Scott Sumner On The Government's Response To COVID-19 And The Future Of Level Targeting

by

Summary

David Beckworth: Scott, welcome back to the show.

Scott Sumner: Thank you for inviting me, David.

Beckworth: Oh, it's great to get you back on the show. And last time we chatted, it was late February. Well, a lot has changed Scott since that time, particularly with the economy, COVID-19 has ravaged upon the U.S. economy. We know first-quarter GDP is down 4.8%, probably be revised down even farther when it's all said and done. We know there's an unemployment rate, close to 15%, 35 million have claimed unemployment insurance, benefits. Retail sales is down. Even the CPI came down in negative territory for the first time since 2008. So a lot has been going on and I'm wondering, have there been any surprises for you or is this more or less what you expected?

Sumner: Well, it certainly wasn't what I expected a few months ago. Like a lot of people, I think I was caught off guard by the severity of the crisis. Although I suppose in retrospect we shouldn't have been, but yeah, I mean, it's a lot worse than I would've expected a few months ago. On other hand, once the lockdowns began and so on and in March it was pretty clear. We were going into a deep slump in the economy.

Beckworth: Yeah and the Fed has responded, Congress has responded, some really large engagements by the U.S. Government in trying to provide life support or relief to the U.S. economy as part of the fight against this virus. And I'm wondering, what is your general sense of the response by government?

Evaluation of Government Response to COVID-19

Sumner: Well, there's a number of aspects to it. So you have fiscal policy, which includes certain targeted programs, as well as the more general concept of fiscal stimulus. You've also got credit policies by both Congress and the Fed, trying to expand credit, and then you have monetary policy by the Federal Reserve. So I guess overall I have mixed feelings on what's been done. I would have liked to have seen a more expansionary monetary policy. In terms of fiscal policy, I would have preferred they focus on specific needs and not have focused as much on just sort of a general stimulus approach. So by that, I mean, specific needs might be unemployment insurance or addressing the medical crisis, things like that. Whereas say the $1,200 rebate was probably aimed more at just general stimulus rather than specific needs.

Sumner: And then the credit policies, I don't have strong views on that. It's not really my area of expertise, but I guess I wonder what exactly the purpose here is. It seems like we're giving assistance to some small businesses, but not others. So it's not clear to me why the program is sort of targeted in that way. If there is a desire to expand credit to business or specifically to small business, I guess I'm wondering why they don't just do something like subsidized bank lending to firms below a certain size and have all firms qualify for that program.

Beckworth: So Scott, we've had a number of guests on our show and I'd be interested to hear your thoughts on this, but the argument they've been making is that Congress should have done more of the heavy lifting. So the Fed has taken on a lot of activity. You can broadly think of two categories the Fed has been doing, a large amount of liquidity support, so I would call liquidity facilities. Everything from, its support of repo markets to support of money markets, primary dealers, commercial paper, even the dollar swap lines. You can think of all of those as supporting liquidity to the financial system and maybe even more precisely, liquidity to the shadow banking system. The Fed by law is supposed to support banking and the stability of the financial system.

Beckworth: So they're doing that, but then all these other facilities, which you could loosely call credit facilities, they'd be our second bucket, is kind of an innovation. It's a new area for the Fed to go into and a number of our past guests have suggested that may have been better done by having Congress do that. So for example, maybe Congress should have set up a facility or have done more through the SBA to reach small businesses. That the Fed is being asked to do something that's not well designed to do. Do you share that concern?

Lending Facilities: Congress vs. the Fed

Sumner: Yes. Well, I would prefer the Fed would focus on monetary policy. It's not so much that I think the Fed is less adept at doing that than Congress, but I think it sort of takes their eye off the ball in terms of what the most effective way of dealing with this crisis is. So for me, the best way to ensure liquidity is to have a monetary policy expansion or enough that we get adequate growth in nominal GDP in the medium term. I certainly don't think there's much they could have done in the second quarter of this year, but as we look forward to 2021, you'd like to have adequate growth in nominal GDP. And I think that sort of monetary stimulus would have a broad effect on the economy, helping many firms, many financial institutions in a much more comprehensive way than trying to just get in there and loan money to this or that company.

Sumner: So overall I would prefer the Fed focused on monetary stimulus rather than what you might call credit stimulus. And if credit stimulus is going to be done, some other branch of government probably should be doing that task. So yeah, I guess I'm in broad sympathy to some of your other people you've had on in that regard.

