George Selgin on the Past, Present, and Future of a Real-Time Payments System

George Selgin is the director of the Cato Institute’s Center for Monetary and Financial Alternatives and is a returning guest to the Macro Musings podcast. Today, George joins Macro Musings to talk about recent developments in the payment system. Specifically, George and David discuss the history of attempted payment system solutions, the challenges and costs facing the implementation of a real-time payment system, and why we should care about this issue today.   

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: Our guest today is George Selgin. George is the Director of the Cato Institute's [Center for] Monetary and Financial Alternatives. And as a previous guest of the podcast, George joins us to discuss recent developments in the payments system [and] also what is happening with the Fed's review this year, and finally, the latest developments in the Fed's operating framework. George, welcome back to the show.

George Selgin: Always nice to be here, David.

Beckworth: Well, it's good to have you back on. You are the reigning champ on Macro Musings in terms of show appearances. I believe this is your fifth appearance.

Selgin: Hey, whoopee!

Beckworth: So, you are like the Mike Munger of econ talk, that's what you are for Macro Musings. So, he's, I think, like a regular, reoccurring, almost a co-host. So, you're like the number one appearance on this show. If you haven't heard any of George's previous shows, go back and check them out. But, today, I have brought George on to talk about the payment system. We will probably spend most of the time talking about the payment system, and then we'll get into the review, and finally, the operating framework as it relates to some of the developments in the money markets and in repo. But, there's been a lot going on with the payment system, George, and you've been a big part of that conversation recently, you were at a Senate hearing on facilitating faster payments.

Beckworth: Lots of real-time payment discussions, we've had Aaron Klein on the show, we've had a few others. But, I want to really dig into the weeds today and learn what's going on, and why does it matter? Because, for many people, including myself a few months ago, this is an esoteric technical subject that, it didn't get much attention. It's probably not on most people's radar, but it's very consequential. So, talk us through, why should we care about the payment system? And then, after that, maybe tell us what's been happening.

What’s Going On With the Current Payment System and How Can We Fix It?

Selgin: So, David, for a lot of people, the payment system seems just fine. So, you get paid, and you get your funds reasonably quickly for your needs and et cetera. But, there are many people who actually have reason to be very unhappy with the way the payment system works. Merchants sometimes have to wait a while, maybe for an extended period of several days for their funds to actually come through, and payments are recorded, but the actual final payment doesn't clear for a while, and it's only then that they have funds credited to their accounts that they can use. But the people who suffer the most from the slowness of payments in this country, and it is a problem, particularly in the United States, are poor people who live from paycheck to paycheck, and this is something that Aaron Klein has been particularly eloquent about.

But the people who suffer the most from the slowness of payments in this country, and it is a problem, particularly in the United States, are poor people who live from paycheck to paycheck.

Selgin: And the fact that it can take several days for a paycheck to clear for them can be very costly. Costly because they have to resort to payday lenders, or check cashing services, if you like. Costly because they incur overdrafts in their accounts, because the money isn't there yet. And that is really the most serious, but not the only adverse consequence of the fact that we have a relatively slow payment system in this country. So, we don't all notice it, we don't all care. But, plenty of people do, and there really is no good reason for it. And especially when you consider that there are many, many countries that have payment systems that are very quick or practically all payments are completed and credited within minutes, if not instantly. So, we have some serious catching up to do.

Beckworth: So, we're behind other countries when it comes to real-time payments.

Selgin: Many countries, yes.

Beckworth: Okay. Which is surprising and you'd think we'd be at the cutting edge, but, like U.K. and Canada and Europe, they have much faster payment systems there.

Selgin: Indeed, yeah. I'd say that many of these countries are a decade ahead, conservatively. So, partly, it's the fact that our legacy payment system, in some respects, was pretty good, and so, there wasn't the urgent need that was felt, at least, widely enough to drive reforms. Some countries, in a sense, benefited from the fact that the payment systems were so slow that they knew they had to overhaul them, and they did. And now, they're ahead of us. But, we also have a challenge that other countries don't have in having so many banks and credit unions, not mention other payment services. And therefore, coordinating reforms that require that all of these firms be operating on common networks, is a lot more difficult in the United States than it is elsewhere.

Selgin: So, it's not an easy task that can be done to make our payment system as fast as the payment systems of other countries that have more concentrated banking systems, and are smaller, to begin with.

Beckworth: Yeah, we recently had your colleague, Diego, on the show, and he talked about one defining characteristic of the U.S. banking system is it's fractured. There's just so many different banks in many places, and you've written about the history of this yourself, but that's a big part of the story here, right?

Selgin: Yes, it is. So, we're talking about roughly five to 6,000 banks and almost equal number of credit unions. And then there are other kinds of depository institutions and other kinds of payment service providers. And many of them are very, very small. So, coordinating them, getting them all, just, you could establish a network that could handle fast time payments, you're only a third of the way down or something like that, because then you've got to get everybody to join. And as you'll see, when we get into this a little bit further, that can be a real problem, but it's important because networks can be much more effective, are much more effective, if everybody's in them.

