Joe Gagnon on Inflation Progress and the Path Ahead: Breaking Down Jerome Powell’s Jackson Hole Speech

Joe Gagnon is a senior fellow at the Peterson Institute for International Economics and was formerly a senior staffer at the Federal Reserve Board of Governors. Joe is also a returning guest to Macro Musings, and he rejoins the podcast to talk about Fed Chair Jerome Powell’s speech at the Jackson Hole Economic Symposium. Specifically, Joe and David talk about the future direction of r star, what current inflationary trends mean for the Phillips curve, the Fed’s commitment to a two percent inflation target, and a lot more.

Read the full episode transcript:

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

David Beckworth: Joe, welcome back to the program.

Joe Gagnon: Good to be here.

Beckworth: Great to have you on. I'm excited to have you on because I want to talk about Chair Powell's speech at Jackson Hole. We were all anticipating that. What's fun about it, Joe, is we really didn't know what he was going to say, a lot of speculation, but we were all waiting for it. He gave the speech, and I want to use that as a motivation or a launching pad to talk about a number of issues that he brought up in this speech that are very relevant to the macro environment today. Let me just list them here and then we'll work through them. First topic I want to bring up motivated by his speech is, what is the neutral rate, or R-star, and where is it heading? There's been some discussion over that. Also, why has inflation come down? He touched on that. What does that mean for the Phillips Curve? A lot of debate on that that we'll come back to. Finally, will the Fed revisit its inflation target? Chair Powell has very firmly committed to 2% presently, but what does that mean for the future?

Beckworth: We'll cover those, Joe, and use Chair Powell's speech as a launching pad, as I mentioned. Let's start with R-star and its future path. Where do we think rates, equilibrium rates, real rates will be headed… the trend path of rates going forward? Let me quote from his speech, because he touches on this a bit, and he said at his Jackson Hole speech, "We see the current stance of policy as restrictive, putting downward pressure on economic activity, hiring, and inflation, but we cannot identify with certainty the neutral rate of interest, and thus there is always uncertainty about the precise level of monetary policy restraint." Then later he says, "As is often the case, we are navigating by the stars under cloudy skies." So he's referring to R-star, and as you know, Joe, he gave a speech, I believe, in 2018 where he talked about navigating by the stars and expressed some skepticism that we really don't know these with precision. This is a big question, where are rates headed? Do you have any thoughts on that? Where do you think equilibrium rates are going?

Breaking Down the Future of R-Star

Gagnon: Well, to cut to the chase, I think, probably, R-star is going to be a little higher over the next few years and possibly longer than it was pre-pandemic, but it's not going back to where it was in the '80s or '90s. I think the demographic factors are too strong to keep it down, but I think the thing pushing it up is fiscal policy seems to be a lot looser, especially in the United States, but maybe in some other countries and perhaps it will continue to be both for defense reasons. If defense budgets are going up because of the Ukraine war and because of climate change, if we're going to be spending a lot of money on climate investment, particularly government borrowed money, those things would tend to push up the equilibrium rate. I'm not an expert on this. I've been reading different people and I think it seems like the demographic factors are still stronger than that, but that may partially reverse.

Gagnon: Anyway, there's a lot of uncertainty, as you know. Two of the main measures that people are looking at… in fact, David, you posted this recently on Twitter, now, X, whatever that is, showing that the measure for the Richmond Fed and the New York Fed, which were not too dissimilar for a while, are now diverging, with the Richmond Fed one going up and the New York Fed one staying low. The reason is, I looked into this just to understand it, and the New York Fed, Laubach-Williams one basically is mostly driven by the trend growth rate. With labor force and productivity growth still relatively low in recent years, it's not surprising that R-star is still low. The other measure, however, looks at inflation. It says R-star is the level of interest rates that stabilizes inflation. Since inflation went up, I think that's why it's gone up lately. The Laubach-Williams measure does not look at inflation. So, it all depends on what you assume, and that's a big difference in assumptions. I think both have issues. I think there are theories about the natural rate that are more complicated than that. I'm just not sure what to think. I think somewhere in the middle might be right, but not because between the two they have everything covered. It's just my gut feeling about fiscal policy.

Beckworth: Yes. I share your view on the demographic issue. In fact, John Williams, in a talk he gave, he stressed that as well. He said, look, R-star could go up, but he believes these demographic factors, which were there before, and are even more pronounced now, will go in the other direction. Of course there's those who say, actually, demographics will move it in the opposite direction because older people will start spending out of their savings. But, let me go back to the two measures of R-star that you mentioned. The Laubach-Williams out of the New York Fed, and the other one is called the Lubik-Matthes out of the Richmond Fed. I put this on Twitter and it received a lot of commentary.

Beckworth: One thing I put on Twitter with the one from the Richmond Fed was the standard air bands. Some people said, "Well, they're not that different. They're within the confidence bands." The problem when you say that is that the Lubik-Matthes confidence bands go from 1/2% to 4 1/2%. It could be anywhere, so there is a lot of uncertainty. The other thing I would note, if you want another measure, is to look at market estimates. By that I mean look at TIPS measures and I like to look at TIPS measures five years, five years out, so the average real rate over five years, five years from now. That has gone up quite a bit.

Beckworth: That I believe is over 1%. Joe, as you have told me and corrected me on previous shows, markets often get things wrong. Markets could be just overreacting right now to what's going on with the Fed. This raises questions about theories of interest rates. What really drives them? Going back to that comment you just made, we think of things like trend growth being driven by productivity. Productivity affects it, labor supply affects it, people's time preferences affect it, and it's just hard to know in real time what is going on. You see, I guess, and I think I do too, demographics is probably the biggest part of that story.

