Oct 24, 2022

Peter Ganong on the Dynamism and Resiliency of the US Economy

By reducing housing costs and promoting smart welfare policy, the US can take concrete steps to bolster dynamism and resiliency within the macroeconomy.
David Beckworth Senior Research Fellow , Peter Ganong

Hosted by David Beckworth of the Mercatus Center, Macro Musings is a new podcast which pulls back the curtain on the important macroeconomic issues of the past, present, and future.

Peter Ganong is an associate professor at the Harris School of Public Policy at the University of Chicago. He joins David on Macro Musings to talk about his work on the dynamism and resiliency of the US economy. Peter and David also discuss the income convergence story in the US, how to address increased housing costs, the economic effects of pandemic response measures, and a lot more.

Read the full episode transcript:

Note: While transcripts are lightly edited, they are not rigorously proofed for accuracy. If you notice an error, please reach out to [email protected].

David Beckworth: Peter, welcome to the show.

Peter Ganong: Thanks so much for having me.

Beckworth: It's great to have you on and like many of my recent guests, I think I've interacted with you on Twitter. Of course, you've written great papers, some cutting edge papers, and I'm excited to get into them because they do deal with the dynamism of the US economy, convergence in the US as well as some of the lessons we've learned about resiliency, what works, what doesn't work during the pandemic and the Great Recession. So I'm excited to get to those, Peter. But before we do that, tell the listeners a little bit about yourself, your journey into economics, what you're doing at the University of Chicago and all the other good stuff.

Ganong: So I think I would briefly just highlight two moments from my journey. The first is that I started out of college planning to probably be a lawyer, maybe a rabbi, and I had to take this required economics class and I wasn't excited about it and I found a substitute. And the substitute class was a class on applied price theory where Ed Glaeser taught, basically, Gary Becker price theory to undergraduates. That class was absolutely transformative for me because it showed me and a couple hundred other students that the tools of price theory are fantastic for understanding the world and identifying solutions to problems. I ended up working as an undergraduate research assistant for Ed Glaeser and then several of the papers, but in particular the regional income convergence paper that we're going to talk about later, really came quite directly out of being brought into the price theory tradition at age 19 by Ed Glaeser.

Ganong: The second journey moment that I would identify for you is a moment when I was physically very cold. So I graduated college and I took a job working for Christina Romer in economic policy in the Council of Economic Advisors in 2009, right as the Great Recession was starting, moved to DC, wanted to live somewhere relatively close to downtown with the salary that I had, which was quite good, but also not that much compared to the housing prices. So what we ended up doing was me and another friend from college rented a one bedroom apartment with a closed in porch. I know you used to live in DC, David, so you've seen these closed in porches. Sometimes they're heated and sometimes the porches are unheated. So we happened to get an unheated porch. And for various reasons, I don't quite remember the allocation rule, I ended up being the one who was living on the unheated porch and my roommate lived in the actual bedroom. Now, the good news about this one bedroom apartment is there was direct bathroom access without having to go through the roommate's bedroom. The bad news is that it was through a window. So if I wanted to go to the bathroom in the middle of the night, I had to climb through a window.

Beckworth: Wow.

Ganong: It got really cold and my parents got me a heated blanket because they thought that I should have heating. And my girlfriend at the time, now my wife, refused to ever stay there because she's like, "Why would I stay on an unheated porch?" So you'd think that, "Oh, maybe she's living in better housing." No. So she was living in a two bedroom apartment in DuPont Circle where they took the living room and put up two more walls. So there were four women living in a two-bedroom apartment. So quite literally doubled up. And at some point if I'm spending a lot on housing and living in an unheated porch and my girlfriend at the time is in a four-bedroom apartment that was supposed to be a two-bedroom apartment… We have relatively good jobs. Housing is really expensive for everyone who's living in DC.

Ganong: When you're a 22-year-old, that's not a big deal. When you're trying to raise a family, that's a big deal. When you don't have the salary of a technocrat working for the federal government, that's a really big deal. So you need to be careful about the role of personal experiences and research because you don't always want to over extrapolate from your own experience. But in this case, that unheated porch and thinking about if housing prices are causing me to make these choices, they must also cause a lot of disruption in DC for other people, especially for families, was the starting point for a lot of my early research.

Beckworth: That is very interesting, Peter. So you both had a formative intellectual experience and then a practical living experience come together and make you the economist that you are. And you had a couple big names thrown in there as well, Ed Glaeser, Christina Romer. So very fascinating. I was delighted to hear you talk about price theory and you did so with enthusiasm. And hey, you're back at the University of Chicago, so I think the circle is complete now. You're back there doing research using price theory. And the story about housing in DC is very, I think, relatable to many people, many listeners.  And so with that said, let's get into your first piece I'm going to cover, and that's an well-known article of yours, you co-authored. And the title is, *Why Has Regional Income Convergence in the US Declined?* from the Journal of Urban Economics. It has to do with housing as part of the story, but walk us through the basic facts that motivate this story about income convergence and what's changed about it.

The Basic Facts of Income Convergence in the US

Ganong: I coauthored this paper with Daniel Shoag who's at Case Western and one of our advisors, Robert Barro has a very famous paper about regional convergence of states where he showed that the poor states were catching up to the rich states. This a Brookings paper. I think it roughly covers the period of 1890 to 1980. And Danny's joke about this is that this is the cleanest regression you'll ever see in macroeconomics. Pretty much to a state, the poorest states grew fastest and the richest states grew slowest for any 20-year period pretty much that you pick between 1880 and 1980. So Delaware, Rhode Island, Connecticut, Massachusetts grew slowly, [and] Mississippi, Arkansas, Alabama grew very quickly over that period. And the natural thing you do in grad school after you replicate a result is you say, "Well, let's look again." So we looked again and we thought we'd made a mistake because the iron law was broken and there was almost, it depends exactly on how you do the specification, no relationship between the income level of a state and how much it grew from 1980 to 2010.