Selgin: So, if I want to send a payment to somebody, if that person's bank is in the same network, that can make it a lot easier, and may not be possible at all for my network to send him a payment quickly, if his bank isn't in the same network.

Beckworth: Okay. So, that's the backdrop, the story to why we care about this and why it's an issue in the United States. We have an antiquated payment system, it takes time, it's costly for people in the lower half of the income spectrum, and for merchants as well, for businesses. So, there is a need for change. And there's been several developments, there's been two big stories, I think. One is the private sector's attempt to solve this with pressure from the Federal Reserve. And now, the Federal Reserve is attempting to address it itself. So, walk us through the history of attempted solutions. Selgin:  So, really, things got started back in 2014, and the Fed quite rightly took the initiative by convening a task force. A task force that had many participants from the private sector, particularly, and interest groups. And the Fed challenged or charged this task force with coming up with ways to improve payments. Particularly, the challenge was to come up with the so called solution to the problem of providing so called real-time gross settlement services for retail. Now, that's a real mouthful. So, the essence of real-time payment is that, the same message, the same, if you like, computer clicks that deliver the payment instructions, also deliver the funds, it's all done instantaneously, so there is no delay between a payment instruction being sent, and actual funds being transferred to the payee's bank, and thus, to the payee. It all happens instantaneously.

Selgin: Now, that's quite different from conventional payments, retail payments, which, depending on what form, the specific form they take, can take quite a long time, because, the payment order is one thing, but the actual transmission of funds from one bank to another is another. There is a process of clearing, which is, reconciling the different payment messages, and finally, a process of settlement, which is, when good funds, so called, think of them as bank reserves, get moved from the payer's bank to the payee's bank. And it's the clearing and settlement process, and particularly, the settlement itself, that's what can take a long time, and that's where those unfortunate delays come in.

Selgin: And so, with real-time payments, that is a solution. I'll try to emphasize that it is not the only solution, but it is certainly a solution. It is the fastest solution because it does away with the gap between a payment order being made and instruction being given, and actual funds getting where they need to get.

Beckworth: Okay. And so, the Federal Reserve encouraged the private sector to come up with a solution, and so, what was the private sector's solution?

Selgin: So, right, they convened this task force in 2014, and they got a number of proposals out of this task force, and they're all very interesting. But, only one of them really scored very high points. The Fed and others, they graded, they actually graded, they gave them like a report card on the different proposals. And one in particular, which was a proposal from The Clearing House, TCH, a private payments organization that's been around since 1853, based in New York, their proposal got very high marks all around, it was an A+. And this had to do with security, reliability, potential for becoming ubiquitous, which is to say, a universal, everybody could, in principle, join, equity. There were a number of criteria, got very high marks.

Selgin: Well, TCH, in fact, went ahead and set it up. They spent something like a billion dollars, and by 2017, they had the system up and running.

Beckworth:  What was the coverage like by that point?

Selgin: Well, of course, when it first was established, it may have had some preliminary subscribers, but, eventually, it had membership that covered about 50 percent of all bank deposits. That number's probably gone up some now, but it consisted mostly of the large banks, and which, of course, it doesn't take that many large banks before you're at 50 percent. They did not get that much membership from smaller banks, and they were struggling to do so. But, I suspect and they believe that over time, they would get more and more members, because this was a desirable service that smaller banks would want to be able to offer and not just leave to the large ones.

Beckworth: The bigger network grows, the more it is in your advantage to join the network. So, they're on this trajectory of this growing network of real-time payments, and then, a big shock on the road appeared before them, and what was that?

Selgin: Well, the shock was, let me step back a bit to make clear that, up to that point, to 2017, and for some time later, the clear, a fairly an unambiguous signal the Fed was sending was, "We're going to see if the private sector can do this. If it can, that'll be that."

Beckworth: So, the Fed was cheerleading this process.

Selgin: They were very much so, and they gave every indication that if the private sector solved the problem, they had not suggested that they might come in and offer this service themselves. And this is all, by the way, quite consistent with the spirit, if not the letter of the 1980 Monetary Control Act, and with the Fed’s own rules, which hold that it should not offer any new payment services, that, a service that the private sector is capable of providing. So, the idea is, if the private sector can do it, the Fed won't do it. The Fed will only do it if the private sector can't do it.

Selgin: So, TCH, with this understanding, had passed with flying colors, the competition to come up with a solution for real-time payments, set up its system, which is called RTP, for real-time payments, and had it running in 2017. And as I say, by that time, or, by sometime in 2018, had about 50 percent of the market, when the Fed, I think this was in August 2018, suddenly sent out a request for comment on the possibility that it would set up its own real-time gross settlement retail system, and also requested comment, and this is very significant, on whether it should provide 24/7/365 days a year settlement services, or some kind of settlement service. And that other part of that request for comment is very important, and I want to get to it.

Selgin: But, the first part, where it said, "Oh, we want to know whether we should enter in the real-time space," that was the bump in the road or the shock that TCH suddenly found itself confronted with, because now they're saying, "Good gracious, we've just spent a billion dollars, we're halfway, we've got half of the bank deposits signed up, and now, the Fed has really thrown a wrench in the works. Because, first of all, the possibility that they will set up a rival service, means that banks may wait and see about that service, because they won't want to join both. It's expensive for a bank to join any of these services. There are hardware costs, there are other costs that have to be expended by the bank, in order to link up, so to speak.

