Scott Sumner on The Money Illusion

Scott Sumner is David’s colleague and the Ralph G. Hawtrey Chair of Monetary Policy at the Mercatus Center. Scott is also a returning guest to the podcast and joins David on Macro Musings to discuss his new book, *The Money Illusion: Market Monetarism, the Great Recession, and the Future of Monetary Policy.* Specifically, David and Scott discuss common misconceptions about the 2008-09 Recession, why bubble narratives too often miss the mark when explaining rising asset prices, whether the Fed’s adoption of average inflation targeting signals that it is moving toward a level target, and much more.

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: Thanks for inviting me, David.

Beckworth: Well, it's a big day for the show and it's a real treat to have you here because your book is being launched. We're going to go to an event today here in town, in DC, but you have a new book out and we're excited to promote it and talk about it. This book is the book of your journey, right? Really since the Great Recession, and we'll get into that, but several other reasons, this is a big deal, at least for me, is this is the first podcast we've done in person back in the studios at Mercatus Center, Scott.

Beckworth: You have a very select crowd here. You're the first one back recording in studio in person. This will also be show number 300. This is the third hundred show we will have issued or ran since we started back 2016. So, it's exciting to have you on. I believe the last person we had in person, the last time I was in the studio, it was Paul Schmelzing. I don't know if you saw that episode. He has that paper showing the 700 year downward trajectory of interest rates.

Sumner: Yes, I remember that.

Beckworth: Yeah, very interesting. I was just as curious, so back when we talked with him, the tenure was like a 1.6, so now it's at 1.3. So, his theory still seems to be holding up for the most part, but we're here today to talk about you and your book, and you have a great book. It's called The Money Illusion: Market Monetarism, the Great Recession, and the Future of Monetary Policy. You have six sections in your book. I'll just mention them briefly: value of money, the dance of the dollar, never reason from a price change, how to think about macroeconomics, the Great Recession, and what does it all mean?

Beckworth: Those are the big sections, but instead of going through those, I want to maybe talk to you from a perspective of your journey in writing this book. How did this book come about? The stages of your experience from the Great Recession period, right around then, and then to the present. Let's start with what you were working on right prior to the Great Recession. So, you've been doing some research, and I believe in three different areas. Talk about those.

Sumner: Great. Yeah, the book is sort of based on my blogging, but the blogging itself is based on research that I've been doing since before the Great Recession began. I felt like, once I got into it, that my research gave me a unique vantage point because I was looking at three separate areas that all sort of came together in the Great Recession and gave me certain per of that some other people maybe didn't have. Those three areas of research were work on the Great Depression of the '30s, which has obvious parallels to the Great Recession of 2008.

Beckworth: Yep.

Sumner: The Japanese liquidity trap of the late 1990s and also using market indicators as a guide to monetary policy. I was very interested in ideas like creating a nominal GDP futures contract to guide monetary policy, or market indicators more broadly. Then, in late 2008, a bunch of things happened that seemed to me very closely related to the research I'd been doing. I saw a lot of parallels to what had led to the Great Depression. I saw market indicators flashing red and were seemingly ignored because of reliance on more of a sort of tailor rule approach. Inflation was high during 2008. So, I got very concerned about where policy was going around the Fall of 2008, and I think it was partly based on the research that I'd been doing.

Beckworth: You were early to the party of sorts warning that this could turn out really bad, this looked like another Great Depression. I should mention your work on the Great Depression was cited by Gauti Eggertsson and his great papers that he wrote on the Great Depression in the AER. You had all these interesting papers and projects going on, and you said you saw signals in 2008. What were specifically some of the signals that really had you worried?

Sumner: I guess there were several signals, but probably the two that were of greatest concern were the sharply falling stock market and the dramatically shrinking tip spreads. So, the tip spreads, as you know, are a sort of crude indicator of market inflation expectations, and they went negative in the Fall of 2008. But there were many other indicators all over the global economy showing a sharp contraction hitting, during that Fall quarter. Particularly, I think right after Lehman failed, and that was in September 2008, the Fed did not cut interest rates.

Sumner: That's when I became sort of radicalized. Over the previous two decades, I'd been pretty content with Fed policy. For instance, I visited Harvard University. At that time, I was living in the Boston Area, and I spoke with one New Keynesian economist and had a good conversation. At one point, I asked something like, "Doesn't the Fed’s see what's going on?" And he responded, "Oh, they see it. They just don't know what to do about it."

Sumner: That really had a profound impact on me because I had read a number of papers by well known economists, including Ben Bernanke, that were highly critical of what Japan had done in the '90s and early 2000s, and it offered suggestions as to what monetary policies should be doing, that it insisted that the zero bound was not a barrier. Lars Svensson had the foolproof way out of liquidity trap and so on.

I thought those papers were very good and I thought we had a roadmap for what to do if we got in the same situation as Japan, and then suddenly, in the Fall of 2008, discovering no, we don't have a roadmap. That was very disillusioning for me. So, I became radicalized and I started blogging and so on.

Beckworth: It's hard to think of you being radicalized. You're such a nice person, Scott. But you were definitely beating the drums that the Fed should be doing more and they were slow to respond. If I understand correctly, you were almost giving up on monetary policy research. Up until this point, you thought everything was kind of settled or done. You still had interest in the futures, but you didn't see yourself doing this journey, going on this path. You're going into some other areas of research, is that right?

Sumner: Yeah. I'd been doing some work on neoliberalism and policy, sort of more development oriented policy issues. Interestingly, I had just been working on an idea for paper arguing that it's impossible to predict recessions, especially demand side recessions, because if we could predict them, we could prevent them. James Hamilton said something similar to that at one point I remember. The basic idea is we shouldn't waste a lot of time trying to come up with models that would predict recessions, because if we actually had a model like that, the Fed would take steps to adjust monetary policy until a recession was no longer expected.

