Macro Lit Review 4: Highlights from Mid-2023 with George Selgin

George Selgin is a senior fellow and director emeritus of the Center for Monetary and Financial Alternatives at the Cato Institute. George is also a frequent guest on Macro Musings, and he rejoins the podcast to talk about some of the recent developments in the monetary and financial policy space. Specifically, David and George discuss the history and present developments surrounding FedNow, the future of real-time payments, how to revise the Fed’s operating system, whether the Fed is currently delivering on a soft landing, and a lot 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 [email protected].

David Beckworth: George, welcome back to the show.

George Selgin: Great to be back on, David. It's always a pleasure.

Beckworth: It's great to have you on and as listeners will know, you are a regular on the show, but this is the first time, George, I've had you on as a regular where you are recording from Spain and I am recording from the United States. This is a first for our ongoing discussions that we're having on the podcast.

Selgin: Well, it's very hot here, as it is in many other parts of the world. For me, it's four o'clock in the afternoon and that is close to the peak temperature time. I'm very happy to be indoors just doing a podcast.

Beckworth: Nothing more delightful than discussing monetary policy on a hot summer day. Now, you are, again, in Spain. We're here in the US. It's always great to be in a world at a day and age when we can communicate with people across the world, which we do a lot on the show. Now, last time we chatted, George, you were giving us an update on your book on the Great Depression. Maybe give us a recap of what the book's about and where it is in the production process.

Selgin: Sure, yeah, thanks for asking. It's a book about how the US recovered from the Great Depression. It's about the recovery, not the cause of the depression, which makes it a different contribution from the more common ones on that subject. It's specifically about the New Deal and to what extent it helped or did not help to hasten the US recovery. That's what it's about and it is just about done. I am at the stage of getting, it's a long book, it's a big book. I'm at the stage of getting feedback from experts on the various chapters. There's about 30 of them. And that's been going very well. So, I'm hoping to deliver the manuscript to University of Chicago Press on time in September.

Beckworth: Well, fantastic. Well, we'll have you back on the show when it does come out and we'll discuss the book. We've had a previous podcast, we'll provide a link to it in the show notes for our listeners, where we go over the book, but we'll do another show once it comes out. George, I love getting you on the podcast to discuss recent developments, happenings in the space that you and I care about, monetary policy and financial policy. We often do this routine where we will go through three articles or three recent developments that we want to discuss. So, George, you are the guest. I'll start with you. What is the first topic that you want to bring up in our conversation today?

The History and Details of FedNow and the Future of Real-time Payments

Selgin: Oh, that would be FedNow, David. I think most of your listeners will know that FedNow is the real-time retail payment system that the Fed has been developing for several years now and tomorrow is its launch date. This is a big deal and there's been a considerable controversy about FedNow, and some of it is, I think, off-base, a lot of it. But I myself am a critic of FedNow, and I have been since the get go. In fact, I got a, I think it was in 2017, maybe even before that, I got a memo from a banker saying, "The Fed is thinking of doing blah, blah, blah and you need to look into this." I said, "All right." I didn't have any preconceptions. But once I read up on what was happening, the history behind it, I became very critical and wrote comment letters and testified to Congress. Anyway, I've been on top of this thing from almost the very beginning of the Fed's announcement of its intentions or potential plan.

Beckworth: So, we are recording this on July 19th. Tomorrow is July 20th when the FedNow comes out. FedNow is a real-time payment system and I'll let you go through the details of it, but it does have some banks that have agreed to participate in this opening. Also, the Treasury, the fiscal service, which may be a big deal. Love to hear your thoughts on that. Maybe walk us through the history of the FedNow and then what you see is happening to the future of real-time payments, these two competing systems.

Selgin: Sure. The basic story is that for some time other countries have been developing payments arrangements that could allow payments to be accomplished in real time. That is, you order a payment to be made from your bank and it's instantly done and the money is transferred. The accounts are settled all in a moment and there's no delay. There are no delays in settlement, which is the actual transfer of Fed funds or reserves in our country to the recipient bank. In the United States, we didn't have anything quite like that on the national level. Instead, the various legacy payment systems that the Fed operated all involved delays in settlement and could involve very long delays over weekends and holidays or both combined, so that even if some somebody's bank received a notice of a transfer, the money may not be credited to the recipient's account for some time.

Selgin: And these kinds of delays could be important and costly for all kinds of reasons for businesses, but also for people who are living paycheck to paycheck as it were, and will have to go to some payday lender to cover themselves over during the settlement delays. Anyway, this was a big problem. The Fed, in 2014, set up a task force with many participants from the private payments community, consistent with its rules, longstanding rules now, where it doesn't supply payment services itself if the private sector can do it. I think that was a good rule. It's a rule that now the Fed seems to have thumbed its nose at. So, it had this task force and it solicited proposals. One of them was from TCH, The Clearing House, which is the new name of a very old institution.

