James Bullard on Life as a Fed Bank President and Monetary Policy in 2017

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: Jim, welcome to the show.

James Bullard: Great to be here. Thanks for having me.

Beckworth: Oh, it's a real treat to have you on the show. I like to get started by asking how did you get into macroeconomics?

Bullard: Well, I had an undergraduate major in quantitative methods and information systems, which was computer science plus operations research, plus business. And then I added another major, which was economics. And the basic plan was to somehow do computer science and economics, which I think is roughly what has happened, although not exactly. And then later went to graduate school at Indiana University and then I was off and running. I just think it's totally fascinating to study the macroeconomy, and I think that has boiled me through my career.

Beckworth: All right. So you went from Indiana to the St. Louis Fed and there you served for a number of years until you became president in 2008. Is that correct?

Bullard: That's right. So I was on the research staff for about 18 years here. It's like an academic job you're expected to publish, you're expected to contribute to the monetary policy debate. So that's what I did during those years.

Beckworth: Well, tell our listeners, what is it like to be a president of a regional federal reserve Bank? What's the day in the life of a president?

Bullard: Well, compared to the research world, it's very different. In the research world, you're very busy if you're an active researcher because you have many projects and they have deadlines and there're conferences coming up and everything takes a long time because you're dealing with difficult ideas. But when you become president, then it's kind of a, it's a different world. You have, first of all, a lot of people reporting to you. I went from eight people reporting to me to 1000.

Beckworth: Wow.

Bullard: So the scale difference is very different. So, I think, you have to schedule in order to get things done. I think that's something you don't have to do at probably lower levels in the organization where you can just walk over to your neighbor and ask them, what's going on on this project or that project. So, I think that took some getting used to.

Bullard: I'm also an executive within the Federal Reserve System, so it's a federal system, so, there are committees, there are other ways that the banks interact and the Board of Governors, and so there's a lot of aspect there that has changed over the years, and I've adjusted to that as well.

Beckworth: So, how do you find time balancing between preparing for these FOMC meetings, I know you do a lot of speeches in your region. And I noticed you also still do research. So you mentioned scheduling but you must burn the midnight oil to get all these things done.

Bullard: Well, scheduling is an issue and I would say about a third of the time I'm doing things that are directly related to being the CEO of the bank. And about a third of the time, I'm doing things as a system executive. And the other third of the time, I'm the face of the bank, doing public events. And also, my research time has to be carved out of that. But I also have great coauthors and great network of friends within macroeconomics, so that helps me a lot.

Beckworth: Yeah. So, one of the other challenges that I think would be hard for me if I were in that position is being able to be careful about what I say because you guys discuss information that can change markets. I think back to 2006 when then Fed chair Ben Bernanke talked to the CNBC anchor, Maria Bartiromo about interest rates over dinner. The next day she reported, it kind of rattled the market. So, is that a challenge for you? Was that easy to make that transition, you'd be very mindful what you say around whom and in certain contexts?

Bullard: Well, we adapted a policy that I, I'm always on the record because people could record you in any situation, or report what you say in any situation. And so, I try to say the same thing, whether I'm talking to someone one on one or whether I'm talking to a group or whether I'm giving a speech, the theme's are always the same and they're also basically my thoughts on the macroeconomy, my thoughts on monetary policy. So I'm not trying to describe what the chair's going to do or what the committee is going to do. I'm just trying to tell you, here's how I see things right now.

Bullard: One thing I like about that is it produces a record for my own news about how my own views have evolved over time. And then I have a track record about what I was saying at different junctures.

Beckworth: Yeah. I bring this up now because as you obviously know, yesterday, the president of The Richmond Fed resigned over some information that was shared through him, and it was almost an inadvertent admission. He confirmed some information. I guess I feel some sympathy in the fact that I could see myself easily doing that, and it would be tough to always be mindful of what you're saying.

Beckworth: Let's move on to some economic questions now. I want to look at approaches to monetary policy, in particular here, and look at the Fed's 2% inflation target. And before I get into the specifics of that, I do want to ask more basic question. What drives inflation? There's been several wall street journal articles, one recently, one about a year or two ago, where they come out and they say, hey, there's this internal debate going on inside with the Federal Reserve. Maybe they over generalize about what's really driving inflation, inflation has been low, persistently low. And there's just some questions. They quote Janet Yellen saying that she's unsure, and I'm not sure how much of this take is truth.

Beckworth: But my question would be to you then, what do you see as the key determinants driving inflation and what role does monetary policy having in affecting them?

