Russ Roberts

Sumner on Money, Business Cycles, and Monetary Policy

EconTalk Episode with Scott Sumner
Hosted by Russ Roberts
PRINT
Burgin on Hayek, Friedman, and... Topol on the Creative Destruct...

Scott Sumner of Bentley University and blogger at The Money Illusion talks with EconTalk host Russ Roberts about the basics of money, monetary policy, and the Fed. After a discussion of some of the basics of the money supply, Sumner explains why he thinks monetary policy in the United States during and since the crisis has been inadequate. Sumner stresses the importance of the Fed setting expectations and he argues for the dominance of monetary policy over fiscal policy.

Size: 32.0 MB
Right-click or Option-click, and select "Save Link/Target As MP3.

Readings and Links related to this podcast

Podcast Readings
HIDE READINGS
About this week's guest: About ideas and people mentioned in this podcast:

Highlights

Time
Podcast Highlights
HIDE HIGHLIGHTS
0:33Intro. [Recording date: March 19, 2013.] Russ: I know you are working on a book on the basics of monetary policy, and I wanted to tap into that as well as discuss some recent posts at your blog, The Money Illusion. So we are going to talk about the basics, and ideally we'll get, toward the end, to the interaction between monetary and fiscal policy, and why you are so skeptical of some of the claims of Keynesian supporters in discussing the stimulus. So, I want to start with some very basics, though. You recently wrote that it's hard to argue that the business cycle is caused by real shocks. And you gave some historical examples. Why is it hard to argue that real shocks cause recessions and unemployment? Guest: Well, first of all I should say that I'm talking about countries with large, diversified economies. Like the United States. If you have a very, very small economy dependent on like one agricultural product, obviously a real shock could make a bigger difference there. Russ: Like a bad rain season, a drought. Guest: Yeah, exactly. And also, I'm focusing more on employment fluctuations than Gross Domestic Product (GDP). You could measure the business cycle either as fluctuations in real GDP or fluctuations in employment. And in my view it's fluctuations in employment that are really the key stylized fact about business cycles. We think about the mystery of a depression--why are so many people unemployed when there's lots of things that need to be done and the people want to work, and so on? So that's I think the biggest mystery that needs to be explained, big fluctuations in employment. And I don't think real shocks can do that in the United States, by and large--can cause large fluctuations in employment. I think they are basically nominal or monetary shocks, although not necessarily changes in the money supply per se, but some sort of failure of monetary policy to keep nominal spending growing on a steady path. And as far as the historical examples--yeah, in the blog posts I mention just a few things that people need to think about. I remember the 1987 stock market crash, and there was a lot of talk at the time of how similar the 1929 crash, going into a recession. Not only was there no recession--there wasn't even a tiny slowdown in the economy. Not even a blip. And I think one of the things that points to is--I don't think that something like a stock market crash by itself would cause a recession, unless it was accompanied by a failure of monetary policy. I mention the Japanese tsunami--it had almost no impact on the unemployment rate in Japan. Doesn't show up at all. These are big shocks. That's the biggest natural disaster to hit a developed country probably in my lifetime. And the 1987 stock market crash was certainly the biggest crash over a short period of time--a period of about 6 weeks, I think the stock market lost about 45% of its value, something like that. Russ: Those are just two examples. So I guess the next thing to do would be to look at the big swings in unemployment and ask whether they were preceded by a supply shock. And it's hard to see what those would be. Guest: Well, one thing some people point to is the energy shocks. And in my view it's a little hard to disentangle the monetary aspects and the energy aspects of the 1970s. We know that monetary policy was to blame for the high average rate of inflation during the 1970s. But the energy shock certainly created some variation in the inflation rate. But even there, I think monetary factors were probably the biggest problem, in terms of fluctuations in employment. There were certainly lots of other things that made the economy less efficient, like price controls and so on. But I don't think those are as important for the business cycle. Let me also point out, especially for your listeners that are sort of free market oriented, like myself: I think there is a tendency sometimes to engage in wishful thinking. If you don't like a lot of bad government policies that are happening at the time, your tendency is to attributed economic distress from the business cycle to those bad policies. But we had, under Lyndon Johnson, an enormous increase in the size of the scope of government, and yet the 1960s were a boom decade. So, in the long run some of those policies come back to perhaps hurt us in terms of efficiency; in our medical care system; and so on. But I don't think there's much evidence that that sort of thing really explains the business cycle to any great extent. Russ: And you also mention, which I think was a nice point--I think it's often forgotten because people don't pay careful attention to dates--that the collapse of the U.S. housing market between 2006 and 2008 was not followed by a sharp rise in the unemployment rate. Guest: Was not accompanied by. Right. So, if you take the peak of the housing boom--was about January of 2006, unemployment was 4.7%. And then 27 months later, in April of 2008, housing construction had fallen in half, but unemployment was still 4.9%. Which is really a very low rate. And the reason for that is: The rest of the economy is doing fine. So as jobs are being lost in housing construction, we are picking up jobs in commercial real estate, in exports, manufacturing, services--all sorts of other sectors. But the overall unemployment rate wasn't changing that much. I mean, GDP growth was slowing a little bit; but it was still positive. And then in the second half of 2008 when nominal spending fell very sharply--in fact at the sharpest rate since the Great Depression--that's really when it spread to other sectors: commercial real estate, manufacturing, services. All of them started losing large amounts of jobs. And that's when unemployment doubled from 4.9% up to 10%. So, I think that if you look closely at the data you find that even a sizeable sector for the economy--like residential real estate, which was 6% of GDP at the peak--a decline in that sector really isn't enough to create a recession. If you have stable monetary policy that keeps nominal spending growing at a slow but steady rate.
7:23Russ: It just comes back--the way I think of this puzzle of the labor market and its variability and how hard it is sometimes to find a job--in the 1990s, if you got laid off or got tired of your job and you quit or you were a senior in college going out on the job market, there were lots of jobs. And that was true up until about 2001. That being laid off or quitting was not a traumatic experience. It's not fun, but you were pretty confident--I think most people would be confident in those times. True in the 1980s as well--well, I'll find another job. But right now, it doesn't feel that way. And I don't think it is that way. Something is different. Guest: Right. And I think--I attribute this to the nominal spending or nominal GDP shock that hit us in 2008 and 2009. Over a 12-month period we had a 4% fall versus a trend rate of 5% increase. So really our nominal GDP growth was about 9% below the trend between mid-2008 and 2009. And that just had a devastating effect on the job market. I like to use a metaphor of a game of musical chairs. So if you visualize that--if you have 100 people and 100 chairs, everybody has someplace to sit down. But the music stops. Russ: It might take a while to find them sometimes. They might not be in a neat circle. Guest: Right. Russ: But they are around. Guest: Well, yeah. That's a nice supplement, because that fits in to how the job market does have this information problem, search process. But they eventually do find them, usually fairly quickly. Now if you suddenly take away, say, 4 or 5 chairs, then a bunch of people are going to be sitting on the floor. And the way I look at the job market is that nominal wages tend to be kind of sticky or slow to adjust. So, when there's less total spending in the economy, or total net nominal income in the economy, but people are paid about the same amount per hour, there's going to be fewer hours worked. And that's your recession, basically. So you have roughly the same pay per hour, a smaller amount of total money circulating, expenditure in the economy--that is, a smaller amount available to pay workers--there will simply be fewer hours worked. Russ: Instead of the alternative--that wages would fall and people could keep their jobs with smaller hours. Guest: Right. But that really doesn't happen very smoothly. It certainly happens in some sectors. You can find really hard-hit sectors where workers have taken pay cuts. But there are many, many sectors that are quite large--like my own, where we haven't had any pay cuts at all; we've just had a slowdown in the rate of increase. I think, not just education but health care and many other sectors that aren't hit as hard have relatively inflexible wages. And so the overall wage level in the economy does not adjust when nominal spending grows 9% less than trend over 12 months. What we've had since then is some adjustment in wages, and as the adjustments occurred, the unemployment rate has slowly fallen from 10% at the peak to 7.7%. So you could say maybe it's halfway back to normal. And I think that along current trends we would probably see further adjustments in wages, further slow growth: we're seeing 2% growth in wages instead of the normal 3 or 4%; and that is gradually healing the labor market. But it's healing it more slowly than if we'd had steady growth in nominal spending.
11:01Russ: So I need to ask some clarifying questions. It's slightly embarrassing. Well, it's embarrassing in lots of dimensions--because I've had you on before and I've probably asked some of these questions before. But I like to think I get a little smarter as the months go by--not always. That's the EconTalk Illusion, maybe. But anyway, when you say that nominal income or nominal spending--first let's clarify for some listeners: That means denominated in some actual dollars. Not corrected for inflation or anything like that. So when the total volume goes down--you know, you are talking about it like it's an independent thing. Isn't that just a different way of saying we're in a recession? It sounds like you are saying that recessions-- Guest: Oh, good question. I get that a lot. But that's actually not the case. Because--okay, a couple of examples. We all know about the big hyperinflation in Zimbabwe, recently. Or in Germany earlier on. Now obviously their nominal GDP was going up dramatically. And yet if you look at the real side of the economy in Zimbabwe it was doing very, very poorly. You could probably call it a recession. So, I don't think that you can simply look at ups and downs of nominal GDP and say: Well, that's automatically a business cycle, so that I'm just talking in terms of a tautology. A lot of people think that when I first explain that nominal shocks cause business cycles. And they say: Well isn't a shock in nominal GDP just a recession, so-- Russ: It sounds like you are saying a recession causes a recession. Guest: Right. It sounds that way. But actually a recession would be better described as a fall in real GDP. And nominal GDP I think is better described as a stance[?] of monetary policy. Now, it is true that in the long run these two are sort of independent of each other. So, printing money doesn't really in the long run make a country richer. And reducing the money supply probably does not make it poorer in the long run. But what happens is, because wages and prices are sticky, you get these short term cyclical effects from fluctuations in nominal GDP. So, if you believe that this correlation is very strong--and your question actually presupposes a stronger correlation than even I'm claiming. So, you're sort of saying: Aren't real and nominal GDP always closely correlated with each other? And I would say to that: Not quite always. There could be hyperinflation where a breakdown, or there could be such an enormous real shock that you could have a fall in GDP, real terms, even without a fall in nominal. That could happen. And it does happen in small countries. But what I would say is this: For the United States, yes, they are so closely correlated that that raises the question of then: Could we then smooth out the fluctuations in real GDP by smoothing out the fluctuations in nominal GDP? And then the second part of that question is: Can monetary policy deliver slow but steady growth in nominal GDP? Russ: Yeah, those are the two key questions. Before we get to those, let's do a little bit of history of economic thought. Very recent history of economic thought, the last, say, 60 or so years. So, we go back to the early work of Milton Friedman. He would argue--I think he argued that the quantity of money in the short run had real impacts on the economy. I'm rephrasing, I think what you just said. In the short run, it could have real effects. And in the long run, it had very little. Or, it could be destructive; but it was intended to work through the nominal values, through the prices, not through real output changes. And inflation in and of itself can be destructive to decision-making, so that's the sense in which it can have real effects. But what he really was saying is that short run effects of money are real and long run effects are only nominal. Is that fair? Guest: Yeah. And my views are very similar to Friedman's in many respects. And the one difference would be that he focuses more than I do on the actual money supply. I focus on what is sometimes referred to as the money supply times velocity. So, some of your listeners may have seen the famous equation of exchange, MV=PY. And the total spending in the economy is the amount of money in the economy times its velocity of circulation. So, the early monetarists like Friedman tended to assume that velocity would be fairly stable if monetary policy was stable. So, they just had M grow at 4% a year, that would keep total spending at a pretty stable path. Russ: Because V would be stable. Guest: Right. Russ: And then PY, the nominal value-- Guest: The other side, which is the nominal value of income or expenditure, would be also stable. And our group--sometimes called 'market monetarists'--tend to focus instead of on the money supply itself on just the total amount of spending, M times V. And so what we're trying to do is we're not trying to stick to a steady growth path for the money supply, but rather adjust the money supply as needed to offset any changes in velocity, in order to keep that total spending path of nominal GDP pretty stable. And that's still sort of in the spirit of Friedman. It still has sort of the same goals. It has the same assumptions about in the long run money is neutral and it only leads to inflation and doesn't affect real variables. So, much of the structural model of monetarism is still there. It's this more modern version we're working with. But we focus more on total spending rather than just on the money stock itself.