Beckworth: What do you think of the more general point that the Fed is kind of forced into providing liquidity to the financial system outside the traditional banking system? So a lot of these facilities that we just talked about, they're geared toward getting funding towards shadow banks, to money creation in firms that exist outside the traditional banking system. And so they're engaging in activity under 13(3), again the primary dealer credit facility, money market liquidity facility, commercial paper facility. This is a way for the Fed to support liquidity through this shadow banking environment. And I'm wondering, is this the best way to do that? Or you think just injecting through other markets is a better way to go?

Sumner: Well, I think that if we had a more expansionary monetary policy that would indirectly improve the liquidity in many of these markets you're talking about. Now, of course this crisis is different from anything we've seen before so I'm open to the possibility that at least for the next few months it's not optimal to try to maintain stable nominal GDP growth. Because it would be too distortionary, especially with all the lockdowns and so on. And if we make that decision that we don't feel we can do enough monetary stimulus to maintain adequate growth in nominal GDP, then we may decide we want to do more specific things to support specific sectors of the economy for this brief period of time, hopefully brief, when nominal GDP declines.

Sumner: But again, looking out towards the more medium term, I really think the best way to address this is to provide liquidity. It isn't just a question of the specific markets you intervene in. It's also a question of the indirect effects. Even if the Fed buys a lot of treasury bonds, the injection of money and the purchase of treasury bonds will indirectly support other markets, such as corporate bonds. You don't have to directly buy corporate bonds or common stock in order for the monetary policy to have a supportive effect on those markets. So I wouldn't necessarily rule out some of these programs, but I would put them second behind, first doing just ordinary monetary stimulus to a greater extent that the Fed has done that already.

Beckworth: Okay, Scott, so you have a policy brief that actually lays out how you would like to see the Fed be more aggressive and you've co-authored it with Patrick Horan, our colleague at the Mercatus Center and it is titled: *Reforming the Fed’s Toolkit and Quantitative Easing Practices: A Plan to Achieve Level Targeting.* And you go through several steps, but walk us through this policy brief, how can the Fed do more? How can it be more aggressive in traditional monetary policy that would, therefore, make it less urgent to do these other secondary policies?

How the Fed Can Conduct More Aggressive Monetary Policy

Sumner: Right, so we started with a basic principle that the Fed has two ways to boost aggregate demand or nominal spending and those are increasing the supply-base money and reducing the demand for base money. So QE is the most famous example of a program that increases the supply of base money. And in terms of reducing the demand for base money, you can do that in a number of ways. One is reducing the interest rate on bank reserves, even if necessary making a negative. But I would prefer a second approach, which is to do what's called level targeting. So under level targeting, you promise to come back to the previous trend line when there's a dip in inflation or nominal GDP. And in a crisis like this level targeting has the effect of raising expectations for spending growth in the future, for inflation in the future. And that tends to reduce the demand for base money as well.

Sumner: So I think level targeting is probably the single most important thing that the Fed should be doing that it is not doing so far. Although they've been talking a little bit about that idea. And I think a lot of economists focus too much on what are called concrete steps, like QE or interest rate changes and not enough on the importance of setting the right target. I think switching to something like level targeting is actually far more powerful than the things the Fed has done so far, such as quantitative easing and cutting interest rates. So that's where I would put my emphasis at this point.

Beckworth: Okay. Now, do you think it would be very easy to do a level target right now? I imagine a better time would have been before this crisis erupted but we are here now, so how would you guide the Fed into adopting a level target? What would it take in the midst of this crisis for the Fed to adopt a credible level target?

Sumner: Right. So although I prefer to use nominal GDP, I'll talk about prices because I think realistically, they're more likely to do a price level target. So they might do something like promising to get the price level back to a trend line of 2% a year growth, starting in say, December of last year. So by the end of 2021, that's a couple of years out, we would want prices to be 4% higher than at the end of last year. And I think the credibility problem is overrated. My view is the markets are very good at seeing what the Fed is likely to do. In fact, in most cases the markets have a better understanding of what the Fed will do than the Fed itself. And this has been the case in recent years with predictions of interest rate changes and so on. So if the Fed actually does intend to do level targeting, I'm very confident that the policy will be credible or at least will become credible in a fairly short period of time.