Ganong: And so then that for us begged the question, what was driving convergence before? The Barro model focuses on capital mobility as a driver of convergence. We saw no reason why capital mobility would've changed around 1980. And so that led us to focus on a different factor of production, which is labor mobility. So there's also a consistent relationship... It's not quite as strong as the income convergence, but pretty strong, where the richest states in the union from 1880 grew the fastest. And rich here is defined on a per capita basis. So high income per capita, places like California, Nevada, Connecticut are places that grew a lot in terms of their population.

Ganong: And then that's matched by no growth or even outflows from places like Mississippi, Arkansas, Alabama. So people are moving on net from low income places to high income places and that meant that migration was an engine of economic mobility, because if you're in Mississippi and you want to raise your income you can always have the option to move and you can move to a place where incomes are substantially higher. So that relationship broke at exactly the same time that the convergence relationship broke. Poor places were catching up to rich places as people moved from poor to rich places. On net, people stopped moving from poor places to rich places at the same time that the poor and the rich places stopped catching up with each other. That's just a time series. It's not causal, but they could be related and that's sort of the possibility that we try to explore in this paper.

Poor places were catching up to rich places as people moved from poor to rich places. On net, people stopped moving from poor places to rich places at the same time that the poor and the rich places stopped catching up with each other. That's just a time series. It's not causal, but they could be related.

Share this

Beckworth: So just to recap, you retell the story about convergence across the US. So poorer regions are catching up with richer regions in the US economy and at the same time mobility patterns of where people move to, match that story, they're moving from poor parts of the country to richer parts. So it's a possible explanation. Is that fair?

Ganong: Yep. We say there's a time period when there was convergence. There's a time period of migration, they both go away and that's like section one of the paper. Section two of the paper asks why and studies the role of, this is going to be a mouthful, but it's important to get all the words in, skill specific migration and housing. So let me tell you a little bit about skill specific migration, a little bit about housing and feel free to just keep pushing with more questions.

Beckworth: Sure.

Skill-Specific Migration and Housing

Ganong: So in the time period when people were moving from poor states to rich states, that was a fact that was true for low skill workers and high skill workers. However, from 1980 onward, roughly something different happens in the following sense. High skill workers where high skill here is defined as having a high school education, if you're looking at, say, 1940, defined as having a college degree. If you're looking at say 2000, high skill workers kept on moving to those high income places and low skill workers stopped moving to those high income places.

Ganong: So to go back to the narrative that I talked about before and make it very concrete, I was living on that unheated porch in 2009 and I was a high skill mover moving into a high opportunity area in Washington DC. But the more important story wasn't that I was paying a lot of rent, it's the missing worker who never came who had less education, who, had this been 1940, she would've had found a lot of economic opportunity in DC. Come 2009 when I moved there, she didn't come on net. And so I didn't even get to meet her. And I said, "Oh, I was worried about the people who were facing high housing prices there." I was, I am, but I didn't even get to meet this person because she didn't come because just didn't make any economic sense for her to do so according to our calculations.

Beckworth: In your paper you used the illustration of a janitor versus a lawyer. So maybe walk us through that example.

Ganong: Great. So let's imagine someone who's a janitor or a lawyer living in the deep south and making a migration decision. Do I want to move to the tri-state area, New York, New Jersey, Connecticut? Wages for lawyers in 2010, if you just look at wage income, are 46% higher. Wages for janitors are 28% higher. So if you only look at wage income, it seems like the both people should make this move. Let me clarify, we do one calculation in 1960 and one calculation in 2010. So in 1960, wages in the tri-state area, New York, New Jersey, Connecticut are 38% higher for lawyers and 84% higher for janitors than they are in a group of states that we describe as the deep south. After you take into account housing costs, the premia are almost exactly the same, 39% and 70%. Now, let's fast forward 50 years to, roughly, when I was living in DC. The wages of lawyers and janitors are 46% higher, for the lawyers, 28% higher. The fact that the janitors premium is already lower, tells you something about changes in technology and the national wage structure. That's important. I don't want you to think it's not important, but the key point of what we found was the housing prices.

Ganong: So after you adjust for housing prices, the wages of lawyers are 39% higher in New York than in the deep south. So the lawyer should still make that move. The wages of the janitor are 7% lower in New York after taking into account housing prices. So it's quite literally a reversal. It's not worth it anymore to make that move… if your occupation is a janitor and you'd get the average wage in either place, you pay the average housing cost in either place. It's an erosion of economic opportunity. There was one way that people could exert effort and sweat to move from a place with a rough economy to a better economy. And that way is just not available in the same today in the US and the way that it was 50 years ago.

It's not worth it anymore to make that move… if your occupation is a janitor and you'd get the average wage in either place, you pay the average housing cost in either place. It's an erosion of economic opportunity. There was one way that people could exert effort and sweat to move from a place with a rough economy to a better economy. And that way is just not available in the same today in the US and the way that it was 50 years ago.

Share this

Beckworth: And what you're saying is that the key part of the story is the change in housing costs, is that right?

Ganong: So I want to be clear that I think both changes in the wage structure and changes in housing costs are very important. However, when we wrote our paper, there were a number of excellent papers already about changes in the wage structure. And I think what the point of our paper, relative to those papers, was trying to bring housing to the table and say, "Look, you can't get there all the way just with thinking about changes in wages, you really need to think about housing markets as well."

