Patrick Harker on Real-Time Payments and the Future of the Federal Reserve

Patrick Harker is the Rowan Distinguished Professor, Professor of Operations, Information and Decisions at the Wharton School of the University of Pennsylvania, and the former president of the Philadelphia Federal Reserve Bank. He joins the program to discuss his time at the Fed, the FOMC, Amish banking, his amazing academic career, and much more.

Subscribe to David's Substack: Macroeconomic Policy Nexus

Read the full episode transcript:

This episode was recorded on September 10th, 2026

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: Welcome to Macro Musings, where each week we pull back the curtain and take a closer look at the most important macroeconomic issues of the past, present, and future. I am your host, David Beckworth, a senior research fellow with the Mercatus Center at George Mason University, and I’m glad you decided to join us.

Our guest today is Patrick Harker. Patrick is the former president of the Philadelphia Federal Reserve Bank, and he joins us today to discuss his time at the Fed, the FOMC, and his amazing academic career. Pat, welcome to the program.

Patrick Harker: Thanks for having me.

Beckworth: This is long overdue, and thank you for coming on, and we’re recording here at Wharton. You were formerly a dean here. Now, you’re back after the Fed. You were also the president of the University of Delaware. You’ve had an amazing career and incredible time also at the Fed. You were there during some important developments. 

We met recently at the Atlanta Fed Financial Markets Conference, and we had a lot of interesting conversations outside the formal presentations. One of them, we were sitting around a table, and you told me this incredible story about the Amish and how a certain bank works with them. When I heard that story, I said, “Man, I’ve got to get you on the podcast to tell our listeners about it.”

Harker: I’m glad you are, yes.

Patrick’s Career

Beckworth: It is such a fascinating story. We’ll get to it. Before we do that, tell us more about yourself because, again, you were not just an academic. You were a dean. You were the editor of a prominent journal. You were a university president. Tell us about yourself.

Harker: I’m an accidental academic, I would say. I think it’s an important lesson. I also often tell young people this story. When I was a young student here at Penn, I was on the football team, and I got hurt my junior year. I was hanging around the department. I was an engineering student. A woman in the department, one of the administrative assistants, she saw me hanging around, and she was the chair’s administrative assistant. She said, “Pat, you look bored. I’m going to set you up with a meeting with the chair. You should take that meeting.” Now, when you’re a student athlete, time management is everything. When that constraint of the athletic part gets released because you’re not playing anymore, you do have a lot of free time. 

John Lepore, he was the chair. He hired me as a research assistant. I won a student paper competition, a national student paper competition, and then he started talking to me about getting a PhD. This never crossed my mind. I’m from a family of steamfitters and pipefitters. I was going to go and be an engineer. This was the furthest thing from my mind to be an academic. 

I tell students this because good advice doesn’t always come from the mentor with a title. It comes from people who know you and care about you and will tell you the unvarnished truth. You have to be coachable enough to take that. Again, my plan was not to be an academic. I ended up staying on for my PhD. I submatriculated here my senior year into a PhD program. I finished a PhD in two years after that, and also a degree in economics. My PhD is in engineering, but also I was ABD in economics as well.

I was all set to just become an academic. I was living the grant life, the soft money life. I had a lot of research money and so forth. Initially started to teach at the University of California Santa Barbara. Everybody says, “Why did you ever leave the University of California Santa Barbara?” We came back here to Wharton. I was a faculty member for quite a while. 

What changed me as the next opportunity I had that was really life-changing in many ways, which was in ’91 and ’92. I was a White House fellow under George H. W. Bush. That’s about building leadership. I realized—and I worked for the director of the FBI—that I had skills in leading organizations, managing organizations. When I came back here, I was a department chair. I was then interim dean, and then I was talked into being dean. This was not in my plan to be dean. I did that for eight years, and then went on to a challenge at the University of Delaware for another eight years. Then I was recruited to go to the Fed. I spent the last decade until June of last year as president and CEO of the Philly Fed.

Beckworth: What an amazing career. You weren’t just a professor. You were one of the youngest endowed professors at Wharton here. You were the dean from 1999 to 2007. I’m curious, what is it like to run a renowned business school, an Ivy League business school? Now, I know one of the characters of business schools—you’ve got faculty you’ve got to deal with, you’ve got budget constraints, you’ve got marketing, you’ve got to keep the reputation, high in the ranking. How do you do all that?

Harker: A really good team. What’s really important is to have a really world-class team. I had great people. One was a guy named David Schmittlein, who was my deputy dean, who then went on to be long-term dean at MIT Sloan, who then died recently, tragically, with cancer. I had a great team of people around me.

Beckworth: All right. Now, next, you were the president of the University of Delaware, 2007–2015. At a previous conference where we attended, it was the Hoover Monetary Policy Conference, I was lucky enough to sit next to you at a dinner table. We started talking about your time there. You mentioned your involvement with the NCAA. My ears, I perked up. It’s like, “Wow, we got someone who actually was on the governing body of sports.” You left before we got to the NIL and the transfer portal, and now the outright direct payments of athletes. You saw some of that coming in your time?

Harker: Oh, absolutely. When I was there, I was on the Division I board, so the large institutions. We were voting on governance changes to Division I athletics. It was all around football. By the way, step back. The NCAA was created to try to control football. I played college football, so I’m not opposed to football. The NCAA is created to rein in college football. The issues right now are college football and college basketball.

You could see that the governance changes then were going to lead to what we have today. It was consolidating power among what’s called the Power Five conferences, the major players. That’s where all the money was. You could just see eventually that the student-athletes were going to get paid. We’re now in pretty much of a mess. It’s hard to call these student-athletes students when they have no intention, many of them, to actually get a degree.

