Alan Auerbach

EF: As you know, the federal debt stands at $36 trillion, more than 120 percent of gross domestic product. While high federal debt isn't new, it has grown enormously during the pandemic and post-pandemic eras. Should we be worried?

Auerbach: Yes, I think we should be worried.

I do have the problem of having said we should be worried a long time ago, when the situation wasn't as bad as it is now. I would say I think even more strongly now that we should be concerned about it.

One factor that clouds the issue is that some of the warnings that we've had — not from me — about huge spikes in interest rates, runaway inflation, and things like that haven't really happened. We haven't seen the sudden bad outcomes that some people might have expected.

Some people have argued that the debt is just not an issue. I think one of the problems is that it's not an issue you have to worry about until you do. And when you do, it's too late, really. At least, it's much more difficult to do things because by that time, you've gotten to a point where you really have to start cutting in very painful ways instead of making adjustments over a longer period of time that can be more subtle.

EF: Is that the bad outcome — that interest debt servicing displaces other priorities?

Auerbach: Yes. There are different ways that debt can lead to bad outcomes in countries that are less central to the world economy than the United States and don't have a reserve currency and are historically less trustworthy. It can cause a crisis in terms of lack of access to capital markets and things like that.

That's not what I anticipate for the U.S. What I anticipate more is just a gradual tightening of the vise, where more and more of the revenue we raise goes to debt service. And we're in less of a position to raise taxes because they're already creeping up. And our spending commitments are growing faster than our ability to tax.

One of the reasons why we haven't done more politically about the debt in recent years is that until the last couple of years, we've had low interest rates relative to our growth rate. They came down for several years below what was expected. And so it made people more and more sanguine. But if you look over the longer reach of time, such favorable interest rate outcomes are not something that one can anticipate.

It's better to start dealing with it now when we have a little bit of wiggle room than to wait until we're really up against it.

EF: An optimist's argument might be that productivity growth is going to be great, and we're going to be able to grow ourselves out of this situation. What is your reaction to that?

Auerbach: I think part of the problem is that historically interest rates and growth rates tend to move together. In the shorter run, of course, that's not true, but there are good reasons why stronger economies with faster growth would have higher interest rates. There are more opportunities for investment.

That means that, at least over the longer term, the government's not likely to come out that far ahead. If the growth rate picks up maybe in the short run, it will. So there's been a lot of emphasis and thinking about the difference between interest rates and growth rates.

EF: Can we take comfort from the fiscal situation in Japan, where public debt exceeds 250 percent of GDP?

Auerbach: I think not. First of all, some of the difference is that a lot more of the Japanese government bonds are held within government accounts. If you look at net debt-to-GDP ratios, which exclude debt held by the national government, Japan is still substantially higher than the U.S., but I don't think the gap is quite as big.

I think more importantly, the institutional differences between Japan and the U.S. make it easier for Japan to have a big debt-to-GDP ratio. Almost all Japanese government debt is held domestically, which is not true of the United States. So in terms of thinking about having willing holders of the debt, that's more true in Japan than it is in the U.S. Second, I think much more of the debt is held by financial institutions in Japan. The government's not simply going into debt markets the way it does in the U.S. A lot of it's held in financial institutions. It's not necessarily that they're required to, but it is part of the Japanese culture or custom that the debt is held that way.

And again, I think it means that the ability of the government of Japan to issue debt is higher for a given debt-to-GDP ratio. That won't necessarily always be true. Japan could also encounter serious problems at some point and it's hard to know when.

Also, if you look at some of the things that are going to press on the national debt, they're more problematic in the U.S. In particular, we spend a lot more in the U.S. on health care as a share of GDP than Japan does, and health care expenditures, both private and public, are growing faster than GDP. We're at 19 percent of GDP or something like that on health. That's substantially higher than Japan and at least the government component of it is growing and occupying a larger and larger share of our federal budget. That's adding a lot of pressure in addition to the debt service coming from the debt that's already been issued.

EF: Do you anticipate fiscal pressures will lead policymakers here toward so-called financial repression — measures to push Americans and American institutions to hold public debt, such as capital controls and regulatory requirements for financial institutions?

