Jon Hartley on the Legacy of John Taylor and his New Measure of R-Star

Jon Hartley is a macroeconomist and affiliated scholar at the Mercatus Center. Jon returns to the show to discuss the most recent Hoover Monetary Conference, the legacy of John Taylor, why central banks should be using his new measure of r-star, the status of debt management at the US Treasury, and much more. 

Subscribe to David's new Substack: Macroeconomic Policy Nexus

Read the full episode transcript:

This episode was recorded on August 26th, 2025

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: Welcome to Macro Musings, where each week we pull back the curtain and take a closer look at the most important macroeconomic issues of the past, present, and future. I am your host, David Beckworth, a senior research fellow with the Mercatus Center at George Mason University, and I’m glad you decided to join us.

Our guest today is Jon Hartley. Jon is a macroeconomist and an affiliated scholar and colleague of mine at the Mercatus Center. Jon is also a returning guest to the show, so check out his previous appearance if you haven’t already done so. Now, Jon, along with Michael Bordo and John Cochrane, have a new paper on John Taylor’s contributions to economics, and this paper was presented earlier this year at Hoover. Jon joins us to discuss this paper as well as give us an update on r-star and debt management at the US Treasury Department. Jon, welcome back to the program.

Jon Hartley: David, an honor to be here.

Beckworth: It’s great to have you on. Now, quickly remind the listeners, you also have a podcast, great podcast I listen to as well. Tell us about that briefly in case listeners don’t know. I suspect many of them do already, but just in case they don’t, what is it?

Hartley: Absolutely. It’s the Capitalism and Freedom in the Twenty-First Century Podcast. It’s hosted at the Hoover Institution. It’s the official podcast of the Hoover Institution Economic Policy Working Group.

Beckworth: Check it out, folks, if you haven’t already. These are the two podcasts every week you should check out and make sure you’re up to speed on your macro. Now, you have this great paper, Jon. You presented it at the conference. Actually, there were two conferences. There was a John Taylor Conference on one day, and then right after that, there was another conference, the Hoover Monetary Policy Conference. It happens every year. You were part of this paper during that first conference on John Taylor.

Now, you worked with Michael Bordo, John Cochrane. It’s a really neat paper. You even collected a lot of statistics on his citations, a number of articles. It’s a really fascinating paper, and I want to get into it. Before we do that, Jon, I wanted to do something I did with you last time. Last time we were together, we had just finished the Shadow Open Market Committee Conference, and we reflected on the conference a little bit. Any highlights from the Hoover Monetary Policy Conference you want to share with our listeners?

Hartley: Absolutely. I think, one, there were many fantastic talks on both the John Taylor celebration and at the Hoover Monetary Policy Conference. In reality, both of these were celebrating John Taylor and his contributions, which are quite incredible and substantial when you think about it in the long run. I would preface it, and the beginning of our paper says this, which we’ll talk about in a little bit, but John Taylor truly is one of the greatest macroeconomists of the past 50 years.

I think a lot of these papers and talks across both days really explain why that is. I think, obviously, for those that follow monetary policy, the Taylor rule’s become incredibly popular. There are some panels that were focused on that. For me, I really appreciated both Ed Nelson’s dinner talk, which connects the Friedman monetary economics legacy to John Taylor. 

I also really appreciated Dr. Condoleezza Rice’s talk as well. She spent a lot of time talking about John Taylor’s contributions to public service. John actually served in Washington in several administrations. Most recently, he served as undersecretary for international affairs during the George W. Bush administration. Some people may not be aware of this, but John spent a lot of time actually helping set up the central bank in Iraq after Saddam Hussein was deposed. There’s a lot of John Taylor’s career, I’d say, outside of just his academic contributions, which are important. Also, just thinking about where John Taylor fits into the history of monetary economics, I think, was well appreciated. I appreciated those talks for those particular reasons.

Beckworth: When we get to this paper, I want to share my story because John Taylor was my boss at the US Treasury Department during one of those stints. Condoleezza Rice mentioned a fascinating story. I hadn’t heard about John Taylor, and I think it was Michael Boskin, were important in helping the Polish currency survive the transition from communism to democracy. 

Before we get to that interesting conversation, just a few thoughts here from that conference. Again, I love this conference. Every May, it happens. Encourage folks, if they can, to get out to it. Probably the most exciting panel I heard there, and I’ve actually had some podcasts follow up because of this, was the digital money, digital payments panel. In particular, Luis Garicano was on one of them. His presentation was so mind-blowing for me. Dealt with dollar-based stablecoins, how it was overtaking Europe. The Europeans are freaking out. He came on the podcast later. We’ve had a fun conversation. I did a Substack on it. In fact, I’m doing my own interview later this week about stablecoins in Europe and other issues. That was really fascinating. 

Then I have to mention, as well, Jon, Peter Ireland’s paper, because he talked about nominal GDP. His point, and I have to mention this, is that even though the Fed is now a flexible inflation targeter, and maybe time permitted we can talk about this new framework, but now that they’ve returned officially to flexible inflation targeting, his point was they could always cross-check themselves by looking at how fast the dollar size of the economy is growing. I think that’s just a real practical point. A lot of fantastic panels there. Of course, the fiscal panel with John Cochrane is very important because we’re getting closer and closer, I believe, to the cliff here in fiscal dominance. 

John Taylor’s Contributions to Economics

Let’s segue into your paper. It’s a great paper. We’ll provide a link to it in the transcript. I want to mention upfront what you just said. You said he was the greatest macroeconomist since Milton Friedman, at least over the past 50 years or so. In your paper, Rich Clarida says that the two most important monetary papers were Taylor’s 1993 paper where the famous Taylor rule emerges, comes out, but then also Milton Friedman’s 1968 presidential address at the AEA meetings, also a very important paper. Those are two very consequential, heavy-hitting, long-lasting influence papers. It was neat to see Rich make that point. 

Why don’t we step back and tell us, how did you guys get together and come up with the idea for this paper? Who said, “Hey, let’s dig up all the records, let’s collect and summarize his impact on the profession?”

