Transcription
Uh, it's a great honor and a delight to be here and to deliver this year's Felstein lecture. As, as, as Jim pointed out, I was educated at Princeton and MIT, so I wasn't a student of Marty's yet. Marty, nonetheless, had a profound influence on my life and career.
Uh, I was a freshman at Princeton in 1977, and I then took introductory microeconomics from Harvey Rosen, who was a superb teacher. And he, Harvey, later hired me as his research assistant. Now, Harvey was a recent PhD student of Marty's. Uh, so even though I didn't know it at the time, I entered the economics profession as Marty's grand student.
Four years later, as a first-year student at MIT's PhD program, I took a couple of courses from a promising young assistant professor named Larry Summers, making me Marty's grand student yet again. Then, in the summer of 1982, President Reagan nominated Marty to chair the Council of Economic Advisers, and that's when I got a call from Marty. That was the first time we ever talked. On Larry's recommendation, uh, Marty offered me a job on the council staff. I quickly accepted and spent academic year '82-'83 in Washington.
I worked for Marty a second time in 1985 after joining the Harvard faculty. Marty was then head of EC10, the full-year introductory course at Harvard. At the time, it was standard practice for new assistant professors to teach a section of EC10 along with many graduate students. Most assistant professors disliked the assignment, and the practice was soon abandoned, but I loved it. EC10 served as a great reminder of why I fell in love with economics in the first place, covering a full year of micro in the fall and macro in the spring. Years later, when I sat down to write my introductory textbook, Marty's approach to the subject was firmly implanted in my brain. In some ways, I wrote the book that Marty would have written if he had ever taken the time to do so. When the book was published in 1997, I was delighted that Marty's EC10 was among the first courses to adopt it.
The topic I'd like to talk about today was close to Marty's heart. As Jim pointed out, the stance of fiscal policy and the path of government debt. Throughout his career, Marty advocated for greater saving, both private and public. As President Reagan's chief economist, he warned about the adverse effects of large government budget deficits, much to the chagrin of some other Reagan administration officials. If he were here with us today, I have no doubt that he would be concerned about the fiscal path the United States is now on.
Now, some years ago, the Wall Street Journal ran a cartoon that goes: A small child is coming home after getting off a school bus. As he opens the, the, uh, the door to the house, he shouts to his parents, "What's this I hear about you adults mortgaging my future?" Now, I like this cartoon not because it's funny. It's not really. Well, most of the Wall Street Journal cartoons aren't funny. Uh, but I like the cartoon because it succinctly summarizes the economics of government debt.
Courses in microeconomics and macroeconomics examine how government debt affects interest rates, capital accumulation, trade deficits, and so on. But the starting point for all that analysis is a transfer of income between generations. And that's exactly what this cartoon highlights. In their personal capacity, parents cannot choose to live beyond their means and leave negative bequests to their children. As voters and citizens, however, parents can do exactly that, and Americans are now doing it in a big way.
Now, historically, large changes in the debt-to-GDP ratio follow a simple pattern. Here's, here's a graph that I'm sure the macroeconomists in this audience have seen many times. The typical debt spikes up during crises, such as major wars, deep economic downturns, and the COVID pandemic. Then, when normalcy returns, the debt-to-GDP ratio gradually declines. This approach seems reasonable. Debt-financed spending during crises makes sense because it provides some stabilization of aggregate demand during downturns and it prevents large temporary tax increases when spending needs are extraordinary. The policy also ensures that the cost of crisis is shared among current and future generations.
The situation we now face differs substantially from this historical pattern. For those who follow economic policy debates, it's all too familiar. After massive budget deficits during the Great Recession of 2008-2009 and the COVID pandemic of 2020-2021, the government debt as a percentage of GDP is near the historic high reached at the end of World War II. By itself, that's not necessarily alarming. In the, in the few decades after World War II, the country managed to significantly reduce government debt relative to income through a combination of economic growth, some inflation, and fiscal prudence. But the trajectory ahead of us is not so benign.