Key Determinants of Inflation and Role of Monetary Policy

Bullard: Well, my reading of the modern monetary theory literature is that it does come to a conclusion on this, which is, roughly speaking, today's inflation is mostly driven by inflation expectations in the future. And that is what has made monetary economics such a fascinating area over the last 25 years. The literature has tended to de-emphasize in various ways the Phillips Curve kind of story where the economy is too strong or too weak and this is driving inflation. That story's been badly strained I think in the last few years. And if you look at the empirical evidence on it, it's pretty weak.

Bullard: So, the literature has an answer for us, but it's kind of a disconcerting answer to say that inflation depends only on unexpected inflation or mostly on expected inflation because then that comes back to haunt you, what does that mean for how you should actually implement monetary policy day to day and how you should we react to the various pieces of data that we see describing the economy?

Bullard: So I think, actually, of the questions we're going to talk about today, I think one of the very key issue in all of macroeconomics is what drives inflation, and how is that, how's our perception of the theory changing how we do monetary policy, both today and going forward?

Beckworth: Recently, there was a paper presented at the US Monetary Policy Forum by Cecchetti and some coauthors. And this paper was also highlighted by the Wall Street Journal, and The Wall Street Journal titled their article, Everything the Market Thinks They Know About Inflation Might be Wrong. And they go on to talk about this article and I went and read it myself. The article looked at the ADF idea from 1984 to the present, so starting the great moderation period up through the great recession and through today. And they find that things Like the output gap, standard measures of money, and even the inflation expectations aren't highly correlated with actual inflation. Did you get a chance to look at that paper or did you hear about it?

Bullard: Oh, yeah. I definitely looked at it. It's well done I thought. And I think it's typical, if you don't believe that paper, you can go look at many others that have tried to do the empirics of inflation. I can think of one by stock in Watson and Jackson Hole a few years ago. And the answer that you get from various angles is always the same, that inflation doesn't seem to be related to the variables that we think it should be related to.

Bullard: But if you read the modern theory, you get an explanation for this, which is that if the monetary policy is run fairly well, and you have an inflation target, then what will happen is that you'll adjust your policy instrument appropriately as shocks come in, and then expected inflation will just stay at the inflation target, and actual inflation will just stay at the inflation target. So then when you go to run your regression on what determines inflation, it looks like it's not related to anything else in the economy.

Bullard: So this is a bit of a paradox about empirical work, which is that under well run inflation targeting policy, then inflation never really moves that much, and therefore you can't see the statistical correlations in the data that you might otherwise expect to see.

Bullard: So I think this is very challenging and something that has not really been fully absorbed by either the academic community or the business community, but it does show up in this empirical work, and especially in this paper by Cecchetti et al.

Beckworth: Yes. And I agree with your conclusion, your point too. It's a sign of success, the fact that inflation expectations are not correlated with, it's a sign that the Fed's actually doing a very good job in anchoring and keeping inflation bounded. I use this analogy because this article came up and the casual observer would say, see, the Fed really has no influence. Really, it speaks to the opposite.

Beckworth: In fact, Cecchetti, they have a blog, and they made this point too, it actually speaks to the success. But I think of it this way as an analogy. If the Federal Reserve is driving the economy, so the economy is a car, the Federal Reserve's a driver, and let's say that the inflation target is a speed limit. As the Fed's driving that car, it's going to hit some hills, some valleys and it's going to adjust the gas pedal. So the pedal might be the instrument, it's lever for monetary policy. And it's going to hit number of hills, a number of shocks you can think of, and it'll go up and down them. But the speed is going to remain constant, inflation is going to be relatively stable.

Beckworth: So if I'd looked at the relationship between movements in that gas pedal, and the speed, I'd find none, but yet it's the Fed actively using it to get that stable speed.

Bullard: I think that's that type of analogy is exactly right. And so, it shows that you got to think in general equilibrium terms to get this right about what's driving inflation.

Beckworth: Let me ask you about another theory of inflation, and this is a fiscal theory of the price level. There's been discussion on that as well over the past few years since the crisis. What are your thoughts on the fiscal theory of the price level? Does it have any value and explaining the path of inflation?

Bullard: Well, I'm open to it, but whenever I've seen this discussed in academic settings and I've seen many papers on it, it really leads to a lot of confusion about what is being said. And I think the fiscal theory really needs something more of a smoking gun. A key empirical relationship or a key theoretical finding that would more tellingly suggest what it is about fiscal policy and how it works that would be related to inflation.

