Lukasz Rachel on Non-Ricardian Macroeconomic Policy and Its Implications for Inflation

Lukasz Rachel is a former Bank of England economist and currently is an assistant professor of economics at the University College of London. In Lukasz’s first appearance on the show he discusses his big career breaks, the implications of secular stagnation in the industrialized world, what is next for R-star, what non-Ricardian macro policy looks like, his policy prescriptions for the US, and much more. 

Subscribe to David's new Substack: Macroeconomic Policy Nexus

Read the full episode transcript:

This episode was recorded on October 29th, 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 Lukasz Rachel. Lukasz is a former Bank of England economist and is now at the University College of London. He has written widely on secular stagnation, r-star, and monetary-fiscal interactions, and he joins us today to discuss them. Lukasz, welcome to the program.

Lukasz Rachel: Thank you for having me, David. It’s a great pleasure to be here.

Beckworth: It’s great to have you on. Now, when I first met you, it was back in January 2020 in the warm and wonderful San Diego. It was the American Economic Association meetings, and you and Larry Summers were presenting a paper on secular stagnation. I remember it so vividly because I was blown away by it. We’re going to come back to it, but one of the key claims you made in it is, man, were it not for the budget deficits and the large stock of debt, the neutral rate would be even lower than it was. Back then, it was pretty low. We’ll come back to that, but it really stuck with me, made me think about this issue a lot more.

It’s great to get you on the show five years later, and no more COVID. That was a crazy time because it was the calm before the storm. Little did any of us know what we’d be stepping into in the next few months. Before we get into your paper or your research, tell us about yourself. You worked at the Bank of England. Now you’re an economist doing research in the academy.

Lukasz’s Career

Rachel: Yes, absolutely. Let me first just say that I remember this meeting in San Diego. It’s a very memorable conference because it’s sunny San Diego a couple of months before all the lockdowns began. That was a momentous occasion. Also, I presented the paper in the panel that was chaired by Raghu Rajan, joined with Larry Summers. It was some of the young economists’ dreams come true, I guess. It was a very special time.

I’m from Poland originally, in case listeners are wondering about the accent. I was drawn into economics already as a teenager because I observed really massive structural transformation of the Polish economy from the centrally planned economy during the communist times to a market economy that’s now very nicely connected to the rest of the world. I arrived in the UK when Poland joined the European Union, so a good 20 years ago. I studied at the London School of Economics. My first job was, as you mentioned, a job at the Bank of England. That was a really special time for me because I started at the bank in the late summer of 2008.

It was not just a special time for me, but a special time also for the institution because, of course, this coincided with the meltdown of the global financial system. It was a very fascinating time to start at a policy institution like that. I realized, suddenly on my first job, just how countercyclical a job of an economist is. We mostly needed our expertise, it’s mostly valued when the world, as we know it, is changing. That was really fascinating.

The bank has prepared an elaborate set of induction activities for us for the new economists joining the bank. We saw senior people being pulled out from some of these briefings, from some of these meetings, in order to make some more important decisions during these tumultuous times. It also meant that much of the knowledge that we had or the understanding of the economy was suddenly questioned. My first job at the Bank of England was looking at UK productivity and investment dynamics. Productivity started dropping very fast during the Global Financial Crisis in the UK and elsewhere as well.

Little did I know that nearly two decades later, that’s true, we’re still going to be talking about the productivity puzzle and the persistent hit to the productivity levels since then. That was a very formative time for me. It was very interesting. I’ve met an awful lot of colleagues and wonderful managers at the Bank of England. I have then progressed through the bank through a couple of jobs and started working more and more on longer-term research and longer-term issues.

One issue that really drew me in was this overarching topic, which is a very deep macro topic of the neutral real interest rate. It’s a very interesting topic because that’s variable. The neutral real interest rate captures all sorts of trends and changes in the economy, both on the household side, on the firm side, also on the government side. With that longer-term work, more research-style work, I was drawn more and more into academic style of research and writing. Then I also started to do a PhD at the LSE.

Subsequently, a few years down the line, changed from being in a policy institution from the Bank of England to being in academia. The set of questions and the set of issues that fascinate me and that constitute a good chunk of my research remain the same. I’m motivated by the same set of puzzles and facts. Yes, very happy to be here to discuss some of these important issues.

Beckworth: Yes, I can see that interest in the neutral real rate running through the papers we’re going to talk about today and the other research you have done. I got to ask quickly, though, how did you connect with Larry Summers? That’s awesome. You two could write a paper together, and he can help elevate your work.

