The nationwide debt is not $39 trillion. One economist says it is really $100 trillion

According to Kent Smetters, faculty director of the Penn Wharton Budget Model and one of the country’s most respected fiscal economists, that $39 trillion number is a polite fiction. The real tab, he argues, is closer to $100 trillion.

It has to do with the accounting distinction between explicit obligations — legally binding debts the government must repay — and implicit “pay-as-you-go” obligations — expected future spending commitments that carry moral or political, but not legal, force. “What we call implicit obligations are twice the size of explicit obligations,” Smetters told Fortune in a recent interview, referring to the unfunded liabilities buried inside programs like Social Security and Medicare.

If the U.S. government were required to report its finances under the same accounting rules as a publicly traded corporation, Smetters pointed out, the debt-to-GDP ratio wouldn’t be the current level of 100%, which is bad enough. “We’d be reporting a debt-to-GDP ratio closer to 300%.” The gap between those two numbers, he warned, is not a rounding error — it is the deliberate product of federal accounting standards designed to keep the full picture hidden from the public.

‘A shell game, not a Ponzi scheme’

Smetters is careful about the language he uses. Critics of Social Security have long compared the pay-as-you-go structure to a Ponzi scheme, in which early investors are paid with money from later ones. Smetters rejects that framing.

“It’s not a Ponzi scheme,” Smetters said, “it’s a shell game,” insisting that this distinction matters. A Ponzi scheme implies fraud, but Social Security never promised a higher return than market returns. It’s more of a shell game because Social Security and entitlement reforms can be moved obligations “off book”, from explicit Treasury obligations to implicit pay-as-you-go liabilities, because federal budget rules do not account for implicit liabilities. In 2001, Treasury Secretary Paul O’Neill attempted to explicitly book the value of pay-as-you-go liabilities but several events later that year – including the 9/11 attacks and the Enron meltdown – shifted the focus toward more immediate challenges. The newer accounting framework was outlined in a 2003 book published by the American Enterprise Institute.

In Washington’s case, Smetters said, that misdirection is written directly into federal law. His Beltway credentials include a stint as an economist at the Congressional Budget Office and as Deputy Assistant Secretary for Economic Policy at the U.S. Treasury, and PWBM is widely used in Washington DC to analyze the fiscal and macroeconomic effects of federal policy proposals. So he knows what he’s talking about when he rattles off, from memory, accounting legislation from 1985 as the origin of this shell game.

By statute, Smetters pointed out, the Congressional Budget Office is required to analyze Social Security on what’s known as a “scheduled benefits basis” — meaning its models assume the program will pay out full promised benefits indefinitely, even after the trust fund is completely exhausted. The problem, of course, is that current law also requires benefits to be automatically slashed the moment the fund runs dry, while the CBO is legally barred from modeling that reality.

This accounting quirk has enormous political consequences. It allows lawmakers to craft Social Security “reform” bills that appear, on paper, to close the program’s long-term funding gap — while actually doing nothing of the sort. Smetters points to the Social Security 2100 Act as a prime example. The bill was widely celebrated for eliminating the program’s 75-year actuarial shortfall. In reality, he argued, it actually worsened the long-term picture by increasing implicit debt by more than $1 for every $1 of reduction in explicit debt. Smetters said he believes this was an unintentional effect; it can just happen when working with an incomplete accounting framework that focuses on explicit federal debt.

This bill simply shifted costs from the explicit ledger — where they show up in official debt projections — to the implicit ledger, where they vanish from public view entirely. “They reduced these explicit liabilities, but it actually increased implicit liabilities by even more. And so instead of helping future generations, instead of incentivizing saving and labor, it ultimately, in the end, did the opposite.”

Handcuffed accounting

As a longtime budget watcher, Smetters attributed the issue to something almost mystical in the gears of government, “remnants” of legislation that remain even as Congress keeps writing new bills to supposedly correct the errors from previous regimes. In particular, he pointed to powerful “remnants” rooted in the 1985 Gramm-Rudman-Hollings Balanced Budget Act, which effectively prevent the agency from projecting discretionary spending to grow faster than inflation over the budget window. During his own three-year tenure at CBO, Smetters claimed, staff recognized this constraint as unrealistic and effectively required CBO to consistently produce debt projections that were too optimistic

“One of the things that’s left over from Gramm-Rudman-Hollings,” Smetters explained, “is that CBO by law is not allowed to grow discretionary spending faster than inflation over the next decade, over the budget window. And if you were to ask them, does this modeling make sense? They’ll be the first to say, ‘No, this absolutely makes no sense.’” In practice, this means the CBO’s long-range models have a structural problem built into them. “They constantly underestimate the growth in the debt,” Smetters said.

Congress passed the 1985 law in a moment of genuine fiscal panic: deficits had exploded during the early Reagan years, and lawmakers desperate for a mechanism to enforce discipline created a statutory framework that required automatic spending cuts — “sequestration” — whenever the deficit exceeded preset targets. To make that system work, the law needed a standardized, apples-to-apples definition of what counted as a “deficit,” so Section 257 codified the rules governing CBO’s official budget baseline. The political logic was straightforward: by locking in assumptions about what programs were “supposed” to spend, Congress could measure any new legislation against a fixed ruler.