17:07Russ: Well, one of the virtues of that approach, of your approach, is that M--it's a nice thing to write down in print and a textbook, but what is it exactly? So, one of the problems with using the quantity of money to measure whether monetary policy is doing anything or makes a difference is it's hard to know how to define it and then how to measure it. The different measures of M--M1, M2, M3, M4, etc. And some people think that those have become outdated. The Fed doesn't even publish some of those any more. They are still collected by some under-the-radar folk. The puzzle for me, then, is: How do you know whether you are doing it well or not? Normally--this is a great mystery to me; I've asked many guests about this; maybe I'll finally understand it after you answer--but, outside of you and me, everyone else seems to think we should look at interest rates. So, if interest rates are low, then monetary policy is expansive. Expansionary. It's 'loose.' And this then leads to a conclusion by some people, mostly Keynesians, that monetary policy is ineffective, because it can't influence--interest rates can't go negative, so they're close to the so-called 'zero bound', which is just a way of saying zero. And therefore monetary policy is ineffective. Nothing is left. We have to turn to fiscal policy. We've talked about this before, but explain again why interest rates are misleading. And then if interest rates are misleading, what am I left with? Guest: Well, first of all let me just say that although you say everybody just thinks in terms of interest rates, you could almost do a very interesting study on why that is, because our textbooks say that interest rates are not a reliable indicator of monetary policy. And the head of the Federal Reserve, Ben Bernanke, says that interest rates are not a reliable indicator. And so on. And Milton Friedman said they aren't a reliable indicator. And so actually there's a long tradition in economics among even not just monetarists like Friedman but, again, in our textbooks; and Ben Bernanke is sort of a New Keynesian. It's not at all generally accepted that interest rates are a reliable indicator. Interest rates tend to be really high during hyperinflation; but nobody really thinks that's tight money. So, I think we have to start with a fact that, although when I started my blog and started talking about interest rates being misleading, it was somehow seen as a contrarian view. It really shouldn't have been seen that way. But for whatever reason, it was at the time. And then the other part of that is: interest rates really need to be seen as conditions in the credit market that reflect conditions in the economy. So, generally speaking there are two things that will make interest rates low. One is low inflation and the other is a weak economy. There are other factors as well, but those are the two main ones. And, obviously we have both right now. We have relatively low inflation and we have a very weak economy. So that's why interest rates are low. Interest rates tend to be high during the peak of a business cycle and they tend to be high when inflation rates are high, like the 1970s. So that's my starting point about interest rates. Then if you look at monetary policy, you've got this problem: A tight money policy is likely to do two things. It will make inflation lower. And a tight money policy will push an economy into a recession. Well, both of those things tend to reduce interest rates. Right? Low inflation and recession. So, Milton Friedman once said: Very low interest rates like in Japan--talking about Japan at the time--are a sign that money has been tight. And I know that sounds contrarian, but it's very much consistent with basic economic theory. Even to some extent Keynesian theory. Russ: So, how do you reconcile that with the fact that if you asked, I think, Ben Bernanke or most observers they would say that monetary policy has been wildly out of control and expansive? Guest: All right. Well, what I would do if I was asking Ben Bernanke is I would quote him from 2003, where he said that the only reliable way of looking at monetary policy is to look at the rate of growth in nominal GDP and inflation. That's what he said. And then I would point out that since 2008, those two on average, if you average the two, have been the lowest since Herbert Hoover was President. So, using his own words it's inescapable that money over the last 5 years has been tighter than at any time since Herbert Hoover was President. And then I would ask Ben Bernanke: Why have you now changed your views and why are you now saying money is unquestionably accommodative, when in 2003 the criteria you laid down suggest it's actually very tight? So I would put him on the spot. Russ: And how would he answer that? He's not a fool. He's a smart person. Guest: I don't know. I wish a reporter would ask him at a press conference. Quote him from 2003. Russ: In 2006 I asked Milton: Why is it if it's the quantity of money that matters--and again, you can debate whether we can measure it or not, whether there are offsetting changes in velocity, but Milton basically was a quantity guy--I said why does the Fed always talk about interest rates? And he said: It makes them feel better. It makes them feel important, that they are doing something; they are manipulating, they are steering the interest rate. Do you think Ben Bernanke got that disease once he got into the chair? Guest: No. Well, look. I think that it's a little more complicated than that. First of all, the Fed does have a short term impact on interest rates. And it is true that when they do an easy money policy, it's often the case that in the very short run, short term interest rates will fall. Sometimes long term interest rates go up at the same time, so it's kind of confusing there. So there's a grain of truth--obviously if there wasn't a grain of truth in it they wouldn't be talking this way. The other thing is, I think they overweight how important interest rates are to what is called the 'transmission mechanism.' So, if you ask the average person: Why does monetary policy affect the economy? Most people either wouldn't have a clue, or the only answer they would be able to give is: Well, if the Fed cuts interest rates it's more likely that I'll go out and buy a home. Or borrow money to start a business. Russ: That's what the business community, the business journals write over and over again. It's why they say monetary policy is ineffective: Interest rates are low, that discourages investment but they can't drive it any lower. But that seems ridiculous. Guest: Well, it is. Often people--people at some level know it's ridiculous, so they'll often say both things. So, in the Great Depression for instance the business community had two objections to easy money. One objection was: it's pushing on a string, it won't do anything because interest rates are already near zero. And the other was: Well, if Roosevelt keeps doing that we'll have hyperinflation like Germany had. So those were their two objections: they can't do anything and it does too much. Now, in case you think that's just the Depression, the Bank of Japan has said both things. They've said: Well, we can't do any more. And: if you force us to do what you are asking us to do, we'll get high inflation. Russ: And it's the same thing right now. Guest: It's a contradiction.
24:58Russ: Let me ask you two more questions on this theme. Which is: So you are saying that monetary policy has been tight but not loose. The policy has been contractionary. Despite the talk. Despite the low interest rates. But you look at the Fed's balance sheet--it's gone nuts. Base money--base money, high-powered money--is so big. The Fed's been so active. There's been all this quantitative easing [QE]. How can you say that monetary policy hasn't been active or loose? Guest: All right. Well, let me point some things out. First of all, so your listeners don't think I'm completely crazy here, there's a pretty general consensus among economists today that monetary policy in the early 1930s was contractionary. That period of time when we had big deflation. This is Friedman's view, Ben Bernanke's view; the view in the textbooks. It's pretty widely accepted that the monetary policy was contractionary. Now, what you don't often hear is that under Herbert Hoover, they did QE--they did quantitative easing. When interest rates fell close to zero they started to do big open market purchases of bonds, in 1932. And between 1930 and 1933, the monetary base rose very sharply. Not as sharply as recently--but then, they weren't paying interest on reserves, either. That's one of the things that is a quirk in the modern system. All this money the Fed has supposedly been "printing" actually hasn't been printed. Almost all of it is an electronic entry in a bank, a deposit at the Fed that the banks earn interest on. Russ: That's a new-- Guest: And really what the Fed is doing is they are swapping one kind of interest-bearing liability of the government, called Treasury Securities, for another interest-bearing liability of the government called Bank Reserves at the Fed. They are just swapping one for another. If they were truly printing money, the supply of cash in our wallets would be going up a lot, like it did in the hyperinflation in Germany. Now that has gone up some. There has been some increase in cash in circulation. But that's really a reflection of the fact that when interest rates fall to zero and you can't really earn anything in the bank anyway, people do prefer to carry a little more cash than during normal times. But it's been a very small increase in actual cash in circulation. It's been mostly this reserve game they are playing of shuffling between two assets--Treasury Securities and Bank Reserves. Which are both interest bearing. Russ: But isn't that-- Guest: That's not printing money in the hyperinflation sense of Germany. That was not interest bearing. Russ: I understand. But doesn't your observation then just prove the critics' point? You are saying they haven't really done anything. Isn't that just another way of saying monetary policy is ineffective? They've done all this "stuff." It's just on the books; it's not real. And so monetary policy has had very little impact. What would you have had them done? Guest: They've done the wrong stuff. Russ: What should they have done? Guest: The way you have to think about monetary policy is sort of do the reverse process from everyone. Everyone thinks in terms of gestures and then waiting around to see if they succeed. The way you think about monetary policy, in my view, is you start with success and then work backwards and ask: What do we need to do? So, let's say the target is 5% growth in nominal GDP. So, you start by announcing you are going to do whatever it takes so that the markets expect 5% nominal GDP, [?] to occur. Whatever it takes. Even if you have to buy up all of planet earth. Now, you are not going to have to do that. In fact, if you announce that target you probably will end up doing less QE than we've actually done. So, the countries that have the most inflationary policies today, like Australia for instance, have the smallest amount of money printing going on. And the ones that have the most deflationary policies, like Japan, have the largest amount of money printing. Which I know goes against common sense, but it reflects the fact that when you expect faster growth of nominal spending, you don't want to sit on cash as much. So, the amount of printing of money the Fed has to do is really determined by what the public wants to hold. I'm going to back up because this is kind of hard to grasp. Let me give you an analogy with something your listeners may have heard about, like the Gold Standard. Let's take that as an analogy. Under the Gold Standard, the way it worked was the government would set a price of gold and basically say: The public can hold as much cash as it wants; it's up to the public; we will supply whatever they want in terms of bringing gold to the central bank and asking for cash. So it will be completely demand-determined. Does that make sense? Russ: Yeah, keep going. Guest: Okay. So, what you are doing there is you are targeting the price of gold and then letting the market determine the money supply and interest rate. That's a gold standard. Now, what I'm proposing is instead of targeting the supply of gold, you target something like a nominal GDP futures contract. Or at least you target market expectations. So you say: Look, we're going to watch the markets and we are going to supply money until the markets expect nominal GDP growth to be at least 5%, if that's the target. Now, the next question I always get is: How much money would they have to print? They've already printed so much. And it hasn't done much. My answer would be: It's very possible they'd have to print less. Because when people expect faster growth they have less incentive to just sit on cash that doesn't earn interest. Russ: But you are assuming-- Guest: By the way: I would stop paying interest on reserves, too. I would tell the banks: You are not going to earn any interest on reserves. We are going to get faster growth in the economy. If you want to sit on that cash and earn nothing, that's fine. We'll supply all you want to sit on. But we are not going to pay you interest on it. And instead we're going to target nominal GDP. You'd find that banks actually wouldn't want to sit on as much cash if they expected faster growth in nominal GDP.
