Patrick Harker, president of the Federal Reserve Bank of Philadelphia, talks about the U.S. economy and the impact of fintech.

The U.S. economy is growing at a good pace, and the labor market feels “really tight” with job levels running below the natural rate of employment, said Patrick Harker, president of the Federal Reserve Bank of Philadelphia, at the recent “Fintech: The Impact on Consumers, Banking and Regulatory Policy” conference. During an interview at the event, Harker told Knowledge at Wharton editor-in-chief Mukul Pandya that he sees the central bank raising the fed funds rate target in December and then three more times in 2018, “assuming that inflation comes back.” He also noted that the already low inflation rate still could be overstated, which would mean that productivity growth has been underestimated.

As the Fed starts unwinding assets on its $4.5 trillion balance sheet in October, Harker expects the impact to be a 25-basis point increase in the fed funds rate over a four- to five-year span. “We’re not talking about a very large impact,” he said, “but we don’t know for sure.” As such, the Fed will move cautiously and see how markets react. Harker also said the Fed should continue paying interest on excess reserves, which it only started doing in recent years, because it is the “only tool” the Fed has along with the overnight reverse repo to manipulate the fed funds rate.

When it comes to financial technology, or fintech, Harker said banks have been innovating with technology for decades. This time feels different because many innovators are not financial institutions or other incumbent players, but outsiders. He noted that the Fed is watching developments in fintech carefully but it is not ready to embrace digital currencies as yet without further study. One positive contribution of fintech is its potential to broaden access to finance for all parts of society, especially the unbanked. But a sticky issue is to ensure that the most vulnerable in society aren’t overlooked as algorithms take over many tasks. “How would you even build an algorithm to have that sense of fairness?”

 An edited transcript of the interview follows.

Knowledge at Wharton: This week, Fed Chair Janet Yellen said that the Fed may have overstated the strength of the labor market, and the rate of inflation has come down, too. What do you think is happening on the inflation front, and are rate increases justified this year or next year? What are you views?

Patrick Harker: I’ll put the inflation picture in context. GDP growth continues along just as we pretty much have anticipated according to our forecasts, running slightly above trend, around 2.3%, 2.4%. You can pick a number somewhere north of 2%. Q1 was weak. Q1 is always weak. You should recognize that we have a seasonal issue with measurement of GDP in quarter one, and so take that with a grain of salt. Q2 has just been revised up to slightly over 3% — 3.1%. So in terms of GDP, things seem to be going along just fine.

Now, of course, the devastation that we’re seeing from these hurricanes is horrific devastation. It’s obviously terrible for the people who are involved, and our thoughts go out to them. But in terms of the economy as a whole, the best estimate [of their impact] right now is that we’ll see about 100 basis-point decline in GDP this quarter, and then it will bounce back to about 1.3% in the subsequent quarter — some number around there. So it will have an effect, but it will be a transitory effect, as that works through the economy. So that’s on the GDP front.

“The labor markets feel really tight.”

With respect to employment, we’re running below what most people would think is the natural rate of unemployment. The labor markets feel really tight. And that’s not only in the data, in terms of unemployment, but you look at the JOLTS (Job Openings and Labor Turnover Survey) data — the job opening and quits data — things are really tight. You’re hearing this … over and over again, of the skills mismatch question. Companies can’t find the skilled workers. We’re not talking about Ph.D. economists and computer scientists. We’re talking about truck mechanics, electricians, plumbers. You can go down the list.

Now, we get to the inflation question. So if that’s true, where did the Phillips curve go? What happened to it? Did it just disappear? Well, the conventional wisdom is it’s flat. But it’s been flat for a while, and that relationship between unemployment and inflation, we’re just not seeing. What’s happening with wages? Well, they are flat, but some recent work that was done by our colleagues in San Francisco point out the fact that if you look at somebody who’s staying in a job through a sustained period of time, their wages are going up well above what we’re seeing on average. It may be just a compositional shift. Part of the story of the wage issue is that we, the boomers, are retiring — high-priced boomers are being replaced by lower-priced millennials. And that is keeping the average down for now. Again, there needs to be more work in this area, but we are starting to hear and see some wage pressure.