Hartley: Absolutely. I’ve been working at Hoover with John Cochrane as well as with Michael Bordo for a number of years now. One thing that’s often come up, and it’s so wonderful to see, is we have so many wonderful economists at the Hoover Institution, and they truly deserve a celebration of their work. We did a celebration for Bob Hall’s work last autumn. We just did one for John Taylor, which we’re talking about now, and we’re doing actually one for Thomas Sowell in October.

One, these are economists that certainly deserve a Festschrift-like celebration, and that’s how all these came together. In particular, with this paper, I think, really, I’m an empirical macroeconomist. I love using data to understand things, certainly macroeconomic dynamics, and using identification. That’s my style with things. I think sometimes it’s just fun. I think if we’re macroeconomists, obviously, schools of economic thought are important to think about a little bit. In my mind, quantifying the impact of economists is very interesting, an interesting way to think about the impact of ideas over time.

That’s how this all came together, and my contribution to the paper with John and Mike was really putting together a lot of this data. We really just try and summarize a lot of the other papers being presented about John’s career, whether it’s academic or public service-related. Some of the graphs that we have are wonderful because they show, for example, John Taylor’s citations over time, and they also show, for example, John’s influence in FOMC deliberations. We went back to all the FOMC transcripts and really just looked at every single mention of John Taylor or the Taylor rule, and just counted those.

It turns out that it’s actually been increasing quite a bit, in particular during the 2010s, and it’s been increasing pretty steadily. I think, really, just getting back to the citations and John Taylor’s major academic impact, there’s really two main areas. We talked a little bit about the Taylor 1993 rule. There’s also the Taylor 1999 rule. That’s a slightly bigger output gap. That has quite a few citations as well. John Taylor’s citations, unsurprisingly, explode with the 1993 Taylor rule. Really, it’s just a beautiful way to describe monetary policy, that interest rates are both dependent on inflation and an output gap or some deviation from, say, full employment.

One, just qualifying that the stance of monetary policies is really critical. Obviously, there’s an r-star component there as well that goes into the Taylor rule. What I think is also fascinating and perhaps maybe not as well appreciated about John Taylor in terms of how he fits into the history of macroeconomic thought is, John Taylor also played a very important role with coming up with sticky prices basically in the early 1980s. To give you a full and maybe very condensed version of macro history in the past, maybe 50 years or so, you had this period of rational expectations, revolution in the 1970s, lots of inflation.

Basically, old Keynesianism had died. Some people still believe, and a lot of people still believe, that money was not neutral and that monetary policy could have real effects in the economy. This was a problem because rational expectations models didn’t have monetary policy producing real effects. What did John Taylor think? He said, “Well, if you have some nominal rigidities, we can create real effects.” That’s where Taylor’s 1980 paper, or Taylor contracts, comes in, and it’s one of his most cited papers. It’s really important.

Now, a lot of macroeconomists, a lot of New Keynesian modelers today use Calvo contracts that came out a few years later, in 1983, when the Calvo contracts paper was published. There’s also Rotenberg contracts as well. Now, Taylor’s contracts are based on fixed time intervals, whereas, long story short, Calvo pricing has basically fixed probabilities. It turns out that the Calvo pricing method was a lot easier, more tractable to use a model to produce basically closed-form solutions.

It’s a long story short, but John Taylor Contracts were there in 1980, a huge contribution in terms of early New Keynesian modeling, how to get real effects, how do you have monetary non-neutrality in macroeconomic models. John Taylor was there at the very dawn of New Keynesian economics. Very important. Then obviously, fast forward to 1993, there was also this problem going on in the 1980s, which was, is it money? Is that the key policy instrument or interest rates?

John Taylor really came in, demonstrated that interest rates were really the tool that central banks use to set monetary policy. There were several other papers that were, around that time, trying to get at rules-based policymaking and how to describe interest rates, but John Taylor was really there first to do this. Obviously, one thing I guess some people are maybe not familiar with, but it both has this ability to describe policy but there’s also a theoretical component as well that makes it very nice or tractable, which allows you to get equilibrium determinacy.

Very wonky, but that theoretical contribution as well, I think, is part of what made the Taylor 1993 rule so popular and so famous. I know a lot of people, when they hear the Taylor rule, they think, “Well, the Fed doesn’t really follow the Taylor rule,” and there’s been all these debates about how closely it follows it or not. If you run through that empirical regression, and obviously, what r-star we’re using matters, whether you’re using Laubach-Williams or some other measures matters, what kind of output gap does the Fed use? How are you even measuring the output gap? That matters too.

At some level, this whole idea of ensuring that central banks don’t necessarily follow a Taylor rule mechanically to a T, and to be fair to John Taylor, he’s never really advocated for that, but I think it’s worth thinking about. For example, in recent years, the Federal Reserve, in the inflation of the early 2020s, deviated quite a bit from the Taylor rule, I think, regardless of what assumptions you’re putting in there. I think the Taylor rule perhaps looks pretty good coming out of that.

There’s a paper that was presented by Emi Nakamura at Jackson Hole this past week that takes a counterargument to that, that the Fed was wise to look through the inflation of the early 2020s as a supply shock. But I think at some level, if you look at the long-run history of interest rate setting in the US over the past 50 years, and you look at what the Taylor rule says and prescribes, it’s an excellent way of describing monetary policy and the Fed’s monetary reaction function.

Beckworth: Jon, there is a lot there to unpack, so let me do it piece by piece here. Let’s first go back to those charts you mentioned, and I encourage listeners to go look at these charts that you put a lot of work into, I can tell, Jon. But really fascinating, as you mentioned, a lot of John Taylor’s big contributions, they begin to be seen later. In 1993, he has that prominent paper, and then the citations to it blows up, and you mentioned his later balanced approach also blows up. It’s a really cool chart. You see his lasting influence. 

