- | Monetary Policy Monetary Policy
- | Mercatus Original Podcasts Mercatus Original Podcasts
- | Macro Musings Macro Musings
John Taylor on the Taylor Rule, the 2008 Crisis, and Fed Reform
John Taylor is the Mary and Robert Raymond Professor of Economics at Stanford University and the George P. Shultz Senior Fellow at the Stanford University's Hoover Institution. John Taylor's research has had a major impact on economic theory and policy. He has served a number of times in the President's Council of Economic Advisors, as well as the Under Secretary of Treasury for International Affairs. John Taylor joins host David Beckworth to discuss Taylor's famous monetary rule for central banks in setting interest rates in response to changes in inflation and output. They discuss how Taylor discovered the rule and how it has performed over time. Taylor also shares his thoughts for improving current Federal Reserve policy.
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: John, welcome to the show.
John Taylor: Good to be here, David. Thank you.
Beckworth: Thanks for participating. You're probably best known for your famous Taylor Rule. Most students at some point in their career if they're studying in economics, people who follow the markets, are familiar with the Taylor Rule. Could we begin by you explaining to our audience, what is the Taylor rule?
Discussing the ‘Taylor Rule’
Taylor: Sure. It's a guideline or a procedure for central banks to follow, to determine what their interest rate settings should be. It's guided by just a couple of factors in its simplest version, the inflation rate and the state of the economy compared to its potential. Most importantly, I think it says by how much the interest rate should change, when those two factors move.
That quantitative aspect I think, is what's made it useful to people. It's a way to explain monetary policy to new central bankers, it's a way to compare policy over time, and it's turned out to be more useful than I could have imagined at the beginning.
Beckworth: You're saying it has some key components. It has a measure where the central bank would respond to changes in inflation that are outside of the central bank's target, as well as changes in the slack of the economy or the output gap. Is that right?
Beckworth: One question I've always had about the Taylor rule as I've studied it and as I've seen it used in practices is how did you come to the realization of it? Was it a eureka moment? Did you slowly put the pieces together? I know it was published in 1993 in a Carnegie‑Rochester paper. How did you actually come together and make the Taylor rule?
Taylor: It's not a eureka moment. It's a long process of...I've actually always been interested in policy rules for monetary policy. It goes back to when I was an undergraduate. I did some work on finding different rules that the central banks could use. I think it's a development that came out of these new stochastic models of the economy.
I studied it further in graduate school, and been interested in it for a long time. The models got better I think, the ability to analyze policy rules within models got better. Going into the period before you just mentioned in the 1980s, we built I think pretty reasonable models to try out different policy rules.
We didn't know exactly what it would look like. There was an emphasis on money growth rules at the beginning, or other ways to determine it. I think through that process of trying out different rules in models and then thinking about how they'd work, I came to this general thinking about it.
It became pretty clear by the late '80s I'd say, or maybe the early '80s, that the rules were quite complicated that were coming out of the models, and they really wouldn't work in practice. There's too many variables, too many things. Then the effort was to simplify it.
I think it really involved judgment as well as the models, looking at the models to see what variables didn't matter much, what was more important, and through that process came down to the simple one. It was designed to be simple, the simple one that people now call the Taylor Rule.
Beckworth: Were you surprised at how successful it's been, not only in fitting the data during the great moderation period, as well as, just its widespread use?
Taylor: Yes. The original publications I mentioned, actually I delivered it at a conference in Pittsburgh, it was an idea that I thought could be useful to people. I made the case in a very low‑key way that this is something central bankers should think about.
I had, obviously, no expectation that it would become as useful, or talked about as it has. I think though the main reason that happened early on, is some financial people began to see that it was a good description of the effect going forward. Some of the moves subsequent to the publication of the rule were predicted by it pretty well.
That got people in financial markets quite interested in it. I think that probably brought more interest as well, and the fact that it was designed for the United States. It seemed to work pretty well for other countries that generated interest. No, I couldn't possibly have expected the kind of interest that came about.
Beckworth: Now you have the Taylor Rule, you also have the Taylor Curve, and the Taylor Principle named after you. You must be doing something right with your magic touch with economics.
The original model was a 1993 paper, and in that you put a weight on the, we'll call it "the inflation gap." That's the difference between the desired inflation and actual inflation. As you mentioned earlier, "If inflation is going at a pace where you don't want it, do you respond?"
There's a way that determines how much you do, and the same thing for the output gap. In the original Taylor Rule you had 0.5 as the weight. Is that correct?
Taylor: Yes. Actually, if you think about the reaction of the interest rate, the inflation rate per se is 1.5, but the reaction to the output is 0.5, as you say.
Beckworth: It fit the data remarkably well when you published that paper. It did all the way up until the early 2000s. We'll get to that in a bit, the housing boom period. I think it said something about what the policy wasn't doing.
There's been some other versions that have come out subsequent to that, and I would just like to hear your thoughts.