The agency also faces restrictions on dynamic scoring — the practice of modeling how a policy change ripples through the broader economy — for any measure that affects fewer than 1 million people. “This is a pretty nutty one, too,” Smetters said. This blinds budget analysts to hundreds of billions of dollars in potential tax revenue generated by high-skilled immigrants over their lifetimes, he explained. Even high-skilled STEM workers who pay much more in taxes than they receive in benefits are scored as losing the government money on average, he explained, pointing to a 2024 analysis of this published by PWBM from former CBO director Douglas Elmendorf and MacArthur “genius grant” awardee Heidi Williams.

Setting aside the politics, Smetters argued, the economics suffer. “Suppose that we were to shift immigration away from low-skilled to high-skilled workers … even though we know those high-skilled STEM workers, for example, they’re going to pay a lot more taxes, CBO is not allowed to account for that.” The result is a federal budgeting apparatus that, through an accumulation of outdated rules and political compromises, consistently produces rosier projections than reality warrants.

Six years to a Social Security reckoning

For Social Security specifically, the law directed CBO to assume that scheduled benefits would be paid in full, indefinitely. It was a conservative modeling choice at the time because the trust fund was healthy following the 1983 reforms. What nobody fully reckoned with was that this assumption would become permanently embedded in law, transforming what was meant as a neutral accounting convention into a tool that could make structurally insolvent legislation appear fiscally responsible on paper.

At one time, according to Smetters, there simply wasn’t a “budget problem” in the United States, because there wasn’t an “automatic stabilizer” like in the Gramm-Rudman-Hollings legislation to keep benefits from growing at the same rate. If you just look at the data, he added, “discretionary spending clearly grows with the size of the economy. I mean, it grows clearly faster than inflation,” but it can’t do that by law. Gramm-Rudman-Hollings was never explicitly overturned, just parts of it bit by bit, “and so CBO just is not allowed to grow discretionary spending faster than inflation.”

The stakes of these accounting games are about to become very real. The Social Security trust fund — which covers Old-Age and Survivors Insurance — is now projected to be depleted by 2032, one year earlier than the CBO estimated just months ago, a revision driven in part by recent tax legislation. When that happens, the program is legally required to cut benefits to whatever level incoming payroll taxes can support — a figure currently estimated at roughly 84% of scheduled payments. Tens of millions of retirees would face automatic cuts with no congressional action required to trigger them.

That looming deadline should be forcing urgent reform debates in Washington. Instead, Smetters said, turning inwardly critical, although he said he doesn’t believe PWBM is a bad offender in this regard, years of alarmist rhetoric about crisis from think tanks and budget hawks have left lawmakers cynical and disengaged. “They’re kind of tired of the boy who cries wolf,” he says. They’ve been “saying the sky is falling for the longest time,” but the sky hasn’t exactly fallen yet, he said, likening it to predictions that the world was supposed to be underwater by now. “That’s a dangerous approach,” he said, adding that “what really needs to be done is really serious, thoughtful discussions of modeling and explaining to people.”

That fatigue is dangerous, because time is the one resource that fiscal reformers are rapidly running out of. The longer Congress waits, Smetters warned, the narrower the range of available solutions becomes. A structural fix implemented today could be spread gradually across generations — modest adjustments to the retirement age, phased benefit recalibrations, incremental revenue increases. A fix implemented in 2031, under pressure of imminent insolvency, will look very different: steep, sudden, and politically brutal. Think a steep tax hike, rather than a gradual, structural adjustment.

The deeper problem

Underlying all of it, Smetters argued, is an epistemic failure in how Washington evaluates fiscal policy. The simplified models and politically convenient scoring rules that dominate budget debates don’t just misrepresent the debt — they actively mislead the policymakers who are supposed to fix it.

The Treasury’s own Financial Report of the United States Government puts the 75-year unfunded shortfall at $73.2 trillion, driven entirely by Social Security and Medicare. The reason the distinction matters is compounding urgency: the longer these implicit liabilities remain off the official ledger, the less pressure lawmakers feel to address them, while the actual cost of closing the gap grows larger each year. 

The CBO already projects that deficits will hit $1.9 trillion in fiscal year 2026 and balloon to $3.1 trillion by 2036 under current law — and that projection still assumes Social Security pays full benefits even after the trust fund is projected to run dry around 2032, obscuring just how much worse the trajectory actually is.

Without transparent, micro-founded economic models that capture the true generational transfer of costs, he warns, lawmakers will keep falling for proposals that shuffle liabilities from one column to another without changing anything fundamental. “They’re just going to get duped into simplistic solutions that really don’t do much,” he said.

The United States is not, Smetters insisted, on the verge of imminent collapse. The debt is manageable — in theory. But the window for a managed solution is closing. And the accounting rules that were supposed to help Washington navigate a path forward are, by design, pointing in the wrong direction.