Jul 19, 2021

Scott Sumner on What Milton Friedman Would Think of Monetary Policy Today

Milton Friedman’s critiques of 20th century Keynesian economic policy continues to shape monetary policy discussions today.
David Beckworth Senior Research Fellow , Scott Sumner Ralph G. Hawtrey Chair of Monetary Policy

Hosted by David Beckworth of the Mercatus Center, Macro Musings is a new podcast which pulls back the curtain on the important macroeconomic issues of the past, present, and future.

Scott Sumner is the Ralph G. Hawtrey Chair of Monetary Policy at the Mercatus Center. Scott joins David on Macro Musings to discuss Milton Friedman's views and what he might say about some of the recent developments in monetary policy. Specifically, Scott and David talk about nominal interest rates as indicators of the stance of monetary policy, fiscal austerity as means of reducing excessive aggregate demand, Friedman’s critique of the Phillips curve and wage and price controls, what Friedman might have said about the recent inflation numbers, and much more.

Read the full episode transcript:

Note: While transcripts are lightly edited, they are not rigorously proofed for accuracy. If you notice an error, please reach out to macromusings@mercatus.gmu.edu.

David Beckworth: Scott, welcome back to the show.

Scott Sumner: Thanks for inviting me, David. Good to be here.

Beckworth: Well, it's great to have you back on. The bigger context for the show is that you've been reading Ed Nelson's magnum opus on Milton Friedman. This is his two volume book titled “Milton Friedman and Economic Debate in the United States, 1932 to 1972.” You've written several pieces on that, you're working on a longer paper on Milton Friedman's work based on Ed Nelson. Listeners of the show will know we had Ed Nelson on the show back in March of this year to discuss the book.

Beckworth: So, if you haven't listened to that episode, I encourage you to go back and check out our conversation with Ed. Ed is at the Board of Governors. He's a great monetary economist. One of the last pure monetarist probably around, and maybe the last monetarist at the Board of Governors, the Federal Reserve System, but he's written a really fascinating book on the life and time of Milton Friedman's from 1932 to 1972. Scott, what was your overall impression of the book?

Sumner: Well, I really loved the book. It's a joy to read. I think it's must reading for anyone seriously interested in macroeconomics. Ed Nelson is unusual in that he's really good at two different things. One is knowing how to write about the History of Economic Thought. He has a very deep knowledge of Friedman and the other people he interacted with.

Sumner: Also, and this is not always true of historians, he has a deep knowledge of the nuances of monetary economics and theory, and so on. So he's able to correctly interpret Friedman's position relative to others in the debate, which is not always easy because Friedman can sometimes appear to be a little bit all over the map in various times in his career. But Ed really did a great job of finding the common threads that run through Friedman's reasoning throughout the decades. It's really an interesting book to read. I enjoyed it.

Beckworth: Yeah, it was a great book to read. But it's also a large book to read, two volumes. If I recall, the first volume was like 700 pages plus, the second volume, 500 pages plus. I had a hard time getting started on this book, because of the size, I finally did. I was glad I did. But really large, but great books. Again, I encourage listeners to go back and check out the previous show we did with Ed Nelson.

Beckworth: Yeah, I've enjoyed Ed Nelson's work, his own work over the years. He does serious macro work. He did a lot of interesting work recently on what Milton Friedman would have said about the QE programs. He's done these models that kind of link Friedman's ideas to what the Fed did with large scale asset purchases and we'll come back to some of this moving forward.

Beckworth: One of the big takeaways I got from the book, and there's many and this is not doing justice to the book, but one of the big takeaways is how much Milton Friedman changed his views over time. I was really surprised to see where he went from 1940s to where he ended up. So in the 1940s, Scott, as you know, he had some very unique views that Ed Nelson's described as being similar to other learners' functional finance views.

Beckworth: So Milton Friedman wanted to see basically deficit financed, budget deficits and have it financed entirely by reserves. So in the limit, get rid of all government bonds. Just the pure money, supply creation or reserve creation to be precise. However, he did say in this context that he'd want to balance that over the business cycle. So you would increase reserves, the deficit during the recession. Then you would wind it down during the boom year. Ed Nelson said he was still in his mind targeting some measure of money, but in a very post Keynesian way. I found that really fascinating. He wanted to go to 100% reserve creation that would finance the deficit entirely. Were you surprised by that? Maybe you were familiar with that already. I wasn't.

Sumner: I was only vaguely familiar with that. Because I think I had heard about it but I hadn't paid a lot of attention to it because I knew he moved on from that. But you're right. He was influenced by the Keynesian view at the time, which is very powerful in the 30s, late 30s and 40s. You're right, it had a role for money that was sort of special. It's kind of like I guess Congress taking over a monetary policy and trying to do counter cyclical monetary policy or reducing the money supply and booms and vice versa. Later he went to having the Fed run monetary policy and just a cyclical target, money growth at a constant rate. So I haven't paid a lot of attention to it partly because he moved on and partly because I don't think it's really workable for Congress to try to do that kind of monetary policy.

Sumner: But it is an interesting thing to consider, and it helps us understand where he began from. Ironically, although it might be considered a little bit extreme as a proposal today, when Friedman did move on into his monetarist phase in the early 1950s, his reputation actually went down quite a bit within the profession. This is something that Nelson talks about a lot, he had sort of a period out in the wilderness where he was viewed as a little bit of centric, a little bit of a nut. Then as he got into the 60s and 70s, his reputation revived quite a bit for various reasons that we can discuss. But when he first started espousing his pure monetarist ideas in the 50s and was really rejecting Keynesian views of fiscal policy, he was viewed as a little bit of a crackpot at the time.