Sumner: I gave up on that project, because in the Fall of 2008, I saw that policy was set at a position that was expected to lead to a recession basically. We were not doing the things that would've been appropriate if we wanted to avoid a recession. We would've cut interest rates sharply after Lehman failed, if that was possible.

Beckworth: Even if we knew a recession was coming, your point is we weren't doing what was necessary to respond. Even if we had that perfect foresight, there was still the problem of responding to it.

Sumner: Yeah. In one of my blog posts, I use this metaphor. I say we shouldn't expect a highway engineer to forecast a bridge collapse. We should expect a highway engineer to prevent a bridge collapse. If they knew that bridge in Minneapolis was going to fall into the river a few years ago, they would've buttressed it before it fell. Right?

Beckworth: Right.

Sumner: You don't really want to predict a bridge collapse. You want to predict situations where you have to take a policy action to avoid a bridge collapse. So, you want a monetary policy that essentially is expected to avoid recessions. Now, recessions still may occur, but you don't want to be in a position where you're basically forecasting a recession because then you have monetary policy set at the wrong position with the important caveat that there might occasionally be a recession that monetary policy can't do anything about and is predictable.

Sumner: Now, would that occur ever? I doubt it. The COVID recession was certainly something that could not have been prevented by monetary policy, but that was also pretty much not predicted. You'd have to have sort of a predictable supply shock that you couldn't do anything about with monetary policy. I don't know what that would look like.

Beckworth: Yeah. That's a tough nut to crack. A big part of your presence was your blog, which was also called TheMoneyIllusion, and your book is kind of a compilation of your thoughts and ideas over this time. How did you get into blogging? You were frustrated, you were concerned, you felt policy makers weren't responding in a timely fashion. Is that what prompted you to start blogging and why blog? Why choose the blogging medium?

Sumner: Oh, well, I wrote a few op-eds that didn't get published, and I'm not very good with computers, but someone helped me set up the blog. It was just a way to get my ideas out there. When I started blogging, I didn't expect very many people at all to read the blog. It did sort of catch on for a while, I think, partly because it was a very topical issue and a lot of people were talking about it. I had a little bit of a fresh perspective. I had a perspective that didn't fit neatly into the standard narratives on either the left or the right.

Sumner: I was arguing policy was too contractionary, which wasn't necessarily friendly to people on the right. I was arguing that we should be using monetary stimulus, and lot of people on the left thought only fiscal stimulus would work in that situation. It was sort of a third way approach. I think that carved out a little niche for myself, and of course, you were blogging at the time also.

Beckworth: Yes. You were called the blogger who saved the world. Remember that? I think it was 2012. Was it The Atlantic? Had an article that, the blogger who saved the world, because your work, the momentum it created, helped contribute at least to QE3, QE2. I forget exactly. What is your recollection of that story?

Sumner: Yeah. I think it was ‘saved the economy,’ which is itself a lot of hyperbole.

Beckworth: Hey, take it. Claim it.

Sumner: The world would be even more absurd, but obviously, it's flattering to get those stories. I don't think they're accurate, but I believe it was 2012, I think it was the Fed policy announcements. I may be incorrect on this, late in the year where there was some movements like QE3 and forward guidance. Do you remember there was going to be a fiscal cliff at the beginning of 2013?

Beckworth: Yeah.

Sumner: Late 2012, the Fed announced a number of initiatives like QE3 and some pretty aggressive forward guidance. It was partly done to offset the anticipated slowdown due to the fiscal austerity at the beginning of 2013. I think there was a positive stock market reaction to the announcement and sort of a little bit of a euphoria in the financial press about this move. So, there were a couple articles that mentioned my name in this period. But I think that realistically, people like Michael Woodford have more influence.

Sumner: As I recall, Michael Woodford gave a talk at the conference in Wyoming, Jackson Hole, and mentioned nominal GDP targeting as sort of a good second-best policy, good compromise policy, level targeting specifically. There was a number of people that were pushing the Fed more aggressively in the early 2010s. I don't think it was my blogging per se, maybe had a little bit influence on the debate.

Beckworth: Well, it was an important piece of the puzzle about that. There was a big puzzle that led the Fed to be more accommodated during that time. We know Christina Romer, who was at the CEA back then. I remember she had an op-ed in the New York Times that said, Will Bernanke Have His Volcker Moment? And she called for the nominal GDP targeting. But you were definitely an important piece of that momentum, that push to, to respond.

Beckworth: I know also we've looked back and we've read Bernanke's book. Even if Bernanke wanted to follow everything you prescribed, it was very difficult internally with regional Fed presidents who disagreed, much more hawkish. You had Congress coming down on him. There's a lot of moving parts there, but you are definitely one of them. I think you played an important role. Now, just briefly before we get into some of the substantive parts of your book. So, you were blogging, and you continue to blog, I should say that it's pretty amazing. You continue to blog, continue to respond to comments on your blog, but at what point did you come to Mercatus? You were working at Bentley University. What year did you begin working as a senior research fellow at the Mercatus Center?

Sumner: I think it was 2015. As I recall, I started maybe six months on a sort of temporary contract of some sort and then went full-time, maybe at the beginning of 2016, roughly. At that time, I was just finishing up teaching at Bentley University. I'd been there for more than 30 years. So, I sort of retired from teaching and went full time with Mercatus, but I was basically working out of my home. For family reasons, we decided we wouldn't move. My daughter was in high school.

Beckworth: Sure. Yeah. So, you've been here. You started blogging out of Bentley University 2008 or 2009?

Sumner: Beginning of 2009.