Beckworth: So there's been a lot of work done on the decline in labor mobility in the US and even some other papers that aren't directly related, but speak to this. So for example, the famous China Shock paper where they talk about these communities hit hard by China entering the World Trade Organization and engaging in trade with the US. And for me, one of the big takeaways is not the shock itself, it’s that people didn't leave those communities, people that were hit hard didn't migrate to some other part of the country where there might be opportunity. And your findings try to explain some of that, maybe most of that through the cost of housing. Now, you're stressing, there are other things as well going on. There's automation, other things affecting the demand for their services in a big city. But housing is a big part of the story. So walk us through what's happening to housing that's causing this change to take place.

Housing Costs and Labor Mobility

Ganong: What happened over time is that the relationship between housing prices and income got even steeper. What that means is that higher housing prices in those high income places are going to erode the benefits for migration. So why are housing prices so much higher in these high income places? The channel that we focus on is regulation. We have, admittedly, a very coarse measure of regulation, which is we look at how many lawsuits there are that mention words like ‘land use’ by state. You can see that the number of lawsuits that mention land use by state goes up a lot over time and it goes up most in the places that more reliable surveys like the very famous [inaudible] land use survey identify as the highly regulated places. And so we understand there's having been a supply shift where... Let me give you another very concrete example. So my dad grew up in Albany, California. My grandma was an architect.

Ganong: She bought an empty plot of land and designed and then built her own house on Albany Hill in California in the East Bay. At the point when she moved there, it was really easy to buy an empty plot of land and put up a house. It's extraordinarily difficult now to put up a house in Albany, California, Berkeley, California, frankly anywhere in the Bay Area. So the idea that you could show up somewhere and build a house at a reasonable cost at a place that was offering economic opportunity as the Bay Area did for my grandparents when they moved there, that has more or less stopped... I mean, we can talk about the YIMBYs later, but as of 2010 when we wrote this paper, there was no option to just show up, buy an empty plot of land, and easily build a house the way my grandma did and the house my dad grew up in.

Beckworth: So housing supply is not responding to the increase in demand that would be there were these people migrating to the city. So they see that housing is expensive because supply is not rising proportional to the amount of migration.

Ganong: Sure. So let's just be a little more precise. My view is that the demand was always there and what happened was that the supply conditions changed. So we used to meet the growing demand to live in these productive places by building more housing. When we stop building more housing then it's going to get capitalized into prices and many people will get pushed out once it's capitalized into prices. So same demand conditions, change in supply conditions is my summary of how I think about it.

We used to meet the growing demand to live in these productive places by building more housing. When we stop building more housing then it's going to get capitalized into prices and many people will get pushed out.

Share this

Beckworth: We've had several guests on the show that explain and go behind why supply has become so constrained.

Ganong: Those are some of my favorite episodes.

Beckworth: Oh, great. So we won't spend too much time on that, but I do want to ask this question related to your paper, maybe an extension of the paper, where we are today. So one of the big things that we've been through is the pandemic and we'll get to your research on that in a bit, but one of the big shocks that have come out of that is there's been a lot of people moving away from cities now. So housing has gone up in other places of the country where they've migrated to. There's also more work from home. How do you see these dynamics playing out and affecting this pattern that you guys found in your paper?

Ganong: I don't know, but let me speculate with one piece of data from a neat paper by John Mondragon who's at the San Francisco Fed and Johannes Wieland, who's at UCSD. So they have a paper where they study the change in housing prices across different areas in the US and they make the point that housing prices rose most in the areas that had the most remote work going on pre-COVID. So they measure remote work 2015 to 2019. Now, we can link to this paper in the show notes. They show that the places that had the most remote work then are also the places that had the biggest housing price pop during the worst of the pandemic.

Ganong: So to me, although in principle you could have imagined that remote work would be undoing some of the expensive coastal elite cities that, just in principle it could have. In practice, what I take away from Mondragon and Wieland paper is that that's not what happened. Instead, what we had is more demand for space, more people moving to the suburbs, greater overall demand for housing. Those places are not elastic housing supply, so prices go up even more. It's possible we really don't understand how remote work is going to work, what it's going to look like in five years from now. But in principle, the immediate impact could have been to undermine some of these trends. I don't see any evidence. If anything I'd say it exacerbated some of these trends.

Beckworth: So Peter, what do you think about the future of housing in the United States? There's been a lot of discussion on this issue. You've mentioned some of the previous guests that were on the show and there's a lot of things happening both on the left and the right. I think there's this coordinated effort to increase housing supply. Are you hopeful for the future on this front? Can we fix housing? I mean, housing has been identified as a key problem in US economic dynamism. I mean, there's that famous paper by Enrico Moretti showing that maybe 9% of GDP is lost because of it. You think of productivity gains. You think of just all the lost opportunity. We could be such a more dynamic place if we had this problem fixed. Are you hopeful that one day we will be there?

The Future of Housing in the US

Ganong: I always hope that we'll be there. I think it's really hard and I just want to point out two types of frictions that maybe are not fully appreciated by listeners, although maybe they are. So first, I think it's the campaign to convince DC think tank nerds and fellow travelers of DC think tank nerds like myself that housing regulations are a problem, has been wildly successful. So you can get executive orders from the Obama Administration and the Trump Administration and similar noises from the Biden Administration, that housing supply is a real problem. But that hasn't turned yet into substantial change on the ground. First, there's a point about federal government versus state and local. But then second, I think that there's very real political economy problems… to do more UChicago, Let's talk about Coase. How are you going to pay off the losers from relaxing housing regulations?

Ganong: There's a generational challenge and there's a challenge of who are the voters versus, we talked before about how I imagined in a childcare worker who was not able to move to DC because of the housing prices. She didn't move to DC. She also therefore doesn't get a vote in DC about increasing the supply of construction there. So I think the political problems are very, very real. And convincing think tank nerds, it's always insufficient, but it's especially insufficient here because the political economy problems are just so challenging. So I feel some success that Ganong and Shoag and I think there's many other excellent books and papers, have persuaded the nerds and I think now it's time to get to more work or to roll up our sleeves and get to work on some of the political economy problems. And having identified the economic problems is just completely insufficient here.