According to the Congressional Budget Office, which is the projection you're seeing on the screen right now, the debt-to-GDP ratio will, under current law, continue to rise over the next three decades, reaching 156% in 2055. There's more of a no ending in sight to this. If they continue past 2050, it just keeps going up. Even more worrisome, this projection is optimistic. It assumes that the US economy will experience normal economic growth without a crisis like a major war, a deep recession, or another pandemic, which would put debt, push debt even higher. And this projection here does not, does not account for the so-called "big beautiful bill" that President Trump just signed into law, which will steepen the ascent of government debt.
Herb Stein once wisely said that if something cannot go on forever, it will stop. And I have no doubt that the path of rising debt-to-GDP ratio will stop at some point. The open questions are how and when it will stop. That's what I'd like to talk with you about today. I think logically, there are only five ways to stop this upward trajectory. They are: first, extraordinary economic growth; second, government default; third, large-scale money creation; fourth, substantial cuts in government spending; and finally, large tax increases. I'd like you to encourage you now to try to assign probabilities to each of these outcomes. Individually, each of these outcomes seems highly unlikely. But the probabilities you assign must, must sum to at least one. And I say at least because we could get more than one of these things, but we have to get at least one of them. So let's consider these five possibilities in turn. Okay?
So I want to start off with extraordinary economic growth. That would surely be the most benign of the possible outcomes. When the CBO makes its debt projections, it assumes future productivity will grow at about the rate we've experienced historically. Is it possible we're entering a new golden age of more rapid growth due to new technologies like artificial intelligence and advances in biotechnology? Yeah, it's possible. It's possible. The, uh, money manager Kathy Wood, for example, CEO of Arc Invest, has suggested that because of these developments, economic growth will soon accelerate from the 3% historical average to 6% to 8% going forward. Some people call this the technological singularity. My first thought when hearing such projections is that's nuts. Over the past few decades, we've seen the internet revolutionize how people work and lead their lives. Yet, economic growth has not been extraordinary. The effects of today's nascent technologies will likely be similar: life-changing, but not so transformative as to establish an entirely new growth path. In reaching this conclusion, I, I've been influenced, and maybe too much so, by the work of Robert Gordon on the rise and fall of economic growth and the work of Nicholas Bloom and his co-authors on the hypothesis that ideas are getting harder to find. I hope I'm wrong, and I hope Kathy Wood is right, but I wouldn't bet on it. It would surely be imprudent for fiscal policymakers to assume that rapid growth will come to their rescue.
The next possibility is that the government will default on its debt. For many people, such an event seems inconceivable. US government bonds are often considered among the safest of assets. But that view is, I think, much too sanguine. History offers many examples of sovereign default. Spain defaulted more than a dozen times between 1500 and 1800. More recently, we've seen defaults in Russia in 1998, Greece in 2015, Venezuela in 2017, Argentina in 2001, 2014, and 2020. The United States is not so immune to the political and economic forces that can make default an attractive option. Recall that Alexander Hamilton, the first Treasury Secretary, argued forcefully and successfully against default on the Revolutionary War debts. But other prominent figures at the time opposed Hamilton's plans and were more open to the possibility of partial default. In particular, James Madison thought that speculators who had purchased the debt from the original lenders at a deep discount should not be rewarded with full repayment.
More importantly, the United States has, in fact, defaulted on its debt. What I have on the screen in front of you is this brilliant book by Sebastian Edwards, "American Default." In the 1930s, many US bonds had gold clauses that ensured their value in gold bullion. When President Roosevelt decided to pull the nation off the gold standard, he recognized how expensive these gold clauses would be. So, he decided to abrogate them. Not surprisingly, the decision to unilaterally rewrite these bond contracts led to a court case, and the battle went all the way up to the Supreme Court. In a five-to-four vote, the court sided with Roosevelt. In the midst of the Great Depression, that outcome may have been desirable, but without doubt, it was a partial default, as the title of Edwards' book suggests.