Beckworth: Okay. Beginning of 2009, but you were writing off op-eds, you were visiting Harvard faculty trying to ask them what's going on. You started this conversation really, I'd say late 2008, but blogging 2009, and then this added to your notoriety, your fame, if I can say that, and then Mercatus Center picked you up. Other people were interested in your work as well. I just want to say, as someone who works with you as a colleague, I'm very grateful for all that, because I wouldn't be here, Scott, if it weren't for you and your work.

Beckworth: You're the reason we have a Monetary Policy Program at the Mercatus Center. The reason we have this podcast right now is because you started this monetary policy program. We've brought on Chris Russo, another scholar, and we have several people working with us. We appreciate all your labors, all that you've done, and we continue to appreciate what you've done, and that includes your new book, The Money Illusion.

Sumner: And I appreciate what you've done. I think you've raised the program to a higher level since you've been here. Certainly your podcast is one of the most important parts of our program.

Beckworth: Well, let's just say this. I still pinch myself that I get paid to talk to interesting people like you and others, so it's been a real treat for me too. But your book, The Money Illusion, again, it encapsulates all that you've done, you've worked on over the past decade plus. I want to step back and tell your story, your interpretation of what happened in 2008. Really, I guess you could say 2006 or 2009, because you have a very different view than the standard account of what happened in the Great Recession.

Beckworth: Some people prefer to call it the “Great Financial Crisis.” They see it as a financial crisis, some see it as a housing crisis or a credit boom bus cycle. But you have a very different story to tell. Maybe not entirely different, but you definitely focus on something else that's more important, and these are kind of like secondary effects. Maybe walk us through your interpretation of what happened during that time.

Misconceptions about the 2008-09 Recession

Sumner: Most people have a vision of business cycles where policymakers such as the Fed and fiscal policy are like firemen that come in and put out fire. In this vision, the private economy is inherently unstable and it creates problems, and we have to fix those problems with Federal policies of various sorts. The monetarist perspective that I come from sees policy makers as more like arsonists. This is a little bit cruel, of course, and I don't mean this disrespectfully because they're doing the best they can, but policymakers unintentionally create fluctuations in nominal GDP growth.

Most people have a vision of business cycles where policymakers such as the Fed and fiscal policy are like firemen that come in and put out fire...The monetarist perspective that I come from sees policy makers as more like arsonists. This is a little bit cruel, of course, and I don't mean this disrespectfully because they're doing the best they can, but policymakers unintentionally create fluctuations in nominal GDP growth.

Sumner: When there are these fluctuations, if it's very high, you get an inflation problem, and if there's a sharp slowdown in nominal GDP growth, you get a severe recession and often a financial crisis. These are predictable effects of a slowdown in nominal GDP growth for two reasons. One is nominal wages are sticky and the other is people contract nominal debts. For that reason, when the flow of nominal income in the economy drop sharply, there's less money to pay workers and less money to service debts. So, you get high unemployment and you get financial stress.

Sumner: And this occurs almost every time. There's a sharp slowdown in nominal GDP. But because it's hard to see the connection between monetary policy and nominal GDP, people reverse the causation and they see what's going on as instability in the private economy causing a recession, and then policy makers coming in to fix the problem. The indicators people use for monetary policy like money growth and interest rates are not really reliable indicators.

Sumner: Therefore, they don't really see how Fed policy could have actually caused the Great Recession by allowing nominal GDP growth to slow sharply. But that's in fact, my view of what happened, and I think over time, that has become strengthened by lots of other things that I've learned that I didn't even know when I started blogging. The work of Kevin Erdmann on housing. I learned that the whole idea of a housing bubble was probably a misconception.

Sumner: The housing prices were not unreasonable in 2006. The level of housing construction was not in 2006. I think we know that now from some of this recent research. The slowdown in housing construction between beginning of '06 and '08 was not associated with a sharp rise in unemployment. The recession was really caused by a broader drop in nominal GDP growth that affected all industries, not just housing construction. The banking crisis didn't really get severe until the Fall of 2008. That severe banking crisis took place about nine months after the recession started.

Sumner: Rather than causing the recession, it was itself a response to slowing nominal GDP growth. Another misconception is that the Fed was doing all it could in 2008. Not only was it not doing unconventional stimulus in 2008, it wasn't even doing all the conventional stimulus. It refused to cut interest rates after Lehman failed in September of 2008, holding them at 2%. A lot of people, I think, sort of misremember the early stages of the Great Recession and have this sense that there was this wildly unstable speculative of economy that collapsed, and the Fed had to come in and clean up the mess. That's just not what happened when you look closely at the data.

Beckworth: Going back a few years before that, when the Fed had all those rate hikes from, I guess, mid 2004 through 2006, do you think that was another example of them bonding to bubble pressures, bubble talk? I mean, why were they tightening back then? Because that surely contributed to this.

Another misconception is that the Fed was doing all it could in 2008. Not only was it not doing unconventional stimulus in 2008, it wasn't even doing all the conventional stimulus. It refused to cut interest rates after Lehman failed in September of 2008, holding them at 2%. A lot of people, I think, sort of misremember the early stages of the Great Recession.

Sumner: Actually, I think that those interest rate increases were appropriate in my view. I believe they should be targeting nominal GDP growth now, exactly where is debatable, but no GDP was growing pretty briskly in 2004, '05, early '06. The equilibrium interest rate was probably rising during that period. The Fed probably needed to raise interest rates to prevent inflation from overshooting at that time. I think that the real mistakes were made, maybe from late 2007 all through 2008, but especially in the second half of 2008.