Bullard: So I think what do you often get in these kinds of discussions is that maybe there's multiple equilibria, a lot of things could happen. So, therefore, maybe the inflation path that we see is being driven by fiscal policy. I don't think that's enough all by itself. You do have the very high inflation economies, and in those cases, they're almost always associated with fiscal irresponsibility of some sort. Even today, you've got places like Venezuela or Zimbabwe or other, Argentina, other places where they have tried to use monetary finance as part of their regime, and they ended up with a lot of inflation.

Bullard: So in that sense, there's certainly, it's certainly long been known that there's a tight connection between fiscal and monetary policy. But for countries that don't have those kinds of issues, the thought has been that you could isolate monetary policy separately and you didn't have to worry about the fiscal backdrop.

Bullard: So, I think the fiscal theory group still has a ways to go, it's an interesting theoretical possibility, but it remains only that right now.

Beckworth: Okay, let's move on to the Fed's inflation target, it's 2%. It was introduced in 2012. But there's been several studies that have shown that implicitly, the Fed was following something similar to that long before. I've seen studies early to mid 90s, the Fed was implicitly aiming for around 2%.

Fed’s 2% Inflation Target

Beckworth: I got a question, a clarification question. I had this exchange actually with two of your economists, David Andolfatto, Steve Williamson, just this past week. I made the contention that the Fed's preferred measure of inflation was the core PCE deflator. And they pointed out in writings and that could put me to some documents that it's actually headline inflation. I get the sense from from journalists, I get the sense from some of Janet Yellen speeches and FOMC statements that the FOMC is more concerned about the core measure. And I know you're not enthusiastic, you have this talk rotten to the core. What do you think the Fed actually is is going after more, is it the core measure or is it the headline PCE measure?

Bullard: The official target is the headline PCE inflation rate. There's good reason for that. We have our constituents here in the United States and they have to actually pay all these prices, including prices for gasoline or home heating oil or food. And therefore, it's not really right to go back to your constituents and say, well, we stabilized prices, but just not the ones that you have to pay the most often or have to face the most often. So, it is the headline that we're most concerned about, that is the goal and really the only goal.

Bullard: The only reason that you would go to some other measure is to try to extract signal from noise. And that I think would be reasonable, but the core measure is such an ad hoc way to do that, that I think we should have abandoned it long ago. And you should get a little bit more sophisticated if you want to extract signal from noise. One way to do that is the Dallas Fed Trimmed Mean measure, and there are other ways that you could go about this. There are all of kinds of paradoxes about just throwing out the food and energy.

Bullard: One thing that's kind of comical is one of my colleagues a few years ago looked at food, and over some time period, food was actually the least volatile of all the components. So, you're throwing it out because it was volatile in the 70s but it wasn't always volatile over every time period. We have a lot of sophisticated time series, type techniques that we can apply to this. I think the baseline should be the Dallas Fed Trimmed Mean, it's about 1.9% year over year.

Bullard: So, on that kind of a basis, the signal would be that you're pretty close to target inflation, headlines just moved a little bit over 2% now on a year over year basis. So, pretty close to target on inflation.

Beckworth: So just to be clear, my impression is wrong then. People aren't as concerned about core PCE inflation?

Bullard: They are, and I'm telling you, they shouldn't be. I think this has become embedded because people built models that worked on core and they would have to recalibrate and re-estimate and everything. And they're not too keen to do that.

Bullard: Another problem with core is that ironically, throwing out food and energy doesn't really throw out all the effects of changes in energy prices. So, core still reacted to the big oil price decline in 2014, even though supposedly you're throwing out the energy prices, but there are relative components that aren't directly energy that don't get thrown out that are affected by the energy price movements. And therefore, core actually sunk a little bit, maybe a couple of tenths on a year over year basis just because the energy price move was so big. And now it's come back some because energy prices have stabilized.

Bullard: So, one of the ironies is that it it's not really as smooth as you would think based on just kind of the naive idea that, oh, I'm just going to throw out some of these prices.

Beckworth: If we look back at the PCE measure, PCE deflator measure, it has been persistently low. Now I know you mentioned that the headline is finally kind of, like it's hit 2%. I mean, it has, it's gone over 2% a little bit, but if you look at the average since mid 2009, we finally got the bottom, the great recession. The PCE measure has been on average below 2%. And this has puzzled a lot of observers, maybe even some Fed officials. What is your explanation for why inflation for a while at least seemed to be below its target?