Gagnon: Yes, I think so, and certainly the longest lasting. Fiscal… who knows where fiscal could head? We might go back to a lot smaller budget deficits. For the time being, it seems we've got a huge increase in debt.

Beckworth: Yes, and this is an issue that's going to be debated and we'll see it play out. Of course, it has implications for the path of Fed policy. If, in fact, R-star is still low, meaning the long-run version is low, then the Fed will eventually have to lower rates down, otherwise they would have created a severe recession. On the other hand, if it's higher, then we'll have rates higher, and that will have implications for everything from mortgages to interest expense on the national debt. But, this is a debate going on and there's an article in The Wall Street Journal recently titled, *Why the Era of Historically Low Interest Rates Could Be Over.*

Beckworth: They allude to the issues we've touched on, the deficits. Some people think the deficits are going to be permanently higher. I'm not sure we can make that assumption. Some of them also take the opposite view on demographics that retirees in industrial economies will start spending their savings. This, to me, speaks more to a short-run R-star, but they noted that economic growth is now running well above trend. If you look at the Atlanta Fed's GDP Now measure, it's almost at 6%. That's true, but I think that's more of a short-term phenomenon versus more of a medium to long-term phenomenon.

Beckworth: This is an interesting debate to follow for sure, and I think the demographic question is the most important in my view, and it will be interesting to see going forward whether our interpretation of it makes sense or the other interpretation… people start spending out of their retirement. Let me just summarize why I think our view is right, because I think there's two things that work. One, if you live longer, you're going to be saving longer and have to prepare a bigger cushion for retirement. Secondly, the older you get, the more risk averse you get, and that means, on balance, your portfolio will be weighted more towards safer assets versus risky ones. That would be my argument for it.

Gagnon: I would just make a little note that my colleague Olivier Blanchard has noted too. Growing longevity, especially if it's concentrated above the retirement age, so that working age population isn't affected by this, but the retiree population is growing because people are going to live longer, that doesn't affect growth rates because it doesn't affect the working age population. Without any change in the growth rate, if people are living longer in retirement, they're going to want to save more. I think that's what you just said. That is separate from the low birth rate part of it. I think that's a point my colleague Olivier has made, and people who've made the point that, well, retirees are going to start spending more as they get older, that's not true if they're living longer than they expected.

Beckworth: Lots of discussions even now about moving the age of Social Security back, right? Beyond 65. If that happens, what do you have to do? You have to work longer, save more, and so that changes the dynamics, indeed.

Gagnon: I would note that you might counter argue, well, but longevity is not increasing, in the US at least, which is a bit of an anomaly, but that's entirely among people below 65. It reflects drug overdoses and what people call deaths of despair or whatever. That's a different thing and it's not going to change this thing. Once you get to 65, even in the US, your longevity is increased.

Beckworth: Great point, and an important distinction to make there, so thanks for doing that. Let's move from that point about the future of R-star. We could spend a lot of time there. Again, I think time will tell, but we'll find out five years from now what the true path of R-star is. Let's go to another point that Chair Powell made at his speech at Jackson Hole. This deals with the question of why inflation has come down, why the disinflation from last year. I'll just read an excerpt from his talk here. He says, "Good morning. At last year's Jackson Hole Symposium, I delivered a brief direct message. My remarks this year will be a bit longer but the message is the same. It is the Fed's job to bring inflation down to 2% and we will do so. We have tightened policy significantly over the past year and inflation has moved down from its peak, a welcome development. It remains too high." He then goes on to talk about three particular forms or three particular components of inflation, goods inflation, housing services, and non-housing services.

Beckworth: Then he says, "Turning to the outlook, although further unwinding of the pandemic related distortions should continue to put downward pressure on inflation, restrictive monetary policy will likely play an increasingly important role. Getting inflation sustainably down to 2% is expected to require a period of below trend economic growth." He's acknowledging success, a lot of things happening, supply side things happening, but he is suggesting that moving forward there's going to be a bit more need for monetary policy tightening and below 2% growth. Joe, let me take those comments and then just put them side by side with the actual developments with inflation. Just to recap, last year we had the CPI peak around 9%. The PCE measure which the Fed prefers was around 7% and now we're down to about 3.3% for both of those measures or close to that. Core PCE and core CPI have come down a little less, so core CPI is around 4.7% and core PCE is around 4.2%. So, there's still a ways to go with this core measure which we typically think identifies the trend path going forward. Progress has been made, still we have a ways to go. What is your assessment of it?

Assessing the Recent Inflationary and Growth Trends

Gagnon: I share Powell's assessment. It seems that… and I guess we'll get to this, but there's different factors behind the inflation which we'll talk about later. Some of them were very clearly transitory and some are more longer lasting, and we're seeing that as the transitory ones play out but the longer lasting ones persist, and I agree with the way he breaks it down into core goods, housing, and then other services because I think that's a sensible way. We know housing is going to bring us some good news finally in the next few months or more, because we saw that comes with a lag. We see the new rental price indexes have been falling for quite a while. That's good. The thing that probably doesn't get as much attention as it should, I think, is core goods. That actually fell in July, and I think there's room to go. I did a blog post last December in which I really played up that core goods, especially auto prices, are way out of line with wages, and once you control for energy and input/commodity prices, those things, they have a markup on wages that is way out of line with trends and there's room for it to fall.

Gagnon: I think those prices are a little less sticky than service prices, and they can fall, where service prices never fall. I think there's more potential for good news in core goods prices than people are thinking about. I think that would be great because it would help wages catch up on inflation. Then the sticky piece I think that's going to be much lower to come down is the non-housing services, because those are basically wages. If you plot the earnings cost index, appointed cost index and the non-housing services, they're parallel lines. They move very much the same. That's going to take wages to slow down, and that's going to take a while, but hopefully, it'll happen, especially if core goods prices fall, because then workers will feel like [they are] catching up. It will give us all some comfort that we can have less inflation and less wage growth without losing ground to inflation, so that would take time.