Sumner: Now in case I'm wrong, the Fed also has to adopt what's called a “whatever it takes approach” to asset purchases or money injections. So the Fed has to tell the markets essentially, that they plan to buy enough assets to raise inflation expectations up to their target. And this is something they haven't done. As you know, inflation expectations have fallen sharply during this crisis. And there's no economic model that I know of where it's appropriate for inflation to fall during an adverse supply shock. If anything, we should be having above 2% inflation in this kind of environment. So that's the low hanging fruit that the Fed really needs to address, is a commitment to come back to that previous trend line for either prices or nominal GDP.

Beckworth: And you mentioned in your policy brief that if necessary, Congress should allow the Fed to buy other assets, right? Beyond traditional treasury securities.

Sumner: Right. So there's an interesting story here because I've been advocating this idea for several years and I had no concept that the COVID-19 crisis was on the horizon. So here we are in 2020 and the Fed is sort of doing what it seemed like I was asking them to do. But I'm actually skeptical of what they're doing right now because my idea was that Congress should give the Fed the ability to buy essentially unlimited risky assets in a crisis, if necessary. To hit the nominal GDP target or the inflation target.

Sumner: And the idea was they would only do this if necessary, if they'd run out of traditional ammunition. Conventional policies, like buying treasury bonds or switching the level targeting. Some of the ideas that don't involve buying risky assets. So risky asset purchases were going to be a last resort. The reason I favored that ability was to make it clear to markets that the Fed had the ability to hit its targets because it could do whatever it takes. And in an emergency, it could even buy risky assets, if necessary to hit its target and that that would add credibility. So that's your earlier question about credibility, that's one of the purposes of allowing the Fed to buy a wide range of assets. So, while I'm happy Congress has essentially granted them that permission. I still would prefer that be done more as a last resort rather than a first resort.

Beckworth: So they are getting ahead of themselves by putting all this emphasis on the corporate credit facilities as opposed to what you would want. And that is more force on traditional monetary policy.

Beckworth: They should announced, say, “We’re going to return to 2% inflation or return to some kind of price level path. We'll buy as many treasuries as it takes. And if need be, then we'll buy corporate bonds.”

Sumner: Right. There's a very close analogy to the second half of 2008, which you remember well.

Beckworth: Yeah.

Sumner: In the second half of 2008, the fundamental problem with the economy was crashing nominal GDP. But the Fed did not actually know that at the time. And so they thought the fundamental problem was a banking crisis and what they needed to do was provide help for the banking system and that this would fix the problem. In fact, the banking crisis was worsening precisely because nominal GDP was falling fast. So what the Fed did is they provided a lot of liquidity to the banking system but they were worried that this would increase aggregate demand in the economy and create inflation. So they sterilized all those injections by paying interest on bank reserves.

Sumner: As a result, the economy continued to collapse sharply into early 2009. And then around March 2009, the Fed realized that the problem really wasn't the banking system. It was a collapse in aggregate demand like the early 1930's. And then they switched to aggressive monetary stimulus programs like QE. But by that time, it was too late, we were already in a very deep slump. And this isn't an exact analogy because the real shock this time is much bigger, obviously. So there's no amount of monetary stimulus that can prevent a recession when you have lockdowns and social distancing and so on.

Sumner: But nevertheless, I think there's too much emphasis on thinking that the real problem now is not enough credit being provided or not enough liquidity in financial markets. Whereas we're moving more and more towards a situation where the real problem is going to be just a deep recession caused by a lack of nominal spending. And so I'd like to see more done on that. And what I'm proposing in level targeting is not that radical. The Fed itself has been thinking in those terms, considering that option. It's been supported by a lot of prominent people like Ben Bernanke and so on. So I think the Fed really needs to be more aggressive in addressing the macroeconomic problem. And not assuming that just fixing these financial problems will solve the more fundamental problem.

Beckworth: Okay. Now I want to touch again on the purchase of corporate bonds. So the point you raised is that there should be an order to what the Fed is doing. And purchasing corporate bonds should be the bottom of that list, long after it adopts the level target, long after it buys vast amounts of government securities and then if needs to, resorts to these corporate bonds. So let's just say for the sake of argument, we do it in that order, we get there. There's going to be some observers out there who say why is the Fed jumping into credit markets? Why is it intervening? Why is it becoming a credit allocator?

Beckworth: In other words, some observers may think that the Fed is becoming very non neutral or more non neutral by getting into markets other than treasuries and there's actually a debate on this. There's a debate on what does it mean to be neutral? So we've talked about it before on the show but on one hand, there's the view that's held by most Americans and in the UK. And the view is that central banks should only purchase government securities or largely purchase government securities. And the idea is, from a consolidated balance sheet view, it's kind of a wash. So all the treasuries are being issued, are being bought up by the Fed. And the footprint of the government, from a consolidated perspective, is very small. And therefore, it makes sense to buy up government securities because it minimizes the government's footprint.