31:39Russ: I agree with all that. The problem is that you assume--and you may be right, but you may not be right--that the promise is credible. You assume that because Ben Bernanke and our imaginary world in 2008, instead of doing what he did he had said: We're not going to let nominal GDP fall. We're going to make sure that it grows at, say, 5% a year. You're assuming that people would say: Oh, I don't want to hold money and have it just sitting around. I'm going to invest it. But maybe the reason they are not investing it--and by the way, Scott, I'd love to believe what you are saying; I think it's lovely. It's got a certain aesthetic appeal to it. But it could be that the reason they are not using that money, the banks, is the economy stinks. There are not many good, attractive investments. And so Ben Bernanke can talk all he wants; he can try to change expectations. Guest: Yeah. Russ: How do we know he can change them? Guest: Okay. I've got three sort of interrelated answers to that. One is there's really no example in all of human history of a fiat money central bank trying to inflate and failing because of the lack of credibility. That would be my first answer. My second would be: There are techniques that can be used to sort of lock in credibility. Now Roosevelt used devaluing the dollar against gold. I would not suggest that today because gold doesn't play the same role in the economy. I've talked about nominal GDP futures markets. But even if they don't go that far, it's surprising how little the central bank has to do to change expectations. Not how much. Let me give you an example. Over the last, say, 3 months, the Bank of Japan has not even come close to doing what I'm proposing. They are sort of being dragged kicking and screaming into a world of a 2% inflation target. In fact, they don't want to do that. And the government is trying to push them in that direction. I would actually want the central bank to enthusiastically endorse a 2% inflation target, because that would be far more credible. So even in this worst case of Japan, where you have a dysfunctional central bank that's not cooperating, isn't doing the right things to create credibility, what they've been able to do is, in 3 months, depreciate the yen dramatically--and we've seen the market respond to specific announcements coming out of the government on monetary policy. So, we know they are linked. Speeches by the new Prime Minister, and so on. Stock market's up, what, 45% in the last 3 months. Phenomenal increase. And again, the stock prices have been strongly linked to various public statements from the Administration there proposing a higher target of 2%. First, as you know, they've had mild deflation in recent years. So, here's a case of a central bank doing far less than I would ask them to do, in a situation where I would even be doubtful if it would have any credibility, given how little enthusiasm they have for this. And yet the markets seem to have treated it very seriously, and actual exchange rates and stock prices and other variables have moved very strongly in just a hope that they'll do what the government wants them to do. There's no promise out of the central bank they are actually going to hit a 2% inflation target. Russ: So, let me restate my question. I'm going to say this in a different way. You are saying if the Central Bank of the United States, if Ben Bernanke had done exactly the same thing that he actually did--which was to buy up huge amounts of mortgage-backed securities from the balance sheets of banks, injecting reserves into the banks, which they now hold at the Fed--as you say; and they bought some Treasuries, too; those are just accounting things that they did. They increased the reserves that banks held at the Fed by buying up their assets. If they'd have done that and at the same time not paid interest on those reserves, and at the same time said they had a different policy--they wouldn't have had to do as much of it? And the economy would be healthier? Guest: Yes. Absolutely. I think that if they had had a more robust policy, especially back in 2008, a policy like the Australians had, keeping nominal GDP growth, at least long term, along a trend line, I think it's very possible--first of all, it's possible interest rates never would have gotten to zero. Because if your nominal GDP is expected to grow at 5%, your demand for credit is going to be much stronger. So, Australia, interestingly, was the one developed country that didn't have a recession this time around. Everybody else did, pretty much. And guess what? Their interest rates never fell to zero. Almost all the other countries saw interest rates fall virtually to zero. Now, if you really believe low interest rates were easy money, you would have to believe that the one country that had the tightest monetary policy in the world out of the developed countries somehow avoided a recession in 2009. In fact, what actually happened in Australia is they had a stronger nominal GDP growth track. A little higher than I would recommend for the United States, by the way. They've averaged, oh, sort of 6 or 7% nominal GDP growth over a decade. And it did dip a little bit, but then it came back. So they've been about to maintain that trend line pretty well. And as a result they actually haven't had to inject much money at all. There, what's called the monetary base--that's the money actually printed by the government plus bank reserves--it's only 4% of GDP in Australia. It's 18% here; 23% in Japan. So, the irony is, the central bank that has the most expansionary monetary policy, which is Australia of the developed countries, looks like it is doing the least. And it's actually doing the most. It has the smallest monetary base and higher interest rate. And the country that seems to be doing the most effort--Japan, which has had interest rates near zero for 15 years and has a huge increase in their monetary base, even larger than the United States--they've had deflation for 15 years.
38:23Russ: But how do I know what's causing what? You've got to be arguing that in Australia, the Central Bank had this clear policy and everybody knew it was going to happen, so they didn't have to do as much. Whereas here, we didn't understand monetary theory, monetary policy, and we tried to just print the money without changing the expectations and so nobody believed it and it kind of just slumped along. How do you know if there were real things in the economy that were causing those changes? Guest: Well, there definitely were real things that caused monetary policy to go off course. So that's true. In other words, let's take the housing bubble, for instance. I'm not trying to argue that the entire crash of the housing bubble is due to mistakes in contractionary monetary policy. I think the second half of the decline, in 2008, 2009, was that the first part was due to, you could call it a real shock, and there's all sorts of theories about what went wrong with the housing market. I don't have anything interesting to say on that subject other than that when the crash did occur, it did tend to depress what's called the 'natural rate of interest'. So, one of the reasons that interest rates hit zero in the United States was what I regard as overly tight money that led to low expectations of nominal GDP growth; but another big reason that interest rates hit zero was that the demand for credit fell sharply with the crash of the housing sector. So, that was a real shock. As it pushed interest rates toward zero, we discovered that the Fed actually didn't have a Plan B for operating at zero interest rates. Even though Ben Bernanke, in his academic writing, said: Well, even at 0 rates the Fed can do lots of things, so it's not necessarily a big problem. But I think what we discovered is Ben Bernanke and the Fed are two different entities. So there's the academic Ben Bernanke that was telling the Japanese they should have done all these things at zero rates. And then there was the Federal Reserve, which is a big institution; it's conservative, it's hard to move in a radical way in terms of new ideas, new concepts. So, once we got to zero interest rates, it was almost immediately apparent, around late 2008 that the Fed really didn't have a backup plan. They had no plan for preventing nominal GDP growth from plummeting and staying--this is very important--from staying well below the trend line. They had no plan for getting a recovery in the economy, other than keeping interest rates low and crossing their fingers and hoping that would work, even though after 15 years it hasn't worked in Japan. Russ: But you make it sound like they are doing nothing. They weren't doing nothing. They were massively intervening in asset markets. They were buying up stuff-- Guest: You're right. So they did buy a lot of securities. And, I don't know how much of the failure of that to work was the reason that it didn't work in Japan, which was just bearish expectations, and how much of it was their decision to pay interest on reserves. It was some combination of things. Russ: Fair enough. Guest: And for whatever reason, it didn't. But they also did it in such a way that was almost calculated to fail. For instance, they said: Now, don't worry; we're going to pull all this money back out of circulation at the slightest sign of an uptick in inflation. Well, actually, you'd be better off when you have a severe slump of trying to get back to the trend line. And to do that, that's going to require inflation to be a little bit above 2% in the recovery phase, just as we had deflation between 2008 and 2009. But the Fed decided they wouldn't risk that. So they consistently told the public that they are not going to allow inflation above 2% until--and now I have to back off. A few months ago, they changed policy and said that they'll tolerate inflation up to 2.5%. And that's still not the best way of doing it. They should be targeting nominal GDP. But that policy change probably was significant. That does seem to have changed expectations a little about monetary policy going forward. And we'll probably get to that later as we get into the policy questions. But in terms of what went wrong back in 2008 and 2009, their big mistake was not to do what Ben Bernanke told the Japanese to do 10 years ago, which was do something called 'level targeting'--promise to get back to the same trend line that you left during the recession.
43:18Russ: So, let me give you a different version of this story about how money matters. And it's a different story about the ineffectiveness of monetary policy. And that's Steve Hanke's. Which I associate in some degree with Milton Friedman's, maybe more than yours. And you can try to see where your story fits in with these other stories. If I have the story right. Which is the following. Hanke on this program a few episodes back said: Well, it's true the Fed has expanded the monetary base dramatically. But it's not been enough to offset the collapse in what he called 'bank money'--that is, I assume he was referring to the collapse of the shadow banking sector in 2008 and 2009. That this enormous reduction in liquidity as banks went either bankrupt or insolvent and had to be bailed out, as they pulled in their loans, as they deleveraged, as people stopped borrowing, as people paid off loans, that the money supply, properly measured--he uses M4--shrank. And although the government was very active, it didn't do enough. And similarly, you argued that the Fed was active in 1932 but Milton would say they weren't active enough. The money supply still fell dramatically through the early part of the 1930s. How does your story fit in with those stories? Guest: Well, it fits in very closely. The only difference is they look at the impact of the money actually produced by the Fed, which is called the monetary base and how it impacts the broader aggregates, like M2 or M4. I look at the money produced by the Fed, the monetary base, and how it impacts nominal GDP. But the basic story is essentially the same. All three of us believe that the money created by the Fed itself, the base, wasn't done in such a way that it stabilized whatever the target variable be. Whether it be M4 or nominal GDP. So, we just have a slightly different view on what's the optimal target variable. But the essential underlying analysis is very similar. Russ: But your--the difference, if I understand it correctly--is that in your story, it's the setting of expectations. It does seem kind of magical, Scott. It's not the actions that they took that failed; it's that they didn't talk about it the right way. Is that fair? Guest: Yeah, but talk is the most powerful action they have. The reason why is very interesting. Here's something people don't think about. What really matters is not what the Fed is doing now but what it's expected to do in the future. Let me give you a quick example. Probably your listeners have heard about what's called the 'quantity theory of money'--that if you double the money supply, the only thing that happens is you get inflation. Prices would double. Russ: Yeah, roughly. Guest: Nothing real would change. Everything would be twice as expensive. That's the quantity theory of money. But it only applies if the monetary increase is permanent. Because think about this. Suppose the Fed announced that we're going to double the money supply for four weeks, and then we're going to bring it back down to the original level. I don't think there's any quantity theorist on earth who would say: oh, the price level will double for four weeks and then it will fall back down to the original level. Because think about it: Would you buy a house that went up from $200,000 to $400,000 for four weeks and then drop back down to $200,000? Obviously not. No one would pay the $400,000. So, when you have a monetary injection, it's not going to have much effect on the price level unless it's expected to be permanent. And that means that what really matters when we set asset prices--house prices, gold prices, stocks, bonds--every asset out there is priced based on what people think the Fed will be doing in the future. Now they may not realize that. They may think in terms of vague phrases like confidence: Do I have confidence that the real estate market will be strong over the next 5 years? But they are really thinking about nominal GDP growth even if they've never heard that phrase. Or: Am I confident my income will be growing over the next 5 years? So what really matters is confidence in a stable but positive monetary policy going forward over a number of years. And that's why talk is the most important thing the Fed does. It's talk not just out of thin air, but talk about specific things the Fed will be doing in the future. And since the specific things they'll be doing in the future determine the value of asset prices today, that's the most important part of Fed policy. Signaling, communication, whatever you want to call it. And what it does this day, this week, this month is much, much less important than what it is expected to do 1, 2, 3, 5 years out into the future as a path of policy over time. Russ: I certainly agree that expectations matter a lot. Let's shift gears and we'll of course come back to this. Guest: That's what talk is about. It's about expectations. Russ: Right, but talk is cheap. And the question is whether you can talk in ways that are credible. You have to talk about you have to act in ways that make your talk credible. Guest: You know, they are very credible. The central banks a few decades ago said: Oh, we blew it back in the 1970s. Now we understand the Taylor principle, so we are going to start targeting inflation around 2%. And almost all the major central banks started doing that. And they basically succeeded in giving us an average inflation rate of about 2%. So people know that central banks, when they say they are going to do something, they know they can do it. And again, with the Japanese case very recently, we've seen very vague statements having an immense impact on Japanese markets. Which means people absolutely do pay attention to what the Central Bank says. Russ: Well, I think that's true.