Beckworth: Yeah. This is the response I've gotten, because I share your view. I've written op-eds. I wrote one in the New York Times. I mean I got so much hate mail from that. The argument I made was with Ramesh Ponnuru. You remember this, is that, the way we looked at it, and I think you shared the same perspective, is the Fed cut rates to 2% and everyone will say, man, the Fed cut rates from like, what was it? 4% or 5% all the way down to two. It was massive rate cuts and you're telling me the hasn't done enough?

Beckworth: The problem is though they stop in April and they don't do anything until October. That equilibrium rate is actually declining, declining, declining, and they're keeping the policy rate fixed at two, so there's a growing gap between what the actual rate is and what the equilibrium rate is. But moreover, the Fed was talking up rate hikes in the first half of 2008. If you look at speeches, even I recall the August FOMC minutes talk about how participants believed that they would have to raise a Titan policy because inflation was taking off. In September, which is probably the most egregious example, Lehman had collapsed, and they said in their statement that there's equally concern about output collapsing and inflation taking off at that time. That's a long period from April to October to kind of sit in their hands in terms of monetary policy.

Sumner: Well, yeah, and I would even go further. I don't think those rate increases mean, rate decreases, sorry, mean what people think they mean. When people think about the Fed cutting interest rates, they visualize an easy money policy. But rates can move around for many reasons. Because of the sharp slowdown in the housing market in '07, the natural interest rate was falling.

Beckworth: Good point.

Sumner: Now, the Fed was actually preventing market interest rates from falling as fast as they normally would have. They weren't pushing rates down. I'll give you an example. From about August 2007 to May 2008, a period of about nine months, the Fed didn't increase the monetary base at all. That's very unusual because normally the base creeps up a little every year during that period. That sharp, slow down in the monetary base was done to prevent rates from falling even faster.

Sumner: We have this sort of cognitive illusion. We see the Fed cut rates every few months, and we think, oh, they must be doing easy money. The rate's going down stepwise fashion. But if the equilibrium rate is falling faster, that same Fed policy may be holding them up above the equilibrium rate. There have been some studies that estimate where the equilibrium interest rate was, and they show the rate falling faster than the Fed cut rates. I think it's very misleading to look at how the Fed is tweaking rates up or down and drawing any implications about the stance of monetary policy.

Sumner: Now, then later, they of course, did increase the monetary base sharply when we got into the QE programs, but by that time, it was sort of too late to prevent a severe recession. But I do think that if they'd been more aggressive with monetary policy in late '07 and into 2008, the recession would've been far milder.

Beckworth: It's just hard to see that and to recognize that there's both the actual target rate and then the equilibrium rate, and you got to look at the gap between the two if there is one, and that's the stance of policy, not what the absolute level of the rate is.

Sumner: In a sense, it's sort of even worse than you'd think. It's not just that the interest rate is kind of unreliable or sometimes unreliable. As a general rule, when the Fed is cutting interest rates, monetary policy is getting tighter. That's because when they're cutting them, it's usually a period when they need to be cutting them faster, it's when we're going into a recession. As a general rule, when the Fed is raising interest rates, they're usually falling behind the curve. The equilibrium rate is rising faster and they should be raising rates faster.

We have this sort of cognitive illusion. We see the Fed cut rates every few months, and we think, oh, they must be doing easy money. The rate's going down stepwise fashion. But if the equilibrium rate is falling faster, that same Fed policy may be holding them up above the equilibrium rate...I think it's very misleading to look at how the Fed is tweaking rates up or down and drawing any implications about the stance of monetary policy...As a general rule, when the Fed is raising interest rates, they're usually falling behind the curve. The equilibrium rate is rising faster and they should be raising rates faster.

Sumner: A lot of the business cycles historically in the US has been the Fed moving rates around more slowly than the natural rate. It's almost exactly the opposite. I sometimes use the analogy of when wages are falling. They're usually above equilibrium, because wages are sticky, slow to adjust. When wages are rising, they're usually lagging behind equilibrium. It's sort of counterintuitive in that sense.

Beckworth: But this is consistent with your call for a future's market. I mean, what you're saying is the Fed typically is behind when it's adjusting rates, whether up or down. The Fed is effectively following where the fundamentals are taking interest rates, and often it's too slow to do so. I think your argument for the markets, relying on market signals more, and in particular your futures contract, is that it would help the Fed get over that problem. The Fed would be much more nimble and responsive to the changes in the economy.

Sumner: Exactly. So, if you want 5% nominal GDP growth, you want to set interest rates, not according to some formula like the Taylor rule, you want to set interest rates at a level that the market believes will lead to 5% nominal GDP growth over, I'll say, one- or two-year horizon.

Beckworth: Yeah. Okay. The nice thing is all your research agendas all come together and fit a nice big picture here. One other angle on this is financial crisis story. As I mentioned, some people call this the great financial crisis. I like to call it the Great Recession. And the people who do that, they view this through the prism that it was a financial crisis. The financial system was crashing, shadow banking had a run on it. The institutional money markets had a run on them.

Beckworth: And the Fed had to step in. In fact, the Fed opened up its first dollar swap line facilities. Actually, they've been around longer than that, but they opened them up to multiple central banks. People focus on that side of story, but you stressed, is that before that happens, there has to be a slowdown in nominal income growth that makes it harder to fulfill obligations and financial contracts. Is that right?

Sumner: I think that's usually the case. Obviously it's possible to have a financial crisis during a period of stable nominal GDP growth. But historically, as far as I know, these sorts of financial crises tend to occur, almost always, when there's a sharp slowdown in nominal GDP growth.

Beckworth: I mean the Great Depression would be another example. Right?

Sumner: Great Depression, Argentina around 2000. There's many other examples like that. Of course, Europe, by the way, it's interesting that Europe had a recession the same time as we did, even though they didn't have the subprime situation. Well, they point to other debt problems like Greece, but that's very predictable when European nominal GDP growth plunged, even more than it did in America, it's natural that the weakest borrowers are going to struggle in that kind of environment. Early on, there was, almost a little bit of gloating in Europe about how America was being hit by our reckless cowboy capitalism in early 2008.

