Robert Hall on GDP Measurement and the Long Slump

Robert Hall is a professor of economics at Stanford University and a senior fellow at the Hoover Institution. Robert has also written widely on macroeconomic issues since the 1960s and is the chairman of the National Bureau of Economic Research’s Committee on Business Cycle Dating, which maintains the chronology of U.S. business cycles. He joins Macro Musings to discuss the difficulties of measuring gross domestic product and dating the beginning and end of recessions. David and Bob also talk about the pros and cons of nominal GDP targeting and price level targeting. Finally, Bob shares his thoughts on why our economy has performed so lethargically since the Great Recession.

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: Bob, welcome to the show.

Bob Hall: Thank you.

Beckworth: I always ask my guests when we get going as to how they got into economics. What was your journey into this field?

Hall: Thanks. It was very similar, astonishingly similar to many economists in my generation, which is, when I started college, I thought I wanted to be in physics. Physics at that time ‑‑ this is the early 1960s ‑‑ was sort of the classy subject. Then I discovered that I wasn't as good as some of my rivals at doing physics, and it wasn't making use of other talents that I had, more on the verbal side, so I sort of shopped around.

Hall: I was a philosophy major briefly, but that didn't work. I enrolled in a mathematical version of microeconomics, and bingo. That was definitely what I wanted to do. I've never regretted it, and I've never deviated from it.

Beckworth: You must have had a great teacher, too, to have kept you on it. Because I know sometimes, teachers can make the difference in whether one sticks with a course or not, particularly in economics. I've heard many students say that they either love or hate economics, and it often goes based on the teacher.

Hall: Yes. It was Sidney Winter, who's quite a prominent economist.

Beckworth: Huh! Interesting.

Hall: I've always been grateful to him. I'm not sure what would have happened without him.

Beckworth: Your work is amazing, getting ready for this show, I went over your data and the breadth and the depth of your work, again, goes everywhere from growth to optimal taxation, labor markets dynamics, fiscal policy, monetary policy, technical innovations, new money economics, slumps. It ranges from the late '60s if I'm not mistaken to the present. That's just amazing. I'm dying to know, I'm sure our listeners are too. How do you stay intellectually curious for so long and keep plotting along?

Hall: I don't know. I'm not sure that it's possible to answer that. I guess the boring answer is I didn't know what else to do. When I turned 60, I asked myself whether I wanted to become an old fogey. An old fogey is somebody who says, "Well, I'm just gonna leave the details to the younger guys and just talk about the big stuff." I decided not to. That have been bland, so, I continued to burrow into things in detail. I do a lot of joint work now with young hotshots, so that's a big help.

Beckworth: Yeah, looking at your work, you're still doing pretty rigorous, Dynamic Stochastic General Equilibrium models. You definitely don't shy away from tough topics. It's encouraging and if you have some secret sauce to share with us. You haven't seemed to burn out. You kind of remind me of Milton Friedman was sharp until the very end.

Hall: Absolutely.

Beckworth: It's encouraging to see people like you. No deep secret to how you do it, huh? You just plot away.

Hall: I don't think so. Very conventional life in a whole lot of respects. There's nothing special that I can recommend, but my father was somewhat the same. He was an academic, and he was active from...He died at age 80, and he had been going to the office every day and writing stuff. There's some heredity working here.

Beckworth: Let's move on to an area you're well known for. That's the Recession Dating Committee. What is it like to serve on that committee? How do you call a meeting? I've learned just in doing research for this program that you are the only chair of that committee. You've been there since the beginning. Is that correct?

The NBER Recession Dating Committee

Hall: Yeah. The National Bureau of Economic Research, which is the highly respected organization that maintains the chronology, has actually done that since the 1920s, but it was done more informally by a small group that didn't seem to require a chair. When Martin Feldstein took over the national bureau in 1978, he thought that the process ought to be formalized. He was probably right, because soon after that, starting in 1980, we had a couple of recessions, one not so big and the other very big.

Hall: It was probably a good thing that he sensed that maybe it was a good idea to make it a little more formal and more transparent, which ‑‑ we initiated the process of explaining what we were doing, having an announcement, a memo that we write every time we have a meeting that explains to the world why we met and what we were thinking about. Sometimes if we take action, if we need to enter a new date into the logbook then we explain that thoroughly. I was put in charge of that process and I've been doing it ever since.

Beckworth: What is it like during a meeting? Are you the one who calls the meeting and if so, do you wait for some catastrophe, what appears to be a recession going on? Is there a regular meeting of the committee?

Hall: No, it does not meet regularly. It's very much event‑based. We get the most attention when it's become reasonably obvious that the economy has started moving downward which it usually does in a pretty well‑defined way. But we also are responsible for figuring out turning point, what we call "trough."

Hall: The committee gives a date of the peak of what we'll call economic activity, which is mostly real GDP, and then the date of the trough. We get more attention, though, because the onset of a recession is kind of a big deal event. The determination that we've finally hit bottom and now we're growing again is not quite as much...receive quite as much attention.

Beckworth: Typically you announce the official beginning of a recession well after it's started, right? You want to be sure.

Hall: Oh yes.

Beckworth: You want to be careful.