Beckworth: So Peter, this is an important issue and if I had to, I would put my top two concerns about the US economy over the long run, housing and the decline in the population growth rate. Population decline has been happening. It's been increased really since the Great Recession. The rate keeps dropping and we know fertility is not the answer. It's hard to change that. Immigration is the answer, but we're having a hard time making any progress on that front. So population matters to long run economic growth because both, it's the idea generation, kind of the endogenous growth stories of Paul Romer and stuff. If you want to innovate, be creative, we got to have more people, more ideas. And if you want to be also a force for good in the world, Matt Yglesias has his book One Billion Americans. And he lists a bunch of arguments. So one of the things he notes that I think is very interesting and important is if you want to be a force for good in the world, you got to have the biggest market in the world. Movies, corporations, they cater to the biggest market and we can set a tone for the world as opposed to say China being the biggest market and it dictating the terms of what kind of messages and branding go out-

Ganong: From a fixed cost of innovation, you just got to be the biggest.

Beckworth: You've got to be the biggest. And then on top of that, just you got to have, I think, the bodies, the ideas to support economic growth. But with that said, let's move from the long run growth and dynamism discussion to short term and questions of resiliency because you've written on that too. You've written both about the Great Recession as well as the pandemic. I want to briefly spend some time on the Great Recession because you have several papers that you wrote about the mortgage issue, household wealth versus liquidity. So maybe give us a bird's eye view of what you covered and found regarding the Great Recession in housing.

Housing and the Great Recession

Ganong: Great. So just very briefly, when I was in that unheated apartment in DC, Pascal Noel who's my longtime co-author, was also in DC and an apartment hopefully with heat. And he was in the trenches working on housing policy in the White House during the Great Recession. And I think most people would agree at this point that the Great Recession response to the decline in housing prices and the wave of delinquencies is generally considered to be a failure. What we've done since we went to graduate school is try to unpack some of the pieces of that failure and try to understand, where did it come from, how can we do things differently? We've tried to unpack and understand where mortgage default comes from and what do you do about it. So first, let's talk about where mortgage default comes from, then we'll talk for one more second about what do you do about it.

Ganong: So mortgage default almost always coincides with a decline in household incomes. And one thing that we did is we took bank account data and mortgage servicing data both from JP Morgan Chase. So people have Chase bank accounts and mortgages serviced by JP Morgan Chase. And we asked what happens to your income for people who default on their mortgage? On average, it falls about 20% in the year leading up to mortgage default. Now, that's just a fact for context setting. Here's the remarkable thing. You take borrowers in the Great Recession who were underwater on their mortgages. So that means that the debt they owed on their house was more than the mark to market value of their mortgage. The conventional wisdom at that time and frankly in the academic literature for a long time afterward said, "We're going to be strategically defaulting. They're going to walk away from their houses."

Ganong: One corny name for this is jingle mail. Mail in your keys because your home's not worth it anymore. There was an idea that a huge number of people who were behind on their mortgage debt or underwater on their mortgage debt were going to walk away from their homes. And what we found is that, actually, the people who fell behind on their mortgage payments who were underwater had the exact same decline in income as the people who fell behind on mortgage payments who were above water, which is to say that pretty much everyone, maybe 94%, 97% of people who defaulted on their mortgages in the Great Recession had a liquidity problem. Their income had just gone down prior to when they defaulted. And it's inconsistent with the view that there were a lot of strategic defaults of the sort of jingle mail type that I talked about before. So that's part one of the research. Where does default come from?

Beckworth: Can you define liquidity problems? What is a liquidity challenge for a household?

Ganong: What I mean when I say a liquidity challenge for a household is, you need to make your mortgage payment and you don't have the money this month or you don't have the money for the next few months, but you want to stay in your home and not just you want, in the abstract, to stay in your home. Almost everyone in the abstract wants to stay in their home. But quite literally, if you had a slightly different financial situation where you were able to borrow money that you're not able to borrow, you would've kept making your mortgage payments. Or if there was a little bit of flexibility in your mortgage payments today, it would become possible for you to make some catch up payments down the road, meaning you want to stay there and you're willing in a monetary sense to put your money where your mouth is about staying there, but you can't put your money there today. You need the mortgage lender to be a little bit patient, which is obviously a challenging thing for any financial institution to orchestrate.

Beckworth: Okay. Go into your second point that you're going to make.

Ganong: Great. So paper one, where does default come from? Paper two, what do we do about it? And in this paper what we did was we went into the guts of some of the different mortgage modification programs that the government ran and also that the private sector ran and what we found was that changes to mortgages which reduced the debt outstanding, so debt forgiveness, had no impact at all on mortgage default whereas restructuring of mortgages which provided liquidity to borrowers at no net cost to the lender, because the lender got paid back down the road, generated a dramatic reduction in defaults. And that's totally consistent with what I told you about the income series. So it's two totally different methods but a similar conclusion from both studies, which is that default problems arise, it appears largely because of liquidity and liquidity provision. Now, that we've defined the term, I can use my new jargon, generates substantial reductions in default. I don't know if we'll get to this later in the podcast, but this has very, very clear lessons which have played out in the context of the recent economic crisis from the COVID-19 pandemic.

What we found was that changes to mortgages which reduced the debt outstanding, so debt forgiveness, had no impact at all on mortgage default whereas restructuring of mortgages which provided liquidity to borrowers at no net cost to the lender, because the lender got paid back down the road, generated a dramatic reduction in defaults.