For example, Janet Yellen, in a 2012 speech, mentioned your, "Other Taylor Rule." I know you disagree. You replied to her. But she pointed to a 1999 version of the Taylor Rule. How is that different than the original one? You say it's not your rule. It's just a different exercise using a different form.
Taylor: That was a paper I wrote in the late '90s. It was a historical paper to see how different rules would work. One was what people call the Taylor Rule. The other was a rule that had been a modified Taylor Rule that was used at the Federal Reserve. I wanted to compare those two and run them through.
The other one that was used at the Fed had a higher coefficient on output. It was more reactive to the upturns and the downturns in the economy.
You're right. In 2012, Janet Yellen indicated that was a preferable rule. I'm not sure she really used the word "preferable," but considered it. It's just more reactive to the downturns. In a deep downturn you cut the interest rates by more.
There's lots of debate over the years about how reactive it should be. I think the value, if it gets too big, then you've got too much volatility in the interest rate. It's very hard to measure potential GDP, there's uncertainty in that. That calls for a smaller coefficient.
I think the philosophy is similar, but this is not something that we can pinpoint exactly, but too much of a reaction I think leads to problems.
Beckworth: OK. What about the Glenn Rudebusch Rule, from San Francisco? He's taken it a step beyond that and has used the unemployment rate, is that correct, in a Taylor Rule?
Taylor: Quite a few people have noted the close connection between the output gap, or the deviation between low GDP and potential versus the unemployment rate. It's not perfect, and it shifts over time, but that means that you could substitute that in.
I don't think it's a major difference. I think the focus on output as a whole is attractive to me because it's a broader measure and easier. There are people who have substituted output with the unemployment rate.
Beckworth: Let's go back to your original 1993 Taylor Rule, and let's assume that there's no output gap. Let's assume inflation is right on target. What is left in the rule? Isn't it just that intercept term that the natural federal funds rate, is that what would fall out of the model?
Taylor: Yeah. That's part of the model. Assume, back when it was two percent real federal funds rate and equilibrium, so it's an inflation target which we also assume to be two percent. That's actually something that's been used since then by many central banks, including the Fed.
The 2+2, the inflationary rate plus the real interest rate, gives you an equilibrium federal funds rate of four percent, and it's in kind of a steady state. That is the assumptions that were made at the time. I think they're still pretty good, but that's a long time ago.
Beckworth: I think it's interesting that if everything is going according to plan, the business cycle is no longer there, inflation's on target, what you're left with is basically this natural market clearing interest rate, the intercept.
Beckworth: Let's take that Taylor Rule and let's talk about some history. I want to begin by talking about the period from the mid '60s up through the early '90s, 1980s, we call, "The great inflation period." I know you've done work on that.
For our listeners, this was a time when inflation expectations become unanchored. You actually saw inflation go into double digits. In the early '80s, I think it hit close to 15 percent. Interest rates were going really high during that time. It's a very different environment than what we have today, where we have what seems like a low inflation problem. Back then, it was a big problem.
How does your Taylor Rule help us make sense of that great inflation period?
Explaining the ‘Great Inflation’ period
Taylor: What people did after the Taylor Rule was written down was they went back and looked how it would've performed in this terrible period in the late '60s and '70s. It didn't perform well because policy wasn't even close to it. In a sense that was, I think, a partial validation that there was something to it.
I think the first people who did that were at the San Francisco Fed. They basically said, "Look, there's something here, because in the period when policy was really off, the performance was off, as well." That's what happened then.
Now, you can go back and say, "This is another reason why you need a framework like this," because they didn't have one then. It was very discretionary, stop‑go policy, and it led to poor performance.
Beckworth: After that period with Volker, and after the post‑Volker period, you find that the Fed is following that Taylor Rule much closer, is that correct?
Taylor: Yes. That's another thing which we really discovered afterwards. Again, some people at the San Francisco Fed, Bill Poole at the St. Louis Fed, he was president at the time, noted that close correlation during this great moderation period, where the economy was performing well, and the performance policy was pretty close to that description.
Beckworth: That suggests that at least some part of the great moderation, we can attribute to better monetary policy.
Taylor: I think so. There's great debate about that, but the timing is pretty close and people have written about it. There's other factors, certainly, monetary policy isn't everything. But I think it's a factor, definitely.
Beckworth: I want to look at one person who pushed back against the application of the Taylor Rule in the 1970s, and that's Orphanides. He had a paper which creates tension for me, because it makes some interesting arguments, but there's some areas where I'm not sure I understand.
He went back and took the Taylor Rule and said, "Look, given the information Fed officials knew at the time, they were following the Taylor Rule." The issue, he argues is that they didn't have good information in real time. They didn't have the true potential, or the output gap of the economy. Is that a correct assessment of his work? How do you respond to it?
Responding to Critiques of the ‘Taylor Rule’
Taylor: Orphanides, I think it was a constructive exercise. He was trying to figure out why the Fed was so off of the Taylor Rule in this period, and decided that it was because they had an estimate of the potential GDP, which was way off. Rather than just a discretionary stop‑go policy, if you took that into account, they are actually close to the Taylor Rule.