Sumner: And they'd been more responsible in their banking behavior. Well, at that time, almost no one expected that the European recession would be far worse than the United States. Well, let's say you take this hypothesis that it's the subprime mortgage that caused the American recession, could you explain why the recession was worse in Europe based on that? Obviously not. But suppose you felt it was due to monetary policy and how it impacted nominal GDP growth, now there is a big difference. The ECB raised interest rates in July of 2008, making their initial slump even steeper than the US.

Sumner: And unlike the Fed, the ECB raised rates twice in early 2011, immediately followed by a double dip recession in Europe, which we didn't have in the United States. So, there's a direct correlation between monetary policy errors by the ECB and their deeper, double dip recession, and the somewhat less bad response of the Federal Reserve and a milder recession here. On the other hand, if you take the subprime mortgage theory, there's no way to explain why Europe was hit much harder than the United States.

Beckworth: Going back to 2008 as an example, I mean, if you look across a country, Australia also had a high amount of household debt leverage there, a lot of high home prices. So, all the same symptoms the US had, but it had a very mild experience. I'm not even sure if it had an outright contraction, just it maybe slowed down a bit. But they had a much more accommodative monetary policy.

There's a direct correlation between monetary policy errors by the ECB and their deeper, double dip recession, and the somewhat less bad response of the Federal Reserve and a milder recession here. On the other hand, if you take the subprime mortgage theory, there's no way to explain why Europe was hit much harder than the United States.

Sumner: Right. So, they had a higher trend rate of growth of nominal GDP. For various reasons, their economy is a little bit faster growing, has more immigration, has a little higher inflation target than other Western countries. All those came together to produce about 6.5% trend growth in nominal GDP in Australia at that time. What that meant is that nominal interest rates in Australia were historically higher than in other Western countries. One advantage Australia had is they didn't face the zero bound problem, and they maintain nominal GDP growth at a healthy enough level. It did slow during the global recession, healthy enough level to avoid a technical recession. There was a slow down in Australia, and they actually went from the early '90s and all the way up to COVID before they had a recession, about 30 years roughly.

2020 as a Case Study for Level Targeting

Beckworth: Yeah, that's amazing. No recession from 1990 up until 2020. I mean, that's just quite the run. Hats off to the Australians. Now, the other example I want to bring out that lent support to your thinking about financial crises is this past year, 2020. We saw no banking crisis. We saw some stress in the financial markets in late January and February. Stock markets were contracting quite a bit, but once the Fed stepped in, things looked much better. But one could make the argument, the reason we didn't have a severe financial crisis last year is because of all the nominal income support from the government, both through monetary policy and fiscal policy, that nominal incomes were, not only stabilized, they went above trend. And that meant people could make their mortgage payments, they could make their car payments. In nominal terms, at least, the financial obligations were met, the nominal debt constraint wasn't binding.

Sumner: Exactly. Both of us have been promoting this notion of level targeting for quite a while. The idea is, whether you're targeting prices or nominal GDP, when you have a slump, you want to come back to that trend line as quickly as possible. As you recall, in the Great Recession, when we had the deep slump in both nominal GDP and inflation, we simply set a new and lower trendline. We didn't come back to the original trendline, and that made the recovery very slow from the Great Recession. This time around, nominal GDP is already back close to the trendline. You can draw the line in different ways, but we're basically back close to trend.

Sumner: In terms of the price level, we're actually slightly above the trendline because of the supply shortages. In the nominal sense, nominal spending has recovered very strongly. As a result, I think that's the major reason we haven't had a debt crisis and we've had very aggressive, both monetary and fiscal policy during this period. But I also think the expectations channel has been important. I think the markets have bought into the Fed's commitment to do average inflation targeting, and so make up for any inflation shortfalls, and that's boosted confidence. Also, the Fed's willingness to sort of do whatever it takes and buying a lot of assets to stabilize things has helped. There's certainly been a lot of fiscal stimulus, which has helped average people service debts, as you say. Yes, I think that does explain why the financial crisis hasn't occurred this time.

Beckworth: Yeah. I want to stress that point about average inflation targeting being an important part of that story because it's the Fed saying, look, we're going to do no harm. We're going to allow the recovery to occur at a healthy pace, get back to the trendline, and we're not going to get in the way. Whereas, in the past, they would try to nip the recovery before it even really took off because they were concerned about inflation forecasted looking forward. I think the Fed is an important part of that, understanding that story. I stress this because, I'll just bring up some recent things that have been happening, Jay Powell was up for nomination again, as you know.

In terms of the price level, we're actually slightly above the trendline because of the supply shortages. In the nominal sense, nominal spending has recovered very strongly. As a result, I think that's the major reason we haven't had a debt crisis and we've had very aggressive, both monetary and fiscal policy during this period. But I also think the expectations channel has been important. I think the markets have bought into the Fed's commitment to do average inflation targeting, and so make up for any inflation shortfalls, and that's boosted confidence.

Beckworth: Many progressives have been attacking him because he hasn't been really strong, or he's at least dialed back financial regulation. They don't think he's been enough to stabilize the financial system. You can have that argument, that conversation, but what that conversation is missing, in my view, is the fact that this new framework probably does even more for financial stability than tweaking financial regulations do, because it's a commitment to preserve the flow of nominal income into the economy. You do that, you get, for the most part, financial stability.