Selgin: And these two services, and this is very important, there's no indication, there was never any indication, and there still isn't from the Fed, that its rival service, which has since been given the name FedNow, will be interoperable with the RTP system. Which means, basically, that, if you want to send a payment, and your bank is on RTP, to someone whose bank is on the Fed system, when that gets set up, that may not be a fast payment because they're not connected, they're not interoperable. So, that meant that, of course, every bank that hadn't signed up for RTP, now has to worry about what's the right thing to sign up for. So, it had a chilling effect on membership to RTP, that the Fed might join, might compete with it.

And these two services, and this is very important, there's no indication, there was never any indication, and there still isn't from the Fed, that its rival service, which has since been given the name FedNow, will be interoperable with the RTP system. 

Selgin: And, of course, it faced RTP with the prospect of, The Clearing House, with the prospect of having to keep compete directly with the Fed, which is never something you really want to have to do. Because, and this is extremely important, we are economists, and many of your listeners will also be economists, so, of course, we like competition. But, competition with the Fed is not plain old competition, the Fed has lots of ways to cheat, to put it bluntly. First of all, they're regulator, as well as a competitor. So, they regulate firms they compete with, and that is a conflict of interest that is extremely problematic, extremely problematic.

Selgin: Second, although the Monetary Control Act requires the Fed to recover its costs for any service it provides, essentially, so that it can't cross-subsidize, it can't use its monopoly rents from issuing currency over which it has a clear monopoly. It can't use those, for example, to underprice other services. At least, it can't, according to the law. But, enforcing the Monetary Control Act is not easy. It's not always clear whether there are cross-subsidies involved and we rely on the Fed's own accounting procedures to determine whether it's recovering its costs, and those procedures haven't been subject to an outside audit since 1984. So, it's very loosey-goosey, to use a phrase that my good friend Bill Trumbo likes to use. And so, you have a lot of issues.

Selgin: You have a third issue with respect to the Monetary Control Act in this case, in that, the Fed is merely speculating that it's going to satisfy the MCA requirements, right? It doesn't know, but if it... I should say that I've jumped ahead a bit. Of course, ultimately, the Fed did decide to go ahead with FedNow. But it, in doing so, implicitly asserted that it would be able to meet the Monetary Control Act's provisions, even though it doesn't really know how much money it's going to make, it doesn't even know what it's going to charge. Because, unlike RTP, it hasn't resolved, made up its mind about a fee schedule, and that's another important issue. Anyway, this was all very disconcerting to the folks at The Clearing House, understandably.

Selgin: And they joined with some others in responding to the comment, request for comment, naturally, quite critically. And I also wrote critically on this, saying that I did not think it was going to be helpful for the Fed to compete in this space, and I did not think it was going to be necessary. So, there were a few people who wrote against the FedNow, what became the FedNow plan. On the other hand, they were also thousands of letters, mostly rather perfunctory, from community bankers, mainly who'd been urged to chime in in favor of FedNow, which they did. So, ultimately, what the Fed did, was to say, "Look, the overwhelming number of comments have been favorable for FedNow, therefore, we're going to do it."

Selgin:  And that was a correct statement, if you just went by the sheer volume of comments that were favorable, even if they weren't very substantive, and many were not. But, there were many comments, fewer, of course, but there were many very substantive comments that criticized the proposal and said the Fed shouldn't go ahead with it. So, anyway, they went ahead, and now, you have two potential rival systems, one of which is already in existence, and the other which will be coming, according to the Fed, in four or five years, which is a very long time. We can talk about that too.

Beckworth: Right. So, let me summarize what you've said. I want to use the analogy here of a sports game. Let's say is football, all right? So, the Fed got this team together, called the private sector, and they're saying, "Hey, go play that other team out there. The other team it's the lack of a real-time payment system. We want you to play and beat them, and we're going to cheer from the sidelines. We're going to cheer loud, hard, we're going to encourage you." So, the private sector gets out there, The Clearing House takes the lead. It's playing and it's making progress, it's getting out to near the end zone. And all of a sudden, they hear this loud whistle, and they look in the stands, and suddenly, the cheerleaders say “Get off the field, we're going to play."

Beckworth: And they're like, "Well, we've almost won." And they're like, "No, we want to play, and moreover, we're going to rewrite the rules. If you don't like it, we're going to rewrite the rules, we're going to get in, and we're going to take over." Which seems kind of strange. And I have to ask this question, were some of those cheerleaders in the past, now supporting the Fed? I mean, specific people, like Dan Tarullo, I imagine he must have been one of the people who supported the TCH's plan, did he change his tune? I mean, how did this conversion take place, I guess? Going from cheerleader to proponent of the Fed’s own system, there's got to be a story there.