Rachel: Yes, absolutely. I mentioned this more longer-term work that I started doing at the Bank of England. We published a paper with my colleague, Tom Smith, while we were at the bank. The paper was this relatively long, relatively detailed work on the various drivers of the r-star, of the natural real interest rate. In that paper, we used a desired saving, desired investment framework, which stayed close to my heart because I think it’s a very useful lens through which we can think about r-star.

It so happened that Larry Summers was very interested in this topic at the time. Larry picked that paper up and wrote an op-ed in the FT about it. We were, of course, extremely happy that that happened, and that’s what allowed us to draw a connection. Two years down the line, I was already doing the PhD at the LSE, and I got the opportunity to spend some time in the US as a Fulbright scholar. I spent that time at Harvard. Then Larry was very kind to spend some time with me talking about the economy and talking about neutral real rates. From these conversations, it transpired that we have enough material and enough research to start thinking about writing a paper together. Those were the beginnings.

Then, yes, the rest is history. For the next year or so, we worked hard on this paper on secular stagnation that you mentioned. We published it eventually in the Brookings. Again, a fantastic conference. That’s the one I presented as well at the American Economic Association meetings.

Secular Stagnation in the Industrialized World 

Beckworth: Yes. Let’s talk about that paper. It’s a 2019 paper. It’s titled “On Secular Stagnation in the Industrialized World.” As you mentioned, you and Larry Summers wrote it. Give us the executive summary, and then we’ll jump into the details.

Rachel: Absolutely. We wrote this paper during the time when, over the past decade, most of the developed world was stuck at the effectively zero lower bound or effective lower bound. Deficits were relatively high. Debt ratios were rising. Yet activity and the recovery post Global Financial Crisis has been below expectations. It has been slower than many had expected.

Many papers out there by that time have pointed out that there are structural forces in the economy that translate into high demand for assets or high propensity to save by households and low investment demand by firms. That was where the literature was going, in terms of trying to explain the phenomenon of relatively weak investment, a relatively weak demand in the economy, and low interest rates. 

Against this backdrop, our main point in the paper with Larry is that not all forces push in the same direction. Our main substantive point of the paper was that the fiscal policies and the social policies that we have observed across a block of advanced economies have tended to push interest rates in the other direction. All the empirical research and also the OMO modeling would suggest that the rising debt-to-GDP ratios, increased provision of social security, healthcare spending across major advanced economies, would have all else equal pushed interest rates up. Of course, we have observed this long-running trend of long-term real interest rates trending downwards. The main point that we were making in the paper is bringing these forces together in the quantitative framework to try to assess how important they were quantitatively.

What we found is that basically these, what we call public sector forces, so again, the rising debt-to-GDP ratios, rising provision of social security spending, these forces were quantitatively very important. They were masked by some private sector drag on the neutral real interest rate, which is what we think about in terms of equilibrium real interest rate in the longer term in the economy that’s operating at full employment and which has inflation broadly on target.

Beckworth: Yes, it’s a really fascinating paper. What really caught my attention at that AEA meeting, that conference where you presented, was this claim that the neutral rate was down by 700 basis points. The private sector neutral rate was down by 700 basis points, but for the deficits. Had it not been for the deficits, social spending, it would have been down that much.

It was only down roughly, I think you estimated, 300 basis points. That 400 basis point difference meant that we weren’t in deeply negative territory for interest rates. That’s the most interesting thing. Let’s do the right counterfactual here. Had it not been for all the deficit spending, you’re arguing, we would have been in a very deeply negative interest rate world. Now, help listeners understand why that’s important. What is a secular stagnation world? What would it have meant if we had deeply negative interest rates? In other words, how much worse could it have been compared to just running big deficits and all the worries we had about those?

Rachel: Absolutely. Our concern has been that the low level of interest rates makes it harder for monetary policy and for macroeconomic policy in general to stimulate the economy at the time of need. Exactly as you pointed out, David, the counterfactual, once we take into account the upward pressure on interest rates from these drivers of debt, deficits, and social security and healthcare spending, would have been for interest rates to be significantly lower.

Why is that a problem? Well, chiefly because monetary policy is naturally constrained by the effective lower bound. With natural rates deeply in the negative territory, if that indeed was the world we lived in, monetary policy would find itself more often at the zero lower bound, which of course the research has shown that there are mechanisms that this can become a self-fulfilling problem. The agents, the households and firms in the economy, could start expecting the monetary policy to be ineffective on occasion and therefore revise their expectations accordingly.

