Henry Curr on Inflation, the Phillips Curve, and A New Monetarism

Inflation is losing its meaning as an economic indicator, and central banks need to respond.

Henry Curr is the economics editor for The Economist magazine, and the author of a special report by the magazine on the phenomenon of low inflation now facing the global economy. Henry joins Macro Musings today to outline this report and the big questions surrounding low inflation. David and Curr also discuss the persistent low inflation of the present around the globe, why the Phillips Curve has broken down as a policy tool, and how technology may be causing inflation to miss its target set by central banks.

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: Our guest today is Henry Curr. Henry is the economics editor for The Economist magazine and the author of a special report by the magazine on the phenomenon of low inflation now facing the global economy. He joins us today to discuss this special report and the big question surrounding the low inflation. Henry, welcome to the show.

Henry Curr: Hi, it's great to be on. Thanks for having me.

Beckworth: Great to have you on. This is a really neat read. I read through it, had a lot of questions answered, and just a great discussion of the low inflation now facing the global economy. I want to get into that in a minute. Before we do that, I want to talk about you and your career. How did you get into economics and ultimately, into journalism?

Curr: Sure. Well, economics, goodness. When I was a teenager I was really interested in politics and current affairs. That was what I was really enthusiastic about. And then I supposed I went on a bit of a journey, where I came to see economics as the prison through which I liked to understand the world and understand how it worked, I suppose.

Curr: I studied at university, philosophy, politics and economics. And then, I didn't have an inkling for a while that I would go towards the economics, but I ended up specializing in it quite strongly, and then I stayed on after my undergraduate studies. This was at Oxford. I stayed on to do the MPhil and the master's degree in economics. But I didn't want to be an academic. I never really aspired to do that in part because I'm not sure I quite got the concentration span of an academic to work on, on the very long projects that academics tend to work on.

Curr: I like to work on a wide range of things. And I went into a consulting for a while. I was an economics consultant for a year in London and I was doing mostly microeconomic stuff, antitrust cases, a bit of public policy, some regulatory stuff. And I got into journalism kind of by chance. I'd always read The Economist magazine and they advertised for someone to write about the British economy.

Curr: And I thought that sounded like a really interesting job and opportunity. So I applied and I was successful and then it's gone from there really. After a year of covering Britain, I moved to the US after Greg Ip moved from The Economist to The Wall Street Journal. I plugged that gap in the US and then last year I moved back to London to become an economics editor.

Beckworth: Okay. And my understanding is you are the youngest economics editor ever, which is quite an accomplishment since this magazine has been around since 1843.

Curr: Yes. That's what they tell me. Yes.

Beckworth: So if we're talking a very long rich history magazine and you're the youngest economics editor. So it's a real honor to have you on our show today. And no big expectations, but we have a legend with us today. Okay. So we're going to talk about something that's been a puzzle for many people, including myself and others. And I would say even more so for central bankers, like you mentioned this in your lead piece that all the major central bank heads…

Beckworth: …So Jay Powell, the Federal Reserve, Mario Draghi, the European Central Bank, Mark Carney, the Bank of England, they've all said, "Man, this low inflation is a puzzle. Why are we here? How did we get here? What's causing it, what does it mean for policy?" And you have this great special report that gets into all of this. And I want to read your introductory paragraph in the lead article that you wrote because it is really almost bizarre how far we've come from the previous problem where we had too much inflation to the point where we have maybe not enough inflation.

On The Low Inflation of the Present

Beckworth: So I just wanted to read this because you have some great history you draw upon here, and then let you run through it after I do this. So here it goes: “Inflation used to be the scourge of the world economy and the bane of American presidents. In 1971 amid an overheating economy, Richard Nixon took to the television to announce freezes on ‘all prices and wages throughout the United States’. A board of bureaucrats ruled on what this meant for everything from golf clubs, memberships to commodity futures.”

Beckworth: “Gerald Ford, Nixon’s successor preferred grassroots approach. He distributed buttons burying his slogan ‘WIN’ short for Whip Inflation Now. Ronald Reagan, running for office four years later, amid another surge in prices declared inflation to be’ as violent as a mugger, as frightening as an armed robber, and as deadly as a hitman’”. Love that. Next sentence: “Today that lethal assassin has gone missing.” So explain that to us. How did we go from a lethal assassin to a missing one?

Curr: Well, the first thing to say is I enjoyed you reading that. If you'd like a job with our audio edition, I'm sure there's one waiting for you.

Beckworth: Okay, thanks.

Curr: Yeah. It is remarkable, isn't it? That transformation and indeed it's remarkable writing the piece, the experience of telling people about it and sort of pitching the idea internally to people who don't think about economics a lot. They're always quite surprised that you might see low inflation as a problem because they remember the old days of those policies you just read out and of high inflation being a big problem.

Curr: So what happened? Well, one thing that's happened is over that very long period is the emergence of inflation targeting central banks and the independence of those central banks. And I think that's basically no question that that's been the dominant change in the macroeconomic paradigm that has got rid of high inflation over the long term.

Well, one thing that's happened is over that very long period is the emergence of inflation targeting central banks and the independence of those central banks. And I think that's basically no question that that's been the dominant change in the macroeconomic paradigm that has got rid of high inflation over the long term.

Curr: But then we also have this really curious behavior of inflation over the last decade, where it hasn't performed as people expected it to. Central banks have undershot their targets. They claimed to be doing all they can to meet them, but they persistently had to revise their forecasts because inflation hasn't been as high as they thought. And that's more unique to the last decade. And I would say that it's not something that's well understood. Some people would say, "Well, nothing's really changed in the economy and all this shows is that central banks haven't been ambitious enough."

Curr: And I don't think that's quite right because I think things have changed and we'll get on to what some of those things might be, I think later on. But something different has been happening in the last decade. That's point one. And then there's another group of people that say, "Well, things have changed so fundamentally that we need to completely rip up the textbooks. Inflation isn't a monetary phenomenon anymore. We may as well sort of give up."

Curr: I think that's wrong too. And I think that's actually a sort of dangerous path to go down because the main mess and over longer time horizon is, look, these independent central banks have done very well at containing what used to be a problem. So there's both sort of remarkable long-term success, really. And then a curious short-term failure is how I put it.