Bullard: It has run very low. And like I said, I think the best model of inflation involves mostly inflation expectations. My feeling on this was that we came out of the crisis, the inflation rate fell and then it bounced back, which is exactly what you would think would happen after the shock. And everything was fine, but then it started to fall again. And then in 2010, in the middle of 2010, the year over year inflation measures, all of them were quite low. 1% year over year basis or even lower.

Bullard: And at that point, we did the QE2 program in the second, we talked about it in the second half of 2010, and then implemented starting in November of 2010. And if you look at what happened to inflation after that, I mean, QE2 worked like a charm. The inflation expectations went up and actual inflation followed right behind. And by the time you got to the end of 2011, beginning of 2012, PCE inflation year over year, both core and headline were right on target.

Bullard: So I thought at that point that this whole issue was over and our work here is done. But as it turned out, after that, inflation started to drift down again, and since then, has really only come back now. So, you've had quite a quite a few years here with inflation below target. I think that's definitely been a challenge of US monetary policy. You do have a global environment that has very low inflation. Maybe that has something to do with that. You've got period where you had zero interest rates for a long time. You could have Fisherian effects that that might have been taking hold.

Bullard: I think it's really been a challenge to explain the low inflation during this period. For people that just think low nominal interest rates are going to lead to high inflation, boy, we sure didn't get that over the last eight years.

Beckworth: True.

Bullard: And the same in Japan. For years and years, we've had low nominal interest rates. Supposedly, that's supposed to lead to a lot of inflation, it didn't happen. So people have to be reexamining their theories about how they think these things work.

Beckworth: So another part of the conversation related to this particular topic is the nature of the Fed's inflation target. And recently, the 2016, the Fed had a new statement on its long run goals. It came out in early part of the year and it added the term symmetric to its discussion of its target. And you yourself have said, I quote here, "The target should be definitive from above and below." So you were sympathetic to that idea but you actually voted against that statement for the reason, and correct me if I'm wrong here, but you weren't happy with that document because it was suggesting some overshoot might be appropriate to make up for past misses. Was that your objection in that document?

Beckworth: And the reason I bring this up is because a lot of critics have said why isn't the Fed's target more symmetric. It seems like 2% is more of a ceiling. But yet, if I understood you correctly, you were reluctant to want to try to purposely overshoot and make up for past misses. Is that fair?

Bullard: I did dissent on that the first time around. I did make a statement on it at some point. But I guess I want to just stress, I've never really been against symmetry. I think symmetry is good. I think it's the proper way to think about it. All the models I work with, you would have shocks that would sometimes leave you with inflation above target and sometimes below and that would be fine. So I think the idea of putting symmetry in there was a good idea. I'm not objecting to that.

Beckworth: Okay. You have also noted in your research, and we can segue into that, the benefits of level target, what's the price level, target nominal GDP level target. One of the things of having a very flexible inflation target would do, it would approximate something like more of a level of path target. I want to talk about that. What are your thoughts on level targeting in general, then maybe we can move into the work you've done on nominal GDP level targeting?

Advantages of Level Targeting

Bullard: I like the idea of price level targeting and nominal GDP targeting. I've come around to it a little bit more than I would have been several years ago partly based on my own work. I also like the idea that in the New Keynesian model, it's really price level targeting that's the optimal solution.

Bullard: And what I like about that is that it's a very simple thing just to check if you're on the path or not. And the only trick about doing the path is when you're going to start the path. I think the right place to start the path for the US economy is about 1995 because it took a long time for the Volcker disinflation to occur, and it took a long time to actually get inflation down to 2%, which actually happened in 1995 or so. And then if you take that as kind of being the date at which the Volcker shock was finally rung out of the system, then from there, you would expect a 2% price level path to be the one that we would shoot for.

Bullard: I gave talks a few years ago that showed that we were actually right on that path. So, from the New Keynesian perspective, with that very simple metric, you could say that US monetary policy was close to optimal. Since then, I think probably the day after I gave those talks, I think we started drifting off that path. I've been thinking about, at some point, revisiting that and looking at where we are because I think we're below now because inflation was below target for so many years.

Bullard: So, I do think, I do have some sympathy for this. I think that price level targeting is probably optimal in a wider range of models than is currently appreciated by the profession.

Beckworth: So, one of the implications of a price level target is that you would have to tolerate some overshoots, some undershoots from the 2% rate or from time to time if there's certain shocks. On average, you would still maintain 2% growth path. One of the questions I have though, is that something that can be tolerated politically? For example, if the Fed had been doing something like a price level target prior to the crisis, and then there was this period of below 2% inflation, and the next year or two maybe we'd have had 3%, 4% inflation. As a Fed official, do you see that as even doable? I mean, politically with blowback from Congress, from your constituents, be able to see, to appreciate what you were doing?