I think that was one of the main concerns that we were thinking about. Of course, low natural real interest rate, if it’s not delivered by monetary policy, will show up as this equilibrium somewhere else in the economy, so as even lower investment rates and therefore would translate in time into lower growth rates of the economy. The essence of the secular stagnation hypothesis is that the overwhelming desire to save and not much of an investment demand create this deeply negative real interest rate that is problematic from this cycle management and long-term growth perspectives.

Of course, in the historical episodes which we actually did observe, the social security did increase and the GDP ratios did rise, and so interest rates did fall over time, but perhaps not as much as they would have otherwise.

Beckworth: This is so interesting because we had all these conversations back then about negative interest rates. For example, should the Fed, should the Bank of England, should the ECB use negative interest rates as a policy tool, should we try to go down there? There are challenges with that, with physical cash, with potential disintermediation in the financial system, but it’s a real challenge, and we would have had even more decline in growth and slowing and stagnation. That’s the concern.

The other point I want to highlight from your research, which I’m sympathetic to, and some people they push against, but this was driven by structural forces. There’s some who might say, “Oh, it was the central banks who lowered the rates,” and I always say, “no, the central banks were following the fundamentals down to zero.” They had to go down to zero, and they got stuck at zero, as you point out. Maybe explain to our listeners the structural forces that were explained back then for why rates were going down and why central banks had to follow them down.

Rachel: Yes, absolutely. Again, for thinking about the structural forces, I find it very useful to think about the demand for assets or the capital supply in the economy, so basically the saving side of the economy, so households’ decisions about how much consumption to postpone to the future and how much to save today. The balance of that relative to asset absorption in the economy—which comes from investment demand by firms, and also of how much a government is borrowing from the household, so government deficits and debt—this balance is in constant movement, so changes in the economy affect both the decision of the households and also the decision of firms and for the government. 

The main forces that we think are operating in terms of affecting households’ decision to save are the demographic factors. So the fact that 50 years ago, the average person expected to spend around six or seven years in retirement, whereas today it’s more like 20 years. That makes a massive difference to how much people want to save for that later stage of life.

Another demographic factor is, of course, the slowdown in population growth. That slowdown means that we need less capital investment to keep up with the population numbers. We need less investment to keep the capital-to-worker ratio at the desired level. Demographic factors have been very important, and there are several papers that try to quantify their impact, and my work also contributes to that endeavor.

Another important factor in both saving and investment decisions is the future of the economy, is essentially the expectations about future growth of productivity. If households expect productivity growth to be high, they know that they will have a lot of resources to draw on in the future for their consumption, and that limits their desire to save. At the same time, if firms expect productivity growth to be high, they want to invest because they want to have machinery and capital in place to basically make use of that productivity down the line.

Higher productivity growth would act both on the saving and on the investment side to push interest rates up. Of course, in the event since the 1970s, we have seen productivity growth expectations being revised downwards, so in the other direction. That has meant that households have been becoming less and less optimistic about the future, and so have firms. Households reacted by increasing saving, and firms reacted by cutting investment. In our macroeconomic models, certainly, that tends to be one of the most decisive factors. 

There are other trends in the economy that are important, too. Research has highlighted the rise in markups and market power of firms, that has important implications both for investment in physical capital but also in the value of firm equity, which capitalizes on the future profits that will be important for this balance of capital demand and capital supply. Of course, as I mentioned, there are other forces that come mostly from the public sector side. Public sector forces, policies such as government debt management and social security, that we already discussed.

Beckworth: The decline in expected productivity growth that shifts desired investment back. But then on the other side, desired savings has gone out because of demographic factors in the world. You mentioned the aging population, the decline in population growth. All those things contribute to a declining r-star or neutral real rate. 

Now, a question I have for you, Lukasz, is like your measures, I believe in your paper, you show this trend occurring from the 1970s going all the way down. If I look at some popular measures of r-star like the Laubach-Williams measure, what you see after the Great Financial Crisis is a sudden drop-off. It’s like it’s going down and it falls to the floor.

Some measures, they go back up recently. Just let’s go back to that almost a discrete decline during or after the Great Financial Crisis. Is that overstated? A better way to think about it is, it was always coming down, maybe we just recognized it after 2008, or did something actually exacerbate, make it worse in 2008?