Hall: The headline in "The Wall Street Journal" is invariably "National Bureau Finally Determines Date of the Onset of Recession." It's on page three. It's never front‑page news.

Beckworth: It's got to be a political job of sorts, right? Because there's certain people, I'm sure, are pressuring you to make it official. Others don't want you to. Do you ever feel that?

Hall: Not recently. Early there was just a whiff of that. I won't go into the details, but it was when I was in DC one day and someone of a long‑ago administration for political reasons wanted to have a date that was the other administration. The outgoing administration instead of the incoming one. [laughs]

Beckworth: That's hilarious. Unsurprising, but hilarious.

Hall: It was done very subtly, and of course I brushed it off. It's really just not been a factor.

Beckworth: You mentioned you look at GDP and I went to your website. For listeners out there they have a nice website that answers a lot of questions and you provide data. But you look at, among other things, you mention GDP but you also look at GDI growth, domestic income, employment.

Beckworth: Then as an aside you say you also look at real cells and industrial production. They aren't complete measures of broad economic activity. How do you weight GDP, GDI, and employment? What emphasis do you place on each of those? Any way to let us into the decision‑making process, how it's done?

Economic Factors of Recession Dating

Hall: The first remark is that we are fundamentally hanging on something we call activity, which we generally affiliate with either real output or real income. The output‑income identity says that except for measurement issues, output and income are the same. We exploit that.

Hall: You mentioned GDP and GDI. We often, and many other people do too, look at the average of the two. There's been some studies that suggest that that is a more accurate measure and that what you should not do is neglect GDI, so we don't.

Hall: Those two measures are at the top and they get equal weight, because we literally just weight them equally. Down from there, we look at the labor market. We'll look at unemployment, we'll look at employment. Two measures of employment, two different surveys of employment.

Hall: Then as you mentioned, if we want to track down some special factor that might be in some sense distorting a date, usually only by a month or two, for example, in the census years, it actually makes a difference. The government hires so many census takers that it actually shows up as a blip in the economy.

Hall: That's the sort of thing we want to think about. More than anything else, we study activity defined to be output and income equally weighed. It's not formal. It's very much trying to think about how we've done things in the past, and do them consistently, but also intelligently, as the economy evolves.

Hall: Sometimes, the point is made that you can come close to replicating what we do in some very simple‑minded ways. It's not a problem, because of course we are doing something that is guided by a published number, real GDP here, real GDI, but we actually spend some time thinking about it.

Hall: Typically, you ask about what a meeting's like. Meetings, in the last few times, has always been on the phone. Then we circulate spreadsheets and charts, so that everyone's looking at the same thing, and talk about it. The ability to communicate data rapidly over the Internet has made the task much simpler, since it means we don't have to have a physical meeting.

Beckworth: Now, there is this rule of thumb. When I first got into economics, they said, "Oh, a recession is, as you know, this rule of thumb, two quarters of negative growth in real GDP." But that often isn't the case, and I suspect that's not really what you rely upon. Do you know where this idea began, that you could have two quarters of negative GDP growth?

Hall: Actually, I don't. It comes up all the time, and the FAQs on the website you mentioned do go into them. I've got to say, it's not completely crazy. There's been times when it looked like it came reasonably close to replicating what we do, but there was a very special experience in 1980 that didn't fit that mold, and yet clearly we wanted to call it a recession, just a very quick recession.

Beckworth: 2001 was like that as well, right?

Hall: 2001 was very tricky. One thing I should mention is that, although the official real GDP, GDI numbers are quarterly, four times a year, you can actually put together from the information that the government publishes, monthly numbers. We actually look at the monthly numbers, but the monthly numbers in the 2001 recession really didn't give a very definitive.

Hall: Normally, you receive a very definitive, the economy went up, and up, and it turned around and started going down, and that's the recession beginning. That just wasn't true in that one. It was a very special thing. On the other hand, the Great Recession, starting at the end of 2007, was just very clear.

Hall: There's never been any doubt about the onset date, and the ending date, too. 2001 was kind of special, then we got back with a much more serious recession that began at the end of '07. We were back into what we regard as the normal mode, but you're right. It was much less clear than it usually is, when it began.

Beckworth: Do you ever sit back and think about all these studies that have been done based on your dating? All these probate models that predict when a recession's going to occur, they take your committees dates, and they plug that into a model, and they take the yield curve spread. Do you ever sit back and go, "Man, my work is being used to help further the understanding of business cycles"?

Hall: It must be, because it's very prominent activity at the national bureau. It's one of the things that many people hear about first. It's part of a flagship of the bureau's public facing activities. I think it's a help to have something that's respected, that's more than just plugging in the GDP numbers themselves.

Hall: It does focus on this thing, which does seem to exist, a business cycle. When you look at the economy, most of the time, the economy is growing, sometimes rapidly, sometimes not so rapidly, but it's growing. Then there are times, usually much briefer, when the economy is shrinking, often very rapidly, as it did in 2008, especially after the crisis in September of '08.

Hall: There really is something there, and I think it is a help researching it, but on the other hand I'd say the majority of research is done more on just looking at the data directly, rather than putting in gray bars, which is what we contribute, the gray bars that people put in these spots.