Beckworth: I'm glad to see core goods coming down as well, and I made a call a year or two back that this inflation would be transitory for that reason, and it's been proven wrong. It lasted longer than I thought it would, but I believe in capitalism, I believe in globalization. I know we have reshoring or friendshoring, but still, capitalism by its nature with tradable goods is going to drive down prices to the lowest cost producer. That's just the fundamental that has to reemerge, and like you said, it will feed into wages and other parts of the inflation measure. On the wage front, Joe, we also heard some good news this week from the JOLTS data, job vacancies are down quite a bit, and then the quit rate, which some say are an even better piece of information from the JOLTS data, is down to its pre-pandemic level.

Beckworth: They say this is closely tied to wage growth, so that may be another sign that things are cooling down. The only caution I would put out there is, I think I mentioned this already, is the GDP Now measure from the Atlanta Fed, last time I checked it was getting close to 6% for the third quarter, which is way above trend, and that in turn translates into income. Its spending becomes someone's income and that would portend to maybe higher wages. Any thoughts on what's going on with the growth of the real economy, apparently so rapid?

Gagnon: That's just one measure and I think I saw Paul Krugman mention there's, I think, leading indicators going the other direction [inaudible]. There's other indicators that show some slowing. I don't know what's going on. I think the labor market is my… I would focus more on that, and as you say, on the one hand unemployment is very low, but on the other hand, as you say, vacancies coming down, quit rates and some other measures showing some bit of cooling, so fingers crossed, but the probability of a soft landing, if anything, has gone up a bit in my estimation.

Beckworth: Yes, for sure. If we do have the soft landing, and again, you could argue we are already in the midst of it, kudos to the Federal Reserve for pulling this off. You could say it's luck as much as it is good policy combination of the two.

Gagnon: There was a bit of a debate, how much credit does the Fed get for this? Because are the channels that monetary policy would affect to cause this really operating? The answer is not really very much. A lot of this is stuff that was just temporary. It was going to reverse, the Ukraine war and commodity price spike that has run its course and some COVID disruptions that have been unsnarled and untangled, so people are willing to go back to work, lots of things like that. That has nothing to do with the Fed and it was just unwinding from Ukraine and COVID, but I think the way to think about this is that the Fed was ridiculously loose a year or two ago.

Gagnon: It was just way too loose for the economy. It was contributing to the inflationary pressures, and it has reversed that mistake and unwound that mistake and it needed to, or things would've been even worse. In that sense, it certainly is helping. Also, the other piece, which I think we'll come to more later, is a big part of this is expectations, and I know people poo-poo the idea that workers and firms literally have rational expectations and can predict inflation with any model. But when I say expectations, I mean people's sense of what's normal and what's going to continue to be normal.

Gagnon: It's based on their lived experience plus their interpretation of the institutional framework that they live under. Even if they don't listen to Chair Powell's speech, they know that the Fed's in charge of this, and people say good things about the management of the Fed. That seeps into their consciousness, plus the fact that they lived through 20 years of low inflation, those two things really have anchored people's view of what is normal and kept price increases in check, and made it possible for inflation to come down quickly. If that weren't true, if we were in Argentina where people didn't have that view, this would have turned out very differently, of course.

Gagnon: And it has turned out very differently in Argentina because people don't share that view of stability both historically and institutionally that we have. I think that's so important. As macroeconomists, we have to credit that as being critical for advanced economies, why advanced economies typically can handle these things better than some emerging markets, because they have better institutional frameworks and better history of policy and outcomes that people rely on and believe in and it affects their behavior in deeply rooted ways. It's not about pure rational expectations, but it is about expectations.

Beckworth: Yes, I'm sympathetic to that as well and if we compare this period to the Volcker recession or disinflation period, there had been 15 years of rising inflation, whether it was the Fed's fault or it was supply shocks, nonetheless, it got baked in and so it was much less credibility coming into that. The Fed had this high sacrifice ratio. It had to exact a lot of pain because there wasn't that credibility you just outlined. So, we can think of the Fed tapping into its credit at the credibility bank and using it to make it easier to come down from this high inflation.

Gagnon: Absolutely. I think that's so important. They use that credibility, but they also don't want to waste it. It's a tricky road to follow, but I think you've got to give them credit for doing, I think, the right thing. After making a mistake two years ago and being too easy, I think they're now on the right path.

Beckworth: Yes. Okay, let go back to this point you were making earlier about the fundamental drivers of the inflation. Supply side, demand side, it sounds like you're pointing to more supply side disturbances overall.

Evaluating the Fundamental Drivers of Recent Inflation

Gagnon: Well, I think the inflation that's come down lately is supply-side inflation mostly, but I do think there's demand side. I think the demand side inflation has been the more persistent inflation. I think there are different views as to whether this bulge of household savings has been spent or not. I think you've seen there's different studies at the Fed on that. I think you recently posted something on bank deposits or something that still looks like the most liquid of assets is above trend. That harks back to the money supply views of all of this. I think that people still seem to be willing to spend a lot despite high-interest rates. I don't know how much of that is still money in the bank that they can afford or whether the strong job market is making them feel comfortable, I don't know. It's interesting how robust the economy's been. I also think fiscal policy is turning out looser than people expected, which is supporting growth.

Gagnon: A lot of states had these big pots of money from the fiscal packages still, rainy day funds from the fiscal packages that they can tap into. I think there's been a lot of underlying sources of demand that have supported strong demand in the US. I think that is the main thing that still is going and the Fed is worried about. That's what Fed policy should hit and it did hit housing for a while. I think the worrisome thing is that house prices haven't fallen despite huge rises in interest rates. Does the Fed need to do more? That's the tricky thing. I don't know.