Selgin: I'm sure there is, David. By the way, I really like your analogy, because the Fed has, indeed, announced it's coming on the field itself, and it's going to be the umpire, or the referee. And I also like it because it stresses the fact that, although people have said, "Oh, we can't let The Clearing House run a monopoly," the payment space is very crowded with different providers, some of which do provide instant payments, but for small networks. And TCH is competing with all these other established payment services, including the legacy services that are not real-time. So, it's not as if there's no competition, and I'll talk about how the Fed could have confronted TCH with much greater competitive pressure, without doing what it did, when we get around to talking about 24/7 settlement services.

Selgin: Anyway, I really like the analogy, I think it's correct. And I completely forgot what you were asking about.

Beckworth: What I was going to ask, so, the interesting part of this-

Selgin: Oh, whether the change, how did the... Yeah.

Beckworth: How did you go from being a cheerleader, to just saying, "I want to take the field"? I mean, that's a pretty big move, right?

Selgin: It is, indeed. And as I said, it caught a lot of people by surprise. I wish I could tell you what the inside politics of all this consisted of. The only prominent Fed official I know who does not, who has argued that the Fed's entering directly into this space is not a good idea, is Randy Quarles. I don't know about Tarullo.

Beckworth: Well, he's no longer at the Board of Governors, but I'm just wondering like how his views of evolved…

Selgin: I don't know what his views are.

Beckworth: Or Lael Brainard, I mean, I know she was pretty strong advocate, I wonder where she stood back in 2014 in all of this.

Selgin: Yes. I don't know, I don't know. I do know that the major source of this decision for the Fed to create its own system, a lot of that is coming from the Federal Reserve Bank of Kansas. Which, the economists at that bank have made a specialty of real-time payments. And clearly, for quite some time, I think really predating the task force of 2014, have been, as it were, looking forward to the Fed moving in and establishing its own real-time payment system. So, I suspect that a lot of the pressure came from the Kansas City Fed.

Beckworth: All right, let me play my analogy out a little bit more then. So, you have these cheerleaders, the Fed in the stand of the game for starts, and there's a bunch of fans out there, all the different Fed officials, but there's a small group of those Fed officials sitting quietly with their arms folded, that's the Kansas City Federal Reserve Bank. Everyone else is cheering, "Rah, rah, rah, RTP," and they're just grumpy, they're upset, and they slowly work the crowd over and say, "Hey, it's time for us to take the field." Maybe that's part of the story then, huh?

Selgin: I suppose. Like all analogies, this one's starting to get a little bit frayed at the edges.

Beckworth: Okay, fair enough.

Selgin: But I do think that you had a sort of technocratic impetus, where, you have people who have been thinking about these things for a while, and they really like get their hands dirty, running a system, they think they can do it. And you also, let's face it, have the usual bureaucratic motives. For a long time, the different Federal Reserve Banks, their main function, we think of them as most people think of the Federal Reserve Banks, other than, apart, I'm not talking about the board, I'm talking about the 12 banks, think of them as places where you have some expert economists, and then you have a president who takes part in the FOMC, perhaps, and helps make decisions about monetary policy. Or, you also have bank supervisors, et cetera.

But I do think that you had a sort of technocratic impetus, where, you have people who have been thinking about these things for a while, and they really like get their hands dirty, running a system, they think they can do it. And you also, let's face it, have the usual bureaucratic motives. 

Beckworth: Well, let me give them a hearing here.

Selgin: Sure.

Beckworth: I'll take their side, a little pushback on that point. So, Thomas Hoenig now, who's a colleague of mine here at the Mercatus, and I had him on the show, and he mentioned that he had to actually lay off a bunch of people because of this whole technology change, in terms, we don't carry cash, checks, around much anymore. And so, there were people who were let go from the Fed, but you're saying, the incentive is always there to find a new reason for existence, to be relevant.

Selgin: Yes. If they could have found something quick enough, they wouldn't have had to lay off those people, and Tom wouldn't have had that unpleasant experience. And so, they're trying to be forward-looking, and, of course, I didn't mean to imply that they don't actually want to see their budgets grow, which is also part of the standard bureaucratic model. So, there are a lot of bureaucratic factors that I think play in here. I don't think ultimately it matters what the Fed's motives are, what matters is whether their intruding the spaces is necessary and whether it's the best thing they can do to promote faster payments. And that, I'm sure, the answer to the second question is no. And that's what I want to talk about, if you don't mind.

I don't think ultimately it matters what the Fed's motives are, what matters is whether their intruding the spaces is necessary and whether it's the best thing they can do to promote faster payments. And that, I'm sure, the answer to the second question is no.

Beckworth: Yes, please do.

Selgin: I want to talk about the other part of the Fed's request for comment back in 2018, and what resulted from that.

Beckworth: So, just to summarize, the first part was a real-time payment system.

Selgin: Yes.

Beckworth: And now, let's jump into the second one. Tell us about that.

Selgin: Yeah. So, the second thing the Fed asked about, was whether it should set up, or offer 24/7 settlement services, and to assist the liquidity management in private payments networks. Or, in the established legacy networks. Now, let me give you an example of how this reform could help RTP, for example. But, first, let me step back and say, at present, and for some time, the Fed has two wholesale settlement services it offers. They're the, Fedwire is one, and the other one is the National Settlement Service. Now, I don't want to go into any real deep details about those, but these are services that essentially see to the final settlement I was talking about earlier.