Beckworth: Now, Scott, what I found really fascinating, and really interesting about that proposal, it sound very post Keynesian, his counter cyclical budget balanced approach that would use reserves. If you take that idea and you match it with his optimal quantity of money, his Friedman rule which would put interest rates really, really low, if you put those two together, it sounds very similar to what some post Keynesian are advocating today. The Fed had 0% interest rates forever and financed deficits. But he moved on, so we'll move on as well.

Sumner: Just interject one point on the on the optimal quantity of money argument. Now, he favored getting there through tight money. I think some of the people that advocate the zero interest rate policy today advocate doing sell through expansionary monetary policy. So in a sense, I think unless I'm mistaken, Friedman's view on how to get towards zero interest rates was closer to NeoFisherian view.

Beckworth: That's interesting.

Sumner: You have a very low inflation rate, very low growth in nominal GDP. That's what drives interest rates to zero. I don't think that is actually being advocated by some of the people on the left that are attracted to a zero interest rate policy.

Beckworth: So, that's a good distinction, a good point to highlight. So that was the 1940s. The 1950s, as you mentioned, he became a quantity theorists. 1956, he had this famous restatement of a quantity theory. Ed Nelson says though he was already changing his views before then, 1956 was just a marker in terms of publication. Ed Nelson traces it to his affiliation at the University of Chicago, his NBR project, his work with Anna Schwartz from the monetary history. He even says the federal court in 1951 had some influence on Friedman's thinking as well, all these things made him what he became in terms of a quantity theorist. But if you keep going down through his career, you go to 1968, he had this famous AEA presidential address where he says the Phillips Curve in the long run is bunk. It doesn't work.

Beckworth: We'll come back to this and the piece that you wrote. He set the conversation for the next half century about how should the Fed and how should macro policy respond to business cycles, really put forth this natural rate idea. A few years before that though, he had his plucking model. I want to come back to that as well, maybe later in our conversation, but he had very interesting, very different ideas. Then you mentioned his constant money supply growth rule, which he advocated. I went back and tried to find where he first proposed this. What I found were a couple of testimonies he gave to Congress in the late 50s in the mid 60s on targeting money.

Beckworth: I don't know if that strikes us when he's first introduced that's what I found. He held that view for some time. Then in his book, Money Mischief, which came out in 1992, he came around to endorsing Bob Hetzel's proposal for targeting the breakeven inflation on the difference between tips in a regular bond. Now, at the time, those didn't exist. But he was very excited about the idea and so we should pursue it. So his thinking continued to evolve even I think to the end and on some areas. He was just a very successful, I think, person in terms of influencing the debate on economic policy.

Beckworth: So let's talk about his influence and success. I want to go to an article you recently wrote based on your reading of this book, I think it's part of your bigger paper that you're writing. But you have four Keynesian ideas that Friedman rejected, and we'll provide a link to that. So why don't you walk us through the four ideas that Friedman attacked, and you say, successfully attacked during this time?

Four Keynesian Ideas that Friedman Rejected

Sumner: Right. Before doing so, let me frame this issue by saying that a lot of people think of Friedman in terms of his policy proposal for stable money growth. I don't think that's the best way to think of his views, as you point out later in his life, he became more interested in ideas like inflation targeting. He favored targeting money growth for pragmatic reasons, he conceded there might be other better policy options out there. When inflation targeting seemed to work pretty well under Greenspan, he sort of endorsed that. So rather than focus on money supply targeting is the key Friedman idea.

Sumner: I think his real influence was his critique of Keynesian economics. So I'll mention four key ideas from the late 60s, which were by the late 70s and 80s, becoming incorporated into new Keynesian thinking on macroeconomics. So these are, first of all, that interest rates are not a reliable indicator of the stance of monetary policy. This is partly because of the distinction between real and nominal interest rates.

Sumner: At the time, during the 60s, most economists focused on interest rates as the indicator. So a tight money policy was high interest rates, easy money was low interest rates in their view. Friedman said, "No, you have to look at the money supply growth rate. If money is growing very fast, and you have inflation, interest rates may be high, but you still may not have tight money." A second view was that fiscal policy is not appropriate tool for controlling inflation, it's not going to be very effective. A third is, as you mentioned, his critique of the Phillips Curve. He said there's no reliable long run trade off between inflation-unemployment. In the long run, the Phillips Curve is vertical, that is unemployment will tend to go back to the natural rate regardless of what the rate of inflation is, in a long term sense. Finally, he was critical of cost/push theories of inflation, and specifically the policy recommendation, wage price controls that came out of that theory of cost/push inflation in the late 60s, early 70s.

Sumner: So in my view, all four of those critiques were broadly accepted by the new Keynesian economists by say, the 1980s. What you saw was a series of events in the late 60s and early 70s where Friedman's view on all four issues seemed to be confirmed by events. So if we start with the first one, interest rates, his claim that these higher interest rates in the late 60s were not going to slow inflation, it turned out to be correct. Inflation continued to accelerate through the late 60s and into the 70s, even as the Fed raised interest rates. So the broader profession began to rediscover Irving Fisher's idea of the distinction between real and nominal interest rates. They started to pay more attention to the money supply, even while never fully accepting Friedman's obsession, if you will, with the money supply targeting.

Sumner: A lot of this also came out of Friedman’s and Schwartz’s book on monetary history of the US, which came out in 1963 and turned out to be very influential in changing views about how to think about monetary policy. So when inflation continued to accelerate, people started to pay more attention to monetarists like Friedman.

Sumner: Then in 1968, President Johnson raised income taxes as a way of slowing inflation. This was a policy option that many Keynesians favored, but it didn't seem to work. So we actually went into a budget surplus in late 1968. Despite that, inflation continued to accelerate. This is probably the event more than any other that led people to rethink whether fiscal policy is really the right tool.