Beckworth: Yes, it is more challenging in the US when you have 30-year mortgages with fixed rates, getting that interest rate effect. Although we do see evidence of housing slowing down because of the Fed's tightening. That's something Powell mentioned in his speech as well, that housing is slowing down, at least that sector from the Fed tightening.

Gagnon: Housing starts have slowed, yes, but house prices, I'm surprised that they have not fallen as much, but I guess housing starts is what's more relevant than inflation.

Beckworth: To the extent we do have this aggregate demand-driven inflation that's hanging around and it was a part of the inflation story, what theory should we use to interpret them or to think about them? I'm going to throw four theories out there, Joe, you tell me your favorite or maybe a combination of them. One theory might be the old hydraulic Keynesian theory, an undergraduate textbook story that we had too much stimulus, the size of the 2021 stimulus, ARP, relative to the size of the output gap.

Beckworth: The output gap was around 400 billion, and we added over a trillion dollars. Another theory would be the New Keynesian theory that talks about keeping rates below the equilibrium or R-star. You could tell that story, the Fed kept rates too low for too long during this period when the economy was recovering. A monetarist story might say there's too much money. Of course, the money was dropped via helicopters, so it wasn't a pure, maybe, monetarist story, but still, the quantity of money was a lot. Then finally, I'll throw out the fiscal theory of the price level that says large persistent primary deficits, currently unexpected, are going to drive the price level up, in this case, higher inflation. How would you make sense of the demand-driven side of inflation we've seen?

Gagnon: I would go in the reverse order. The fiscal theory, I simply don't believe is relevant in the United States. I don't believe people are at the point yet where they think the Fed is going to be forced to do something because of fiscal policy. The Fed's job is to control inflation. Until Congress acts, the fiscal theory of the price level can't rule, and I don't think anyone expects it to come and kick into gear. I rule that one out. Then I think the other three all have some relevance, but I think the first two that you mentioned, one is all the stimulus and overhang of Keynesian pump priming measures, and then the Fed being too loose, are both correct and are not inconsistent. I forget the third one of--

Beckworth: The monetarist one.

Gagnon: Oh, right. I'm not sure what that adds, and I worry about in an era where money now pays interest, I really think the old, looking at money aggregates isn't the same anymore. I don't know that that adds anything, but anyway, it's not inconsistent. I think we had way too much stimulus, but at the time I thought I didn't see the supply shock coming and I just thought, "Well, yes, we'll have a little bit more inflation, but we'll have an economic boom.” We didn't because the supply curve actually shifted the other way, so we got more inflation and not more output, so that was unfortunate. Initially, I think we were facing a crisis and it's a once-in-a-century shock and better to do too much than too little. I firmly believe that. I think it was better, what we did in 2020, 2021, than what we did in 2009, so 2010 we didn't do enough, then we did too much later, but I would rather have too much. I think the problem is that the Fed should have been tightening two years ago instead of one year ago, I just think the Fed was behind the curve, it should have seen this coming.

Gagnon: The whole private forecast… I think we even talked about this on my last podcast, David. I just don't understand how the whole professional forecasting community missed this, the Fed and the professional forecasting community both share equal blame in just missing… as you said, the output gap was much, much smaller than the fiscal stimulus, even more than you said, I think, because I think the fiscal stimulus is more than a trillion dollars today. There were reasons to think the multiplier would be somewhat lower than normal, but it was so much bigger than that, that I just don't know how the forecasting community missed this one, I still don't understand.

Beckworth: Well, let me throw something out there to help the forecasting community and those of us like myself who made an inflation forecast, it was way off back then. Maybe it goes back to what we were talking about earlier, the credibility of the Fed, maybe just kind of taking it for granted, the Fed has consistently kept inflation close to its target, therefore, it will going forward, no reason to expect this to be any different, and it was different. This time it was different that the Fed did fall behind, so to speak, and again, it's just the uniqueness of the situation, I think, played into that as well.

Gagnon: It's a strange equilibrium outcome. It's like, well, people didn't expect it because the Fed was in charge and so it was going to take… they believed the Fed would handle it and the Fed didn't see it coming because people weren't worried about it. They both were looking at each other and no one was… and only a few weirdos like me, but also Larry Summers and actually, at Peterson, a number of them, Adam Posen, Jason Furman, Larry Summers, and we all have connections with Peterson, we were the four that I remember saying,… and Olivier Blanchard; five [of us saying], "Wait a minute, the numbers just look too big to us."

Gagnon: I called forecasters up and said, "What are you thinking?" Basically, what I was told was, "Well, we believe it's going to be saved, we just believe the saving rate is going to go way up. People aren't going to spend this money for various reasons, they just don't want to go outside their door, because of COVID blah, blah, blah,” so it was unusual. They just believed in a huge increase in saving rate, which I guess… it's not like they didn't think about it, but anyway, I don't want to belabor the point, because nobody saw the inflation going as high as it did, even Larry didn't see that.

Beckworth: Yes, and again, we've come a long ways from that point. Again, CPI was 9% and now we're down close to 3%, that's a huge drop in inflation over the past year, close to six percentage points, a lot of progress has been made, a lot of that is due to luck, or these things working themselves out, supply side issues, and some of it is due to the Fed, but there's a lot more work to do for the Fed. Alright, let's take that disinflation success and apply it to another controversy, and this is actually something Chair Powell noted in his speech. If you looked in his speech, he actually had a footnote where he referenced a nonlinear Phillips curve. I know you're a big fan of nonlinear Phillps curves. In fact, we may have talked about this in the last show, I believe, we talked about the paper you did on nonlinear Phillips curves. So the Phillips curve, as most listeners on the show know, is kind of like the workhorse mechanism that most macroeconomists think through when they think through, how is inflation generated, at least in the short run. I think, in the long run we can talk about other things, but short run is the Phillips curve.

