The US Federal Debt
$40 trillion and growing. What it costs, who holds it, and the paths from here.
(≈ $117,000 per American)
(above the 1946 peak; only 2020-21 was higher)
(more than defense spending)
The raw magnitude
The United States federal government owes just over $40 trillion. That number is larger than the annual economic output of any single country on earth, the United States itself included. About four-fifths of it (~$32T) is held by the public - private investors, foreign governments, pension funds - and the rest (~$8T) is held by federal trust funds like Social Security, essentially one pocket of the government owing another.
More revealing than the dollar figure is the ratio to the size of the economy. Debt grew faster than GDP through most of the last 25 years. The ratio jumped in the pandemic, fell back as inflation lifted GDP in 2021-22, and has crept up again since 2023.
Before the 2010s, gross federal debt had passed 100% of GDP only at the end of World War II, peaking at 119% in 1946. That ratio then fell for 35 years - not because the debt was paid down, but because GDP grew faster than the debt. The question today is whether that can happen again under current conditions.
How the US compares
"High debt" only means something against a reference. Most developed economies face some version of the same problem: the IMF's advanced-economy average was 108% of GDP in 2025 and rising. Here is where the biggest economies stand on the IMF's measure of gross government debt for 2025:
The US sits in the middle of this crowd. The dollar's reserve-currency role buys room the UK didn't have in 2022. But the debt burdens everyone is carrying today are historically high, and the usual escape routes - letting inflation run higher than the central bank's target, pressuring domestic banks and pension funds to hold government debt at unattractive returns, slow currency depreciation - are open to all of them at the same time. When everyone runs the same trick at once, the trick works less well.
Who holds the debt
Not all creditors are the same, and who owns the debt shapes how a crisis would actually play out.
Of that total, about $8 trillion is owed by the government to itself, mainly to the Social Security trust funds and to retirement funds for military and civilian federal workers. They are legally required to park their surplus in government bonds. The other ~$32 trillion is held by the market: American banks, insurance companies, pension funds, individual investors, the Federal Reserve, and foreign buyers.
The big picture: American savers ultimately carry most of the debt, either directly (through their pension funds, 401(k)s, and bank deposits) or indirectly (through the Fed, whose losses eventually become taxpayer losses). Foreigners held nearly half of the publicly held debt in the early 2010s and have held about 30% since 2022. Their holdings still hit a record $9.5 trillion in February 2026. A more domestic buyer base is easier to pressure by regulation, which would make the financial-repression path (below) easier to pursue and more painful for American households.
What it costs
Debt becomes a pressing problem when paying the interest on it starts crowding out everything else the government does. For most of the last 15 years, that wasn't a live concern - interest rates were near zero, so even a very large debt was cheap to carry. That era ended.
Net interest on the federal debt nearly tripled between fiscal 2020 and 2025, to $970 billion, and passed $1 trillion in the first 11 months of fiscal 2026. It overtook defense spending in fiscal 2024.
The federal debt has to be continually renewed. About a third of the debt sold to investors comes due within a year, and when an old bond is paid off the government just issues a new one in its place. If interest rates have risen since the old bond was sold, that replacement happens at the new higher rate - meaning the interest bill climbs even if no new debt is added. In fiscal 2025 the government took in $5.2 trillion and spent $7.0 trillion; CBO expects the gap to reach about $2.1 trillion in fiscal 2026. Under current law it projects debt held by the public to rise from 101% of GDP this year to 120% in 2036, past the 1946 record of 106%, with net interest doubling to $2.1 trillion.
The paths from here
There is no costless way out. Every historical example of a country reducing its debt involved one of the paths below - usually a blend. Each is listed with the question that matters most: will the political class actually take this path, or is it a fantasy option on a slide?
Grow out of it
The economy grows faster than the debt. The postwar playbook: from 1946 to 1974 the US cut debt held by the public from 106% of GDP to 22% without paying it down - largely by outgrowing it.
Will it happen? It needs no hard votes, which is its appeal. It also needs sustained growth above 3% a year (the US averaged 2.4% in 2016-2025; CBO projects 1.8% for 2027-2036) or a big leap in how productive the economy is. AI may deliver one - the Fed called productivity growth "strong" in September 2026 - or may not. On official projections, growth alone does not stop the debt ratio rising.