Sumner: Friedman argued that even if you have a tight fiscal policy, as long as you're increasing the money supply rapidly, as they were in the late 60s and early 70s, you'll continue to have high inflation. Fiscal policy alone cannot control inflation, if money supply is growing rapidly. Again, that seemed to be confirmed by events. I might add that today, there are some people again bringing back fiscal as a way of controlling inflation. I think MMT, that's part of their view. But Friedman's prediction on that tax increase seemed to come true.

Friedman argued that...fiscal policy alone cannot control inflation, if money supply is growing rapidly. Again, that seemed to be confirmed by events. I might add that today, there are some people again bringing back fiscal as a way of controlling inflation.

Share this

Sumner: A third was the Phillips curve. It was believed that there was a trade off between inflation, unemployment. In the 1960s, it very much looked like there was. Interestingly, Friedman and Phillips also predicted that that wouldn't hold true permanently at a time when it looked like it was working still in the late 60s. Then in 1970, by the end of 1970, we had high unemployment, over 6% and yet inflation was still running at 5.5%. So a lot of people were scratching their heads, how can this be happening? This is inconsistent with Phillips Curve Theory.

Sumner: But Friedman and Phillips had predicted this a few years earlier. So that really boosted Friedman's reputation. A lot of Keynesian schools like Harvard and MIT by the 70s, economists were saying, "Well, Friedman looks like he was right about that issue, about the shifting Phillips Curve."

Sumner: Then finally, the price controls. That had not necessarily been a traditional Keynesian policy option. But because of the frustration with the fact that we had high unemployment and high inflation at the same time, there was this frustration among Keynesians that it didn't seem like traditional aggregate demand tools were working. So that's what led them to reach for what's called an incomes policy or wage price control as a way of controlling inflation, instead of tight money and fiscal policy.

Sumner: So that was adopted by President Nixon in 1971, it seemed to work pretty well for a period of over a year. But then it all fell apart in '73 and '74. Shortages developed as the controls were gradually phased out. We accelerated much higher inflation. So by the mid 70s, most economists including most Keynesian economists, had begun to lose faith in wage price controls as a tool for controlling inflation. Friedman was actually, believe it or not, somewhat of an outlier among economists in 1971 when the wage price controls were instituted.

Sumner: Today, we would think of that as a fairly radical policy proposal. The government setting wages and prices in the economy. But in 71, when it was imposed, it was imposed by a Republican president and was fairly popular among Democratic economists at the time. So I think those four areas where Friedman is a critic of Keynesian economics seem to be correct on a number of issues, eventually got incorporated into the new Keynesian model, and are really the areas that distinguish I would say, The New Keynesian Model from traditional Keynesian economics of the 50s and 60s. That, to me, is Friedman's true legacy, not his proposal for targeting the money supply growth rate.

I think those four areas where Friedman is a critic of Keynesian economics...eventually got incorporated into the new Keynesian model, and are really the areas that distinguish I would say, The New Keynesian Model from traditional Keynesian economics of the 50s and 60s. That, to me, is Friedman's true legacy.

Share this

Beckworth: Very interesting. I want to go through those four with you, flesh them out a little bit more, but just to go back on his constant money supply growth rate rule. I had this question for Ed Nelson. I asked him, what if the Fed and government had targeted a constant money supply growth rule, say during the 60s and 70s? If that had been the case, maybe then velocity wouldn't have been so unstable, the relationship between money and nominal income would have been more predictable, more stable, more steady. Ed Nelson said yes. In fact, in the book, Ed Nelson quotes Alan Blinder who said, "A rule like Friedman's constant money supply growth rule would have been useful during those two decades."

Beckworth: So I think something people often miss when they critique that rule, and I'm not advocating, as listeners know, both Scott Scott and I are for nominal GDP level targeting. But one of the critiques against this constant money supply growth rate target is that well, velocity is not stable. Well velocity may be endogenous to how stable policy is. I think Friedman would say, "Well, if the Fed had been targeting something like a constant growth rate rule, the money supply, then things would have turned out very differently." So just want to put that out there. But let's go through the these four areas where you say Friedman was victorious and they got incorporated into new Keynesian macro model, macro thinking, which is the workhorse model that's being used at most central banks. Let's start with the first one about nominal interest rates not being a good indicator of the stance of policy. So how do we see that in new Keynesian analysis today?

Nominal Interest Rates as Indicators of the Stance of Monetary Policy

Sumner: Well, if you think of like the Taylor rule approach to monetary policy, its distinction between real and nominal interest rates is embedded right within the rule. So if inflation is rising, say by one percentage point, the model might suggest you should raise nominal interest rates by one and a half percentage points to get ahead of the curve, so to speak, that is to make sure that even the real interest rate rose. In that way, you'd actually have a contractionary effect on the economy which would slow the inflation.

Sumner: The basic mistake that was made during the 60s and 70s is yes, they did raise nominal interest rates somewhat, but inflation expectations were rising even more rapidly. So they weren't really having an effectively contractionary policy, the real interest rate wasn't being raised in a way that would control inflation. So that distinction is very much a part of sort of the Taylor rule approach to New Keynesian policy, not to suggest the Taylor rule is identical to New Keynesian economics. But that way of thinking about interest rate control, what is the interest rate relative to the natural interest rate? It's very much cognizant of the difference between real and nominal interest rates.

The basic mistake that was made during the 60s and 70s is yes, they did raise nominal interest rates somewhat, but inflation expectations were rising even more rapidly. So they weren't really having an effectively contractionary policy, the real interest rate wasn't being raised in a way that would control inflation...that way of thinking about interest rate control, what is the interest rate relative to the natural interest rate? It's very much cognizant of the difference between real and nominal interest rates.