Beckworth: Right. Your legacy is there every time I open up the Fed database, and I see the gray bars, or I'll open an article, and I see the gray bars. One last question about the recession community before we move on to our next topic that is a question of GDP. Is it properly measuring the economic activity as well as it used to?

Beckworth: By that, I mean there's a number of things in the past that were recorded in GDP. For example, things like encyclopedias, maps, maybe your stereo system. All of those can now be found as apps on your phone, on your smartphone. They're either free apps, or they're much cheaper.

Beckworth: To some extent, there might be some measurement issue here. GDP's not capturing some of the effects, some of the developments that have emerged with our economy going increasingly more digital. Do you think that is making GDP a less precise measure, or is it still fairly reliable?

Hall: First of all, that's an issue that's been raised long before the Internet came into being.

Beckworth: Fair point.

Hall: After all, broadcast figures, the same point could have been made when TV got started in the 1940s. We're suddenly providing something that people really like, that they got for free as long as they were willing to sit through ads. Of course, ads are the same thing on all the apps that you're talking about, they'll typically expose you to ads, too.

Hall: It's not a new problem, but it's an issue. There's an interesting new national bureau working paper by Chance Iverson at the University of Chicago that goes into that issue in detail, you've probably seen it, and says it's actually a big deal. Even if you accepted the argument that you've got to plug these things in, in some kind of estimated value, it's very small compared to output.

Hall: It's very interesting. It's added a lot to people's lives, but the amount of economic activity that goes into it, compared to a lot of boring stuff that is big, construction, mining, all those sectors that move around are pretty small. It's an issue.

Hall: No one's ever said that the GDP system was perfect. There are many things that give rise to people's satisfaction, that are done at home, that aren't touched at all, but it doesn't pretend to do that. It does something that seems to be useful, and that's market production of goods. It's useful in that it gives people value, but it's not total economic satisfaction by any means.

Beckworth: That study that you mentioned, I think there's another one from Brookings that came out, that made the same argument. I was frankly surprised by it, but it does say what it does in terms of the measurement issue not being that big of a deal. Let's move on to some work you've been doing lately on the long slump, or the persistent slump that we've experienced since coming out of the bottom, the trough of the business cycle that ended in June of 2009.

Beckworth: You've had a series of papers, and I probably didn't even see all of them, but I want to mention a few of them for our listeners. You had a paper in 2011 that was actually your presidential address to the American Economic Association called "The Long Slump."

Beckworth: You've had a more recent one, "Anatomy of Stagnation in Modern Economy." You've had another paper, "The Macroeconomics of Persistent Slump." You've really been looking in at this topic, and in trying to make sense of what we've been through these past few years, since the end of the Great Recession. Let me begin by asking you a very basic question. What is a slump, and how is that different than the contraction?

Slumps vs. Contractions

Hall: The contraction begins when the economy turns around, stops growing, and starts shrinking. Then you get down to the trough. It's typically a period of 6 months, 18 months, or maybe even longer, of contraction. Then, typically, as I mentioned before, a period following that, the recovery, when the economy, at first, in particular, tends to grow slowly, and then it picks up steam, gets back to normal, and then sometimes goes beyond normal, like it did in, say, 1999.

Hall: Focusing on rates of change, when the rate of change is negative, the economy is shrinking. We call that a recession. When it turns around, and starts growing, we call that an expansion. That's the slope related terminology.

Hall: Then, the other terminology, which is very useful, especially for what the economy's been through in the last eight years; is a slump is a time when the performance of the economy is substandard, means that the unemployment rate is high, growth is typically low, income is lower. Things are just not very good, and that occurs. The slump begins, usually pretty early in the recession, and continues long after the economy turns around.

Hall: The thing I remember very clearly was, the day that we decided that June of 2009 was the end of the recession, the President of the United States of America said, "What do these people know? Look around, this economy is still in terrible shape. How can they say the recession is over?"

Hall: We didn't say the recession was over. We did say the recession was over, but we didn't say the slump was over. We actually said the slump was it its worst. Now, it looks like it's going to gradually recover, which it did. This was confusion between the slump, which is basically bad times, and a recession, which is worsening times.

Hall: It's very, very difficult to get that point across. In fact, many, many economists will refer to bad times as a recession, even though the economy is growing, as it has been since 2009, which is a long time. There's not much we can do to overcome that, but one of the reasons I got slump into the title was, that was an attempt to steer some people in the direction of getting their terminology right.

Hall: You mentioned the gray bars. The gray bars are the contraction times, it's not the slump times. It has occurred to me that it might be of interest to try to have another chronology with, I don't know, blue bars that define slumps. There would be periods when, say, the unemployment rate is above six‑and‑a‑half percent, or something like that.

Hall: There would definitely be bad times, but will include these time is when things are bad, but getting better. It's still a slump. Then, we avoid being excoriated be the President of the United States of America.

Beckworth: Right. You recognize the times are tough, and just that the economy has turned a corner. Let's talk about the prolonged slump that we've been through. In a number of your papers, you get to driving forces that guide us into the slump, and some of them seemed to be tied to, I'll say cyclical, but these developments that aren't necessarily long‑term developments.