You might naturally ask, "What about today? Might any modern-day president ever entertain the possibility of default?" Seems kind of inconceivable, doesn't it? Okay. Well, here is Donald Trump back in 2016 when he was initially a candidate for president. He was asked by a journalist, "How are you going to handle the debt?" And here's what he said: "I'm the king of debt. I'm great with debt. Nobody knows debt better than me. I've made a fortune by using the debt. And if anything doesn't work out, I renegotiate the debt. I mean, that's a smart thing, not a stupid thing." The journalist asked, "How do you renegotiate the debt?" And Trump replied, "You go back and you say, 'Hey, guess what? The economy crashed. I'm going to give you back half.'" Now, if President Trump's second term has proven anything is that he's willing to expand the Overton window, the range of policies and arguments deemed acceptable in political discourse. Remember this exchange the next time someone tells you that default on US government debt is unimaginable.
Now, it's sometimes said that a nation with debt denominated in its own currency never needs to default because it can always print the money it needs to repay its creditors. That is true, but I don't find the thought nearly as reassuring as some who advance it. We have lots of historical experience with what happens when central banks use monetary expansion to finance risky fiscal policy. The German hyperinflation, German hyperinflation of the 1920s is perhaps the most famous example. But more recently, you've seen a similar story play out in Zimbabwe. From 2006, the typical unit of currency in Zimbabwe went from 50 Zimbabwe dollars. You can see it as a picture of what a typical unit of currency looked like. And just three years later, the typical Zimbabwe currency looked like this: 100 trillion Zimbabwe dollars. And even a hundred trillion Zimbabwe dollars was soon worthless. I remember seeing a picture at the time taken in a Zimbabwe restroom. Here it is: "Cautioning people not to use the toilets to flush newspapers, cardboard, or Zim dollars." It is a well-known theorem of monetary economics that when people must be told not to flush their cash down the toilet, monetary policy is not optimal. I think you can find the proof of that in Woodford's books somewhere. Uh, um.
Now, such hyperinflation is, of course, a form of default in the sense that bondholders are paid back in worthless currency. But it's an especially destructive way to default. High inflation wreaks havoc throughout the economy. Given the choice, it may be better for the government to default explicitly rather than embark on an implicit default in the form of hyperinflation. Nonetheless, hyperinflations occur when fiscal policymakers don't want to come to grips with their own folly, and monetary policymakers are too weak to resist the pressures from fiscal policy. Such a regime, sometimes called fiscal dominance, doesn't always lead to hyperinflation like those in Germany and Zimbabwe. There are more moderate cases, like the 75% annual inflation that Turkey has experienced in recent years. That outcome is better than hyperinflation, but it's, it's hardly desirable.
It's worth noting in this context that Donald Trump has made clear that he believes the president should have more authority over monetary policy, an idea that most economists reject. Last month, Mr. Trump even publicly mused about appointing himself to the Fed. That's, here's the headline suggesting fiscal dominance. And he has consistently pushed for more expansionary monetary policy. Over the next few years, the conflict between fiscal and monetary policymakers could well become a defining event. It is unclear whether future Federal Reserves will have the fortitude to stand up to a demanding and belligerent president. So, I wouldn't rule out the high inflation scenario.
The next way to put fiscal policy on a sustainable path is to enact a substantial cut in government spending. Many people favor this alternative, at least until they consider the details of what it means. President Trump began his second term by empowering Elon Musk and the newly created Department of Government Efficiency, DOGE. That initiative has led to one of the largest reductions in the federal workforce in US history. I'm personally troubled by the chaotic approach that DOGE has taken. It seems to be following the famous Silicon Valley injunction, coined by Mark Zuckerberg, to "move fast and break things." This mantra may work well in a startup, but it's not the right way to run one of the world's largest and most important governments.