I debated that, because people at the time, at the Fed, they were certainly making mistakes in retrospect. They were not the kind of people who would be misestimating the state of the economy. Arthur Burns was the Chairman. He was a very good economist, I'm not saying his policy was good, but that doesn't seem too plausible to me as an estimate.
Then you have another part of the Administration. Alan Greenspan was the Chairman of the Council of Economic Advisors at the time, and he's also not the kind of person who would be off on that. It didn't seem, to me, to be plausible as an underlying motivation for policy, but it's an important thing to point out that that estimate can be very uncertain.
Beckworth: I think that's a great point. On the one hand, that's what's interesting about it, to me. It speaks to this knowledge problem. It's hard to know in real time, output gaps. There's certain things that central bankers simply don't know.
On the other hand, I wrestle with it because we know, looking back, you've written about this, others have, there were policies being pursued that simply weren't sustainable. President Johnson pursued the Vietnam War and Great Society. You had Fed officials who believed in cost‑push inflation, or the income policies. Robert Hetzel has written a lot about this.
There were clearly bad policies back then. In my mind, it can't just be a measurement error. There has to be some fault laying at bad decisions made at the Fed.
Taylor: Absolutely. Even the use of the very old fashioned Phillips curve was still there, and there were lots of other things motivating the policy. The general idea, remember that we had wage and price controls at that time. It was a different world.
Beckworth: This is interesting, because I went to an AEA session. For our listeners, that's the American Economic Association, annual meeting's every year, that economists get together. John, you were giving a paper along with your co‑author, and Alan Blinder was there. This was probably, I don't know, five or six years ago.
You were revisiting the question of, "Why did we get high inflation in the '70s?" Alan Blinder has this paper, he argues it was these oil shocks. He revisited that idea and made the case for it. You, and your co‑author, said, "No. It was poor policy."
You get up, you both make your compelling cases, and then a third paper was presented. It was presented by a German economist, and they described what happened in Germany. They showed they did not have the problem that the US had. To me, that clinched it right there.
One party said, "It's supply shocks." The other says, "It's the Fed." Germany was also subject to the same oil shocks, it was a global phenomenon, yet they somehow managed to pull through in a much better fashion. I think international evidence definitely supports the idea that there were some poor policy decisions made during that time.
Taylor: Yes, I remember that session at the meetings. I also thought the timing was important, because the policy seemed to go off before these shocks. In addition to inter‑country comparison, international comparison, you had this really poor timing that was much better.
Plus, you had other shocks later. I think another example is Japan, actually at those times they were much less affected by the shocks.
Beckworth: Along these lines, you recall Ben Bernanke's co‑authored piece where he went back and looked at what caused the problems during this time. He argued it wasn't the supply shock, it was how the Fed responded to the supply shock. What do you think of that argument?
Taylor: I'm not sure I know what specifically he's referring to. My sense is that you don't want to accommodate these supply shocks too much, or at least at all, to some extent. You want to keep your policy steady, and a steady rule.
Keep the same kind of rule, same kind of policy rather than, "There's this shock, we need to increase the money supply by huge amounts because prices have risen." That's the wrong response.
It's not just that the policy is helping to cause the inflation, it's the response to these oil shocks. If it's too accommodative, that's the word that some economists use, you really have got to stick with your...If it's money growth, stick with your money growth, or stick with your interest‑rate rule, and it'll work much better. That's what the models would show at the time.
Beckworth: Let's move to the housing boom period. You made quite a splash in 2007 at the Jackson Hole meetings. The Federal Reserve gets together every year. It's a big deal in central bank circles.
You gave a paper on what was unfolding at the time, the beginning of the crisis. You argued that the Federal Reserve's interest rate policies in the early to mid‑2000s was a key driver of the boom, and, of course, subsequently, the crash we tie to that, as well.
You argued in that paper, as well as in a little book, "Getting Off Track," that the Fed played a huge role in creating the housing boom. Can you share that argument with us?
The Fed’s Role in the Housing Boom
Taylor: Sure. Actually, I was asked to present at the Jackson Hole meetings, where, you're right, all the central bankers come, about the housing boom. I suspected that it was these low rates, and people were talking about it at the time that were part of that.
The low interest rates of the Fed made it easier to get the mortgage at low rates. It created a part of the boom, those extra‑low rates encouraged a lot of the excesses, what we call, "search for yield."
Then tried to look at that by connecting the deviation from a policy rule of the kind that Bill Poole, he was President of the St. Louis Fed, was looking at, and noted it was quite a deviation, and that historically would explain that movement in housing prices.
I don't think it's the only thing, because I think there were some regulatory lapses at the time. It was also mostly the Fed or the other regulators. But those two things together, I think, were a big factor.
If rates had moved up in the same way that they would have moved up using the procedures of the '80s and '90s, then I think we would have avoided a lot of that excess up and therefore the excess down.