49:36Russ: So, I want to shift gears as I said, but it's really just a variant on the conversation. Which is that you wrote recently that Keynesianism, this whole idea that there's a possibility even of a fiscal stimulus, is a fraud. You said there's no evidence that the fiscal stimulus of 2009 had any impact, and you say it's scientifically empty, is how I would describe it. But you called it, I think, a fraud. What was your point there? Guest: Well, first I need to explain the underlying reasoning here. We have to start with the fact that during normal times, when interest rates are positive, almost all mainstream macroeconomists agree that monetary policy is the proper tool for stabilizing the economy, not fiscal policy. A lot of your listeners who may have studied Intro to Economics a few decades ago might have been taught that the Keynesians believe in using fiscal stimulus during recessions. Actually, that was abandoned by the late 1990s. The so-called 'New Keynesians' decided: No, the Fed should steer the economy with a 2% inflation target and we shouldn't use fiscal stimulus any more. The reason is very simple. Both of those tools are aiming at Aggregate Demand, or total spending in the economy. If you are already using monetary policy to get the optimal path of spending, then there's no possible role that fiscal policy could possibly play. You've already determined the path through monetary policy. So fiscal policy becomes completely redundant. It's also less efficient; there's long lags; Congress isn't very good at smoothing out business cycles; it runs up deficits. There's just all sorts of reasons that monetary policy is better for stabilizing the economy. Now, that part is not controversial. Here's where it gets controversial: When interest rates fall to zero, is there once again an argument for using fiscal stimulus? Paul Krugman and others say yes. Now, at zero interest rates there's a powerful argument for using fiscal stimulus because monetary policy is ineffective. However, Paul Krugman and others also say: Well, monetary policy actually could do a lot more if the Fed were really willing to be more aggressive; but they are just too conservative to do the things they really need to do. Therefore we need fiscal stimulus. And my response to the Keynesians is: Then you are sort of arguing for fiscal stimulus on the basis of incompetent monetary policy. Which is defensible. But if you look more closely at the Fed and ask in what way are they incompetent, they are not incompetent in the right way in my view to make fiscal stimulus work. Let me explain that. Russ: Fiscal or monetary? Guest: Fiscal. In other words, what I talk about is what's called a 'monetary offset.' The easiest way to see this is if fiscal stimulus really did boost aggregate demand, it would raise inflation. At least a little bit. Now, if the Fed is targeting inflation it will just tighten monetary policy to offset that. It will prevent inflation from rising and it will thwart the intentions of the fiscal stimulus. So, in order to make fiscal stimulus work, you need an incompetent central bank. You need one that isn't effectively targeting inflation. Or whatever goal variable it has. And instead just sort of passively lets fiscal stimulus move inflation up and down according to the whim of Congress. But I would argue that this vision the Keynesians have of monetary policy becoming passive at a zero interest rate is simply flat out wrong. Yes, the Fed has been too cautious; they've been too conservative to promote recovery that both Krugman and I would have liked to see. But they've been passive in a very specific way. Once we fell into recession, they were very clear in the sense of, like: This much recession and no more. So the Fed started to do very aggressive things every time the economy seemed to falter. First QE1. Then QE2. Then they did something called 'Operation Twist.' Then they did QE3. Then they did something called the 'Evans plan for signaling inflation and unemployment targets.' So they've done like at least five major unconventional actions that they adopted at various points when the monthly data seemed to show the recovery was faltering. And by doing all those specific actions they've been able to keep nominal GDP growth growing at about a 4% rate, pretty much regardless of what was happening to fiscal policy--whether it was expansionary or contractionary. It didn't make a lot of difference. And this year a lot of Keynesians said: Now we've had these tax increases; now the economy is really going to slow, because our models say that's a fiscal austerity. And at the beginning of the year I said no, probably GDP is going to grow just as fast in 2013 as in 2012, even though they've raised the payroll tax and they've raised taxes on the rich and they've cut some spending recently. Russ: Sort of. Guest: The spending is debatable. But there's no question at the end of 2012 they did enough fiscal tightening, according to the Keynesian models, to knock about 1.5 points off GDP. That was at least what I'd been seeing in the Keynesian articles. Now, we'll see how it plays out. But it wouldn't surprise me at all if GDP is just as strong this year as last year, because the Fed's actions, which were taken partly to offset this Fiscal Cliff, will essentially neutralize the effect of it. And by the way, you can find statements by people at the Federal Reserve pointing to the dangers of the Fiscal Cliff as one of the reasons for the fairly extraordinary actions they took late last year. Or at least actions that were sort of unprecedented in technique, let's say. I wouldn't say their actions were that powerful. But they were certainly a move in the direction of setting some explicit targets for the economy. And in my view the expansionary effects of that are probably large enough to offset the contractionary effects of tax increases. And if there are spending cuts with the sequester, which we'll kind of have to wait and see; but you're still going to get the same sort of growth. In fact, recently I've seen some articles suggesting growth is even picking up a little bit in the last few months. Which is exactly the opposite of what should have happened if the fiscal austerity model was correct. So, no, I don't think fiscal policy has an effect because I think the Fed neutralizes it. They have their own target. Russ: And do you think they neutralized the effects of the 2009 stimulus, then? Guest: That's a harder thing to know for sure. But here's the argument that I made recently. What if we had done no fiscal stimulus at all? So, you recall in early 2009 the stock market had crashed, unemployment was soaring, there were a lot of articles about another depression coming; the banking system was in severe trouble. Everything looked extremely bleak. The stock market was I think less than half of its current level. What if the government announced: Well, fiscal policy will do nothing to save us from a depression. Russ: We can't afford it. Guest: So, all the attention of the economics profession and the journalists and the pundits would have been on the Fed as our only hope. And they would have dug up immediately Ben Bernanke's writings that the Fed can prevent another depression. The Japanese should have done all these things. And we should have done all these things differently in the early 1930s. And they would start asking the Fed now: Why are you not doing the things that Ben Bernanke told the Japanese they should be doing? Like level targeting. Which to this day the Fed has not done. And I think the Fed would have been under enormous pressure. And I think Ben Bernanke himself would have wanted to do a lot more in those circumstances. Now, I can't say for sure but if the Fed had adopted something like a level targeting plan, which Bernanke recommended on the Japanese, I think that would have been much more effective than the actual policies they did plus the fiscal stimulus. We'd be today at a higher level of spending in the economy if Bernanke had done what he recommended the Japanese do about a decade ago. Instead of the actual policy of the Fed plus the fiscal stimulus. And I think it's plausible they would have done that because they would have seen the emergency we were in. Now, you can say that's speculative, and I perfectly agree with that. I can't be certain. But my point is: fiscal stimulus is not a scientific concept, because it's based on assumptions about Fed response, offset, that are completely hypothetical, psychological, mindreading. And they present these multipliers as if they are some sort of parameter of nature, like a scientific model. Which they are not.
59:36Russ: So, I'm sympathetic to that view. And I'm sympathetic to your view. And I'm also sympathetic to what I would call a micro view of recessions. Which is not quite the same as a supply shock story. Let me just throw this out there and let you react to it. The Keynesians--I hate to call them that; I'll just say Paul Krugman and others who support increased government spending to help the economy--they argue that, well, something went wrong in 2008 and we didn't have enough spending; consumers lost confidence, businesses lost confidence; so government has got to step in and make up the difference. That's their story. Your story is--well, my story is: government doesn't spend money well; I don't have any reason to think that causes prosperity; and I don't assume that you can hold everything else constant while it does that and you just get a boost, in, say, nominal income. Your story is: Something collapsed in 2008 in the nominal side of the economy and the money side; the Fed should have tried to make that up; they sort of tried; they didn't do it effectively; they weren't credible; and as a result, we're in a slump. The third view says: All this macro stuff is just talk; there are real things going on underneath the economy. I recently interviewed Ed Leamer for a project I have called the Numbers Game, animated conversation shorter than EconTalk, where he says: manufacturing employment has been falling in real terms due to productivity and globalization--manufacturing employment not manufacturing output. Manufacturing employment has a long negative secular trend. Big decrease in number of manufacturing workers and so recessions aren't getting the bounceback they'd normally get if the trend were positive. Casey Mulligan on this program argued that we changed all kinds of incentives to keep them out of the labor market, and that's why unemployment is so high. How do I distinguish between these three different stories? The world is a complex place. As a Hayekian I like to admit that it's a complex place. As a human being, with natural tendencies toward thinking I understand stuff, I want to be able to say: Here's the theory. You claim to have the theory. Do you think it's possible in your coming book or elsewhere to make the empirical case that would convince skeptics? Of these other two stories? Guest: That's what I'm going to try to do. Obviously there's a lot to say there. I would say, the easiest one to dispose of I think is the secular decline in manufacturing. That's true, but it doesn't really explain the cycles. So instead of a straight line trending downward, which would be jobs in manufacturing, we have sort of a staircase downwards--big drops in recession, then level during the recovery, then big. You know what I'm saying? Picture a staircase going down to the right. Russ: Yeah. I think it's actually negative. I don't think it's so staircase-y. It's more intense during the recessions. Guest: Okay, maybe even slightly downward during the boom. But that doesn't really explain the cyclicality. It just explains the trend. Obviously over the long run we've been able to make up those jobs in the service sector. We had very low unemployment in 2007, as recently as that. And yet we'd lost millions of manufacturing jobs over decades. And before that, a hundred years ago we were losing millions of farming jobs, and people were moving to the cities. So, we constantly get this churning. But it's done in a gradual way that doesn't really by itself create business cycles. Now the Mulligan argument is a little bit tougher. There is a bit of truth in what he's saying. During the recession the incentives to work were reduced by things like greatly extended unemployment insurance. Almost two years at one point. Now it's a little bit less. So that probably increased what's called the 'natural rate of unemployment'--the rate that exists because of people being between jobs; they'd be a little more picky, be willing to wait a little more before taking that job because they could rely on more unemployment benefits. And other things like minimum wage could be mentioned and so on. But what I would say about that is that that doesn't really explain how we got into a severe recession. Because a lot of that was done in response to the recession. So there is a sort of entanglement between supply and demand for the economy. When you have very bad demand side policies and the economy crashes into a depression, the government reacts often in counterproductive ways and makes it worse. But still we have to prevent the cyclicality that causes that in the first place. One of my favorite examples was in the Great Depression. Or in the 1920s, the United States had a relatively free market economy. And as we went into the Great Depression because of mistakes made by people that today would be called 'conservative'--we had too tight a monetary policy, and a big crash in spending--we reacted to that failure of demand with counterproductive supply-side policies under, first, the later period of the Hoover Administration and then under Roosevelt, that made the recovery much slower than it needed to be. And Mulligan would have been pointing to those mistakes as to why we have such a slow recovery, and he would have been partly right. But I still think that the primary focus should be on stabilizing demand, because it's those demand failures that trigger things like the extended unemployment insurance in the first place. So that's kind of how I'd respond to the Mulligan argument. Russ: Are you a Keynesian? When you say, 'demand,' is there a difference between nominal and-- Guest: No, I use the term 'demand' synonymously with nominal GDP. Which is a little different from Keynesians. They don't. But I just feel that sometimes it's easier to talk about demand and supply shocks. Other times I use the phrase 'nominal and real shocks.' I mean that synonymously with supply and demand. But Keynesians define demand a little differently. They tend to think about it more in a real sense of how many goods and services are being bought. I tend to think about it in a nominal sense of like: What's the dollar amount that's being spent on goods and services in the economy? That's how I think about demand. Nominal GDP. Russ: So, my last question, which I should have asked you before: So, what happened in 2008? You argue that nominal GDP targeting is not a tautology. Why did nominal GDP fall in 2008? Guest: It was kind of a perfect storm of bad luck. So, I think you have to look at it in parts. Once interest rates hit zero, the Fed didn't know what to do. So then we want to back up and ask: Well, how did we get in that position? I think the end of the housing bubble is part of it. So, when the housing bubble burst and spending slowed very sharply, the natural rate of interest started to fall. It wasn't at zero yet; it was about 2% when Lehman failed, the interest rate. But it was falling very, very sharply. And in fact it should have been even lower. The Fed did not cut interest rates after Lehman failed, in their next meeting. It was still 2%. That was a mistake. And the reason they made that mistake was the other big problem, which is we happened to have this housing crash right at the same time as a huge oil price spike. So, because of the oil price spike of 2008, where gasoline shot up from $2 to $4 a gallon in just a couple of years, our headline inflation number was high; the Fed was frightened about inflation. So, even as nominal spending was slowing they were really grudging in easing their policy. So, by refusing to ease aggressively, they let the economy slide, partly under this oil shock and partly because of the housing crash. And by the time they realized, oops, we're sliding into a deep recession, interest rates had fallen close to zero and they didn't have their normal tool, their familiar, comfortable tool of manipulating interest rates to control the economy. And then we found the Fed didn't know what to do when it couldn't use its conventional tool of adjusting interest rates. And when the markets saw the Fed was adrift and didn't have a backup plan, the markets crashed. And that was the big crash in the second half of 2008, when markets realized: We are going into a deep recession, the Fed doesn't have a Plan B; it's going to be really bad. And the market forecasts were correct, by the way. The recession would be bad. Russ: Scott, it's always interesting to talk to you. You tell fascinating stories. And they could be right. I look forward to hearing more about them and I look forward to your book.