Bullard: I don't think we're there right now to be able to do that kind of thing to implement a price level target because too many questions are still out there about what is price level targeting, why would you do that? At the grassroots level, I think people just see the higher prices and they say you're not doing your job and I don't want to pay these higher prices. So I think it's difficult to come back with a highfalutin theory and say, well, this is really the best think.

Bullard: So, I think for those communication reasons, it's been better to just have the inflation target and say we're trying to hit the inflation target, and that's that. It would be hard since we only recently officially put down an inflation target, the thinking has been that it would be hard to all of a sudden drop that and go to an inflation target. Also, I think that there's some people that kind of still think there's a permanent trade off between inflation and unemployment or output. And therefore, by accepting higher inflation get more output and things like that.

Bullard: And so, I think it would get convoluted with the group that wants to do that kind of thing, and it would be seen as a foot in the door for those that just want higher inflation because they think there's some kind of permanent trade off.

Beckworth: Well, on the marketing front, I think this is one appeal of a nominal GDP level target. You can say, look, we're trying to stabilize your dollar income, your average dollar income growth. Take the focus off of prices, instead focused on your nominal, you wouldn't want to use the term nominal, but you'd say, focus on your dollar income. But it's interesting you mentioned the whole communication challenge in implementing something like level target.

Beckworth: Ben Bernanke in his book mentioned that, was discussed, nominal GDP level targeting was discussed in I believe 2011. You don't want to change horses midstream, it would just be really tough to get on that. With that said, there's been a number of other regional bank presidents, yourself included, but John Williams from the San Francisco Fed, Robert Kaplan, Charles Evans, all of these individuals have expressed some sympathy for a nominal GDP level target. But again, I think the concern is the one that you raise, it'd be difficult to transition into anything, any kind of level target, whether it's price level or nominal GDP.

Beckworth: So let me ask this question, if you could start over, let's say, we're starting a whole new world and we started over, there's no baggage, no issues following us. Do you think that FOMC officials would be more open to moving to something like a nominal GDP or price level target if you didn't have these other issues coming along?

Bullard: Yeah. I think from just a theory basis, I think there is more sympathy than there would have been a few years ago. I don't know, you'd have to ask others exactly what they have in mind. But I don't think we're quite, we're not quite to the point where it would get serious discussion. I think, also, you're in a situation here where basically, we're at our unemployment and our inflation goals. So, policy's looking pretty good right now, we're not in recession or anything like that.

Bullard: At some point in the future, we'll probably have more problems than we have today. And that's usually the stimulus to think about other types of approaches to policy. But I think there's sort of a simmering dissatisfaction with the ordinary ways we think about monetary policy because they haven't worked that well post-crisis and they haven't made that much sense post-crisis.

Bullard: And so, to the extent you could go to a level targeting kind of model and make more sense of our current situation and also communicate that effectively, I think that might be a wave of the future, but it remains in the future I think today.

Beckworth: All right. Well, let's turn to your research that you've done on this issue. And you've written several, you've coauthored several papers on motivation for nominal GDP level targeting. And it's motivated by this idea of incomplete credit markets. So, can you speak to that and what's the insight your paper shares?

Incomplete Credit Markets and NGDP Targeting

Bullard: Yeah. So, in the paper, we get rid of sticky prices. So we allow flexible prices, but we have a friction somewhere else in the model which is in the credit market. And this has been something that's been talked about for decades in macroeconomics that, just to put it in practical terms, people sign mortgage contracts, which are a big deal in their lives, and they sign them based on nominal magnitudes. And they're not state contingent.

Bullard: So what theory would tell you is they should be signed in real magnitudes and they should depend on your ability to repay in the future as to how much you would actually repay. But we don't do it that way maybe for good reasons. So, if you have this nominal contracting problem and you have aggregate shocks, we have an aggregate shock to productivity, then what's happening is every time there's a shock you're reallocating some to borrowers or lenders. And this is sort of messing up everybody's use of the credit market because in our model, most of the use of the credit market is to pull consumption forward in the life cycle. 30 year olds and 40 year olds when they want to buy houses and consume housing services earlier in their life, then their income really allows, and that's why they borrow.

Bullard: But shocks hit the economy, and then this messes up the allocation of resources. The good news is that the monetary authority can fix that through an appropriate nominal on GDP targeting policy. When the shock hits, then you have to change the price level in just the right way in order to restore the allocation of resources that would have occurred had everyone done the right kind of contracts instead of these nominal contracts.