And I think that's actually a sort of dangerous path to go down because the main mess and over longer time horizon is, look, these independent central banks have done very well at containing what used to be a problem. So there's both sort of remarkable long-term success, really. And then a curious short-term failure is how I put it.

Beckworth: Yeah, that was a really nice frame [that] you put in your article, you mentioned there's a long-term story, a short-term story, the long-term one you just mentioned, it's kind of been the long fight against inflation, central banking, they won a major battle, the war against inflation. And then the short-term story is this past decade and we'll get into that later. But the breakdown of the Phillips Curve relationship and trying to make sense of that. And that's probably where the central bankers are really perplexed more recently.

Beckworth: But just some of the facts that you outline in your article are a little startling to me, I didn't realize how pronounced the low inflation problem was. I mean, I know that’s the story for the US, you hear the European Central Bank, you hear Japan, but when you put those all together, you realize it's actually a good chunk of the world economy. And there's other places like that. So you mentioned 28 of 43 inflation targeters have inflation below target or the low end of their target range. So that's quite a big miss and you put by GDP, 91 percent of the inflation targeting world is below target. So there is quite a big miss going on throughout the world.

Curr: Yes, that's right. It's absolutely something that's happening everywhere. It's worth saying that some of those inflation targeting banks I’m counting there are emerging markets. And in emerging markets, the inflation's a little higher because the inflation targets are typically higher, but it's still the case that they have been part of this long-term transformation. And it's still the case that some of them are undershooting. And when you add it all up, it very much does this look like a global phenomenon.

Beckworth: Yeah. You say that all the advanced economies basically are falling short of their targets or the low end of their targets in the cases of advanced economies. And then about half of the emerging markets are doing the same thing. So emerging markets are a little bit different, but they're falling into the same traps, maybe at the slower pace, but we're already there. So this is an interesting phenomenon and it has huge consequences. And that's why you wrote this special report. One of the things you bring up, the one I want to ask you, why is this such a big deal? Why do we care? Why write this story? We beat inflation. What are the costs to not thinking about this carefully?

Curr: I think that's a good question. And that speaks to what I was saying earlier about some people wondering whether it's really a problem at all. I'd say that there are three costs that I set out in the article. The first is that if your inflation's below target and you buy the textbook model of how the economy is generating inflation, what it means is monetary policy hasn't done its job and if monetary policy had done this job and had provided more stimulus...

Curr: …then you would've had faster growth and you might've hit your inflation target too. So if your inflation's below target you've left something on the table because you haven't pushed the economy as fast as it might have otherwise been able to go. That's one thing. The second thing goes back to the reason you have an inflation target in the first place, why is it helpful?

Curr: Well, it's helpful because it makes inflation predictable and that brings all kinds of benefits to the economy. It becomes easier to strike contracts, for instance, if people have a good idea what inflation is going to do. But if central bank stop being credible because they're constantly undershooting their inflation targets, then inflation expectations might start to do strange things. And that makes it harder to strike contracts.

Curr: And I think this is a point that is a very old one, but it distributes resources in an arbitrary way between lenders and borrowers, especially if inflation goes off target. So if inflation is lower than people expected when contracts were struck, then debts shrink more slowly in real terms. And that is bad for our borrowers and good for lenders.

Curr: That's the second thing. And then the third thing, which is I think probably the most important point here is the way in which low inflation can be self-reinforcing insofar as once your inflation expectations start slipping, that bears down on your real interest rate. And then if you're at the zero lower bound, as many economies are and other economies are close to, to the zero lower bound-

Curr: -You may not be able to offset [those] still lower nominal rates. And then you get more into the sort of secular stagnation type story of sort of getting locked or liquidity traps-like story in which you get locked in this sort of disinflationary world and that's real sort of downside risk. So I say there are three pretty clear costs that are worth worrying about.

Beckworth: Yeah. Your second one I think is really interesting because there's been recent talk in Europe for example, of how do we address the fact that the ECB has been undershooting this target by quite a bit more so than the Fed has. So the ECB's target is close to two percent and it's been closer to one percent in practice and some have suggested over there, some prominent individuals that they just lower the target. Let's just change the target to fit the actual outcome. But that's a dangerous precedent, right? Once you start doing that, what happens if inflation drifts down to half a percent to 0 percent.

Beckworth: If you keep changing it, you definitely undermined credibility. This ability to make these long-term contracts, as you mentioned. And I do think it's important to keep monetary policy effective and anchoring on the nominal size of the economy. You've got to have a credible target. So I do think this is a huge deal. So again, an important report you've written here. Let's move onto the next article in this is called *Finding Phillips* and it's all about the Phillips Curve. And maybe why don't you describe for our listeners who probably know already, but go ahead and describe what is the Phillips Curve and what is it supposed to be used for?

The Phillips Curve and its Use

Curr: Sure. It’s actually a slightly harder question to answer than you would think. One of the things I like about writing is it imposes such discipline on your thinking. And sometimes you don't know really what you think about something until you try and write it. And when you sit down to write an article about the Phillips Curve, you immediately have to decide which one you're talking about. There's the accelerationist Phillips Curve from Milton Friedman. There's the New-Keynesian Phillips Curve, which is quite different.

Curr: And then there's a question of whether you're talking about prices or wages, but the general idea behind the Phillips curve is that whichever flavor it's in is that low unemployment should on the whole be associated with higher inflation because it indicates that an economy is operating with less give, with a slack. And there should be competition for resources. And that's the general idea that sort of percolates through whatever Phillips Curve is your favorite.

Beckworth: Okay. So the general idea is that the amount of slack in the economy tied to the inflation rate, the greater the slack, the lower the inflation rate and vice versa. And that's kind of the dominant view of how we think about inflation. Is that right?

Curr: I would say so, yes. I mean, that view sort of dominates central bank communications, it's implicit in the models that you're taught in economics courses. I mean, it's a little bit more complicated than that because the Phillips Curve comes in these different flavors. I mean, it's called the Phillips Curve after Bill Phillips and his observation was empirical. It wasn't theoretical when it concerned wage inflation, it didn't concern price inflation.

Curr: And there's a sort of simple story in which there's this old Phillips Curve that policymakers thought they could exploit. And then they were proved wrong by that hyperinflation of the '70s and '80s, and then we had to move to this new world. But I would say that on the whole, yeah, it's a dominant macroeconomic way of looking at things.

