As a Federal Reserve official, Joseph Gagnon played a critical role in providing the intellectual justification for the Fed’s quantitative easing (QE) programs. Now a senior fellow at the Peterson Institute for International Economics, Joe joins Macro Musings to discuss the events leading up to the decision to implement QE and its consequences. He and David also discuss how the Fed’s QE compares and contrasts with the QE implemented by the Bank of Japan and the European Central Bank. Finally, Joe shares some of his thoughts on Brexit’s wider implications.
Read the full episode transcript:
Note: While transcripts are lightly edited, they are not rigorously proofed for accuracy. If you notice an error, please reach out to [email protected].
David Beckworth: Joe, welcome to the show.
Joe Gagnon: Hello, thanks. Good to be here.
Beckworth: I want to begin as I always do it with my guests, and ask how did you get into macroeconomics?
Gagnon: Going way back, I was very interested in current affairs. In high school, I would read "Time Magazine" regularly and so I was interested in what's going on in the world. I could've gone many different directions but in college I thought my economics classes were the most interesting.
Gagnon: They were taught by people I had actually read about, Otto Eckstein in particular who was one of the Times regular contributors way back then. He was one of the lecturers in ECON 10, which was the introduction. I was thinking about science but that was sciencey enough and had some current events interests. That became my passion and I like macroeconomics because I really like the broad focus on the big picture.
Beckworth: How did you get into the Federal Reserve?
Gagnon: After college, I got a job as a Research Assistant at the Federal Reserve Bank of Boston. I went to graduate school and reaffirmed that I did want do macro and so interviewed for jobs in academics and the Fed. That was a tough decision actually.
Gagnon: I thought the Fed in Washington having a bigger pool of people, that I could talk to and work with and do research with. Most university departments are quite small. You might only have one or two colleagues that you could talk to and I felt especially at the young age I would want to rather have more.
Gagnon: The Board has 200 PhD economists here in Washington. Most of them are interested in macroeconomics. It just seemed less restrictive. You mentioned the Berkeley thing. I did teach for a year at Berkeley and I was thinking about, "Do I want to go back into academics." That's when I really felt like, "No, I missed being in Washington. I missed actually working on policy stuff." So, I never looked back.
Beckworth: You've got a really productive run at the Federal Reserve Board of Governors. In the many years, you were also very productive with the publishing. What is it like to work at the Board of Governors? When you get up in the morning, you go in. You have chats with your fellow economists. You do briefings, research. What is a typical day in the life of a Board of Governor economist?
Gagnon: It changes over time. If you get promoted and you stay then you do different things. That was one of the reasons I left in the end, because I had gotten far more management duties than I used to have. In the end, I just preferred actually doing economics and not going to meetings and having to give performance reviews and...
Gagnon: That kind of stuff. That's why I left. But as you start out they give you quite a lot of time to do research. Basically you're allowed to finish. Most people maybe who haven't finished their PhD they are given some time to finish it. If you have finished your PhD, you're given time to turn it into articles. Submit it or revise it or do follow‑ups to it.
Gagnon: This is sort of a honeymoon period of six to nine months where you aren't asked to do very much at all. Then gradually you're slotted into the work program but you still have plenty of time to do research. A lot of it is self‑directed in the end. I think people over time either succeed at research, and defend their turf to do more research. If you to research that's productive and people like you, you get more ability to do more of it.
Gagnon: Some people find it frustrating or they think it's not their thing and prefer working on policy work and people self‑separate and of course many people leave to go into academics or in to "Wall Street."
Beckworth: Let's move on to an area you're very well‑known and that's the Fed's quantitative easing program or QE and that's it's large‑scale asset purchase program for one, I went out and bought lots of treasuries and mortgage‑backed securities. You were part of the Federal Reserve during, at least, part of these programs. Tell us what is QE and how is it supposed to work?
Quantitative Easing and How it Works
Gagnon: Let me give you background and then we can...
Gagnon: We were heading towards a zero policy rate. The federal funds rate is the normal policy tool that were used to think about and it was going to hit zero. Then what? We all thought they couldn't go below zero, which is interesting because now we're seeing that maybe it can go below zero but at the time it didn't even occur to us that it could go below.
Beckworth: Sure, that's reasonable.
Gagnon: I mean we get into why that is but anyway we saw that as a bound. Then the question was what next? A bunch of us were asked to do some research into various options. This is now public information for our guests on the Peterson website I posted. If you go to the real‑time blog and look under my post you'll see a post where I posted the transcripts or the actual papers that were written back in 2008 and that were given to the FOMC in December 2008. When the five‑year period was up for secrecy, I got a FOIA request and got them to release that and that was priceless.
Gagnon: Now we have, you can see the 21 papers that were written by various staff economists, mostly in New York and Washington and including three by me and it was on different options ‑‑ my options. Some of them were studying other countries, Japan and some were at different options. I can't even remember them all. Mine were about buying longer‑term assets and pushing down term premium and buying riskier assets or different assets, like mortgage‑backed securities and pushing their rates of return down and lowering mortgage rates to households. That ended up being what we called them the QE1.
Gagnon: Now there was another thing that we did that was meant to be big, but never turned out that big, which was called the, Term Asset‑Backed Securities Lending Facility, I think, TALF. That was also meant to be $1 trillion.