Share this

Beckworth: Yeah, it's a great finding and I should mention the name of the paper so that our listeners will know and provide links to it in the show notes. But the first paper is, *Why Do Borrowers Default on Mortgages?* A QJE article from 2022. And then the second one is an AER article titled *Liquidity Versus Wealth and Household Debt Obligations: Evidence from Housing Policy in a Great Recession.* So two great papers and the key takeaway at least for me is liquidity really matters here. And if you can do something to make it work, more flexibility for households, they can weather some of these storms better. What you just said, we saw that in the pandemic. In fact, I was just looking at a chart from the New York Fed, it's a household credit report that comes out every month, maybe every quarter. I forget exactly. But it shows the delinquencies and it's striking. You see this massive spike in delinquencies 2007, 2009 and there's practically nothing the past two years. And yet the potential is certainly there, right? One of the sharpest recessions on record, high unemployment.

Beckworth: So the policy response was very different and we'll come back to that in a minute, how things were different. But you mentioned in making these papers that you used JP Morgan Chase data. Let's use that as a segue into the next set of papers I want to talk to you about and that's your pandemic papers. You were able to access this data through JP Morgan Chase and I understand your efforts helped create a new database that's available to the public. Tell us about that story? How did you even know to access this data in the first place and how'd you get access to it and what's the results of all this activity?

Creating a New Economic Database

Ganong: So in graduate school, I mentioned before, my co-author Pascal Noel in the heated apartment. We decided that we wanted to study household economic distress. We went to every major bank in the US as well as several credit unions, submitted many proposals. They all said no. Not surprising. I mean, who's going to say yes to a graduate student? And then we basically got lucky, which is that someone that we'd worked with in the past, her name is Diana Farrell, was setting up a think tank and she said, "Are you interested in helping to set up this think tank? It's going to be a think tank situated within JP Morgan Chase." So we started working with her, actually, three other amazing Chase coauthors who joined in time where Fiona Greig, Chris Wheat, and Daniel Sullivan. And now there's also more co-authors on the Chicago side as well, Joseph Vavra, Damon Jones. At this point, it's a huge team effort, but it's an amazing team to be a part of.

Beckworth: And you collect data from checking accounts, transactions? Explain to us this database that you've built.

Ganong: So I'll quote Diana here. So Diana's mantra going in, and I think it still governs how the JP Morgan Chase Institute, to be clear is the name of the think tank, runs, is study the world, not the bank. So the idea is that there are some private sector research initiatives which are designed to say, "Look at how great our business is. Look at how much our business is helping people." That's not the point of this initiative. The point of this initiative is, "Oh by the way, as part of doing business with 20 plus million checking accounts and many more credit cards and mortgages than that, that we can learn a bunch of things about what's happening in the real world in households’ economic lives.” So it's deposit data, it's mortgage servicing data. There's also credit card data. We occasionally bring in external data. So we've done some work trying to understand racial inequality and consumption smoothing and so that requires some external data. But fundamentally it's not about studying the bank, it's studying the world through the lens of the records of one bank.

Beckworth: But you're able to look at all these transactions from households based on different characteristics. Is that right?

Ganong: That's exactly right. So just to give you two concrete examples. We can see if you have a mortgage in distress and say, "Well, what happened to your household's income in the 12 months leading up to that?" We talked about that paper. We can also see households get unemployment insurance payments. And that's really important because, originally in my dissertation we studied how households choose their spending behavior during unemployment. And then when the pandemic hit, we all looked at each other on Zoom, so proverbially looking at each other and said, "Oh my gosh, we've had a big expansion of the safety net," particularly an expansion of the unemployment insurance system to a record size. Let's understand how that's affecting households, how it's affecting their spending, how it's affecting their job search. And that's how I spent the last two years of my life.

Beckworth: So tell us the name of this database, make sure that the listeners get it straight because I don't think I have the correct name. What is it?

Ganong: So the best way to say it is the JP Morgan Chase Institute has assembled a set of databases or data assets and there are a number of researchers who use this data. So there's another researcher at Purdue, Lindsay Relihan, another at Harvard Business School, Olivia Kim. And I don't think the database has a name independent of JP Morgan Chase Institute.

Beckworth: Okay. We'll provide a link to the listeners. They can check it out. So you use this data both for those papers we just discussed that covered the Great Recession, but also the pandemic. So let's move into the pandemic and let's talk about what you have found. And maybe just for our listeners, remind them what were the big policy innovations that were introduced during the pandemic to help get us through it.

An Overview of Pandemic Policy

Ganong: So there were several different new policies that we tried. Frankly, I think it's probably like the most rapid policy innovation that we've seen certainly in my lifetime and I'd have to do more historical research to go back before my lifetime. We'll talk a little bit later about widespread mortgage forbearance. It was really remarkable. First, let's talk a little bit about unemployment insurance. So unemployment insurance was expanded in three ways, one very similar to prior recessions. We made benefits last for longer.

Ganong: Two novel expansions of unemployment insurance, which I think are really interesting and we're still figuring out the lessons from. One expansion was that there was a supplement to weekly unemployment insurance benefits that sometimes was as much as $600 a week, which came close to tripling the typical unemployment insurance payment. Sometimes it was $300 a week, which came close to doubling the total unemployment insurance benefit. So that's about the level. And then there was a sea change in terms of who was in. So the states supported by funded and supported by the feds relatively quickly set up a whole new program to provide payments to people who are not eligible for regular unemployment insurance. It was called the Pandemic Unemployment Assistance Program. We'll, I guess, dive in on each of those now in turn, but that's the overview is longer, higher, broader.