Beckworth: You can tell several different pieces of the Phillips curve puzzle, it could be nonlinear. Also, inflation expectations may be an important part of the story, which you touched on earlier. Phillips curves also have a supply shock term in them as well, so you have got several different pieces where you could tell a story through a Phillips curve going forward. Of course, there's been many who have commented on the fact that we've had unemployment relatively low, inflation dropping dramatically, and maybe you could go tell, through one of those pieces there, the story, but I asked this question to Ricardo Reis, and he was on the podcast a few weeks back and I said, "Can we can we still use the Phillips curve? What's your take on this?"

Beckworth: So, I want to read his reply because it generated a pretty long conversation on Twitter between him, Olivier Blanchard, your colleague, Paul Krugman, and others, and I just want to read his part and then you can respond because I think you've read this exchange as well. It goes on, back and forth, and Ricardo has some fascinating things to say, and then Olivier. It's fun to see in real time, this debate in macroeconomics by big names like Ricardo Reis and Olivier Blanchard. I asked him, "Can we still claim the Phillips curve?" He says, "David, let me start by saying that I am the Phillips Professor, London School of Economics, so I have to defend the Phillips curve," which was a great remark.

Beckworth: He goes, "I can never say it's obsolete, otherwise I would fall on the floor since my chair would become obsolete. So, it's definitely present, but with that account, let me make three observations. The first one is that, the way I understand monetary policy is whereby tightening monetary policy, a central bank is able to bring inflation down. In the same way that when I go to the doctor with an infection with a bacteria of some kind, antibiotics are the way to kill the bacteria and cure me from that. However, a side effect, and I emphasize, let me say it slowly, a side effect of raising interest rates is that you also cause a recession. You also lead to an increase in unemployment. In the same way that a side effect of taking antibiotics is that they tend to wreak havoc on your GI system.

Beckworth: Note that this is not a channel. It's not by taking antibiotics and screwing up my intestines that I, therefore, kill the bacteria. No, it's a side effect. Likewise, raising interest rates lowers inflation and has a side effect of unemployment, but it may not lower unemployment the same way that you may go through a course of antibiotics and be perfectly fine with your gastrointestinal system. The fact that unemployment has not gone up does not in any way discredit the way in which monetary policy works." I'll stop there, but he's making this point, and I think you've got the gist of it, that we shouldn't necessarily expect unemployment to have to change with the drop in inflation. Then Olivier Blanchard responds. They get into this discussion, what is the key issue in the price level in macro? That's a long setup for a question, Joe, but I'm going to ask you, since you're a Phillips curve fanboy, how do you make sense of what you've seen over these past few years with disinflation and low unemployment?

Grappling With the Phillips Curve in Today’s Macro Environment

Gagnon: So, I read that discussion, and I think Ricardo… I think it was unfortunate. I don't like the way he said it's just a side effect. I think that's wrong, and I'll come to that. Otherwise, I think they were actually talking a little bit past each other and not disagreeing that much except for that one point. The Phillips curve that now… the standard New Keynesian Phillips curve has several components, one of which is expectations, and I think that's what Ricardo was talking about a lot. Another of which is the actual effect of unemployment on inflation net of expectations, and that was what Olivier was talking about, and I think that's not a side effect. I think it was wrong for Ricardo to say that was a side effect. No, that is the main mechanism in some sense. Then, of course, as you said, there can be controls for supply shocks which are outside the model. Those are the three parts of it.

Gagnon: I think what we’ve had were supply shocks of a type that weren't just controlled by food and energy prices or commodity prices because we had the supply curve breakdowns, and we also had a withdrawal of the labor force. I think the COVID shock, especially in the US than maybe other countries, is the first time U-star, the natural rate of unemployment, rose for non-demographic reasons in the post-war era. Normally in my work, U-star, the natural rate of unemployment, which a lot of people don't believe in, and they don't believe it's very stable, it is quite stable, has been stable historically. It's explained very well by the aging of the labor force in that it rose in the '60s and '70s, and then it's fallen since then as the labor force has aged. Once you do that, you get a nice stable U-star and the Phillips curve fits pretty well. COVID was the first withdrawal of people from the labor force for a non-demographic reason in the post-war period, and it was pretty big, but it seems to be over now, and so it raised U-star temporarily and then lowered it.

Gagnon: I think the big piece of this that people forget is that the expectations term has been very stable. I think the one that makes most sense in these models and fits best is the longer run, the real long-run expectations, which as we've seen the data on long-run expectations have been remarkably stable. We're only back to where we were before the Great Recession. They drifted down a bit and then they come back up, but they really have been stable, if you think about it. In that world, in which people have faith in the Fed in the long run, and they don't believe that we're in a new world, I think they thought… they realized COVID was a big shock and Ukraine was a big shock, but they didn't think it was going to permanently change inflation. I think that's what the evidence says. They knew it might temporarily change it for a year or two, but they didn't believe it would permanently change it. That has really been important. Because in the Phillips curve, if you look at some plausible measure of where unemployment was relevant to the natural rate, that would not have pushed inflation up to 7% or 9%. It wouldn't have. Even with a fairly steep Phillips curve, it might have done 3% or 4%.