Share this

Beckworth: Yeah, if I had to think of a summary view of the new Keynesian interest rate thinking it would be maybe Michael Woodford in his textbook. He basically makes the case, it's pretty standard that you want to look at the expected path of the Fed funds' target rate, what the Fed is going to do with it, and compare that to the path, expected path of the neutral interest rate. So that would also be capturing Friedman's ideas?

Sumner: Yeah. I don't recall whether he used terms like neutral rate a lot. But that's implicit in the way he thought about interest rates. That is if interest rates are raised to 8%, but inflation was 10%, then Friedman would say, "You don't really have a contractionary monetary policy," because the real interest rate is negative two. So that's obviously, and I should say, Friedman's critique of viewing monetary policy in terms of interest rates wasn't solely based on this distinction between real and nominal interest rates. He also talked about the income effect from monetary policy. So for instance, a tight money policy can drive the economy into a recession. In a recession, even the real equilibrium interest rate will fall. This is called what Friedman called the income effect. It's sort of an intermediate effect.

Sumner: So liquidity effect is short term, easy money lowers interest rates. In intermediate term, you have the income effect, where monetary policy changes the business cycle and changes the equilibrium interest rate as a result of the business cycle. Then the long run effect is the Fisher effect, long run inflation expectations factor into long run interest rates. So Friedman actually looked at all three of these factors. He regarded the short term nominal interest rate as a wildly inappropriate indicator of the stance of monetary policy, partly due to the real nominal interest rate distinction, but also partly due to the income effect of monetary policy, which can change the equilibrium real interest rate.

Beckworth: Okay. Let's go onto number two about fiscal austerity not being sufficient to reduce excessive aggregate demand. You mentioned the 1968 case where President Johnson increased taxes, they ran surpluses. Yet the economy was still was growing too hot, and it was a case of overheating there. So how would you tie fiscal policy then to the inflation? Is there any link? By that, I mean, the following: if the Fed is creating a certain amount of monetary base, has a certain path implied about its target, interest rate going forward, can that be influenced by what we think the government's going to do in the future with deficits? So if the government's going to run excessive amount of deficits in the future, can fiscal policy indirectly affect inflation via that route? Would Friedman make that case?

Excessive Aggregate Demand and Fiscal Austerity

Sumner: It could. He tended to believe that more earlier in his career and moved away from that view over time, at least to some extent. So he certainly recognized that in some countries, let's say Zimbabwe, that they're very closely linked because the central bank is forced to monetize, budget deficits, that sort of thing, Venezuela more recently. So he knew that. For the United States, he tended to think that the Fed was under pressure to some extent to monetize fiscal deficits. But I think over time, he became less enamored of that view. I think it's partly because events didn't really seem to support that theory very well. In fact, although I mentioned the budget going into surplus in late 1968, if you look at the 1960s and 70s, in general, there really weren't large budget deficits.

Sumner: I think a lot of us have been taught that there was this guns and butter policy during the 60s, and that created the high inflation. Then we were taught there was the supply shocks during the 70s and that created even higher inflation. But when you look at the numbers, neither of those claims are true. The budget deficits in the 1960s were not particularly large. Real GDP growth in the 1970s was perfectly normal, around 3% per year. In fact, both the 60s and 70s, the cause of the high inflation was exactly the same, accelerating growth in nominal GDP. The growth in nominal GDP was not being driven by monetary policy to finance budget deficits. It was just being done for various other political reasons, maybe in the short run to lower unemployment or confusion over real and nominal interest rates, or whatever your theory is.

A lot of us have been taught that there was this guns and butter policy during the 60s, and that created the high inflation. Then we were taught there was the supply shocks during the 70s and that created even higher inflation. But when you look at the numbers, neither of those claims are true.

Share this

Sumner: The Fed was really under no pressure to finance budget deficits in the 60s and 70s because they were much smaller than what we've had more recently. Even in the 1980s, when inflation came down sharply, the budget deficits were substantially larger than in the 60s and 70s. So there really isn't much empirical evidence to support the notion that the deficits were the motivation for the printing of money. As Friedman saw more and more data to support that position, he lost interest in the view that fiscal policy was the underlying factor. He started to more directly blame the federal reserve for mistakes.

Beckworth: So he wasn't an advocate of fiscal dominance. He definitely thought the Fed had more ability to shape the path of the price level.

Sumner: Exactly. Maybe you can find a few of his statements earlier in his career where he talked about how fiscal deficits could pressure the Fed to monetize them. But he tended to move away from that view in the in the 60s and 70s and took what I think of as more of a modern view that the Fed can and should control the rate of inflation, for any reasonable debt. Now, of course, if the deficits became very large, there would be political pressure on the Fed. But the national debt as a share of GDP, I don't remember the exact numbers, but I think it was in the range of 30 to 40%, 30 to 50%, say, of GDP in the 1960s and 70s.

The Fed was really under no pressure to finance budget deficits in the 60s and 70s because they were much smaller than what we've had more recently. Even in the 1980s, when inflation came down sharply, the budget deficits were substantially larger than in the 60s and 70s. So there really isn't much empirical evidence to support the notion that the deficits were the motivation for the printing of money.

Share this

Sumner: That's a fairly low ratio as a percent of GDP. That number was trending lower. So think of it today. I think it's over 100% of GDP today and trending much higher. So we didn't have the kind of deficits that we have today or even the deficits we had in the 1980s under Reagan. But that is the weird version taught in the history books. That it was fiscal policy, and then it was supply shocks. The data just doesn't support either claim.

Beckworth: Hey, look. I'm guilty of that myself. I think I've even said on the show several times like the mid to late 60s was a period where they did guns and butter, the Great Society and Vietnam. Maybe someone taking that view would advocate that maybe the deficits weren't that large at that point in time, but maybe they were expected to create a path of higher deficits, and they never materialized.

Sumner: Yeah. I think that's probably John Cochrane's view of fiscal theory of the price level.