Selgin: That is, they see to it that, after payment instructions are sent from one bank to another, that the funds get transferred from the sending bank to the receiving bank, and so, the receiving bank can give credit to whoever the payee is. And so, that's done either using Fedwire or using the National Settlement Service or both. Now, here's the thing, both of those services have limited operating hours. They're not open 24 hours a day, and that already restricts the number of payments. It restricts the potential for payments to be settled within the same day. There's just something called the ACH system, Automated Clearing House, and depending on when payments are sent through it, there are two payment windows during the day, and if they're sent in time for those windows, they're settled on the same day.

Beckworth: Why is that? Why don't they have a 24 hour system?

Selgin: That's a good question. But, if payments don't come on time for that second window, they don't get cleared until the next day. And that is a lot of payments are processed through there. And again, the problem here is, Fedwire's hours are limited. It's extended its hours somewhat over time, but I think it closes at nine o'clock now. 9:00 P.M. Eastern Time, if I'm not mistaken. And, Fedwire and the National Settlement Service are closed on weekends and holidays. It's because of these limited Fedwire and National Settlement Service hours, that payments can sometimes take days to settle. It's because of those.

Beckworth: Which is hard if you're a poor person.

Selgin: Your average poor person doesn't mind, isn't harmed that much, I should say, right? They're not harmed that much by payment that takes hours to settle, but it's still settled the same day. They're not even harmed that much if it gets settled the next day, but they’re harmed some. But they really suffer if it's a weekend or with holidays. But that's all because these established existing Fed settlement services don't run 24 hours a day, seven days a week, 365 days a year. Other countries have services, a central bank operated settlement services, that do run all hours.

They're not harmed that much by payment that takes hours to settle, but it's still settled the same day. They're not even harmed that much if it gets settled the next day, but they’re harmed some. But they really suffer if it's a weekend or with holidays. But that's all because these established existing Fed settlement services don't run 24 hours a day, seven days a week, 365 days a year. Other countries have services, a central bank operated settlement services, that do run all hours.

Selgin: Okay. So, the other component of the Fed's 2018 request for comment, was, should we provide this kind of extended wholesale settlement service, either by increasing the operating hours of these existing services, or by providing what they called a new liquidity management tool? And the response to that component to the comment letter was unanimously in favor. Yes, the Fed should do this. And that's not surprising, because, actually, various people in the payments industries have been pushing the Fed to do just that for years, and the Fed has known. It has talked about doing these extended hours for years, but it keeps dragging its feet.

Selgin: Even a third payment window was supposed to have done it some time ago, and it dragged its feet, didn't do what it needed to do, and now, it's delayed for almost two more years. That's just to get a third-

Beckworth:  Why? Why this delay?

Selgin: Well, it's a great question, David. Because, think about it, the Fed has managed, in a very brief period of time, to convince itself that it has the know-how and ability to go forward with their new, completely new technologically sophisticated real-time payment service. It was able to say, "Yes, we can do that." Because that was the result of that part of the comment letter to determine that there was enough support, they can do it. But, apparently, they need more time how to increase the hours. And this is the thing, it's absolutely shocking, if you go to the Fed's 2019 announcement, I think it was January 2019, where they said, "Okay, here's what we've decided to do in response to this feedback." The first part of that announcement is, "We're going to go ahead with our real-time FedNow system."

Selgin: The second part, which is kind of buried in there, but I urge your listeners to go ahead and look this up and read it. You can provide a link.

Beckworth: I'm going to put a link to it on the show, yeah.

Selgin: Well, look for the paragraph where it says what it decided to do about 24/7, and this is absolutely scandalous. They say, "Well, as for that, we're going to keep thinking about it. We need to explore that possibility further." And this is a possibility, remember, that's, they were urged to pursue for years. And then, at one point, way back, they said, "Oh, yeah, we're going to eventually get around to this." And this one, this no brainer of a decision, they say, "We have to explore further." And just to add insult to injury, they say, "We may get around to requesting comment on the possibility."

Beckworth: Comment on a comment.

Selgin: Yes, comment on a comment. This is a response to a comment request that asked for comment on extending the hours of Fedwire and National Settlement Service, and now, they're saying, "Well, in response to the universal comment in favor of doing that, we've decided that we may seek comment about this possibility." It's absurd. Now, this is where things get really nasty. So, remember that, the main, not the only, but the main impetus here, and most important reason for wanting to make change is the poor people who are suffering from, not just from delays of a few hours in settlement, but delays of a day, or two days, or three days, or four days, and depending on holidays.

Selgin: This simple, relatively simple reform would be able to solve that problem, eliminate those delays, on just using existing or legacy payments services, and arrangements, and networks, and could probably do it. I can't imagine why it would need to take more than two years. But, the Fed has decided not to take that step, where the cost-benefit ratio, you would think, the benefit to cost ratio would be pretty darn high. But it did find it easy to decide to undertake a much more ambitious reform, the necessity of which is hardly clear because it would replicate the services of an already established real-time payment network, and which will, according to the Fed's estimate, not be up and running for another five years. And you can bet that adding two years to that is a conservative proper adjustment.