Beckworth: There's a German view, which is very different, which says the Federal Reserve should buy up assets in a manner that replicates or at least best replicates what the typical investor has in their portfolio. So in other words, central banks should buy a little bit of all the different assets that people are buying so that it's not disruptive. And in fact, the German view would argue by only buying government securities, you're being very distortionary, leaving a huge footprint and you're not being neutral. So two very different views of what is a neutral footprint or a neutral intervention by the central bank. One says only government securities. One says have a portfolio that mimics what the average investor holds. So I guess my question to you is where do you come down on that? And how would you respond to someone who says the Fed's getting involved in credit markets by buying corporate debt?

The Debate Over Central Bank Neutrality and Helicopter Drops

Sumner: Well, I guess I don't really agree with the German view you described. First of all, I don't think that that proposal would be neutral either. Unless I'm misunderstanding, if you buy corporate debt and corporate stocks, you're favoring big business over small business. Unless I'm missing the point there. But the Fed is, in a fiscal sense, part of the Federal Government and its profits go to the Treasury. And money creation creates seigniorage, which is a source of essentially, tax revenue for the Federal Government. So it just seems reasonable for the Fed to be buying back treasuries when they inject money. But if it actually were necessary to buy corporate bonds in order to hit the target for prices or GDP, then I would favor doing so with a provision that once the crisis is over, those bonds be sold back.

Sumner: So there would be no permanent investment of the Fed in the corporate bond market. And if traders knew that, knew it was just a temporary purchase, then hopefully the prices would not get too out of line with fundamentals. So it's a lesser of evils. I would be willing to listen to other proposals as ways to inject money in the economy other than buying treasuries first and then say, mortgage backed securities, second. Corporate bonds, third and so on. But I haven't seen any plausible arguments for other ways of injecting money that are more efficient. You do have of course, proposals for helicopter drops but I've argued those are very inefficient and essentially wasteful ways to get money into the economy.

Beckworth: Why don't you share with our listeners why you think helicopter drops are an inefficient way to get money to the economy?

Sumner: Well, these injections are presumably in some sense, temporary. Either the increase in the monetary base will later be withdrawn from circulation or at some point in the future, you'll have to pay interest on the reserves. That's to some extent what we did of course, in 2016, ‘17, ‘18. Either way, there's a burden on the Fed. And if it simply drops the money into the economy, when the money has to be withdrawn later to prevent high inflation, then you'd have to actually raise taxes to pull the money out of circulation. And that would require distortionary taxes in the economy and would be much less efficient than if you had simply injected money by buying assets. And you could then later sell those assets to withdraw the money from circulation. And not have increased the national debt, not have to raise taxes at some point in the future.

Sumner: So I think that used to be at least a fairly standard view. And I think that what's really going on here is the calls for helicopter drops are driven by a perception that monetary policy is out of ammunition. Now, if that were actually true then you could make that argument but as I said, I don't think the Fed is actually out of ammunition. They have a lot of things they could do, which would be much more powerful than what they're doing right now. So it seems to me, a helicopter drop would be the absolute last resort, even behind buying risky assets.

Beckworth: Interesting.

Sumner: And risky assets, perhaps when you sell them a few years later, you've lost 10% on your investment. If you do a helicopter drop, you've immediately lost 100% on your investments and that's a future liability for taxpayers. And even if interest rates stayed at zero forever, that would just mean the change in the nature of the economy gave the monetary authority a huge one time windfall that they could do with whatever they wanted. Perhaps of several trillion dollars. They would never have to pay it off but in that case, you'd want to ask yourself, what would be the best thing to do with this multi-trillion dollar windfall? Should we drop it out of a helicopter? Should we build high speed rail? Should we provide aid to homeless people? It's not obvious to me that dropping the money out of helicopters is the optimal way to hand out the money. And to me, that's almost what we did with the $1200 rebates.

Sumner: I think there’s no logic in why this group of people making up to almost $200,000 a year were the best use of the money. If we're going to do that, why not give twice as much to people with half as much income, for instance. I'm a little bit unclear as to what exactly the motivation is for some of these programs that are being adopted.

Beckworth: Well, let me ask this question. If the helicopter drop was tied to a target, let's say it's the Fed's inflation target or the price level targeting proposed, and it was only used to the extent that the target needed to be hit… Wouldn't it be permanent in that situation, and not need to be pulled back out, and therefore not as costly?