Friedman’s Critique of the Phillips Curve

Beckworth: Let's move on then to your third point about the Phillips Curve. So you mentioned Phillips, you mentioned Friedman. They both had a 68 presentation papers that really undermined the view that over the long run, there's a tradeoff. I just wanted to explore that and push that out to its implications for trend inflation. So what you could salvage from that is that there's a short run trade off, right? The long run trade off isn't there and there's a short run that may not be very stable relationships. In the extremes, maybe even the short run isn't there, hyperinflation environment. But what I wanted to ask is this, even in the best case scenario, which may be a short run Phillips Curve, Phillips Curve is not a theory for long run inflation, right? Phillips curve is at best a theory for short run inflation, is that what Friedman would say?

Sumner: Yeah. I'm not even sure if he would say it's necessarily a theory for inflation at all because he tended to view the causality differently from a lot of the Keynesian interpretation of the Phillips Curve.

Beckworth: Okay.

Sumner: The Keynesian view tends this goal causation from real output or employment to inflation. So an overheating economy causes the high inflation. Friedman talked in terms of inflation being above expectation causing low unemployment or inflation being below expectation causes high unemployment. If you're having trouble wrapping your mind around that causation, one way to think of it would be like a sticky wage model. Suppose you had sticky nominal wages, inflation falls sharply, unexpectedly. Now you've increased the real wage rate, right? So that would cause work companies to want to employ fewer workers.

Sumner: Friedman, I think, said that he recognized the causation goes both directions. But the reason he liked to think in terms of inflation shocks affecting unemployment was that inflation is a nominal variable. Ultimately you think the fundamental shock in the background was actually monetary policy. Money is also a nominal variable. So he was really sort of thinking in terms of nominal shocks having real effects in the economy.

Sumner: If you think of the Phillips Curve that way, nominal shocks having real effects, it makes more sense to think in terms of unexpected inflation impacting the unemployment rate than vice versa. But again, he did recognize that causation went both ways. It's very complicated. Of course, inflation is sticky, slow to adjust. So I'm over simplifying his views a little bit here. But that's, I think, something that distinguishes his views from others.

Sumner: If you think of it in terms of nominal shocks, having a real effects, you're not going to have a real theory of long run inflation, right? So your theory of long run inflation is going to be whatever's coming out of monetary policy. So if money growth is very rapid, like it was in Latin America during his lifetime, then you'll have high trend inflation. If it's very slow, like in Switzerland, you'll have low trend inflation. Then when you talk about variations in inflation, that's going to be associated with the business cycle. So the accelerating and decelerating money growth will affect both inflation rates and real variables like output and employment.

Beckworth: Let me ask one question that I was going to ask later in the show, but I think it's relevant to bring it up now. That is what would Friedman think of the popular view that I see, that okay, the Phillips Curve doesn't work, especially between the price level and unemployment or any measure, slack, output gap, whatever it may be. But it does seem to work or there's something there at least statistically between wage inflation in measures of slack, unemployment rate. For example, I think a popular one is looking at the ECI, which is employment cost index, growth rate and the employment to population rate. Those, I have some friends, you have some friends, Scott, that really advocate that is the way to do the Phillips Curve today. What would he say about those folks? What would he say to those folks who are using that to help think about inflation?

Sumner: Well, that's a good question. A lot of his work was done before the big supply shocks of the 70s. And so his working assumption, in his ‘68 famous paper on the Phillips Curve, was that wages and prices were moving in tandem, I think he assumed a constant markup model or something in that paper. So he had very much a demand side view of the business cycle. So in that sense, he had a Keynesian view of the business cycle, what we would call Keynesian today. He didn't distinguish that much, I think between those two variables, the wage inflation and price inflation. But I think, and this is something I'd have to look up, if you do bring supply shocks into the picture, then obviously, wage and price inflation can diverge quite a bit.

Sumner: I presume, in that scenario, Friedman would probably see wage inflation is more reliable than a Phillips Curve type short run analysis, but I can't be certain about that. Now, Ed Nelson's book, I should point out, goes up until 1972. I think there's going to be another volume that looks at his later work. So I'm sure there'll be a lot of discussion of that issue. But at least in his core period where he was discussing monetarism, he tended to do some wage and price inflation are similar.

Sumner: I might add that the failure of the Phillips Curve in the 70s also failed if you use wage inflation, like wage inflation got very high in the 1970s as well. Unemployment was considerably higher than in the 60s. I think you're right that in a cyclical concept, context, the ups and downs of the business cycle, wage inflation probably works better than price inflation, I prefer using wage inflation myself. So I agree with you there. But in terms of the long run vertical Phillips Curve, it's still long run if you use-

Beckworth: It's still there.

Sumner: Wage inflation on the vertical axis.

Beckworth: Whatever you use, it's still going to be vertical.

Wage and Price Controls and Cost/Push Inflation

Beckworth: All right, Scott, let's go to number four, where you talk about cost/push inflation and the implications that flow out of it for wage price controls. I wonder if you have any thoughts on some of the growing interest in that today? So we often hear people talk about, well, we don't see any really, really strong growth in wages. So we know inflation won't take off. They look to wages as an indicator. I'm not sure that's entirely wrong. You mentioned previously how nominal income grew rapidly in the 1970s. But that doesn't mean it's necessarily a cost/push phenomena. It could be demand driven, right, that's causing wages to go up as well. Then I guess the last thing though is the the implication of wage and price controls. What would Milton Friedman say to those people today who are advocating them?

Sumner: What would Friedman say about those advocating wage price controls today?

Beckworth: Yes, yeah.