Gagnon: What you've got is, you had a supply shock that boosted inflation way above what the Phillips curve would've done, but once that supply curve recedes, you go back to where the Phillips curve would do, because expectations didn't change. The Phillips curve would've put inflation at 3% or 4%, given where unemployment is, and the natural rate. It's a bit tricky because knowing what the natural rate did in COVID is hard, but I think we're back to where we were pre-COVID, which is a, maybe, natural rate around 4%. We're basically, with unemployment, close to, slightly below the natural rate, so inflation should be slightly above 2%, and so maybe that says 3%, and maybe that's where we're heading. The Phillips curve seems to explain, pretty well, where we are, and it seems to me, once you've taken account of supply side and including the fact that people didn't want to work. Maybe I missed something, but that's my interpretation of what happened.

Beckworth: So, the Phillips curve has held up, and you just explained it through those different pieces: the supply term, the actual expected inflation term, and then you have that output gap term or the labor market term. What about the nonlinear element of it? How important was that?

Gagnon: Oh, very important, because people who had estimated a flat linear Phillips curve, it was so flat that no amount of unemployment would have any noticeable effect on inflation. It was ridiculous. If you push people who believe that hard enough, you realize that they actually, implicitly, always had a nonlinear curve. Because if you said, "Well, are you saying literally that if unemployment gets down to 1% that there's going to be no inflation?" They all said, "Well, no, no. Something would change. Something would break." Well, that is admitting that there's a nonlinearity. You're admitting that. If you think inflation would start to rise by more at some point, what you mean is the curve isn't linear, that there's a nonlinear piece to it. I think everyone knew that, and even people who don't like Phillips curve models sort of believe that.

Gagnon: I think, to me, that's all evidence for a nonlinear Phillips curve, the fact that we had more inflation. It may be that when Phillips drew the Phillips curve, it was a hyperbola. Think about it. If you remember your geometry, what a hyperbola was, well, it goes from being perfectly horizontal for very high-end rates of unemployment to being perfectly vertical for very low rates of unemployment. The statistics in his paper were pretty much nonexistent. It would never get published these days, but that's what he drew. Something like that may be at work. I've just found a simple linear kinked Phillips curve, where one part is very flat and one part is moderately steep, that there may be an even steeper part which we might have hit briefly, is one way of thinking about it, but it's not well-estimated because we just haven't had as tight an economy as we did a couple years ago in the post-war period. It's been a long time since we've… the Korean War may have been the last time we had something like this, so it's just hard to estimate.

Gagnon: Oh, there's one other piece, though, that I think I'm trying to work on for my future research, but I don't know. It's not coming together too quickly. If you think that there's a nonlinear price adjustment mechanism, don't call it a Phillips curve, just say price adjustment mechanism, nonlinear Phillips price adjustment mechanism that goes industry by industry, sector by sector, and it's of this shape that we think that the hyperbola, or whatever, that we think that low rates of unemployment or high rates of capacity utilization have a bigger effect than excess capacity. So excess demand has a bigger effect than excess supply. If that's true, then when people shift demand across sectors, which is what we saw in COVID, we saw the biggest shift in post-war history from services to goods, and now going back, bigger than any time, again, since the Korean War.

Gagnon: That, by itself, even if there was no increase in total demand, just the shift of actual demand, of steady demand, across sectors, is inflationary. Because if you're starting out from equilibrium, the sectors that are getting more demand are going to have big price increases, and the sectors with less demand will have very small price declines, and the net effect is going to be inflationary. I think that is a piece that the aggregate Phillips curve doesn't capture well, but may have played a role in COVID. I'm trying to write this up and get some results, but it's been tough. It's tougher than I'd hoped. There's one empirical regularity that supports it, which is if you look at the shift of consumer demand between services and goods, the two biggest shifts in post-war history were in 1950 and 2021. Those were the two biggest spikes in inflation in post-war history, and short-life spikes in inflation, and so it says that maybe there's something to this. Maybe this nonlinearity in the Phillips curve across sectors is a piece that our aggregate Phillips curve doesn't capture well because aggregate spending doesn't have to increase for this to operate, you see? So, the aggregate Phillips curve, aggregate unemployment might not change, but if it's across sectors, it's going to look like a supply shock to your aggregate Phillips curve.

Beckworth: So, the challenge in uncovering that is that you have got to find the spending data for each of those sectors, as well as the cost, and that can be a challenge, empirically.

Gagnon: Yes, and I've just looked at, broadly, just services versus durable goods. What I told you was true for that shift. Yet, when I try to put it into a regression, we haven't been having much luck, so I don't know. Maybe, then, it won't pan out, but this is my-

Beckworth: You're hopeful.

Gagnon: -next project. I'm hopeful.

Beckworth: I like your framing, a nonlinear price mechanism. Is that what you said?

Gagnon: Yes, because Phillips curve gets so many people upset-

Beckworth: So, we'll start calling it the PM instead of the PC, you know, the Price Mechanism. I guess, let me approach the Phillips curve this way: can it also be seen as another version of the short-run aggregate supply curve? It's just kind of restated, right?

Gagnon: I think so, yes.

Beckworth: Yes. So, I think some people who have hesitation of the Phillips curve, it can be thought of simply as a form of a short-run aggregate supply curve that's nonlinear. Most undergraduate textbooks draw it as nonlinear. The only thing that is linear is the long-run aggregate supply curve, and we can shift that if capacity changes over time. The other comment about this, and this goes to the flat Phillips curve that we had prior to this, I think a lot of it has to do with identification issues. That is, because the Fed has been so effective, pre-2020, in keeping inflation close to its target, you're not going to find a relationship in the data between unemployment and inflation. If the Fed is in fact responding to shocks that could cause inflation to go off target or full employment to deviate and systematically responding to those, it's going to be hard to find something systematic between inflation and unemployment. That's the identification issue.

