And, this was written from the perspective of again, focusing on Japan, which was the one country at the time at the zero bound.

Sumner:  So, he came up with a proposal that involved manipulating the foreign exchange value of the yen. And, I won't get into the details of it, I'll give you a simplified version, just to give you the intuition. The simplified version is this, for decades, Japan has had inflation rates of a couple percentage points below the US. And, they've had interest rates, on average, a few percentage points below the US. And, you can explain the lower interest rates through the Fisher effect, right? If you have a little bit higher inflation, you tend to have a little bit higher nominal interest rates. And, that's what we've seen in the US. Well, what if Japan were to peg the yen to the dollar at a fixed rate, as other places like Hong Kong do?

Sumner:  In that case, according to purchasing power parity, in the very long run, you should have a similar inflation rate in Japan in the United States. Now, you might want to depreciate the yen a little bit first before you fix it to the dollar. But, the basic idea is, in the long run, if you fix the yen to the dollar, you'll have a similar inflation rate in the two countries. And, that should then raise Japanese inflation closer to US inflation.

Sumner:  But, here's what's interesting about it, and I haven't seen anyone else notice this, in his proposal, the nominal interest rate goes up in Japan immediately and persistently. And, that's because of interest parity condition. So, because America has higher interest rates than Japan, if the Japanese pegged their currency to the US currency, Japanese interest rates immediately rise to the level of US interest rates. Now, normally, we think of higher interest rates as a tight money policy. And, Svensson assures his readers, he says, "Don't worry, the real interest rate's actually lower in this situation, it's just the nominal interest rates going up." But of course, when I saw that, I immediately recognized something that didn't even occur till a decade after he wrote the paper, which was this, Neo-Fisherian idea that a truly expansionary monetary policy actually involves higher nominal interest rates.

Sumner:  Now, my own view is, both the Keynesians and the Neo-Fisherians are wrong. I don't think low interest rates are expansionary, I don't think low interest rates are contractionary. And, vice versa, for high interest rates. I think that's what's called reasoning from a price change. The effect of any change in interest rates depends on what's causing the change in the interest rate. But, what Svensson showed, is there's a very plausible monetary policy regime that you can write down on paper, sketch it out, and it will clearly raise inflation over the long run, and it will clearly raise interest rates immediately and persistently. And, that's kind of a surprising result. That's essentially the claim of the Neo-Fisherians.

My own view is, both the Keynesians and the Neo-Fisherians are wrong. I don't think low interest rates are expansionary, I don't think low interest rates are contractionary. And, vice versa, for high interest rates. I think that's what's called reasoning from a price change. The effect of any change in interest rates depends on what's causing the change in the interest rate.

Sumner:  Now, that doesn't mean that if tomorrow, the Fed raises interest rates, that's going to increase inflation in the United States, anymore than buying an umbrella will cause it to rain or whatever, that's reversing causality. But, there are monetary policy regime changes that will both have the effect of raising inflation and have the effect of raising interest rates. And, in a few cases, like the one he cites, even immediately raising interest rates. So that’s kind of interesting.

Beckworth:  So, is it fair to say that Svensson's foolproof escape way is a special case where Neo-Fisherian theory holds?

Sumner:  Yes, it is. And, I would go much further than that. I don't think Neo-Fisherian theory, itself, is all that important. I think why it's important is, it's an interesting critique of professional economics. The real problem I see here is this, we've consistently misdiagnosed the stance of monetary policy by looking at interest rates, and to a lesser extent, QE. And, a lot of these are very reactive to changes in the economy. So, in other words, you'll get a tight money policy, that will depress the economy, interest rates will fall. And, because interest rates fall, people think, oh, money's easy, and yet the economy's depressed, so monetary policy doesn't work. So, they're confusing cause and effect.

Sumner:  And, if you look at Svensson's paper, you can see that we can't actually draw this connection. We can't assume a low interest rate policy's an easy money policy. And, if it's not working, we can't assume that it means monetary policy's ineffective. He showed a, quote, foolproof way for Japan to get out of the liquidity trap that would actually involve higher interest rates. So, it would look to the average person like a tight money policy, but would actually be far more expansionary than what the Japanese are doing right now.

I would go much further than that. I don't think Neo-Fisherian theory, itself, is all that important. I think why it's important is, it's an interesting critique of professional economics. The real problem I see here is this, we've consistently misdiagnosed the stance of monetary policy by looking at interest rates, and to a lesser extent, QE. And, a lot of these are very reactive to changes in the economy.

Sumner:  Now, just as an aside, your listeners may wonder, well, why don't the Japanese do this? And, I think the reason is, the US government won't let them. So, the US government will say, that's currency manipulation, we're not going to let you do this, we'll put trade barriers on your exports and so on. But, in a technical sense, it would work in a very effective way. So, what the Japanese really need to do is, do something that mimics that, but without making it look like they're directly intervening in the foreign exchange market. That's the trick they have to look for, a price level rule with a whatever-it-takes promise to buy domestic assets. Say, at a high enough inflation rate, so they don't have to buy up the whole world, like Bernanke mentioned.

Beckworth:  And I will mention again, that in that full proof paper, he mentions an implication, is a permanent increase in the monetary base. Okay, very fascinating work. And, Scott, we look forward to you finishing this paper and getting it out. We will spread it and circulate it. But, for now, our time has come to an end on the show. Our guest today has been Scott Sumner. Scott, thank you for coming on the program.

Sumner:  Well, thank you for inviting me.

Photo by William Thomas Cain via Getty Images