Bullard: So, I think that that's a pretty robust, more robust than I thought, initially. I actually discussed the paper by Kevin Sheedy that some listeners might know at Brookings years ago.

Beckworth: Yup, I remember that.

Bullard: And initially, I thought that his results would not extend to a model with more heterogeneity than what he had. But when I did it my way, it turned out that it worked just as well, and the intuition was very similar. So that made me think that this kind of result is maybe more widespread, would work across a wider variety of models than people think.

Bullard: Also, I think that this kind of story that the main problem in the economy is not sticky prices but instead is credit markets where the contracting is suboptimal, I think that that's at, a gut level, I think rings true, I think you see all this mortgage debt out there now running 12 to 13 trillion, that's trillion with a T. It's all nominal contracts, every time there's a shock, you're misallocating. So, nominal GDP targeting would be exactly the thing to fix that. It would provide the missing insurance in those markets, and you get better allocation of resources.

Bullard: This general theme I think is an interesting one. I'm hoping to write more papers along this line. But obviously, this is a relatively new type of theory relative to what's been done in monetary economics up to now.

Beckworth: Yeah, it's very interesting. Your papers, I think you got at least two on that topic, right?

Bullard: We did a new one with more stuff about labor markets.

Beckworth: And then you mentioned Kevin Sheedy's work and then Evan Koenig from Dallas that has also done some similar work on that. I think it's a very compelling argument. If most of our contracts because of incomplete markets are denominated or nominal contracts, and it makes sense to want to keep nominal income stable. You can still have real adjustments, but you want to minimize the distortions created by these incomplete credit markets.

Beckworth: Let me go back to something you hinted at I believe earlier, another justification or motivation for a nominal GDP target. And that is maybe a knowledge problem, right? So, if you think of an inflation targeting framework, you think of maybe a Taylor rule kind of guides the behavior or maybe explains how a central bank would react in an inflation targeting environment. One of the big issues is knowing in real time the output gap. So I know you know Orphanides works from the 1970s, where he shows that, lo and behold, the Fed was following the Taylor rule and was doing the best, it could just didn't have the best real time measures of the output gap.

Beckworth: One argument that's been presented is, one way around that is not try and to get at the output gap. Target nominal GDP, pick that level, go with it, and you don't have to play God anymore, you don't have to try to guess and pick and choose, what's the output gap, what's going on. Do you see that as another argument for nominal GDP targeting, it simplifies monetary policy, it makes it easier on you as a monetary policy official?

Bullard: Yeah, I think, I mean, the Orphanides stuff was very good. And any model that is going to be based on gaps, and it's really, it's the real economy gap, maybe distance between unemployment between what you think the desired level would be or distance between actual output and the flexible price level of output. Those kinds of things get very problematic very quickly and I think that was the value of Orphanides' work in saying how that contributed to the volatility in the 70s and the run up of inflation in the 70s.

Bullard: I think the rap on nominal GDP targeting has been that you could name, let's say, a 4% nominal GDP target, and the growth rate might slow down, real growth rate might slow down secularly. And then all you end up with is zero real growth and 4% inflation. And then if there's a welfare cost of inflation, then now you just penalize the economy more than you have to in a world where the actual growth rate has slowed down to zero. So those people would not be very happy in that kind of a world, zero real growth, and plus they have to put up with a higher inflation.

Bullard: So I think you want to be able to adjust for trends in the real growth rate of the economy in a way that sort of respects what's going on. One of the things about Orphanides is that he said that, there was a productivity slowdown in 1973. And policymakers kept attributing that to the output gap itself, and therefore, kept monetary policy too easy because they didn't recognize the productivity slowdown. It took quite a few years before it became conventional knowledge, probably four or five or even six years before it became sort of accepted that there had been a productivity slowdown and that this had led to probably a policy mistake in the mid 1970s.

Bullard: So, I think the Orphanides work shows that you got to pay attention to the underlying real growth rate economy when you're thinking about nominal GDP targeting. And just let me add on this. In the thing I described, productivity growth is bouncing around every quarter because it's a, and there's a shock on that parameter. So it's moving around stochastic. You don't get fooled by the idea that you've got the wrong trend growth rate in that model, the way you would in some of the Orphanides type stories. So, it's a little bit different in that sense.

Beckworth: Okay, well, let's move to current Fed policy. And one of the issues that might be emerging this year or next year would be the Fed's balance sheet. The Fed this year has raised interest rates again and many FOMC officials are talking up more interest rate hikes. Part of the FOMC's plan is to raise interest rates first to some point and then begin to allow its balance sheet to be reduced. You have pushed back against that approach. You've argued that they should start reducing the balance sheet sooner.