You see how a paper, when it first comes out, may not have as big of an impact until later, and this paper clearly does. It has this lasting influence. You mentioned Jim Bullard from the minutes or the transcript at an FOMC meeting where he talks about how influential John Taylor’s Taylor rule has been. It’s worth noting there’s the Taylor rule, there’s the Taylor principle, which says you need to raise interest rates more than one-for-one with increase in inflation. Then there’s the Taylor curve. He has three fairly well-known ideas in macroeconomics named after him. Very, very successful.

Hartley: The Taylor contract.

Beckworth: Taylor contract. Make that four. I want to go back to that, the staggered wage setting of that. You mentioned Harald Uhlig in the paper. He had his own discussion at this conference, and he made the case that maybe sticky wages or staggered wages might be more relevant. We defaulted on Calvo pricing because it was easier to work with, basically, is what he argued, mathematically more tractable, but maybe more realistically, as we go, something like wages.

Let me step back from this, Jon, and ask you a question. You’re a young macroeconomist. You say you like empirics. What frictions do you see as most important? Is it sticky output prices, sticky wages? Some might say sticky information. Some might say sticky nominal debt contracts. Where do you think is the more important frictions that prevent markets from clearing instantaneously at the macro level?

Hartley: Absolutely. I think it’s fair to say I’m definitely a bigger fan of, I’d say, using sticky wages or sticky information. This is part of a forthcoming paper of mine, which is, I think sticky prices has been a bit of a misadventure in the sense that, one, if you go back to Keynes’s general theory, he really talks about sticky wages for the most part. I don’t know if this was prophetic or not. 

One, I think the reason why we largely have sticky prices in the Keynesian models today is because of this model tractability. It’s a lot easier to simply just have Calvo pricing. You can get some very nice closed-form solutions pretty easily. Something that I think is pretty obvious to someone living in 2025 is that in many costs, prices aren’t that sticky anymore. Whether it’s your Uber rides, your Lyft rides, or your Amazon prices, they’re changing quite frequently and more frequently than offline prices, but now, even for what it’s worth, if you look at some of these statistics, you’ll find that about 20%, almost, of all retail sales in the US are now online.

Many costs have gone to almost zero for about 20% of US retail. About 50% of Chinese retail, for example, is online. Now, what you’re also seeing at stores is you’re seeing things like electronic shelf labels. Walmart, for example, has said that they’re going to put electronic shelf labels in half their stores worldwide by next year. These are the little electric or screens that are effectively now price tags, have replaced paper or cardboard price tags. Now you can see them in many grocery stores, that they’ve become quite commonplace. In my opinion, that means that many costs are falling quite a bit.

For just an empirical reason, I think we should start thinking about are we reaching the end of sticky prices? Are we reaching effectively zero money costs? I think that’s a good reason perhaps to really shift more toward sticky wages or sticky information of some sort. Behavioral macroeconomics is very much focused on the sticky information idea. There are older papers that have focused on sticky information, think Mankiw-Rice. There are many people calling for shifting to sticky wages over the years.

There’s some quite influential models that do this. I think the time is now to make that shift. If I can in any way convince people that there’s an empirical case to do it, even though our models may be harder to solve, I would be happy to do it. I think that that would be a worthwhile effort.

Beckworth: You’re saying in a world of increasingly widely used dynamic pricing, where literally demand-driven changes in prices do take place because of technology, AI, all those things, we need to abandon the Calvo ferry and use these other tools, these other approaches. You mentioned sticky information. That makes a lot of intuitive sense. I think we saw that during the pandemic. There was an interesting paper that was done that showed that most people really don’t pay much attention to inflation until it hits some kind of threshold where they’re triggered. They’re information-insensitive up to a point, and so suddenly it matters when it matters. 

Those examples, I think, underscore the sticky information point. I like myself, also, the sticky fixed nominal debt contract story. Why haven’t, for example, tips taken off widely? Why do we still see a lot more in nominal Treasury debt contracts? I have to infer that, for some reason, people like fixed nominal units on their bonds, on their money. That friction may always be there. Then sticky wages as well.

Interesting conversation, Jon. We look forward to you swaying the conversation in that in future years. You mentioned the Nakamura paper at Jackson Hole. She mentioned, for example, the ’21/’22 deviation. Maybe it wasn’t such a bad thing. Maybe that’s just in a world where you have a huge war effort, so to speak. 

Let’s go back to the papers that were presented at Hoover several years back by George Hall and Tom Sargent. They compared the pandemic to World War I, World War II financing. You have to finance it somehow, and you break the norms. In fact, you had a paper by Robert Barro. You mentioned this in your discussion where he called the Taylor rule a contingent rule. In normal times, you use it, but outside of normal times, it’s like the gold standard. We suspended it during the Civil War. What’s wrong with that thinking? Why not say, “Okay, Nakamura, she made a reasonable point.” Why would you push back against that?

Hartley: I think this comes back to this question of what caused the inflation of the early 2020s. There’s several different explanations. I think, for me, it’s both a supply and demand explanation that, sure, yes, there was a supply constraint, but you also had a surge in demand as well. I think it’s fair to say that stimulus payments had something to do with that. There’s different, essentially, observationally equivalent models.

You could take the framework of, say, Larry Summers and Jason Furman, who said, “Well, there’s all this stimulus that was much bigger than the Keynesian output gap that you would measure, and that should create inflation.” You could take the fiscal theory at the price level, thinking of John Cochrane, and you could say, “Well, at some level, there was no plan to pay it back, and so it’s an inflationary event, and this worked through some expectations channel that expected future surpluses weren’t there, and so there has to be some inflation to basically make the fiscal theory equation hold.”

You could just take the monetarist approach, the MV=PQ. M jumped a lot. In my mind, I think it’s pretty clear that the Fed acted too late in that, sure, we can celebrate now that, sure, we didn’t get some recession following this tightening, but at some level, we still had this 20% increase in the price level, or greater than that. I think that’s not something to celebrate in any respect.

I think if the Fed had acted earlier, in October of 2021, when it was very clear that the inflation story was not something about used cars or something that’s very one-off and strange, that it was very clear that owner-occupied rents were rising, that at that point, that that was the point to begin tightening, not six months later when the Fed ended up doing so. I think the whole narrative of team transitory was so strong, and it was carried by people in the media. It gave the Fed this cover.