Beckworth: Yes. I got my PhD at the University of Georgia. When I was there, I taught a lot of athletes. It was a great experience, I had a lot of fun. I’m a big Georgia Bulldog fan. The difference between when I went there and what the stories I hear now—Carson Beck, for example, he got a lot of money with the imaging license. He bought a Lamborghini to drive around the University of Georgia campus. I just can’t imagine seeing your quarterback with a Lamborghini. The difference between him and the rest of the students who are just trying to get by, pay their tuition.

Harker: Again, in some ways, you could say that’s good for the student-athletes. They’re getting their fair share. It raises the broader question of why are universities in this business to begin with. I’m old school about this. The goal of athletics really is not just to be a farm system for the pros because the vast majority of student-athletes do not become pro—the vast, vast majority.

It’s to build leadership capabilities, build grit, build determination, build character. When you lose that as the goal, and the goal is for universities to make money, the student-athletes to make money, you do have to question why universities are in the business. Now, the Ivies, to their credit, the Ivy League is always stuck with a much simpler model of college athletics. I hope they stay with it.

Beckworth: Okay. Just a few other stories on this big transition. You left again before it all took place, before the transfer portal—

Harker: Yes.

Beckworth: —before the NIL. Okay. There are some great stories that did come out of this, at least it seems to me. Vanderbilt, they got a great quarterback through the transfer portal. Now, they’re competitive. They used to be the worst team in the SEC. There are stories like that. Then there’s other stories.

I taught previously at Western Kentucky University, a regional state school. I believe in 2015, in one of the last years I was there, they were in the top 25 ranking. Today, they never would be because any good player is going to leave after a few years and probably go to some other better school, get paid. That, to me, is a tragedy, that a school like Western Kentucky University is going to have a hard time competing nationally.

Harker: Again, you have to ask yourself, why is the university in this business? It’s really important for your listeners to understand. The vast majority of colleges and universities in America lose money on college athletics, a lot of money. In the case of Virginia, the Commonwealth of Virginia, it actually is a line on your bill at a state institution of your subsidy as a student to student athletics.

That number is not small. Again, you have to ask yourself, and a lot of students on campuses are asking themselves, “Why am I paying for this? What am I getting out of this? I don’t even get to go to the games because the tickets are sold out.” I think colleges, universities, and leadership there really has to step back and ask themselves why they’re doing this at all.

Beckworth: We also have this demographic cliff for enrollments. There’s this cost of sports, as you said, for most schools, the students pay for it. Where do you see this all ending up? When the dust settles, and the sorting takes place, what’s going to be the future of college sports?

Harker: I think the broader question is what’s going to be the future of the colleges in America. There, the Philly Fed has a very robust research program that we started when I was there around higher education, the financing of higher education, the economic impact of eds and meds on regions in the country. It has a database, three years of data on 520 regions in the U.S. where you can drill in and see how important these institutions are to these regions economically. A whole robust set of data. 

What we’re seeing right now is unprecedented. We are seeing merger and acquisition activity in higher education or simple closures because of two things. It’s not just a demographic cliff. It’s also international students are less likely to come here because we’ve become less welcoming to them. Many of these smaller institutions, smaller liberal arts institutions, were relying on those international students to make their enrollment. You’re starting to see lots of college closures and the implication that has for these communities all across America where they are the main game in town. It’s well beyond college athletics. It’s really about the livelihood and the success of those communities all across the country.

Beckworth: What do you see? Just big flagship universities surviving?

Harker: Yes. The mega universities, the flagship universities will do okay. There’ll still be niche programs for small liberal arts colleges who will find uniqueness in their offerings. Right here in this commonwealth, in Pennsylvania, we’re already starting to see consolidation of a lot of the mid-tier state universities because there’s just not enough students.

Beckworth: Let’s go to your work at the Fed. That’s how I know you as the former Fed president, the Philadelphia Fed, 2015 to 2025. Is that right?

Harker: Correct.

Beckworth: Okay. Tell us what’s unique about the third district of the Federal Reserve System.

Harker: It’s the smallest district geographically. Although, if you take upstate Maine, Boston would be smaller. They have upstate Maine, so we are the smallest district. Very diverse in many ways, though, because Pennsylvania itself is very diverse, very agricultural when you get outside the main cities. There’s a couple of things about it. 

One is it’s the home of the credit card industry because of Delaware. Delaware, it’s the place where credit cards were nationalized. Philadelphia has always had a deep expertise around everything consumer. It started off with what was called the payment card center. We morphed that into something called the Consumer Finance Institute at the Philly Fed because Philly not only has that expertise because of understanding consumer debt broadly, but it also runs the stress test, the CCAR, on the retail side. It does all the consumer modeling. It has a deep bench of expertise around mortgage lending, auto lending, credit card. That provides the system with a deep understanding of consumer expenditures generally, which is really important given that consumer spending is 70% of GDP. It really creates a really incredibly valuable asset for the Federal Reserve System.

Beckworth: One thing I like about the Philly Fed that I’ve used and benefited from is your data that you provide. In particular, the Survey of Professional Forecasters has been so useful in my research, and then all the real-time data that’s provided there. A lot of work, a lot of resources go into that, so thank you for your support of that.

Harker: It’s worth pausing on the real-time data because people don’t understand sometimes what Philly has. Philly keeps the data that policymakers see at the moment they make a policy decision because it’s easy to look back and say, “Well, they made a dumb decision,” but you’re using data that’s been updated several times since. What policymakers have in front of them when they make a decision is the data they have and not the future data. It’s really important, as you said, for researchers to have that data at the moment where people are deciding.