This simple, relatively simple reform would be able to solve that problem, eliminate those delays, on just using existing or legacy payments services, and arrangements, and networks, and could probably do it. I can't imagine why it would need to take more than two years. But, the Fed has decided not to take that step.

Selgin: And in the meantime, what are all these poor people to do? They're going to have to put up with exactly the costs, and delays, and hardships that they've been putting up with for all these years, for another five, six years. It's disgusting. Finally... Sorry, I'm on a rant.

And in the meantime, what are all these poor people to do? They're going to have to put up with exactly the costs, and delays, and hardships that they've been putting up with for all these years, for another five, six years. It's disgusting. 

Beckworth: Please do, this is great.

Selgin: I hope your listeners will forgive me, but this really is terrible. The other point that's very crucial here is, if the Fed had in fact decided to take the steps to make its existing settlement services operate 24/7/365, that alone, of course, would have confronted RTP, The Clearing House system with an important source of competition. Because, it would have meant that ordinary payments would be much faster, there would be no really serious delays. And they wouldn't be as fast as real-time payment, that's true, but the difference would have been small enough that, for RTP's ability to overprice its service, which is what you worry about a monopoly doing, right? Its ability would have been very limited.

Selgin: Because, if the fee goes up just a cent or more in the RTP arrangement, and we need to talk about fees more.

Beckworth: Yeah, absolutely.

Selgin: But, their ability to extract rents would be limited because of the greater contestability of the payments markets, where contestability isn't just a matter of having more than one provider of the exact same service, but of having multiple providers of close substitutes or near substitutes, that's enough, that would have done a lot. So, the Fed could have, as it were, by making the opposite decisions it made in response to the comments that it requested, by deciding, "We're not going to do RTP, but we are going to do the 24/7 thing with our settlement services." It would have provided relief for the poor much sooner, and it would have provided an important source of competition to RTP, and therefore, it would have had the best of everything.

Beckworth: Yeah, the best of both worlds.

Selgin: And it did just the opposite. And I believe it did so because of completely unjustified bureaucratic motives, also encouraged by misguided perceptions in the part of the public that it did nothing to allay about the potential for abuse from RTP. And I do want to talk about that.

Beckworth: Yeah, this is very interesting. I wasn't aware of the second part of the story. I knew the RTP story somewhat, but the second part of the story is interesting. So, that, the second part of the story, so, the legacy or the ordinary payment systems that we have, could have been dramatically improved. I mean, and you said, they talked already about having a third window, but just simple fixes on existing frameworks could have been tweaked, could have been fixed. And because they hadn't been fixed, there was this push to do real-time payment system. So, the real-time payment system is a symptom of a deeper problem that could have been fixed easily, and so, they're going to take a whole different approach, a whole different tack, which is kind of mind blowing, really.

Beckworth: I mean, why take such a difficult course when there was a much shorter direct path to accomplishing the goal of helping getting faster payments?

Selgin: Well, I don't know the answer to that. I do want to say that I don't want to imply that having a retail payment system, and even one that could cover most of the country was an unnecessary goal or unworthy goal, I think it's a fine goal, it's a fine objective, but it isn't the whole story about how to improve payments in this country, and it doesn't make sense to have left the legacy payments arrangements in their decrepit backward state, when they too could be improved. And so, I'm all for real-time payments, but I think that the Fed already had accomplished much of what it needed or perhaps all that it needed to accomplish with those, by getting the private sector to step up to the plate and establish a system, which has been in place since 2017.

Selgin: What the Fed should have been doing, first and foremost, is healing itself, right? Doctor heal thyself. Fix up your existing payments, you have a monopoly on settlement services. The Fed is the only institution that can push reserves from one bank to another directly. And I can explain how RTP handles that in a moment, and I probably should. But, RTP, ultimately, is also dependent on the Fed for funding. The Fed has this monopoly on settlement services, which it runs inefficiently, or, at least, inadequately, from the point of view of all the payments providers and who, unanimously, see that it should be open for more hours. They could have provided three windows a day for ACH payments, 24 hour service, and weekend service, holiday service, and provided a vast improvement for it would have made a huge difference. There'd still be scope for real-time payments.

Selgin: Because, some people really do need, it really matters a lot to some people, mostly to merchants, the difference between it taking a couple hours and it taking seconds, that can matter to a lot of people. And real-time payments can also be very useful for inventory management and information purposes, because the message is used to also include other information about the transaction. So, real-time payments is great, I don't want to put it down at all. But, if the Fed were responsible public servant, it would have improved its settlement systems, first of all, and then, it would have seen if it could also encourage the development of real-time payments, and it would only have intervened to do that if it was absolutely sure that private sector couldn't.

If the Fed were responsible public servant, it would have improved its settlement systems, first of all, and then, it would have seen if it could also encourage the development of real-time payments, and it would only have intervened to do that if it was absolutely sure that private sector couldn't.