Gagnon: It was meant to unclog the lending system basically by saying that if you originate a loan for credit cards or student loans or auto loans is it a government guaranteed or backed by an asset, by car or something, you could package it and sell it to the Fed and the Fed would buy it and therefore if you're a bank you would have to hold on your balance sheet and you could free up your ability to lend.
Gagnon: That was another one of these big programs, but it didn't turn out to be that big. The Fed charged a premium over what the market normally charged and so there was less demand for it. What we mostly did was what I did.
Gagnon: Then the question was, if we buy lots of long‑term bonds will it have any effect? The very efficient markets' theory says, it shouldn't. It says that people will see that this washes through any premium the Fed's earning on these assets will be revenue for the treasury, which will lower future tax rates and has no effect on people in the aggregate. For an efficient market as well it shouldn't work.
Gagnon: There weren't many studies of it, I mean people who had done research. There were some people like Ben Friedman who'd gotten fairly large estimates that it would be a big effect, but his results were just not very robust. Interestingly there's a little‑known fact that the person I had to go and say. "Well, what is the best guess?"
Gagnon: There was all this work on Operation Twist but when I looked into it, I said, "Well, they actually never did much so why would you expect there to be an effect." That wasn't very helpful. Then Ben Friedman has some regressions and he found some big effects but they were not very robust.
Gagnon: Then Jeff Frankel interestingly in a paper in which he tried to argue for efficient markets and said, "These effects are small," actually got some interesting results. They were small but they were non‑zero and they were I thought more credible. Based on them, I said, "Well, if you bought $1 trillion you might get ‑‑ I forget ‑‑ like 20 or 30 basis points, which didn't seem like much for $1 trillion. Maybe, I think I said more than that. That's right. Because I went back and looked and I think I was still a bit optimistic even though I told you a real low number. We had nothing to go on and...
Beckworth: What about Japan?
Gagnon: We hoped it would work. Japan had done QE, and they said they were buying long term bonds, JGBs, including 10‑year JGBs. I was puzzled. There was some disagreement about how much effect it had. Some people said it had an effect, some people said it didn't, and it was a bit puzzling. Later, after this, a year or two after we did QE1, I found an interesting paper that said, "Ah, we looked at what bonds they did buy back in the earlier QE in Japan, back in the early 2000s." They only bought 10‑year JGBs that were about to mature, so they had one year left to maturity, Basically, they were buying T‑bills.
Gagnon: The thing is, if you are buying short‑term government liabilities, and issuing short‑term government liabilities, or money that is a very close substitute to that, you're not going to have much effect. You need to buy something. You need to take that money and buy something that's different from money that has a higher rate of return than money.
Beckworth: Otherwise, it's a simple asset swap.
Gagnon: Yeah, otherwise it's too close to a perfect substitute. Afterwards, we learned that was the problem, and then the new QE that Japan did in 2013 was quite different. The final point I would make is, if you look at the FOMC statements when they were doing QE1, they were very reluctant to claim that this would push down bond yields. They said it would make your credit more available, and ease credit conditions.
Beckworth: Right, credit easy.
Gagnon: Yeah, but they never actually said, "We are going to lower the term period. We are going to lower long term bond yields." Now they do, but they didn't in the past. There weren't events, and understandably so. It's hard to exaggerate how seat‑of‑the‑pants this was. We didn't know how to calibrate this.
Beckworth: Let's talk about some of the channels through which QE is supposed to work. There's the portfolio channel, and then there's the signaling channel. I think your work is focused more on the portfolio channel. I'm I correct in that assumption?
Gagnon: Focus is probably not the right word because some people would say what we're finding is the signaling channel. We believe it is about portfolio channel, but other people have criticized us and said, "Well, how do you know it's not the signaling channel?"
Gagnon: In particular, Glenn Rudebusch, and a couple co‑authors of his, Bauer and Christensen in separate papers, have argued that at least in the US, there's at least as much signaling as portfolio, although they find in the UK it is more portfolio.
Gagnon: I think that the differences in the US/UK results they have are not fully believable, and I'm not sure I fully buy the results. I'm not a good enough economist to really parse, explain exactly. We acknowledge signaling as some of it.
Beckworth: Can you explain the two channels to our listeners?
The Portfolio vs. Signaling Channels of QE
Gagnon: Sorry. I'm going to all these details that no one knows. When the Fed is pushing down, when it hits zero bound on short rates, and it wants to hit down long rates, it can talk about, one thing it can do is give forward guidance where it says, we plan to hold the short rate low for longer, and that will push down long rates, because long rates include expectations of future short rates.
Gagnon: That, everyone thinks has worked to some extent. Some people think that buying long term bonds increases that affect, because by putting your money where your mouth is, in other words you've committed yourself, you've bought this long term asset. If you don't hold the rates for long, as you say, you will lose money on it because the price will fall if rates go up.
Gagnon: It's a commitment device, if you believe the central bank actually cares about its profits, which is another issue. Anyway, Rudebusch and others say that is a big part of the channel by which when the Fed bought long term bonds, it pushed down the long term bond rate, because it enhanced the credibility over keeping rates lower, for longer, going out.
Beckworth: This is a signaling channel?