Beckworth: That's also an issue, I think, when we look at those money measures you referenced. If you go back and look at simple-sum measures of money, there appears to be a breakdown in the '80s, '90s and to the present between them and inflation. I think it has a lot to do with identification issues. If the Fed is systematically doing its job, it's effectively responding to money demand shocks. Even though it's not targeting money, per se, it is responding to shocks in a way that money supply and money demand offset each other, so you're not going to find anything. In fact, Milton Friedman calls this the thermostat example, that there's going to be… empirically, it's hard to see it, even if there's structurally a deeper relationship there. I would just throw that out there as well, that even if the Phillips curve, empirically, doesn't look present with a simple scatterplot, that’s because you have an identification problem in front of you.

Beckworth: Okay, let's go from Phillips curves to the last part I want to draw on from Chair Powell's speech, and that is the Fed's commitment to 2% inflation as their target. Here is a quote from his speech. He says, "It is the Fed's job to bring inflation down to our 2% goal, and we will do so." Then later he says, "2% is and will remain our inflation target." Your colleague, Jason Furman, had an op-ed in The Wall Street Journal right before this, calling for a 3% target. I was not surprised that Powell, of course, came out and clearly said, no, we're not going to do this. They want to maintain that credibility we've talked about, and I think that's important for them. I think, just in general, if you're going to raise your target or change your target, you want to be at that target for some time, you need to show that you're capable of doing it, you’re not just changing it for convenience purposes. With that said, we do have a Fed review coming up in 2024/2025. We've had FAIT, which may have received some bad press because of the high inflation. What do you see happening at this review? And maybe lay out what you would like to see happening as well.

Looking Ahead to the Fed’s Next Framework Review

Gagnon: Sure. Well, there’s a lot there. My colleague, David Wilcox, had a blog post today, by the way-

Beckworth: We'll link to it.

Gagnon: -August 31st, in which he said Powell's speech is very clear, two means two, and they're sticking to the 2% target no matter what. He had two points. One is what you said, which is, well, they should get back to two before they change it to prove that they can do it and not lose credibility. The other point, which I think he made even stronger, was that they should be done in the context of a framework review. I agree totally with the second point, they should be in the context of a framework review, but I'm not sure I agree with the first point, especially if during the framework review, they're not all the way back to two yet, I don't think that they should wait. If they think that the right thing to do is to change the target, they should… in the context of the framework review, they should change the target as appropriate. I would point out that in the December press conference of the FOMC, last December 2022, Chair Powell was asked about this and he said clearly, “we're sticking with 2%. Changing it is not on the table.” Then he said something that made my jaw drop, hasn't got much play. He said, “this is not the time for that,” i.e., changing the target. “This is not the time to change the target. There may be a project on that at some point.” Then he went back to, “but we're sticking with two.”

Gagnon: The point, I thought, and this is December '22, was that in the long run, when he said long run, he was thinking about the framework review, which would be two years ahead at that point. For the press and everything like that, two years is pretty long, far ahead. He opened the door at that conference that it could be reconsidered in the context of these frameworks. This is 2023. The framework review hasn't even started yet, let alone finished. He cannot change the target now and should not, and should stick with two, and should say, "We're sticking to two, and everything's up for discussion in the framework target, but unless and until we decide otherwise, we're sticking with two. We have no reason to believe that will change." He could even go further and say, "I don't expect the framework review will change it, but it might." But, it might. I mean, it should not be taken off the table like it was the last time. It should be discussed. I feel very strongly about that.

Beckworth: One of the motivations for Jason Furman's call for a 3% inflation target in the recent Wall Street [Journal] op-ed he did is the concern about hitting the zero lower bound more often going forward. That is something I think many people would agree with prior to this inflation surge. I suspect many people today would be less worried about it. I guess this circles back to our earlier conversation, what is the future of R-star? If we live in a world with low R-star, there's a chance we end up in the zero lower bound again. I guess this all comes back to this question of uncertainty. How certain are we that we will be in a low-interest rate world, in a low-inflation world going forward. If we are, I think the argument is stronger for a higher inflation target, but if, in fact, we are in a world with high R-star, higher interest rates, it just strikes me that it's a less compelling case. Is that fair?

Gagnon: I think it's fair to say that the lower R-star is, the stronger the case is for raising the inflation target, yes, but I think that even at the old R-star, two was too low of a target. There was probably a higher chance of hitting it than we thought at the time, even with the old R-star. I think the case is strong even if we go… and I don't think we're going back all the way to the old R-star anyway, but even if we go some way back, I think the case is still very strong. My research shows that there's a second reason, there's a zero-bound reason, on the interest rate policy, which I definitely think is probably the number one concern, and a higher R-star will slightly reduce that urgency, but it won't eliminate it. The other thing is downward price and wage rigidity.

Gagnon: That is true regardless of what R-star is. As inflation gets low, and 2% is low enough to see this, you start to see truncations in the distribution of wage changes, and I think if you looked at service prices, you'd see the same thing too. You get this truncation in the distribution, which indicates an unwillingness to lower certain wages and prices, which impedes the good functioning of the economy. And I think a small increase just from two to three is enough to really avoid most of that, and really to get us most of the way to a more smoothly functioning labor market and product markets. Even aside from the zero lower bound on interest rates, there's a zero lower bound on price and wage adjustment, which we don't want to be too close to. For both reasons, I think a 3% target would really go a long way to, I think, address the concerns, without being too high of an inflation rate. I think people wouldn't really complain or notice 3% inflation as opposed to a 2% inflation in the way they did 5%, or 6%, or 10% inflation. 4% seems to be a dividing line between which people think about inflation, and we want to be on the low side of below 4%.