Beckworth: Great data, great work there. 

Amish Banking

Now, I want to segue into the story that you told me at Amelia Island about the Amish because the Amish are a unique sect of Christianity, and they have some traditions that prevent them from using traditional banking. Tell us about that.

Harker: One of the challenges I think we’re facing in the banking industry is the small banks, the community banks. They feel that they don’t have the scale, they can’t compete, especially scale due to technology. The Bank of Bird-in-Hand, Pennsylvania, is the exception to that in that a lot of people are saying, “Well, community banks aren’t going to exist just because they can’t compete.” If you really understand your customer, if you really deeply understand your customer, you can not just survive, but you can thrive. They are the perfect example of this. In fact, I have the CEO come to my class last year, and I’m going to ask her again to come this year. It’s an incredible story.

They are the first de novo. That is the first startup bank after the Great Financial Crisis in America, the first one. They don’t cater exclusively, but very heavily, to the Amish community. They understand the Amish community. There are a set of things that they had to work through with their regulators, the FDIC being one of them, to be able to serve this community. They generally don’t have credit scores. They don’t know your customer. They don’t have IDs often. They had to get a waiver so they could use the genealogical records that they keep, very meticulous records, to say know your customer and have people attest that is Earl. Earl’s okay. You can bank with Earl. They don’t do a whole lot of other things, but they’re great business people. 

The Bank of Bird-in-Hand figured out, and they worked through all this. Another example is they have what are called gelt buses, G-E-L-T, which is Pennsylvania Dutch for money. These are mobile branches. Again, they had to work with the FDIC because they move these around. At every location, it’s considered a branch location. If they change the schedule, they have to go back and get that approved again, but they figured out how to fast-track all that to serve these communities. They’ve done a tremendous job. Very few write-offs. They’ve lost very little money. 

The Amish are very good business people. They’re just incredible in terms of their character. They serve what the Amish call the English as well. We’re English. If you’re not Amish, you’re English. It’s a great example. Even down to the drive-through at the branch, they have two drive-throughs: one for the horse and buggy and the other one for cars. They really understand this community.

As a result, they’re getting requests to go to many other states where Amish communities are and pockets of Amish communities in New York state, Ohio, you name it. They’ve got to figure out, strategically, whether they want to do that. I use that as an example of a bank that really, really knows their customer and what they need. That is the alternative. They’re not alone. I gave a talk recently to a group of bankers here on campus. I use a couple of other examples, a bank in Wyoming that really caters to ranchers. If you know that customer base well, you can, as I said, not only survive, but thrive.

Beckworth: Community banks are very important and will continue to be so for these communities. That’s so fascinating. Now, you mentioned, to identify who’s a legitimate participant, they opened up a family tree book, and they point out on the tree who this person is.

Harker: People come and attest to who this person is.

Beckworth: They come and attest, okay.

Harker: Because they have no other records.

Beckworth: That’s remarkable.

Harker: Also, when you give them a loan for a farm structure or a house, there’s no insurance on those properties because the community ensures it. If your barn burns down, the community comes and rebuilds your barn. Again, for most commercial banks, there’s this steam that comes out of their ears when you hear something like that. What are you talking about? I can’t underwrite a loan like that.

Beckworth: Error, can’t compute. That’s incredible. Kudos to this bank for making the effort to connect with this community and providing those services. Just as an extreme on the other end, the Amish would never use stablecoins or tokenized deposits.

Harker: I’m not sure of that. They are very good business people. Yes, I think there’s a caricature of what the Amish are in reality. No, they don’t use technology in some ways, but they hire people to use technology.

Beckworth: Someone might have a phone for them who can have a digital wallet on it, and they can receive payments.

Harker: They don’t drive, but they hire people to drive them if they need to.

Beckworth: Where I was going with this is the Amish have their needs, and this bank has met them. A lot of the conversation now, and you were involved in, was with payments, settlements, FedNow. We’ll come back to stablecoins and tokenized deposits and such later. When you were at the Fed, something else you worked on was FedNow and payments. Tell us about that experience.

FedNow and Payments

Harker: The two big committees in the Fed—by the way, everybody focuses on the Fed, monetary policy. Absolutely true. Where you spend most of your time as president is not sitting at your desk and musing on whether the balance sheet’s too big or too small, or whether we should go up or down 25 basis points. Most of your time as a president of a Fed is spent running the system, the operating side of the Fed. That’s the part I think people don’t realize. The Fed is a big bank. It’s the nation’s bank. It’s the fiscal agent for the US government. It’s a bank to banks, et cetera. 

The Fed runs major payment systems, major IT systems. I chaired, for quite a while during my time there, the IT committee for the Fed, overseeing IT. You think, “That’s not very interesting.” It’s about a third of the Fed budget overall. The Fed spends a lot of money on information technology, like any large financial institution does. 

I was not involved directly in designing FedNow, but I was making sure they were resourced appropriately to be able to pull off what they did. Also, we had to break some of the governance rules. The Fed is a very complex organization, some would call Byzantine in some of their decision-making when it comes to these operational issues. We cleared the way for FedNow. We changed a lot of the procedures and so forth. We set them basically up as a skunkworks in some ways so they could get their work done. I was, by the way, not a fan. I’ll be completely honest. I was not a fan of starting FedNow. 

Beckworth: That is so interesting.