Of course, nobody wants to necessarily be the Paul Volcker that causes a recession. It’s a very difficult challenge, and at the time, I think it was a very difficult narrative to fight against. This was just something born by this one-in-a-100-year pandemic. At some level, I do think it’s hard to ignore the politics of some of these things. I think it’s very, very difficult at some level to want to raise interest rates, to be the one that does that, like the Taylor rule in this case, would prescribe.

A careful maybe thought experiment, I think if Donald Trump was president in April, May of 2021, and had there been a big inflationary uptick, I think the narrative might have been a little bit different. I think people would have, for political reasons, tried to attach the inflation narrative to Trump at that time, called the Trump inflation, rather than some team transitory supply shocks narrative. Perhaps that’s a somewhat cynical way of thinking about things. I think the reality is political narratives matter for economics to some degree, in trying to understand how we think about the macroeconomy.

We don’t have the time to really do these sorts of excellent studies with excellent identification on the fly. Policymakers really have to think about things very quickly and not necessarily have perfect answers. In my conversation with Rich Clarida, he’s made these sorts of comparisons where there’s the difference between being an academic, where you’re getting a very precise answer, versus being a policymaker, where you have to have an answer that’s good enough. Anyways, I think it would have probably been better to follow the Taylor rule a little bit more closely. I think at some level, had some thought beyond the groupthink, we wouldn’t have had as worse of an inflation that we had.

Beckworth: I’m sympathetic to your point that a lot of the inflation was demand-driven. I would argue maybe even more so than you would, and I think fiscal policy, monetary policy, they definitely played a role. In fact, I have a working paper out that I’ll link to in the transcripts. I also think, though, that there was a lot of uncertainty, say in 2021. The vaccines hadn’t come out. I think you could view the largest of fiscal policy as an insurance policy. You just mentioned in real time, a lot of uncertainty, they threw everything against the wall.

In retrospect, yes, we could have maybe calibrated that better. I want to be gracious in the sense that, in real time, I think it’s a little trickier. Again, I view this as the Taylor rule is wonderful, but it’s contingent. If you’re in a major global crisis, I think you can suspend that, given the world you’re in. Now, I think the counter-argument you just made is a good one, though. At some point, though, you leave that world. You have to start following the Taylor rule again. I think you fairly point out they could have started tightening sooner. 

Let’s move on, though, to the other parts of your Taylor paper. We’ve talked about the Taylor rule, we’ve talked about his work on macroeconomic theory and modeling, but he also did a lot of work on international macro. Tell us about his work there.

Hartley: John Taylor spent a lot of time working around the world, essentially as a consultant with many central banks. John Taylor, in particular, has made a lot of friends at the Bank of Japan and around the world. I think what’s really important with John Taylor, and this, I think, gets back to what I said earlier, John Taylor knew how to sell an idea. I think it’s a great model for academics who, out there, are trying to sell ideas to policymakers. He was a policymaker himself, but he also, over time, built a network of monetary policymakers around the world.

For example, when we were at the Hoover Monetary Policy Conference a few years ago, we both met Kuroda. It’s amazing seeing some of the inner workings of some of this. The Hoover Monetary Policy Conference is, in many respects, a culmination of John Taylor’s networking across central banks around the world. One, it’s wonderful to see central banks from countries around the world come to this conference, but at some level, the Taylor rule is something that I think has had a good degree of influence in central banks around the world.

Where did this all begin in engaging with central banks? It really began at the Fed or with the Federal Reserve. I remember talking to Don Kohn and how, in the early 1990s, there was, I think, a seminal moment where Don came to Stanford and met with John. It began this process of really John Taylor’s amazing, I think, engagement with policymakers. He served as undersecretary for international affairs, which is a very critical role in terms of international economic policy. There’s too many to list.

Again, as I mentioned earlier, helping to set up Iraq’s central bank. Truly, John’s project of monetary policy rules really was an international one. I think that’s really what I would, I think, emphasize is that the influence of the Taylor rule is international. It’s not just specific to the US. Sure, there might be slightly different weights and different central banks, different policymakers have different reaction functions, but the Taylor rule is an international and global phenomenon.

In my opinion, that’s why we say at the beginning of the paper that he’s one of the greatest macroeconomists of the past 50 years. In my opinion, if there’s one macroeconomist who we’ll be talking about or people will be talking about in 50, 100 years from now, I would be willing to bet that John Taylor is one of those macroeconomists.

Beckworth: Oh, definitely. You get to work with him, Jon. You get to interact with him even now. I, too, have been blessed to cross paths with him. My story, real briefly, I went to the US Department of Treasury right out of grad school. John Taylor was looking to hire some PhDs, and he took a chance on me. I didn’t come from a great pedigree university. I go in there, and it was just great. He wasn’t my immediate boss. He was the undersecretary. I was several layers down. He engaged with me, very encouraging.

Just being able to connect with him, I think, was consequential because after I left Treasury, I did a paper with a friend on the international linkages of monetary policy. John Taylor really liked that. He cited it several times in his ongoing work on international economics. Then, Jon, I invited him onto my podcast. Back in 2016, when it first started, he was, I think, one of the first two or three guests on the show. He agreed to come on because he knew me from Treasury, and he loved the paper that I did.

We had a great time. It went well. Then, a few days later, I get an email from him. He says, “Hey, David, would you like to come out to the Hoover Monetary Policy Conference and present if you have something on international economics?” I was like, “Yes, definitely.” I got the invitation to present at the wonderful conference you just mentioned, the Hoover Monetary Policy Conference. My co-author, Chris Crowe, and I, we recreated a paper, or updated a paper that had looked at the composition of assets that foreigners hold on the US. We issued dollar bonds, repoed bank accounts. We just showed the US as this banker to the world. We updated this paper by Hélène Rey and Pierre-Olivier Gourinchas. 