Regardless of one's views of the DOGE downsizing initiative, there's always reason to believe that its impact on the overall budget would be limited. The compensation of civilian employees of government, of civilian government employees, makes up only 4% of the federal budget. Moreover, contrary to some people's perceptions, the size of the federal workforce is not bloated by historical standards. And you can see that on this slide here. Federal civilian employment made up 4.5% of the economy's total non-farm employment in the 1950s. Today, it's under 2%.
When thinking about the federal budget, it's best to recall a quip from Peter Fisher, a Treasury official in the George W. Bush administration, who once called the federal government "an insurance company with an army." Defense spending constitutes 13% of the federal budget. More than half of federal spending is on Social Security and health programs. That percentage has risen over time and is projected to keep rising in the years to come as more of the baby boom generation retires and starts drawing benefits. Now, enacting large cuts in these entitlement programs is politically treacherous. When Paul Ryan was Speaker of the House, he endorsed some modest cuts in these programs. In response, the opposition party ran television ads showing an actor who resembled Ryan pushing a grandmother in a wheelchair off a cliff. Oops. I'm sorry. Here we go. Here's, here's a, here's a still from that, that famous ad. My sense is this ad campaign was pretty effective. That explains why President Trump has said throughout his political career, including as recently as February this year, that Social Security and Medicare are not going to be touched, other than to investigate fraud. All this leads me to conclude that, in light of what Americans expect from their government, substantial spending cuts are probably out of the question.
That brings me to the last way the United States may respond to its unsustainable fiscal trajectory, which is raising taxes. I view this as the most likely outcome in the long run for two reasons. First, each of the first four ways I've talked about—extraordinary growth, government default, high inflation, or massive spending cuts—seems either implausible or unacceptable. We might well get some of these, but we won't get enough of them to put fiscal policy on a sustainable path. Second, the United States is now a low-tax country compared with its peers. In most aspects of life, regression toward the mean is a strong and pervasive force. And this context is probably no different. According to the OECD, governments at all levels in the United States collect only 28% of GDP in tax revenue, compared with 35% in the United Kingdom, 41% in Sweden, 43% in Italy, and 46% in France. The OECD average is 34%.
A natural question is how much taxes must increase to close the impending fiscal gap. Larry Kolikoff does a lot of these numbers, and he looks at the present value of the infinite stream of future spending and taxes, and he estimates a gap of 7% of GDP. My own rough calculation suggests a somewhat smaller number. The CBO estimates the primary deficit averages 2% of GDP over the next 30 years. Add to that about 1% of GDP for the "big beautiful bill" that was just signed. And assuming that interest on the government debt exceeds the growth rate by 1 percentage point, and another 1% of GDP to service the existing debt, which is roughly 100% of GDP, that'll stabilize the debt-to-GDP ratio. This yields a fiscal gap in total of 4% of GDP, rather than the seven that Larry, Larry estimates. So this gives us some sense of the magnitude of the task ahead.
To close a fiscal gap of 4% of GDP with only increased revenue, the United States would need to raise overall tax revenue by about 14%. That's a huge tax hike, but would bring us only about halfway toward the level of taxation that prevails in the United Kingdom. US taxes would remain below the OECD average and well below the levels in France, Italy, and Sweden. From a strictly economic standpoint, that is entirely feasible. To be sure, most European nations use their higher tax revenue to pay for public services that Americans often finance privately. The most significant example is, of course, healthcare. If the United States were both to close this fiscal gap and provide universal government-provided healthcare, a much larger tax increase would be required, and that would bring the US tax burden close to the levels in Italy and, and Sweden. In this sense, the future of fiscal policy is intertwined with the future of health policy. But for now, let's set aside the possibility of major health reform and, and just focus on the, uh, the current fiscal gap.