Harker: The clearinghouse had a private sector RTP that could provide that service, even though there’s a question of ubiquity in terms of small banks and others having access, which is real. I think there was a strong case to be made that FedNow needed to be started for that reason, if none other. I thought we should just leapfrog to what the new technology is going to be. That would be tokenized whatever. You can think of a central bank digital currency as a tokenized dollar or a tokenized deposit that really leapfrogged the technology to the next thing. I lost that argument.

Beckworth: That is so interesting because it was you and Randy Quarles, as I understand it, were probably the ones pushing against it.

Harker: There were a couple, but we decided to do it. Then I was all in.

Beckworth: You’re a team player. Let’s go.

Harker: I’m an old football player, man.

Beckworth: You don’t like to play, but you’ll do it.

Harker: I was not the glory guy on the field. I was a defensive lineman. When you decide something as an institution, you can have an argument internally, but then when you decide to do it, you got to make sure it’s done right. I was involved from that perspective. Going forward, I think it does play an important role. I am quibbling on some of the design issues right now in some of my academic work.

Beckworth: That’s a question I have because both as a former Fed president, and as an expert in operations management science and engineering, one of the issues I see is this real-time payment system is a utility, and you need large economies of scale to bring the cost down, but now you got two of them. It’s like having two sewer systems. You can’t get the cost curve down low enough, or is that something that’d be surpassed?

Harker: Two isn’t always a bad thing. Let’s abstract away from payments and look at cloud computing. Most organizations have two cloud providers, maybe three, but they have two because of redundancy needs. Now, they need to be interoperable from your perspective of running the organization, but if one goes down, you got to try to figure out how to keep your business running. Having two payment systems like that isn’t necessarily a bad thing as long as there’s an interoperability with them. It reduces risk.

Beckworth: That currently is not there, right?

Harker: They’ve got to keep working on that.

Beckworth: You’re hopeful, but you see areas where they need some improvement going forward.

Harker: One of the questions is what do we mean by real time? One of my mild obsessions right now in terms of my academic work is fraud on these networks. There’s a saying in the payments business: Faster payments equal faster fraud. The payment people hate that, but that’s actually a quote from the guy who is running the US Payments Council, right? Because the fraudsters are going to quickly move in, move out, try to find the path of least resistance, and there’s various ways of doing this. Some of these payment rails have gotten hit with some pretty large fraud.

The problem is we don’t actually know how deep the fraud goes because it’s not always reported or publicly reported, but it’s big. You ask yourself why real time? I started thinking about this, and I partnered with a colleague here at Wharton in the finance department in thinking about this some more. They were doing work on the Brazil Pix system, and their approach was more on the liquidity side. Both argue for you want a little bit of delay. Real time shouldn’t mean real time. If I can just take an example of a calibrated model I ran recently: If you just add 16 seconds into the payment, you can reduce fraud because now you can better look for that fraud, right?

On the liquidity side, if you, say, double that to 30 seconds, you can now start netting more payments, and the banks can have lower liquidity, or they have to hold in the network, which has an implication for their risk-taking because if I have to hold a lot of liquidity on the payment side, and I want a certain return, I’m going to take more risk on what’s left. That’d be the argument. You don’t want to do that. Now, the argument against this, by the way, is, well, these real-time payments are eventually going to compete with the card networks; 30 seconds, 16 seconds is not possible.

You’re not going to stand there at your Walmart line and wait for 16 seconds for your payment to be approved, right? Milliseconds work there. There’s ways around that. There’s ways of engineering the system. For most people, real time doesn’t have to mean right now. A little bit of delay actually buys you something. The point I’ve been trying to make with some of this work is everybody’s rushing to get even more and more real time, and I’m saying, well, maybe we should hold off here. Maybe it’s better to build a little delay in for lots of reasons, whether the liquidity reason or the fraud reason.

Beckworth: As a good economist, you’re saying there’s tradeoffs in everything.

Harker: Absolutely.

Beckworth: There’s new corner solutions. There’s never 100% in or out. There’s some margin that’s going to be some tradeoff. You have a paper. We’ll provide the link in the transcript titled “Speed, Fraud, and Bank Liquidity: The Operational Architecture of Real-Time Payment Systems.” Check that out for more. That is so fascinating. Now, I know Darrell Duffie recently had a paper on netting as a way to reduce the demand for all this extra reserve sitting on banks’ balance sheets. There’s lots of implications that flow out of this.

Harker: Absolutely. My colleague in finance, along with some colleagues at MIT and so forth, their paper actually looks at this and says, well, if you’re going to take more risk in order to make up for the liquidity you have to hold for the payments, net, that’s a bad thing for the system.

Beckworth: Okay. There’s tradeoffs even to Darrell Duffie’s proposals. The practical side, I guess, of real-time payments—and just for listeners, most of them know already what this means. I’ve had guests on, like Aaron Klein, who’s from the Brookings. He’s real hung up on the person doesn’t have a lot of money in their checking account. They get direct deposit on Friday. They don’t get access to it Tuesday, let’s say. For them, that could be a big deal. With real-time payments, 16 seconds is not a deal-breaker for them. They can get their payment Friday, same day as direct deposit.

Harker: No, I get it. I get that if you’re living paycheck to paycheck, you want to get that money in your account because you have bills to pay that week. As I said, I’m from a family of union pipefitters, so I get that. That’s the world my family lived in. We can do that pretty easily. It’s really more these payments out. That’s where most of the fraud is.

Crypto Assets, Tokenized Deposits, and Stablecoins

Beckworth: Stepping back, what do you think of this big movement in general toward crypto assets, tokenized deposits, stablecoins? There’s this big push to lower costs, to be quick. Thoughts on where we’re going?