Gagnon: This is called a signaling channel of portfolio purchases. It's like forward guidance, but it's forward guidance through actions, as opposed to just talk. It's actually buying things that give you an incentive to carry through on your talk. I think there's undoubtedly some of that, but I've always thought that it's not very credible for central banks to make commitments about where their policy rates are going to be more than two or three years ahead.
Gagnon: We just can't forecast that far ahead. There'll be turnover on the board. The chairman's term never lasts more than four years, so you can't commit future people to do things. It's even hard enough to commit yourself to doing things. Going beyond two or three years, I've never thought is credible. I always thought that when forward rates beyond two or three years go down because of purchases, I just think, prima facie, that must be evidence of a different channel, which is a portfolio channel. By the way, there is some evidence which the Rudebusch et al people cannot explain.
Gagnon: When we did things like Operation Twist, where we actually sold two year Treasuries and bought 10‑year Treasuries, that, according to the signaling channel, should've lowered all rates, everywhere, across the board, all the way out. It didn't. It raised two year rates when it lowered 10‑year rates, and that was because they were selling two year bonds when they were buying 10‑year bonds.
Gagnon: A signaling channel cannot explain that, so I point to that. I think that I'm pretty comfortable that the portfolio channel is part of it. That's just saying that if you make 10‑year bonds scarce, and there are some people who want to have them, for whatever reason, you will push their yield down. That's not possible in these really efficient markets views of how financial markets work, but if you relax those assumptions, then yes.
Beckworth: The basic idea is that there's these investors out there, you're taking out these long‑term treasuries from their portfolios, and they're going to rebalance their portfolios. They've got to get back and rebalance them on a risk return basis. You got rid of the Treasury, so now they need to go buy something similar, or close to it, so they buy some safe corporates.
Beckworth: Then the people who sold the corporate bonds have to go by their own security, and you have this effect that spreads. There's a massive rebalancing throughout the economy, which eventually leads to more aggregate demand growth, higher inflation. Is that the idea?
Gagnon: That's the idea.
Beckworth: What I like about this, when I first saw this, and Bernanke mentioned in some of his speeches, it reminded me a lot of the early Monetarists. Milton Friedman, and MJ Schwartz had a paper in 1963. I don't remember the exact quote, but the quote is very similar to what you just described. Also, some other Monetarists had similar thoughts, as well.
Beckworth: In some ways, I thought you guys were resurrecting some of the old Monetarist thinking on what would happen. Now, you did a study, it's well‑known, it's highly cited. Do you know how many cites you've gotten off of that paper?
Gagnon: I don't know, but it's pretty high.
Beckworth: It's a career maker, if you were going for a 10‑year high or something, if you had something like that. It's a great paper. Tell us what you found in that paper, in terms of the effect. Also, maybe the other literature that's fallen. What is the consensus view of how many billions of dollars, how much yield do you get out of it?
Gagnon: When I left the Fed in late 2009, and I'm thinking that there would never be any more QE, that they were talking about exit, and the forecast is looking OK. I thought, "I'll leave on a high note." I've already been offered a job by Peterson's, too, and I put them off for a while because of all of the stuff that was going on, and I left.
Gagnon: I was immediately asked by people at the Federal Reserve Bank of New York, with whom I had worked in managing the first QE. They were the ones who actually did the buying, in New York, if I wanted to work with them on doing research on what the effects were. I thought that was a great idea, and I'm so glad that they approached me on that, and that led to that paper.
Gagnon: What we did, and it was the first paper that came out that actually measured QE, but it was quickly followed by others. We had two things we did. One was, we looked on days when the Fed announced increases or changes in how much long‑term assets they might buy, and there were several days in which they either announced new news, or changed it, or disappointed markets because they didn't make any further decisions.
Gagnon: We looked at all those days, and looked at changes in bond yields in narrow windows, like basically from the day before, to the day after, something like that, on the change, on the day of bond yields. On these specific days where we think most of the news was about quantitative easing, and we added up the changes, and came up with an estimate of how much the total QE1 package lowered bond yields.
Gagnon: We got about 100 basis points for QE1, which was not quite a $2 trillion thing. If you add it all up, it was a $1.75 trillion. That was the first thing. We did another way of looking at it, which was just ignoring QE, the whole QE period. All of this should work in response to, say, Treasury changing the maturity of its issuance.
Gagnon: We looked at what the government's net supply of long‑term bonds to the private market is over time, and as a share of GDP, and in a model of explaining long‑term bond yields, adding that in with other factors, and seeing how much that would move long‑term bond yields.
Gagnon: Then we can calibrate that coefficient by the size of what QE1 was, and say how much it would've done, and we got about 50 basis points, which is about half of what the other part of our paper did. The interesting thing we've since come to believe is, you may hear people say, "Ah, QE1 worked. It was a market panic, it was great, but after that, there's diminishing returns, and there was no effect."
Gagnon: Well, yes and no. Yes, we believe that the larger effects we got in the event study reflected the panic of the time, and the newness of the policy, and that's all good, but we think that the smaller effect, the 50 basis point effect, which is about half of what we got, is not diminishing.
Gagnon: That is the non‑panic period effect, and we think that there's no evidence anywhere in the world, or in the US since, that that's falling. Yes, there's a panic extra effect, but it doesn't mean that there's no effect now. It just means that we're down to the more normal time effect, which is pretty steady, and not diminishing.
Beckworth: That second part was a regression over a time period that went way before QE, correct?