Rachel: That’s a very interesting question. I think one that requires us to think about the concept of r-star that we are considering. One important distinction in the literature is about the horizon in question. You can think about equilibrium real interest rates as a business cycle frequency. Here and now, what is the level of real interest rates that would make monetary policy neutral, given all of the cyclical shifts in the economy and given all of the structural changes that we have observed?

The research that I have done, including the paper with Larry that we have discussed, focuses more on the longer-term drivers. These are the structural forces that are beyond the business cycle frequency. What the concept of r-star consistent with these models tells us is where the r-star will ultimately settle once the business cycle frequency shocks subside. 

Going back to your question, David, in terms of what happened after the Global Financial Crisis, I think absolutely it has been the mix of both. This was a big cyclical shock, certainly a big shock to the confidence stemming from the events in the financial sector, but of course, just broadly the confidence of firms and households, with massive uncertainty depressing demand.

It has also been a medium-term shock with the realization of necessary deleveraging on part of any sectors of the economy. It has also been a bit of a realization about the long-running trend of these driving this imbalance in saving and investment that we have talked about. I think all three factors contribute in the short term. Then once the cyclical factors abate, we are left with this longer-term concept that I think these structural forces are the main driver of. Structural models that you mentioned can really speak most directly and most clearly to those.

Beckworth: Now, that is a great answer. There are several things we struggle with in macroeconomics. One is identification because there is always general equilibrium effects. Another one is to distinguish between cyclical versus trend or structural forces. Your answer is we need to be careful not to confuse the cyclical for the structural. Structural is still there. Cyclical will come and go. It is always right in front of us. In the real world, we see cyclical screaming in front of us. 

What Next for R-Star?

We tend to think that way, which is a nice segue into your next paper, I want to chat with you about, another Brookings paper. This one came out this year. It is called “What Next for R-Star? A Capital Market Equilibrium Perspective on the Natural Rate of Interest.” You actually help us flesh out this notion. How can we think about r-star going forward? Because, as you know, Lukasz, if I look at tips, real measures from the markets, they have gone up quite a bit. Ten-year Treasury is up quite a bit. On the surface, it sure looks like r-star has gone up. You say, “Hey, not so fast. Let’s dig deeper and look at what is going on.” Summarize this paper for us.

Rachel: Absolutely. I just want to convey the fact that this is a fascinating juncture to be thinking about this question. If we look at the financial market measures of real long-term interest rates, they have declined continuously, basically for three decades, up until the COVID episode. Together with, of course, during the COVID episode, the central banks were fighting the post-COVID inflation. Short-term interest rates have gone up substantially. Exactly as you said, David, the long-term interest rates have also increased by a lot. When we look at the picture, the figure of real long-term interest rates, it does look like a significant break from trend.

A natural question to ask from the perspective of someone who has done research on this before the pandemic is whether this is indeed a strong structural break and basically the end of secular stagnation as we thought about it in the run-up to the COVID pandemic. Or perhaps it’s a cyclical thing that for some reason, is also showing up in long-term real rates.

In this paper, what I tried to do is basically shed some light on this question, also acknowledging just how complex and difficult that question is. Rather than providing a definitive answer to this multitrillion-dollar question of where r-star is going to settle, I tried to be helpful in providing basically a toolkit that might be useful for analysis of this question. This toolkit comprises of three parts. The first part is this capital markets equilibrium in the long run. I already mentioned this, but in this paper, I go further than before in constructing a capital markets equilibrium framework that is directly derived from a general equilibrium macroeconomic model.

That is very helpful for several reasons. First, it allows us to think about the various drivers of neutral real interest rates together with what they imply for wealth-to-GDP ratio. A fascinating fact that we should mention to our listeners now is that over the past 30 years, while the interest rates have been on this declining path, we have seen a dramatic increase in wealth-to-GDP ratio across advanced economies. That ratio has increased from about three to roughly about six over the space of the past 40 years or so. It’s basically a doubling of wealth relative to income, together with the decline in real interest rates. It’s precisely this capital markets equilibrium framework that gives us the discipline to think about the drivers of both trends simultaneously. That’s the first part of the toolkit that I propose in this paper that I think is useful.

