James Bullard Discusses Nominal GDP Targeting

This 15-minute podcast was released April 19, 2019.

How does nominal GDP targeting differ from inflation targeting? And what would be some of the advantages and disadvantages of using nominal GDP targeting? St. Louis Fed President James Bullard says: “The biggest advantage is this idea that you would really cement inflation expectations around the target. This would give investors and financial market participants, households, businesses, the confidence that the central bank really was going to deliver on what it said.”

Transcript

Karen Mracek: Welcome to the Timely Topics podcast series from the St. Louis Fed. I’m Karen Mracek, your host for this podcast. With me today is St. Louis Fed President Jim Bullard, and we will be discussing nominal GDP targeting. President Bullard, thank you for being here.

James Bullard: Thanks for having me.

Mracek: Before we jump into specifics of nominal GDP targeting, can you tell us a bit about how the Fed maintains price stability now, and why we are talking today about a different approach?

Bullard: My view of this is that, beginning in the 1990s, there was a coming together of the academic literature and practitioners in central banking around the concept of inflation targeting, which, roughly speaking, meant that central banks named an inflation target and conducted monetary policy in such a way as to hit that inflation target over the medium term. And this is commonplace today, but at the time was a big shift compared to the ’70s and ’80s where there weren’t any inflation targets, and it wasn’t at all clear, relatively speaking, what the various central banks were doing.

I have to say, in the big picture, that inflation targeting has been crazy successful. Inflation has been much lower and much closer to these inflation targets across the countries that adopted them. Inflation’s been much less variable than it was in the ’70s and ’80s. The inflation expectations in these various countries have become much less volatile and much more clustered around the inflation targets. So, inflation targeting has been a great success story, and the question now would be, can you improve on that?

Mracek: Can you explain, then, what nominal GDP targeting is, and how is it different from inflation targeting?

Bullard: In inflation targeting, we would name the inflation target, but if we missed the inflation target, either on the high side or the low side, you wouldn’t be too worried about it. You’d say, “Well, OK, we’ll try to hit it again next year or over the medium term, and we won’t worry about the fact that we’ve missed it this year or maybe several years in a row.”

With price-level targeting or its close cousin nominal income targeting, you would worry about past misses, and you would try to make up for past misses in such a way that you would stay on a path for the price level or a path for nominal GDP, depending on which route you went. But they’re closely related.

So the main difference is that private-sector investors would understand that you were going to make up for past losses, and they would understand that if you missed to the low side in the past, this would mean that future policy would likely miss to the high side for a little while. And, vice versa, if you missed on the high side for a while, then you would probably miss on the low side in the future for a little while.

And, in so doing, at least in theory, this would further cement inflation expectations—even more than they’ve already been controlled by the inflation targeting regime that’s been in place the last 25 years. You would further pin down inflation expectations, and therefore, get even better monetary policy than what we’ve had over that period of time.

Mracek: To clarify, when you talk about nominal GDP targeting, are you referring to targeting the level of GDP or the growth rate?

Bullard: At least in the simplest examples, it would be the level of nominal GDP. You’d have a projected path for that level, and you would be right on that path all the time, every year. I mean, practically speaking, that’s not what would happen, but ideally that’s what would happen.

Mracek: How would nominal GDP targeting work in theory?

Bullard: I do have this paper called “Optimal Monetary Policy Masses,” which is only one out of a sea of papers on this issue. In that paper, there are big private credit markets as there are in the real world. But in the model, those credit markets don’t quite work the way they should because all the trading takes place in nominal terms and it is non-state contingent.

And because of that, the monetary policy can follow this nominal GDP targeting rule, and that fixes the problem that would otherwise exist in credit markets, and therefore gets a more smoothly operating credit market than you would otherwise have.

So, this would be an improvement to the equilibrium of the economy and everyone in the economy would be much happier than if you didn’t do this. And so that provides one rationale for nominal GDP targeting in a theoretical setting.

That’s just one possible way you could talk about this issue inside of a model, but I think it’s an interesting one.

Mracek: Can you give us an example of when the monetary policy response under nominal GDP targeting might be different from inflation targeting?

Bullard: Yeah, even right now, when the Fed looks back on the inflation outcomes since 2012, we’ve generally missed to the low side of our inflation target. And, as I was saying earlier, under inflation targeting, that’s just treated as a bygone, and you don’t worry about that, which means that the way the Fed is treating it right now is that, yes, we’ve missed in the past, but we’re just going to promise to hit the target in the future, and then that will be that.

But under nominal GDP targeting, you wouldn’t be satisfied with that. You would say, “I want to make up for the fact that I missed to the low side over several years,” and therefore it’d be OK if you allowed inflation to be higher for several years in order to have it average out over a longer period of time to hit your 2% inflation target.

In some ways, at a very simple level, there’s just more commitment to the idea that you’re actually going to hit the 2% target instead of saying that, “Well, we’ll try to hit it, but if we miss it, then there’s no guarantee.”

This nominal income targeting idea would be a way to make sure that you hit the target over longer periods of time.

Mracek: What would be the advantages of using nominal GDP targeting?

Bullard: The biggest advantage is this idea that you would really cement inflation expectations around the target. This would give investors and financial market participants, households, businesses, the confidence that the central bank really was going to deliver on what it said. It was going to deliver this 2% inflation rate. They could use that in their planning, and they could be reasonably confident that that was going to be the actual outcome in the economy over longer periods of time. This would help with getting the best allocation of real resources that we can get. So, that would be the principal advantage.