Beckworth: That's the basis for saying, "Hey, this is not just a stressful financial crisis period."
Gagnon: We didn't even include QE. It stopped before QE happened, so it really wasn't contaminated by it.
Beckworth: It's pre‑crisis?
Beckworth: That's interesting. Let me give you some of the pushback that's often seen.
Gagnon: Do you want to? I thought you were going to follow up with what people find for other countries.
Gagnon: Just this year, I put out a policy brief on the Peterson website, where I did not even that thorough, but a moderately thorough review of studies that have come out. I did a thorough review about three years ago, and then I updated it, but probably not up to the standards of general economic literature.
Gagnon: Anyway, I got over three dozen, almost four dozen studies. Actually, more than four dozen studies, but about 30‑some studies that I can compare. I can find, people do very different things, and sometimes it's hard to compare effects across studies because there's actually very different things being measured.
Gagnon: I've got about three dozen studies that you can actually compare percent of GDP, long term bonds bought versus change in bond yield from that. You can compare, most of them are for the US, but a number are for the UK, for the Euro area, for Japan, for Sweden. Everybody finds an effect. Everybody. There is a tendency for these event studies to actually be bigger than the regression effects, especially at the beginning, but the effects differ a bit across countries, but perhaps in sensible ways. They're pretty close for the US, UK, Euro area. They're a little bit larger for Sweden, and a little bit smaller for Japan.
Gagnon: Perhaps that suggests, if you look at the bond markets, Sweden has a smaller bond market, and Japan has a larger bond market. Maybe you should be normalizing by the bond market, not by GDP, and maybe that would give you an almost identical effect across countries, which I think would be really nice. By the way, our original study is right in the middle of the range, of the people, the effects people have looked at.
Beckworth: Many central banks are doing QE now. The European Central Bank is doing it, we'll talk about that a little bit later, as well as the Bank of Japan. We'll come back to that. One of the areas that QE gets confusing, I think for folks, is that the intention is to lower these long term yields, right?
Beckworth: But going on at the same time, we've had this weakening of the economy, the global economy. Long‑term yields on safe bonds have been going down anyway. For example, just yesterday, we learned that the 10‑year government bond yield in Germany went negative, which is striking.
Beckworth: I know the ECB is doing QE, but if you look at this trajectory, you look at safe yields, US, Germany, UK, they've been going down since 2008, before QE, after QE. Sometimes it's hard for people that, just the eyeball test, to step back and disentangle what's driving what. Yes, QE helped, but what's going on? Can you speak to that? Is it the safe asset shortage problem? What's driving this long‑term decline in yields beyond QE?
Explaining the Long Term Decline in Treasury Yields
Gagnon: There's several pieces. As a part of background, I would say that even before the crisis, people had noticed that real long‑term bond yields had been on a downward trend for 30 years. They peaked in the early '80s, and then they fell, and they fell gradually, but to a level that's even lower than they had been in the '60s.
Gagnon: The '70s was an unusual period where inflation inflated away bond yields, but we don't think that people really, it may reflect some poor expectations of inflation. Basically, bond yields have been coming down for 30 years from a peak in the early '80s, and they are now at record low levels.
Gagnon: That was a trend that started before the crisis, and I think has continued. I think that will get us into a whole bunch of issues about secular stagnation, and what's going on, but there are three observable things which I think our priority would lead you to expect lower interest rates. One is demographics. Fewer working age people, the growth of working age population is way down around the world. That means less need for factories, and houses for family formation, and jobs for creating capital for workers to work with. There's less need to invest.
Gagnon: Also, as people approach and enter early retirement, they save a lot. Once they get deep into retirement, they should dissave, but the facts are that people don't really dissave as much as you would expect in the early stages of retirement, so there is that bulge of people entering retirement that still means a lot of saving.
Gagnon: Then productivity growth has been weak, lately. We don't know if that's a weak trend or not, but the data showed that low productivity growth would tend to reduce demand for investment, less new things to buy. Finally, my point, which doesn't get mentioned as much, but I think is very important, is that emerging markets used to borrow from the advanced economies to develop, and now they lend.
Gagnon: They actually stopped doing that, and they have a different strategy, which is basically lending to hold down their currencies, to get exports or to have trade surpluses. Dani Rodrik's written about this, other people have. Practically, it seems to have worked. As economists, there's lots of criticism we can make of it, but people say, "Look at China."
Gagnon: Anyway, the bottom line is that governments in emerging markets used to borrow, and now they lend to advanced economies. That's a big shift, because it's actually gone from a percent of world GDP borrowing to one or two percent of GDP lending from emerging markets. That holds down interest rates in advanced economies, as well. They're all three reasons. There are empirically verifiable reasons why you'd expect lower long‑term yields.
Beckworth: Let me throw maybe a fourth reason in there. If you look at, at least nominal long‑term yields, they were going down, but the trend accelerates after 2008. One story might be that, we had the great recession, and I think the incredible shock that it was, risk premiums went up, people became incredibly risk averse.
Beckworth: We never have fully healed from that crisis. We've had a spate of subsequent bad news shocks. We're just getting out of the Great Recession, and the Eurozone crisis emerges. That thing never really has been resolved. It seems to every few years reappear.