Beckworth: Yeah. So what I see is the go to model in central bankers minds, even though, like you said, [there’s] different versions and it's often ridiculed. It's still widely used, and I want to present a different perspective. And in fact, I'm going to maybe use an analogy. So imagine there's a coin, I'm going to call it the inflation coin. And on one side you've got the Phillips Curve model, which you just described. On the other side, and I'd like to think from this other side, but I think they're just two sides of the same coin.

Beckworth: And then it's kind of a simultaneous story you could tell, but there's the money supply, money demand story on the other side of that coin and that prices or inflation are driven by the imbalance or difference between money supply and the real demand for money. And I think you can think of it along those lines. And that's kind of like an old monetarist story, but I think it's useful.

Curr: Yeah, I was going to say, I think you can reconcile-

Beckworth: Absolutely. Yeah.

Curr: ... those two. Yeah, because what the Friedmanite view of the Phillips Curve would be, well, the important thing about money is that a monetary policy cannot be used to drive real variables in the long run. So therefore if you try to drive unemployment too low, you're only going to end up having an inflationary impact. And that's a story that is sort of rooted I think in that second side of the coin you were talking about because money supply and money demand is ultimately going to generate inflation. And so if you try to do other things, you're going to kind of go, "Hey, why?" I don't know if you'd agree with that.

Beckworth: No, absolutely. And I think it's useful to have all the tools you can to think through these issues. I think not always resorting to the Phillips Curve could be helpful in some circumstances. And I'll give an example, but I want to also be clear, both approaches have challenges. With the Phillips Curve, you've got to know the natural rate of unemployment or what's the output gap?

Beckworth: If we can think more broadly, how do you measure slack? And that's a very hard thing. We don't observe it directly. So it's this unobservable but important part of the story in the Phillips Curve side, on the money supply, money demand side is we don't observe real money demand directly either. So these both have these unobservable that make it tricky. This is where I think the money supply, money demand story becomes important.

Beckworth: And we'll talk more about it when we get to the global factor story. But I look at the safe asset demand story. There's this demand for safe assets. There's a relative shortage of it also called the global savings glut as the demand for liquid safe assets that could be loosely coined as money. So broad money demand, including treasury bills, commercial paper repos, all that stuff.

Beckworth: And there's been this demand, an increased demand for it. And that effectively has caused some of the low inflation around the world. And you only can think about that if you're looking at the money supply, money demand side, it's harder to think about that when you view it from the Phillips Curve side. So just for what it's worth, I think it's useful to have both approaches at your disposal.

Curr: Yes. That's interesting. I think that's right. I suppose in that world you might be looking at a shift in the Phillips Curve caused by the kind of shock you're talking about. I need to think about it.

Beckworth: Yeah, no, I think so. So you would see simultaneous movements, but it's easy for me, at least in my brain to think about, well, there's this demand for safe liquid assets. Effectively it's a broad money demand shock and that has been one of the sources driving down low inflation, not the only one but one of them. And there's been a number of causes. We'll come to this later. Why has there been a demand for safe assets and why does it continue to persist?

Beckworth: Alright, let's move on to the Phillips Curve. The Phillips Curve is the dominant view here and we need to think through the issues surrounding it. And that's what this second article is about. So in the second article it goes on to talk about how the Phillips Curve has failed. And you've mentioned this earlier in the past decade in particular. So this is like there's two stories, there's the long-term story and then there's the short-term story for inflation. And the short-term story is the Phillips Curve has not done a great job both on the period where we should've had deflation and had more inflation. So walk us through those two stories and what does it imply?

Curr: Sure. So throughout the decade, the Phillips Curve has failed. And in the article we have a chart that demonstrates this. It used to be that you could see it, for example, in the data for the United States quite clearly. But what happened after the financial crisis is unemployment rose, but inflation everywhere didn't fall as much as the Phillips Curve models were suggesting it should.

So throughout the decade, the Phillips Curve has failed. And in the article we have a chart that demonstrates this. It used to be that you could see it, for example, in the data for the United States quite clearly. But what happened after the financial crisis is unemployment rose, but inflation everywhere didn't fall as much as the Phillips Curve models were suggesting it should.

Curr: So you had a sort of a lot of head scratching over that. You had the IMF saying, "There was a deflationary dog that didn't bark." You had some scholars saying that “we think the natural rate of unemployment has gone up a lot and there's been a supply side shock. And that explains why we haven't had as much deflation as the models would have predicted.” And then things went along for a while.

Curr: We had this big oil price fall in the middle of the decade, which did cause a deflationary shock. So that deflation, if you like, came a little late. But by that point, your labor markets are recovering. For the last few years in particular, rich world labor markets had been on this extraordinary tear where we've seen employment rates surge to new highs. So for the working age population, two thirds of OECD countries are a record employment high, so a record proportion of the population have jobs.

Curr: And that's an extraordinary labor market boom. In the US it's taken the unemployment rate to the lowest since the late 1960s. That hasn't created the degree of inflation that the Phillips Curve models predicted that it should. So both after the crisis, when labor markets were really weak, and today when they're really strong, the model seems not to have performed, at least in the way that was expected.

So both after the crisis, when labor markets were really weak, and today when they're really strong, the model seems not to have performed, at least in the way that was expected.

Beckworth: Yeah. And there's all these attempts to salvage the Phillips Curve and we'll talk through some of the explanations that you provided in the article. But it is interesting to see people really working hard to save the Phillips Curve. It's there, it's just hidden in the data. Or maybe there's some careful way to interpret it. It's nonlinear. So let's work through those explanations. Let's start with the first one. It's a statistical artifact that we don't see it working the way we thought it would work.

Curr: Yeah. So this idea, which I think is analytically sort of indisputable, really, is that when policymakers try to exploit a relationship it might not appear the same way as it did before. Now, this will probably sound very familiar to a lot of people as the sort of insights of the 1970s, the late 1970s, the Lucas Critique idea is actually subtly different from that because what the Lucas Critique says is that there were some relationships and if you try to exploit them, the relationship breaks down. You can't exploit it at all.

Curr: This statistical artifact idea says that the underlying causal relationship that the model says is there might be there, but once central banks start exploiting that it might just get really hard supporting the data. So it would say for instance, at the start of the decade, inflation got high because there was a commodity shock even when labor markets were weak, but central banks were willing to tolerate that high inflation because the economy was so weak. Whereas if you get that sort of supply side shot now with labor markets really tight, central bankers might get a bit more hawkish.