Sumner: Yeah, I think that's right. From my perspective, the Fed's role in monetary policy is by far, its most important role. I do think there are some issues with financial regulation related to the moral hazard in our system from FDIC and so on. I have an open mind on what the regulations should be. Maybe you could argue for a little bit higher capital requirements, so things like that. That's not my area of expertise, but for monetary policy, I think Powell’s done an excellent job with the new policy of average inflation targeting. And he's also very skilled at signaling and working with Congress and so on. I've been pretty happy with his performance.

Beckworth: Scott, we've been talking about your book and what you discuss in it. Now, the book was finished in 2018, is that right, 2018?

Sumner: Mostly finished in around 2018. So, it was written before the COVID recession, which is a little bit unfortunate.

Beckworth: You do recognize it in the book though.

Sumner: I do. I do mention it. After it's written and goes to the publisher, you do have a chance to come in and add a couple of paragraphs to reflect current events. So, it does look to readers like it was written recently, and some of the data was updated, but the core manuscript was written earlier. I should mention that I think it was unfortunate because this is a book about how nominal shocks cause recessions. We're hit, right as the book is coming out, with a once in a hundred-year real shock recession. The timing is not exactly ideal, but on the other hand, I feel good about some of the other aspects of the book that maybe we can talk about.

Beckworth: Yeah, let's talk about some of the claims, the ideas you argue for in the book, starting, and we've touched on this a little bit, but start with your critique of the traditional bubble story.

The Traditional Bubble Story

Sumner: I'm a believer in the efficient market hypothesis. Not that it's precisely true. No social science model is too, but I think it's a useful way of thinking about markets and how they aggregate information and provide an optimal forecast. I think asset prices are usually relatively efficient based on fundamentals. I'm very dubious of people who claim that such and such a market is obviously overvalued. Most experts, I think, believe that the tech stocks in 2000 were obviously overvalued, or housing prices in 2006 were obviously overvalued.

Sumner: For several years, those claims, against the EMH, looked pretty plausible, right? But today, the tech stock prices of 2000 don't seem very high. In fact, do you recall people saying things like those stock prices only make sense if you think American internet firms will eventually dominate the global economy?

Beckworth: Well, they do.

Sumner: Well, they do now. Or the 2006 housing prices would only make sense if you think interest rates will get lower and lower and NIMBY regulations will stop new construction. Well, both of those things have happened and we're now at a new normal of much higher housing prices in America. I think these markets we're picking up some long term trends that really did change the traditional fundamental price earnings ratio or rent price ratio in housing. I think people are too quick to dismiss things as bubbles that may have underlying fundamental causes. I think the recent run-up in asset prices has strengthened my argument that maybe those earlier prices were not in fact too high.

Beckworth: In fact, if you look around the world, that's the similar story.

I think people are too quick to dismiss things as bubbles that may have underlying fundamental causes. I think the recent run-up in asset prices has strengthened my argument that maybe those earlier prices were not in fact too high.

Sumner: Yeah, all over the world now, housing prices are higher and there's a growing acceptance that real interest rates will remain really low. You mentioned that 700 year study and so on. In that environment, a flow of rents or dividends is just valued higher than it would have been in the 20th century.

Beckworth: Okay. That's the first idea that's been vindicated is your view of the bubble story. The other thing though, is the Fed seems to be doing what you called for, some version of level targeting. You feel good about that too?

Is the Fed Moving Toward a Level Target?

Sumner: Yeah, I think there's actually two areas that I would cite where Fed policy is moved in a direction I feel more comfortable with. One is the average inflation targeting is pretty close to level targeting. That is, is pretty much the idea that, when there's an inflation shortfall below their 2% target, they allow overshoot for a period of time until they get back up to the previous trend line. That's what they've been trying to do, and I think they've done that pretty successfully during this COVID recession. Now, just as an aside, because it is such an unusual recession, I don't want to claim victory. Like, it may be that just this recession is so much driven by the virus that we're overestimating what Fed policy has done, but at least it looks successful as best we can see so far.

Beckworth: Well, let me ask a question on that particular point. Do you think it was unfortunate or was it fortunate that the Fed implemented this new framework during this time? You could say it's unfortunate because all these supply shocks kind of cloud the inflation interpretation, right? Maybe we shouldn't be looking at inflation the way we should, otherwise it would if it were demand driven. The fortunate part might be, they came in with a big tailwind from fiscal stimulus. I mean, they were in a good position to try to hit their 2% target. So, there's both, I guess, two views to that.

Sumner: Yeah, it's really hard for me to say. I think that a regular recession would have been a cleaner test. Actually, the test really would have been if we could have moderated the business cycle. My view, which is a little bit out of the mainstream is that during normal times, at least monetary policy alone can stabilize the business cycle. If we do NGDP targeting and level targeting, or even average inflation targeting, we'll have fewer recessions and milder recessions even without fiscal stimulus. But of course, in the COVID period, the shock was so large that it was inevitable we'd get a lot of fiscal stimulus.

Sumner: It's a little hard to disentangle how much was done with monetary policy and how much related to the fiscal. So, it wasn't really a clean test of monetary policy in my view. I think that the better test will be going forward, when there's more ordinary shocks, how this policy works in that environment. That's one area where policy is definitely moving in what I think is a good direction. I think I'm associated with nominal GDP targeting, but in a sense, I'm more committed to level targeting than I am to NGDP.

Sumner: The level part of it is, to me, the most important part. Then another part of what's called market monetarism is using market indicators as a guidepost rather than sort of a computer model of the economy where you plug in historical data on unemployment and output gaps and set policy on that basis. I think we had a pretty clean test of the two, head to head, in early 2019. A year before COVID, the economy was slowing due to the China trade war and other factors, and there was nervousness in the markets. Some pretty big drops in the stock market. And other markets were showing indications of a slowdown.

Another part of what's called market monetarism is using market indicators as a guidepost rather than sort of a computer model of the economy where you plug in historical data on unemployment and output gaps and set policy on that basis. I think we had a pretty clean test of the two, head to head, in early 2019.