Sumner: Well, the concern here would be that it might be not costly in the short run, but then a few years later when interest rates rose, it would become very costly, because once interest rates rise above zero, you basically either need to pull that extra base money out of circulation or else you have to start paying interest on it through an interest on bank reserves program. If interest rates stayed at zero forever then you could argue there's no cost other than maybe an opportunity cost of what you do with the money.

Beckworth: I was thinking more along the credibility lines that you mentioned earlier. If you've got a credible target maybe you don't need to pull it out, but I guess you'd say if you've got a credible target, then why do the helicopter drop in the first place? This becomes a chicken and the egg-

Sumner: Here's where I think I could agree with you on this. If we arrange things in terms of what you do first, what you do second, and as you get increasingly desperate, you do increasingly desperate things. I'm of the view that there's almost nothing worse than a deep depression in the economy. If we ran out of every other option and the only option was helicopter drops for a deep depression, I would certainly support helicopter drops in that case, but I worry we're rushing too quickly. Basically we go right to fiscal stimulus before we've even tried any serious monetary stimulus.

Sumner: How can we say we're serious about monetary stimulus if we haven't even done things like level targeting, which so many leading experts, including the former head of the Federal Reserve, Ben Bernanke, have advocated? I don't think there's even people within the Fed would argue that there's really very strong arguments against doing level targeting at a time like this. You and I both know they've been talking about it in their review of options going forward that's been ongoing for the last two years or so. If we haven't even done that, and we have a lot of treasuries and mortgage backed securities that are relatively safe assets that we haven't bought, why would we even be talking at this point about doing fiscal stimulus?

Sumner: It's not clear to me what the argument is. Maybe there's an argument. And just to be clear, I do believe we have to run larger budget deficits this year for ordinary fiscal reasons; tax revenues fall, unemployment payments rise, and so on. It's appropriate to run a bigger deficit during a deep slump. I'm talking about just doing spending for its own sake, for no other reason than to get them to boost aggregate demand. That's the thing that doesn't make any sense to me.

Beckworth: That's a fair critique, that monetary policy hasn't done more, just narrowly defined as what monetary policy normally does; commit to a level target and buying up as many assets as needed to hit it. The Fed hasn't done that, which kind of is ironic because on one hand, Congress is leaning heavily on the Fed to do more, but the more that it's asking it to do is much more in the credit area, which we would call fiscal policy. The Fed is being asked to do things that's probably better suited for Congress to do, or therefore fiscal policy, and yet no one is asking the Fed to go ahead and adopt a level target.

Beckworth: I've always thought it ironic for example, that President Trump as he berates the Fed for not doing enough, he never once has called for a level target. Maybe he doesn't know about the level of targeting idea, but you're right, if the Fed has not been fully exhausted on its normal operating space, why ask it to go to a secondary place where it's less experienced, less comfortable probably, and really better done by Congress? So, those are all fair points.

Sumner: Yeah, no disrespect to Congress or the president, but you almost wouldn't expect them to have a very sophisticated understanding of things like level targeting. This is a real sort of inside baseball idea, and even a lot of economists don't follow this issue very closely. That's why I was, for the last three or four years, calling on the Fed to go to Congress and explain to them why they needed permission to buy more assets during a crisis, and maybe go to Congress and talk about the advantages of level targeting, and explain these things.

Sumner: I personally think level targeting involving the price level would be relatively uncontroversial because you'd still have that 2% number out there so people that don't follow it closely would still view it as a 2% inflation target. I don't really expect people outside the Federal Reserve to have a very sophisticated understanding of all the options for monetary policy, and I'm not at all surprised that Congress would look at the Fed more from a credit perspective.

Sumner: An institution that can help provide loans, because the Fed itself talks about its role in banking terms as a central bank, as something that works through the banking system. Even though in my view, it's really a monetary authority that is quite separate from banking. It's providing enough money to keep nominal spending on target, and banking is sort of a side show in my view, but that's not the standard view. I'm not surprised that in an emergency Congress would look to the expertise of the Fed, and call on them to provide help in getting credit flowing to businesses.

Sumner: Congress is probably aware of all the gridlock in Washington, and aware that there is a lot of expertise of the Federal Reserve. So even though, as I said earlier, in principle you'd like fiscal policy to be doing these things, to me, it's not that surprising that in an emergency an institution like the Fed would be called on to help out in this. It's also not surprising that Congress would not be pushing the Fed very hard on monetary policy because most people and even many economists think when you've cut rates to zero, there's nothing more the Fed can do. And so, I really wouldn't expect them to push much harder on that issue.