A big question is whether a large, a tax hike large enough to close this fiscal gap is politically possible. I'm reminded of an old chestnut about Washington politics. It's been said that the United States has two political parties: the stupid party and the evil party. Sometimes the two parties get together and do something that's both stupid and evil. They call that bipartisanship. In that vein, there's now a bipartisan consensus about a central tenant of tax policy. The Republicans don't want to raise taxes on anyone except universities with large endowments. The Democrats want to raise taxes only on the richest 1%. So the two parties essentially agree that 99% of Americans should not have to endure higher taxes. This bipartisan consensus is the roadblock between where we are and where we need to go.
If a sizable tax increase is inevitable, as I think it is, we must look beyond the top 1% of the income distribution. Typical high-end taxpayers living in places like New York or California, where a lot of them live, are already taxed very heavily. Adding together federal income taxes, state income taxes, payroll taxes, and sales taxes, the marginal tax rate on ordinary income of the highest earners is about 55%. Of course, some loopholes that benefit the richest Americans are natural targets for reform. There are many, but the ones that come to my mind immediately are the taxation of carried interest, the treatment of capital gains in opportunity zone investments and qualified small business stock, and the section 199A deduction for certain pass-through businesses. That's a really geeky kind of tax policy think list to give you, but look, Google it, and you'll probably, I think you'll agree with me that I think they could use serious reform. But we shouldn't expect to get 4% of GDP in additional revenue just from those at the top. Attempting to do so would be highly inefficient given the marginal tax rates most of them already face. It's probably not even feasible to raise enough revenue from the small group. According to my back-of-the-envelope calculations, increasing the marginal tax rates on the richest 1% by 15 percentage points, so the top rate goes from 55% to 70%, would raise only about a quarter of the revenue needed. And this calculation assumes that people won't change their behavior in response to higher taxes, which is, of course, optimistic and unrealistic. In practice, increasing taxes on only the most affluent would raise much less revenue. That's why closing the fiscal gap will require broadly shared sacrifice.
The natural solution, I think, is a value-added tax. Most nations around the world have value-added taxes. And you can sort of see in this picture which ones do and which ones don't. The United States is almost alone in not having it. Among OECD countries, VAT revenues average 7% of GDP, which is more than enough to close the fiscal gap. One virtue of a VAT is that it taxes consumption rather than income. My reading of the vast literature on optimal taxation is that consumption is the better tax base because taxing it does not distort the margin between consumption today and consumption in the future. That would, however, distort the labor-leisure margin, but absent lump-sum taxes that can't be avoided if governments are to raise more revenue. Ed Prescott suggested that higher tax rates in Europe are the main reason that Europeans work less than Americans. That's a view that I find plausible. Yet others have proposed other explanations for the high levels of European leisure. Alberto Alesina and his co-authors emphasize the role of unions. Olivier Blanchard says there are different preferences on two sides of the Atlantic. If the United States ever institutes a sizable value-added tax, it will provide a natural experiment to test, to test Prescott's hypothesis. Excuse me. If he's right, sorry. If he's right, and Americans start working less, we'll need a somewhat larger tax increase than I've estimated. Sorry.
As of now, there's no obvious support among our political leaders for a value-added tax, but the idea is not completely beyond the pale. Back in, oh, thank you. Thank you, Nancy. Back in 2009, Nancy Pelosi briefly floated the idea of a VAT when she was Speaker, though she did not introduce any legislation. In 2016, when Paul Ryan was Speaker of the House, he advocated for a destination-based cash flow tax, which in some ways resembles a VAT. At times, I've even wondered whether Donald Trump's unconventional views on economics might lead him to favor a VAT, though not for the reasons I would advance. He has argued that value-added taxes of other nations are a trade barrier, like a tariff. That's not true, of course. Back in 1989, Marty Feldstein and Paul Krugman wrote a paper debunking that fallacy, and they weren't the first to do so. But simply, a VAT is trade-neutral because it applies equally to imports and domestically produced goods. Nonetheless, uh, nonetheless, Mr. Trump's misunderstanding, together with his affection for tariffs, might make him open to a US value-added tax. Oh, excuse me.