Now there’s some recent work done by Mike Dotsey and his colleagues here at the Philly Fed research department on the Phillips curve. The conclusion of that work is, whether it’s dead or not, for the last several decades, it has not been a very good predictor of inflation. There may be other metrics and other measures that are better predictors of inflation than what we’ve traditionally thought of with the Phillips curve. … If you think of our dual mandate, [the one thing] that is problematic is inflation.

The job market seems to be incredibly healthy on average. [But] there are pockets where we need to bring people off the sidelines into the labor force and help them get the skills they need. That’s something we’re doing in Philadelphia with what we’re calling the “Economic Growth & Mobility Project.” It’s working with communities, trying to do research and outreach in that area [to offer] things like apprenticeship programs. We just put out a report on apprenticeship programs that are not your mother’s and father’s apprenticeship programs. They’re no longer just in the trades. We’re seeing it in IT and health care, financial services. Apprenticeships are now becoming, as the market has gotten tight, one way for companies to get the talent they need.

“We may be overstating inflation … [and] we may be understating productivity growth in the economy.”

On inflation, we are running below our 2% target. Only for bankers and central bankers is this a concern, that, “Oh, my gosh, we’re behind on the 2% target.” The average American is saying, “I have a job. Inflation is 1.5%. Life is good.” … So, what’s wrong with this picture? Everything’s great. Well, of course there are risks — bankers would like higher margins obviously, for [their] spreads. And we have the risk of, if and when another negative shock hits the economy, having the leg room to be able to move rates down. That is a problem.

So inflation is the one area that does give us a little bit of pause. It gives me pause. And it’s a complicated story. It may be that the Phillips curve or Phillips curve-like mechanisms will reassert themselves. But there is also work being done here in Philly — led by Leonard Nakamura that is pretty persuasive — that we may be overstating inflation, even as low as it is. And as a result, we may be understating productivity growth in the economy.

Leonard’s favorite example is the iPhone [and its many features]. We no longer have [to use a separate] GPS [or get a Gameboy]. … It’s all in the phone. … But we don’t know how to capture that shift in the statistics. We can do it hedonically, in retrospect, but in real time for policymakers, we’re having a hard time figuring out how to capture that. That leads me, as a policymaker to say, “Well, given that, let’s just let things play out a little bit. Let’s just see how the data evolves over the next several periods.” [The latest] PCE (Personal Consumption Expenditures) report was weak, so we’ll have to see how this plays out.

I think hitting pause for a little bit on the Fed funds path is appropriate for two reasons. One, you see the data play out. Second, we are unwinding our balance sheet, and that process will start … in October. It’s time to just let that play out. … We want to view [the process] in a way that is as exciting as watching paint dry. We want to lay this out very clearly. The caps are very low. And the best estimate that I’ve seen is that will have an impact the equivalent of a 25 basis-point increase in the fed funds rate over a five-year or four-year period.

We’re not talking about a very large impact, but we don’t know for sure. And I think prudence would call for just letting that happen, see how the markets react. I’ve still penciled in an increase in [the fed funds rate in] December, and three increases for 2018, assuming that inflation comes back. But I do emphasize the words “pencil in.” We’ll just have to see how things evolve.

Knowledge at Wharton: To follow up on the last point you made about the winding down of the $4.5 trillion balance sheet. How long do you think the process will take?

Harker: It will take several years. The question is not so much the timing, but what are we going back to? That’s what you really need to know. And there, if you just do the math — so our currency liability, say, call it $1.5 trillion, and then we have about $400 billion in the Treasury account, plus some other foreign reverse repo, and there are little things here and there. So we get up to some number — call it $2 trillion.

The question is what do we need in order to … execute monetary policy? Well, this gets into a debate about whether you want a floor system or a corridor system, and you need to know that elasticity of the demand for reserves. We’ve been way off on that curve for now quite a while, so we’re not quite sure in this economy what that elasticity is. We’re on the flat end, for sure. And there is some view — and I’m sympathetic to this view — that a floor system makes some sense. The issue is we don’t know when we stop being in a floor system and go into a corridor system. That is when the elasticity increases.

Again, there is some recent research by some of our economists here that said, “You know, that number could be as high as $1 trillion. We just don’t know.” And others will say, “Well, it could be a low as $100 billion.” … You pick some number in the middle, and it would get you to a balance sheet of $2.5-ish trillion, somewhere around there, going forward. But we’ll know it when we get there. We’ll be able to see how the market is reacting. We’ll be able to respond appropriately.