Sumner: I think he would say it's treating the symptom and not the underlying cause of inflation, which is excessive money growth. Here's one way I like to think about Friedman's views on a wide range of issues. You can think of Friedman's monetarist career as a thought experiment like what would happen if you permanently increase the money supply growth rate and the trend rate of inflation to a higher rate than before? Okay. What would that look like in a modern economy? If you think through that thought experiment, all four of those critiques I just mentioned are related to that thought experiment. The real nominal interest rate distinction becomes important if you increase the trend rate of inflation. The Phillips curve starts shifting. Right?

What would Friedman say about those advocating wage price controls today? I think he would say it's treating the symptom and not the underlying cause of inflation, which is excessive money growth.

Share this

Beckworth: Right.

Sumner: If the trend rate of inflation permanently rises. The other two, it's more subtle on fiscal policy and cost/push inflation, what Friedman would say is, at best fiscal policy could cause like a one time increase in velocity. Maybe it makes velocity go up two or 3%. But that's it. It's a one time shift in the price level. But it's not a persistent change in the trend rate of inflation. Same thing with cost/push, maybe monopolies and labor unions push the price level two or 3% higher than it would be under a less monopolistic economy. But that's a one time increase in the price level. Right, or an oil shock. That's a one time increase in the price level.

Sumner: To get a higher trend rate of inflation, you really have to have persistently more rapid money supply growth over time, over years, over decades. If you have that money supply growth persistently rising, and this is really important, anything you do on the fiscal front or on the cost/push inflation front will be mostly ineffective. At best, you'll win, maybe a year of lower inflation, like the price controls worked for about a year, little over a year. Then they stopped working.

Sumner: Fiscal policy might slow inflation briefly. But if it's actually being generated by rapid money growth, the force of that on the nominal variables like inflation, the nominal GDP will eventually overwhelm any other non monetary influence on the price level because all those non monetary factors are things that only cause a one time change in the price level. Even if it's a permanently more expansionary monetary policy. I'm sorry. Even if it's a permanently more expansionary fiscal policy, it only raises velocity one time. If you make a permanently more expansionary monetary policy, it raises inflation from now until the end of time. So they're fundamentally different types of variables. Friedman understood that, he worked through the implication of faster, permanently faster money growth in those four areas I cited. I think that's why he was successful, he turned out to be correct, that inflation was actually being driven by more rapid money growth.

Beckworth: That covers I think most of the points you make in your other article we were going to go to, “Inflation is a Nominal Phenomenon.” Anything else you wanted to add from that piece before we move on?

Sumner: No, I think that's yeah, that was the-

Beckworth: We covered it, okay.

Sumner: In a way, the phrase, "inflation is a nominal phenomenon" is kind of an obvious point. I think of some of these other theories of inflation is trying to explain it using real variables. I think of fiscal policy and price controls as real tools to control a nominal problem. I don't know if that-

Beckworth: Well, let me ask this question based on what you've just said, what Friedman thinks, what you think, how would you make sense of this past year? So particularly the past few months, we've had higher than expected CPI, NPCE prints, we've seen inflation a little bit larger than was expected. But even the expected amount was higher. Would you view this more as a one time pop or of a trend change?

What to Make of Recent Inflation Data

Sumner: Yeah. That's hard to say. My own personal view is [the recent spike in inflation is] probably a one-time pop in prices. I base that on the fact that some of the financial markets like long term bond yields and tip spreads don't show a lot of inflation going forward. On the other hand, I think we have to be careful or modest about making predictions in this area because I'm old enough to remember when the grade inflation got going, there were a number of articles saying, "It's temporary, it's just this product or that product." So what I would say is, yes, I do understand that a lot of the inflation is being driven by things like use car prices and so on. But I don't find that totally reassuring. To the extent I'm reassured about inflation, it's more based on the financial market forecasts.

My own personal view is [the recent spike in inflation is] probably a one-time pop in prices. I base that on the fact that some of the financial markets like long term bond yields and tip spreads don't show a lot of inflation going forward.

Share this

Sumner: But then even there, someone can argue the other side and point out that to some extent, the Great Inflation was not predicted by the bond market, right, at the time. But I think it's the best or the least bad tool we have to forecast is to look at what the consensus market forecast is. This is such an unusual period, such an unusual business cycle that I don't think that the insights of Milton Friedman really are very useful. But I also don't think my own earlier writings on macroeconomics are particularly useful for this business cycle either.

Beckworth: Well, I think this speaks though to the bigger critique that you and I have that a nominal GDP target would be useful for this very situation, right? We don't know in real time what inflation is doing, what's driving inflation, supply shocks, temporary, permanent. We just don't know. But what we can look to is nominal GDP, look at the forecasts of it, and that would avoid some of this confusion. All right. I want to do a round with you now, Scott, where I'm going to ask question, the question, what would Milton Friedman say about…?

What Would Milton Friedman Say About …

Beckworth: I want to bring up a number of questions related to current events, and you do your best to answer them. Okay? So here we go. Question number one, what would Milton Friedman say about the current spike in the M2 money supply? So this is something that many people will bring up when they're worried or they want to make a case for something crazy is about to unfold, the 1970s is here again. Just look at that surge in the M2 money supply. How do you think about it or how would Milton Friedman think about it?

Sumner: Okay. There's two ways of answering that question. The obvious way is that he would be worried that the faster money growth would lead to higher inflation one or two years down the road. That's certainly based on his way of looking at data during the 1950s, 60s and 70s. There is though an additional way of thinking about it, which is what if Friedman had lived through the last couple of decades and had seen everything we've seen. How might he have reevaluated his views on some issues?

Sumner: We know that in the 90s and early 2000s, he moved away from money supply targeting towards interest rate targeting. I wrote a paper a few years ago about, it was titled something like, "What would Friedman have thought of market monetarism?” This is in an Oxford volume on Milton Friedman. In this paper, I looked at a lot of the things he said throughout his career. I argued that he probably would have evolved in a more market monetarist direction. In fact, was evolving in that direction late in his career. So I think that he might not be as concerned about the spike in the money supply, as he would have been in the 1960s and 70s if he had lived this long and saw many of the things that we've seen in recent decades, including the unusually high demand for money in a zero interest rate environment.