Inflate it away
The Federal Reserve tolerates inflation at 3-5% for a decade. Existing debt - where the interest rate was locked in when the bonds were issued - slowly loses value as prices rise around it. No new law required.
Will it happen? In some measure it already has: consumer prices have risen more than 2% a year in every month since March 2021 (3.4% in August 2026). It's a common historical route because it's invisible - no politician votes for it; the central bank just misses its target for years. Savers, wage-earners, and pensioners pay the bill; people who own assets (homes, stocks) fare better. Against a long run of it: the Fed raised rates in September 2026, and bond prices imply inflation of about 2.3% a year over the next decade.
Financial repression
Regulations push banks, insurers, and pension funds to hold government bonds at returns lower than the rate at which prices are rising. Combined with moderate inflation, the real weight of the debt slowly erodes. Used in the US from 1945 to 1980.
Will it happen? Possibly. It appeals because voters don't see it; it shows up as bank rules and "safe asset" mandates, not tax increases. Pieces exist: regulators eased big banks' leverage rule in November 2025 to free room for Treasury trading, and the 2025 GENIUS Act will require dollar stablecoins to be backed by cash and short-term Treasury bills. But the defining feature is missing for now: 10-year inflation-protected Treasuries pay about 2.9% above inflation.
Raise taxes
Federal revenue is about 17% of GDP. Counting all levels of government, US taxes were 25.6% of GDP in 2024, against an OECD average of 34.1%. Raising federal revenue by 3% of GDP would cover the gap before interest costs and about half the deficit.
Will it happen? Not on a normal day. No president has won on raising middle-class taxes since 1980, and the last big tax law went the other way: CBO scored the 2025 budget act as cutting revenue by $4.5 trillion over ten years. Real tax increases historically require a crisis to force the conversation - WWII, the 1990 budget deal, emergency legislation after a shock in the government-bond market. Without a crisis, this stays at "tax the other guy" symbolism.
Cut spending
About three-quarters of what the federal government spends is locked in by law - Social Security, Medicare, Medicaid, interest on the debt. The rest, which Congress votes on each year, is about $1.9 trillion of $7.4 trillion in fiscal 2026, defense included; cutting all of it would not quite close the deficit. The arithmetic forces the conversation back to the big programs.
Will it happen? Only in part. The 2025 budget act cut projected federal health spending, including Medicaid, by about $1.1 trillion over ten years, but made no broad cuts to Social Security or Medicare benefits; every serious attempt to cut those since Reagan has been buried. The main Social Security trust fund is projected to run out in late 2032; unless Congress acts, benefits then drop to the 78% that incoming taxes cover. Deep cuts stay politically unlikely until a crisis removes the choice.
Extract from abroad
Historically, great powers under fiscal stress have turned outward. Colonial resource extraction, war reparations, post-war asset seizures, weaponizing the dollar (blocking adversaries from the dollar financial system) - most great powers have reached for some version of this. The modern US versions: sanctions, tariffs, pressure on allies to pay more for their own defense, and collecting a small "tax" on every country that holds US bonds just because the dollar is the currency everyone uses.
Will it happen? It already is, in part. Tariffs lifted customs revenue from $77 billion in fiscal 2024 to $195 billion in fiscal 2025, but the Supreme Court struck down the emergency-powers tariffs in February 2026, and most of the $166 billion they raised is being refunded (CBO). A New York Fed study found US firms and consumers bore nearly 90% of the tariffs' cost. Whether this path looks like soft pressure (tariffs, sanctions, export controls) or something harder depends on who is in office. Both parties have shown appetite for different flavors of it.
Kick the can
Add about $2T per year to the debt indefinitely. Interest compounds. Credit-rating agencies downgrade. At some unpredictable point, the government holds a bond sale and fewer buyers show up than expected - interest rates jump, the dollar stumbles - and the adjustment comes as a crisis rather than a choice.
Will it happen? This is the current path - the default, because it requires no new votes. Moody's removed the last top-grade US credit rating in May 2025. In September 2026 the 10-year yield hit 5.18%, the highest since 2007, and a five-year note sale drew its thinnest demand since 2018, though it was still bid twice over. The 2022 UK crisis showed how fast confidence can shift. A similar event in the US bond market is not a prediction, just a real possibility.