Curr: So what you get as a result of that, what if central bank preferences are different, in one set of economic circumstances and another while that could create a sort of artificial correlation between inflation and unemployment rather than the reverse correlation than you'd expect. And I think that's probably something to that. Why I think it fails is the why wasn't this also true in the 2000s, that we can see a Phillips Curve in the data in the 2000s. And I also think it ultimately, has a limit that it still doesn't imply that unemployment can fall forever without you getting inflation.

Beckworth: Yeah. Let me run an analogy by you and ask if this is the same point being made. So let's assume the economy is a car. So we’ll use a car analogy here, an automobile. The gas pedal is the stance of monetary policy. So when I hit the gas pedal I’m adjusting maybe the target interest rate or however you want to measure the stance of monetary policy and speed of the vehicle is my inflation rate.

Beckworth: So let's say I'm aiming here in the United States where speed limit of 65 miles an hour. So I want to keep it steady and what happens is I'm driving along the road, I hit some hills and I got to push the gas pedal even farther down. To keep 65 miles an hour because I'm going uphill, I'm fighting gravity. But then on the downside of the hill ease up on the gas and I'm driving along and I encounter all kinds of these hills, maybe these hills represents shocks.

Beckworth: Maybe there's sudden things that happened to the economy, that effect unemployment. So maybe my gas pedal’s responding to the… maybe unemployment changes, maybe other shocks to the economy. But the end product is the speed limit is relatively stable. So I'm going to see no correlation between my gas pedal and the inflation rate. And so I might see maybe a correlation between the gas pedal and the unemployment rate.

Beckworth: And if I try to look just at the surface of those two things, I'm not going to see a relationship there that I thought would be there. Like maybe you might expect when I hit the hills, it affects your speed. But because monetary policy is kind of responding, it’s offsetting all these shocks, you don't see the relationship in the data. Is that a similar story?

Curr: That's a similar story, yes. I'm not sure it's a perfect analogy. Partly because, of course monetary policy isn't hitting its target consistently. It's not staying at the speed limit in this decade. Inflation is sort of moving around, it's just moving around unpredictably. So maybe you need a little bit of noise, you need to driver who can make a mistake or something like that. But yes. Yeah, it's basically the same idea that I think that illustrates it nicely.

Beckworth: Yeah. So maybe instead of driving 65 miles an hour, he's driving 55 miles an hour. So instead of hitting two percent inflation, we’re hitting 1.5 percent inflation, so a fair critique. Okay. So then the first point is that there's an identification problem that we just don't see it in the data, but it actually, it's there. In fact, I went to a Brookings conference few weeks ago and there was someone who got up and made this very point, a very powerful econometric exercise. But at the end of the day, I think there's a limit to using this approach, right? Because policymakers want to use this tool. So is it a tractable, usable tool? So let's go on to the second explanation. Anchored inflation expectations. Tell us about that.

Curr: Yes. So this is the idea and this has come through in pieces of research in Europe, out of the ECB. It's come through in research from the Fed and some other places as well, that once your central bank becomes credible and once people expect maybe not exactly two percent inflation or whatever your target is, but once people expect inflation to be low that that's going to really start interfering with the Phillips Curve.

Curr: Because well, one of the stories that's told is if people just don't believe that the central bank is ever going to let inflation get out of hand, then they don't need to pay so much attention to economic data. And so they're not reading the newspaper and thinking what's inflation going to be on? What pay rise do I need to demand? Or anything like that.

Curr: And some of these mechanisms are thought to in theory lie behind the Phillips Curve. The public's meant to be anticipating what's going to happen to inflation and then the expectations become self-fulfilling. The more anchored are your inflation expectations, the less powerful that sort of amplifying effect of the public anticipating things going wrong is going to be.

The public's meant to be anticipating what's going to happen to inflation and then the expectations become self-fulfilling. The more anchored are your inflation expectations, the less powerful that sort of amplifying effect of the public anticipating things going wrong is going to be.

Curr: And so some people have suggested that that would make the Phillips Curve flatter. And that also makes sense as a story of something changing over time. Because of course, inflation targeting central banks have been getting older during this period. So maybe that inflation expectations have been sort of becoming more anchored as well.

Beckworth: Do you think this is a story of too much of a good thing? Central banks have conquered inflation, they've done such a great job, that they've created a straight jacket where inflation expectations can't budge even when they need to?

Curr: Well, it's kind of a paradox because if… can you be a victim of your own success in this area? I don't know. If you get inflation expectations really anchored, well, what should that mean? It should mean that you could drive the unemployment rate sort of unnaturally low using monetary policy for quite a long way before people wake up and realize what you're doing and everything becomes unanchored again.

Curr: It becomes tempting for you to want to exploit this more favorable trade off that the anchored expectations are getting you, even though those expectations were anchored in the first place by the central bank promising not to do that. I don't really know what I think about that, but it's a bit intriguing.

Beckworth: Yeah. I mean, this issue of credibility, it's hard earned credibility. You don't want to blow it, but you don't want to also be artificially constrained by it as well. But credibility is something that once you lose it, it's hard to gain back. So I completely understand it. I once had a colleague describe the credibility of inflation targeting central banks like a marriage. You're faithful to your spouse, you develop trust over years. If you come home late a few nights, it's not a big deal, but once you're caught in an affair, that credibility is blown. It takes a long time to build it up again. And a very kind of graphic analogy.

Curr: Sure, yeah.

Beckworth: But the point is that that's the kind of credibility you've got to have. Could your spouse trust you again? And if you break that credibility, it's hard to come back in a short meaningful time. It will take a long time to build it back up. And so it's something we want to be careful with, but we also maybe want to think, has it become artificially constrained? And that's why I think level targeting, as we'll talk about later, I think it could be very helpful. But let's move on to the next point. And that is a nonlinear relationship might be behind the breakdown in the Phillips Curve.

Curr: Yeah. So this is the idea that the Phillips Curve’s still there. Just the way it operates is that it's flat for a long time and then suddenly inflation will take off very rapidly. And the people who think this is the case like to point to the late 1960s for example, as a case study of when this happened. And I have some sympathy with this view because ultimately it cannot be the case that you could keep driving down the unemployment rate forever. That you could keep pressing the accelerator forever and not see a response.