Beckworth: For example, in your AEA Presidential address, you mentioned the overhang of housing and consumer durables, household debt, the financial frictions. In your more recent papers, you've mentioned things that might be more of a permanent nature like productivity declines, labor force participation rates that may not come back. How do we make sense of these driving forces? What's a way to think about them, and what does it mean for the slump?

The Driving Economic Forces of a Slump

Hall: The first thing I would say is, I think we can easily divide them into what we've called cyclical forces. A lot of them, in the most recent experience, were unleashed by the financial crisis, which seemed to put households and businesses into a very constrained situation, where they pulled back in consumption, spending by households, and employment, job creation by businesses.

Hall: That type of thing is, I think, pretty well measured by unemployment. Unemployment shot up to 10 percent, from its normal level of around five percent, and then it worked its way down, and in fact it's back down to its normal. Unemployment is now 4.7 percent, which is right around its normal level.

Hall: A lot of those things have to do with short term spending decisions in households and businesses. Then, the driving forces, since those forces have abated, unemployment's back to normal. Then we've seen some others that are special. Some of them, I didn't even mention in the presidential address, because they haven't become so apparent.

Hall: The most important one is the decline in productivity growth. Productivity is rising, but at a very disappointingly low rate compared to its historical increases. That's the most important one. The second important one is, it interacts with the cyclical forces, and that is that during the financial crisis, and afterwards, capital formation in the US economy was way below normal.

Hall: That has a follow‑on effect, in that capital formation forms capital, and we got into a situation of capital depletion. We didn't have as much new plants, equipment, and other types of capital, and therefore produced less. That's actually, as of today, that's the second most important.

Hall: The third one is, and it's gotten a lot of attention, and I think a lot of misunderstanding, is that the labor force has shrunk. The labor force is people who are either working or looking for work, and we've had about a three percentage point shrinkage relative to normal trends, and relative to projections that were made before the recession.

Hall: At first we thought, "That's easy to explain, because there's a problem of job availability." Now, with the labor market looking very normal, and some signs of, especially on the employer side, that they are having trouble finding workers. Yet, it's not drawing people back into the labor force. The labor force is surprisingly low, and that's the third factor. Those are the big three that are driving forces that are very active today, even though the basic indicator of unemployment is totally back to normal.

Beckworth: Do you think any of those structural driving forces were influenced by the length of the recovery? Someone like Larry Summers would argue, our economy basically has gone through economic atrophy. Just like a muscle that doesn't get used regularly, it's potential is going to shrink.

Beckworth: These resources, the labor force is going to get smaller because people have dropped out. Even elderly people who might otherwise be working have just decided to drop out. Capital formation is not there because there's been a sustained decline in demand for it. Productivity would be higher. In other words, there's some kind of endogenous side to the supply side of the economy. Do you think there's any merit to that argument?

Hall: First of all, Larry and I have talked about this personally, quite a bit. He spent the month of March here, and we had an extensive set of discussions on the point. I don't think we're too far apart. I've made more of an attempt to dig into the quantification of those things than he has, so I can point to some measures.

Hall: On the labor force, it's still not clear. The surprising contraction in the labor force actually started before the financial crisis. You can see a drop further, a more rapid drop of participation after September of '08. It's clear that there was some impact. How much lasting impact there is, I think is still a research topic.

Hall: On the other two, I think we have some hard data that confirms Summer's opinion. I mentioned before that we have this capital depletion, which is a term I would use for the fact that we didn't build capital stock nearly as much as we would normally have done, starting at the end of '08, but especially in '09 and '10.

Hall: We're suffering from that. We can't produce as much without. We don't have planned equipment. The other thing is unproductivity; there was just a complete collapse of R&D spending in the US economy. It's not gotten as much attention as probably it deserves, but actually it began in 2000.

Hall: We had this big upward spike of R&D, intellectual property formation, during the dot‑com bubble. Then it collapsed, and you can see the bubble collapsing. Then it started collapsing down to astonishingly low levels, and it hasn't turned up. This is something that's now included in the most broad measures of the capital stock.

Hall: It's another form of capital depletion. Now, there's areas to think that also means that at least some parts of productivity arrive magically. They arrive because people start thinking, and they get charged off as R&D workers. They do something, they create something, and then it becomes useful, broadly, so it shows up in productivity. It's not just part of their own capital.

Hall: There's every reason to think that the decline in R&D spending...the thing is you don't see any...there's no sign at all that R&D spending was particularly impeded by the crisis. It looks like it was a process that began before that. It certainly continued after the crisis. Now, of course, I don't think that Larry believes that everything that's wrong with the economy today can be blamed on the crisis. In that respect I think the signing is very supportive of the position that he's taking. There's a very big question of just what policy can do about it, but I think it is...everything points in the wrong direction.

The Prime Age Employment Rate Story

Beckworth: Let me ask you about the labor market question. One indicator that I've always struggled with relative to the standard headline unemployment rate is the employment to population ratio or the employment rate for prime age worker, 25 to, I think, 54.

Beckworth: Lo and behold when I was looking through your vita you actually had a recent paper on that with the San Francisco Fed. One thing that struck me about that is these are prime age workers. These aren't people who are retiring, people who should be actively in the labor force. That thing it fell and it's recovered about halfway. The fact that it has...