Beckworth: Well, let me restate what I said earlier then, real-time payments is great, and there's good reasons to pursue it, but the reasons that have been listed, or argued for it have been the poor story, those in need. And what you're saying is that it could have been fixed a whole lot easier, and there's this more direct way of doing that. And I think there should be a cautionary tale for all those out there, who've argued, "Let the government do it, because they're the ones that they're going to care about the poor, they're the ones they're going to solve the problem." This should be a very cautionary tale that sometimes they don't get it right. And we could have had the best of both worlds, we could have had a system that does help the poor, as well as a real-time payment system been developed by The Clearing House.

Selgin: Yes, absolutely.

Beckworth: So, let's move on to some of the challenges with the real-time payment system in terms of costs, because, our time is running out here, and it looks like, listeners, we’ll have to have George back on to talk about the Fed's review.

Selgin: Yay, I've been preserved by status.

The Cost and Implementation Challenges Facing a Real-time Payment System

Beckworth: Yeah, bring you on for your next show. But, let's talk about some of the challenges in implementing the real-time payment system. So, I know you've touched on this cost issue, but one of the requirements of that 1980 law is that, whatever the Fed gets into, it has to recover its costs, when it does something. So, if it does a real-time payment system, it's going to have to cover its cost. It can't just have a freebie, where it hands out their service to the public.

Selgin: That's right, unless it cheats.

Beckworth: Unless it cheats. And what's interesting is, one of the implications that comes out of this, and this is where I've been really puzzled, and maybe you can explain why this is the case. But, if the Fed has to cover its costs, there's a good chance it will have to start offering volume discounts, so that, someone who does a whole lot of use of this payment system gets a cheaper price, which is typical, you see that in any walk of life. And so, who will that be? The big banks. The big banks will get cheaper prices than the small banks, most likely, if this cost rule applies, and so, small banks themselves, may be harmed by the Fed stepping into the real-time payment space.

Selgin: That's exactly right, David. I think the small banks, the community bankers have been sold a bill of goods. And here's how the story actually developed, when the RTP launched, they did so with a contractual commitment to their customers, to charge a flat fee. That is, no volume discounts. And that's a legally binding contractual commitment that they have. When the Fed announced that it was prepared to compete with them, or maybe after it decided that would compete, then RTP said, "Well, we may not be able to continue to offer this flat rate fee commitment, if the Fed enters."

Selgin: Now, they were very good reasons for the RTP to do that, because, RTP knew, The Clearing House knew, from experience, that the Fed was likely to offer volume discounts, and then it would have to follow suit, because it would lose the business of the large banks, and including, ironically enough, including many of the owners of The Clearing House. But, they, the bankers, are going to take the cheapest solution. And, in fact, that's exactly what happened in the Automated Clearing House space. Today, the Automated Clearing House network is jointly run by, ACH payments are jointly provided by the Fed and The Clearing House. Originally, The Clearing House charged a flat fee, then the Fed turned to volume discounting, to compete, not with The Clearing House, but with Visa, which was a big player and was offering volume discounts.

Selgin: Then, as ACH, sorry, as The Clearing House became a more important player or wanted to preserve its market share, it, too, resorted to volume discounts. So, The Clearing House has good reason to know that, if you're competing with the Fed, you might be competing with a firm that offers volume discounts, and you'll have to match. That's the only reason why they announced that their commitment to flat fees was contingent on whether the Fed ended up competing with them or not. Well, the Fed, it was a clumsy announcement, because it played into the Federal Reserve's hands, and the Federal Reserve and its supporters, they started saying, "See, you can't trust TCH. Look at them, they're planning now, to renege on their commitment to flat fees, they're going to screw all the community bankers. This is why you need us to intervene."

Selgin: But, the reality is, that TCH would have had no reason to renege on those commitments, and has no reason now, unless it's to be able to compete effectively with a Fed that's itself charging volume discounts. So, volume discounts are probably going to be the outcome of all of this, whether the community bankers like it or not, but it won't be TCH's fault. And this relates to another point, the TCH, at least, has made a now, qualified commitment to flat fees, right? They're going to be there until the Fed makes it impossible. And I think they'll stick to that, which is to say they'll stick to it till the Fed actually gets its system up and running.

Selgin: What about the Fed? The Fed has been urged to say what its pricing policy is, will it commit to a flat fee, if ACH does? I mean, if TCH does. And the Fed absolutely refuses to do that. When, at the hearing, you mentioned, they were pressed, Esther George was pressed to say, "Well, what is your pricing strategy?"

Beckworth: Now, who's Esther George? For our listeners.

Selgin: She's the president of the Federal Reserve Bank of Kansas.

Beckworth: Okay.

Selgin: She was pressed to ask, "Well, what are your pricing plans? RTP has its pricing schedule and its commitment that's qualified, what about yours?" And her answer, shockingly, was, "Well, we haven't figured out what rates we're going to charge." Now, you tell me, how did the Fed determine that this was a good thing for it to get into, and that it could do it and meet the obligations of the Monetary Control Act, if it still doesn't know what fees it's going to charge? Wouldn't it need to have thought about that, in order to determine whether this was a worthwhile project, and whether it could recoup its costs? I think so.