Beckworth: Yeah. I wonder too by asking the Fed to do so much with credit policy, and the most likely disappointing outcome that's going to emerge from it we've talked about with previous guests, that the Fed is not very good, its hands are tied by the Federal Reserve Act, there's probably going to be a lot of disappointed stakeholders when it's all said and done, that that disappointment may harm the ability of the Fed to do traditional monetary policy.

Sumner: Yes, because it becomes politicized.

Beckworth: Yeah, politicized. Our previous guest Lev Menand, we talked a lot about some of the legal challenges the Fed has with its policies, with the credit facilities in particular, and it just seems like it's not only going to be politicized, but actually setting itself up for maybe lawsuits in the future. Typically, the courts have avoided the Federal Reserve.

Beckworth: Usually when someone has attempted to sue the Federal Reserve, they don't have standing in the eyes of the court, but it could be the case that there'll be some small business or some trade association of small businesses that might feel they got left out, and take the Fed to court. If that does occur and they do get a day in court, it opens up a whole can of worms that the Fed may not want to have to deal with. Again, it erodes the ability to do its traditional monetary policy role.

Sumner: Yeah, that's definitely a danger.

Beckworth: Well, going back to the level targeting idea, I wonder how you feel about it. Where do you think it's going to be going? We've talked about it for a long time, Scott. You and I, we talked about it from the last crisis coming through this crisis. The Federal Reserve is doing a review, and last we heard they would end it in June. I wonder though if this crisis has kicked that can down the road, and just in general, I don't get the sense, and I think this is what you're saying with our conversation so far, that there's really a lot of interest in talking about something like level targeting right now.

Beckworth: In fact, I was on Twitter recently, I got into an exchange with a good friend about this review and wondering whether it would actually take place, and he said, "Well, what's the point? Doesn't the Fed have bigger fish to fry right now?" And I was like, "Well, no, actually. You want the Fed to have level targeting so that when the recovery takes place, the Fed doesn't feel handcuffed by 2% inflation." Because I really do have this fear what's going to happen is as soon as the recovery starts to take place and we get close to 2% inflation, the Feds are going to start feeling pressure from the outside, maybe even from the inside, to dial things back.

Beckworth: If it had a level target, it would have more degrees of freedom and could continue to support a robust recovery. And so, I do think it's a huge deal. I do think it's consequential that we have a level of target, but I don't hear many people talking about it. I think that's the point you're making. I'm just wondering, how do you feel about the future of level targeting? What is your sense of how near it is?

The Future of Level Targeting

Sumner: Well, I'm a little more pessimistic than a few months ago because I would have thought in this sort of crisis, the Fed would have already done something along these lines. Ironically, I think the thing that makes it so difficult is although a lot of outsiders don't think it's an important issue, I think within the Fed their reluctance is due to the fact that they understand it is a very, very powerful tool. It's so powerful that it's a little bit frightening because under level targeting in a way everything is different, now the Fed has ownership over the path of some sort of nominal aggregate like prices or nominal GDP.

Sumner: There's nowhere to hide if they're off course, and they would no longer be viewed as an institution making a series of gestures that might or might not work, they'd be viewed as steering the ship and they'd be judged on that basis. They would have a very strong obligation to do whatever it takes to over time get back on track. That's one of the things that makes it so powerful, but at the same time, it makes the Fed reluctant to commit to something like that because they know that it would increase their responsibility quite a bit.

Beckworth: Okay. Let's move from level targeting to specific types level targeting. You mentioned price level targeting, there's nominal GDP level targeting, and then there are many forms of price level targeting. You can think of average inflation targeting or temporary price level target by Ben Bernanke. Do you have a sense of which form of level targeting is most likely to emerge?

Sumner: My fear is it'll be watered down like average inflation targeting, which is sort of halfway there, but not really committing to level targeting. I do think Bernanke's idea is a good one, of doing it temporarily during a crisis because this is really when you need it. In normal or good times, I think inflation targeting is probably almost as good as price level targeting, but I fear that it will be something watered down because that's how big institutions are, they compromise, they're reluctant to take a leap into the dark.