Curr: It's very easy to reject that because if that were the case, we'd essentially got endless free resources at our disposal. That central bank can just print a load of money and start buying stuff up for the state or employing everyone. It just doesn't hold, inflation has to take off at some point. I suppose the additional part of the picture in the nonlinear story is that it's going to get very steep and then suddenly we're going to be in a world where it's hard to control because it will happen suddenly.

Beckworth: Yeah. If this weren't true, then The Fed could buy up the whole world. Right?

Curr: Exactly. Yeah.

Beckworth: We know that that can't be true. But the nonlinear relationship, I mean, I think that does make sense. Again, it goes back to the credibility point that you can exploit at some point and it’s going to blow up and lose that credibility. So maybe that's a key part of the story. Maybe the identification problems part of the story.

Beckworth: And I think all those things have some part in the inflation expectations. So the Phillips Curve is struggling, it's still alive. The Fed still uses it. It's still the tool that people invoke, but it has its challenges. So let's move on to the next article in this special report. And this deals with technology in your clever title for it is, “Alexa, how much is it?”

Technological Disinflation and the Amazon Effect

Curr: So this article - I'm glad you like the title - is partly about the Amazon effect. The Amazon effect is this idea that inflation's gone away because retail in particular and because of the rise of Amazon obviously, has become more competitive and everyone's having to slash their prices to compete with Amazon. And we know from various studies that prices online tend to be a little bit lower than prices in brick and mortar stores.

Curr: So this has something to do with what's going on with inflation and the strange performance of inflation. I actually don't find this very convincing for a number of reasons. One is that the retail has been getting more competitive for decades, think of the rise of Big-box retailers in the 1990s and the rise of imports from China in the 2000s, that also drove down the price of goods.

Curr: And we haven't seen much goods inflation for a very long time. So the fact that Amazon is doing it in this decade, it doesn't seem to be particularly important to me, but it is something that people speculate about. But then there is a lot of other technological stuff going on in the economy, which I do think is relevant. And what the article discusses is the idea that there might be these productivity increases around the provision of free services around the provision of search engines, YouTube, WhatsApp and smart phones that aren't captured in economic statistics.

Curr: And this is often presented as, like I said missed productivity or missed growth. But the flip side of that is that you're actually missing some deflation or some disinflation in the figures because these sorts of free products and technological advances are replacing things that people used to pay money for or used to pay more money for. And so that's what the bulk of the article is about.

Beckworth: Yeah. Now this article brings out this point very well and illustrates that through some examples, through apps on your phone that you use to pay for the same service. For example, if I had to buy an encyclopedia set before, now I can look up something on Google or if I pay for a camera, now it's in my phone, or maybe pay for, I don't know, Sony Walkman. Everything's on my phone now. And so the costs of those things have fallen dramatically. And I think there's, there's a point in this argument.

Beckworth: Here's my question. There was some research done by several economists and at the University of Chicago, I think Chad Syverson, maybe some out west as well, where they tried to estimate how big this was. So maybe this missing consumer surplus. And they came out with the conclusion that it really isn't that big, it doesn't explain a lot, but what you covered in your article suggested otherwise. So I guess, is there a consensus on how much this can explain or how big is the missing productivity that actually is there we're just not measuring?

Curr: I don't think there's a strong consensus. I think one of the key issues here is how much things have changed because you might say, "Well, it's always been the case that productivity growth has been somewhat under counted in the statistics." And the studies I've seen, I think I have seen the one to which you refer, but I am not familiar with it enough to speak for it. But the studies I've seen, that are skeptical of this in fact often say, "Well, we think there's been missing technological advances in the inflation statistics for a long time." And so nothing's really new.

Curr: I think [what is] said against that on the other side is this experimental evidence, which I referred to in the article showing that consumers say they place very high value indeed on some of these services. And I think this is a source of active debate as to how significant it’s been, because some people say as well, these services are sort of inherently non-market.

Curr: So if you think of a Tinder or a dating app, for instance, it's never been the case, that many people have paid for dating services in the past. And so some of these things, I think that the phrase that John Fernald at the San Francisco Fed uses is that these are conceptually non-market items. Other people say, "well, regardless of whether or not they are conceptually not non-market, you can just see the consumers place a very high value on them.”

Curr: And therefore even if you say there's nothing wrong with the statistics it's still the case that there's a lot going on that is not to be counted in some sense that is relevant from, I suppose, a welfare perspective. And that's where I think the interesting debate lies.

Beckworth: Yeah. And when I read these articles that came out and said, "Look, there hasn't been that much change, maybe a little bit." In my mind, that debate wasn't settled. And you mentioned some of these experiments [that] Erik Brynjolfsson has been doing from MIT and some of his coauthors and they really were fascinating. I encourage the listeners to read the article, but what would they have to pay someone to give up social media? And there were some pretty high figures in there. So there is value people put on these free services.

Beckworth: So it might be part of the story. I remember Janet Yellen some time ago talking about this Amazon effect, is this the reason inflation is undershooting its target. The question is, is it a persistent enough growth in productivity to be a big part of the story? So you can imagine like a one-time increase which would have a onetime effect, temporary change in inflation, but has there been at a continual growth and productivity that wouldn't have contributed meaningfully to the low inflation we're seeing. So where do you come down on that?

Curr: What I would say is that I don't think that technology can strongly explain the inflation shortfalls we've had and that's part. But I do think that it's increasingly the case that there are large parts of the economy, that aren't well captured by existing economic statistics. So the implication of that is that were we measuring it properly? You might get a different story. And in my opinion, the most likely story is the inflation is in fact lower than the measured statistics say. So the shortfall we're seeing is in fact worse, that would be my view. So it's not an explanation. It actually makes the puzzle, if you like, even worse. Although if your statistics are bad, perhaps it's less of a puzzle why central bankers wouldn't hit their targets.

I don't think that technology can strongly explain the inflation shortfalls we've had and that's part. But I do think that it's increasingly the case that there are large parts of the economy, that aren't well captured by existing economic statistics. So the implication of that is that were we measuring it properly?