Beckworth: We've got panics in China. We have concerns about political uncertainty in certain parts of the world. It's as if we've had this massive wound that just hasn't had a chance to heal in the economy. The German bond yield, for example, going below negative, all the newspapers are saying, "This is because of Brexit." In other words, you have all of these uncertainty shocks that have continued unabated, and the economy just hasn't had time to heal. Is that a possible story?
Gagnon: Yes, but I would overlay it on a story. I think you'll like this, and maybe it'll lead to the Neo‑Fisherian order. To me, I look at the US and the UK versus Europe and Japan, and what it tells me is that actually, there's a paradox here that if you are more aggressive in easing policy and pushing down rates of return across the economy faster in a crisis, you will give recovery, and you will bounce back sooner. You'll go back to normal sooner.
Gagnon: You actually will have higher rates after a time, because you succeeded in easing monetary policy. This gets to the whole people's view of, how can you tell if monetary policy has been easing or not? Over time, ultimately, to ease the monetary policy should lead to higher rates.
Gagnon: This is why people get confused about QE, because they look at US QE. Initially, we saw lowered rates, but now rates are higher in the US than Europe. What happened? Well, it worked. It got us on a better inflationary path, trajectory than Europe did.
Gagnon: Now, Europe obviously had other issues, but let's not forget, this crisis initially started in the US. We had the big housing bubble, Europe had him small, peripheral bubbles, but nothing like the US. Everyone in Europe was convinced in 2009 that this was US‑centered recession, and they wouldn't be that much affected by it.
Gagnon: Yes, it turned out that there are weaknesses in the Euro area that caused them problems, but it still is true that the big, bad loan issues were in the US, the big housing issues were biggest and worst in the US, and yet it's odd that the US didn't suffer more than Europe, and it suffered a lot less. Why is that? I would argue that the strong response of monitoring fiscal policy in the US, stronger than Europe, led to a better recovery, and then we could get back to higher rates, sooner.
Beckworth: I'm very sympathetic to that story. I have, in fact, a new working paper coming out in the Mercatus Center on the mistakes the ECB made. They actually raised interest rates in 2008, and twice in 2011, and I think you can trace that a lot to some of the other problems that they had.
Beckworth: The Eurozone, in my view, is a flawed currency to begin with. It was bound, in my view, for some kind of problem, but I think the ECB policy making up until Mario Draghi was highly flawed, contractionary. The sovereign debt crisis, austerity, all those things got their teeth from the mistakes the ECB made. I'm very sympathetic to that point.
Beckworth: I like the point you made earlier. If QE does work, and if it's applied properly, you should eventually see higher long‑term yields, unless the economy recovers, then inflation expectations go up.
Gagnon: The thing is, that's a Neo‑Fisherian view, but it's just where I disagree with those guys is on the timing, and the causality.
Beckworth: Absolutely. I guess what makes it interesting, and also challenging as a policy maker is, you have to sell this as, "We're going to lower rates." Then, if it works, public rates go up, and then you get criticized, "What are you guys doing? It doesn't seem to be getting the effect you want."
Beckworth: I have been supportive of QE policies, and more supportive, actually, as they went along, because I thought QE2's time calendar was a constraint on how effective it could be, and QE3 was much more data dependent, condition driven. I think that's the way we should do it.
Beckworth: In my old age, I joke, but with a little bit of time, I've begun to wonder if the effectiveness of QE was constrained by one thing. That is, I'll say the Fed’s, but ultimately it is the body politics commitment to low‑inflation. If we look at mid‑2009, when the recovery actually begins, inflation, core PCE inflation is averaged to one‑and‑a‑half percent.
Beckworth: Other measures have been above that, but the preferred measure of the Federal Reserve is averaged one‑and‑a‑half percent, its target is two percent, and I know the target doesn't begin until 2012, but implicitly, they were targeting two percent all along.
Beckworth: If Bernanke or Yellen had tried to allow inflation to go up a little bit, their heads would have rolled on Capitol Hill. Everyone now expects the norm is one‑and‑a‑half percent inflation. It's Congress, it's people I know at home, it's the body politic. I don't think Bernanke, even if you wanted to generate temporarily higher inflation could do that.
Beckworth: There's the models have estimated if you wanted to have rapid catch‑up growth in 2009, 2010, you would've had to tolerate a little bit faster than normal inflation, maybe three, four percent, until you got back up to where you were pre‑crisis, and you could have grown normally, and had two percent inflation.
Beckworth: My point is this, to cut to the end of this long‑winded discussion. QE was tried, but ultimately, wasn't it muzzled by concerns about inflation? In other words, if QE starts to get the economy going, the Fed sees inflation getting near two percent, going above that, they pull back. Is that a fair critique, or am I bridging too much into it?
Was QE’s Effectiveness Impacted by Inflation?
Gagnon: Certainly some people criticized, even back in QE1, that this would be highly inflationary. There were articles in "The Wall Street Journal" saying that. I don't think the Fed paid a lot of attention to those viewpoints. I think the Fed was more concerned about, these are big numbers, this is unprecedented.
Gagnon: There's just a natural tendency if something is new and scary, not to do too much of it. We sort of knew that if you took our best estimates literally, what this would do, and how it would work, that you needed to do even more than they did, but they were doing such big numbers that the case for doing even more was just not persuasive to them.