Auerbach: I don't. The U.S. went through a lot of financial deregulation. We've had financial repression in the past, but it was many decades ago, and it's hard to imagine imposition of capital controls or other requirements that essentially force lower interest rates on households to help finance the federal budget.

There hasn't been any movement in that direction in the political sphere from either side. I haven't heard any mention of it, and so I'm kind of doubtful that that's going to be one of the channels we use to deal with the federal debt.

EF: We've been talking about the federal debt broadly. When you think more specifically about Medicare and Social Security, do you see a crisis on the horizon for either of those programs?

Auerbach: The problems in those programs are a little bit like the problems with the federal debt itself. There is one important difference, which is that Medicare — at least Medicare Part A, the hospital insurance — and Social Security have trust funds. By law, Social Security, for example, can't pay benefits once the trust fund hits zero; they can only pay benefits that can be financed by current revenues, which would be substantially lower than the benefits that are currently promised.

The Social Security trust fund is projected by the Social Security trustees to run out of money in less than a decade. If that continues to be true, and it hasn't really changed much in the last few years, then we're going to get to a point where either there has to be a change in the Social Security system or benefits have to be cut.

I doubt that benefits will be cut across the board. That's what would happen if nothing were done. So in that sense, you might say there's a manufactured crisis in store. The same thing is true of Medicare Part A, which has a trust fund that also will eventually run out of money.

That said, I'm not as confident as some other people that this will lead to a reform of these programs. It's true that in 1983, which was the last time the Social Security trust fund was nearing exhaustion, we had the Greenspan Commission that recommended changes in Social Security, which were then adopted, which raised the retirement age very slowly and increased payroll taxes. That put the Social Security system on a better financial footing for many decades.

That could happen again. But it could also be the case that Congress and the government don't have the appetite for providing this kind of bad news to people in the Social Security system. They could just say, well, we'll use general revenue funding to cover the shortfalls of Social Security. We already do that for Medicare Part B, the health insurance, and Medicare Part D, the drug benefit. They are not self-sustaining; we have premiums paying for a small part of the benefits and the rest comes from general revenues.

Some of the traditional supporters of Social Security say it's good to have it be a self-financing system because it makes people feel that they have a stake in it when they're paying their payroll taxes and so forth. But if the choice of the government is to cut benefits, raise payroll taxes, or use general revenue funding, given their behavior in recent years, I'm fearful that they'll choose general revenue funding and just kick the can down the road.

EF: General revenue funding meaning, implicitly, debt funding?

Auerbach: Yes, that's exactly what it means. Right now, Social Security is walled off from the rest of the government in the sense that it has dedicated funding that including taxes on benefits as well as payroll taxes. That supports the system and, although we include Social Security in the unified federal budget, it is self-sustaining for the moment. Whether it remains so, we will find out in the not-too-distant future.

EF: Also related to retirement, you found in your research that the federal tax system and federal programs discourage the elderly from working. In what way?

Auerbach: Both the additional taxes that they pay when working and the benefits that they lose.

We always think of taxes discouraging work with increases in taxes as people work more. That's certainly one of the things that discourages work. It's true for the elderly just as it's true for everybody else. But in addition, especially for the elderly, there are some pretty large benefit programs that are means tested. This includes Medicaid, for example. We think of Medicaid as a program for the poor, but a large share of Medicaid benefits go to the elderly — for example, through coverage of long-term care. It's not covered by Medicare, but it is covered by Medicaid. But if your resources are too high, you don't qualify. And so if you have more income and more assets, you may not qualify for Medicaid. Supplemental Security Income is another transfer program that the elderly benefit from that is means tested, and of course there are others.

There are potentially pretty big disincentives to work if you are at risk of losing some of these benefits. They can swamp the effects of just the explicit taxes that you pay.