Beckworth: Okay. And there were also stimulus checks, child tax credits. I mean, you could throw in the PPP program. I mean, people may not have felt that directly, but the firms kept them employed because they got that, right? And there was a number of forbearance opportunities. So there were a lot of things that happened. Do you want to shed any more light on those before we get into your work?

Ganong: The one thing I want to say is that there are times where we took existing programs and scaled them up quite effectively. And there are times where we tried to build new programs from the ground up. So you referenced the Paycheck Protection Program, which I think, the consensus in the literature seems to be that it was a boondoggle, meaning that it did not preserve many jobs and generally paid benefits out to relatively higher income households that weren't in need of income stabilization. Another way to say this is, there were many layoffs, but not many of those layoffs were prevented by the Paycheck Protection Program. So to me that's an example of the foibles of standing up a program in real time. So most European countries had a very different labor market response. So most European countries paid firms to keep their workers employed and workers got the same paycheck as before.

Ganong: We tried to do that through the Paycheck Protection Program and largely failed in part because we didn't really have the government infrastructure that those European countries had. And then instead, because we couldn't really keep people paid through their firms, we instead expanded the unemployment insurance system in several ways. And so the whole US response is shaped by under investment in the public institutions relating to the labor market. I think later we're probably going to talk a little bit about the ARP, the American Rescue Plan and inflation. And when you think about that set of issues, it's really important to be thinking about how did we get here and why couldn't we do something more targeted? And the Paycheck Protection Program stands out to me and is an example of the costs of under investment in labor market infrastructure.

Because we couldn't really keep people paid through their firms, we instead expanded the unemployment insurance system in several ways. And so the whole US response is shaped by under investment in the public institutions relating to the labor market...And the Paycheck Protection Program stands out to me and is an example of the costs of under investment in labor market infrastructure.

Share this

Beckworth: That is a great point. If we don't have these institutions in place during normal times, it's hard to think of how we could effectively design them in a panic, in the midst of a crisis. And one of the things I remember very vividly from this experience is just the lack of infrastructure. Even unemployment checks, a lot of people had a hard time getting them because the computer systems that the states were using were outdated. The infrastructure for PPP, even they went through banks. It was the only thing we had in place. So lots of infrastructure questions and I think it's a great point. We need to be thinking about these issues well in advance of a crisis, but let's move on from that and get into your work. So you have this huge NBER working paper titled, *Spending and Job-Finding Impacts of Expanded Unemployment Benefits: Evidence from the Administrative Micro Data.* So you dig into the data and you find both implications for spending and employment. But walk us through what you do in that paper and what are the big takeaways from it?

The Spending and Job-Finding Impacts of Expanded Unemployment Benefits

Ganong: The amazing thing about the bank account data is the usual way that we understand how government programs work is we ask how much money went out the door and maybe we run a survey of a few of the recipients. Here, we can literally trace the money from the states into people's bank accounts and we were interested in ways that this expansion of unemployment benefits affected people's lives. We were interested in two of those ways that we could see through the lens of the bank account data after we see the real money hit. So the first is the spending side. So what we do is we just trace out what happens to spending for unemployed households compared to employed households. And we saw something that I've never seen before in my research life, no other researcher has ever seen, and I think it's possible we'll never see again, which is that the spending of unemployed households during the worst of the pandemic went up. So normal time spending of unemployed households goes down. During the pandemic, I'm willing to bet, David, that your spending went down. However, on average the spending of unemployed households actually went up.

Beckworth: Interesting.

Ganong: Why did it go up? It went up by, depending on the month and depending on the measure, between 10 and 20%. This is in our view, largely related to the large supplements to unemployment benefits, the $600 a week that was being paid out during that time. So that money both was sufficient to raise spending immediately. In addition, it was such a big transfer, it also raised bank account balances persistently for unemployed households. It's not a big part of the story for everyone because bank account balances also went up for employed households. But the unemployed households went up even more because we basically pushed out so much money in such a short period of time.

We saw something that I've never seen before in my research life, no other researcher has ever seen, and I think it's possible we'll never see again, which is that the spending of unemployed households during the worst of the pandemic went up.

Share this

Ganong: So the second question we ask is, let's follow the money and ask about people getting jobs. To what extent are people less likely to take jobs when they're getting generous unemployment benefits, much more generous than usual? And the answer is, if you're null hypothesis is that no one's behavior is going to be affected, that's clearly rejected. There's a statistically significant effect that the more generous benefits reduce job finding and the magnitude of the effect is smaller. The extent to which the benefits reduce job finding appears to be smaller than any of the other studies that… I don't want to say in any other study ever, but then any other study of a recent literature review that was done right before the pandemic. So people did take jobs less often when the supplements were being paid out, but the magnitudes are really, really small, meaning the work discouragement effects were much smaller than the effects that we see during normal times.

Ganong: So I want to be careful. What I'm not saying is that unemployment benefits have permanently changed in their nature. I'm saying the pandemic was a special time and the policies were special. So the special time will never, I hope, never be able to resurrect again. And that's one of the reasons why we think the disincentive was small. But the policies also had a different flavor. In particular, the supplements were temporary. So when they were authorized, they were authorized not to last forever, but just to last for at whatever was our best guess or our best hope of when the critical period would be over. And so one lesson we take away is that temporary supplements are a particularly powerful tool for macro stabilization, even in a normal recession without a pandemic. Let me try to explain a little bit why, and let's think about this… the limiting case of a temporary supplement, David, is a severance payment, right?

Ganong: Because you just get a severance payment for one week and then you continue on. So a severance payment is going to have a small or no disincentive effect depending on what the model you use is. But it's still going to help keep households whole and it's also going to have a benefit in a normal recession of stimulating aggregate demand. In the paper, we take our empirical estimates and combine them with a model and we ask a very simple question. Suppose you're a policymaker who didn't care at all, intrinsically, about unemployed households, all you wanted to do was manage aggregate demand and support aggregate demand [inaudible]… Now, I'm going to call this and you can't see if I'm doing air quotes, but a “traditional recession.”