Beckworth: Yes, there was some interesting work done on this these past few years as inflation has taken off that people really start paying attention around 4%, somewhere around there. In fact, your colleague Olivier Blanchard referenced an article that I went and read, where the people looked at Google searches for the term inflation versus the actual inflation rate. It's kind of flat, there's nothing there, up until close to about 3.5% to 4% and then, boom, it takes off. There is some threshold that when you hit it, people become very cognizant of it and then they start worrying about it. Well, you want to be in a place where people don't worry about inflation. They aren't worried about the Fed and what it's doing with its inflation target. That's a good point. You want to increase it just a little bit, but still keep it below whatever that threshold is, which is a fair point. So zero lower bound, and if I can summarize your second point, money illusion. We suffer from money illusion, and therefore-

Gagnon: We do.

Beckworth: -we need to have this extra buffer in place to deal with money illusion. Maybe in the future, Joe, when we're all perfectly rational, homo economicus, and we index everything, we won't need to worry about that. Until then, until we maintain humanity as it is, zero lower bound, and then the money illusion.

Gagnon: Yes. Well, look, if you want to be at a place where there's probably no money illusion, again, look at Argentina. People are very used to thinking about inflation and inflation-adjusting. And I think… I don't think we want to live in that world. I think we'd rather live in a world in which inflation is so low and stable that people don't start thinking that way.

Beckworth: Yes, for sure. Okay, so last point on this framework review, and Joe, you know I have to bring this up, of course, as a nominal GDP level targeter. I hope that it at least gets some discussion, some consideration, maybe even more than that. I will mention that we're having in a conference, some events, some policy briefs coming out leading up to this review, so that nominal GDP targeting or nominal income targeting will get some consideration during this time.

Beckworth: But one thing I do hope we take out of these past few years is the benefit of makeup policy, and that we don't miss the lesson and look at just the high inflation. I mentioned this to Adam Posen on Twitter. I said, "I hope we don't throw out the makeup policy baby with the high inflation water." My concern is that we don't look at the win, we just look at the loss or the inflation problem. Yes, it's true we overdid it, we overshot, but it's also highly remarkable that we returned the trend path so quickly, so rapidly. That wouldn't have been possible without, at least implicitly, makeup policy. Maybe it was explicit with FAIT allowing it. And yes, again, we need to learn how to better fine-tune that, make it more precise, but there's something there, I think, a lesson learned. My worry is we'll go 180 degrees in the opposite direction and never touch it, whatever form it might be, price level targeting, nominal GDP level targeting, but we're going to get burned by this experience. Are you worried about that too?

The Importance of Makeup Policy

Gagnon: I am. I'm very worried. Although, if you had a higher inflation target or a higher nominal GDP growth target… and I agree with you. I'd rather focus on nominal GDP growth, or a level, or some nominal GDP target that has some memory to it, put it that way. I don't know if it's a permanently sticking to a level absolutely, or it's a growth rate with some lags and-

Beckworth: Some adjustment over time, yes.

Gagnon: Yes, some adjustment over time, I don't know, but some memory, absolutely. As you say, what FAIT was talking about would be better with nominal GDP than with price level, I agree. When people say that the move the Fed did towards FAIT, which was a small move, that is somehow responsible for all we've seen and shows that it was a big mistake just because we had a burst in inflation, that's just crazy. We had the biggest shock in 100 years, the biggest pandemic in 100 years. This was a global thing. No other country had adopted FAIT. They all had inflation. That had nothing to do with the inflation we saw. It's ridiculous to blame it on that and to say that that was a mistake. The mistake was not seeing the inflationary implications of what was going on soon enough, but FAIT didn't cause that. I totally agree. FAIT was an appropriate step, a small step in the right direction, and should not be abandoned. If anything, we should enhance it. I just totally am with you here. This is a once-in-a-century shock and a once-in-a-century mistake that the Fed made.

Beckworth: Even if people don't blame FAIT, some might still say, "What's the need of FAIT, given that we have high inflation?" I bring that up because I had Zac Gross, a former Reserve Bank of Australia economist, on the show, and he noted that when they did their review, which came out this year, they left their flexible inflation target as is. There was no makeup element added to that. Now, the Fed added a makeup element. The ECB had some language about symmetry, which is maybe a watered-down version of makeup, at best-case scenario.

Beckworth: The RBA didn't go anywhere near it. They talked about it, but he said they decided not to go there because inflation was high. He said it was simply a matter of timing. I'd say a matter of dumb luck. Had the RBA done its review maybe a year or two ago, before the inflation took off, it may have also incorporated makeup policy. That's my worry, is some might say, "Well, we have high inflation. We don't need to do makeup. We've got it all taken care of," but I guess the onus is on us to make sure we make this point that makeup policy made a difference. That's why we got back to full employment so quickly. Again, we can do better next time. We can try to be more proactive and raising rates when we need to and dialing things back.

Gagnon: Look, I think a good counterargument to these people would be to say, well, how would FAIT now, the FAIT policy now, change… how would changing that policy change what you would do? It wouldn't, because FAIT is only about if you've been at the zero-bound for a while, if you've been underperforming your target for a while. We haven't been, so it's not going to be relevant. If we come to a time in the future when we have been at the zero bound for a while, we have been missing our target for a while, then it continues to be appropriate to deal with. It doesn't change policy at all now or for the foreseeable future, and yet it leaves in place something that in the future, at some point, it could be useful and if it's not useful, so what's the harm in keeping it, right? I see no cost right now to keeping the FAIT policy since it is a moot point for the foreseeable future, yes?

Beckworth: Well said. Well, with that, our time is up. Our guest today has been Joe Gagnon. Joe, thank you so much for coming back on the show.

Gagnon: It's been great to be here, David. Great to talk with you.

Photo by Natalie Behring via Getty Images