I think the question about both nominal GDP targeting and price-level targeting is whether the additional gains that you would get are going to be that big compared to what you already are getting from inflation targeting. I started this off by saying inflation targeting has been crazy successful, and I think that’s the right view of the last 25 years in the countries that have adopted inflation targeting. I said it’s been a way better experience than the ’70s and ’80s where you had inflation expectations bouncing all over the place and a lot more volatility. You got both nominal volatility, but also more recessions and more frequent recessions.

Most of that has smoothed out, the financial crisis notwithstanding. We’ve managed to have more moderation in real variables and nominal variables in the inflation targeting era, even though we had the big financial crisis. But that’s only one part of the last 25 years.

Mracek: And are there any disadvantages to nominal GDP targeting?

Bullard: Some people say it hasn’t been tried, and it would be hard to communicate. And I think one way to convey that idea is that it really relies on private-sector expectations understanding the policy, and because they understand the policy, they expect inflation to be right around 2%. And because of that, you get good things to happen in the economy.

It’s the kind of thing where you might say, “OK, we switched to nominal GDP targeting.” Nobody notices in the entire economy. No one pays any attention whatsoever, and you don’t get any of these effects at all. I think that would be the kind of thing that is very practical and could possibly happen, because private-sector people might say, “Well, I don’t understand it,” or, “I don’t see what the difference is between this and inflation targeting. It’s all too subtle.”

The theories are relying on these things being really tight and really affecting these expectations a lot. But the reality might be a lot more distant than that.

Mracek: One common criticism is that stagflation is consistent with nominal GDP targeting. What is your thought on that?

Bullard: Stagflation is a term from the ’70s when both unemployment and inflation were high—at one point both in double-digit ranges in the early 1980s—and this came to be called stagflation. It was considered a bad outcome.

There is an aspect of this with price-level targeting and nominal GDP targeting because good times are the times when inflation would run below target, and bad times are the times when inflation would run above target. And that’s different than what we’re used to.

But when you think about it, bad times are times when you want the somewhat higher inflation because that’s supposed to drive the real interest rate lower, and that is what is supposed to get you out of the recession and help you during recessionary times. So it does have a certain logic to it, even though we had this sour experience from the ’70s.

I don’t think anybody’s talking about that kind of inflation or that kind of volatility that we had at that time. That’s not something you’d want. You’d want sort of mildly higher inflation in bad times, and mildly softer inflation during good times. So I think it’s more contained than anything that happened during the ’70s, which was quite volatile.

Mracek: I think you touched on this, but have any central banks used nominal GDP targeting?

Bullard: No, not to my knowledge. And one of the criticisms is that we barely got everybody converted over to inflation targeting, and now you’d be switching again. The U.S., in particular, only named its inflation target in 2012. Japan’s another country that didn’t come around to inflation targeting until relatively recently.

And so it’s not clear that you want to, then, make another change. Although I would say, if you’re going to make a change, you should make it during good times, not try to improvise when it’s a very volatile area or high recession risk or something like that. If you wanted to make the change during calm, successful times for the economy, that’s probably the time to do it.

I do think that one of the advantages of inflation targeting was that smaller countries took it on first and experimented with it, showed how it could be done, and then you had, eventually, the founding of the ECB [European Central Bank] coming on in the late ’90s, and then the Fed following through later. And so you kind of tested it out in other places before you had the world’s leading economies adopting it.

If the U.S. moved first and went with price-level targeting or nominal GDP targeting, that would be a different kettle of fish, and that would be something that has to be thought about carefully, I think. You’d be setting a trend in a global environment, and other countries would likely follow.

Mracek: That make sense. Would the average consumer notice a difference if a central bank switched from inflation targeting to nominal GDP targeting?

Bullard: I don’t really think so. I think the person on the street would regard everything as being about the same. But again, this is really more about cementing expectations in financial markets and the way bond markets, in particular, would operate under this kind of a regime as opposed to the one we’re using now. And so I don’t think a typical person would see much difference.

Mracek: Theoretically, if the Fed were to adopt nominal GDP targeting, would it capture both sides of the dual mandate—that is, stable prices and maximum sustainable employment?

Bullard: It would, because—at least in theory, if this was the true optimal policy—this would mean you’d get the best allocation of resources, which would include labor resources so that labor would be supplied exactly at the rate that households would desire. And this would be the best employment outcomes that you could get for the economy. So it’s all supposed to work at least on paper. Whether that actually works in the real world is certainly an area that people like to debate.

Mracek: Got it. Is there anything else you want our listeners to know about this topic or how the Fed thinks about maintaining price stability?

Bullard: Well, I think it’s under review in 2019, and it’s something the Fed will consider, I think, as part of its review. I think it’s good to be thinking about possible innovation in monetary policy frameworks as time goes on. Surely, the framework we’re using today will not be the same one we’re using 50 years from now. And, in order to be able to evolve at the right moments, we have to have regular reviews and think about these issues. And I think that’s the purpose of this framework review in 2019.

Mracek: Great. Well, President Bullard, thank you for joining us today and sharing your views on nominal GDP targeting.

For more of his remarks on this topic and others, you can visit our website, stlouisfed.org, and click on “From the President.” To listen to more of our podcasts, go to Timely Topics.