Ganong: So in that traditional recession you would pay out about $2,000 in severance to every unemployed worker before you even did a single dollar in stimulus checks. I'm flagging this both for the number, it's an indicator that temporary supplements to the unemployed are potentially a useful macro tool for future recessions, but it's also a reminder about the availability of alternatives. So I think we're going to talk later a little about the American Rescue Plan and some of the strengths and weaknesses of it. But one lesson that we take away from the unemployment supplements is that it's possible to do a bunch of targeted policies building on the experimentation that we saw during the COVID-19 pandemic.

Beckworth: So something else you bring out in your paper is a lesson about the marginal propensity to consume for certain households. What did you find there and what are the takeaways?

Ganong: So we studied the marginal propensity to consume, or in layman's terms, I would say the spending response of unemployed households as opposed to the average household in the economy. We found that unemployed households tend to spend money at a higher rate, have a higher marginal propensity to consume than the average household. Most of those households are employed at any point in time. You might be thinking, "Wow, this guy's like a real dummy. He found that unemployed households are more willing to spend." Isn't that obvious? You really need to be a University of Chicago professor to figure that out. The key lesson, which is actually like a research lesson is that even holding constant characteristics, so even after you net out, because we can do this in the model, the fact that the unemployed households have temporarily low liquidity.

Ganong: Net that out, you still find that even in these exact same economic conditions, as if the unemployed households were not unemployed, those households would still be more willing to spend. So the fact that you became, hypothetically, David, became unemployed during our recession, is itself a tag that you are a good person to send money to. From the perspective of some of a policymaker who's just trying to support aggregate demand, unemployed households are more willing to spend money regardless of their circumstances. That's the, we're not total dummies lesson, that came out of that part of the research.

The pandemic was a special time and the policies were special. So the special time will never, I hope, never be able to resurrect again. And that's one of the reasons why we think the disincentive was small...And so one lesson we take away is that temporary supplements are a particularly powerful tool for macro stabilization, even in a normal recession without a pandemic.

Share this

Beckworth: So for me then, a big takeaway is it's important to have targeted relief, send it to the right households that can use it as opposed to some of the criticisms we have seen about stimulus checks going to everyone. Let's segue into that. So there's a number of things we've already touched on that were done during the pandemic. The unemployment benefits were extended, the stimulus checks. We talked about PPP. We already shared some of your concerns there. There's also forbearance on student loans and mortgages. I don't know if you've thought much about this, but if you had to rate or to rank them where would you put them?

Ranking the Pandemic Response Measures

Ganong: That's a great question, David. I don't have a global ranking. Thankfully, I was not working in the government and so I didn't have to come up with that ranking. I want to make two points. The first is I want to draw a contrast between two different types of forbearance that happened. So we had mortgage forbearance and we had student loan forbearance and the similarities end literally there. Let me walk you through how each of them worked and try to pull out some of the design choices where we had different choices and ultimately made different choices.

Ganong: So for student loan forbearance, everyone was in. Regardless of their economic circumstances. For mortgage forbearance, we had a policy that was night and day compared to the Great Recession, but still required households to opt in. So households had to sign a document where they attested to the fact that they were experiencing economic hardship because of the COVID-19 pandemic. Altogether, of households holding mortgages in the US, our best guess is that 8% of households raised their hand and said, "Me, I think I'm going to need help." Ultimately, some of them actually made all their payments, but it was quite easy to access and 8% of households raised their hand. And you might say, "Oh my gosh, 8% of households potentially fell behind on mortgage payments." But you could also say 92% of households said, "I don't need help," and stayed with the status quo of where they were before.

Ganong: So what happened is we had basically an opt-in system for mortgage forbearance and we had an opt out system for student loans. And the most recent estimate I've seen for federal loans during the pandemic, specifically for federal student loans was that 99% of borrowers stopped making payments as opposed to the 8% who stopped making payments for the mortgages. So what I see here is that somewhere in between the really difficult to access, expensive mortgage modification programs that we ran in the Great Recession, where it may potentially cost the lender or the government $50,000 a pop to include someone in the old style relief programs, and the blanket relief that we saw for student loans.

Ganong: I think that it appears to me like there's a power and a success of opt-in programs with low hurdles to participation. So the mortgage forbearance was designed this time around thoughtfully with basically what I would describe as an incentive compatibility constraint in mind. So what I mean by that is that those 8% of households who raised their hand did not get a large gift in financial terms from participating. Why? Because the payments that they didn't make today, they still had to pay back in the future. And that's related to what I said before about temporary liquidity problems. Now, my coauthor, Pascal Noel, has calculated that a mortgage borrower during the pandemic who did forbearance for a typical mortgage, saved about $15,000 on their mortgage.

Ganong: The cost to the lender in net present value terms, including all this sort of foregone interest that I just mentioned, was $7,300. So there was some wealth transfer, but small compared to the immediate payment reduction. In contrast, in the Great Recession, households that got a $5,000 payment reduction, so a smaller payment reduction, it cost the lender and the government together about $45,000. So in one case, just to give you the numbers, borrower receives $16,000, lender or government loses $8,000. That's the pandemic. During the Great Recession, borrower receives $5,000 in the first year and costs the government or the taxpayer $45,000. So what that means is an example of a policy that was really thoughtfully designed so it’s incentive compatible, only the people who really need liquidity are going to sign up for this. And it was a success in my view.

Beckworth: That's an amazing number. 92% did not sign up or opt into this program. It's also amazing that it worked out so well. But for the 92% that did not opt in, do you think the other programs were part of the reason? So in other words, they could pay their mortgage because of the unemployment.