Sumner: I think the yield curve inverted briefly during early 2019. I think, during normal times, we would have had a recession in 2019. The Fed is traditionally behind the curve in reacting to this kind of situation where the equilibrium interest rate is falling. It's hard to observe. We don't observe it directly, but we can infer the equilibrium rate is falling because essentially, the bond market is signaling a likely decline in interest rates in the future and so on. Well, what the Fed did this time is ignore the fact that the economy was booming, that unemployment was only 3.5%.

Sumner: They ignored their computer models and they looked at these market indicators and they cut interest rates three times in 2019. That was something people didn't expect at the beginning of the year, based on the strength of the economy. Lo and behold, by early 2020, the economy was continuing to do well, but it was not overheating. We were not seeing indications of high inflation, and so we dodged a recession that I think would have occurred. In fact, by early 2020, we were already in the longest expansion, more than 10 years.

Sumner: I think that's a beautiful example of something that isn't very noticeable because it was like the dog that didn't bark in the detective story. You don't notice it. People don't notice there was no recession in 2019, but I think that's because the Fed took steps to prevent what would have normally occurred with a more, a backward looking monetary policy that just looked at inflation and unemployment historical data. There was no need for a rate cut if you just looked at historical data at that time. It was all based on this forward-looking market data, as far as I can see, at least.

Beckworth: Yeah, that's a remarkable period for several reasons. One is that the Fed turned around so quickly because 2018, they were talking about doing even more rate hikes. They had nine rate hikes between 2015 and 2019. And now they were talking about doing more. In fact, you mentioned the yield curve, I remember John Williams, Lael Brainard talking about, well, what's the big deal with a little inversion of the yield curve? I actually made a t-shirt during this time that said, “we have the nerve to invert the curve.”

Sumner: I remember that.

What the Fed did this time is ignore the fact that the economy was booming, that unemployment was only 3.5%. They ignored their computer models and they looked at these market indicators and they cut interest rates three times in 2019...Lo and behold, by early 2020, the economy was continuing to do well, but it was not overheating.

Beckworth: Yeah. Going from that, sometimes it's hard to change your mind when you say you're going to do something, but the Fed was nimble, and I give some credit to Powell for helping navigate the Fed away from that potentially disaster situation. You attributed to them taking market signals more seriously or observing what's going on. Let me take a more traditional view here and what someone might say who wants to stick with the computer models, the new Keynesian framework, they'd say, look, you're right. If we looked at headline numbers, we would have caused a recession, but the problem was, it wasn't the model that was our inputs. It's like we misestimated what the natural rate of unemployment was, we misestimated what our star was. And only after the fact, that we realized it was much lower. So, if we had those numbers right now, what would be your reply to that if they said that to you?

Sumner: My reply is that's what the market indicators are implicitly doing. They're estimating those. In a perfect world where you had an NGDP futures market or something like that, if you're targeting the NGDP futures price at 4% or 5% growth, then whatever interest rate comes out of that policy is the natural interest rate, right?

Beckworth: Yeah.

Sumner: At least you'd hope it is. So, we don't have any direct way to measure the equilibrium or natural interest rate in real-time. All we can do is guess what it is by looking at indicators. My argument is that the market indicators are more timely in estimating movements, and that the natural rate of interest can move very quickly in a crisis. If it starts to get worse and the Fed is behind the curve in responding, the mere fact that the Fed is behind the curve creates expectations of a downturn, which further depresses the natural rate of interest. It's almost impossible, in my view, to do monetary policy without looking at market indicators at all.

Sumner: In fact, the Fed has always paid some attention to them, They've throughout history, that 1987 stock market crash, even the 1929 stock market crash, the Fed cut interest rates right after the crash. So, it's never that they've ignored market indicators. It's more that they haven't paid sufficient attention to them until more recently. I think 2019 was, for me, the most beautiful example of them paying just the right amount of attention.

Sumner: I think, without COVID, they would have engineered maybe the first soft landing which I define as a period of time when unemployment stops falling for multiple years, but you don't go into recession. Now, you think that would be no great achievement, right? Unemployment fall, as you recover, then it should level off at a low level, and you should go a few years without recession. But if you look at the unemployment time series, we never see that.

Beckworth: That never happens.

Sumner: It goes down, down, down, and then there's a recession.

My argument is that the market indicators are more timely in estimating movements, and that the natural rate of interest can move very quickly in a crisis. If it starts to get worse and the Fed is behind the curve in responding, the mere fact that the Fed is behind the curve creates expectations of a downturn, which further depresses the natural rate of interest. It's almost impossible, in my view, to do monetary policy without looking at market indicators at all.

Beckworth: Boom, yeah.

Sumner: Other countries, interestingly, do have soft landing so it's not impossible. Britain had a soft landing in the early 2000s. Unemployment in Britain went down in the '90s, just like America, but they didn't have a recession in 2001. It just unemployment leveled off at a low level in Britain from '01 to '07, or whatever, somewhere around there. Australia has had soft landings. It can be done, but the fact that we don't, indicates there some fundamental problem with monetary policy, that when we get to the peak of the business cycle, the Fed almost always mishandles things. They're behind the curve when sentiment turns negative and they cut rates too slowly, and we never seem to get these soft landings. That's sort of, to me, the holy grail of successful monetary policy.

Beckworth: No, this is great. Jay Powell may be seen as someone who ushers in this era on a more sustained basis, where we do look more forward at asset prices, market signals. Going back to your answer to my question, what would a standard new Keynesian say to your critique? Well, just, we need better measures of the output gap, but the problem is you can't. You can't get real-time measures of these variables, and the best thing you can do is look at market signals. I mean, there's papers done how part of the 1970s problem was mismeasuring the output gap, or de Finetti's paper says, look, if you take a Taylor rule and you apply it to the 1970s Fed and you plug in the real-time output gap estimates, they were doing things just fine.