Harker: Actually, I think it’s really good that we’re trying to innovate in this space because if we can get more efficient—what you always worry about here is, as you said, there are scale economies in all these. People tend to extract economic rent when they get scale economies. How do we do this in a way that creates more competition? Let’s take one example, tokenized deposits versus stablecoins. The stablecoin market is there. We have it. It’s in legislation.

The banks, of course, are responding in kind. They’re not going to just sit still, and they have one advantage of a tokenized deposit versus stablecoin. They can pay interest. They have a built-in advantage of tokenized deposits. We’ll see how this all shakes out. The market has to do what the market’s going to do. It seems to me that if I can do everything I can do with the stablecoin, and get some interest, I’m probably going to do that. It’s tokenized deposit rather than with a stablecoin, right? Again, but we’ll see how this all plays out. I think cryptos are completely different. Some view it as an asset class. I’m not quite sure.

Beckworth: Definitely not a transaction.

Harker: It’s not a transaction tool. I think all this innovation is critically important. There are still areas, even with all that that I think are frustrating to Americans. It starts with when you swipe your credit card at a merchant. There’s still that 3%. That is, I think, one of the big challenges for the merchant or for the customer that is still out there. If you think about it, the people who pay cash, if there’s not a separate price for cash or credit, the people who are paying cash are subsidizing the people who are paying with credit.

It’s usually lower-income people paying cash who are subsidizing higher-income people who are paying credit card. That, to me, is where some of the innovation needs to occur because that, relative to other jurisdictions in the world who cap those rates, let’s say 50 basis points, we’re not even anywhere—

Beckworth: Really? They do that in Europe? 

Harker: Yes.

Beckworth: Interesting.

Harker: Now, there are costs to that, too. Like any regulation, there’s costs.

Beckworth: Tradeoffs.

Harker: There’s tradeoffs. That is probably one of the biggest frustrations that merchants have, for sure, and customers do, too, when it’s explicit on the bill, like I’ve got to add that 3%.

Beckworth: That is so interesting. We recently have talked to some ECB officials, and they’re going down the path of having a public digital currency, a CBDC. One of the motivations, not the only one, but one of the motivations is they want to have a domestically based payment system because they have major credit cards over there dominating this space. Maybe some of it is they want to get away from those fees, but also have a domestic provider.

Harker: That’s Pix, too. That’s the Brazil Pix system. It’s wildly popular in Brazil. It’s popular for two reasons. One, it’s becoming ubiquitous with lower costs. Second, given a country like Brazil where the population is spread very far and wide and in very difficult places to serve with, say, cash, it’s like why the Bahamas created the SandDollar because it’s really hard to move that cash to all those islands. There’s an operational advantage to it, as well as, again, these network effects. Once everybody’s in on the system, it runs more efficiently. Pix is a great example.

Beckworth: This is great. I just actually did a podcast with the governor of the Bahamas Central Bank[MH1] . We were talking about the SandDollar. It helped me see CBDCs in a new light because I’ve been very like, I don’t know about privacy, but I’m thinking this very much from a US or big economy perspective. When you have these smaller economies, he told me, “Look, it’s costly to get cash to these islands.” They may be banked, but they’re not really banked because the banks don’t service them very well. This public utility will provide people who are less fortunate the chance to have access to the payment rails. Same thing in Brazil. It’s a more nuanced, complicated story than CBDC is bad. It would actually provide a service to those who are in need. An interesting conversation. 

Consolidation at the Fed

Going back to your time as Fed president, because you alluded to this, you oversell the IT, and so you managed back offices. I want to bring up something that’s been a little contentious lately. This has been a call by Governor Chris Waller to consolidate or centralize some of the back-office operations. What exactly is going on there? From your perspective as a former president, what are your thoughts on it?

Harker: Let’s start with the big rocks in the Fed. The payment rails have already been consolidated. They’re under the authority of a payment executive, currently Mark Gold, who’s retiring at the end of this year. That’s already been consolidated in the system. The IT is already moving toward consolidation. Supervision and regulation is a delegated function from the Board of Governors, so they have every right to consolidate more and more.

You still need local bank examiners, so you can’t consolidate everything. You’ve got to have a local staff. If you look at where the big bodies are in the Fed, and that’s where the cost is, it’s in the bodies. They’ve already either been consolidated or on their way to being consolidated. I understand what Chris is saying, but at the same time, I’m not sure how much more juice there is to squeeze. Yes, you could squeeze HR, but you’re not going to save a lot of money from the HR function.

Beckworth: It’s created a bit of a ruckus, though. I’ve seen a lot of Fed people like, “Oh, no.” You’re saying, “Look, we’re already there.”

Harker: We’re moving that way. My time was up before I finished the consolidation of IT, but it was already in train. We were already moving toward that.

Beckworth: You mentioned one of the things that’s been consolidated is banking supervision. Michelle Bowman, for example, whoever the executive is at a regional bank, they report directly to her in terms of bank supervision. Is that how it operates?

Harker: It’s complicated.

Beckworth: Explain it to me. How does that work?

Harker: Yes, it’s a matrix. The head of supervision reports to the president, but also has a dotted line to—

Beckworth: Oh, a dotted line.

Harker: Yes. Because they are technically employees of that reserve bank, not of the Board of Governors.

Beckworth: I bring this up as a concrete example. March 2023, San Francisco Fed, people called out Mary Daly, but who really did the supervisor report to? Her or to Michael Barr? How do we think that works?

Harker: Yes.