Bullard: Yes.

Beckworth: So, can you tell us about that and your reasons for why?

Bullard: Yeah. Originally, I tried to argue that we should, when we're trying to exit, that we should let the balance sheet run off first so the balance sheet would be shrinking. And then at some point later, come with the lift off and get the policy rate off zero and then start to normalize the policy rate.

Bullard: That was in the original exit principles put out by the committee. And then later, that got switched so that we were going to lift off first. And then after that, some of the rumors were six months later that we would allow the balance sheet to shrink six months after we started lift off.

Bullard: Now, as it has worked out, we didn't lift off till December 2015. And then that was after kind of a year of hesitation. And then at that point, we said, well, probably 2016 will have four rate hikes. That didn't work out either. We only raised the policy rate once in 2016 as well. And so, you got quite a ways down the line here. I think the issue of run off of the balance sheet got pushed to the backburner during that period.

Bullard: Now that we're moving rates a little bit more, this issue can come back on the table and we can get on with the business of reducing the size of the balance sheet. I do think it's important. I think our policy right now is not completely coherent in my view because we're raising the policy rate, which is raising interest rates at the short end of the maturity structure.

Bullard: But according to our own rhetoric, the big balance sheet is putting downward pressure on other parts of the yield curve. So you really got this twist going on in the yield curve with everything else equal. And I don't really think that that's the current intentional policy. Normally, when you're trying to raise the policy rate, then you want all the rates to go up more or less in tandem all the way out the yield curve. And with this policy, we're kind of distorting that process.

Bullard: So I am a little bit concerned about that I think if we allowed run off of the balance sheet, then you can get the normalization to occur in a more conventional way, all the way across the yield curve, and I think that's important.

Beckworth: So you're concerned that the current approach may put upward pressure, maybe invert the yield curve, it might be destabilizing for financial firms and other entities.

Bullard: Yeah, I don't know. Invert the yield curve is maybe a little bit strong. This kind of thing is, when I'm talking like this, it's always like taking everything else as given. And of course, everything else is not given and there are many other things that affect interest rates and especially longer term interest rates. But to the extent those things are held constant, what is the thrust of our policy? The thrust of our policy is to have the policy rate moving up, short end of the curve moving up and the medium and longer end of the curve not moving up as much as it otherwise would. In an extreme case, you could talk about yield curve inversion, I don't think we're quite going that far. That would be that sort of the logical conclusion that you'd reach in the end.

Strategies for Shrinking the Fed Balance Sheet

Beckworth: Let me ask this. So the strategy for shrinking the balance sheet is to passively, allow it to passively run out. So as these securities mature, you would simply not invest the funds. If you passively shrink the Fed's balance sheet, it's not going to be a smooth, continuous process, there might be some years where there's a large amount of securities that mature, maybe the next year a smaller one. Would it make more sense to actively manage that reduction of the balance sheet for that reason?

Bullard: Yeah, I think you could do that and I think that would be a fine way to go. You could put out a plan for the pace at which you want the balance sheet to shrink. And then you could ask the New York desk to undertake operations so that it did shrink at exactly that rate. Whether that's really necessary or not, I don't know. If we felt it was, we could certainly implement that. Another way to go would be just to allow the runoff and just go from there.

Bullard: It is a little lumpy and I think I'd like to hear from people in markets if they thought that it was going to cause more volatility than it's worth, then we could manage it down a little bit. But I think some of the thinking has been that the month to month variation is probably not that big. That it would cause significant problems in these markets, but you could argue about it for sure.

Beckworth: All right, so here's another technical question on the Fed's balance sheet. So right now the interest on excess reserves is a bit higher than what a one month treasury bill yield is. And it's been persistently that way for some time. Now, given that observation and also given the new regulatory requirements, the liquidity coverage ratio which banks have to hold certain amount of high quality liquid assets, and now that new regulatory burden treats bank reserves and treasuries as equivalent as safe assets.

Beckworth: If that gap persist, that spread between interest on excess reserves and treasury bills persist, at the moment the Fed's trying to unwind its balance sheets, it seems to me that could present some problems in this way. So the Fed's beginning to pull out bank reserves and it's pushing treasuries back into the market. But bank reserves are earning more because interest on reserves is higher. Banks may not want to let go of those bank reserves. Could there be a problem there if interest in reserves remains significantly higher than treasury bill rate?