“I’ve … penciled in an increase in December, and three increases for 2018, assuming that inflation comes back.”

Knowledge at Wharton: Looping back now to the topic of fintech, one thing that has struck me in listening to all the wonderful speakers is sometimes it seems like we almost have this tsunami of financial technology innovation going on everywhere. But to what extent do you see it as being revolutionary and truly disruptive? And to what extent is it evolutionary and something that can be integrated into the existing financial system?

Harker: It can be both. It doesn’t have to be one or the other. There are clearly … important innovations happening in the market. That said, financial services have been in the business of innovating with technology for a long, long time. This is not something new to financial services. It feels different now, because it’s not the traditional banking institutions and other institutions that are doing it. There are new entrants into the market. But … I think people are finding out … that [the disruptors may be] very good at innovation to the customer but then [they have] all this compliance [and other] stuff … to worry about. And so it may be better to partner with the institutions that know how to do that, and also get scale.

This is not very different than many other industries. … There are going to be the Amazons of the world that emerge. But then, others will figure out how to adopt this technology or partner with people and be successful. [Also,] the fundamentals don’t change very much. People still want some basic things out of the financial system — access to credit, they want to be able to pay their bills, they want to have safe and secure funds. And they want to have other services that can help them succeed, whether they are a business, a small business, or a consumer. Those fundamentals are still there, it’s just how we deliver them that’s changing. And that will continue to evolve. So I think it can be both.

Knowledge at Wharton: The other big theme that has come up in the past couple of days is what are the key benefits of fintech, and how do you see the trade-off between the benefits and the biggest risks of fintech?

Harker: I enjoy having access to all the mobile platforms and all the ability to do that. That’s the advantage that we’re seeing across the board — and so hopefully we’ll also see … people who are marginalized in the financial system being able to get access to the financial system. For me, that’s more aspirational right now, but it’s really important. And it’s important for the system that we continue to have ubiquity, that everybody can have access and plug into the financial system. In particular, if there are entrants who are coming into the market who can bring the unbanked into the market, help educate them about how to create a sound financial future for themselves, we’re all better off as an economy. The large risk is that many — not all — but many of the new entrants were born in a period where they’ve never seen a deep downside of a business cycle.

Banks generally have been through that. They know how to manage that risk. If there’s a concern I have, it’s that the institutions make sure they have the proper risk procedures and policies to protect themselves when that happens. … We will have that downside of the business cycle [at some point]. And to make sure that institutions are protected, and hence consumers are protected in that process, I think is critical.

“Whether cryptocurrencies and other forms of currency emerge, the basis of that has to be [this:] How do they create that trust?”

Knowledge at Wharton: Do you believe the creation of digital currencies will threaten the ability of central banks to maintain control of money supply?

Harker: I’m not sure yet. At this point, I would be on the side of sort of doubting that for a couple of reasons. One, I agree with the Acting Comptroller [of the Currency Keith Noreika] that we should let the market evolve and see how this technology evolves. … That said, I think some of the fundamentals haven’t changed. For the economy as a whole, the paper that’s in your pocket that we call “money” only has value because we believe it has value, because we believe that the government stands behind it. It’s all a trust issue.

So whether cryptocurrencies and other forms of currency emerge, the basis of that has to be [this:] How do they create that trust? In the case of a central bank-issued digital currency, it is the government standing behind it, the sovereign state standing behind it. You could have other models of that, but they have to play out over time, because again, everything can work in good times. It’s when the bad times hit [that’s a question].

The other issue I am personally concerned about is — whether it’s a government-issued or central bank-issued digital currency or another large player comes in and issues this digital currency — when things really go bad, where do Americans turn? … Well, they’re going to come back to the government. That’s been the history of the country in recent vintage. And so, we’d have to think through that. And again, this is one where we’re not ready for yet. We’re seeing it evolve, but it is something we continue to study, and I continue to study and learn from the experience of others who have done this.

Knowledge at Wharton: To make sure that things don’t go bad, the role of the regulator is very critical. The question is how can you regulate in such a way that you mitigate the risks that you are talking about? And even more fundamentally, how do you regulate an algorithm?