The second part of the toolkit is thinking about the transitional dynamics between where we started and where we’re heading to, sort of between the steady states of the economy. As we discussed just a few minutes ago, some of the changes in interest rates can be pretty abrupt, but the literature so far has usually analyzed the transitional dynamics in this relatively simple way, assuming perfect foresight on the part of agents. Basically, what that assumption implies is that people in this model economy already know about future shocks that are hitting the world and therefore adjust their behavior already in advance. That usually results in relatively smooth gradual decline in r-star estimates. In this paper, I wanted to go beyond that because that, of course, is not very satisfactory or realistic assumption. I wanted to go beyond that and propose this limited foresight transition path where basically I do put extra care into constructing agents’ expectations over time, and how these expectations have shifted and when in relation to these various shocks. That, I think, provides us with a much better, more realistic take on the dynamics of r-star over time. I show in the paper that has quantitatively some meaningful implications.

Then the third part of this toolkit that I think is useful is thinking about sensitivities that this macro model embeds or implies. Why are sensitivities useful? Why are they unuseful? Because if something now changes in the environment, if US administration changes significantly one of the policy levers, then we want to have a tool that gives us the multipliers from that policy change onto what we might expect to happen in r-star going forward.

Coming out of these sensitivities, we can combine these sensitivities together to form various scenarios that are often discussed out there in the markets among market participants and economists. We can form scenarios that correspond to these narratives and basically quantitatively check roughly what these scenarios imply for r-star going forward. That is basically a summary of where I think that paper is useful. We can, of course, also talk about what that framework and toolkit actually imply, so what the numerical results that I’m getting are.

Beckworth: Let’s go ahead and mention them. If I read correctly, you have your r-star close to 0% today, started at 5% in 1970. It’s still a similar story to your first paper. There’s been this long downward decline in r-star. Just to go back to this point you made, in the previous paper we discussed, we took more of a, I think we could call, flow-based view, desired savings, desired investments. In this paper, you’re taking more of a stock-based view. This capital wealth asset model, it provides added richness and understanding to it.

I want to go to your second point, and this is limited foresight. I think this is really powerful insight. Your point, I believe, is that because it’s hard to distinguish the cycle from the trend in real time, we have limited foresight. We really can’t see well beyond what’s in front of us. Maybe we are overestimating how high r-star is in the long run. Is that a takeaway?

Rachel: It’s partly a takeaway, but I don’t think it’s a unidirectional. We might be over and underestimating r-star if we use a structural model for thinking about r-star, and if the information and forecasting capabilities of firms and households in our model are unrealistic, then some of the results that we’re going to get out of that framework will be unrealistic, too.

Specifically, if there are big shocks on the horizon that the agents expect, that households and firms expect, they will change their behavior already today. This is the more of modern macroeconomics that expectations matter. What I think is useful for quantitative answers about r-star is to make sure that these expectations that the agents have are plausible and realistic. I do not want firms and households in my economy to be perfectly forecasting the Global Financial Crisis in the early 2000s.

With a more realistic set of expectations, we can also get the households to be surprised by some of the shocks, even as they are transitioning from one steady state to the other. Basically, what I try to emphasize is that framework is capable of handling both, for example, a demographic transition that is well underway. Agents expect to some extent that demographic trends will continue, but at the same time, they are surprised by some other shock. For example, a reduction in long-term growth expectations that is spurred by the weak recovery following the Global Financial Crisis. It’s about that mixing of those different expectations and different forces that can deliver what I think is quantitatively a more meaningful answer to the question of how r-star has evolved and what has driven it over time.

Beckworth: We can go both directions. I don’t want to overstate my case, but if I could apply this to what we’ve seen, after 2008, maybe it overstated how far down it went, and now maybe we’re overreacting to the upside. Is that a possible application of your model?

Rachel: What comes out of the limited foresight transition is that growth expectations have changed, but only several years after the crisis. To document that, I go back to looking at some of the forecasts for GDP growth at the advanced economy level and for each of the major economies individually. I document this pattern that for a long time following 2008, institutions such as the IMF were projecting growth to ultimately come back up, and in the medium to long run, go back to the pace which we have observed even before the Global Financial Crisis. It’s only after a few years of disappointing recovery that these expectations were revised down.

In the model exercise that I carry out, it’s those revisions of those expectations later on in the recovery, so in the mid-teens, that really impact the medium to long-term r-star then. However, if we put all of these forces together—of course, growth is one of them, demographics is another, changes in markups, and all the forces that we have discussed so far—the model implies that the long-term real neutral rates have declined a lot since 1980s, but they have never declined as far as what the market rates would have suggested in the run-up to the COVID pandemic.

On the one hand, there is a significant decline in real rates, but this decline is nonetheless smaller to levels that are closer to zero, rather than minus 1% or minus 2% that we have observed in market implied real interest rates in the run-up to the pandemic. In terms of the interpretation of the recent pickup in market-based real interest rates, some of that pickup might be interpreted as a correction toward what the model would have suggested real interest rates should be. Some, but not all of it.