Now, you tell me, how did the Fed determine that this was a good thing for it to get into, and that it could do it and meet the obligations of the Monetary Control Act, if it still doesn't know what fees it's going to charge? Wouldn't it need to have thought about that, in order to determine whether this was a worthwhile project, and whether it could recoup its costs? I think so.

Selgin: So, what essentially George was admitting, though, perhaps unwittingly is, the Fed has no idea whether it can do this in a way that allows it to meet the Monetary Control Act, and it has no idea whether it's going to charge volume discounts or not, and it's unprepared and unable to commit to not doing so. There will be volume discounts, there's no question about it. But, again, it won't be RTP's fault.

Beckworth: Let me ask this, the Fed now has made a decision to pursue it, has the authority to pursue it, I guess, may it still opt out of it? I mean, it says, "Look, we have made a decision we can do it." Could it say, "We've looked at this, we've decided maybe it's not in our interest, we'll let The Clearing House have the RTP, and maybe we'll just focus on the existing legacy payment system"?

Selgin: The only way that could happen, David, is if there's such an outcry against what the Fed is up to, which will require people to give a lot more thought to the whole question of The Clearing House's role and whether it's okay to have a private organization operating an important payment service. If enough people were to study this matter, you could potentially get sufficient outcry. There are already some representatives, mainly Republicans, who are concerned about the Fed's decision. That, such an outcry could cause things to change. But I think it's very unlikely because of the misperceptions out there.

Selgin: At the hearing, again, you had Sherrod Brown just essentially identifying TCH with big banks, and saying, "Look at all the bad things big banks have done, we don't want these big banks to ruin the payment system." And, of course, he's correct insofar as TCH is owned by 20 big banks, but not all those banks were bad. And TCH is not responsible for what Citibank did, or what some of the other big banks did. Furthermore, the ownership and governance of RTP is separate from that of The Clearing House itself. It has its own little association, et cetera, and it has representatives from community banks on its advisory board, I think three of them. And finally, what people don't know about RTP is that, it's always operated its payment service is like a public utility.

Selgin: It doesn't seek profits, it pays no dividends, it gets no dividends from any of its payment services. And that's because it was formed, in the first place, back in 1853, as a club for banks to provide services that would themselves be beneficial to its members. It doesn't need to make profits from the services. The fact that the banks can use the service is enough of an incentive. Finally, there was a lot of silly talk about how this could be dangerous. You can't trust them to have enough safeguards, et cetera. But, The Clearing House's record, again, going back to 1853, is better than the Fed's. On safety, it's never had any failures on any of the several payment services it supplies, its outstanding. And, of course, if anyone had found anything bad about TCH in its history, they'd have brought it up. Nobody has, they can't, they just make claims out of thin air, as it were.

Beckworth: So, the RTP, or the real-time payment system that it set up, it's owned, there’s shares that are owned by the big banks, but it's run separately.

Selgin: Yes.

Beckworth: And let me ask this question, so, let's say, one of the big banks did have problems. The big bank's balance sheet is not overlapping with RTP. So, you could have a bank have problems, and RTP still would run just fine, is that right?

Selgin: Absolutely, yeah. RTP is fully funded. So, the way it operates is very simple. For listeners who are aware of the history, the Suffolk System in the 1820s in Boston, it's just remarkably similar. There's a joint account at the Fed, and this funded by all the banks that have joined the RTP system, that are members. And they have to put money in that account. And all the clearing and settlement is done on the books of this one account, just transferring funds, or credits, debit here, credit there, and it's fully funded. Now, that means they have to preload the account or the payments won't go through.

Selgin: So, that's where Fedwire and the National Settlement Service, that's where they put a wrench in the works, because, the account has to be pre-funded enough by the participants, to last through the weekend and last through the holiday. Of course, they wouldn't have to, they'd be more liquid and wouldn't have to pre-fund as much otherwise. So, that adds a little bit to the cost. But, anyway, it's fully pre-funded, there's no risk of failure in the settlements, because they are essentially backed, the payments are all backed by 100 percent reserves, and the reserves move… real-time gross settlement, you can't have a failure. It's not like where you're waiting for the funds and something could happen before the end of the day or something. So, there's no risk here, there's essentially no risk.

Selgin:  And as for security and all that, I dare say, the folks at The Clearing House know more than the Fed does, about how to get those things right.

Beckworth: Okay. Well, with that, our time is up. And, again, listeners, we will have George back on for, I believe, his seventh show next time, when he comes on and then we'll talk about the items I mentioned earlier, we didn't even get to the review and the Fed's operating system. But, George, thank you so much for coming on and enlightening us on this debate over the payment system.

Selgin: Oh, thank you very much, David. It's a pleasure to be here again, and I really appreciate the opportunity.

Beckworth: Macro Musings is produced by the Mercatus Center at George Mason University. If you haven't already, please subscribe via iTunes or your favorite podcast app. And while you're there, please consider rating us and leaving a review. This helps other thoughtful people like you, find the podcast. Thanks for listening.