Beckworth: Yeah. And there's always been some concern among central bankers that the CPI and other price indices overstate inflation because of this problem, right? There's a substitution bias. New things are constantly being put into the consumer's basket that weren't measured before. So one of the reasons two percent was first introduced… there's multiple reasons. But, one of the reason's two percent was first used as a number for an inflation target is many thought that would kind of capture some of that bias. So maybe we are closer to 0 percent in practice inflation rather than one and a half to two percent.

Curr: Yes, it is a longstanding concern. It was something that Alan Greenspan used to like to raise in front of Congress in the 1990s. I should add that there are people who make precisely the opposite argument. Who say that the hedonic adjustments as they are called in the inflation basket, which account for things getting better have only been introduced lately and do too much. And so actually we're understating inflation. I've seen a couple of people make this arguments.

Beckworth: Really? Interesting.

Curr: Yeah. I do not think the balance of the evidence is in that direction. I come down the other side, but it's certainly shows how this area is a source of rich debate.

Beckworth: Yeah. So lots of work out there for budding young economists. So get out there and do the research. Okay, let's move on into the next article. And this is prices without borders. So this is the global story. What factors from around the world have contributed? And you cite a study in there from some economists at the World Bank, which is really fascinating. And you note that they have this study that looks at a global trend.

Global Factors Influencing Inflation

Beckworth: So you look at all the inflation around the world, there's kind of a common component among all the different paths of inflation we're seeing. So it looks like there's some kind of global story and they find, and correct me if I'm wrong here, that global trend can account for 50 percent of the advanced economies or rich world's inflation and about a third of the poor countries. Is that right?

Curr: Yes, that's right. And in fact, this is in a book about inflation-

Beckworth: A book, okay.

Curr: ... that the World Bank brought out not too long ago and it focused on emerging markets actually, but they do look at those advanced economies as well. And it contains a lot of really good evidence on inflation around the world. I'd recommend it to your listeners.

Beckworth: But that's quite a big number, 50 percent-

Curr: Yes, it is.

Beckworth: ... of inflation in the US, for example. But I know that's on average, this to say it also holds up for the US specifically. But that's a lot of variation explained by this global trend.

Curr: Yes. Now, what's really interesting about this debate is that it's quite clearly polarized between a group of people who say, "The global factors really matter for inflation." And a group of people who say, "It's all the shift towards common monetary policy." So it looks like it's a global phenomenon, but really it's just local trends going in the same direction. Are the central banks behaving in the same way across countries? It's also a debate. The debate being whether or not inflation is a tool globally that has been around a while.

Curr: So it's a lot of people are asking the question in the 2000s, or after China joined the WTO and there were a load of cheap imports into America from China, and people started to think about the global influences on inflation then, and people on the whole, I'd say sort of said this wasn't significant. There was a really technical theoretical paper by Michael Woodford, which showed the in New-Keynesian model, global factors can possibly matter. But suddenly in this decade with the persistent inflation undershoots, it's come back around again and a lot of people asking the question again.

Beckworth: Yeah, that's interesting. So one way to look at this, if we take the view that it's just common monetary policy or similar monetary policies around the world, one maybe more benign interpretation if that is true, is that, man, all these central bankers should become good inflation targeters, they're doing their job, they're staying focused on low inflation, maybe around two percent. And kudos to them, job well done that they're acting independently, but it looks like there's something common to all of them.

Beckworth: And really they're all very good at their job. That maybe is the nice story. A less friendly story, maybe a more sobering story, is the Federal Reserve is effectively setting monetary policy for much of the world. So there's this story about the Fed being a monetary superpower that Helene Rey talks about the global financial cycle, that's more about business cycle, but maybe there's some role here also for the Fed in setting global inflation trends as well. Any discussion on that?

Curr: Yes, I mean, I think that's right. And I think that the integration of global capital markets is one area where the case for inflation being set somewhat globally is strongest in so far as if your capital markets are integrated, then yes, the Fed monetary policy is going to be really important. But also your global interest rates are going to synchronize. The global r-star, as it's called, the interest rate the central banks have to set to hit their inflation targets.

Curr: That's ultimately set in a global market balancing savings and investments. And if r-star gets driven down everywhere and central banks don't adjust enough, then that's, yes, that would be a simultaneous monetary policy error causing inflation to be low. But it would have a fundamentally global cause, which is the so called glut of global saving, that Ben Bernanke first referred to in 2005, causing interest rates to be low everywhere. And that's where I think the global story behind low inflation is strongest.

Beckworth: Yeah, I agree with that. I think there's two different takes here. One is the Fed is setting monetary policy and it's causing all this action in all of these developments to occur, including low interest rates. But I think there's another story, which is the, you called the global saving glut. I'd call it the safe asset shortage problem. I called it earlier a global money demand shock, which I think is independent of the Federal Reserve. It's independent of central banks.

Beckworth: It's due to the things you've talked about in the article and elsewhere, but the demographics, emerging markets not having the capacity to create their own safe assets, policy uncertainty from trade wars, all those things I think independent of what the central bank is doing is driving this demand for safe, liquid, money-like assets and is driving rates down and causing low inflation across the globe as well. But I could see one making the story that the Fed also has some global role to play as well.

Curr: Yeah. I think the Fed's role is particularly important for emerging markets simply because there’s such an impact on their currency from Fed policy moves that it determines their inflation rates and also the strength of their economies to a significant degree. So yeah, I suppose they are separable. The sort of role of the Fed in driving the global financial cycle and the sort of preferences that we're seeing for safe assets as you say, or to save which are driving interest rates down everywhere.

Beckworth: That's a great point. I hadn't really made that clear distinction, just thinking out loud here. But if you look at real interest rates, Kristin Forbes had this paper at the Jackson Hole Symposium this year. She showed kind of the average short- term real interest rate for emerging markets and then for advanced economies, and it's really striking. You see what you think you'd see for advanced economies. It's going down. But for emerging markets it's been relatively stable.

Beckworth: So something very different is going on there. And the story you just mentioned, well, the emerging market, the demographics, all these pressures for safe stores of value are being focused on the advanced economies that can provide them. And thus they're the ones bearing the disinflationary cost of that phenomenon, whereas the Federal Reserve’s global muscles are being flexed on emerging markets, and that's where that phenomenon is being felt. So I'm sure there's some interaction between the two, but that's a nice, I think, way to separate out those two stories.

Curr: Yeah, that's a good point. Yeah.

Beckworth: Okay. What about commodity prices? How consequential do you think those are to global inflation?