Gagnon: I think, with hindsight, I would hope, because they ended up being pushed to do more. If they had done everything they did in the end, all three or four rounds of QE earlier on, we would've had a better result, and maybe some of these criticisms wouldn't be around, because it would have been seen as more effective.
Gagnon: I don't think inflation, fears of excess, I don't sense fear of excess inflation was ever the main thing. It's more like, "Jeez, our balance sheet's so big. Could there be some unintended consequences? Are we going to cause a lot of risky behavior as you push people out of bonds?" Just a sense that we don't know what this could mean, or maybe it'll be more powerful than we think. Just that kind of thing.
Beckworth: There's the uncertainty of trying this new tool of monetary policy.
Beckworth: Related to that critique, there was work done by Paul Krugman in his 1998 Brookings paper, Michael Woodford and Eggertsson had a similar Brookings paper in 2003, I believe. They argued that QE would be, in effect, irrelevant to the result. Ultimately, it's temporary.
Beckworth: Going back to Japan's 2001 and 2006 example, they said, "Look, the market knows this is ultimately temporary. That's why inflation has not taken off." If they thought it was permanent, it would've been huge. It would've been 1970s style inflation. The central banks are so credible with their low‑inflation targets, that no one believes this is going to be permanent.
Beckworth: Even Bernanke, Yellen and their exit strategy have all said, "At some point, we want our balance sheet to go back down to a pre‑crisis value, whatever that would be." Do you think the Woodford, Krugman critique that QE ultimately was not a permanent injection made it less effective than it could have otherwise been?
Gagnon: Yes, in theory that's certainly right. I don't know how you could credibly make it permanent. I fear that would not be a good way to run the economy, because this gets at the helicopter money idea, right?
Gagnon: You print this money, and you spend it, and you promise that you will never mop it up. You will never pay interest on the reserves it generates, or you will never, it's back to the old type of money that never paid interest in, and you'll never sell off the assets to shrink the money supply, and so you'll just always have a bigger money supply that pays zero interest, and is therefore inflationary.
Gagnon: Yeah, that would be more stimulated. The problem is that in the recession, when people aren't spending, the amount of that that you'd have to do would later on cause a lot of inflation. It's this time problem whereby you would have to promise credibly to generate double digit inflation for a few years, in order to get decent growth in the recession.
Gagnon: You don't know when that inflation would show up. It's kind of random. We don't really understand that process, when people will suddenly switch back to being more normal. This is the problem, money demand is hugely cyclical, and in a way that we don't understand very well. The amount of money you would have to create in the crisis, I think would lead to inflation down the road that would be unacceptably high. Better to do a policy that's less forceful, but that's more under the control of the authorities, is what I think.
Beckworth: What about a level target? Now you can do a price level target, or in my case I would prefer a nominal GDP level target. Wouldn't a level target address some of those concerns? It would constrain long run, nominal stability. It would keep inflation stable in the long run. In the short run, it would give a little bit more flexibility to do those kind of things.
Gagnon: With a level target, you could not do helicopter money because helicopter money implies a credible promise to have access inflation down the road, which would violate your level target.
Beckworth: If you're below the level target though, right?
Gagnon: Yeah, but calibrating that, how do you calibrate how much money is going to give you enough to get you back to the level target without giving you too much later? The problem is, what's enough in the short run is too much in the long run, and if you're promising never to mop it up, then you're virtually guaranteeing that you're going to go off your level path.
Gagnon: I do think that a level target does have a nice property in a recession, which is that if you're undershooting for a while, then that builds up inflation. You've got to get more inflation later. If people see that that's helpful, that is not quite as powerful as helicopter money, but it is something that I think is useful.
Gagnon: I, myself, am struggling with how much of a level versus growth rate target do we want? I like nominal GDP targeting. I guess I'm thinking of something like, not a pure level, but not a pure rate of change per year. Rather, some sort of backward averaging rate of change, or something. I don't know, somewhere between a level and a growth rate. I think that has a nice property.
Beckworth: Let's move on to Japan's QE. We already talked about the early QE, it was ultimately a temporary one, but now they're doing Abenomics. Tell us about Abenomics. How is it different? You've argued, actually, you wrote an article saying that "Mr. Kuroda needs to be more bold." Tell us about what's going on over there.
The State of ‘Abenomics’ in Japan
Gagnon: Japan is such a frustrating country to study. I actually was Japan desk, and worked for a while, for early jobs at the Federal Reserve, and Treasury. I think an interesting case study, you mentioned getting political buy‑in for inflation. Well, Shinzo Abe ran on a platform of raising inflation.
Gagnon: He specifically said, "We want two percent inflation." He actually said that before he was elected. He actually made a platform. It was actually surprising that a politician would literally say, "I want to change the governor of the Bank of Japan. I want to change the Bank of Japan mandate, and I want to get more inflation."
Gagnon: People elected him, and he...four months later was able to install a new governor at the Bank of Japan, who was, to my surprise, incredibly forceful, and did a new kind of QE like we have done, not like they had done earlier. Buying true 10‑year bonds with 10 years of remaining maturity, pushing down the term premium.
Gagnon: Massive quantities, bigger, as a share of their GDP, than the Fed did. If you look at core inflation in Japan, and measures of inflation expectations, they go, they soar. When he comes in, they go way up. Core inflation, properly measured, the one that the Bank of Japan likes, which excludes fresh food and energy, and the consumption tax increase goes from minus about three quarters of a percent, to plus one‑and‑a‑quarter percent.