Ganong: Yes, absolutely.

Beckworth: Okay, speak to that then.

Ganong: Sure, so, the US economy has never had as strong of a safety net response as we had during the COVID-19 pandemic. So we talked already about student loan forbearance, unemployment insurance, stimulus checks. The situation in terms of household incomes never deteriorated. And so there were many fewer households that needed the type of specific relief of mortgages. But I would venture that even if they had more households that needed the relief, programs like forbearance, which provide liquidity and then you promise to pay it back down the road and then anyone who wants can raise their hand and walk in are likely to be successful. So it's a built in suspender situation where we didn't know what was going to work going in. And I feel lucky that together it all worked and now we need to figure out how to trim and reorganize before the next recession.

Beckworth: Yeah, great story, and as someone who's a big advocate of nominal GDP level targeting, as you've heard many times on the show, which can also be stated as nominal income targeting, I think it's a key insight or key observation to know that all these programs did effectively stabilize incomes in the aggregate. Not everyone equally. And that's key to financial stability. Households make lots of commitments, mortgages, car loans, lots of financial obligations. Same thing for businesses too.

Beckworth: And stabilizing nominal incomes, even if real incomes change, is key to financial stability. And all these programs played some role in doing that. Let's move on from the immediate 2020 pandemic and go to the next year, 2021. And that's when the second big pandemic package, the ARP, was passed. And you touched on this earlier, many people looking back now and I will admit in real time it was probably hard to know when the pandemic was going to be over, the delta variant.

Beckworth: A lot of things were happening in real time. So this is Monday morning quarterbacking for sure, but a lot of people look back, including myself, and say, "Man, I'm not sure that 2021 package was really needed." At least to the extent and scale that it was provided. Several trillion dollars more above and beyond what we saw in 2020. I wonder if you have any thoughts about that? I mean, is this a fair critique and if so, what could we have done differently to have made that second package, that ARP package, more appropriate and arguably less inflationary in terms of what we see today?

Evaluating the Effects of the American Rescue Plan

Ganong: So I think there's no question that the design of the American Rescue Plan was fighting the last war. Meaning, I think there's a broad sense among policymakers that after the Great Recession, fiscal and monetary support was insufficient, and so, well, let's fix our mistake. And frankly, I think I probably could've made the same mistake if I had been in that position at that point in time. You're right that we now have the benefit of further hindsight. And so having already said that, I think I would've made the same mistake. Let me try to think about what we could have done differently or would do a little bit differently in future times.

Ganong: So I've already talked about this question of opt in for mortgage forbearance versus opt out for student loan forbearance. Just to be clear, I would've made student loans opt in just like was done for mortgage forbearance, I think quite successfully. The other thing that I think... the best way to have supported households that are most in need is to do targeted aid on steroids and be comfortable with and tolerate and defend a level of misses and fraud in those targeted programs in the name of not having to do big stimulus checks that go to everyone, or a Paycheck Protection Program where business owners who don't need the funds get it. So I think would've done more of the targeted programs in order to do less of the two really big bazookas, the stimulus programs and the Paycheck Protection Programs. Let me pull out one specific example that comes to mind where I'd like to sort ideally reverse or reread back to you a recent Washington Post headline.

The best way to have supported households that are most in need is to do targeted aid on steroids and be comfortable with and tolerate and defend a level of misses and fraud in those targeted programs in the name of not having to do big stimulus checks that go to everyone, or a Paycheck Protection Program where business owners who don't need the funds get it.

Share this

Beckworth: Sure.

Ganong: So a recent Washington Post headline said, *US Watchdog Estimates 45.6 Billion in Pandemic Unemployment Fraud,* and that this is 3X the prior estimate of the extent of pandemic unemployment fraud. So just to be totally clear, fraud is bad. We need to build systems in advance of recessions where we can do third party verification of income. But at the end of the day, the US government spent half a trillion just on the supplements and I think about a trillion altogether. So let's take that fraud estimate as truth, that terrible headline of 45.6 billion. If for 5% fraud, or even frankly 20% fraud, we can do a lot of targeted supports and then not have to do the broad based supports, I would take that trade every single day and I think you should, and our listeners should be willing to as well.

Ganong: So something I'm disappointed with in the popular discourse is that the fraud is bad and the fraud should be prosecuted. But I feel that it gets the headline wrong because the whole point is we want to get money to people who are in need and we don't want to get so much money out the door that we generate inflation. I think we got that trade off wrong in both of the last recessions. But to get it right, we're going to have to be willing to tolerate some fraud and that's... From an economic policy making perspective, that's fine.

Beckworth: Yeah, there's always trade-offs. One thing you learn in economics, there's always going to be tradeoffs. I think that’s a great point, be willing to go strong and hard on targeted relief and forego the, potentially more distorting, larger stimulus package. So Peter, this has been a great conversation. We're running low on time and I'll just throw in this general observation that this has been a great experience to learn from. We've had some unfortunate developments, inflation, some fraud, but I hope we don't use this to go to the other extreme and in the next recession, opt in for austerity and prolonged recovery. I think that's the lesson from the Great Recession is we don't want to have a prolonged and stalled recovery. We want something better. We had a glimpse of it. In the pandemic we saw that it's possible. So we've got to find this balance between the Great Recession and some of what we saw in the pandemic, and I think your research is a key part of that story and understanding how to get there. So thank you for what you're doing on that front. And again, we encourage listeners to check out Peter's work. Our guest today has been Peter Ganong. Peter, thank you so much for coming on the show.

Ganong: Thank you so much for having me, David.

Photo by Frederic Brown via Getty Images

Support Mercatus

Your support allows us to continue bridging the gap between academic ideas and real-world policy solutions.Donate