During the Q&A, there’s, of course, people asking questions. This one guy raises his hand and goes, “I commissioned that original paper, and I love what you’re doing.” Guess who that was. It was Rich Clarida. Then, later, I got to chat with him. Just from that story, I just met a lot of interesting, fun people. I really owe a lot to John Taylor myself, my own career, just opportunities, connections. I want to recognize his work, too. 

I just want to briefly mention one more thing on the international side. You mentioned this in your paper that he made some important contributions, theoretically, in the sense that he had these big models that looked at cross-country systems. He found that instead of targeting exchange rates or trying to respond to another country’s central bank, if each country had their own version of a Taylor rule, that was the best outcome you could hope for, which is, I think, what you’re alluding to, everyone doing their own Taylor rule. That was a really fascinating story.

Hartley: Absolutely. I think, again, it can’t be overstated that John Taylor’s career, he wasn’t the type of academic that would just write an academic paper and publish it and be done with it. He would actually make the trips to the Bank of Japan. He would make the trips to Europe. He became a consultant to many central banks around the world. He really believed in being both an academic and a practitioner. It’s something that I think is maybe increasingly becoming rare.

I think, at some level, a huge part of how the Taylor rule became so influential was that—there were others out there that I would say had, let’s say, similar ideas, I don’t want to spend too much time on that—but John really followed through. Rather than just and only publishing the paper, he actually went out and engaged with the policymakers. I think that’s an excellent model to follow for those that are trying to sell ideas.

Beckworth: Absolutely. Listeners, again, we provide the link to Jon Hartley’s paper in the transcript, so be sure to check it out. I know it will be published in a volume soon, if not already, correct?

Hartley: Yes. This will appear in essentially the conference volume that’ll be published by the Hoover Institution Press.

Better Measure of R-Star

Beckworth: Be sure to check that out. Now, Jon, I want to move on to some of your other work that you’ve done. In fact, you’ve done an interesting paper that attempts to get a better measure of r-star. There’s a lot of popular ones. You already mentioned the Laubach-Williams r-star from the New York Fed. Then there’s the Holston-Laubach-Williams measure versions of it. There’s the Lubik-Matthes measure. The FOMC, you could also look at that. They have their long-run real federal funds rate, which could approximately be the same thing.

Then I like to look at five-year, five-year forward tips. What does the market think? Of course, those tips have liquidity, premiums. They’re not perfect either. No measure’s perfect. I’m looking right now at the latest data. The five-year, five-year forward is about 2.5% all the way down to about 0.5% or 1%. Help me make sense of this and tell me why we should look to your new measure to help clear out what really is going on with the star.

Hartley: Absolutely. To think about the history of r-star, it’s, I think, quite appropriate that we begin this conversation talking about Laubach-Williams after talking about the Taylor rule, because it’s very much related. It’s worth mentioning that John Williams is a student of John Taylor’s, also a Stanford economics PhD graduate. Essentially, there’s this challenge, which is, what is the neutral rate, and how do we define the neutral rate? The neutral rate is the rate at which there’s neither accelerating inflation nor decelerating inflation. I’d say maybe a very quick summary of it. 

What’s essential with the neutral rate is, it allows us to estimate where this current stance of policy is. Is the central bank’s policy tight or is it loose? Is it accommodative or is it restrictive? We can’t really assess this without some estimate of r-star. It’s a very important metric for central bankers. R-star is something that comes up very often in both monetary policy circles and also on Wall Street. Because Wall Street, for example, if they’re trading bonds, they want to know what the value of their bonds are going to be worth. Certainly, they’re impacted by interest rate risk and what the Federal Reserve does. So, r-star, something that’s very central.

Laubach-Williams, in the early 2000s, was the first really popular way of estimating this. Then there were several follow-ons. Laubach-Olson-Williams did something similar, expands it to several more countries. You have Lubik-Matthes. These are all what I would call model-based estimates of r-star. They’re essentially reverse-engineered out of a three-equation New Keynesian-like model. That’s essentially what’s going on. You need the benefit of hindsight and history to be able to assess this, to estimate them. That’s essentially what’s going on. My critique or challenge with these model-based estimates would be, what if the model’s wrong?

Economists like to say, “What if the model’s mis-specified?” It’s a fancy way of saying, “What if the model is wrong and doesn’t capture reality?” One thing that I think is worth looking at and turning to as really an alternative are survey-based measures. These are, at some level, atheoretical in that they don’t come out of a model. The question is, why not just ask people what they think r-star is, and why not just take the median of that estimate? As you mentioned before, the FOMC has been providing something like this since the beginning of forward guidance in the early 2010s, with the beginning of the survey of economic projections of the SCP.

Remember the dot plots. If you look at the right-hand side of the dot plot, there is this long-run estimate of what they believe the natural rate is. We don’t know whose dot is whose, but we can take the median of that. We can assume that’s the median FOMC participant estimate of what they think r-star is. I think that’s at least heading in a slightly better direction. Something that was created in the early 2010s was something similar, which was in the New York Fed Primary Dealer Survey, they started asking questions to market participants.

Who are market participants? They’re people who are, for example, working at primary dealers, think the Goldman Sachs or the Morgan Stanleys of the world, their trading arms, or think increasingly hedge fund analysts. More recently, the New York Fed Primary Dealer Survey started asking questions to market participants what they thought the neutral rate would be. They started this in 2013. Just this year, they’ve expanded this survey. It’s now changed names. They merged it with a similar survey that was essentially asking market participants, not just primary dealers. It has a new name now. 

Similarly, there’s the ECB. They, in just the past few years, in their survey of monetary analysts, have started asking similar questions about the neutral rate, the Bank of Canada, and the Bank of England. They started something called the Survey of Market Participants. They’re all asking similar questions about essentially what you think the neutral rate of interest is, or what the long-run neutral rate is, or the long-run equilibrium rate. They’re worded slightly differently. All these questions get at, what do market participants think r-star is.