Beckworth: This, again, is a great time when the Taylor Rule was very useful in illustrating this problem, right? The Taylor Rule said, "Hey, you should have tightened sooner, but you didn't." The gap between the Taylor Rule implied rate and the actual rate can explain a lot of the housing boom.
You did a counterfactual model, where you showed that had the Fed set rates differently, you wouldn't have had quite the housing boom.
Taylor: Right. I estimated a model of housing in which the federal funds rate was a factor, and then simulated under the counterfactual that the rate went according to the Taylor Rule. Of course, it was much less of a boom and bust.
Beckworth: Ben Bernanke, former Chairman of the Fed, as well as a well‑known academic economist, he argued that a bigger cause was the savings glut from all this excess savings coming from overseas. You have all these emerging economies, either for precautionary reasons, they were buying up Treasuries, or because they just wanted more safe assets.
He argued that is a key reason for the rates going down. How do you respond to that line of thinking?
Taylor: What I said is that, globally, there was no savings glut. The global savings rate was coming down, so you have to add something else to that explanation. Actually, Alan Greenspan had a different explanation for why my argument was wrong, and that is he couldn't affect the global long‑term rate very much by moving the interest rate.
There are lots of different views that are put forth to show that my analysis isn't right, coming from the Fed. When you look at them, they don't really add up. There's still a lot of discussion about the global savings glut. It's now been brought in as part of the secular stagnation argument. I wrote about it even in that book you referred to and showed the numbers and the charts. It didn't ever add up to me.
Beckworth: I've argued in a paper, and I know you've seen it, because you cited it a few times, that the Fed is this monetary superpower. During this time, even till today, many countries have their currencies either explicitly or implicitly pegged to the dollar, so when the Fed did ease dramatically in the early 2000s, these other countries, as long as they wanted to maintain that peg, they were forced to follow.
Beckworth: For example, if China wanted to maintain its peg, it had to go up and buy up a bunch of dollars. Those dollars were then taken and recycled back into the US. They went and bought Treasuries. They bought the mortgage‑backed securities. I've always thought that at least some of this savings glut, some of this demand for these safe assets in the US was simply recycled US monetary policy.
Taylor: I think that argument is correct. You've made extraordinarily good points explaining it and showing how it's consistent with the data. We have more information over time that central banks tend to do that, tend to follow each other to intervene. That connection is clear.
People at the OECD have shown how other countries have done what you're talking about. The Fed is the most important central bank, and it has huge effects. You do want to take that into account as an explanation, of even the lower rates that we saw at the long end around the world at that time.
Beckworth: Jumping ahead to today, what about now? The Fed is signaling it's going to tighten interest rates. It already has, 25 basis points, and it's signaled it wants to do some more going forward.
Yet the rest of the world looks like it's really weak. There's problems in China. Eurozone is always on the cusp of another crisis. What should the Fed do or what can the Fed do? To some extent, it's these other countries pegging to the dollar. What can the Fed do going forward if it needs to tighten rates now, but these other countries are weak and tied to the dollar?
Taylor: Right now the problem is that it's a very ad hoc set of policies. We're trying to get back to a rule‑based policy.
In this environment, a lot of the spillovers, the impact, what looks to be currency wars to some people, or the large amounts of capital flows that have been noted, I think they're due to the very unusual policies. It's hard to know what the next step is. You can see the debate we've had in the United States.
To me, the answer to this is pretty simple. We get back to a rules‑based policy, like seemed to work in the past quite well.
But not just the US. I think there's a yearning for this, especially in emerging markets. They don't know exactly what the Fed is up to. They complain vehemently about the capital outflows from the US. Now that's going in the other direction, their currencies are depreciating.
You do need to think about this globally, and you do need to think about countries having a strategy which is pretty clear for their interest rates. When we can get back to that, I think a lot of this will be smoother. It'll be operating in a way like it did, for the most part, in the '80s and '90s.
Beckworth: Going back to the early 2000 there in the housing boom period, one of the things that struck my fancy during that time...Full disclosure. I actually worked for John Taylor several layers down at the Treasury.
I remember being in the Treasury. During 2003, 2004, at that time, there was this massive productivity boom going on, in the US, at least. It was in the newspapers. You read about it. You felt it. It's this very positive environment.
During that time, housing had started. It always struck me that if these are legitimate increases in productivity growth, they should have implied a higher natural interest rate, or they should have at least been nudging up what the Federal Reserve was doing.
It's interesting, because, when I was there, we had a meeting. I worked in the Office of Western Hemisphere Affairs. We would organize meetings for John Taylor and other Treasury officials.
One time, John, we organized a meeting with the Brazilian finance minister, and Ben Bernanke came over. I don't know if you remember this. You hadn't even been in the Treasury, so you may not remember this one.
Ben Bernanke came over to give a talk. He was at the Board of Governors at that time, just a regular governor. I made it a point to be the person at the door to shake his hand and walk him to the meeting. I was, like, "I got to meet Ben Bernanke. I've read about him in grad school."