Curr: Very, and we've seen that both in history back in the 1970s with the oil price shock sending inflation up everywhere. But we've seen it in the 2010s as well with the tilt towards global deflation in the mid 2010s. I didn't think anyone would deny that commodities prices are very important for inflation in the short-term. But what's really interesting about that I think is that some economists get very annoyed at the idea that relative price shifts just some things getting more expensive or cheaper, can really be important for inflation.

Curr: So people who don't like the idea that cheap imports from China are behind low inflation would point to that and say, "Well, that's just a relative price shock." But no one denies that commodities prices are really important for inflation, at least in the short-term. So it's something that no one really disputes. So you have to then think, "Okay, I'm going to separate my view on inflation as a monetary phenomenon into the short and long-term" unless you really think that central banks are capable of offsetting even commodity shocks in sort of real time.

Beckworth: Yeah. I may be one of those persons you've just described. Let me put it this way, I have a hard time hearing the Federal Reserve and other officials, the ECB as well, talk about the low inflation being the result of bad luck or misfortune or just a spate of bad things happening. And for Amazon, commodity prices, this is like a whole decade. If you'd go back to when the recovery starts in mid 2009 we've had persistently low inflation. I know they didn't have an inflation target until 2012 but still implicitly they were targeting two percent back then as well.

Beckworth: So we have a whole decade of below, on average, two percent inflation and either central banks are incredibly unlucky and have the worst fortune of any institution on this planet, or there's something systematic going on. So in other words, yeah, there might be some supply shocks from commodities, from technologies, but you would hope central banks would systematically respond to such shocks and set the trend growth paths. And maybe not year to year, but at least over the decade that have some control over it. So how would you respond to that critique?

Curr: Yeah, I mean, I'd agree with that. I think the distinction between the short and the long-term is quite important. And clearly it's always the case that a central bank that's missing its target can go into the CPI data, find something that's doing something haywire and blame it on that and say, "Well, this is an idiosyncratic one off shock. And we're not to blame.” And I agree that, that's not a convincing excuse, especially if your miss is persistent. But well, I would say about commodities prices, especially is that, that is very clearly a global phenomenon.

I think the distinction between the short and the long-term is quite important. And clearly it's always the case that a central bank that's missing its target can go into the CPI data, find something that's doing something haywire and blame it on that and say, "Well, this is an idiosyncratic one off shock. And we're not to blame.” And I agree that, that's not a convincing excuse, especially if your miss is persistent. But well, I would say about commodities prices, especially is that, that is very clearly a global phenomenon.

Beckworth: True.

Curr: The evidence is that the inflation becomes more in sync across borders when commodities prices are moving a lot. So that would suggest that there's something a little off with the idea that central banks can perfectly respond to it. Unless they’re all making the same mistake I suppose. I suppose the way I would look at it is that if the whole world was a currency union and you had a commodity shock, it would probably be less consequential and the central bank be more to blame. But sometimes if you're, especially a small open economy, you can get just get swamped by these external shocks in the short term I think.

Beckworth: Sure. That's fair. And I can just hear critics now pushing back against me. “Oh, come on Beckworth, you know the Fed has been limited in what it can do.” And you would hear this remark, I think, that the Fed finds it easier to tighten policy than to ease policy. I've heard that a lot and I think maybe there's some truth in that. That it's easier for the Fed to cause a recession than it is create a boom. But one thing I would say to those critiques is to look at the Federal Reserve as an example.

Beckworth: Their estimate of the natural rate of unemployment has fallen persistently since 2015 when they started raising rates. So in that period they had one view of what was the sustainable speed limit of the economy before it overheated, and it has persistently fallen. And Fed chair Jay Powell admitted as much to Congress recently in one of the hearings he was a part of. And it begs the question, what would a counterfactual world look like where the Fed hadn't increased rates the past four years? Where would we be in terms of inflation today? And I suspect it would have to had some difference, right? Maybe inflation may not be perfectly at two or symmetrically at two still, but I think I'd have some consequence.

Curr: Yeah, I agree with that. I would've expected it to. And the Fed lift off, as it was called, would've happened a lot later if you had applied back then their estimate of the natural rate of unemployment that they have now. There's no doubt about that. And I think it would have led to less of an inflation undershoot. That said, if we're talking about this in the context of commodity price shocks, that happened. The commodity price shock happened before the Fed started raising interest rates and had an effect on American inflation before that happened too.

Beckworth: Yeah, fair point. Okay. We're going to skip the emerging market article. Listeners, go ahead and rate it for the sake of time. We only have a few minutes left, but I do want to get to the last article, which is really great. It has the title *A New Monetarism.* So I'm just curious, why that title? Was it named after the Star Wars: A New Hope? Any reason for that catchy title, *A New Monetarism?*

A New Monetarism Through Central Bank Reform

Curr: Why? Well, in several cases I have heard the idea that the case for nominal GDP levels targeting, which is one of the things that you'll be happy to hear or would have been happy to read me endorse, I'm sure, in this article that the case for that is the intellectual heir to the original ideas of monetarism, the Friedman money growth stuff. That it's the sort of next generation.

Beckworth: Fantastic.

Curr: I was sort of inspired by that idea. I'm not even sure I'm the first person to have called that a new monetarism, but I couldn't tell you any one person who used it first. And I have used it in an article. I recommend other things as well, but I think that's the root of it.

Beckworth: All right, well, I like that. I like to see monetarism once in a while remembered. So thank you for that wonderful title. Let's get into the piece itself. So you suggest three reforms are needed, walk us through those reforms for central banks under the new monetarism.

Curr: Yes. So I say that there are three things going on here that we need to think about. The first is that central banks need to find a way of getting around the zero lower bound. They've got to have a way of fighting recessions that can cope with a world of low interest rates and of inflation and inflation expectations falling without the ability to lower nominal interest rates more to offset that.

Curr: The second issue is that they've got to find a way to navigate a world where the Phillips Curve isn't doing what they thought it was. And the third is that they need to find a way to coordinate better with fiscal policy. So should I talk through each of those?

Beckworth: Yeah, that'd be great.

Curr: Sure. So the interesting thing as you know David, about the zero lower bound, is that this is an issue that people have been aware of for a long time. We basically, have a pretty good theoretical prescription for what you should do to escape a zero lower bound world if you can. And it's all about getting a grip on inflation expectations. So you have to find a way to get inflation expectations to go up in a slump. If inflation expectations go up, your real rate goes down, even if your nominal rates are bounded at zero and everyone basically recognizes that I think.