COMMENTS (50 to date)
Roger McKinney writes:
Sumner: “It's not at all generally accepted that interest rates are a reliable indicator.”

I think Sumner confuses intention with effect. When the Fed reduces its interest rate to zero, the Fed’s intent is a loose monetary policy. On the other hand, Sumner looks only at the effect: if low rates don’t goose ngdp, then it’s not loose policy. So it’s a difference of definition and Sumner doesn’t make it clear to people that he has a different definition of “loose” monetary policy.

Sumner: “We have relatively low inflation and we have a very weak economy. So that's why interest rates are low.”

But would they fall as low as they have? Clearly not. So when interest rates fall below what people would expect them to be given the state of the economy, they have a clear reason for thinking policy is loose.

Sumner quoting Friedman: “: Very low interest rates like in Japan--talking about Japan at the time--are a sign that money has been tight.”

Again, that is using effect as the definition of policy and not the intent of the bank or the relationship of market rates to expected rates given the state of the economy. Also, it assumes that “tight” policy (using their definition) is the only cause of recessions.

The problem with Sumner and Friedman’s way of thinking is that the central bank cannot possibly know if its policy is too tight or too loose until several quarters after a policy change.

Sumner: “Which I know goes against common sense, but it reflects the fact that when you expect faster growth of nominal spending, you don't want to sit on cash as much.”

Here Sumner promotes his fixation on expectations. But he is guilty of leaving out many variables that affect the economy, such as high commodity prices for commodity exporter Australia.

Sumner seems to think that the public hangs on every word of the Fed and believes the Fed can do what it says it will do. But after watching the Fed try to stimulate the economy with low rates and multiple QEx’s, the public naturally doubts that the Fed can do what it claims it can do.

Russ: “But it could be that the reason they are not using that money, the banks, is the economy stinks. There are not many good, attractive investments. And so Ben Bernanke can talk all he wants; he can try to change expectations.”

Excellent response. Sumner, like most mainstream economists, think people act like robots, that we have no choice in how we respond to Fed policy.

Sumner: “One is there's really no example in all of human history of a fiat money central bank trying to inflate and failing because of the lack of credibility.”

Why isn’t the past 5 years a counter example?

Sumner: “…what they've been able to do is, in 3 months, depreciate the yen dramatically…”

So Sumner is using the financial markets as evidence that the central bank can change expectations. But the financial market is not the consumer, who determines ngdp.

S

umner: “So, Australia, interestingly, was the one developed country that didn't have a recession this time around.”

But Australia lives and dies by commodities, which prices have been quite high.

What Sumner misses is the massive collapse in asset values, which makes people poorer. Creating new money will stop prices from falling and may even cause inflation. But it can’t make people spend more or invest more. People must see the real economy improving in order to do those.
Sumner places too much confidence in the Fed’s ability to affect expectations. He should read Lachmann’s writings about the elasticity of expectations: expectations are inelastic as long as prices stay within an indifference zone. They become elastic only when prices break through one of the barriers.

Sumner: “: It was kind of a perfect storm of bad luck.”

That’s all we get from mainstream econ in general.
It's all luck.

Russ: “Do you think it's possible in your coming book or elsewhere to make the empirical case that would convince skeptics?”

Excellent question. The way Sumner has set up his theory, there is no way to test it that I can see. You can’t test it with historical data because he has defined of loose vs tight monetary policy by the effect instead of the intent, so if ngdp does not respond to Fed intentions, then the Fed policy was too tight even if the Fed thought it has a loose policy stance. And as with Australia, comparing countries doesn’t work because he leaves out too many other variables that determine economic outcomes.

AD writes:

Always glad to see Sumner return, it's a very important debate even if we cover a lot of the same ground.

I'm confused -- as I believe Sumner is as well -- as to why NGDP targeting is not more accepted in the profession or by the Fed. Is it because they are unwilling to accept temporarily high inflation (say 1% real plus 4% inflation to get 5% NGDP)? Why didn't Milton Friedman, for example, advocate it as being consistent with his analysis of the Great Depression and other recessions? Surely he knew velocity wasn't always going to be stable?

Also, related to the last question in the podcast: Is Sumner's view that the housing crash/financial crisis would just have led to a garden variety recession in terms of real GDP if it wasn't for the monetary policy errors?

Silly sailor writes:

Sorry for showing my lack of expertise. Why can't central banks make interest rates negative?
Why can't they loan out a hundred and get ninety nine back?

Silly sailor writes:

Presumably as we don't have deflation this is what happens anyway. A zero percent rate is something like a negative three percent after inflation. So why not go further and set the rate below zero?

What am I missing?

Why can't central banks make interest rates negative?

Because interest rates aren't 'the price of money'. And attempting to lend out 100 and get back 90 would bankrupt the central bank. You'd get hyperinflation.

Though it isn't quite true that nominal interest rates can't be negative. They were slightly negative for a short period in Japan back in the 90s. American banks once, long ago, charged fees for checking accounts, which is kind of a negative interest rate.

Also, the Fed could charge a negative interest rate on excess reserves banks hold at the Fed. But, those banks would simply stop depositing such reserves there.

The way Sumner has set up his theory, there is no way to test it that I can see.

Sure there is. Set up an NGDP Futures Market and trade the securities he wants to create. Robert Shiller has already been experimenting with something like that in housing securities.

Shiller has also proposed a security worth a trillionth of NGDP that he calls a 'Trill' that seems to me could be adapted.

Btw, most of your criticisms of Sumner, Roger, aren't really valid. He's already answered them in his talk with Russ.

Nick Zbinden writes:

Good Podcast.

I, after a long time, have come to accept the market montarist story. Its seams to me the only story that fits the facts.

What I would suggest to you to talk with somebody like George Selgin who more or less agree with market montarism but comes from another direction. The Austrian Free Bankers would explain these theorys from a micro economics standpoint.

Sumner nicly point to historical examples and it all makes sence but without micro something is missing for me.

I would suggest a podcast with Selgin on montary (dis)equillibrium theory.

James Law writes:

Thanks for another great econtalk. I have a question regarding NGDP targeting. In market monetarism as Dr Sumner describes, if government spending were to actually go down (say, due to the sequester, or other means), would the money supply be increased to make up the difference, since NGDP would fall? Or do market monetarists not include government spending in their calculation of GDP?

thanks
James

JVM writes:

@JamesLaw

> if government spending were to actually go down (say, due to the sequester, or other means), would the money supply be increased to make up the difference

Yes.

This means that market monetarists believe the fiscal multiplier is 0.

emerich writes:

This is Econtalk at its best:(1) the topic is hugely relevant to what's just happened and to what's going on right now in the U.S. economy; (2) the guest's diagnosis and prescription are controversial; and (3) he may be right. Or at least mostly right, which is right enough for me.

Scott Sumner writes:

Roger, One can define the stance of monetary policy as you like, but I think it's fair to say you are about the only person whio uses "intent." I follow Milton Friedman, Ben Bernanke, Frederic Mishkin, and many of the other leading monetary economists by looking at effect, not intent.

Australia has run massive balance of trade deficits in recent years (bigger than the US as a share of GDP.), so it hardly seems likely that commodity exports were what kept them out of recession.

One can know the effect of monetary policy immediately, just look at asset prices (TIPS spreads, etc.)

The Bank of Japan has sharply boosted equity and forex prices with merely a rumor of possible monetary easing--so credibility is the least of our problems. No fiat money central bank has ever tried to inflate and failed. And that includes the Fed.

BTW, In the previous 5 years the BOJ was not trying to inflate, so that also supports my analysis.

Robots? I focus on expectations---you should direct your comments to those who don't believe in rational expectations.

Yes, the financial markets are not the consumer, but the consumer is highly responsive to the financial markets.

You said "Australia lives and dies by commodities." That's not even close to being true. Anyone with a passing knowledge of Australia knows they haven't had a recession since 1991, despite sharp commodity price declines in the East Asia crisis, and again in 2001-02

You said:

"What Sumner misses is the massive collapse in asset values, which makes people poorer. Creating new money will stop prices from falling and may even cause inflation."

I focus heavily on asset prices. The tight money of 2008 caused much of the asset price collapse. And how could be possible that monetary stimulus raises consumer prices without affecting asset prices?

Of course my theory can be tested, if you don't think so, then you need to learn more about my theory.

I'd suggest doing some research on countries like Australia and Japan, and not relying on what you read in the newspapers, which is mostly false.

Scott Sumner writes:

AD, If that's their fear, they are mistaken. Right now 5% NGDP growth would lead to 3% RGDP growth and 2% inflation. If trend growth has fallen to 2%, as I believe, eventually those numbers would reverse.

I would have expected a mild recession, at worst. But a long period of sub-par growth.

Silly, Cash keeps nominal rates from going below zero.

Nick. Yes, Selgin is excellent.

James, Not only can monetary stimulus prevent fiscal austerity from reducing NGDP, it has done so (so far) this year. The recent tax increases have not slowed NGDP growth.

Matěj Cepl writes:

This was a very interesting podcast, but I have to say I am not impressed much. I completely agree that expectations matters. I believe they matter a lot. However, these are not expectations about the future central bank policy, I don't think anybody outside of the academic economists cares about that. It seems to me that prof. Sumner assumes quite direct link between the policy and true state of economy, meaning especially future GDP or employment.