I did, I walked him to the meeting. You guys had your meeting, and I felt like, "Wow, I met Ben Bernanke," real excited. I got my nerve up. I emailed him, and I said, "Hey, Dr. Bernanke, you guys at the Fed, you're keeping rates low, and you're saying you're concerned about low inflation. You're concerned the economy is weak."
I said, "Couldn't the low inflation also be driven by the rapid productivity gains, so we shouldn't be as worried about it? Couldn't this be more of a benign, low‑inflation environment, rather than a weak economy?"
He actually replied to me in an email, I think maybe because I had the Treasury extension. He took me seriously, and he actually replied to me. He said, "You could be right. It could be productivity gains, but there's other data I'm looking to."
He actually acknowledged in an email that, yes, productivity could have been the cause, although he wanted to be cautious and take a different approach. Interestingly, I left that email at Treasury when I left. I wanted to hunt it down, so I filed a Freedom of Information Act and I actually got it. I have it. I want to frame it one day.
Long story short, that idea eventually resulted in a paper I had with George Selgin in the, "Journal of Policy Modeling," I know you've seen. You do a great job arguing the Fed kept rates too low for too long.
What we extend or build upon that argument is why. Why did the Fed do that? Our position is they were lulled into that because of the productivity gains. They felt comfortable keeping policy easy, because they didn't see the high inflation. Does that seem reasonable, that the productivity boom threw them off?
Taylor: Your analysis makes sense. I've read it. It's certainly consistent with this idea the rate was too low, but what it adds is a reason. It's quite possible. You never know exactly why the central bank deviates in this way. It's curious to me that different people at the time have different explanations for it. I think yours is right.
Another explanation is, I had this idea that they had done a good job with Greenspan for a long time, and they thought by just doing this extra low rate for a while, they would even do better. It's kind of the perfect became the enemy of the good. I think that's another factor, too, but I think your analysis of productivity is part of the story.
Beckworth: It's definitely an interesting period, looking back over. Let's move forward in time, from the early 2000s to the time of the crisis, 2007, 2008. We've already talked about your paper at Jackson Hole. Let's go to 2008.
I'm wondering what you think of the work of Robert Hetzel and Scott Sumner. They make this claim that the economy was going to go into a recession. There was no way around it. However, it turned into a great recession, something far worse, because of Fed policies failing to act appropriately in the latter part of 2008.
They say, "Yes, housing was a mess. In fact, housing began to contract in 2006. The housing sector was getting smaller for several years. Financial stress emerges in August 2007. There's the run on the shadow banking system." All of this goes on up to the early to mid‑2008 period.
Even though all these things are happening you still see relatively comparatively stable income and employment. Things aren't crashing. They're not doing great, but they're not crashing. It's only in the second half of 2008 where you see things begin to blow up.
What they think is the Fed failed to be aggressive enough, so from about April 2008 to about October 2008 the Fed didn't do anything, and it signaled it was very, very concerned about inflation. They argue the Fed made things worse. How do you feel about that argument?
Taylor: I think in retrospect if the rates had come down a little faster it would have been better. I think that's the point. It's not like rates weren't coming down. It was just the economy weakened, and you could argue it could come a little faster or faster, whatever the word is you want to use.
I think there's a question about what the Fed could have been looking at to make that decision. There were movements on oil prices, there was the dollar change. But nonetheless when you look back at the time it looks like it would have been better.
I do think, though, really the cliff period later in 2008 had a lot to do with the very unusual actions in the credit markets. I think, this is another thing which I've written about, that the handling of the problems in the banks, first with the investment banks, first with Bear Stearns, but then the on‑again, off‑again bailout policy, really threw the credit markets off.
You really see the crash in the stock market really occurred a week or so after Lehman, and that was at the time where the TARP was being rolled out. There was uncertainty about what it was. I look at those other actions, very unusual, as a factor in that point.
But I think this analysis of what actually was going on with the rates at the time is important. There were some unusual...The credit markets were already showing some strange signs as early as 2007. People have written about that.
I wrote some papers with John Williams, he's now President of the San Francisco Fed. He was visiting at Stanford about what was going on in the money markets, and we could see some stress having to do with the credit selling at the time. The Fed really didn't deal with that at all, it tried to postpone the actions. I think that made it worse when the crash finally came.
Beckworth: I know you recall, I think you've written about this, the time when Bernanke and Hank Paulson went and told Congress, "Look, if you don't do this TARP it's going to blow up by Monday morning." That got out in the press.
I'm sure that didn't help matters, either, just those kind of statements. I don't how else you could have communicated it, but it seemed like those actions themselves were creating more uncertainty.
Taylor: I think so. Those were very difficult times, but in a sense they were saying, "You've got to do this," but then other people could say, "Doing this isn't going to work." The layout of the TARP, the original proposal, two and half pages with very little specificity that was, "People, how is this going to work?"