Sumner: There's some benefits of course in that caution, in general, but in this case I think we really do need decisive action. Bernanke used the phrase ‘Rooseveltian Resolve’, talking about FDR's decision to devalue the dollar in 1933, which was very much along a level targeting sort of line of approach to monetary policy. That had a truly dramatic effect in the economy, turning around the economy almost immediately.

Sumner: But, that was Roosevelt. The Fed wasn't willing to do that in 1933, and obviously this is such an esoteric issue that I don't expect Congress to force level targeting on the Fed. They probably shouldn't because they don't have any expertise in that area. So, I'm a little bit more pessimistic than I was a few months ago. I do think the Fed will do something perhaps like average inflation targeting, which probably would have been enough in an ordinary recession-

Beckworth: But not enough now.

Sumner: But in this case, level targeting... And I hope I'm wrong. If the medical situation improves more quickly than I thought, maybe we'll have a strong recovery, but right now I'm very concerned about the high unemployment lasting for quite a long period.

Beckworth: Yeah, absolutely. And just to reiterate, some may question, why are we spending so much time talking about level targeting in a crisis caused by a large supply shock? I think the easy answer is we see inflation dropping, we see nominal GDP dropping. All the things that you mentioned earlier indicate that it's not just a supply shock. This has transformed into a large demand shock as well. Therefore, you need to at least address the demand side that has taken place, the developments there.

Sumner: Right. The demand shock has actually been larger than the supply shock. Otherwise, inflation wouldn't be falling right now.

Beckworth: Right, exactly.

Sumner: Kind of amazing.

Beckworth: Another way I like to tell this to people is that a supply shock definitely makes us poor. We know that's the case. We can't work as much as we could. We can't engage in business like we once did.

Beckworth: But what we want to avoid is a secondary spillover effect that comes from when people have financial obligations that they have to meet. If they've got mortgages they've got to pay. If you're a business, you've got leases and payrolls you've committed to. If you can't make those, then it's going to cause problems, cause stress, cause disruptions to economic plans that are made. So the best thing that the Federal Reserve can do is to maintain a stable level of nominal income, which is equivalent to a stable level of nominal demand or nominal GDP targeting.

Beckworth: So there's an easy way to tell the story, I think, but, as you mentioned, the argument for level targeting isn't being made very much in this crisis, and we may suffer as a consequence.

Sumner: Right. It's hard to really think of an argument in favor of allowing inflation to fall below that 2% trendline over the next few years. I can imagine arguments against nominal GDP targeting when there's a lockdown. If 20% or 30% of the people can't work, keeping nominal GDP on track would distort a lot of relative prices in the economy. But for inflation to fall during an adverse supply shock just makes absolutely no sense to me. I can't think of any argument for monetary policy not being more expansionary.

Beckworth: I would say to people who made the first criticism, "Why would you want to do nominal income stability, or why would you want to have a stable nominal GDP growth with the large supply shock because of relative price distortions?" I would say, look, you can gradually adjust what you think is the level of nominal GDP. You should stabilize and then go from there. There's no reason you can't adapt along with what's happening to the underlying fundamentals of the economy.

Sumner: Right. I mean I think George Selgin's work suggested something like ... Not exactly, but more like stable nominal GDP relative to the workforce per worker or something like that.

Beckworth: Kind of a labor income target.

Sumner: So I mean you could view this as almost like equivalent to a drop in the workforce if you're going to tell people they cannot work. So it's a little different than the normal crisis where, in a normal recession, if you pump more money in the economy and get people spending, then you get back to full employment. At least for April and May, June of this year, that's probably not the case. But I do think that by, say, the fourth quarter of this year, in 2021, we should definitely be aiming to get nominal GDP back up close to the trendline again.

Beckworth: Okay. Now let's move from level targeting to tools. One of the tools that's being discussed is negative interest rates. Just this week, Jay Powell said no one at the FOMC really wants to touch negative interest rate policy. You have a post up that helps us better understand what he's talking about. The title is, *Negative Interest Rates and Negative IOER.* So walk us through that post and how it helps shed light on this issue.

Negative Interest Rates and Negative IOER

Sumner: Okay. So in that post, I did a thought experiment going back to 2006. At that time, there was no interest on bank reserves so you could treat the interest rate as zero. The interest rate on other safe assets like treasury bills was about 5%. I asked, well, what would have happened if the Fed had cut the rate from zero to -25 basis points back in 2006? I argued that almost nothing would happen because there were very few excess reserves. So the amount of interest banks would have had to pay back to the Fed would have been utterly trivial, less than $5 million in total for the whole banking system.