Beckworth: Two observations. Number one, the fact that it has recovered over this long period suggests that some of that is cyclical, but it still, number two, it has a long ways to go. I look at that. I'm like, well, man, there's got to be something unrelated to structural developments, but maybe there is.

Beckworth: I came across your paper. Correct me if I'm wrong, but what you found is in that particular group of those prime age workers it's the high income workers who haven't returned to the labor force. Is that right?

Hall: Not quite. It's very important that this is...I've had a lot of trouble explaining the finding there because the people who don't think it's right interpret it the way you did. But actually what the concept we're looking at there is household income.

Beckworth: Oh, OK. Household income.

Hall: It's an on‑going project and that paper was the result of something we found that we thought was too important to wait until we finished the whole project which could be forever. The finding is that if...

Hall: I think the best way to interpret is that if you're someone in that age group, 25 to 64, who has a mate who is a high earner then you're more likely to be one of the people who decided that they didn't want to be in the labor force in spite of being in a group which normally quite a very high fraction of the people in that age are in the labor force. What it mainly points to is that the decision whether or not to be in the labor force is a household decision, or a family decision, not an individual decision.

Hall: If you just look at almost all the research that has been done on what's happened recently in participation and just looked at the characteristics of the individuals, and if you do that you see quite noticeably that most of the decline has been among the results with relatively lower earnings. In other words putting these two facts together it's the low earning members of households that also have a high‑earning member who were most likely to be the ones who were withdrawing from the labor force.

Beckworth: That's an interesting story, some kind of comparative advantage at home going on.

Hall: Exactly. That's right. It actually fits with the numbers of what's happened to wage changes. We had this widening of the wage distribution, which means that those high‑income people have been doing quite well. But the other members of the family whose income prospects or earnings prospects aren't as good aren't doing as well. Well, surprise, the ones that have had rapid earnings growth love to work and those who are benefiting from being able to consume what the high income people are making but aren't facing the increases themselves, well, they are less likely to work.

Beckworth: Very interesting.

Hall: It's as we like say, the income effect and the substitution effect of these changes go in the same direction as toward less work. I think it's a coherent story. I have to stay tuned.

Beckworth: Imagine we have a great economic boom. I'm not sure what it would be, but just imagine some big boom emerges. Do you think that would change the behavior of these individuals in the households with high income?

Hall: I think it depends. In some past booms, like in the 1960s, we had a closing of the wage distribution so that the benefits of the boom went differentially to the lower income workers. If we had that kind of a boom, of course, that would have a bid effect in the opposite direction. If it turns out to be more of the same, which is the way things have evolved, certainly that's what happened in the '90s and up to today, the trend has been very much toward the high educated, talented people who have been making good money throughout their careers, have enjoyed most of the benefits.

Hall: The wages have been more or less static among those at the other end of the wage distribution. If that goes on we're going to see more of the same, I think. We're not very good at predicting these things. Some magic could occur which would be extremely welcome where it's the low wage people that enjoy the rapid growth.

Beckworth: This discussion reminds of Charles Murray's book, "Coming Apart," where he details quite extensively these two groups that you've just talked about, the upper income successful people who are getting ahead, and then the low‑income ones that have all the social problem and they're falling behind.

Beckworth: One other area you touched on in at least your AEA presidential address and, of course, that was 2011, a little farther back ‑‑ I don't think it's in your 2016 papers ‑‑ but that is the zero lower bound. I'm going to read a quote of yours from your presidential address in 2011.

Beckworth: You say, "These adverse forces were so destructive because the economy was unable to lower its interest rate to stimulate other kinds of spending." I'm very sympathetic to what I think is the implication. I think the implication here is at the end of the day the real problem is zero lower bound. That's why it enabled or gave teeth to these other driving forces as a problem. Is that right?

Implications of the Zero Lower Bound

Hall: Yes, that's right. It was an amplification factor. Normally we'd be able to use an expansionary monetary policy to offset the shock effects of the financial crisis, but since the financial crisis resulted in very low interest rates, on‑going low interest rates, we could not drive the interest rate below zero. That exacerbated...unemployment wouldn't have risen as far as it did if we'd been able to use more monetary stimulus.

Hall: That factor's in principle disappeared now since we're not at the zero lower bound today. Of course, it could happen again. That's something that Larry Summers is very concerned that we're going to be back into the lower zero boundary in the future.

Hall: I am, too. I'm worried that the Fed gets ahead of the recovery. This all depends on where this natural interest rate is, the idea that there's some underlining fundamental rate that would put us near full employment in the market clearing interest rate. The question is, is the Fed going to get ahead of that in the recovery?

Hall: Back in the crisis though I guess the implication I take away from that is if for some reason we were able to get past that zero lower bound, if we didn't have this institutional constraint of physical cash which people would go to when rates started to go negative, the market would have healed itself.

Hall: The idea is that based on the severity of the slump interest rates would naturally have gone down. The Fed definitely plays a part in that too, but rates just naturally go down, but they got hung up. There's effectively a price floor that prevented the natural healing process of an economy through prices from working.

Hall: I guess this is where I'm going with this and that is, what was the real reason, that real ultimate reason for the severity? Was it the debt, was it the typical housing boom/bust story, or was it the zero lower bound?