Selgin: It used to be called the New York Clearing House, it started in 1853. They've been doing payments and settlements and clearings for banks for a long, long time and they proposed a real time retail system. They were given very high marks for the proposal and they went ahead and launched it, they were working on it 2015, I think it was up and running by 2017. They now have several hundred participants and they all clear on the books of a special account set up for the purpose by the clearing house. But all the clearing is done, settlement is done instantaneously on the books of that institution, which is acting like a little central bank for the purpose. Anyway, TCH was well into this and this is very important. It was working towards achieving ubiquity, which wasn't easy, but they’ve got a lot of banks on board. Smaller banks were not happy to participate, and apparently they complained a lot to the Fed. The Fed suddenly announced, to the great surprise of the TCH, that it was going to consider making another system that would compete directly with the private one. They went through the usual dog and pony show of soliciting comments. They got a lot of support, but a lot of criticism, including mine. They also interestingly asked for comments about revamping the legacy payment system so they wouldn't close on holidays and weekends and was otherwise able to achieve not instant payments, but pretty fast payments.

Selgin: Everybody supported that or something like it. Well, the Fed decided to go ahead with what became FedNow, its own parallel clearing, real-time retail payment system. In response to the favorable comments about fixing up the legacy system, which would've been a much easier thing for it to do, and not so redundant as what it has chosen to do, it said, "Well, maybe we'll solicit comments on this," which was an odd way to respond to a bunch of positive comments it had already solicited. Anyway, so I thought the whole thing stunk to high heaven, frankly. Now, here we are, the Fed has been experimenting. It had 57 odd trial participants, many fewer than RTP, the private arrangement now has, it has quite a few participants. I'll stop with this last remark.

Selgin: When you're trying to achieve ubiquity, the Fed's term for a universal network, which is the best kind of payments network, that is one where everybody is part of the same network. It's easier to do when there's only one option. The minute the Fed put its hat in the ring as it were, it created a dilemma for the banks that had not joined RTP. Now, they had to decide whether to wait several years and join the new system or to go ahead and join RTP because the systems are costly to hook up to. You don't want to spend money on both. What the Fed has done essentially is to slow down the progress towards ubiquity, which I think we would have a near ubiquitous RTP system by now if it hadn't been for the Fed's action. As it is, RTP covers about 90% of all bank deposits, and that's pretty good, but it's not quite what we want, what's ideal. Anyway, that's the story.

Beckworth: So, the small banks and financial firms wanted something other than RTP because they were worried about the big banks?

Selgin: That's correct. The big banks owned the clearing house. They saw a conflict of interest even though the clearing house has been doing clearing and settlements, as I said, since 1853 for all kinds of banks. And,  should say, the clearing house also contractually agreed to charge fees for the real-time payments that would not be to the advantage of big banks. There's no volume discounts, it's a flat fee per transaction, and they were committed to that. For that reason, they had the full blessing of the antitrust department, which wrote commending their proposal until the Fed said it might enter. The reason TCH then said, "Well, we may not be able to stick to our commitment to the small banks,” not because they were going to raise their fees on them, but because they feared with reason that the Fed will end up offering big bank volume discounts. There's historical reason for that fear, in which case RTP, the private system, would have to do the same in order to compete with the Fed. That was their reason. Volume discounting is more likely, now that the Fed is in the picture, someday in the future, than it was before. The small banks that worried about dealing with TCH may regret the choice they made in many of them, in lobbying the Fed to compete with that network.

Beckworth: That is the irony of this situation. The small banks clamor for an alternative to RTP because it is owned by the big banks, but in turn, that enters another competitor. Now, you lose some of that scale efficiency and the Fed has to recover cost according to law. So, it may start offering volume discounts and that in turn would then force the hand of RTP. So, the small banks may have shot themselves in the foot when it comes to, at least, cost,  and it's interesting to note that the FedNow, a lot of the early adopters, most of them are smaller financial institutions in addition to the US Treasury. I guess the US Treasury would be the big one. George, two questions. Number one, will the Fed stick to the law that says it has to cover costs, because part of your argument-

Selgin: No, it hasn't even pretended to be doing that. That's not an essential part of the argument though, David. The reason why I think the Fed might go to volume discounts is not to recover its costs, it's to get all the big banks on board, and that'd be the easiest way for it to do that, and incidentally, of course, improve its bottom line. But as for meeting the requirement, not a strict requirement, but as for recovering its costs in the way that the 1980 Depository Institutions Deregulation Monetary and Control Act, the way that that specifies that they should cover costs if they're going to provide payment services.