Curr: And so the central banks, most prominently the Fed who are thinking about ways to change their mandate to deal with this new world in which we live, are considering changes like average inflation targeting, which would let inflation rise a bit higher in a boom. And that would mean that in a slump, if people expected that catch up inflation later on, you would get a bit of stimulus from that. And that would be a good thing. So that's the first thing. I don't know if you want to comment on that.

Beckworth: No, that's great. So basically, an argument for level targeting in some form, an average inflation target is kind of what I consider a weak version of level targeting, but basically some kind of strategy of makeup policy, which I think is true. It's essential. You've seen this study by, I forgot their names, these Fed economists that said we'll be at the zero lower bound 30, 40 percent of the time in the future. So something like makeup policies are essential going forward.

Curr: Yes. I think the issue is that if you keep targeting inflation or move to price level targeting in order to get that makeup, you've solved my first problem, you found a way to boost inflation expectations, but you haven't solved the second, you're still in this Phillips curve world that doesn't seem to make sense. You still might find that inflation is not as malleable as you thought it was because your way of generating inflation is sort of premised on the idea that you're going to look at the unemployment rate and try to steer that as your sort of intermediate goal for ultimately getting to inflation.

Curr: And it's still the case that, to the extent that the global effects are important, commodity shocks or whatever, that drive inflation around in the short-term and might cause you to miss your target in the short-term, it’s still the case that you'll have to decide whether or not inflation's going up because of those factors. The supply side factors or because of the labor markets and the Phillips Curve and that looks pretty hard.

Curr: So that's what I say really makes the case for switching as well as from inflation targeting today to levels that also makes the case for switching from inflation to nominal GDP. That nominal GDP contains the sort of implicit test for whether or not you're experiencing sort of overheating economy, which is that if inflation and growth go up together, your nominal GDP is going to go up. And so that will call for tighter policy. But if you're faced with higher oil prices or something that constrains growth, it will put inflation up, but it's not going to cause growth or rise. So the nominal GDP target sort of patches that for you.

Beckworth: Yeah, no, I love it. Great points. And going back to our inflation coin we were talking about earlier, I'd look at nominal GDP level targeting that is taken seriously both sides of that coin. So nominal GDP targeting can also be viewed as balancing money against velocity. How often is it money demand versus money supply or you can take the other side, the real side of the economy and the price level. So which ties into the Phillips Curve.

Beckworth: So to me it's a nice way to get past some of the challenges of doing monetary policy when you don't know in real time some of these matters. Like you mentioned, it's a simple metric. There's some other issues surrounding it to be fair, there's measurement issues and data revision issues. But with that said, it's a simple approach that handles these challenges that we're now seeing. So level targeting number one, nominal GDP targeting number two. And what's the final piece of the puzzle?

Curr: So the final piece of the puzzle was fiscal policy. Because even if your central bank does all of that, it's still the case that you're relying on basically kind of a confidence trick. I mean, that makes it sound more devious than it is, but you're relying on expectations to get you out of a trap. How do you hit your nominal GDP levels targeting? Even if we do everything that we want them to do, central banks still face the zero lower bound. They still might find inflation expectations misbehave and they might find themselves stuck.

Curr: So I'd say that fiscal policy does need to come in at this point, and we need a way of ensuring that it does. We need a way of overcoming the political economy problems around governments managing the business cycle, which were the reason for saying in the first place that central banks should be the ones doing business cycle management.

Curr: I'd say, well, you could shop in the automatic stabilizers a bit. So that, for instance, there are automatic tax cuts or larger unemployment benefits when recession comes. That can only help you to do that, but it's probably not going to be enough. And it's not going to be well calibrated when a recession comes just to have bigger automatic fiscal stabilizers.

Curr: Ultimately, I'd say central banks are going to need to have some sorts of fiscal mechanism or fiscal monetary hybrid where they can create a fiscal stimulus of their own at their discretion in order to hit their target. And there are various ways you might do that. But one of them would be helicopter money so that the central bank could create new money to fund tax cuts, but they ought to fund handouts to the public and determine the size of those handouts, for example.

Ultimately, I'd say central banks are going to need to have some sorts of fiscal mechanism or fiscal monetary hybrid where they can create a fiscal stimulus of their own at their discretion in order to hit their target. And there are various ways you might do that.

Beckworth: No. Yeah, that's a great idea. Listeners will know, we've talked about some of those recently, a nominal GDP level target backstopped by a helicopter drop or tied to a helicopter drop when you hit the zero lower bound or liquidity trap. So I think those are all great suggestions. My hope is that we see them come to fruition. I'm not sure that we will. My fear is we'll see average inflation targeting and that's it. And average inflation targeting will be in effect, not much different than what we have today as opposed to being a real makeup policy. That's my fear going forward.

Curr: Yeah, I think that's probably right. I mean, I say in the article that these are quite deep rooted reforms that are quite radical. And I agree with you. We're much more likely to see average inflation targeting. And it would probably take a few more sort of policy failures and things going wrong before people come around to the view that more radical reform is needed. But ultimately if you're going to do business cycle management in this world of low rates and low inflation you have to come up with solutions like these eventually, I think.

And it would probably take a few more sort of policy failures and things going wrong before people come around to the view that more radical reform is needed. But ultimately if you're going to do business cycle management in this world of low rates and low inflation you have to come up with solutions like these eventually.

Beckworth: Yeah, and that's why we have these conversations, Henry, to hasten that change to make the world a better place and hopefully improve understanding on the margin that pushes close to that outcome. With that, our time is up. Our guest today has been Henry Curr. Henry, thank you so much for coming on the show.

Curr: It's been a pleasure. Thank you for having me.

Beckworth: Macro Musings is produced by the Mercatus Center at George Mason University. If you haven't already, please subscribe via iTunes or your favorite podcast app. And while you're there, please consider rating us and leaving a review. This helps other thoughtful people like you find the podcast. Thanks for listening.

Photo by Joshua Roberts/Getty Images

About Macro Musings

Hosted by Senior Research Fellow David Beckworth, the Macro Musings podcast pulls back the curtain on the important macroeconomic issues of the past, present, and future.