Moreover, there's a question of whether people really understand the way Social Security works for people who are below the normal retirement age, which is now essentially 67. For people who are retired, you can receive benefits as early as age 62 — unless you're disabled, in which case you can get them earlier. And then for roughly the next five years, you're subject to an earnings test, which says you lose benefits once your earnings go up above a certain amount. What's essentially a secret as far as most people are concerned is that you do get credits for the additional earnings. That is, your benefits go up in the future because you're earning money now. So if I am earning money at age 64, which wipes out all of my benefits, those benefits aren't gone. It's just deferring the benefits I'm going to get. There's an adjustment that essentially gives me the benefits back at a later date when I do fully retire.

But whether people understand that is quite doubtful. The evidence suggests they don't because there's a lot of bunching of earnings just below where the earnings test starts to kick in — which wouldn't be there if they understood. That is a potentially very large disincentive. It's a particularly unfortunate one because there already is in place an adjustment designed so it won't discourage people from working. But given that people don't seem to understand it, I think there's probably room for reform to make it more explicit, perhaps by getting rid of the earnings test entirely.

EF: Do you see taxation of Social Security benefits the same way?

Auerbach: Taxation of Social Security benefits affects people above a certain income, $25,000 if they're single, $32,000 if they're married. It's not indexed for inflation. So more and more people now have to pay taxes on their Social Security benefits.

Not only does that discourage retirees from working, it discourages them before they receive Social Security because if they have higher assets that they've saved, they're going to have higher income from those assets — interest, dividends, and so forth. And that's going to contribute to the income that might cause them to be subject to taxes on their Social Security benefits.

EF: We've had elevated inflation for about five years. You've argued that this has had significant hidden effects on households because federal fiscal policies don't fully take inflation into account. Please explain.

Auerbach: Well, there are different ways in which inflation interacts with the fiscal system to affect the taxes that people pay and the benefits that they receive. It could help them or hurt them; it mostly hurts them.

Some things are not indexed for inflation at all. I just mentioned one, which was the threshold over which you're taxed on your Social Security benefits. That threshold has been fixed in nominal terms since it was implemented. That means that the more inflation we have, the more people are going to be subject to tax on some or all of their Social Security benefits.

Where we do have indexing for a lot of elements of the tax system and benefits, there are delays before the system catches up. For example, once you're receiving Social Security, your benefits go up every year because of inflation. On the tax side, the federal tax brackets are indexed for inflation so that if your income goes up by 10 percent because inflation is 10 percent, it's not going to change your bracket because the bracket's indexed for inflation. However, there's a delay in the indexing. What that means is that if there's a sudden surge in inflation, the first year or so is going to happen before the brackets and the benefits start reacting to it. For example, if we went from an inflation rate of zero to an inflation rate of 10 percent on a permanent basis, that would cause a 10 percent decline in people's Social Security benefits because it would happen once and then we'd be forever one year behind.

The final thing is that capital income — interest, capital gains, things like that — are mismeasured because of inflation. For example, if I buy an asset for $100 and the price level doubles over the period that I hold it, and I sell the asset for $200, my real gain is zero. But I'd be taxable on a gain of $100, because we don't index capital gains for inflation. We don't index interest income. If the inflation rate is 4 percent and I'm getting 4 percent nominal interest, my real interest is zero, yet I'm still taxable on the 4 percent.

So through lack of indexing, delayed indexing, mismeasurement of capital income, as well as similar effects on the benefits side in terms of delayed indexing, people in general — not every person — have a reduction in resources as a result of inflation.

In one sense, that makes inflation a more effective tool for dealing with the deficit. It's traditional to think about sudden inflation as a tool governments use, particularly in less developed countries with very high debt-to-GDP ratios. They often may be tempted to try to inflate some of the debt away. Indeed, the U.S. debt-to-GDP ratio improved somewhat over the last few years, or at least it didn't get worse, even though we were running very large deficits, because we had a surge in inflation. This is an additional reason or channel through which inflation could help the government finance its deficits.

I don't think it's a particularly attractive way to do it because it's quite arbitrary. If you look at the distribution of effects, it varies a lot across households depending on the type of income they have. We wouldn't say it's very well designed.

EF: You've been paying close attention to fiscal policy for quite a while now. When you see the situation with the debt and debt-to-GDP play out, how does that affect you personally? Do you have some sort of gut reaction to all this?