Hall: I guess the Fed answered or else argued that it was the interaction of the two factors. I think the single biggest one, and I think a lot of research now agrees with this, is that the household spending which had been at very high levels, households had been borrowing to spend through about 2006, 2007, and then suddenly were unable to borrow because of the effect on the financial system from the crisis.

Hall: There was a huge swing toward the elimination of this borrowing, and in fact repayment, that is households started saving instead of dissaving, that swing by itself was huge in terms of its effect on demand. It's something that with a super aggressive monetary policy, impossible because of the zero lower bound, but otherwise if they had been there they might have been able to deal with it.

Hall: The zero lower bound took something that was by itself pretty big and made it even bigger by inhibiting the normal corrective mechanism. I think that's a pretty widely accepted story for what happened in the economy through 2010, 2011. What's happened after that I think is not so well understood, in particular the fact that we're back and full employment is not been completely assimilated in many people's thinking.

Beckworth: Let me turn to one of your papers, a paper you gave at the Jackson Hole Symposium by the Kansas City Fed in 2013. I believe it was there. "The route into and out of the zero lower balance," we were talking about then. You gave, I believe, three steps and this route into the zero lower bound. Number one you have this collapse of output demand, which comes from what you just described, the big collapse in household spending.

Beckworth: Then you hit the zero lower bound. But then the third element, which I think is very interesting, the point you brought up, you have to low and stable inflation to get you stuck in that rut in that area. The reason this speaks to me is I look back at the last seven years or so and I see inflation averaging...let me restate that. Core inflation, the core PCE inflation with the Fed targets, its preferred targeted measure, it's been averaging about one‑and‑half percent. It's been really low and stable.

Hall: Exactly.

Beckworth: In order to get the kind of traction to have gotten out of the zero lower bound, we would have needed at least a temporary bout of higher than normal inflation, something that would have temporarily gone up, but ultimately in the long run...

Beckworth: In the long run we want stable, low inflation, but in moments like these we want some kind of flexibility. You've written about this in several places. You may even have done it in your AEA talk. This speaks to the importance of level targeting. You mentioned price level targeting. I'm a big fan of nominal GDP level targeting, but they basically do the same thing.

Hall: I wish. Sometime I'll talk you out of nominal GDP targeting.

Beckworth: Oh, really? I was going to get to your 1994 paper with Greg Mankiw.

Hall: Yeah, that's right. OK, fine. That was a...

Beckworth: Before we get to that let's stick to the whole price level target argument. Explain to us how that would have helped break us out of the zero lower bound.

Price Level Targeting as a Means of Escaping the ZLB

Hall: With inflation running consistently below target it would mean that there was a growing gap between what we thought would be the moderate growth level of the price level and what it actually was. Under a price level targeting that would mandate a temporary period of catch up inflation so we get back on the growth path that has been defined as the target. That's a terrific idea because it's a commitment to do later what you can't do now to expand the economy, but it's also a commitment not to have out of control inflation in the longer run. It's a great idea.

Beckworth: Is there political support for something like that, though? I guess that's something I wonder about.

Hall: It's interesting. I attended a very interesting conference in honor of Michael Woodford's 60th birthday a couple of weeks ago. Woodford is the one who's been pushing this idea the hardest. In fact he gave Jackson Hole paper which became very influential sufficiently that Janet Yellen and actually talked about the potential desirability, not so much of formal price level targeting, but rather of saying let's push the economy harder than we would have otherwise as some kind of a makeup for the long period of inadequate inflation and excess unemployment.

Hall: The expectation that the Fed would do that typically if it gets back into this would be very desirable. It would make a future slump not as serious. Yellen actually articulated that. I'd not realized that she'd done that. There doesn't seem to be...I think with the current makeup of the Federal Reserve Board it's probably not the...it's probably not on to do that very aggressively.

Hall: Not as aggressively as the economists like Woodford, and myself, and probably Janet Yellen if she weren't in an important political office. But it's understandable. It's a fairly difficult concept to get across in a political setting.

Beckworth: Let's say we were granted the gift of having everyone on the Board being in favor of a price level target. Do you worry, though, they would still face an uphill battle in Congress, the public at large? People have a hard time understanding inflation as it is. Again, I like the idea, and I like nominal GDP level targeting even more, but to get to a price level target implies that we're going to allow a period temporarily of inflation going higher.

Beckworth: Now, of course, the trade‑off, the good trade‑off is you would see rapid real growth. You might be less worried about inflation during that time, but it just seems like it'd be a tough...even if you could get the Board of Governors on‑board it'd be a tough political sell.

Hall: Absolutely. I recognize that.

Beckworth: All right. Let's go to a nominal GDP target. You mentioned Michael Woodford and that paper he gave at the Jackson Hole meeting. In other work, previous work, he's talked about a price level target that's adjusted for the output gap. He goes effectively that's very similar to something like a nominal GDP level target. He made that argument in that paper.

Beckworth: Then you have your 1994 paper with Greg Mankiw on nominal income targeting, and in that paper you looked at three versions of that rule, the nominal GDP growth target, nominal GDP level target, and then a nominal GDP hybrid rule. I'm curious to know...first maybe you can share with our listeners what you found, and, secondly, do you still take the view that you shared in the paper back in '94?