I think the expectations which matter are the expectations about the future state of economy as whole. And contrary to what I understand prof. Sumner to believe, there is not much expectations that actually even the central bank’s actions matter that much.

Matěj

David writes:

@Scott Sumner
If our country shifted from fiscal policy to a more momentary dominated policy how much more scrutiny would a central bank be under from the public and politicians? Would this scrutiny change the incentives of central bankers? would outside forces (lobbyist) have more influence on the central bank?

Seth K writes:

Russ, first of all this was one of the best podcasts you've hosted lately. Very interesting discussion.

However, as you seemed to be, I am really struggling with Sumner's argument that by simply telling the general public that there is a 5% GDP target it will happen by some self-fulfilling prophecy.

Certainly, we all know that the market reacts to expectations about future Fed actions. But as you mentioned it presumably does that because it has faith that the Fed can and will meet those expectations, and I just don't see how Sumner's proposition could generate sufficient faith. When the Fed sets an interest rate target, the path to that target is very short and well worn -- they buy or sell Treasuries until the desired rate is reached. However, if the Fed instead sets a GDP target, the route to that target is less clear. One has to assume that cheaper money will trigger more borrowing which will trigger more spending in productive sectors that will then boost GDP. This extra indirection introduces a lot more variables to the problem that should certainly reduce market confidence in the statement. It's also a deviation from typical Fed policy which in itself should make the market nervous.

What further perplexes me is Sumner's insistence that fiscal policy should not be used in combination with monetary policy to ensure that the stated GDP target is hit. It's as if he's making a suspect promise even more suspect by narrowing the toolkit he is willing to use to deliver it. Which is confusing if he genuinely believes he won't even need to use that fiscal policy tool because the promise alone is good enough to have the desired effect. Anyone's individual beliefs about the overall effectiveness of fiscal policy aside, the fact is that the government can directly increase GDP in the short term by putting people to work with economic stimulus, right? So why isn't this an acceptable plan B for the Fed to promise if interest rate manipulation does not have the desired effect?

It seems to me that a much less worrisome commitment from the Fed would be something along the lines of the following: "The Fed is committed to a rate of GDP growth of 5%. All of the tools in our monetary policy toolkit are on the table to help us reach that goal. However, should these tools prove insufficient, we have already reached an agreement with Congress to enact appropriate fiscal stimulus to achieve our goals."

Todd S writes:

Scott, et al,

Certainly fiscal policy has to be within a certain range of "reasonableness" for this idea to work, correct?

If a series of disastrous economic policies were adopted by Congress that are likely to negatively impact real long term GDP growth (90% tax rates, high tariffs, unlimited unemployment benefits), maybe 5% nominal growth is not a good target anymore.

Or do we stick with 5% and simply make up the difference with higher inflation?

Wouldn't it be better to fix the bad policy?

Would appreciate any thoughts.

Thanks.

-todd

Richard Fulmer writes:

I don’t understand how monetary policy can revive the country’s “real” economy regardless of the economic conditions. What about the increased costs of employing workers given the new regulations and Obamacare? What about uncertainty caused by the serial “crises” fomented in Washington (e.g., fiscal cliff, sequester, debt ceiling)? What about the impact of an activist EPA that is pushing carbon taxes? Do things like these simply not impact people’s decision whether to spend their money or sit on it? Or is their impact dwarfed in comparison to the impact of increasing or decreasing money supply?

Also, the Fed is pushing money into the economy not just by buying assets from banks through its QE programs, it’s also buying bonds from the government to finance its budget deficits. And that money is being spent. Granted most of that money ends up in banks along with the money used to purchase bank assets.

Someday, however, people will start spending and borrowing again. How will the Fed reel all the money back in to avoid inflation? Require that banks buy government securities? Raise the reserve ratio? Can the Fed calibrate the timing and the volume of this “reel-back” so as to avoid either recession or damaging inflation?

When the Fed sets an interest rate target, the path to that target is very short and well worn -- they buy or sell Treasuries until the desired rate is reached.
Actually they first announce their targeted interest rate, THEN they add or subtract reserves to influence the Fed Funds Rate. It isn't clear that their announcement alone isn't enough.

Everyone should also keep in mind that Scott has been arguing the same way for almost five years now, and hasn't been proved wrong in all that time. He said paying interest on excess reserves was a mistake; it seems to have been.

He said Operation Twist would be ineffective; it was. He said QEII would be slightly better than nothing; he was right again.

AD writes:

How about getting the other prominent market monetarists on, now that you've got a pretty good handle on it and see how they respond to some of your questions?

You've got Lars Christensen, Marcus Nunes (who just released a book), David Beckworth (also released a book of essays), a bunch of others that I'm forgetting

PhilH writes:
Anyone with a passing knowledge of Australia knows they haven't had a recession since 1991, despite sharp commodity price declines in the East Asia crisis, and again in 2001-02

There are undoubtedly many reasons for this, but principal among them is a very large increase in property prices and borrowing after the late 1990s, growth which has only recently declined.

The importance of this to the Australian economy is demonstrated by the government's response to the GFC, part of which was to double the cash payments made to people buying real estate. That, combined with reduced interest rates, caused Australian real estate prices to rise sharply, at a time in which property prices were falling in many other countries. It seems probable that this supported consumer activity and helped keep the economy out of recession.

Of course, this didn't occur in isolation. The government response to the GFC also included cash payments to low-income earners and a large program of public works. We also got lucky with a large increase in mining investment, which had already started and is only now peaking.

Bogart writes:

I hope the Fed does not target some nominal rate of growth in GDP, I really hope it fires itself and quits because one government bureau controlling the price of anything is immoral. And that is where I really feel the Keynesians and Monetarists fail. There is nothing moral about the backdoor theft of wealth through inflation. And there is nothing worse that trying to argue that a person being robbed is a good thing because it makes them feel like they have a Central Robber who will get them to a 5% GDP target.

I never heard in the talk what to do when you have a positive nominal growth rate and a negative real growth rate? What if the nominal rate is 5% but the real rate is -2%? Do you put the hammer down to get a 6% growth rate? Do you stop the presses and hit the breaks?

But the best part of the talk was the Mr. Sumner's simple yet eloquent description of the Gold Standard. Say what you want to about a Gold Standard and nominal anything. At least with a Gold Standard individuals can not get robbed easily through inflation.

Brian writes:

Why do many economists believe that it is better for the market to set wage rates yet feel confident that Central bankers know the true cost of perhaps the most important element in our economies – the cost of capital?

SaveyourSelf writes:

Mr. Sumner gave a nice talk. He demonstrates a strong understanding in many respects but also gave a few contradictions.

First, he said that he is free market oriented. He then spent an hour talking about how the federal reserve--ie government--could best manage the money supply to solve the problem of episodic, society-wide unemployment—ie. recession. Obviously, a government intervention is not a free market solution.

Second, he claims he is in close agreement with Milton Friedman's monetary positions, but then argues that the Federal Reserve should use NGDP targeting as its modus operandi. NGDP = Real GDP + Inflation. Sumner is proposing, therefore, that the Fed inflate the money supply to mask any real downturns in the economy. Presumably, he is suggesting that if businesses don't realize that demand for products has fallen, they won't know to slow their production/cut back on work hours. Friedman's stance in "Free to Choose" was that the best central bank policy was a ZERO inflation policy. Freidman argued that society could compensate for a known, stable inflation rate over a long period of time but that any such compensation is cumbersome and expensive when compared with a zero inflation system which requires no extra work. By targeting NGDP, Sumner is proposing a constantly fluctuating rate of inflation, which I do not think Mr. Friedman would have endorsed.

On the positive side, I appreciate Mr. Sumner's ideas about Fed policy nullifying the Legislature’s stimulus projects. That makes a lot of sense given the Fed rules over the components of the money supply--the banks--which are known to have a leveraged advantage on the money-supply compared to changes in the quantity of money, which is the portion of the money supply affected by the "stimulus". Sumner did not touch on recent legislation limiting banking activities, which presumably have the same leveraged power to affect the velocity of money as the Fed albeit only in the negative direction. Thank you, Dr. Roberts, for making some references to your pervious podcasts where that fact was discussed.

On the whole, I find it hard to differentiate the Keynsian position on inflation from Mr. Sumner’s. It just seems like he is wrapping it beneath a different name.

Dan Hanson writes:

If I understand Scott Sumner correctly, he's saying that the conditions for growth will be set if the central bank makes a credible promise of x% NGDP growth, and uses expansionary monetary policy to ensure it. The key is to have a stable target and to convince actors in the economy that the fed will continue to carry out this policy.

Isn't that basically the situation the economy faced in the 1990's? I recall Alan Greenspan saying that he could hold interest rates lower because he believed the economy had entered a new 'high growth' phase with no end in sight. The internet bubble was the result of expectations of continued high growth in the internet economy. Everyone believed the good times were here to say, so investment skyrocketed.

It seems to me that what happened was that this near-universal belief in continued high growth simply ran into the brick wall of reality. This is what happens to most bubbles. At some point, there's a disconnect between what you want to believe and what's actually happening, and when the market absorbs this new reality, the bubble pops and investment crashes.

So perhaps someone can explain how Scott would avoid this scenario: The central bank short-circuits the price signal of capital, causing malinvestments. These malinvestments cause the real economy to lose efficiency. Growth falls. Inflation declines. Velocity falls because there's no where to invest reasonably, so the money supply shrinks. NGDP collapses. So the Fed inflates the aggregate money supply. But as it does, velocity falls even more because no one wants what people have to sell.

Or perhaps NGDP comes back a little through price inflation or even more malinvestment due to even cheaper capital. The central banker says, "See? It's working! We make money cheaper, and NGDP grew." But what in fact happened was that the central bank masked real problems of coordination in the economy and pushed the day of reckoning down the road while making it worse.

Or perhaps stated more simply with a thought experiment:

Let's say an economy had a supply side that made nothing but tables and a demand side that wanted nothing but chairs. Because of the disconnect between what producers are making and consumers want, demand falls and the economy declines. Velocity drops and the money supply shrinks. How could monetary policy possibly fix this? If you just look at the aggregates, all you see is demand falling and layoffs in production. How would it help to inject more money into this economy?

Michael Byrnes writes:

Seth K. wrote:

"However, as you seemed to be, I am really struggling with Sumner's argument that by simply telling the general public that there is a 5% GDP target it will happen by some self-fulfilling prophecy.

Certainly, we all know that the market reacts to expectations about future Fed actions. But as you mentioned it presumably does that because it has faith that the Fed can and will meet those expectations, and I just don't see how Sumner's proposition could generate sufficient faith. When the Fed sets an interest rate target, the path to that target is very short and well worn -- they buy or sell Treasuries until the desired rate is reached. However, if the Fed instead sets a GDP target, the route to that target is less clear."

It's important to emphasize that Sumner is talking about nominal GDP, not real GDP. Thus, all that is really required is that the central bank has the capacity to inflate, if nominal GDP is below target, and to rein in inflation, if nominal GDP is above target (as in the 70s ). I don't think there is much reason to doubt either point.

"What further perplexes me is Sumner's insistence that fiscal policy should not be used in combination with monetary policy to ensure that the stated GDP target is hit. It's as if he's making a suspect promise even more suspect by narrowing the toolkit he is willing to use to deliver it."