Gagnon: A 200 basis point rise, in about two years after this happens. Huge success, big, powerful program, very meaningful. The problem is that, just because you need a big shock, and you need a credible commitment and a big policy action, we don't know how big is big enough.
Gagnon: I mean, calibrating this is hard, and the effects seem to be smaller in Japan, because they have a bigger bond market, so you have to do more. Anyway, they got a big response, but they didn't quite get to their two. It seemed to be leveling off late last year, and the market can no expect, they've got to do one more push, one final ramp up of QE to get over the line to get to two.
Gagnon: In January, they announced what they thought would be that push, and it turned out to be a surprise move into negative interest rates. It was a tiny move, it went from plus 01 percent to minus 01 percent. First of all, that's a small thing. They thought it would matter a lot, I don't know why, it's a small move.
Gagnon: Second, there was a huge outcry against it, I a lot of political pushback. Some members of the board even said "We would have rather done more QE, rather than this," so there's some dissents. Not dissents against easier policy, but dissents against negative rates. A lot of people complaining to their local politicians that they don't like negative rates, households and stuff. It just has had a bad public relations thing. Markets now expect that they have no more scope to do negative rates, because it's so politically dynamite.
Gagnon: Suddenly, markets think they're out of ammunition, and expectations of the future inflation are collapsing. It's just a bad situation, and I don't know why they need to move into QE. Here's the thing. With 10‑year bond yields at zero in Japan, there probably is a zero bound on bond yields.
Gagnon: I mean, why would you hold a 10‑year bond in Japan when you could hold banknotes that yield zero in a safe? Safe sales are increasing a bit. You need to buy some other asset now. You need to move on to something else that doesn't have a zero yield. I argue that is equity, and that was a blog post that you mentioned.
Gagnon: I think they should shop on all the market with a huge equity purchase. It's risky, I understand, but it is absolutely critical that they achieve their promised two percent inflation. They put everything on the line about that. If they lose their credibility, it all unravels.
Beckworth: Is the Bank of Japan legally able to purchase equities?
Gagnon: Yes, they are. They do need the government to agree that they need, but I'm sure the government would agree if they said, "This is what we need." Yeah, they can do it. It was just good for them, but the Fed cannot do that.
Beckworth: Yeah, we're constrained by law. It is interesting that Abe was elected on the promise, the commitment to inflation, given that they had been in a deflationary place for almost two decades. I mean, that's striking. On top of that, there's a large, aging population in Japan, which would not benefit from having higher inflation to living on fixed income. It is a remarkable political development, yet it's interesting why they didn't go for more QE as opposed to the politically poisonous, negative rates.
Gagnon: They were surprised at how poisonous it was, so I think that, to be fair, they didn't expect this reaction. I think also, it was odd that they would think such a tiny move would be powerful.
Beckworth: Let's move over to Europe now. What do you think about the ECB's QE program?
The ECB’s QE Program
Gagnon: I think it's good so far. I'm a little more optimistic about Europe. They're closer to their target than Japan to begin with, and they are having effects. Real rates are falling, and bank lending is going up, which is good. It's slower than I would like, but I at least have confidence, it's more plausible that they would get a good result.
Beckworth: What is your take on negative rates in general? The Bank of Japan is doing them, the ECB is doing them, the Swiss National Bank. Is your preferred course to go more QE and avoid negative rates, because of the political fallout, or would you try both?
Gagnon: I would try both, but I think the key point I think about negative rates is that, I think we are approaching the limit. There is a lower bound. I mean, you don't want people to start demanding $100 bills, or in the Swiss case, 1,000 franc notes, which are pretty easy to store.
Gagnon: Suddenly, you're not going to hit. They're not going to see that negative rate anymore, and they're going to be outside the system, and it's going to raise dead weight costs to society from doing this less efficient way of running the economy, and insulating them from your actions.
Gagnon: I think you don't want to push the economy that way. I used to think the negative rates were a good thing, and they could go further. I still think they're a good thing, but I have trouble seeing them going much below minus one, and so I think it's risky. I really doubt they can go below minus two. Maybe they can't even go much below minus one.
Gagnon: If you push too far, once people switch into cash, it's sort of like one of those tipping things, where then they've made that investment, and then they've switched their behavior, and then getting them back will be difficult. You have to reverse interest rates just at a time when the economy might need more stimulus, and so the optics of that would be bad. Fine, go to minus one, and maybe that's even the first thing you should do, but don't think that you can go much further below zero, I don't think.
Beckworth: Given the current institutional setup that we have in most advanced economies, it's not a practical course much beyond what you think, minus one. Miles Kimball has argued for a novel approach where you allow the exchange rate between deposit electronic money and cash to fluctuate, but I think some of the push‑back he's gotten is just the difficulties of doing that. Any thoughts on Miles Kimball's proposal?
Gagnon: It's a beautiful proposal for an economist to look at, but the idea that society would accept that is just, I cannot imagine.
Beckworth: Really what we're speaking to here is money illusion, in a sense. It's easier to do QE, have some high inflation. You're still going to effectively get that lower, real rate. People aren't getting upset because they don't see the nominal. It's interesting, as an economist, to think this is the money illusion at work.