Selgin: But the Fed has pretty much made it clear that it's not doing that. It has the right to not do so if it's in the public interest, which is of course a loophole. You can drive almost anything through if you're a clever government bureaucrat. But more than that, they didn't even have a pricing schedule when they decided to go ahead with this arrangement. They didn't even announce their pricing schedule for many, many months after they were in the process of building the system. When they did announce their pricing schedule, it was practically a copy of RTPs. There clearly was no attempt to do a cost and benefit analysis. Although the Fed pays lip service to cost recovery, they have said that it may take them a long time, which is another way of saying they're not going to really recover costs, including the opportunity costs of capital.

Selgin: If you can take forever to recover your costs, meaning that you get the same number of dollars you spent eventually, you're not really keeping to the spirit of the law. They've already admitted that they're not doing that. They're not going to have a positive present value project. But they may raise their fees to entice the biggest banks, especially, to join them. The other thing is, they have other ways to entice those banks. They have all kinds of ways as a regulator to put pressure on institutions to hook up to their system instead of hooking up to RTP or in addition to doing so. They have all kinds of ways to do it. They have all kinds of cross subsidies they can resort to, they can lose money, and most of all, they never have to go out of business.

Selgin: I think it's very bad for the Fed to compete with private institutions precisely when it isn't absolutely necessary, precisely because the Fed never has to lose these competitions, no matter what happens. It doesn't have to make it economically worthwhile. It doesn't have to be in the black, ever. As long as it has seigniorage that you can use for cross subsidies, which right now is a little bit of an iffy thing, and it regulates The Clearing House too. It's directly regulating its rival. There was one more thing I wanted to point out, which is that if the concern was… and there are only two concerns that I've heard raised that could justify this competition. One is, "Oh, well, if one system breaks down, we have another one."

Selgin: Well, this is a very expensive kind of redundancy. There are a lot of ways to build redundancy into a network other than building another network, which is what they've done here. By the way, the networks are not interoperable. It’s not the case that by being a member of one network you might as well be a full-fledged member of both. It doesn't work that way. Ubiquity requires that everybody be in one network or at least one. Redundancy, there's a lot of cheaper ways to do that and preserve the network economies that are so desirable. The other argument is, "Oh, well, eventually the RTP could use its monopoly power to change its mind and overcharge the small banks or give benefits to the big ones."

Selgin: But here again, there's a million ways the regulators could prevent that. There's the antitrust department, there's the Fed, which has the power to persuade RTP not to do something like that. There are all kinds of ways the Federal Reserve could guard against RTP engaging in any abuse other than by building a completely separate parallel real-time payments system. I think this has all been very, very bad. I also had my doubts about how popular it's going to be. I think it could become a white elephant. We could even end up with two white elephants if other private payment innovations keep on going on, which are bringing us closer to real time payments without either of these other big projects.

Beckworth: Well, that answers my other question, which was what's the future? They may not work, they may not be what they were billed to be and other technology may pass them up.

Selgin: That's right. I think that it's quite possible that neither of these big networks will become ubiquitous, that they won't be used enough, that RTP may have to drop out. The Fed will certainly keep in business whether it's... it really would survive as a private payment supplier or not even with a monopoly. You'll have this expensive underutilized network and in the meantime, other little alternatives, fintech alternatives, will be providing the equivalent of real-time payments for the vast majority of people who need them. I think that that will happen, especially if, and I know this is going into another topic, but especially if the Fed can be convinced to quit slamming doors on fintechs that want master accounts. The other thing I just want to emphasize, it is a disgrace that the Fed has dragged its feet refusing to make its legacy settlement systems available 24/7, 365, something it has been promising to do since at least 2015. Most slow payments are because of that failure, and that would've been low hanging fruit for the Fed to fix. I think, to be cynical, that it hasn't fixed it because it wanted to do this more glitzy and expensive and unnecessary project instead.

Beckworth: Okay. Well, that point you made on Fed master accounts is a nice segue into the article I wanted to bring up, my first one, so I'll use that. That deals with the Fed master account issue. We've had someone on the show in the past, Julie Hill from the University of Alabama, she's a law professor there. On this show we went over a previous article of hers called, *Opening the Federal Reserve Account.* She goes through a number of case studies and some of the challenges. Well, she has a new article out, and for the sake of time I won't spend a lot of discussion on this. I'll just give the highlights and we'll move on to our next set of articles. But her new article is titled, *From Cannabis to Crypto: Federal Reserve Discretion in Payments.*

Beckworth: What she's done in this new article, and it's online, we'll provide a link to it, is she goes through the recent database that was released by the Federal Reserve, looks at all the master accounts that have been issued or applied for, that's by law now required by the Fed to disclose. She looks at that, she goes through legal history and says that the Fed is still exercising an excessive amount of discretion when it comes to making decisions for master accounts. She goes through a number of cases, the Cannabis Bank, the Narrow Bank, Crypto Custody Bank, International Trade Bank. In her previous paper she talked about the American Samoa Public Bank, a number of examples that really illustrate this challenge. I'll encourage listeners to go check it out. Again, it's going to be an ongoing issue, as you alluded to, George. I just want to tie these two points together, though. You talk about FedNow, I mentioned the master account issue, and what they both have in common is a certain regional Federal Reserve Bank, the Kansas City Fed. Is that right?