Beckworth: Lukasz, this has been really fascinating, this second paper on r-star. Could you discuss some potential upside risk or developments that, in your view, could raise this trend, r-star measure?

Rachel: Absolutely David, I mentioned that some of the pickup in market-based real interest rates might reflect the markets going to where the model indicates the r-star is, but markets went way beyond. The pickup in the long-term real interest rates is much larger than this. Motivated by this, what seems to be some market perception that real interest rates could stay higher for longer, I then use my model to think about scenarios that would take us outside of this business-as-usual scenario. What is the business-as-usual scenario? It’s basically thinking about the forces driving r-star in a similar way that we would have thought about them in the run-up to the pandemic, like a few years back. 

There’s a few new forces happening on the horizon. The fiscal situation is becoming increasingly worrisome and stretched. There are increasing investment needs with relation to the AI boom and perhaps a productivity explosion that is waiting around the corner. There are also changing perceptions about how globalization is going to progress over time, and also issues to do with remilitarization and the need for increasing of the military expenditure by many governments across the world. 

In the final part of the paper, I try to quantify some of these upside risk scenarios. Now, the fiscal scenario is particularly relevant because, of course, the government debt-to-GDP ratios are much higher post-COVID. Some of the trajectories that are projected are worrying as well.

There’s the two scenarios that, perhaps, are most worth highlighting here are first, the scenario-based artificial intelligence. Artificial intelligence is a capital-intensive, investment-intensive technology. We see this big investment boom being driven by AI firms right now. It might also bring about a rising productivity growth and perhaps rising concentration among firms in the economy.

Putting these different elements together, I construct this AI scenario which has the potential to boost r-star by about one percentage point. Of course, that is quite a lot. It could be more if the AI boom proves to be even more powerful. I show a summary of different estimates about what AI could do to productivity in the paper.

Another really interesting aspect as well is the spread between safe rates of return on government bonds and more broadly defined returns to capital. That spread has widened over the past few decades. The safety premium or the convenience yield has gone up over time. In other words, the safe r-star has declined by significantly more than the return to capital, our steady state return to capital.

One scenario that I analyzed in the paper is what if this spread has suddenly shrunk? The model says that if that is indeed the case, then that pushes up on the safe r-star basically immediately and quantitatively significantly. That’s another thing that we should bear in mind that this spread between safe and risky rates being driven by the convenience yields or risk premium is an important driver of safe r-star itself.

I illustrate this story in the paper by showing the cumulative returns on safe versus risky assets that the investors have obtained ex post over a long stretch of time in the economy. The recent episode was the episode where the holders of nominal government debt have not done very well. They might question the safety attributes of this instrument, and perhaps the safety premium might go down. That in itself will have an impact on r-star in equilibrium.

Brothers in Arms: Monetary-Fiscal Interactions

Beckworth: So many questions I could follow up to those comments. Interesting paper. For the sake of time, I want to move on to a third paper of yours, the last one we’ll cover. There’s some overlap between what you just said there about fiscal policy r-star, in this next paper. This next paper is one that came out this year as well you co-authored, and it’s called “Brothers in Arms. Monetary-Fiscal Interactions Without Ricardian Equivalence.” Maybe let’s start out with a real simple question. What is Ricardian equivalence, and therefore, what is non-Ricardian macro policy?

Rachel: Ricardian equivalence is a feature of macroeconomic models which says that for households or for agents more broadly in an economy, the government financial policy, so debts and deficits, do not matter per se. There’s a good reason for why we might believe this implication in the context of a model.

If the government sends a bunch of checks to people today, then these people might appreciate, they might understand that these checks need to be funded somehow, and so they might expect taxes to go up in the future. They will take this check, put it in their back pocket, and save it for when they have to pay the taxes that they expect to have. Now, that implies Ricardian equivalence or Ricardian behavior in the sense that consumption and real decisions by households and by firms are therefore unaffected by this kind of government intervention.

Non-Ricardian equivalence, on the other hand, means that when the government increases its deficits, households and firms react to this in a meaningful way. Why might that happen? For very many reasons. Households might expect not to be the ones paying the taxes in the future, or they might be liquidity constrained, or they might simply be receiving deficits and the cash that’s coming from the government as well due to misunderstanding of how government finances work. There’s probably many other reasons why this Ricardian equivalence proposition might break.