That whole rollout testifying to Congress by the Secretary of Treasury and the Fed sure, I'll just add that that's really at the time when things were really starting to fall apart. I think you have to add that element of uncertainly, confusion. Is there going to be a bailout? Is there not? What's next? You've got to add that to the great deal of chaos in the markets at the time.
Beckworth: Let's move forward then to the post‑crisis period. Let's move forward to the Feds QE programs and their forward guidance. As you know we had QE1, then QE2, then Operation Twist, and QE3. We still have this bloated Fed balance sheet.
We had forward guidance, which I think was...I think QE was very ad hoc. They made it up as they went along, but forward guidance was even more haphazard in a sense. They kept extending the date on when rates would be held low. Then they went to an Evans rule. Then they dropped it.
I understand their argument that maybe they needed QE1 to put a floor under the economy. What are your thoughts on QE after that? What did QE2 do? What did QE3 do to the economy? Did it help? You mentioned earlier they made international relations tougher. What else?
Evaluating the Fed’s Post-Crisis Measures
Taylor: Just a clarification. I think of the actions on liquidity during the crisis as correct. There were swap lines to central banks so they could make dollar loans. There were loans to the financial institutions in the US, lender of last resort. That created an increase in the balance sheet by a substantial amount at the time. I think it was 800 billion. That used to be a large amount.
That I don't think is QE1. That was just, I think, a reaction that made sense at the time. Those liquidity facilities started to draw off pretty quickly as things stabilized in early 2000.
But then, to me QE1, is the first large‑scale asset purchase program that was announced in late '08 and started moving essentially in '09. Housing markets backed securities and treasuries. That continued I think to QE2 and the other QEs. I don't think those were very effective.
There's a study of the first one and the mortgage‑backed securities, and with a student here, Johannes Stroebel, who's now at NYU, and we couldn't find an impact of those purchases when you controlled for other things.
I've been teaching students for years and years, "The way the term structure of interest rates work you're not going to have much effect on the term structure, certainly in Treasuries." I never thought that those were good or positive.
There's debate about it. The studies that the Fed has put out emphasize these announcement effects, but they don't take into account the things that occur later. I don't think it's possible to prove what the stock market did. I think there's great debate about that.
It's one of the reasons certainly in markets now or at the time of the Taper Tantrum, because people don't know exactly the impact on the stock market. But overall it doesn't seem to me you can argue that those were a positive thing for the economy.
Beckworth: On one hand we can argue that the QE didn't do that much, didn't make that much difference. At the same time you hear other critics of the Fed saying, "Oh, the Fed is subsidizing the federal government. It's buying up all its debt."
If you look at the data the Fed really wasn't. The Fed is the largest single owner, but it holds around 18, 19 percent of the marketable Treasuries, which compared to, say, pre‑2007 is very similar. It sold off a bunch in 2007 and QE2 I believe its share increased.
Do you think some of the concerns about hyperinflation...? It never happened. What are your thoughts on that, all these concerns about the Fed's leading us to a 1970s period again?
Taylor: First of all the concerns about inflation had to do with increases in money growth which never really happened. The money, for the most part, remained in the banks. Second, the economy just remained very weak during this period.
I think that it goes to the fact that partly these polices haven't worked. But the QE and other kind of policies seemed to me always had a two‑sided risk. One was interest rates, also inflation rates and maybe currencies, but the other was just on the economic side, just slow growth or retrenchment. I think we got the second one and not the first one.
Although on the currencies, if you think about currencies there's been lots of movements in currencies. The impact on the dollar has been reversed to a large extent because other central banks matched the QE.
Our QE in the US 2009, '10 and '11 was matched by Japan's, and they undid the impact on the yen and the dollar, and then eventually matched by the ECB and that's created the effect there. The currency effects, you have to take into account all the other currencies as well.
Beckworth: One thing that has been a bit puzzling and you've suggested is maybe the weak economy. But if you look at the core inflation rate, the Fed's favorite measure, the core PCE inflation rate, it's averaged about one and half percent since the crisis.
It's been surprisingly below the target, and I know recently Janet Yellen has attributed it to low oil prices. But after seven years of consistently below two percent inflation I do wonder if this speaks to the Fed not doing the right thing, not doing it right. What are your thoughts why the inflation has been so low?
Why has Inflation Been so Low?
Taylor: To some extent it's the economy being weaker than the Fed thought year after year after year. Again, it's not just monetary policy. There are other policies that I write about, too. It doesn't really have a booming economy, which sometimes puts upward pressure on inflation and wages from time to time. That's a factor.
To me it adds to the evidence that this kind of policy is not very effective. On the other hand I might say that this one and a half percent, one percent inflation has never seemed like a tragedy to me, by any means.
Beckworth: Sure. Right.
Taylor: This two percent inflation target was what I put in the Taylor Rule a long time ago, and at the point it looked like a pretty good thing to do. Inflation was much higher in the past. Now that it's been adopted by the Fed and other central banks I sometimes worry it's really made...That you have to have higher inflation. What's wrong with one percent and one and a half percent?