Beckworth: You're always going to have this problem. It's always going to be backward looking. It's always going to be unobserved, but what you do have are market signals. So, embrace them, use them. I hope that the Fed can stick with average inflation targeting with the use of market signals. It's not been easy though, Scott. You've seen the critiques that Powell has received and very prominent people like Larry Summers and many other inflation hawks. This is so different. This is so strange, so different. Somebody even called it now the ‘Powell Put’ instead of the ‘Greenspan Put.’ They would actually see the Fed relying too much on market signals and being too easy on inflation, but you're painting a much more optimistic picture here of what the Fed is doing.

Sumner: Yeah. Well, a couple of points on that. One is that it's possible the Fed is being too expansionary, but getting back to the output gap, I want to make one other point on that. The fact is that we are now in a situation where it's probably more difficult than ever in my entire life to know what the output gap is. There are so many strange things going on with the supply side of the economy that I think it's very, very dangerous to rely on traditional models or rules of thumb when a lot of things are changing under COVID.

Sumner: We need some kind of forecast target that we can rely on in a way that we can't rely on traditional indicators like the unemployment rate or employment population ratio, or all these kinds of indicators. On critiques of monetary policy like that of Larry Summers, Summers might be correct that there's too much stimulus. To me, it's hard to say. I'm reasonably content personally, with Fed policy. From my perspective, the markets are signaling fairly low inflation going forward, but it's really hard to say, because during the 1960s, obviously when we went into the high inflation period, you can find parallels where the initial increases were viewed as temporary, and it'll get back to normal soon, and then it really didn't get back to normal, right?

Sumner: I think ultimately what you have to look at is, do the markets continue to have faith in Federal Reserve policy? Does the Fed continue to have credibility on its long run 2% average inflation target? That's really the key thing that we have to watch. If it comes to the point where they start to lose credibility, then clearly the Fed has done too much stimulus. For now, it does seem like markets still expect inflation to go back to close to 2% not to distant future. But certainly the inflation numbers have been a little higher this year, or quite a bit higher than a lot of people expected.

Sumner: It's a little hard to tell how much of that is temporary supply imbalances, like used cars and things, and whether some of that is just a little bit too much demand. Because COVID has just messed up so many traditional indicators, done so many strange things to the labor market, to the supply of parts and other aspects of the macro economy. So, I'm very skeptical of any pronouncements based on real economic data as a guide to monetary policy.

Beckworth: Yeah. All future studies will have to put some dummy variables in for this period.

Sumner: Exactly.

Beckworth: When looking at time series or panel data, we're going to be very careful with what is happening. Okay. Those are things where you've touched in your bubble story. The Fed moving towards level targeting, the Fed relying more on asset price signals than traditional theoretical models to guide policy. One another thing that comes out in your book is that the Fed should think carefully about aggregate demand shocks, but when it arises, aggregate supply shocks as well, and that seems to have come to fruition this past year, too.

Policy Relevance of Aggregate Demand (and Supply) Shocks

Sumner: Right. One point I make in the book is, I think I have a phrase like, I'm not a supply sider or a demand sider. I'm a supply and demand sider. So, I think both sides of the aggregate supply and demand model are important. And both, during the Great Depression and the Great Recession, I believe the initial contraction was caused by negative demand shock. In both cases, I think the recovery was slowed a little bit by some unfortunate supply side problems, more in the 30s than in the Great Recession.

I'm not a supply sider or a demand sider. I'm a supply and demand sider. So, I think both sides of the aggregate supply and demand model are important. And both, during the Great Depression and the Great Recession, I believe the initial contraction was caused by negative demand shock. In both cases, I think the recovery was slowed a little bit by some unfortunate supply side problems.

Sumner: But I think we're seeing, with COVID, a perfect example of the importance of both the supply and the demand side, because clearly there was a big drop in nominal spending in 2020 with COVID, but also we're seeing some real supply problems, especially in the recovery. I don't think it's possible to look at this picture just from a demand side framework. That is, anyone who's looking at the picture and saying, well, unemployment is 5.2%, and it should be three and a half percent and therefore we need more monetary stimulus, I think is missing part of the picture.

Sumner: There are some changes in the labor market that are making it more difficult for firms to hire workers. I think that we have to be cognizant that the economy is being buffeted both on the supply and the demand side. That makes things a little more difficult. Although I'm a believer in nominal GDP targeting, I think during COVID the appropriate policy would have been to aim for nominal GDP to be back on track, out in the future, a year or two in the future, not like month by month.

Sumner: They shouldn't have tried to prop up nominal GDP in April of 2020, because that's just not reasonable policy with a 14% unemployment, right? You couldn't print money to get those people back to work if they're sent home because of COVID. And if you tried to maintain nominal GDP growth, when real output drops that sharply, you would have had really high inflation, and it still wouldn't have created jobs. The reason why nominal GDP targeting makes sense normally is because normally, when there's a big rise in unemployment, it's because people aren't spending enough so there's not enough nominal revenue to get the workers reemployed.

Sumner: But if the workers are being unemployed because of health risks, nominal spending doesn't solve that. I think the Fed appropriately let nominal GDP fall sharply for a few months and then aimed to get it back on target in a reasonable period of time. It looks like now, in the second half of 2021, we are seeing nominal GDP returning to that previous trendline appropriately.

Beckworth: Okay. Well with that, our time is up. Our guest today has been Scott Sumner. Check out his new book, The Money Illusion. Scott, thanks for coming on the show again.

Sumner: Thanks for inviting me, David.