Gagnon: This is all pushing me towards, we need more inflation. We probably need three or four percent inflation. We used to think two was good, but I looked at some of the papers that justified a target of two, and they all say that the zero bound's not going to be a problem. We now know that's not true, so therefore, if two was good before, and now we know something new, it has to be higher.
Beckworth: It's interesting. Caballero and a couple of his co‑authors have a paper recently out where they make the case that the zero lower bound trap is contagious. That you get to Germany, you eventually hit that minus one percent, investors will go to the next best safe asset.
Beckworth: What you'll see is this chain reaction where all long‑term safe yields will go down together. We see this happening. At some point, it may be that the central bank's hands are forced. They're going to have to do more QE. All right, let's move to, I guess ‑‑ and the minutes we have left ‑‑ to the global economy as a whole.
Beckworth: The IMF lowered its forecast of the global economy. Oil prices have been falling sharply since mid‑2014. There's financial stress earlier this year. You and your colleagues at the Peterson Institute came out with a book with various chapters. You wrote one that says that the global economy is not as bad as the pessimists are making it out to be. Explain that.
Outlook for the Global Economy & Brexit
Gagnon: I think China is not. I think people are really focused on China a lot, and worried about the meltdown in China, and how that would affect the world. One option is they'd have a big depreciation, and they'd have to basically rely on export for that growth again, and the rest of the world couldn't take it, because they're so much bigger than it used to be, and they have to drag the whole world down.
Gagnon: I don't think that's going to happen, because I think the Chinese continue to have more levers of control over their economy than we normally think of in advanced economies. That they are determined not to let that happen. They have massive fiscal space, and they own lots of, not only do they have very low debt, but they actually own lots of companies that either are or could be profitable, that are valuable.
Gagnon: Obviously, some are not, steel and coal, but that's not everything by any means. They have tons of physical capacity to bail out some bad loans, and things like that, that could get them into trouble. They are transitioning, I think slowly, in the right direction. I wish it was faster, but I ultimately believe that they will not collapse.
Gagnon: I think that's a big part of it. I think a lot of emerging market commodity exporters have borrowed a little bit too much in dollars, and have been a little bit too optimistic about commodity prices. They're having to service debts that are a little bit burdensome. I think most of those effects have been seen. If anything, there might be positive news going forward.
Gagnon: Brazil has had a horrible recession, but there's some hope that the new government is actually taking this opportunity to do some major reforms, if they can get them through Parliament. There are some good people that are talking about some good ideas which, if they happen, would be really positive for Brazil. India has been doing well. Indonesia is doing well. There's bright spots, and there's hope, there's upside risk. In Europe and Japan...We need the Bank of Japan to step up to the plate, now. I hope they do, but they didn't today. They just released a statement today which was disappointing. That remains a worry for me.
Gagnon: I cautiously optimistic about Europe. I think Mario Draghi and ECB are in their own slow way committed to doing the right thing. Not as fast as I would like. I don't see collapsed. I just see a bit more sluggish growth and melting food than I would like, but I just don't see collapses everywhere. It's hard to have deflation or re‑spirals downward. Money illusion and downward price would really help in a way, keeps us from just collapsing.
Beckworth: One of my concerns has been the strong dollar. It's gone up over 20 percent. It's come down a bit but since mid‑2014 it rapidly shut up. It has to do with the policy divergence between what the Fed is doing, what the ECB is doing.
Beckworth: My concern and maybe not that it doesn't pretend collapse but definitely a drag on global economy is the strength of the dollars. This is a rapid climb. The Chinese currency is loosely tied to that. It also went up a lot, another tinkering on the margins with their currency. Do you see that us a big problem that it's through the dollar?
Gagnon: I see it as a problem. It's surprising in a world in which US actually has a better recovery because we had better policy earlier on then people look to us to invest in. It's normal that that should put up repairs in the dollar. What's interesting is that that's had a noticeable effect on US economy already and has helped the Fed back. There has been an interesting dynamic more than most recoveries that has kept us at low rates.
Gagnon: I think the right thing for the Feds to do. I've actually become convinced...I thought of the left off was defensible last December and I guess it wasn't a big deal either way. I'm increasingly thinking that partly because of the strong dollars you say that we should be very reluctant to tighten further because we're getting a tightening corresponding dollar.
Gagnon: We need to keep the US economy on track and get back to two percent inflation in our objectives. All the more importantly, but given the weakness of the rest of the world, if everyone goes down and we go down with them and it gets even harder to get out. The policies that will be needed to get out get even more challenging.
Beckworth: All right. One last question. Is Brexit a big deal, if it does happen?
Gagnon: Yeah, it's a pretty big deal, mostly for Europe and UK but it will have negative spill over the rest of us. Those currencies will probably weaken the dollar, will rise more, and investment will drop in the UK. Which means less further bad importance for our exports. I don't see any good of...It's shocking to me that this is such a real possibility.
Beckworth: They are heading a poll, last I checked with the lead group. To be clear to our listeners who don't know. BREXIT is Great Britain leaving the European Union. It is an interesting development. This is the second half of 2016, in many ways is fascinating time to be alive, maybe a little scary for some to see what happens politically in many places. Our guest today has been Joe Gagnon. Joe, thank you for being on the show.
Gagnon: You're welcome. Thanks David.