It's really not distorting. I think that's another factor that we have to think about in the future is, "To what extent is it OK to have movements below of that two percent? How harmful is that?" We really haven't had the downward spirals that people talk about or the deflationary risk that people have been worried about. We haven't had that so I think that needs to be rethought a bit.
Beckworth: Yes. It is interesting. It's been one of the puzzles. I know Paul Krugman has wrestled with this, because according to the CBO's measure of the output gap we've had a long persistent negative output gap. It's only slowly closing, and yet as you mentioned, we still had mild one to one and a half percent inflation. That doesn't seem to fit the basic model.
Let's move to rules and discretion, which we've kind of touched on before, but you've really been speaking to this recently. In 2012, you gave a talk that got published in the "Journal of Money Credit and Banking," and it was titled, "Monetary Policy Rules Work and Discretion Doesn't, The Tale of Two Eras."
Can you tell us about these two eras and the importance of having a rule‑based approach to monetary policy?
Importance of a Rules-Based Monetary Policy
Taylor: Yes, sure. That was a lecture at the Journal of Money Credit and Banking lecture, and I noted that the year I gave it was 30 years after Milton Friedman did.
He basically made the argument that if you looked at history of the US based on his work with Anna Schwartz that you saw this great degree of connection between more rules‑based policy and more discretionary policy, in a sense rules‑based better than discretion.
What I wanted to do in that lecture was effectively update Friedman, and I argued the same thing was true of the subsequent years, because in just the two periods I was referring to then was the late '60s and '70s, clearly high discretionary policy, an emphasis on getting the unemployment rate down and then when it got down inflation picked up, and so you step on the brakes. It's a very chaotic policy.
Then finally the reform, people recognizing it just doesn't work, and Paul Volcker coming in and making really tough decisions to change the policy. Then it became more focused, certainly focused on inflation.
Inflation came down, and ultimately unemployment came down as well, too, and that's this period of the great moderation, and it is definitely a more rule‑like predictable type of monetary policy than we had in the late '60s and '70s. That comparison was pretty clear.
I think what we've learned since then I don't think there's much debate about that point anymore comparing the '70s with the '80s and '90s but now in the more recent period I think it's kind of the same thing, starting to deviate from the policy rules as we talked about earlier in this conversation, in '03, '04, '05. You've indicated some of the reasons for that.
That really pretty much has continued, and it's not been great performance, and I think that's the other part of this lesson.
Beckworth: We need rules not only to make the Fed behave, but to create this certainly so people can plan ahead, better investment plans.
Taylor: Yes. There's a host of reasons to think about a policy being predictable, rule‑like, and you've mentioned some of the important ones. I think in monetary policy there are others. There's this time consistency argument people make, but just the sense of what the policy's going to be is very important.
I might add that the historical paper in the Journal of Money, Credit and Banking there's been quite a bit of follow‑up on that using systematic statistical methods.
David Papell and some of his colleagues at the University of Houston have actually taken those periods, and I think demonstrated, again using statistical discrimination methods, that periods when it was more rule‑like performed better than periods that it was not rule‑like.
There's some work going on here at Stanford that looks at other countries and it's the same phenomenon, so there's something there. It comes out of a theory, but I think it also comes out of the data.
Beckworth: That's a nice segue into the next part I want to talk about, and that is your involvement with the, "Federal Oversight and Modernization Act," the reform act for the Federal Reserve. Can you tell us a little bit about that Act and how it would move us back toward a more rules‑based approach to monetary policy?
Taylor: Several years ago I gave a talk at Cato, and many people had been asking me, "Is there a way to legislate policy rules? We have this evidence that it worked better. Is there a way to legislate it?"
My original work was not about legislating. It was about guiding or thinking about policy, not necessarily the legislation. But I began to think about it, because people kept asking me, "Is there a way to legislate it?"
I went back to look at history of legislation on these issues, and there was in 1977 a requirement that the Fed report its money growth ranges, and that came out of this same kind of poor performance period. People were trying to get an improvement.
I think that was helpful for the Fed for a while, especially during the disinflation, but then that was removed that requirement to report a money growth statistic was removed in the year 2000 and nothing was put in its place.
This did not tell the Fed to follow a particular money growth rule. It just asked the Fed to report what its outlook for money growth was, and the Fed could even decide which aggregates to report.
Based on that experience I came up with the idea, "Maybe the way to handle this is simply to do the same thing with respect to a more modern way to think about policy and that is through these interest rate rules, what the Fed actually does."
That was the idea. If the Fed just describes its strategy or rule, it's its job to determine what it would be, then that would help a lot in terms of communication, in fact in terms of getting back to a rule, because the Fed would have to explain how it deviated.
That idea percolated for a while and we talked about it. Then in Congress people said, "Well, what can we do?" They thought of various things. This is one of the things that they worked on. A little statement like I just made is a much different than the legislation.
A number of people worked on language and tried to get close to that kind of an idea, and that's really what Section Two of the FOMC is. It's a requirement that the Fed report its strategy or rule. There's more detail than what I just said, but that's the basic idea, and I think it makes sense.