Selgin: That's right, yes. The Kansas City Fed seems to be the gatekeeper.

Beckworth: Okay. Well, I feel sorry for them because they got two very hot issues. This might explain in part why they have not been able to get a president yet. So, George, before we move on, do you want to respond to Julie Hill's work?

Selgin: Yeah, sure, David. Well, my first response is that it's excellent work. I've been following Julie's research and writing for some time now, and we share interest in many of the same issues. In fact, I was also on board the marijuana thing very early because I heard from counsel from Fourth Corners, whatever, it's been a while since I've read about them. But they were the marijuana bank trying to get a Fed master account some years ago, and were thwarted with what I also thought were very bad arguments. Although, in this case I sympathize a little bit more with the Fed because the federal laws are behind the state laws when it comes to marijuana legalization. It was an uncomfortable situation for them to be in, to be sure.

Selgin: But ultimately though, it has turned out to be part of the pattern of refusing master accounts to what we might call unusual banks of all kinds. That pattern, I believe, is now standing in the way of some really desirable payments, potential payments innovations. Of course, we don't know what all of them… which ones would work, which ones won't, but what we won't have is the experimentation that lets us find out. The thing is that the Fed seems to have closed the door on these unusual banks when it comes to setting up good payments, alternative payment systems. It has decided to compete head on, I think, which is always a bad proposition, with the big banks when they've tried to do something and it has entered on its own. The big picture here is that the Fed seems to be saying, we want to have payments innovations, but we want to do it all ourselves. We don't want to allow other private market participants to have much opportunity to do it. That's a very different policy than in the past, and I think it's a very, very bad policy.

Beckworth: Okay, let's move on to your next article that you wanted to discuss, so what was that?

Getting Off the Floor: Revising the Fed’s Operating System

Selgin: Well, it was, *Getting Up From the Floor,* an article by Claudio Borio at the BIS, and Claudio's done a lot of really excellent research. But with that particular paper, he's joined the small but very elite set of economists-

Beckworth: Exactly.

Selgin: ... who do not like the Fed's post-2008 operating system, known as a floor system, where it floods the banking system with reserves and encourages the banks to maintain these high levels of real reserve balances by paying interest on reserves, so reserves are no longer scarce, and the way the Fed sets the stance of monetary policy is by changing the interest rate that it pays on reserve, [and] also, some supplementary interest rates that it pays non-banks for keeping money with it through the overnight repurchase facility. That's the floor system, and anytime somebody new comes out for getting rid of it, I throw a little party, sometimes usually by myself. Claudio has now given me an opportunity to celebrate.

Selgin: There aren't many people taking this view, but I think their numbers are growing and I think there's a very good reason for it. If you go back and look at all of the arguments given for switching to a floor system, after the fact, mind you, because when the switch happened during the 2008 crisis, it wasn't planned that way. There was no immediate sense that what had happened would become a permanent arrangement, but ultimately, it did. Anyway, all kinds of rationales were given for making it permanent, claiming that it would keep banks liquid, that it would make monetary control easier and blah, blah, blah. Every one of those complaints has been shown to be false. The system has not proven easy to operate, and the cost has been immense.

Selgin: If anyone had been told at the time when this decision was made to have a permanent floor system, that it would take many trillions of dollars of reserve balances to keep the thing working, I don't think that anyone would have gone for it, nor would they have gone for it if they were told that this system might create circumstances like the present, where after interest rates start going up, the Fed is losing money hand over fist because it is holding on to all of the long-term securities that are backing the large scale reserve balances and it's taking losses on those holdings, or rather it's not booking the losses, but it is reducing its remittances until such a time, if it ever comes, when it can make up for those losses.

Selgin: It hasn't been a very good system. It has killed the interbank unsecured lending market, and that market, economists used to know, played a very important role because it was a way of limiting contagion by encouraging interbank monitoring. Obviously, if banks have to deal with other banks and make unsecured loans or borrow without security, they have to know more about each other or they certainly have to have established some tighter relationships. This was a very important source of discipline in the old days that it's typical… regulators always ignore market-based sources of discipline and regulation, and sometimes because they do that, they kill them. What they get instead is never as good. That's where we are today. Anyway, that's what Claudio's article is. I encourage anyone who's interested in this issue to look it up.