Sumner's argument is that monetary policy can offset fiscal policy. The Fed engaged in QE and QE2 to prevent deflation. Had ARRA been smaller (or nonexistent) the Fed would have needed to take more aggressive action in order to achieve its goal (no deflation). So it would have done so. (Bernanke has steadfastly maintained that the Fed was never out of bullets; in the absence of ARRA he would have used more bullets). Had ARRA been bigger, the Fed would have done less QE - since ARRA would have had a greater impact on GDP, less QE would have been needed.

Michael writes:

Saveyourself wrote:

"First, he said that he is free market oriented. He then spent an hour talking about how the federal reserve--ie government--could best manage the money supply to solve the problem of episodic, society-wide unemployment—ie. recession. Obviously, a government intervention is not a free market solution."

I think there are two issues that tend to get merged in criticisms of NGDP targeting such as the above.

At present, the US money supply is managed by a central bank (the Fed) via its monopoly on the creation of high powered money (currency and bank reserves) and the reserve requirements that banks must meet. The Fed is not just a quasi arm of the government, it is also the monopoly provider of base money.

This presents two obvious questions:

1. Is this a good system for controlling the US money supply?

2. Given this monetary system, what should the Fed's objective be, and what strategy should the Fed use to achieve its objective?

Sumner's argument should be seen as an answer to question #2.

The monetary system in the US is, for better or worse, not really a free market system. US banks do create money, but they are limited in their ability to do so by their reserve requirements and the Fed's monopoly on the monetary base.

If we had a functional monetary system where currency was issued by private banks, then as people opted to hold more currency (rather than other assets), banks would issue more currency. As people chose to redeem more currency rather than holding it, banks would issue less. (Issuing "too much" would lead to insolvency so banks would have an incentive not to do so). The "correct" amount of currency would emerge via market forces, and currency would hold its value effectively. There would also probably be more bank failures, and I don't understand the pros and cons of such a system well enough to have an opinion on whether it would be better than a central bank.

Regardless, we don't live in that world. Like it or not, the Fed controls the monetary base, and through it, the money supply. Given this position, what should the Fed do? Sumner and the market monetarists argue that level targeting, preferably of nominal GDP, is what the Fed should do. Specifically, they argue that the Fed's goal should be a steady, predictable growth of *nominal* spending on output that is consistent with a reasonable long run inflation target. In effect, this is as close as the Fed is likely to be able to get to a functional free banking system. Under such a system, as with free banking, changes in people's preferences to hold currency would not impact the price level.

Hwere is what I think would be a really interesting Econtalk podcast: "Selgin on Market Monetarism". As a proponent of free banking and an opponent of the Fed (see Selgin's podcasts from the Econtalk archives), I think he would offer an interesting - and not 100% opposed - critique.

Michael Byrnes writes:

Seth K. wrote:

"However, as you seemed to be, I am really struggling with Sumner's argument that by simply telling the general public that there is a 5% GDP target it will happen by some self-fulfilling prophecy.

Certainly, we all know that the market reacts to expectations about future Fed actions. But as you mentioned it presumably does that because it has faith that the Fed can and will meet those expectations, and I just don't see how Sumner's proposition could generate sufficient faith. When the Fed sets an interest rate target, the path to that target is very short and well worn -- they buy or sell Treasuries until the desired rate is reached. However, if the Fed instead sets a GDP target, the route to that target is less clear."

It's important to emphasize that Sumner is talking about nominal GDP, not real GDP. Thus, all that is really required is that the central bank has the capacity to inflate, if nominal GDP is below target, and to rein in inflation, if nominal GDP is above target (as in the 70s ). I don't think there is much reason to doubt either point.

"What further perplexes me is Sumner's insistence that fiscal policy should not be used in combination with monetary policy to ensure that the stated GDP target is hit. It's as if he's making a suspect promise even more suspect by narrowing the toolkit he is willing to use to deliver it."

Sumner's argument is that monetary policy can offset fiscal policy. The Fed engaged in QE and QE2 to prevent deflation. Had ARRA been smaller (or nonexistent) the Fed would have needed to take more aggressive action in order to achieve its goal (no deflation). So it would have done so. (Bernanke has steadfastly maintained that the Fed was never out of bullets; in the absence of ARRA he would have used more bullets). Had ARRA been bigger, the Fed would have done less QE - since ARRA would have had a greater impact on GDP, less QE would have been needed.

Matt P writes:

Is there short, beginners-level reading I could do before giving this a second listen? I am so lost.

Recommended reading?

Greg G writes:

Fascinating and thought provoking podcast. I am skeptical that NGDP targeting can do everything that is claimed for it here but I am entirely convinced that this is the most interesting original new economic theory we have seen in some time.

Thanks to Scott for being such a tireless, patient, and articulate advocate for what he believes in and thanks to Russ for asking the kind of challenging questions that helped Scott to explain his ideas in such a clear way.

Even if you agree (and I do) that the collapse of the shadow banking system was a deflationary shock that precipitated the Great Recession, does it really follow that re-inflating credit creation to the level of a bubble economy is the best medicine? Doesn't that carry a high risk of more bad investments and another round of bank failures?

When have we ever seen more investment money chasing fewer promising investment ideas? When have we ever seen our most profitable and innovative companies sitting on more cash without a clue how to profitably put it to work? Can the solution really be yet more credit creation? I am more inclined to see this all as a Tyler Cowan type Great Stagnation where we are at a point in the technology cycle where it is more difficult than usual to create jobs.

Too much income and wealth inequality can be a drag on an economy. A shortage of funds available for investment can also be a drag on an economy. The Left tends to like fiscal solutions because they more easily put new money in the hands of people with lower incomes. The Right prefers monetary solutions because they tend to put new money in the hands of the investor class.

Maybe NGDP targeting would be better than the monetary policy we have had. Even if it is, that doesn't prove there is no useful role for fiscal stimulus in a depressed economy.

uncarvedblock123 writes:

While Sumner claims not to be a Keynesian, and his positions often support this claim, he also seems to rely a great deal on "animal spirits" for his theories. His insistence on success being driven by the message implies that nearly any policy approach can be successful if it's sold properly. While having a clear understanding of policy and direction is always preferable, knowing that bad policy is being employed doesn't increase the likelihood of a positive result. As with any policy, it makes likely a response toward optimal outcome for the participants, which may be completely at odds with the intention of the policy.

mariusz writes:

Russ,

Good podcast, I think Jim Manzi and a tutorial in high causal density should be a required exercise for all guests (including Ed Leamer before he starts peddling education as a cure all potion).

To echo what Dan Hanson so well says above the housing price inflation and its subsequent collapse seems to me like an illustration of what we get with Scott's proposal - everybody believed it will grow forever (at whatever rate), Greenspan put provided institutional support, and then it all imploded.


As regards Ben pre-Fed and in-Fed - it seems to me it is "a skin in the game effect". In-Fed Ben has to convince all the other parties with skins-in-the-game to cooperate and has to consider that he will be held to account for what are effects of the proposed policies. Hence safety first and no more blackboard theories.

Oh, sure, Matt P, ask the impossible!

First, take a breath and congratulate yourself on your courage and honesty. To understand what Scott Sumner and others are actually talking about when they discuss macroeconomic policy--including Federal Reserve policy and "the economy" as a whole--is very hard going. To really get what Scott is recommending requires understanding the Keynesian model, rational expectations, national income and Federal Reserve bank accounting, the historical behavior of the Federal Reserve and its chairmen, and the complex line of empirical testability between economics and politics.

If you are lost or feel like most of what you understand is just because you are riding from crest to crest on some waves of understanding that you can stay atop, rest assured that so are many other listeners just riding those same crests.

Now: what to read to dig deeper? Starting from the top, where did you start to get lost? To suggest more readings for you--in addition to the ones we list in the Readings--I need to know which pieces of the puzzle were hardest for you. Your answer is going to help a whole lot of folks dig deeper.

Michael writes:

Matt P...

Here's a good article from Sumner explaining his ideas:

http://www.nationalaffairs.com/publications/detail/re-targeting-the-fed

And here is a recent post from Sumner on the effects of nominal vs. real shocks:

http://www.themoneyillusion.com/?p=20114

Also, google "David Beckworth Ramesh Ponnuru" and you will find links to a couple of articles on NGDP targeting.

For those asking about further reading, start with Scott Sumner himself.

Pyrmonter writes:

Very interesting, if sometimes unclear, talk. Why isn't this just another version of the Phillips Curve - ie a potentially unstable trade-off? What happens when inflationary expectations move from current levels (I assume 1-2%) to the nominal growth speed; or exceed them?

I find the story about Autralia rather curious - it certainly isn't one that has widespread currency among Australian business professionals. Also find it a bit difficult to reconcile with this:

http://www.rba.gov.au/chart-pack/credit-money.html

Could Australia's success be attributed more to bouyant commodity prices?

Greg G writes:

@ Patrick R. Sullivan

Thanks for that link to FAQ with Scott. Very interesting.

He says that if banks can't find credit worthy borrowers they can just buy Treasuries. Well they "can" do that but that doesn't mean they will do that when Treasures come with huge interest rate risk and returns inadequate to produce a profit.

I still think banks are much more likely to respond by simply making loans that are too risky but have the potential to save management's jobs if they work...or become the taxpayers problem if they don't.

Matt Cres writes:

I think Scott has the causality the wrong way round in his comments concerning Australia.

It is true that Australia has run, for a long time, a large current account deficit. But this does not prove that mining played no role in maintaing the economy throughout the financial crisis. The reason for the high CAD is Australia's low savings, which means we must import capital. The balance of trade, by contrast, is frequently in surplus or in small deficit. However the main reason for the strong nominal income in recent years has been the dramatic increase in Australia's terms of trade. This is particularly evident in the price of iron ore, and to a lesser extent coal, Australia's two biggest commodity exports. This has been documented in numerous places. It has kept corporate tax receipts strong and allowed the government to keep the budget in relative balance (until quite recently).

It is only recently that commodity prices have retreated and this has sent Australia's terms of trade, and nominal income into reverse. See: http://www.rossgittins.com/2013/03/why-tax-revenue-is-falling-short-of.html

Also, I don't think the Reserve Bank of Australia's inflation targeting regime is much different from anywhere else. That is, 'core' inflation of 2-3% over the course of the business cycle. No mention at all of nominal gdp.

The final point about Australia is that its strong fiscal position allowed it to undertake a fiscal stimulus worth about 1% of GDP at the height of the crisis in early 2009. I think that helped.

Changing topics slightly, an alternative and interesting take on Japan is presented by Richard Koo. I urge Russ to consider inviting him on to the show.

Having said all that, another fascinating econtalk.Keep up the good work.

Michael writes:

Pyrmonter wrote:

"Why isn't this just another version of the Phillips Curve - ie a potentially unstable trade-off? What happens when inflationary expectations move from current levels (I assume 1-2%) to the nominal growth speed; or exceed them?"

Isn't the idea of a Phillips curve that there is a stable tradeoff between inflation and unemployment, such that unemployment can always be reduced by tolerating more inflation? Scott Sumner and the market monetarists argue no such thing.

Under a 5% NGDP level target, the inflationary 70s would not have been possible. (During the 10 years before the Volcker Fed brought inflation under control, I don't think NGDP growth ever fell as low as 5%, and it was over 10% for much of the decade).

I don't think there is much reason to think that inflation expectations would approach or exceed the nominal growth rate. This wasn't true even in the 70s, when real GDP growth was about 3%.

In the event of a temporary fall in real GDP, it would be better to keep nominal spending on track (at a cost of higher inflation) than to maintain an inflation target by cutting nominal GDP. Flucutating NGDP is far more destabilizing than short term variation in inflation (around a stable long-term target).