Auerbach: Well, I am sad that the problems that I think are very, very important and should be at the top of the list of things government is dealing with don't interest the government at all.

You might say one of the frustrations of being an economist is that we often see, regardless of the thing we work on — we could be working on environmental policy, where I think there must be an enormous amount of frustration too — is that we have policies we think would work well, which the government doesn't seem very interested in. I think the best we can do is continue putting forward ideas of what we think government should be doing, the problems that we think it should be dealing with. And hope that somebody gets interested in them.

EF: What are you working on now?

Auerbach: I'm working on a few things. One of my most recent papers was on the national debt, looking at projections based on the last century or so and asking what kinds of government reactions to debt will put us on a stable path.

It's the case, as I've said in recent years, that the U.S. doesn't pay any attention to the national debt. That was not true if you go back, say, 20, 25 years or more. If you look, for example, during the Reagan administration as well as the first Bush and Clinton administrations, it was the case that when debt or projected deficits went up, government undertook actions to reduce them, either by increasing taxes or by cutting spending.

That ended sometime in the early 2000s. In the last 20 years or so, it's just not there. If we went back to the way we were behaving then, the kinds of shocks that are going to keep hitting the budget, either because of interest rates or pandemics or financial crises or other things, could be dealt with by those kinds of government reactions.

So it's both good news and bad news. It's good news in the sense that we've been there before. It's not as though we have to undertake an approach that's never been contemplated or practiced. But on the other hand, we lost religion sometime in the last 20 to 25 years. And it's not exactly clear how we're going to get that back because we lost it in a bipartisan way. There used to be bigger constituencies in Congress and in the White House for dealing with national debt, at least when problems became more apparent.

Another paper I've been working on estimates fiscal multipliers, in a broad sense — looking at the effects of, say, a fiscal expansion not just on earnings, employment, and GDP, but also looking at broader measures of social outcomes like mortality, divorce, homeownership, receipt of public benefits, and so forth. This is because my co-authors and I felt that we're taking a too-narrow view of the potential benefits of a fiscal expansion.

These broader benefits are substantial. That is, another dollar of government spending might increase social benefits by maybe 25 or 30 cents in ways that are not accounted for by the way we usually measure fiscal multipliers, that is, looking at effects on income or effects on employment.

EF: What do you think are the biggest unanswered research questions today in public finance?

Auerbach: I would say it's this point we were talking about before: We have a lot of information about the effects of policies and the design of policies, but we seem to lack a way of connecting those to actual policy adoptions.

One example has to do with redistribution; economists for a long time have thought about the optimal ways of redistributing resources in order to overcome inequality. We tend to focus on the outcome, that is, the resources that a household will have. And that's clearly not the way a lot of non-economists think about it. They tend to think about the income that they get before government. So, for example, people would seem to be much more interested in having a job that pays them a higher income than having a job with lower income and a government transfer payment. People tend to think more about what they get in the market as somehow an indication of their well-being and not necessarily equating that with what we give them. That has important implications.

Think, for example, about international trade. We say that free trade can be beneficial for all if those who are losers are compensated. The standard problem with that is we may not compensate people enough. But perhaps the bigger problem is that people may not view that compensation in the same way that they would view having a job. And therefore, as we're now moving away from free trade, governments seem to be more interested in trying to help people in ways that don't actually work through taxes and transfers.

Or think about environmental policy. Every economist thinks some sort of carbon tax would be the best way of dealing with it. We believe in pricing to get people to adopt the right behavior, given the problem of global warming and other externalities. But as much as there's been a bipartisan consensus among economists and attempts to interest policymakers in this, it's been very hard. We've instead adopted policies that are much less effective and much more costly from a social perspective.

So economists need to understand what's missing there — how people perceive problems like this, why they think the approaches that are being adopted are preferable. You might say these are questions of political economy rather than public finance. But ultimately, they are questions of public finance because they involve trying to design policies that are most socially beneficial in ways that can actually be adopted.

Econ Focus interview archive