Beckworth: Yeah, let me read a quote from that article, which I find interesting and speaks to this point you just mentioned about the interbank market drying up because the Fed is flooding the market with settlement balances, reserves. He says, "One of the costs is that the abundant reserve system kills the overnight inter-bank market, and the damage can easily extend beyond the overnight tenor. If banks expect to have sufficient liquidity also in the future, it's not obvious why they should trade with each other for funding purposes beyond the overnight segment either. The damage is long-lasting. If you don't use a muscle, it atrophies. Desks are dismantled, institutional memory withers. In the meantime, not just the total amount of reserves, but their distribution begins to matter." We lose a generation of people who work these interbank market desks and we lose that discipline, and I think that's an important point.

Selgin: Yes, it is. See, what that old system did was it allowed a scarce reserve system to… it allowed a small quantity of reserves, even with reserve requirements, mind you, enforced in ways that changed over time. But nevertheless, it allowed the reserves to suffice for all banks. It was very, very efficient. Now, what we're discovering is that if you don't have that market, by gosh, because of distributional issues, it takes a heck of a lot of reserves to make up for an efficient overnight, interbank, unsecured lending market, a whole lot. Occasionally, as we've seen, we've had problems where even modest attempts to unwind the Fed's balance sheet have led to surprising incidences, unexpected incidences of reserve shortages, of reserve stringencies, of spiking overnight rates.

Selgin: Now, why should we care? Why not just leave all those balances there? The problem is that you create these balances during economic crises when interest rates are down at their lower bounds, at zero or close. The central banks are loading up with long-term securities, mortgage securities or just regular Treasuries. Of course, these securities are being bought when yields are very, very low. If they can't unwind, then recovery confronts the central banks with revenue problems, or rather the value of the assets declines as interest rates rise. But of course, being central banks, they don't have to book the losses, but they do have to quit their remittances.

Selgin: In the old days, this business of making up for losses by booking a so-called deferred asset, which is a way of saying, "Well, we'll make it back later." That made sense back when because there was no reason to think that there would be a systematic tendency for central banks to be paying out more interest than they were earning. But now, we're in a situation that's asymmetrical because QE happens when interest rates are really low, and that's when the assets get purchased, the long term assets. Now, we're still a long way from a situation where the Fed is going to have to go hat in hand to the Treasury, but I think we have an institutional setup between which, and perhaps the possibility of a long run decline in currency demand could undermine the Fed's independence one day. Already, the Fed is under pressure because of the fact that it's not remitting funds to the Treasury. I don't think this is a very good situation from a political economy perspective. I think that now is the time to think about whether something should be done about it, because the more time that goes by, the harder it becomes to fix or change.

Beckworth: Along those lines, George, Claudio Borio in his speech notes that the Central Bank of Norway and the Swiss National Bank, because of the concerns you've just outlined, have taken a step back and adopted a tiered floor system, which is a halfway step between a floor system and a corridor or abundant and scarce reserve system. Is that the future? Do you think more central banks will take a step toward a tiered reserve system?

Selgin: Well, I don't know if others will, but I think that that is the way to go. If you've got a huge balance sheet today and you're a central bank and you really want to get out of this mess. Unwinding is costly because the interbank market has been euthanized, and as a result, attempts to unwind substantially can lead to trouble, even though we have now other arrangements to try to limit that. But the safest way to ease out of a floor system is with a tiered system, where essentially what you have is, at the margin you're back on a corridor system. Above a certain threshold, the interest rate on reserves is below the policy target rate as it is in a corridor system, and it's only the grandfathered reserve balances that pay the higher rate. Japan also has a kind of tiered system. There are three or four of them out there now, not all equally worth emulating to be sure, but that would be the road to reform. Unfortunately, I don't see anybody at the Federal Reserve contemplating any move away from the floor system, which in the space of a decade has become sacrosanct for no good reason because it has not proven to have any substantial virtues and it's had some considerable disadvantages.

Beckworth: Well, with enough losses on the Fed's balance sheet, and maybe with growing recognition that the Fed is paying out really large interest payments to banks through their reserves holdings, this may become more of an issue in the future.

Selgin: Yeah, I think it will. I think will. People have to remember that the Fed made a lot of interest in the short run while the crisis was going on. It's not, strictly speaking, just a question of taking losses, but I think the long run picture is one where the losses are going to become more important relative to the [inaudible].

Beckworth: Yeah, I was thinking more that it's the political optics perspective, even though, yeah, you could sum up total earnings and the Fed's still ahead, just the way it looks currently. But let me segue then into my second piece, and this is an article that actually Peter Stella wrote. He came on the podcast not too long ago and we discussed it and his article's title was, *Should The Fed Have Lost $1 Trillion?* This speaks to the unrealized mark to market losses. We've been speaking to the actual losses, actual net income losses on the Fed's balance sheet. We're going to switch a little bit over to the unrealized losses. As you know, George, as rates went up unexpectedly high, the value of Treasuries and other debt securities have gone down.