The point of departure for this paper that you mentioned, David, is essentially to think about what the lack of Ricardian equivalence, so what this non-Ricardian behavior implies for monetary-fiscal interactions. I guess the background or the motivation that drives this work is that in very many central banks, probably all central banks around the world today, there’s this big push to expand the modeling framework and include models that are based on incomplete markets and heterogeneous agents, so-called HANK models. That is, of course, for very good reasons, that these models have very many realistic features in terms of how the economy responds to various shocks and how various parts of the society are affected differently by various economic phenomena.

Irrespective of the details embedded in those frameworks, what all of these models share is that they break the Ricardian equivalence feature. They break it through the mechanisms that I have outlined, specifically through some agents being liquidity constrained. What we want to do in this paper is think about these monetary-fiscal interactions when Ricardian equivalence is broken, and so in this large class of models that are becoming workhorses not only in academia but also in central banks today.

Now, that is, I think, useful because a lot of the work in economics on the monetary-fiscal interactions has focused on the case where Ricardian equivalence does hold, and we can think about representative agent in the economy representing the demand side of the economy. In this framework, Ricardian equivalence holds, and that implies a very clean and strong implication for monetary-fiscal interactions and their implications for monetary policy specifically.

First, these models tend to imply that Taylor principle is very important, so whether central bank reacts strongly or not to inflation really matters hugely. Second, they imply that fiscal policy might not be inflationary, so purely tax and transfer policy, so governments sending checks to people might not result in any inflation or any boom. Thirdly, they also imply that the fiscal consequences of monetary policy are easily dealt with by fiscal policy. These frameworks also imply a strong degree of independence between the two policy arms. Now, it turns out that when Ricardian equivalence is broken, all of these implications are fundamentally affected.

Beckworth: This is, I think, increasingly relevant in the world we live in. It definitely helps make sense, I think, of ’21, ’22 in the US. Those checks that were sent out, fiscal policy, they signal very clearly this is a wealth transfer. This is not going to be something that we’re going to collect taxes on in the future. When you talk about non-Ricardian fiscal policy, you’re saying the government is not committed or policy is not committed to undoing this in the future with higher taxes.

Lukasz, with non-Ricardian fiscal policy, there’s this lack of backing in the future. You’ve noted it has implications for the interaction of monetary policy, fiscal policy. In particular, there’s this idea called the Taylor principle that if you’re a good macroeconomist, you’re brought up on like, as long as a central bank raises its target more than one-for-one with inflation, we’re good. We’ve got determinancy. Inflation expectations will be anchored. We’re solid. 

What you’re saying is that necessarily isn’t true. That won’t hold. I want to give an example of that. I want to get your response to this. There’s a paper by a gentleman named Eduardo Loyo. He wrote about Brazil in the 1970s and ’80s. He noted from 1980 to 1985, the Brazilian central bank followed the Taylor principle, but you know what? They got hyperinflation. It wasn’t enough. They basically had non-Ricardian fiscal policy, a large stock of debt. Expectations were unanchored. Does this make sense?

Rachel: In our paper, we indeed show that with non-Ricardian demand side of the economy, so with households that are liquidity-constrained or do not perceive future taxes appropriately, Taylor principle is neither necessary nor sufficient for good outcomes in equilibrium. The main intuition from that comes from the fact that the stock of government debt or government stimulus directly impacts on demand.

If this is the case and the government stimulates the economy, that raises demand and inflation, and if the central bank reacts very strongly to that, if they’re obeying Taylor principle and they’re reacting very strongly to that inflation, well, that also has fiscal implications, that higher interest rate payments will always mean that the fiscal situation of the government deteriorates. What that means is further boost to the demand side of the economy because further increases in debt will again spill over into higher spending and higher inflation. With non-Ricardian behavior, the Taylor principle does have this dark side that it can lead to these almost-explosive dynamics of demand and inflation. 

At the same time, the pervasive issue with models with Ricardian equivalence and with representative consumer is the determinacy of equilibrium in these models. These models are very forward-looking. Agents look forward and form expectations about the future in order to make the decisions today. Oftentimes, in these models, if the Taylor principle is not satisfied, John Cochrane has very elegantly showed that agents can form basically any expectations about the future, and this can become self-fulfilling. Monetary policy in these models needs to be very aggressive to combat these sunspot expectations.

In the model with no Ricardian equivalence, with direct impact from debt to demand, there are endogenous channels, which mean that monetary policy doesn’t need to be as aggressive. Basically, what I want to highlight is the Taylor principle is neither necessary nor sufficient for these good outcomes in the economy.