It's the Feds job to choose the rule. It's the Fed's job to describe it. Congress doesn't micromanage that. That would be terrible. It's the Fed's job, but it needs to report what it is. That's a little bit of the history. I think it's promising that the bill actually passed the House of Representatives in November, and there's more talk about it.
Beckworth: What I found surprising is the blowback, the pushback, from Fed officials and other sympathizers of the Fed. It surprises me because this is really not a true binding rule on the Fed. It's simply a benchmark that says, "Hey, Fed, you set up your own rule, and when you deviate from it just explain why."
As you said I think it would really help communication. When Janet Yellen goes to the Hill to testify if they had some kind of common framework from which both sides could raise questions, discuss. Right now there's no rhyme or reason to what the Fed does, and it can justify any action, but if we had a benchmark...
Again I guess my observation is that, man, the amount of resentment towards this non‑binding constraint has just been amazing.
Taylor: Yes, also some people think it's not enough on the other side. But I think it's a natural thing. I'm not saying it's a good thing, but it's natural. If you look at that history I mentioned about the requirements to report money growth the Fed was very resistant to any thought of anything like that and testified against it.
Burns was the Chairman. But they finally saw the handwriting on the wall and then started work with the Congress to craft legislation that they could work with.
I think it's natural. Again, I think the reaction has been overboard, because it really doesn't describe what's in the bill. It say things like, "Oh, we can't follow a mechanical formula." The bill doesn't say to follow a mechanical formula. There's all sort of things.
What I've also noticed, too, David, is a lot of people who talk about legislation don't read it. They read what other people say about it, because legislation's hard to read. There are a lot of pages. That causes a dynamic. I suggest that people read Section Two of the FOMC, and you'll see that there's nothing mechanical about it.
Beckworth: Are you hopeful going forward that it will pass? Do you see light at the end of the tunnel?
Taylor: It passed the House. It's not really completely a non‑partisan bill. There's more Republicans in favor of it. In the Senate there's a similar bill in the Banking Committee. It's quite similar, so you've got two bills.
Right now I think the President would probably veto it if it came to him at this point unless it was changed, but I think it's still promising. These things take time.
The examples of the '70s took a long time. It doesn't need to be partisan, actually. There's nothing partisan about this. I think people will say, "This is, maybe, a reasonable thing to ask the Fed to do. It's not micromanaging." I think it's promising.
Of course, one thing that has to happen is the Fed has to get involved in the legislation, because it would be improved by their knowledge and experience, if they were positive about it.
Beckworth: One thing I've noticed over the past, maybe, month or two, is the Fed has been increasingly talking about a neutral or natural interest rate. Janet Yellen gave a talk on it, Vice Chair Stanley Fischer did, several other Fed officials, and there's been a lot of work on this idea of a natural interest rate, which is nothing new to monetary economics, it goes back to Wicksell.
It seems to be this increasing emphasis, or at least, maybe, motivation for why the Fed raised rates. Janet Yellen gave a talk in December, where she actually put a graph in her paper that shows the range of estimates for this natural rate, based on the Board of Governors staff's model.
What struck me is, if they're making this, and if they're using this, to some extent, to help guide their decisions, why not report that to the public? That might be another way to provide a benchmark against which the Fed can justify its decisions, or explain why it deviates.
They already created this natural short‑run interest rate measure, which is very similar in spirit to the Taylor Rule, and she used it in a speech, so it seems to me to be a trivial matter for them to simply update and publish that along with their other statistics every month. That might make it easier, also, when they go to the Hill, in terms of communicating what they're doing.
Taylor: I agree. I think there's another talk that Janet Yellen gave last year, she took the Taylor Rule and then showed how it would differ, depending on what the equilibrium real interest rate was. As we talked earlier, I assumed it was two percent and she plugged in different numbers.
I think that's a constructive way to think about the problem. You're discussing it within a procedure, or a rule, or strategy, how much difference it makes. Yes, I agree completely. In fact, her talk seemed to me the kind of talk that would happen if you had the FOMC, because you'd be talking about a strategy and how it would be adapted, depending on your estimate of the equilibrium real interest rate.
I do think that we have huge amount of uncertainty about this rate. You're right that the discussion has really multiplied in the last two years. For many, many years the discussions about policy rules was not about the equilibrium real interest rate, it was things like, "What's potential GDP?" The Orphanides work was more about that uncertainty.
It's just in the last couple years that there's just been a real explosion of work on the equilibrium rate. I'm not sure what the reason for that is. I think the main reason is that we've had this zero rate, at the funds rate, and the economy's not booming, and people are saying, "Why? That must mean the equilibrium rate is low." I think there's other reasons for that.
There's other kinds of policies, but that's what people have been doing, and, I think, generates a lot of this work on this subject.
Beckworth: John, it's been fun talking to you today. Our guest today has been John Taylor, of Stanford University. Thanks so much, John, for being on the show.
Taylor: Thank you, David. I enjoyed it very much.