Addressing Unrealized Fed Balance Sheet Losses

Beckworth: And so the Fed has taken an unrealized mark to market loss on its balance sheet, about $1 trillion as of the end of, I think, 2022. And about 800 billion of that is in Treasuries, 400 billion in mortgage-backed securities. Now, the reason I bring up Peter Stella and his article… and I find it so fascinating, he makes the argument that the 800 billion in this unrealized loss is really a wash from the consolidated government budget perspective, and it's really the mortgage backed security loss that should concern us. His argument for the first point is that any Treasury loss in the Fed balance sheet is a gain to the Treasury's balance sheet. To put it differently, from the Fed's perspective, any expected present discount of value of payments the Fed was going to get from the Treasury has gone away and has fallen by 800 billion.

Beckworth: But on the flip side, Treasury now has a smaller liability to pay out of also 800 billion. It's a wash, at least for this unrealized lost. Now, again, political appearances may make this look different, but nonetheless, from an economic perspective, it shouldn't matter. But what Peter notes is that on the mortgage backed security losses, that 400 billion, it's not a wash. The person who benefits… the gain, the 400 billion gain are household owners like myself who refinanced in 2021, really low interest rates, and now rates are much higher, inflation's higher. The real debt burden I'm paying out on my debt has actually gone down below what I was expecting. I have gained, other householders have gained, and the Fed has taken the loss, and ultimately the taxpayer has taken the loss. What do you think about that distinction? It's really these privately held securities that creates the true cost to the taxpayer.

Selgin: I think in the strict sense of cost, that's correct, but I think it's a mistake to completely dismiss the other Treasury security related deferred losses, unrealized losses, deferred assets, same thing. The reason is that our central bank system, ultimately, its political economy is such that it depends on not having to have a budget. Part of that is its independence is partly based on the Fed never having to have a budget that it has to go to Congress to renew. This keeps Congress at arm's length, so to speak. If you think that's a good idea, then you do have to worry about the Fed's own income statement independently of the Treasuries. That is, the consolidated balance sheet view is not appropriate for thinking about Federal Reserve independence, if you think that matters. Now, the Fed is, right now, able to pay its bills, but it's not remitting because it has these deferred assets. In principle though, what can matter over time, what can happen, in the US case it will take some time to happen, but it could. In other countries, it's a more immediate danger of having a floor system. You can have a situation where the Fed can only keep its inflation target, that is, and pay its bills so to speak, cover its operating expenses, et cetera, if it gets a budget. You have to think about that as well.

Selgin: This is a very far off scenario, as I said, in the Fed's case. But if you want Fed independence, you want the Fed to have enough revenue to cover its operating expenses so it doesn't have to print money for that purpose. But beyond that, it's probably a good idea for the Fed to keep the Treasury happy by remitting funds to it on a consistent basis. Again, that's not so crucial here. If you take a smaller country with a system like this or one without a very large demand for physical currency, for the central bank's physical currency, you could much more easily see how a floor system could be incompatible with budgetary independence of the central bank.

Beckworth: Yup. Okay. Let's go to our last round of articles. George, I know where you're going to go to this, so I'm going to kick it off with some recent developments in inflation. Headline CPI last June, June 2022, hit 9%. We just found out this past June it fell to 3%. That's huge; six percentage points down. Core CPI hit a peak of 5.9 last year, and now it's down to 4.8%. Some big declines with no or little change in unemployment. It looks like we're having a soft landing, George. What do you want to say about that?

The Fed’s Soft Landing and Phillips Curve Considerations

Selgin: Well, several things. First of all, and I think we agree on this, David, I've never been one to look at any of these inflation numbers to determine… to rely on any of them, core CPI, PCE, you name it, to decide whether the Fed is doing what it needs to do with regard to inflation. But that's because I think what it needs to do is to get a lid on nominal spending. The reason why those aren't the same thing is because, of course, adverse supply conditions also change measured inflation rates, where the change depends on what measurement you're talking about. But we've had a lot of supply side activity that's been a big part of the inflation story.

Selgin: We have highly volatile prices of certain commodities and things like housing and oil that can give a false impression. For me, the big news is if NGDP growth has not been as excessive as it had been in the past, the part of the inflation story that really is the Fed's responsibility or the Fed's fault, whichever you like, that has been improving, but not as much. I still think there's a little bit of work for the Fed to do there, but things are improving. But I would only say so on the basis of NGDP numbers, not any of these headline or other measures of inflation. That's the first thing. The second is, yes, I think we do have evidence that we're having a soft landing, that this thing is possible and that we might just achieve it.