Beckworth: Yes. Okay, both. Okay.

Harker: There, the loop is the complicated part. You’ve got the supervisory side, which is the local staff or staff of San Francisco, and they do report ultimately to the president of San Francisco, but they also have oversight from the Board of Governors’ supervisory staff. 

There’s a third player when it comes to issues that a bank has, and that’s the general counsel at the Fed because there’s legal issues. That loop is what slows it down often in terms of the debate. Even outside of the reserve bank, within the Board of Governors, the debate between the general counsel and the supervisory staff sometimes takes time because these are complicated issues. They’re not easy.

Beckworth: You told me something else at Amelia Island that when you first became president, you had a veto power over bank supervisory issues, they could appeal to you as a president?

Harker: Yes, that’s changed. It’s changed over my time there, but it was the case. Now, if a bank had a finding, and they wanted to complain about it, and they wanted to rescind that finding, they would appeal to the local reserve bank, and that officer would be the president. Eventually, it would get to Washington as well. I think now the president’s out of the loop completely on that.

Beckworth: Is that a loss?

Harker: No, I still think that local knowledge is really important. I do think that the staff in Washington does need to reach out to the local staff in the reserve bank to understand really what’s going on.

Beckworth: All right. Let’s move from back offices—

Harker: To front office.

Federal Open Market Committee

Beckworth: —to the front office, what makes the headlines, and that’s monetary policy, the FOMC. You were on it for 10 years, pretty amazing. In 2015 and 2019, there was a tightening cycle, balance sheet runoff during that time. How should we characterize or think of that period? Because it comes to 2019, and the tightening stops. There’s a repo crisis moment as well. From your perspective, what was happening during that time?

Harker: It was a complicated time. When I first got there in ’15, just a few months after that, we started lift-off. It was the right time to start lift-off. I was fully supportive of that. We were ready to go. Then we get into, as you said, the period a couple of years ahead, 2019. We’re trying to do this balance sheet runoff. The approach we took was actually due to one of my senior economists who eventually becomes my research director at the Philly Fed, Roc Armenter.

Beckworth: Oh, I just met him recently.

Harker: If you actually look at the transcript, which I did the other day, I didn’t pore through it, but it’s the beauty of AI. I just asked it to look for certain things in the transcript. There’s a quote by Jay in one of the transcripts because people were calling it the Harker plan. I said, “No, if it works, it’s the Harker plan. If it doesn’t work, it’s the Armenter plan.” Because it’s really Roc’s idea. The idea was there is no way you can know a priori what the appropriate level of reserves is.

There were estimates all over the place of what we thought the appropriate level of reserves would be. We had an estimate which was about $1.5 trillion. We thought, “That’s too high. Nobody’s going to believe that.” We knocked our estimate down to a trillion. Other people had it at $500 billion. We eventually settled more on a $500 billion number we had in our heads, and that ended up being around $1.5 trillion.

That’s the number that shows up. The idea was you don’t know what that number is. Come down, and at some point, just level off. Just stop, and just let it ride. Let the liabilities build up. Eventually, you’ll know. We thought, yes, we can stop way short of $500 billion, and we’ll be okay. No, it ended up that was not true. Because you just don’t know. You don’t know. The only way you know it is the market reaction to it. You can’t know it a priori.

The balance sheet runoff issue is always a complicated issue. Look, the Fed moved to an ample reserves regime, and that had everything, tradeoffs. There are pros and cons to that as opposed to the old system. That means we’re running monetary policy with administered rates. Which, by the way, there’s a little quirk in the Fed. The FOMC votes on the fed funds rate. It’s the Board of Governors who vote on the administered rates. By tradition, they’re tied, but they’re only by tradition. You could imagine scenarios where that doesn’t work.

Beckworth: Let’s don’t go into that dark world. There were talks of this earlier in the year that people would be appointed. 

Harker: Yes. We learned a lot out of that experience. It was a humbling experience in that sense, because we just simply didn’t know what that ample reserves level was.

Federal Reserve Balance Sheet

Beckworth: I’m glad to have you here because this is a hot topic now. On the podcast, I’ve talked a lot about this with other guests. We have a Balance Sheet Task Force. They’re actually stepping back and doing a review of the operating system as well as balance sheet policies. I’m just wondering, what are your thoughts on, in a textbook world, ample reserves versus scarce?

To be clear, we can never go back to pre-2008. Pre-2008, we didn’t have liquidity regulations, number one. Number two, we didn’t have interest on reserves. Even in a scarce reserve framework, a normal system still has interest on reserves. You would have had more reserves if you’d had scarce reserves and interest on reserves back then. I’m not suggesting that world, but going forward—and I bring this up because there’s been a lot of interesting research done on this, as you know. 

One in particular I find fascinating is Raghu Rajan and his co-authors. They talk about this ratchet effect that what happens when the Fed does QE, and then it’s all this liquidity injected into the system, but then banks go and extend their own liquidity. On aggregate, there really isn’t any new liquidity. It’s more fragile than it appears on the surface. Every QE gets ratcheted up a little, at least the reserve side of it. As someone who’s both a former Fed official, you’re an engineer of financial systems, what do you make of that, and can anything be done at all about it?

Harker: I think it’s first really important that you recognize that the balance sheet is an operational issue. It is not the major tool of monetary policy. I think we’ve got to take that off the table because it should only be used in exigent circumstances. Now, that’s the problem. Every crisis, and we can put crisis in quotes, seems to be exigent circumstances. 