If you want to buy a house, for example, what really matters to you (and to your lender) is long run inflation.

Pyrmonter writes:

@ Michael wrote:

Isn't the idea of a Phillips curve that there is a stable tradeoff between inflation and unemployment, such that unemployment can always be reduced by tolerating more inflation? Scott Sumner and the market monetarists argue no such thing.

I wasn't meaning to suggest he was (although he does seem to want to use measured unemployment as a macro variable - something I'd have thought a bit of a throwback: there are lots of ways it can vary for reasons quite unrelated to overall "aggregate" economic activity - "structural" unemployment; sudden bad nominal wage contracts that "over-shoot" the market; also the effects of changes in labour market participation reflecting working age people "giving up" for examples). It does seem to me though that his market monetarism assumes some sort of "knowable" basis for determining a relationship between changes in the money base and nominal GDP - something which I wouldn't have thought to be any more stable than "V" in the old "monetarist identity" MV=PQ. Even if there HAS been a stable -ish relationship while the Fed/Central Bank has focussed on non-financial asset and current consumption good price stability, wouldn't you expect a change of regime to change that relationship?

Separately, it seems Sumner assumes assymetric price-stickiness: nominal prices go up freely, but not down (this is fairly standard, but I think, as some of Russ's past guests have observed, only since 1921 in the US, and perhaps a little earlier in the case of the UK, Canada, Australia and other countries that tried to, but never did, disinflate to pre WW1 levels). If that's the case, sure he could see that some key commodity price - eg the price of energy, or in the case of countries smaller than the US, "Imports" as measured in domestic currency - might move by enough that price expectations would rise above his nominal GDP growth rate.

In that case, policy makers would have either to (a) accept a fall in nominal GDP for "demand management" purposes independent of the real effect of the adverse price movement or (b) "cheat". Given the experience of the 70s, and of other countries (UK comes to mind in particular) in getting policy credibility, how long could "credible" policy be maintained?

Finally: how does this square with recent UK performance? I thought the BoE had announced sequentially very very large monetary base increases, and yet the UK seems to be the complete laggard in having not moved back toward its long term growth path.

I ask these questions not because I think Scott's wrong - I find the arguments interesting, challenging, and, as Russ R said, "elegant" by seeming to inject some life into macro models I had thought died at the start of the "Great Moderation", but because I'd like to see more on the topic - an open economy model for example being a good starting point.


Corrie writes:

What exactly is "free market" about central banking? Other than the author alluding to himself in the beginning as a believer in free markets, I heard no evidence of his free market credentials.

Jim Bryan writes:

I had a hard time with the podcast, it seemed to imply that monetary policy can manipulate or drive GDP in an affirmative way. I would contend that FED policy can smother growth, but cannot force or “cause” growth. In other words, the FED can create a growth environment, but this is all. The FED cannot create wealth.

I believe the problems with the US economy have been caused in part by the following:

1. Dependence on foreign energy and the transfer for resources to other countries starting 30+ years ago.

2. Growing government and regulations on the economy.

3. Excessive government spending on military ventures.

4. Trade deficits and out sourcing overseas. This also transfers wealth (see point 1).

These problems having been weighing on the economy for years and we have used government deficit spending and FED policy to avoid really dealing with these problems. Now we are paying the price for NOT dealing with them. I believe this recession/depression was unavoidable. In fact, the deficit spending and FED policy delayed the recession and made the correction larger/ worse. This a period of change or restructuring for our economy, we have over consumed until we can’t any longer. Australia is not suffering because their economy is producing things for export to China, not because of monetary policy. If the need for raw materials in China from Australia drops dramatically, they will be forced to suffer and retool their economy.

What the FED is doing to savers in our economy will turn out poorly, and borders on being criminal. They punish savers and try to push these savers into risk assets. When the bond and stock markets correct, a lot of people will be hurt very badly. I personally have about ½ of my net worth in the stock market and I have tried to be defensive because I don’t believe the stock market levels are based on fundamentals, but the alternative is to let inflation chew away at my net worth.

Message to the FED! The problems are rooted in government policy, regulation, and structural issues. You cannot fix this, but in spite of your actions, it will correct over time.

Scott Sumner writes:

I just have time for a few today:

David, Actually we shifted to a monetary policy-dominated regime back in the 1980s. The Fed was responsible for the so-called "Great Moderation" and fiscal policy played no role in bringing inflation down to 2%.

Seth, I certainly don't believe in "self-fulfilling prophecies". The Fed must adjust the money supply so that it equals money demand when expected NGDP is on target. There's no magic involved, any more than magic is involved with the 2% inflation target the Fed has maintained for several decades. I simply want them to switch from inflation to NGDP, and from growth rate targeting to level targeting.

Once the Fed sets monetary policy at a level expected to produce on-target NGDP growth, what possible role could fiscal policy play? Fiscal policy would only be useful if monetary policy was not able to equate the policy target with the policy forecast, i.e if the Fed ran out of ammo. Unless there's a shortage of paper and green ink I don't expect the Fed to ever run out of ammo.

Todd, the purpose of monetary policy is not to solve problems, especially problems such as those you identify. the purpose is to avoid creating new problems. Unstable monetary policy can create new problems, whereas stable NGDP growth will at least avoid monetary problems. But we'll still have all our other problems.

Brian, I'm certainly opposed to the government setting interest rates.

SaveYourself, You haven't characterized my views or Friedman's views accurately. Friedman once advocated targeting a stable growth path for M. That's much closer to NGDP targeting (stable MV growth) than it is to inflation targeting. In the late 1990s he argued the main problem in Japan was that the central bank had let NGDP growth slow from 5% to 1.5%.

Greg, I agree that the government should not try to reinflate credit creation. In my blog I've been very critical of government programs like FHA, aimed at reflating the housing bubble.

Matt, The RBA doesn't officially target NGDP, but their policy effectively produces much higher NGDP growth than in other developed economies. That's why their interest rates never hit the zero bound.

Dr. Duru writes:

Roger McKinney pretty much summed up how I would have critiqued Sumner's explanations, especially regarding this blurring of cause and effect and convenient ex ante story-telling that dances around other economic impacts in places like Australia and Japan.

I will add that I have listened to all the podcasts with Sumner, and I believe this is the FIRST time I have heard an explicit answer to the criticism that Sumner's recession story is a tautology. It remains unsatisfactory but at least I we finally hear Sumner telling the story of the housing collapse, etc.. as a cause and not the effect of a recession. (I believe Fama on a podcast DOES stick to the tautology story. That is, housing would not have collapsed if a recession had never happened).

I remain baffled by the implicit claim that interest rates fell only because of market mechanisms even though the Fed was at the same time lowering rates. I believe I heard Sumner say that interest rates fell to zero and at that point, the Fed couldn't figure out what else to do. Given the Fed was furiously cutting rates in 2008, I think it is a stretch to suggest that rates were leading the Fed to zero. Regardless, there is some 2-way street going on, and I find it odd that Sumner does not address the ENTIRE series of rate cuts the Fed made in response to economic malaise.

Finally, it is odd to hear someone say that because no Fed has ever lost credibility, then Fed statements will always be credible. Sumner's theories seem extremely fragile to me relying on sentiment and Fed credibility. I would have loved to hear a thought experiment about what could happen the moment a central bank loses credibility. Are we to believe the market will suddenly lose its ability to function properly? I doubt it.

Dr. Duru writes:

Addendum to my earlier comments as I did not see Sumner's response on Australia before writing. I am going to go back to the data to see whether Australia's GDP *growth* is correlated with commodity prices. A collapse in commodity prices need not cause a recession for the country, especially if they are acting to work against it, but the rapid decline in price of a country's most important exports should slow growth as well as *exacerbate* Australia's long-standing negative account balance (it is not just recent I believe). If I am wrong, then yes, Australia has no dependence on commodities. Odd then that the RBA in its statement on monetary policy keeps warning us that the balance of trade has peaked along with commodity prices and the last domino to fall will be a steady decline in investment in the resources sector. By reducing rates, the RBA seems hopeful it can support the non-mining sector enough to step into the coming gap. The RBA believes commodities are extremely important. Not sure why we should not believe its statements on that.

Dr. Duru writes:

Ooops. A correction. I meant ex post, not ex ante. Sorry for all the posts. I also should have added that i am intrigued by the idea of a futures market on nominal GDP. It would be a great way of testing out Sumner's ideas. I wonder whether there would be a way for the Fed to experiment a little before committing...? I am also wondering whether in times of aggressive easing, whether the Fed might figure out ways to try to manipulate THAT market too....

Matt Cres writes:

Scott, you missed my point about Australia. You say the "RBA doesn't officially target NGDP, but their policy effectively produces much higher NGDP growth than in other developed economies."

It's true that the RBA's inflation target may allow a little higher nominal GDP growth (and inflation) than other developed economies. However, the main reason interest rates did not hit the zero bound, and why the economy did not experience recession, is because the economy received a massive nominal GDP boost from income related to commodity exports (in particular through a dramatic increase in terms of trade), not because of monetary policy. The relatively higher interest rates in Australia (and NGDP) are the effect of the higher commodity prices. The causality is this: (1 commodity income boost, leading to (2) higher nominal GDP, leading to (3) relatively higher interest rates (ie. no need to ease monetary policy to the same extent as elsewhere).

There is plenty of evidence for this hypothesis. Even the RBA acknowledges the role of commodity prices in boosting NGDP. In 2011 the Governor of the RBA said: "With the terms of trade at their current level, Australia's nominal GDP is about 13 per cent higher, all other things equal, than it would have been had the terms of trade been at their 100-year average level." (http://www.rba.gov.au/speeches/2011/sp-gov-230211.html)

I would like you to address this point please. Your theory seems to be contradicted by Australia's monetary authorities.

Looks like were about to enter a near laboratory experiment of Scott's ideas, in Japan;

Bank of Japan Gov. Haruhiko Kuroda on Thursday revealed his strategy for ending more than a decade of deflation by expanding the central bank’s purchases of government bonds and allowing it to buy riskier assets.

The bank will “enter a new phase of monetary easing in terms of quantity and quality,” Kuroda said in explaining that the moves will double Japan’s monetary base and the amount of outstanding Japanese government bonds and exchange-traded funds within two years.
So far the market response has been positive.
Billy writes:

Dude, I have been listening for a year and never posted, but this guy just set me off.

One can keep dancing around the fact our GDP is almost wholly made up of fast food, loosing endless wars, and government handouts. One could also keep quoting cooked up government stats to support their ideas, but man oh, man, to actually believe that the end-all solution is lifting stock prices by having an unelected policy maker make a promise?

Doesn't pass the smell test to me.

[Comment edited by commenter.--Econlib Ed.]

SurfinCowboy writes:

Great podcast! I know when they are teaching me something when I have to rewind a couple of times to rehear certain arguments. Very engaging. Keep it up folks!

Lance writes:

I think further explanation is required around the concept that the government deposits in the name of a bank at the fed, and merely pays interest. For example, it is not reasonable to think that any financial institution is not actually paid for mortgages it sells to the GSEs. Potentially during the bailout banks took TARP and actually sat on the infusion of cash taking an interest payment. Sumner should explain how billions are passed through the GSEs to institutions that this is not increasing M1. My last thought, during 2008 there wasn't a lack of demand for credit, there was actually a change in the evaluation of credit worthiness that reduced the availability of credit coupled with a declining property value environment led to the beginning of the decrease in nominal spending.

I think his positions concerning the feds actions and fiscal stimulus dragging out the recovery are spot on.

Comments for this podcast episode have been closed
Return to top