Beckworth: All right. Next question, which is a nice segue into what you just said. That is what would Milton Friedman say about the zero lower bound environment?

Sumner: That's a very interesting question. I think it's fair to say that Friedman probably underestimated the difficulties posed by [the zero lower bound], as did many of us. He certainly felt that the so called liquidity trap wasn't truly a liquidity trap in the 1930s, that the Fed could have stimulated the economy much more effectively if it had been more aggressive in boosting the money supply. Now, others, like Paul Krugman, have criticized that view and pointed to the fact that part of the problem had to do with the difference between the change in the monetary base and the broader money supply. So the Fed actually was increasing the monetary base during the Great Depression, and the money's broader MO and M2 money supplies were going down.

I think it's fair to say that Friedman probably underestimated the difficulties posed by [the zero lower bound], as did many of us. He certainly felt that the so called liquidity trap wasn't truly a liquidity trap in the 1930s, that the Fed could have stimulated the economy much more effectively if it had been more aggressive in boosting the money supply.

Share this

Sumner: I think Krugman pointed to Japan where a similar thing occurred. A lot of base money was injected by the Bank of Japan and not much growth in the money supply. So essentially, that calls into question Friedman's claim that it's clearly true that the Fed can control the broader monetary aggregates at the zero bound. Maybe that's not true.

Sumner: My own view is they probably can do so. But in order to do so, they might have to go beyond what Friedman envisioned as the kind of tools that would make that work. Ironically, they might have to do some of the things that Paul Krugman recommended, the so called promise to be irresponsible and committed to a higher inflation rate in order to get expectations to change in such a way that all that base money is not just hoarded. So again, we're faced with the difficulty of what if Friedman had seen all these events during the Great Recession, how might he have modified his views?

Sumner: But certainly, I do believe that he did live long enough to criticize the Bank of Japan's policies in the late 90s and specifically argue that monetary policy was actually quite contractionary there, despite the low interest rates. So I think he might have been sympathetic to the market monetarist view that the Fed was too tight in 2008, although I can't be certain of that.

Beckworth: Well, I would encourage our listeners to go back and listen to our previous podcast with Scott Sumner. We talked about the Princeton School of Macro Economics. One of the points that you bring out, Scott, that we talked about is the importance of a permanent monetary base injection. This is something that if you look in Friedman's work, he uses this often as a thought experiment.

Beckworth: He says if you permanently increase the monetary base or increase the growth rate of the monetary base, what happens afterwards? I'd go beyond that. He would say, look, you increase monetary base permanently and it's greater than the existing demand for the monetary base, those two things. He would tell a story that's very similar to what the Princeton school would tell in their ideas, their arguments. So I think it's fair to say his views would have evolved but they would have been built upon some of his existing understanding of the dynamics of the price level coming from permanent changes in the monetary base.

Beckworth: Now, the other thing that has changed, of course is interest on reserves which complicates that story a bit. But I think he's not that far off. You look close enough himself, Alan Meltzer, some of the other monetarists, that they all recognize the importance of permanent monetary base injections in their stories, which is very similar to some of the Princeton school arguments. So I think you're right to say he probably would have changed his views of bit, but in a way consistent with what the Princeton school did in the late 1990s, early 2000s.

Sumner: I'd like to mention something about interest on reserves. I think Friedman would have latched on to [interest on reserves] in late 2008 and 2009. Although some of the other monetarists like Anna Schwartz and Alan Meltzer, took a slightly more Austrian view of the Great Recession. I think Friedman as late as 2006 right before he passed away, he was very laudatory about Alan Greenspan's policies. So, he really wouldn't have been able to blame the Great Recession over the expansionary monetary policy, which became the standard conservative explanation of the Great Recession. We stimulated too much in the housing boom, and then we paid a price.

I think Friedman would have latched on to [interest on reserves] in late 2008 and 2009. Although some of the other monetarists like Anna Schwartz and Alan Meltzer, took a slightly more Austrian view of the Great Recession. I think Friedman as late as 2006 right before he passed away, he was very laudatory about Alan Greenspan's policies. So, he really wouldn't have been able to blame the Great Recession over the expansionary monetary policy.

Share this

Sumner: So then that's a problem for Friedman because if you think of a lot of his career, including the whole monetary history of the US, he was essentially trying to rescue capitalism from the accusation or from the claim that capitalism is inherently unstable. He was basically arguing along with Anna Schwartz that no, it was monetary policy that was the problem, not the inherent instability of capitalism. Given he had supported Greenspan's policies up through 2006, he couldn't very well blame the Great Recession on that. I think that the interest on bank reserves, I think you were one of the first to really point this out in your blog in late 2008, was similar to the fed's decision in 1937, to raise interest rate, raise reserve requirements several times.

Sumner: Friedman would have seen this because this was a very important part of his view of the Great Depression. Obviously, interest on reserves is something that's intended to create more demand for bank reserves, just like raising reserve requirements. The fact that that occurred in early October 2008, right as we were plunging into the Great Recession, into the deeper part of the Great Recession, I think he would have viewed that as a mistake and latched on to that because he wouldn't have had many other options to either ... He can't really blame Fed policy before 2006. He doesn't want to blame the inherent instability of capitalism.

Sumner: His whole career has focused on the notion that these severe demand shock recessions are caused by bad monetary policy. So that would have been the thing that I think he would have latched onto. On the other hand, I don't think necessarily, that's a complete explanation of the Great Recession because there's some countries that didn't pay interest on reserves that had problems as well. But anyway, I think that's one thing he might have focused on. It'd be interesting to see how his views did evolve, if he had live. That it's kind of unfortunate that he passed away in 2006.