My argument is simply this: Why not just look at the median estimate of what market participants think the neutral rate of interest is, what they think r-star is? Why not use that in our models, or just in general, if you’re a policymaker, to assess what the stance of monetary policy is? There’s a few things that we learned, and this is all cataloged in my paper on this, “Survey Measures of the Natural Rate of Interest,” is the title. One, we learned that the survey median of market participants follows the median r-star estimates of FOMC participants very closely.

One, it’s hard to know exactly what’s going on, whether, say, FOMC participants are following, for example, the median estimate that’s being provided by the market participants, or whether the market participants are looking to the FOMC participants, what their r-star measure is, or maybe they’re looking at models, Laubach-Williams, Lubik-Matthes, and they’re informing their assessment of what r-star is. 

The point is, say we’re in a situation right now that, let’s say, as of today, the Fed funds rate’s 4.25%, Lubik-Matthes says that r-star is around 4%. Saying that we’re essentially at a neutral sense of monetary policy, or if you looked at Laubach-Williams today, Laubach-Williams or Laubach-Olson-Williams would have r-star at around 2.75%, which would suggest that monetary policy today is very restrictive, with the Fed funds rate around 4.25%. However, if we were to look at the median estimate of r-star from market participants from that New York Fed survey, we’d find that it’s roughly around 3.5%, that their estimate of r-star is somewhat in between the Lubik-Matthes and the Laubach-Williams measure.

It’s also very close to what the FOMC participants think. I think it’s better to follow that estimate than maybe one or another model. The point, broadly speaking, is, one, an atheoretical estimate is something that we should think about. It’s a nice data point to have. It’s only a few central banks right now that actually are measuring this. I’ve been actually having several conversations with central banks around the world. If there are people who work at foreign central banks who are listening to this podcast and would like to create a survey question about the neutral rate of interest or r-star, I would be happy to talk with them.

I think over time, what we’re going to see is we’re going to see more central banks launching these sorts of surveys asking questions of market participants. These surveys ask market participants many more questions than just about r-star. Really, essentially about every macro variable under the sun, but that we can build up a big enough time series of estimates that we can start to rely on these sorts of atheoretical estimates of r-star rather than have to rely on these models that may be mis-specified. It doesn’t mean we should get rid of the models.

I think it’s important that we have them. It’s important that market participants, FOMC members, central bankers, can still look at them to form their assessment of what r-star is. I think it’s nice if we can have a median survey-based estimate off the shelf. That’s the whole idea within Hartley 2024, “Survey Measures of Natural Rate of Interest.”

Beckworth: We’ll provide a link to that, folks. It’s a Mercatus working paper. Jon, it just hit me while you were saying this, I now know why President Trump wants the rate cuts. He’s been looking at the Laubach-Williams measure. Clearly, he’s a follower of r-star himself and believes the right estimate is the Laubach-Williams measure. On a more serious note, though, why not look at market measures? Why not turn to what one might say is a measure where people have skin in the game? They’re literally trading funds. They could lose money. Why rely on a survey versus an actual market measure?

Hartley: It’s a great question. I think the common critique would be something that you just pointed out before, which is that there’s a challenge with risk premium, all sorts of things that can bias these sorts of estimates. You could make an argument, for example, is an inflation break-even for example, really a great estimate of inflation expectations? Sure. I think there’s a good case to be made. If you look at these crisis-like periods where the tips real yields go crazy and nominal rates, and one goes more crazy than the other, then the break-evens really get out of whack really quickly.

Is it that people are changing their inflation expectations suddenly? No, I don’t think so. I think there’s just a huge liquidity shortage, essentially, in the inflation-linked bonds is what’s going on during these periods. It’s a great question. I think also there’s an analog with the term “premium” as well. The term “premium” is another one of these esoteric measures. You could argue that we should have a survey-based term premium instead of something that’s, say, a market-based one to just look at, for example, something like a term spread or another term premium-based model. There’s several of them out there. Some are produced by the New York Fed, that if you want to get away from this challenge of getting risk premium out of your models that can bias them, can bias the expectations, then maybe we should stick to survey-based measures. 

Again, there’s all sorts of biases that go hand-in-hand with surveys, and I’m not saying that they’re perfect, but I think, one, it’s something that’s new. It’s an innovation. We didn’t have these sorts of central bank surveys prior to the financial crisis. It’s actually only in recent years that central banks like the ECB, the Bank of Canada, Bank of England, have started these sorts of surveys, but I think they’re very useful tools.

We can also get measures of uncertainty as well. We get the 25th, 75th percentile of what the market participant estimate of r-star is, but I think for that reason, maybe another good reason to consider these survey-based measures of r-star and the term premium for that matter, rather than, say, market-based measures or model-based measures.

Beckworth: Absolutely. When you turn to market-based measures, there are attempts to fix them, to clean out the liquidity premium, the risk premium. Think of a popular one by D’Amico, Kim, and Wei. They call it the DKW measure, but then you’re back to modeling. You’re back to the very problem you were trying to get away from in the first place, because they have to impose some structure on the data in order to get this refined measure. You’re like, "look, let’s just cut to the chase. Let’s ask people directly what’s going on.”

Hartley: Exactly. Something very simple, hard to do, and central banks deserve a lot of credit for doing this. There’s a lot of other wonderful information that’s in these surveys. The hard work of the paper, for me, is really just putting together all these estimates, going through every single time these surveys have been put out there. In the case of New York Fed, it’s been many years. Really, no one had really compiled these together. That’s the contribution of the paper. Building a time series. This time series of survey-based measures from my paper is available on my website, jonathanhartley.net. You can go to the data section. You can find the latest measure of survey-based r-star for the US, the Euro area, Canada, and the United Kingdom. You can find it there. 

The idea is to build up a time series that people are going to use. That’s, again, something trying to market to central banks around the world, where this would be a good idea, but maybe in the same vein that John Taylor was with the Taylor rule. This is, I think, a much more modest attempt, but really just trying to make the case that, again, three-equation New Keynesian models are wonderful thought experiments, but may not be perfect from a quantitative standpoint.

The same challenge may exist with market-based measures, but at least you have people filtering these things and coming up with their own survey-based estimate.