In retrospect, if I look back at the period with Silicon Valley Bank, we probably overdid it. In the sense that we were very worried about contagion. By the way, really, there was this poor bank here in my district who had a name instead of Republic First, First Republic. They got confused with the two. It’s just a little local bank. The CEO had to come out in public and say, “We’re not them. Please leave us alone.” 

There were definitely contagion issues, but wasn’t as bad as we thought. This is the moral hazard issue. This is the fundamental moral hazard issue, which is really difficult for people in that seat to deal with. You can have a great theory, and you can have a great ideological point of view on this, but when it’s all hitting the fan, and you’ve got members of Congress calling you, this is where you have to say, “No, we’re going to do this and no more,” or at least the Fed should borrow a page from the Bank of England and put time limits on what we do, and think through a little more carefully some of what we are putting in place.

The bank term funding program, for example, with no haircut on collateral, of course you’re going to take that loan. There was no stigma because you were dumb not to take that loan. That was probably not very well thought out in retrospect, but it was done in haste. The issue, if you step back from this, and I’ve been writing a little bit about this on a Substack post, is how do we avoid even getting into this situation to begin with?

I think you’re right. The ratcheting effect is very real. How do we just stop getting toward the cliff? We don’t know where the cliff is exactly, but we should create some buffer. Let me give one example. The countercyclical capital buffer has been zero since its inception in the United States. In the UK, it’s 2% capital. Now, people are saying, that’s forcing banks to hold more capital. No. No. You lower the other capital requirements. There’s still some cost.

Beckworth: As the term countercyclical suggests.

Harker: There’s some friction there, but what’s the beauty of the countercyclical? There’s no stigma, nothing attached to it. When you decide you need to release capital, you release it for everybody at the same time. It goes from 2% to 0. The whole banking industry gets that all at once. We’ve never raised it above 0. This is the engineer in me. I used to do control theory, control models, optimization models.

The goal in stochastic control models, never get to the edge, never get to the turbulent flow in the stream. You always want the laminar flow. Stay away from the edge. This is what I’ve been writing about. There are tools we have that would keep us away from that edge where we don’t have to use QE and so forth so much, but we’re not deploying them. You have to ask yourself why.

Beckworth: That’s on the books. That’s a legal law, right?

Harker: Yes. That’s in there. They can deploy it. I was in the room where we were debating. It’s a Board of Governors decision. I was in the room debating why aren’t we increasing this? There was pushback because of course the banks are going to push back on it. You have to do it in a way that, again, it’s not increasing overall capital requirements.

Beckworth: One more item on this before we move on. Even if one wants to keep the ample reserve system, there still seems to be an openness, I would call a sea change, or a shift in the Overton window at the Fed in terms of lowering the demand for reserves. Everyone can agree on that. Beyond that, there’s disagreement. There’s this big push, and I’m broadly sympathetic to it, but I want to hear you.

Again, you’ve worked with banks. You’re a former Fed president. You’re an operations science guy. You’re an engineer. There’s been this push, at least conceptually, toward making the discount window more accessible, more business as usual, so that the demand for reserves is less. That could be through bringing regulations into the discount window so your line at the discount window could count toward liquidity coverage ratio, things like that, or bring back term auction facilities. What are your thoughts?

Harker: It’s ultimately the stigma issue. This gets publicly disclosed that you went to the discount window, which the market’s going to read as that bank’s in trouble. The reason I bring up the bank term funding program, everybody used it. There was no stigma.

Beckworth: It’s very similar to the discount window, but just no stigma.

Harker: The countercyclical capital buffer, there’s going to be no stigma because everybody’s going to get that. There’s lots of things you can try to engineer around the discount window, but it ultimately comes down to the fact that that’s going to be publicly disclosed, and banks would instead like to go to the reverse repo market or something else.

Beckworth: What about the term auction facility? That was a version of the discount window that was used 2008, 2009.

Harker: Yes. Again, this is in the moment of the crisis where lots of people are going to use it.

Beckworth: A good fair point. I’m arguing for more business as usual. It’s not going to be used because of the stigma. Fair enough. 

Let’s move forward. For the sake of time, I want to transition into some of your current work. There’s a lot more I’d love to ask you about FAIT, COVID, inflation?

Inflation and the Fed’s Mandates

Harker: I’ll tell you one thing very quickly—

Beckworth: Please do.

Harker: —on average inflation targeting, and I think where we are in monetary policy, very quickly. I’ve come to the conclusion that the idea that we have a dual mandate, it’s in law, it’s what we do, but this is what leads us down a bad path. I think we should actually start to think more in a hierarchical fashion. Inflation is job one, and if you’re in spitting distance of the inflation target, then you worry about unemployment. When I step back and think about my time, that’s what got us into trouble, is trying to do this tradeoff.

Let’s take the situation right now, a tradeoff where what’s happening in the labor market has little or nothing to do with things that the Fed can control: AI, immigration. We can go down the list. They have nothing to do with the Fed. They’re affecting the labor market. They have nothing to do with what the Fed can control, positively or negatively, in the short to medium run, maybe in the long run.

They should just focus on inflation. That’s why I’ve become surprisingly more hawkish. I was hawkish coming out of the pandemic, too. I was one of the ones early to say, we’ve got to start raising. Right now, the idea that we’re going to do this tradeoff, no. The labor market, there’s not a lot you can do to affect it anyway with monetary policy. Focus on job one and get inflation to target.

Beckworth: I’m guessing also part of your skepticism is we don’t really understand what’s happening in real time to these real markets either.

Harker: No. Especially the labor market. There are all kinds of forces affecting that.

Beckworth: Measurements are revised, or one measurement says this.

Harker: Yes.