Sumner: So the reason I did that thought experiment was to drive home the point that there's a big difference in principle between, interest on bank reserves that is negative, which is a Fed policy, and market interest rates that are negative. This needs to be emphasized because in recent years, the countries that have negative interest rate on bank reserves also have negative market interest rates.

Sumner: But they're two very, very different concepts. Negative interest on bank reserves is an expansionary monetary policy because it just discourages banks from holding reserves. It reduces the demand for bank reserves. Anything that reduces the demand for money is expansionary.

Sumner: On the other hand, low market interest rates are not, in general, expansionary, rather they're usually a reflection or a side effect of a contractionary monetary policy, which has caused various low growth in nominal GDP and has driven free market or equilibrium interest rates into negative territory. That's what you see in Europe and Japan. Very weak nominal GDP growth for a long period and negative market interest rates on government bonds because of the weak economy there.

Sumner: So if we do a policy of negative interest rates, and I have mixed feelings on that, that is if we pay a negative rate on bank reserves, the goal of that policy should be to raise market interest rates higher, especially on longer term bonds. If longer term interest rates do not rise in response to that policy, then the policy is not effective. It hasn't been embedded in an overall regime, like level targeting, that is going to be effective.

Sumner: So I really like to emphasize the difference between individual tools like QE and negative interest on reserves and the broader monetary regime, like what you're trying to target and whether you have a commitment on whatever takes commitment to hit your target.

Sumner: Often these get mixed up. People see central banks doing QE or negative interest on reserves. They see the economy continue to be weak and they think, "Well, those tools obviously don't work," where, in fact, what's actually happening is the central bank is being defensive. They're responding to a weak economy with a high demand for liquidity by cutting rates and injecting money. But they're not being proactive in actually changing the expected rate of inflation over time. That's what we really need our central banks to do, to be much more aggressive and proactive.

Sumner: And so, when we talk about negative interest rates, there's two types that can be a reflection of a failed contractionary policy or it can be a tool that is used in an expansionary way and to determine which it is you have to look at the whole set of monetary policy actions. You have to look at how markets are reacting to them. Only then can you get a sense as to whether the tool is being used in an effective way or an ineffective way.

Beckworth: Yeah. So this underscores your point about the importance of getting level targeting. Regimes matter and they matter greatly, particularly in moments like this. And so, no matter what tool you use, it's only going to be as effective as the regime it's a part of.

Beckworth: So I make this point myself. Whether you do helicopter drops, yield curve control, negative interest rates, QE unlimited, all of those are tools and they are part of a bigger regime. Right now the regime we have in place is a low inflation target. Again, going back to my concern that no matter what we do, if we get near 2% inflation, it's going to stall the recovery. That's why I think it's important that we get a level target in place ASAP.

Sumner: This is almost the perfect example of a system that's full of cognitive illusions. Almost everything is the opposite of the way it looks at first glance. People think low interest rates are easy money, but they're much more often a reflection of a previous tight money policy. People think QE is easy money, but actually the tighter the monetary policy generally, the lower the inflation rate and the greater the demand for base money.

Sumner: So what you find is when you look around the world, the countries where the central bank balance sheets are the largest as a share of GDP, Japan, Switzerland, and so on, are the places with the lowest inflation. The places with high inflation are those where the central bank balance sheet is the smallest as a share of GDP.

Sumner: So this is all very counterintuitive. So if your goal is to, say, raise inflation, you might, in the short run, increase the size of your central bank balance sheet, inject money in the economy in the hope that this leads to higher inflation, less demand for liquidity, less demand for base money, and ultimately a smaller balance sheet as a percent of GDP.

Sumner: That fact is really hard for a lot of people, I think, to wrap their minds around. And so, there's this persistent pessimism I'm always fighting against, which is that look at what the central banks have done. It hasn't had much effect. Think how much more it would take.

Sumner: In fact, if we had a successful policy, it would take much less. We would have higher interest rates and we would have a smaller balance sheet as a percent of GDP. But it's hard for people to imagine that we could ever get to that place, even though we were in that place as recently as 15 years ago.

Beckworth: Yeah, that's exactly right. That's why we do what we do at the Mercatus Center, in the Monetary Policy Program.

Beckworth: Well, with that, our time is up. Our guest today has been Scott Sumner. Scott, thanks for coming back on the show.

Sumner: Thanks for inviting me.

Original Post

Editor's Note: The summary bullets for this article were chosen by Seeking Alpha editors.