Beckworth: Yes. Use all the information available and use this new innovative approach that Jon is suggesting and has worked on. Check out his website.

The Government’s Debt Management Policy

Now, Jon, in time we have left because we’re nearing the end of the program, I want to return to a topic that we discussed last time we were on this podcast together, and that is the government’s debt management policy. How does it determine the structure of its debt, its average maturity?

When we were together last time, we talked about these claims that the Yellen Treasury under President Biden was going more towards bills, trying to shorten the weighted average maturity. I’m just wondering where do we stand now? We’re in a new administration, a new Treasury secretary, Scott Bessent, no longer Janet Yellen. What’s the latest and greatest on this front?

Hartley: Sure. Just, I guess, a bit of history. The Yellen Treasury in the fourth quarter of 2023 decided to pivot toward more issuance of short-term debt. At the time, if you look at it in a high-frequency window, it caused the 10-year yield to drop by about four basis points in a very short period of time. Some people, particularly within the GOP, were alleging that, one, this was not a good idea, in part because, one, this is what some countries do ahead of some sort of a default-like scenario or fiscal issues, that they’re issuing more short-term debt, and that there’s certain rollover risk with more short-term debt.

That there was a political element to this, that this was like a backdoor quantitative easing to try and offset the effects of the Fed’s then quantitative tightening ahead of the 2024 election. There were a few things that were being said to criticize Yellen. Now, what’s interesting, and obviously with the suggestion that the next administration should reverse that policy, what’s very interesting now is, so far, the Trump Treasury hasn’t reversed this, the Yellen Treasury government debt management policy.

For what it’s worth, when we were having these debates a year ago, when I was on your podcast, I actually defended this debt management decision, even though I certainly was very critical of Yellen on, I think, most other things. The reason why I think having more short-term debt might make more sense is that Treasury bill demand is very strong. We hear about there’s a lot of demand for safe assets. I have this paper that tries to look at this under the microscope with high-frequency data, and quantities, and yields, and looking at these quarterly refunding announcements where the Treasury announces their future debt management policy, the maturity structure, over the next quarter.

When you look at this under the microscope, it turns out that long-term Treasury bonds are a lot more sensitive to these announcements than Treasury bills. The takeaway is, if you have more short-term debt on the margin, you could generate fiscal savings, have, on average, lower interest costs over time. I think there’s a good case to keep that policy. I would commend the Bessent Treasury on doing this. What’s happened so far, we’ve had three quarterly refunding announcements since there’s been a new administration.

Essentially, in June, Treasury Secretary Scott Bessent said in an interview that, essentially, rates look a little too high to issue more longer-dated debt. I think that is a good assessment in the sense that, especially now, with challenges around an increasing amount of government debt, I think we have to be very thoughtful about how we issue that debt and at what maturities. There is this risk, I think, that if we were to go back to the pre-Yellen pivot, there is this risk that the term spread will blow up by some amount.

If you believe that the effect would be the same as the high-frequency Yellen effect was, which was just four-basis-points higher term spread, I certainly don’t think would be the end of the world. There’s always this tail risk that could be something more. I think we don’t want to be in a position where we get something like what we saw in the autumn of 2022 when Liz Truss took to power. It’s a very complicated story. It wasn’t just about, I think, the release of the mini budget, but there was also a UK guilds auction going on at the same time, failed pension demand for longer-term assets.

There’s a lot going on the longer end of the curve. I think there is some risk, of course, if you were to pivot toward more long-term debt, which could potentially put a lot of upward pressure on the term spread. I think, in my opinion, if you put this into a model, there’s a case to be made for more short-term debt. I commend again the Bessent Treasury for continuing on that Yellen pivot toward more short-term debt for the time being.

Beckworth: Am I hearing you correctly that the reason for doing this is lower costs for the federal government? We’re concerned about fiscal pressures, so we want to do this for fiscal costs, or is it more that there’s a demand for Treasury bills?

Hartley: I think it’s both. At some level, we’re talking about relative Treasury demand, really. At some level, if you had to buy a Treasury bond, where along the curve would it be? One, I think there’s just a lot more demand for short-term debt. In general, again, there’s a lot of demand for money-like assets, for safe assets. This also, I would say, speaks to certain theories of term premium that say that there are certain segmented types of investors that are focused on certain areas of the yield curve.

Maybe it’s money market mutual funds are very focused on the short end of the curve, whereas pension funds are focused more on the long end, liability-driven investing. My sense is, in general, again, these debt management issues are about minimizing overall interest costs to the government. Really, that’s what matters most. It’s a risk management type of decision. I think the assessment is that, again, there’s a lot of strong demand on the shorter end of the curve, more so than the longer end of the curve, and let’s not risk it.

To give you a very brief history of government debt management, the Treasury used to do debt management very differently prior to the 1980s. It adopted this stance of stable and predictable. Essentially, the idea is that think about long-term debt management, don’t be focused too much about what interest rates are today, and don’t try to time the market too much. Now, in reality, Treasury is always doing a little bit of market timing to varying degrees. They’re usually issuing more Treasury bills during recessions, and that’s a general pattern.

It’s a general pattern of Treasury debt issuance. I think it’s fair to say, and I wrote op-eds in the late 2010s about this, there’s a pretty hard, roughly speaking, zero lower bound on interest rates. We could talk about negative interest rates a little bit, but there’s a pretty serious bound there. I think there were several people in the late 2010s, including myself and John Cochrane, and others, that said, “Why don’t you just issue a lot of long-term debt now and lock that in when rates are low?”

It’s not what Treasury did, and there were some who argued we should have 50-year Treasury bonds or 100-year Treasury bonds. The Treasury Borrowing Advisory Committee essentially pooh-poohed this at the time and said that there wouldn’t be enough demand for them. There’s this question of, should the Treasury, given today’s interest rate environment, try to time the market or try to think about are interest rates too high or too low. 