Bullard: I haven't thought about that. There is the issue of the total amount of reserves in the system, they have to be held at the system level. Individual banks certainly would have the incentives you're talking about, but the system has to hold on one way or another. And so, the relative desirability of reserves among banks I guess would determine that, but in a system, with reserves, they're going to be holding them one way or the other.

Beckworth: Someone has to hold to them.

Bullard: Someone has to hold them. That's an interesting question and that's a good example of some interaction between balance sheet policy and interest rate policy. I think it's something we'll look at going forward.

Beckworth: Yeah, well, the whole Fed's balance sheet is an interesting experiment of sorts to see how this evolves going forward, will the balance sheet be used as a tool in the future and so forth. Well, let's move to interest rate hikes. There's been one already that we recently just went through, do you see any more coming this year?

Bullard: I'm not sure what the committee is going to do. I mean, you can check the speeches of the other members and they're suggesting more rate hikes ahead. We felt that the outlook for real GDP unemployment inflation just based on hard data is that it's not going to change a lot in 2017. So inflation close to 2%, unemployment fluctuating between four and a half and 5%, growth at about 2%. We think those are numbers that will probably persist through 2017. And for that reason, we're kind of thinking, well, you don't have to do anything with the policy rate really in order to stay near target on inflation and unemployment.

Bullard: So, we're not really suggesting any further moves this year. I could probably be persuaded to make one move or so but I think the biggest issue is, what is the path over the next two years or three years? Is this a situation where we really have to get the policy rate, let's say 200 basis points higher from where it is today? And I don't see that kind of urgency here. It really doesn't look like inflation, even though it has come back to target here, that inflation is going to take off. I'm not really a believer in Phillips Curve type effects. It seems like it would be hard to get the unemployment rate down significantly from where it is today based on 2% growth in the US economy.

Bullard: Now you do have this fiscal policy sitting out there and it's possible that that'll be successful. But on that, I think we can just wait and see what Congress comes up with. And then we can evaluate those changes, most of those changes, if they occur, would lay out there in the future 2018 or later, and they would be drawn out over a period of time. And so, for monetary policy purposes, I think we can, we don't have to be preemptive while we're waiting for the Congress to decide what it wants to do over the next four or five months. I think we can just wait and see what they deliver to us.

Bullard: So just based on the hard data, I think there's not too much change in 2017 for the major variables.

Beckworth: You've mentioned recently in some work in a speech about approaching monetary policy through regimes, and you've mentioned that we've been in a slow growth low interest rate regime. So based on what you just said, I'm interpreting you saying that we're probably still stuck in that regime for some time, we're not going to go into high growth regime anytime soon.

Bullard: Right now, I don't think we have any hard evidence in front of us in that we've all sudden switch to a higher productivity growth regime or that short term real interest rates are suddenly going to be a lot higher. So, I think the short term real interest rate, maybe I'll just talk about that for a second. That in particular, it's a global thing, it's a global object.

Bullard: And when you're doing your Taylor rule, the first thing you have to do is put some kind of number about what the short term real interest rate is. And if you look at a time series of short term, let's say a one year treasury minus either expected or actual inflation, over the last 30 years, that has come down from 4% down to negative 1% over that time period. So that's a long secular trend. And what we're thinking is that it's not wise to predict that that trend is all of a sudden going to turn around and revert to mean.

Bullard: What you should do at least for the next 18 months or two years is you should just expect that that's going to be about where it is, and that's your starting point for your Taylor rule. So you put a -1% as your starting point, then look at output gaps, and you look at inflation gaps, which aren't very big right now. And therefore, you get pretty low policy rates projected out into the future from that. And so, that's part of what we're thinking in the regime story.

Beckworth: Very interesting. So one last question as we are now running out of time, are you hopeful long term the US economy will go to a high growth rate regime? Will we return to more robust levels of growth or are we kind of stuck in this rut for a sustained period?

Bullard: I think that the good news about regime switching is that regimes do switch. We have seen it in the fairly recent past because between 1995 and 2005, we actually had very high productivity growth in the US, and it made a big difference, especially in the second half of the 1990s. So it can happen again. And I also, I'm very optimistic in the sense that there is fantastic technology around. We see this every day but it's the diffusion of the technology into productive processes and into the workplace to make people more productive. That's what you need to see in order to get the economy to grow faster than it has up to now.

Beckworth: All right. Well, we are now out of time. Our guest today has been Jim Bullard. Jim, thank you for being on the show.

Bullard: Well, thanks for having me, David. It's really a pleasure to be on here.