The Prospect of Nominal GDP Targeting

Hall: I have to say that writing that paper was a bit of an eye opener because Greg and I both thought that nominal GDP targeting was a great idea. Then we started looking at the numbers, and the paper's endorsement of the idea is way more limited than we thought it would be when we started writing the paper and looking at the data.

Hall: I've worked very hard on that, on the underlying topic, not a question of what policy should be but rather what the implications would be if you tried to stabilize nominal GDP. The answer is that there's so many things that push real GDP around that you would not want monetary policy to try to offset, in particular fluctuations in productivity growth. With nominal GDP targeting unless you have some fancy way of offsetting it then every time you have rapid productivity growth you're going to have to have deflation.

Hall: You're going to be constantly leaning against a very healthy development because otherwise the healthy development would give you fast GDP growth, nominal GDP growth, and then if productivity growth was substandard as it has been recently then you would be inflating the economy unnecessarily in order to...because you were failing to recognize that. Now one thing you didn't catch, I understand you didn't, but I did a Jackson Hole paper in 1985 back before history began advocating what I still think is the right version.

Hall: It's very close to what you just mentioned that Woodford's been thinking about, but I think it's even better, and that is to substitute the unemployment rate for the output gap so it would be a target. People have talked about doing something like this, but ideally it would a level target where you're targeting some combination of the price level and the unemployment rate. The price level would be allowed to be above target if the unemployment rate was above target.

Hall: It would be...I called it an elastic price standard. I still think I probably should try to spend some time with Mike trying to convince him that would be a preferred alternative. He's too much wedded to this idea that the output gap is something you can really measure, but when you get done measuring...

Beckworth: No, I agree. That's difficult. Orfanidis' work on the 1970s raises some good questions about the challenges in doing that. Since you've pushed back against nominal GDP level targeting, let me take a stab at what you said.

Beckworth: It's right. You're right that there's going to be real changes to the economy that's going to be difficult for the Fed to handle them. The way I view nominal GDP level targeting is that it allows the Fed to be agnostic if it will not worry about temporary short run deviations in the price level.

Beckworth: No matter what happens in productivity if you have a nominal GDP target in the long run you still have a nominal anchor if you have a nominal GDP level target. But in the short run you're absolutely right. If productivity goes up you're going to have maybe a mild case of deflation. I don't think you would ever get the kind of deflation you and in the Great Depression. That was more of a demand shock. Late 1800s we had mild deflation associated with rapid economic growth.

Beckworth: George Selgin, who's written some on this, he calls it a productivity norm. He would have nominal GDP level targeting. His argument is we'll never know what the output gap it. We'll never know what productivity is. We can stabilize nominal income. We can do all these things.

Beckworth: There's been some recent research on the importance of this for nominal debt contracts in a world of incomplete markets. Kevin Sheedy, I believe, has a paper on this in the Brookings Paper, Brookings Institute.

Beckworth: Maybe this is a conversation for another time, but I think one of the virtues is that you don't have to know what productivity is as a nominal GDP level target, if you're comfortable with a little bit of temporary variation in a price level.

Hall: Sure, but you can do the same thing much more preferably, because we measure prices and we measure unemployment. We can rig up a target, which is based on those two things, which we measure pretty well, and it'll work beautifully. There's no reason to get them tangled up in anything even related to output gaps or productivity measurement and so on. You can sidestep all those just by measuring prices and unemployment.

Beckworth: That's kind of where I was going, too, is all you need to do is measure nominal GDP and then ignore everything else. Just let nominal GDP run at a certain pace, ignore everything else. Let me move on, though, to another topic that you've touched on, and that's the fiscal theory of the price level. Can you give our listeners just a brief overview of that? How do you view it?

Beckworth: Is it something that we should be taking seriously? I know your colleague out there, John Cochrane, is a big fan of that. I had him on the show recently. He talked about the fiscal theory of the price level, but I want to hear your take on it.

The Fiscal Theory of the Price Level

Hall: Well, John and I discuss this topic endlessly.

Beckworth: Oh, really?

Hall: I don't think there's any substantive disagreement, because it's well understood in that literature that the so‑called fiscal theory of the price level occurs when the fiscal side doesn't respond to what's happening to the revenue and the build‑up of debt of the national government. It just sits there and does something mindless.

Hall: That means then the price level is determined essentially by the amount of debt that's outstanding, because the only way the government balances its budget in the long run is by inflating away or deflating and increasing the value or decreasing the value of the debt. There's nothing wrong with that as a matter of theory. It's just a question of whether that's the way governments run their fiscal policy. Again, in this literature, that's well understood. This is not anything. John would agree.

Hall: The Cochrane view is that they'll know that it is relevant because, certainly historically you've seen this when hyperinflations break out, it's generally because the government is incapable of balancing its budget to normal tax and spending policy.

Hall: On the other hand, certainly...I have a paper on the subject. It's published in the Germany Economic Review, which shows that the US was unquestionably on the fiscally responsible side. Every time that began to seem to be high, especially after World War II, then and pay off the debt, and not by having extreme inflation, rather by having revenue sufficiently high relative to spending to get the debts to decline.