Some people are suggesting that Scott Bessent is looking toward a regime of government debt management that’s a little bit more focused on market timing than previously existed. Again, he said in this interview back in June that rates looked a little too high to issue more longer-dated debt. I would expect maybe something like that to continue, but I think, in general, it squares with this idea that, perhaps, given where longer-term rates are, we should be very careful about how much longer-term debt we issue.

Beckworth: All right, Jon, in the last few minutes, I need you to help me out here. You’ve got me worried a little bit because you’ve noted that maybe some fiscal pressures is behind this move, lower financing costs. Now, by itself, not a big deal, but in the broader context of what I see happening, I find that a little alarming. Maybe walk me off the edge here. I’m going to tell you what I see, and you can respond to this. You get the final word. Number one, when I see President Trump calling on the Fed to cut rates, his motivation has been strictly to lower the interest costs on the debt.

It’s not like in 2019, he was calling it for more of a full-employment, robust economy. It’s very clearly, let’s lower the cost of interest payments on debt. Secondly, we look at the supplemental leverage ratio tweak. One of the main arguments for doing that is it allows banks to pick up some slack and carry more Treasuries on their balance sheets. Stablecoins, I love stablecoins, but one of the arguments is it’s an extra source of demand for Treasury securities. 

Senator Ted Cruz has called in interest on reserves. He thinks it will open up revenue to flow back into the federal government. I think that’s probably wrong, but nonetheless, all these calls are motivated by fiscal pressures. Now, on top of this, you’re telling me there’s a push for more Treasury bills because it’s a little more cost-effective. I put those all together, and it gets me worried, Jon, that we’re looking at a fiscal dominance train that’s coming toward us, or am I misreading the symptoms here?

Hartley: I would, I think, take the other side of this trade in the sense that, one, I am probably more optimistic about the US than I think most others. One, on debt management, there’s always this challenge of debt management as a question. As long as the government’s issuing debt, there has to be some choice of maturities. I think, one, we should always be thinking about minimizing interest costs subject to some level of interest cost volatility. I think, at some level, being more thoughtful about this than we have in the past, it’s something that I think is something of a positive.

Two, I think, in general, the narrative that there’s a sovereign debt crisis on our doorstep, I would maybe push back a little bit against, just in terms of where long-term Treasury debt’s at. One, 100% of debt-to-GDP is very far away from, say, where Japan is now. We’re on a trajectory that could take us there. Again, depending on how entitlements unfurl in the future. As far as losing dollar dominance and all these other issues, de-dollarization, things that we hear about all the time, I think are promoted often by people who, I guess, have certain, I’d say, interests that are maybe contrary to those in the US.

There’s Bitcoin people that see US dollar’s loss as Bitcoin’s gain. There’s people in the BRICS alliance who would love to see the US dethroned. I think that’s very much a fantasy. There’s a separate paper that I’ve written on this called “De-Dollarization? Not So Fast.” If you look at all sorts of measures, whether it’s the denomination of FX trades, whether it’s the denomination of currency invoicing of trade. If you look at the denomination of central bank FX reserves, whether it’s the denomination of debt around the world, the US dollar has held firm through COVID, through the Russian invasion of Ukraine.

Today, it’s essentially no different than it was, say, just five, 10 years ago. In fact, things used to be actually worse in the, say, early 1990s. There was, roughly speaking, the fraction of all central bank reserves were denominated, the share was essentially 10% less than it is today in terms of US dollar-denominated FX reserves. We’re still, I think, in a very good place. Of course, we’re always at risk of losing this. Of course, I think we do need to think now about having fiscal reforms.

I would be more worried about, for example, there being another recession, say, another Global Financial Crisis, another pandemic that puts another stimulus front and center. Those are the sorts of things that I think would be bad in the sense that they would get us closer towards something that looks like a bad sovereign debt scenario. At some level, the US still has the world’s reserve currency. Treasuries are still the most liquid asset in the world. I think it’s very difficult, in my mind, to bet against the US. I’m very much, I’d say, a US optimist in that respect.

I think there’s just so many things going for the US. Think about the generative AI revolution that’s happening right now. All these sorts of things that could really benefit growth. How do we attack sovereign debt? Obviously, changing your deficits is one thing. Scott Bessent’s very much big on his 3-3-3 plan getting to 3% deficit to GDP ratio, which is, I think, very commendable. There’s other ways to pay down debt as well. Seigniorage is one of those. We had some inflation a few years ago, 20% inflation. That certainly helps reduce the real value of debt.

Really, economic growth is the third way of getting out of sovereign debt issues. I really would look to, and I’m excited or optimistic about how generative AI and related transformations could potentially really be a boon to economic growth that, again, leads to more tax revenues, that allows us to pay down our debt more quickly. Really, generative AI, that story is really specific to the US. No other country really has that innovative capacity. At some level, I’m more optimistic about the US because it has this capacity and ability to potentially and partially grow out of its debt, something that other countries don’t have.

Beckworth: That is a great positive note to end on. I would add to that positive vibe I’m getting from you, Jon, that the Treasury market right now, 10-year Treasury, is about 4.2%, 4.3%, so it clearly is not freaking out and worried about some fiscal crisis. Then one other thing, just really to go with what you’ve said, you mentioned the paper that was presented at Jackson Hall by Emi Nakamura.

There’s another paper presented by Ludwig Straub, if I’m saying his name correctly, and he and his co-authors make the case that there’s actually a lot more demand for assets than there is supply in the US, and the debt-to-GDP could go a lot higher than it currently is before we hit the wall. Those are some great positive notes to end on. Jon, thank you again for coming on the program. Listeners, we have links to all of his papers he mentioned in the transcripts. Jon, it was a real pleasure to have you back on.

Hartley: An honor to be here, David.

Beckworth: Macro Musings is produced by the Mercatus Center at George Mason University. Dive deeper into our research at mercatus.org/monetarypolicy. You can subscribe to the show on Apple Podcasts, Spotify, or your favorite podcast app. If you like this podcast, please consider giving us a rating and leaving a review. This helps other thoughtful people like you find the show. Find me on Twitter @DavidBeckworth and follow the show @Macro_Musings.