Beckworth: Okay, fair enough.

Harker: I had to get that out of my system. 

Beckworth: No, that’s good. If I had time, I’d like to talk to you about nominal GDP targeting as a way to get around that. Tell us your Substack name.

Pat’s Substack and Thoughts on AI

Harker: Load-Bearing. It’s Pat Harker, but it’s Load-Bearing.

Beckworth: Load-Bearing, yes. Check it out, folks. Make sure you subscribe if you haven’t already. I’ve enjoyed reading it. Some of the stuff we’ve talked about, he’s covered on there. You had a piece that really resonated with me because there’s been a lot of commotion about AI. In fact, we’re recording this here. Just in the past few days, there’s been increased talk about AI killing us. Anthropic people have resigned. Oh, Terminator effectively could happen soon. Let’s set that extreme scenario to the side.

Let’s talk about some practical applications or implications of this for people like you and me. There was this Stanley Druckenmiller op-ed in the Wall Street Journal where, it was his ideas, and he just told the AI, “Write this up for me. Here’s point 1, 2, 3,” and it went out. The op-ed itself was fine, but then when people found out it was AI-generated, they lost their minds. Many journalists, many purists, oh, the lost art of writing and such. You had a great post response to this that I had been thinking about, too, similarly, and I think the title was “The Speechwriter I Never Mentioned.” Tell us about that essay.

Harker: For 26 years, I was running organizations: Delaware, here at Wharton, and the Fed. I always had the speechwriter. I never started my speech saying, I’d like to thank Ethan, my speechwriter, for these words. No. But the process was, from years of experience in building up judgment and my own views on things, I would sit with my team. We’d talk about what I wanted to say, how I wanted to say it. They would go off. The research team would make sure the data was correct. They would back it up. Then we’d iterate, iterate, iterate, and I’d hit the podium and say it. I never disclosed that that was the process. I now have to disclose the use of AI. 

Now, how did I use AI? I uploaded all my 100-plus articles that I’ve written over my career, all my speeches that are public on the Fed website, all my op-eds, everything into Claude. I said, here’s me. It learned me. Now, instead of iterating with my speechwriter and my research team, I interact with Claude. Now, I can do that. There’s a counter to this, which is, yes, but the art of writing it makes you think, and that’s absolutely right.

I tend to think more in equations, just given my academic background. I tend to think in models, but that’s true. I still do that. I still sketch out before I go to interact with the AI what I want to say, what are the points I want to make, how I want to make it. Here is the truth. Druckenmiller can do that because he has years of experience in building judgment. I could do that for the same reason. The young person walking on this campus today cannot.

I think the idea that we’re going to have a one-size-fits-all for the use of AI when it comes to writing and so forth, even in my own field, AI is becoming so good at mathematics. The idea in my old field was you got slaps on the back for a clever proof. Those days are going to be gone. AI can do the proofs better than you can. It’s going to be really your ideas and your creativity in terms of the data you bring, the slant you take on a problem that’s going to distinguish you, not crunching the math or actually writing the words.

Beckworth: Absolutely. Like you mentioned, you don’t have this big staff anymore, but you do have an RA called Claude who’s there with you. He can help you think through the issues. When I write pieces, I’m like, what am I getting wrong here? Is this consistent with my earlier writing? What am I missing? What is a critic going to say? It’s immensely useful. I agree from that perspective and also for research as well. 

I wonder also if we’re coming to the cusp of something very similar that happened in the past. That is before the advent of the printing press, we have the famous Gutenberg printing press, people memorized a lot. People memorized the Odyssey.

Harker: The Psalms, they memorized the Psalms.

Beckworth: In Hebrews, it was their tradition to memorize the Torah and then pass that down to their children, and the children memorized. When we got to the printing press, people freaked out. “Oh, we’ll lose the art of memorization. Humanity won’t be the same.” Then we do new things. We couldn’t do advanced math without textbooks with equations to teach us. Linear thinking is possible when you have written text, right? It’s not as easy with this memorization. We give up the ability to memorize books, but we open up new possibilities. That was a painful transition. I’m wondering if something similar is happening now.

Harker: My wife’s a middle school math teacher. The calculator wars went on. Can you use a calculator and not use a calculator? No, pretty much. It’s some mix of the two. That’s my point. For these young people walking around campus, they need to stretch their brain. Their use of AI is going to be different than you and I. Just like that third-grader who’s learning how to multiply, they have to learn that, just the way of thinking about it. Then eventually, when they become a senior in high school, they can just use a calculator. It’s the Boolean logic way of thinking about the world. It’s either all good or all bad.

Beckworth: Yes. Like you said, AI is creating new solutions for math problems, for proofs, and lots of new opportunities opening up. 

We’re near the end of the show here, Pat, but where can people find you? You mentioned your Substack. Are you online, elsewhere?

Harker: LinkedIn and Substack are the two places. Also here at the Wharton School, I’m a co-academic director of something called the Wharton Future of Finance Initiative. You can see some of the work we’re doing there as well.

Beckworth: You never sleep. You keep going. Active person. Thank you for coming on the show. Our guest today has been Patrick Harker. Thank you for coming on.

Harker: Take care.

Beckworth: Macro Musings is produced by the Mercatus Center at George Mason University. Dive deeper into our research at mercatus.org/monetarypolicy. You can subscribe to the show on Apple Podcasts, Spotify, or your favorite podcast app. If you like this podcast, please consider giving us a rating and leaving a review. This helps other thoughtful people like you find the show. Find me on Twitter @DavidBeckworth, and follow the show @Macro_Musings.