Is the federal government's history of enormous deficits proof that it can keep up the borrowing? Some people think so, and legislators sure act that way.

Federal debt held by the public now equals the size of the annual economy – and is growing. Yet investors continue buying Treasury securities. No government bond sale has really failed, no one has suddenly refused to finance Washington's spending, and no unmistakable moment has forced Congress to change course.

That leads some people I cross paths with in academic life to conclude that despite decades of dire warnings, we've managed to go from a 35% debt-to-GDP ratio in 2007 to 100% today with no devastating consequences. Their implication is that we can keep going right up to 175% in 2056, which is the trajectory if Congress fails to reform Social Security and Medicare and if interest rates rise only modestly.

The idea that all this debt won't bring a reckoning may sound tempting, but it's questionable at best.

First, markets have already reacted. The 2021-2022 inflation spike, the worst in four decades, was the price level adjusting to a $5 trillion flood of deficit spending that investors did not believe future fiscal surpluses would cover.

That fear led to inflation. It's hard to believe we won't get another market response when Congress commits to borrowing nearly 16 times what it did during COVID-19, plus interest payments, without any repayment plan.

Incidentally, interest rates are well above what we saw from 2009 to 2021 (back when fiscal doves thought rates would always stay low). Some of the recent increases come from AI investment raising the demand for capital. Also at work, though, is a higher term premium. With inflation still an issue and no fiscal austerity on the horizon, investors – still worried that future debt will be devalued rather than truly repaid – want more compensation for holding long-term treasuries.

But let's just suppose Washington can stay the course without provoking a bond-market response. Assume that Uncle Sam can borrow an additional $138 trillion, including interest payments over the next 30 years – the amount projected by the Congressional Budget Office – and that there won't be a market reaction that raises interest rates or inflation.

How? First, it could be that investors expect Congress to eventually reform entitlements by raising taxes, reducing benefits, or both, so it won't need to increase borrowing. But if future Congresses don't deliver, the reckoning will come.

Second, it could be that despite all the risks in this scenario, investors would still value treasuries as the safest and easiest assets to trade. The world still needs dollars. Given the messes Europe, the UK and other governments find themselves in, some of this could be true.

Third, there's AI, which could supercharge economic growth. Faster productivity growth will mean higher incomes and a larger tax base, making debt easier to carry. But even here, there is no free lunch. AI investment may already be pushing up interest rates; a boom capable of substantially easing deficits would be even more likely to do so. So, as existing Treasury debt matures, it would be refinanced at higher prevailing rates.

Thus, faster growth improves the fiscal arithmetic, but whether it is enough depends on how fast the economy grows, how high borrowing costs rise and, ultimately, whether Washington continues running larger primary deficits.

Even if we assume the best-case scenario, this debt accumulation isn't costless. A bond-market crisis is not the only danger. Government borrowing competes with private borrowers for the economy's available savings, which is another way to end up with higher interest rates. It also means less private investment than we would otherwise have – in factories, software, housing and, yes, the very AI investments we are counting on to raise future growth.

Financing government spending through high and distortionary taxes carries its own economic costs. And, as my colleague Jack Salmon has documented, higher government debt is associated with slower growth. That matters even in a world where AI makes us richer.

Finally, government debt displaces other spending. Annual net interest payments now cost us more than national defense, and older, cheaper debt is being refinanced at today's higher rates. Those are dollars that can't shore up entitlements or help during the next emergency, when the ability to borrow matters most. CBO projects interest payments becoming the largest federal expenditure by 2040, consuming roughly 40% of the government's revenue by 2056.

I wouldn't count on markets to stay calm. But even the calm scenario means less investment, slower wage growth and a budget increasingly spent on past obligations. Without action and perhaps a more obvious cue to complacent congresses, the damage will continue, sometimes quietly, and the danger will grow.

Veronique de Rugy is the George Gibbs Chair in Political Economy and a senior research fellow at the Mercatus Center at George Mason University. 

COPYRIGHT 2026 CREATORS.COM