Just one remark, David, to what you said earlier, which is you alluded to the fiscal theory of the price level and these considerations about when the fiscal side stimulates the economy without any fiscal backing inside, so basically, promising that they will never raise taxes to cover the expenditure.

In the paper that we’re just discussing, with no Ricardian equivalence, there’s, again, a direct link between government policy and demand and inflation. We don’t need to rely on these long-run expectations about government taxing the population or not. Whether the taxes are forthcoming or not will stimulate demand today because agents themselves do not think all the way out into the future.

Beckworth: You have less heroic assumptions about the foresight of the public looking at primary deficits 10, 20 years into the future, and you still get the same result.

Rachel: That’s right.

Beckworth: Again, just to maybe summarize and to make this in plain terms is if we follow a Taylor rule, it assumes in the background that fiscal policy is doing its job in keeping the government solvent. It’s running primary surpluses over the business cycles. The fiscal house is in order for the Taylor rule to do its magic to be successful. Even then, you said, it’s not necessary nor sufficient, in some cases, is that fair?

Rachel: No, that’s right. With the non-Ricardian feature, we do need fiscal policy that is responsible and that reacts to debt sufficiently strongly to stabilize it. Otherwise, these explosive dynamics that I have mentioned might set in. In either model, whether this is a representative agent model or a heterogeneous agent model, there is a need for fiscal policy to be mindful of the trajectory of debt.

In the models without Ricardian equivalence, the clean distinction between active monetary policy, or passive fiscal policy, or vice versa, that is no longer that relevant. It is fiscal policy and monetary policy jointly that determine the properties of the equilibrium and, therefore, determine inflation, but also the debt dynamics. High-level implication of that is that monetary policy needs to be mindful of fiscal policy and vice versa.

Beckworth: Let’s apply it to the real world today, then. I believe in the US, I think currently you could say we’re in a non-Ricardian fiscal policy world. Do you think that’s a fair assumption if we start there?

Rachel: I think there’s ample evidence that it’s the consumers in the economy that are non-Ricardian. When the US government sent checks to people, demand reacted to that. I think that had implications for demand and inflation as we have seen in the last few years.

Beckworth: We have non-Ricardian fiscal policy as perceived by the public consumers’ households, and yet we continue to run deficits. What you’re suggesting your model shows is that those deficits are inflationary. They are contributing and will continue to contribute, as well as if the Fed follows something like a Taylor rule and it thinks through that prism, it raises interest rates to correspond to their increase in inflation. It’s just further adding fuel to the inflationary fire because those interest payments are going back into the economy. It looks like your model does not have a great outlook for the US inflation scenario. Is that fair? Would you be worried about inflation in the US based on what you’ve shown in your paper?

Rachel: We show in the paper that there is this channel that you mentioned, that the higher interest rates feed back onto the debt dynamics, and that further stimulates demand. That is true. I should mention, however, that this channel is overwhelmed by the traditional type of response that monetary policy tightening leads to lower inflation if we have long-term assets in the economy. So in particular, if the government debt is long term, which it is in reality—governments don’t just issue very short-term instruments—then the value, the price of the long-term government portfolio, reacts to changes in interest rates. The price of government debt declines when interest rates are raised, and central bank can maintain the control over fiscally driven inflation. 

More broadly, the key implication of the paper is that monetary policy will have these fiscal implications. If we have a central bank which leans against inflation, as I think it should, then it puts even more emphasis on the fiscal policymaker to behave responsibly, especially during the times of relatively low unemployment and relatively good place in the economic cycle.

Policy Recommendations

Beckworth: Lukasz, in the time we have left, I’m going to give you the final question here. What are your policy recommendations for the Federal Reserve, for Congress with fiscal policy? What would you leave with them based on what you found in your paper?

Rachel: I must say that the normative part of the paper is very much work in progress. We’re working on particularly the optimal policy in the context of our framework. I think the recognition of the interactions when non-Ricardian features are taken into account is the first and necessary step toward better macroeconomic policy.

We have to think about the coordination between the two parts of macroeconomic stabilization policy, the fiscal policy and the monetary policy, without impeding on the independence of monetary policy at the same time. I think that opens up an interesting agenda for the future, both for practitioners and for researchers as well.

Beckworth: With that, our time is up. Our guest today has been Lukasz Rachel. Lukasz, thank you so much for coming on the program.

Rachel: Thank you so much for having me. It’s been great fun.

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.