Hall: Then, it seems like there's a break in the US policy induced by 9/11, which just wiped out fiscal discipline. Now, maybe Cochran's right. [laughs] We do now seem to have the mind with fiscal policy that just adds on debt and debt and debt. It's pretty easy these days with low interest rates.

Beckworth: But, you know what...

Hall: The real...

Beckworth: Go ahead. I'm sorry.

Hall: The real issue is, contrary to my expectations, interest rates go up a lot. That would put the federal government and state and local government too in a real vice. If it doesn't happen, then we can cruise along issuing lots and lots of debt. Very popular around the world, very low interest rates, not much of a burden, but if interest rates rise, then we're in deep trouble.

Beckworth: Let's talk about that, for the last part of our conversation today. That is your explanation for the low yields on safe assets around the world, and this is where I actually first got to talk to you, at that Hoover conference. You mentioned your paper.

Hall: Exactly.

Beckworth: This is an interesting paper. I read it, and you present this phenomenon, in terms of modeling, where you have two types of investors. You have the risk‑averse investors, and you have the risk‑loving ones.

Hall: I would say risk‑tolerant, not risk‑loving.

Beckworth: Right, risk‑tolerant, that was what I meant. Can you explain, using that framework, what is going on with these yields? Just this week, was saw the German 10‑year government bond go negative, which is kind of mind‑blowing. How do you explain it in your paper?

Explaining Low Global Safe Asset Yields and Debt Capacity

Hall: The paper says that when you've got these two different classes of investors trading with each other, then the natural trade is for the risk‑tolerant investors to provide insurance through the bond market, to the risk‑averse ones. You see lots and lots of institutions which accomplish that, in which the basic form of it is that the risk‑tolerant investors borrow from the risk‑averse ones, which means that the risk‑averse ones have repaid, ideally with certainty, but in general with low risk, and that's the way they get insurance.

Hall: If they really like that insurance, and they have to pay for it, the way they pay for it is they don't earn much of an interest rate. If the world is shifting in the direction of more wealth in the hands of the risk‑averse, then that's going to mean we're going to have very low world interest rates.

Hall: The paper doesn't say that's the only explanation, and it does not the time to look at the differences across current, different interest rates, in different currencies. It doesn't deal with the question of why the Germans borrowing euros don't have to pay anything if they're borrowing, whereas the corresponding rate for the US is about a percent‑and‑a‑half, which is an interesting question by itself, but the paper doesn't delve in. It's basically insurance. It treats the bond market as being driven by insurance considerations.

Beckworth: That was a very interesting angle. The risk‑tolerant investors are effectively still an insurance to the risk‑averse investors, and your conclusion is, more of the world's wealth is now in the hands of the risk‑averse investors. In closing, why is that? What has led us to this point that the world seems to be more and more risk‑averse, and therefore has brought these yields down?

Hall: A big part of it is rapid income growth among people whose incomes are still low, but China's a leading example. China has a very high saving rate, accumulates claims on behalf of people whose incomes aren't very high, but are high enough so that they're saving now.

Hall: That's generated a big shift in the world's wealth distribution, in favor of large countries, Indonesia is another one, but China's the leading one, obviously, in which you've got significant amounts of wealth. Just because of the number of such people, even if the wealth per person is not as high, that's just a very big country.

Hall: It's run in a way that emphasizes making very safe investments, but the Chinese don't do what the US does, which is to borrow a lot, and then invest in equities in foreign direct investment, and that sort of thing. The US is a gigantic hedge fund, and borrows like crazy from the rest of the world, and uses the investments on some risky stuff which is very high yielding, and makes a lot of money.

Hall: The US is a huge hedge fund, funded by borrowing from countries like China. Not so much recently, but historically, we're monitoring bases a lot. It's a story that's worth thinking about. I'm trying to propagate it.

Beckworth: I share a lot of those views. As you know, in that paper I presented there at the Hoover Conference, there's some overlap.

Hall: Sure.

Beckworth: One last question. I know we're running out of time here, but one interesting implication of this understanding is that a lot of the growth in our national debt, as well as some of the privately provided safe assets, but particularly the public debt that's considered safe, the US, other places as well, is kind of endogenous, or a consequence in part of this demand from China and other parts of the world.

Beckworth: In other words, the President, Congress, they're facing this price signal which is telling them, "Produce more debt, the world really wants it. They want to hold it." Does this suggest that we have a greater debt capacity than would otherwise be the case, at least in the short run?

Hall: Sure. I already mentioned that. What appears to be a borrowing policy that just has no end, if it doesn't get completely out of control, is a natural equilibrium. In other words, it certainly is not my position that the US, with 70, 80 percent of debt ‑‑ no, of 70 or 80 percent of GDP, depending on how you measure it ‑‑ I don't see that as a particularly bad number. It's coming out of this process, and my paper discusses that briefly.

Hall: Looking forward, there's some concern that the government will get too used to it, [laughs] and then it may not just be possible to put on the brakes, and things could really fall apart then. If the world stopped trusting the federal debt, what would happen? It would be a calamity.

Beckworth: It'd be high inflation again, that's for sure. Thank you so much for being on our show. Our guest today has been Bob Hall. Bob, we appreciate your time.

Hall: Thanks very much. It was fun.