Selgin: I've always argued, and the reason I picked this topic, as I said I wanted to talk about it, was because I knew I could toot my horn a little bit, not that I haven't been doing that this whole time. But I've always thought the Phillips Curve was a badly abused concept. I'm old enough to remember how it was abused in the '70s to argue that a little more inflation is worth it because it's going to keep unemployment low, and we know how that turned out. But I'm also, on record, arguing long ago, I think it was in the late '80s or something, I had a Wall Street Journal article about a market spooked by Keynes, which made the point that the Phillips Curve can also be abused the other way; to claim that you can't have more output or more employment without having more inflation.

Selgin: Now, we're seeing another variation, we've been seeing another variation of the Phillips Curve fallacy. That's the view that you can't bring inflation down without causing more unemployment. Now, all of these are results of thinking of the Phillips Curve as what economists call a reduced or a structural relationship, where the two things are related. It's cause and effect, unemployment on one hand, inflation on the other, instead of understanding that it's a reduced form that's really telling us about a more complicated set of structural relationships, where the bottom line is that you have changes in aggregate spending and nominal GDP, total spending, that when they occur, can manifest themselves either in higher prices and wages or in more output and employment.

Selgin: It's important to understand, for example, that when you slow down the growth of NGDP, it doesn't have to reduce employment. It can instead work mainly by reducing wages or preventing them from increasing as much. Anyway, the other factor that made me believe long ago that a soft landing was possible, and to argue as much, was the high vacancy rate. If you reduce spending slowly, if people expect it to be reduced, and they believe your policy, if it's a credible disinflation policy, and you have a lot of slack, a lot of job vacancies, then a soft lending doesn't look so hard. But if you just are looking at a Phillips curve and thinking, "Ah, that's it, this goes up that has to go down," or whatever, then, you're going to get it wrong. That's what people have been doing.

Beckworth: So let me respond to that, George, because most commentators do invoke the Phillips curve thinking or framing, and I agree with you. I think it's best viewed as a reduced form of relationship that's reflecting something more fundamental, in this case, aggregate demand; what's happened to it and what effect does aggregate demand have on the economy. Is it increasing real activity or is it increasing inflation or some combination of both? And people, the way they think about the Phillips Curve is, there's this relationship, as you noted, and a lot of people today are invoking a non-linear Phillips Curve.

Beckworth: If I can map that argument back into your argument, I think what they're saying is that all of the spending, or a lot of the spending that was added to the economy over the past few years was translated into higher prices or higher inflation. In other words, we were on a very steep or inelastic part of a short run aggregate supply curve. As a consequence, coming down from that won't have much effect on the real economy, because most of the increase led to higher inflation, coming down will also just affect inflation more than the real economy. It's a pretty straightforward, I think, interpretation or mapping from a non-linear Phillips curve into the world that you're thinking about. Is that right?

Selgin: Yes, except I don't like the framing, and to put it tersely, I think non-linear Phillips curves are the macroeconomic equivalent of… or to macroeconomics what epicycles were to the Ptolemaic system of astronomy, that if you manipulate or if you... You can always come up with a fancy enough Phillips curve to fit the facts, but why bother? Why use the darn thing if it leads to so much trouble? I think it is a nice intellectual exercise perhaps to say that we can conceive of Phillips curves doing all kinds of things. You can make them do back flips if you want to, and have them thereby become consistent with the data record. But the question really is, is this concept helpful? Is it really the best tool to use to understand what's going on in the economy? My answer is no. Again, you can do things with it that fit it to the past data record. The minute you try to use it to predict what's going on in the future, you're asking for trouble.

Beckworth: But don't we, as people who like to look through maybe an aggregate demand or nominal GDP perspective, aren't we also invoking something similar to the Phillips curve? We invoke a short run aggregate supply curve that gets really steep going up. How different is that?

Selgin: I think it's very different. I think that the Phillips curve takes too many shortcuts or tries to do too many things at once. That's what a reduced form relationship basically is. I think that thinking in terms of aggregate supply and aggregate demand shifting around is closer to grasping the true structural relationships. Even though it's not particularly complicated, it isn't much more complicated, but it is somewhat more complicated. You have two schedules instead of one, for starters, and I think that it matters.

Beckworth: Okay, let me make one more argument for Phillips's curve folks out there. You can swat this one down again. One argument might be that it's a pragmatic tool. In other words, it provides data that you can see on a monthly basis. For example, the unemployment rate, or some people, as you alluded to, are using the vacancy rates. Governor Chris Waller, Andrew Figura, if I'm saying his name correctly, they looked at Beveridge curves, and there's a San Francisco Fed study that came out recently where they did a Phillips Curve where for slack, they actually look at the unemployment rate over the vacancy rate. But basically, a lot of people are invoking the vacancy rate as a measure. One nice thing about those two measures, unemployment and vacancy, is that they come out on a monthly basis. It gives you maybe a closer real time assessment of where aggregate demand is versus waiting for quarterly GDP numbers to come out.