Beckworth: Yep, I would have loved to have gotten Friedman on the podcast, had he been around long enough. Okay. Let's go on to another what would Milton Friedman say question. Let's ask this one. What would Milton Friedman say about the Fed's efforts last year to save the global dollar funding markets?

Beckworth: So the Fed last year as you know, opened up, its dollar swap lines with other central banks. Some of them already opened, but it expanded number of central banks that could tap into it. It also set up a new repo facility for foreign international organizations, central banks, governments and the like. So it effectively backstop the shadow banking system across the world or euro dollars, whatever you want to call it, but always dollar funding markets abroad.

Beckworth: You could tell a monetarist story here. Look, there's this global supply of dollars. The Fed stepped in and followed Milton Friedman's Maxims, that it kept them from collapsing. So Friedman would have said, "Rah, team. Go Jay Powell." On the other hand, Milton Friedman might have been nervous because here, the Fed is backstopping the global shadow banking system. Where do you think you'd come down on that?

Sumner: Well, I guess if it really looked like it was a serious threat, I lean towards the view that he might have been supportive of that. Relatively speaking for a conservative, Milton Friedman was fairly supportive of what you might call a lender of last resort type roles for central banks. Remember, he criticized the Fed for not assisting the Bank of the United States, which is the first major bank failure in late 1930 and saw the fed's refusal to help in that situation, as making the Great Depression much worse, as a result of the subsequent banking panics. So I lean towards the view that he might have been supportive as long as it clearly did look like it was necessary to prevent a real global banking crisis of some sort. But I can't be certain on that.

Beckworth: All right. So I've had several people ask me this question. What would Milton Friedman say about a nominal GDP level target? Let me rephrase it this way, would Milton Friedman have advocated for a price level target or a nominal GDP level target? So we're now in the world of the fed's average inflation target, other central banks, the ECB, for example, is talking about doing makeup policy. So makeup policy is a form of level targeting. But what kind of level targeting would Milton Friedman have supported?

Sumner: That's a very interesting question. In my Oxford paper, I talked about that quite a bit. So he's definitely on record has been more supportive of inflation targeting than nominal GDP targeting. I think there a few occasions where he was a little bit critical of nominal GDP targeting, at least hypothetical cases that could result from that. Although I believe Nelson mentioned one case where he was a little bit sympathetic to the idea as a possibility.

Sumner: On the other hand, this is something I don't think people have thought about enough. He favored steady growth in the monetary aggregates, say M1 or probably M2 of say, three or 4% per year. He believed that if the Fed did this, velocity would be fairly stable. Okay? So he favored stable growth in M2. He believed that would result in fairly stable velocity. That means, by necessity, that he believed his policy would result in fairly stable growth in nominal GDP.

Sumner: So he clearly didn't view that as a bad outcome. If someone had said to him when he advocated 4% growth in M2 every year, but what would happen if velocity was stable? Wouldn't that be terrible? Friedman would not have agreed with a person. He would say, "Well, actually, it's fine if velocity is stable." But of course, that would mean you'd have a nominal GDP target. Right? You'd have stable growth and nominal GDP. So as an outcome, I think he would view that as generally speaking, relatively a good outcome for monetary policy. I think he could be convinced by some of the arguments that you and I have made that in certain situations like 2008, it would have been quite a bit superior to inflation targeting, but we'll never know the answer to that for sure.

Beckworth: I agree with that, Scott. Both of us have advocated in the past that a nominal GDP target could be called equivalently a velocity adjusted money supply target, right? A nominal GDP target is effectively a velocity adjusted my supply target.

Beckworth: I recall writing out on my blog, I mean, many years ago, back in the glory days of blogging, and while I was so delighted to find out that Michael Woodford picked up that blog post, and he cited one of his papers a talk he gave at the Jackson Hole meetings, it was the 2011, I believe talk, where he said, "What should we be doing?" He came out in support of a nominal GDP level target. He made the case that Milton Friedman would probably support a nominal GDP level target. He invoked my blog post, my argument where nominal GDP targeting is a velocity adjusted money supply target. So why not? Right? That's kind of what you're saying.

Sumner: There's a lot of reason to think that Friedman thought in terms of monetary policy in terms of nominal GDP more than most economists, his 1971 JPEE article, I think you saw that also, someone dug it up a few months ago, is all focused on this question of how money affects nominal GDP growth, separate from how growth in nominal GDP is partitioned between prices and output. He used that as a secondary issue. So that's one thing I could point to.

Sumner: Second, we did see after the Great Recession, a number of people that you might describe as of monetarists like Nick Rowe and new Keynesian, like Woodford and others, moved from inflation targeting towards nominal GDP targeting at least for pragmatic reasons, not necessarily as the optimal policy. So Friedman would have seen the same events as some of these monetarists and new Keynesians that were convinced by the data of 2008 and nine. I think that given everything we know about, how he thought about monetary issues, it would have been quite logical for him to move in the direction of nominal GDP targeting.

Beckworth: Well, Scott, we are near the end here. I will just close with this, that probably the last well-known monetarist, pure monetarist, Bennett McCallum was an advocate of nominal GDP targeting. As I mentioned Ed Nelson as well, Ed Nelson's another monetarist who's still around. But they both advocated for nominal GDP targeting. So further evidence that Friedman probably would have gone that direction, had he been alive, to be a part of this conversation. Well, with that, our time is up. Our guest today has been Scott Sumner. Scott, thank you so much for coming back on the show.

Sumner: Thank you for inviting me.

Photo by Alex Wong via Getty Images

Support Mercatus

Your support allows us to continue bridging the gap between academic ideas and real-world policy solutions.Donate