For a value-added tax to make its way through Congress and onto the president's desk, the minds of many politicians would need to change. In an ideal, well-functioning democracy, a blue-ribbon commission could study the unyielding budget arithmetic, offer a menu of realistic solutions, and convince voters that there are no easy choices. After voters are persuaded, our elected leaders would quickly follow. In the democracy we have, the path to fiscal reform could well be less deliberate and more painful. Change might occur only when the bond market loses faith in US American institutions, political institutions. If one day the bond vigilantes wake up and start viewing the United States as a large version of Greece or Argentina, they will stop buying US debt at normal rates of interest. Congress will have no choice but to face the music, regardless of the political consequences.
So those are the options. How soon might the day of reckoning arrive? Back in 2011, I wrote about these issues in the New York Times. The article took the form of a presidential speech that might be given during a future debt crisis. The United States was about to accept a bailout from the International Monetary Fund, whose headquarters had relocated to Beijing. The conditions for the bailout were substantial and painful cuts in government spending together with higher taxes on all but the poorest Americans. I set the date for this hypothetical speech 15 years in the future. That is next year. That date, I have to admit, was somewhat arbitrary. Like many economists at the time, I saw that US fiscal policy was unsustainable, but I did not see any evidence the bond market was about to hold policymakers' feet to the fire. So 15 years seemed like a reasonable guess.
Soon after the Times published the article, I received an email from Alice Rivlin, the great policy economist and the founding director of the Congressional Budget Office. Here's the email. I found it. She wrote to me, "Great piece in the New York Times on the debt, but I doubt the market will give us 15 years, maybe five." With the benefit of hindsight, we could say that Alice was wrong. 14 years have now passed without a debt crisis. This brings to mind an old quotation from Rudy Dornbusch: "In economics, things take longer to happen than you think they will, and then happen faster than you thought they could." Ernest Hemingway made a similar point. In his novel, "The Sun Also Rises," a character is asked how he went bankrupt. He replied, "Two ways: gradually and then suddenly." I wouldn't be shocked if the United States continued along the path of gradually rising debt-to-GDP ratios for another 15 years. But I also wouldn't be shocked if the bond vigilantes suddenly attack much sooner.
Cracks in the fiscal foundation are already starting to appear. In May of this year, Moody's downgraded US government debt below its AAA status, citing large deficits and rising interest costs. Now, none of the major credit rating agencies give the USA debt. It stopped rating. I began this lecture with a cartoon, so I'm going to finish with another. This one's my favorite one's from The New Yorker. It, uh, takes, oh, here we are, takes place in the Oval Office with the president's advisers huddled around the Resolute desk. They tell him, "Our deficit reduction plan is simple, but it will require a great deal of money."
This is precisely the situation we now confront. Putting the federal government on a sustainable path is, from a purely economic standpoint, relatively simple. If a random group of NBER research associates could be appointed a committee of monarchs, they could solve the problem in a long weekend. In the real world, the solution must come from our elected representatives, who know that any solution will impose significant pain on the current generation of voters. For most politicians, getting reelected is their highest priority. Enacting good policy, that's okay, too, but it's farther down the list. It is possible, perhaps even likely, that the solution won't come until the financial markets give policymakers little choice. That scenario would be unpleasant for nearly everyone.
But if Marty Feldstein were here today, he would likely give us reason to be more optimistic. His copious body of op-eds, many written with Kate, were premised on the conviction that a better-educated public would embrace a more rational economic policy. Perhaps that can occur this time. The United States experienced substantial declines in the government debt relative to GDP without major disruption from 1790 to 1830, from 1870 to 1910, and from 1945 to 1975. Maybe the fiscal future will indeed be so benign. I don't yet see the path from here to there through the political thicket, but I hope it's out there somewhere, ready to be found. Thank you very much.