Selgin: Well, look, I don't want to disparage any of this research. I think it's all interesting and it can reveal a lot. But as far as being pragmatic is concerned, I think the proof is in the pudding. We had a lot of people relying on Philips curve analysis making dire predictions about how disinflation would be associated with big increases in unemployment and possibly another big crisis cycle, and they were wrong. Maybe there were some Phillips curve people who got it right. I'm sure there were, but they had to be thinking beyond the darn little thing. For me, it's just not very useful. I'm pretty sympathetic to a lot of new gains in economics, actually, but I've never been very fond of the Phillips curve, simply because I think it oversimplifies. Like I say, you can have a very complicated kind of Phillips curve analysis, but at that point, I'm not sure it's really a helpful heuristic for that because it has to be too complicated.

Beckworth: Okay, so my last article and tie into all of this is a piece by Richard Clarida and Peder Beck-Friis. This is a piece coming out from PIMCO, and the title is, *Fiscal Arithmetic and the Global Inflation Outlook.* What they show, they do a little scatter plot, and they show that if you sum up and add up all the debt growth by country and then plot it against the cumulative growth in core inflation, there's a pretty strong relationship. They make the argument that the inflation that we did see was largely driven by expansionary fiscal policy. In other words, they're attributing this more to fiscal policy than monetary policy, although they would acknowledge there is some role there for monetary policy. In other words, they're looking at the helicopter drops, the checks, the stuff that were sent out. I guess I wonder, George, where would you come down on that? How important was fiscal policy, doing helicopter drops, versus the Fed keeping rates low, also buying assets for quite a while? How would you break that down? Do you think there is an important role for fiscal policy here?

Is There a Role For Fiscal Policy?

Selgin: Oh, yeah, absolutely. I think fiscal policy is very important. Unfortunately, we have an arrangement where fiscal policy can be creating considerable inflationary pressures adding to aggregate demand. Our setup nevertheless makes it the Fed's responsibility, not the Treasury's, to do something about it. In practice, what that means is that the more the Treasury is boosting demand, the more, other things equal, the more the Fed has to see to it that it crowds out private investment, because that's the lever the Fed has, by raising interest rates. It's not always easy, obviously, for the Fed to generate enough crowding out, if you can… don't mind me putting it that way, to offset the Treasury's expansionary behavior.

Selgin: But as I said, our system puts all of the onus on the Fed. There's nobody at the Treasury's whose job it is to keep inflation from getting too high. That's where we're at. Now, to the extent of thinking that fiscal policy matters, I'm here entirely accord with some fiscal theorists of the price level, like David Andolfatto and also others who insist on the overwhelming importance of fiscal policy. Where I differ from them is my belief that the Fed does have the capacity up to a point to make up for the fiscal authority's extravagance at the cost of squeezing borrowers. But of course, it can overdo it, and that's when you really don't have a soft landing. In this case, I think the Fed has done a pretty good job. I think they should have started a lot earlier, and I don't think they're finished, and I said that too, by the way. I started moaning and groaning that the Fed needed to start raising interest rates in the summer of 2021 by its own standards, and that even by NGDP standards, it needed to do so by the fall of that year.

Beckworth: Yeah, so it's been encouraging to see core inflation fall. We've had this great report this past month, hopefully July will be similar. Hopefully, we'll see more soft landing, and in the data that will mean a continuing decline in vacancies without an increase in unemployment. But the real question is, can this soft landing experience we're having now continue all the way down to 2% inflation? It's one thing to get where we've been. Let's see if we can make it the rest of the way home.

Selgin: Yeah, I would say we don't care if it comes down to 2% inflation, I don't care. I want to see NGDP growth come down to about 4.5%, and that's being generous because, I'm not calling for makeup by that, but I think that that's what I'd be looking at, is the NGDP growth rate getting back down to a long term trend. Of course, David, there's going to be a recession sooner or later, and even if it happened a year from now, all the people saying, "See, we told you there wouldn't be a soft landing," are going to throw their own little parties. But let's face it, if we didn't have a recession after a while, that would be quite unusual, wouldn't it?

Beckworth: Yup. Well, we will hope for the best and plan for the worst.

Selgin: Right. Yeah, I think that there will be a hard landing one day, again.

Beckworth: And with that, our time is up. Our guest today has been George Selgin. George, thank you so much for coming on the show once again.

Selgin: You're welcome, David. I always enjoy it. I'll be seeing you in Washington not too long from now.