Harker: I don’t know yet. I think the answer is we have to continue to study this. As these technologies evolve and the innovation evolves in the market, we need to continue to work and study how this is evolving. The issue of regulating algorithms is interesting. One of my colleagues, a computer scientist, is doing some interesting work on fairness in machine-learning algorithms. … But how do you plug into that some sense of fairness, whether it’s required by a regulatory constraint, or you believe that’s the right way of doing business? That’s actually a lot harder than it sounds, apparently. It’s a very difficult thing to do. Before we even think about how to regulate an algorithm, how would you even build an algorithm that would have that sense of fairness in it? That’s a fairly deep technical question.

Knowledge at Wharton: The fairness issue is especially interesting because of how important financial inclusion is, as a dimension of fintech. How do you see the relationship between fintech and financial inclusion?

Harker: Done right and done well, it can have a tremendous impact. But we also have to recognize that we are human beings. We are flawed human beings, and there are players in the market who are behaving in different ways. … I hear these stories constantly, of small businesses in particular, getting into a situation where they didn’t quite know what they signed up for. And then they walk into their community bank and say, “I’ve got to get out of this deal. It’s killing me. I signed up online for this thing, and I thought it sounded good, but I didn’t really understand it, I guess. And now I’m really in trouble.”

There are good actors and bad actors in any industry. Financial services is not very different than anything else. And I think [protecting against] that is the role of the regulators, and not just at the national level. But even if there was no national OCC charter for a fintech company, there are still state regulatory requirements with respect to fairness to the consumer and so forth. There’s always going to be some form of regulation.

My message to the fintech companies is get involved with the regulators early and often and start forming [a framework that] makes sense to protect the people who want to do it right. I think that can protect … this sphere of innovation that you all need and we want as an economy. But if there are too many bad actors in there, that can stifle that innovation. … We know how regulation goes. … We tend to overreact to a crisis. When something bad happens, the pendulum swings too far. It’s better to be in the conversation right now about what those regulations look like, whether it’s at the national level or at the state level.

“My message to the fintech companies is get involved with the regulators early and often.”

Knowledge at Wharton: Consumers especially need to be protected especially in the wake of what has happened with Equifax. What is your view about the cybersecurity environment and how should consumers be protected?

Harker: One of the big challenges for consumers [is they need] to figure out how to protect themselves. And increasingly, that’s not just [about safeguarding their] financial data, but it’s health data, [their identities,] you can go down the list. … It creates challenges for the consumer and for the financial institutions. … Given how much data and information is now out there about all of us, how do we change the system? How do we innovate to protect my identity — and not just my financial identity?

… I actually worry about this a lot, because if we don’t do this, a lot of the innovation that you are all fostering can be really stifled when customers say, “I’m going back to microfiche and coin, because I’ve been hacked so many times, and I’ve lost so much money, and it’s such a hassle to keep trying to clean up my credit or my health records or my false tax returns that somebody put into the IRS, that I’m just dropping out of this.” That would be a dangerous step for the industry. … Some innovative players have to come in with different ways of thinking about how to protect our identities.

Knowledge at Wharton: Looking to the future, maybe two or three years out, what is your most optimistic and your most pessimistic scenario about fintech? And what is your most realistic scenario?

Harker: The optimistic one is … that fintech not only reaches consumers with deep pockets, but also starts to bring the marginalized consumer into the market, as part of [the effort on] financial inclusion. That’s critically important for the economy. The negative is … there’s continued hacking or there are regulatory issues that stop the innovation in this industry.

If we don’t get this right, and if the consumer doesn’t learn to trust these innovations, [these advances] simply will not happen. It’s all a trust game. The realistic scenario is that there are going to be people who do it well. There are going to be people who don’t do it well, people who succeed, people who don’t. That’s the market. We just need to make sure that we, as regulators, continue to make the playing field level and not get in the way to the extent that we stifle this innovation.

Knowledge at Wharton: What is the biggest prevailing myth about fintech? And if you could say one thing today to dispel it, what would that be?

Harker: Not all technology is equal. With some of this stuff, you’re putting yourself out there at tremendous risk, because it’s not very secure. [My millennial children] don’t think about that. That’s one of the big myths that I know you are all concerned about, and we’re concerned about, as well.