The likeliest outcome is not one path - it's a mix: inflation somewhat above target, some pressure on savers, revenue sought abroad, and continued borrowing, possibly punctuated by a crisis that forces either entitlement cuts or a tax reset. In that mix, savers lose most and asset-holders least. The question isn't "will there be pain" - it's "who pays it, and when."
Where economists disagree
The sections above make the case for concern. It is not the only credible reading. Stephanie Kelton, Olivier Blanchard, and Jason Furman - among others - hold versions of the arguments below, and the data does not rule them out. A reader who only ever hears that the US is in trouble is getting an incomplete picture.
A country that prints its own money can't go bankrupt the normal way
A country that borrows in its own currency is structurally different from a household, a business, or a state government. It cannot be forced into default the way Greece or Lebanon can, because it can always create more of the currency its debts are written in. The real limit is inflation, not bankruptcy. "The government is maxing out its credit card" misses this difference: the government owns the credit-card company and the printing press.
Held by: Stephanie Kelton and the economists associated with Modern Monetary Theory in its strongest form; Olivier Blanchard and Jason Furman in milder mainstream versions.
When the economy grows faster than interest rates, the debt shrinks on its own
In 2019, economist Olivier Blanchard made a simple but important observation: as long as the economy grows faster than the interest rate the government pays on its debt, the debt burden shrinks over time - without Congress cutting anything or raising anything. The math just takes care of it. Historically, that condition has held more often than not.
Still credible? Growth still beats the average rate: the Treasury paid 3.5% on its marketable debt in August 2026, while nominal GDP grew 6.6% in the year to mid-2026, helped by inflation. The catch is the deficit before interest, about 2.6% of GDP in fiscal 2025, which keeps the ratio rising unless growth stays that strong. Periods when growth beats rates and periods when rates beat growth alternate; the situation is not the one-way slide the alarming version of the story suggests.
The dollar's advantage is stickier than critics think
The death of the dollar has been predicted every decade since 1971, when the US cut the last formal tie between the dollar and gold. Every time, it has survived. The reason is simple: the more people use a currency, the harder it is for anyone to switch. Banks, businesses, and governments around the world have built decades of habits, contracts, and financial plumbing around the dollar. Switching is less like changing a vendor and more like asking the world to change its common language.
The alternatives all have serious problems. The euro sits on a fragile group of countries with very different budgets. The yuan is controlled: Beijing decides who can move money in and out of China, which makes it unattractive as a global savings asset. Gold can't easily handle modern payments at scale. The rest are too small or too volatile.
The evidence: Foreign central banks are drifting slowly toward other assets, not fleeing. The dollar was still 57% of disclosed reserves in early 2026 (IMF). As long as the US has the world's biggest, most easily-traded bond market - with strong property rights and no restrictions on moving money in or out - global savings will keep ending up there.
The headline debt number isn't the only one that matters
The total debt number is one way to look at the picture - but not the only way. The government also owns a lot of things (federal land, natural resources, shares of its own debt held in trust funds, the Federal Reserve's balance sheet) and has the largest tax base in the world backing future payments. Other useful ratios: how much of the government's annual revenue goes to interest, or how big the debt is relative to the tax base that stands behind it.
The framing point: Treating the headline debt number as the one number that matters is a choice, not a law. The same country looks very different depending on which measure you center.
It has been done before
The US brought its debt held by the public down from 106% of the economy in 1946 to 22% by 1974, and briefly reduced it in absolute dollars in the late 1990s. It wasn't legendary belt-tightening. It was a mix of growth, mild inflation, and moderate tax and spending choices.
What it implies: Today's ratio is not inherently destined to spiral. If political conditions change, so does the math.
None of this says there is no problem. It says the problem is not bankruptcy, and the timeline is not "crisis this year." The reading that fits both the concerns above and the arguments here: the US has real room to keep borrowing and will keep using it. The costs build slowly and unevenly - on savers, on lower-income households, on public services as interest payments crowd out everything else. And whether this becomes a crisis depends less on the math than on political choices that have not been made yet. Less alarming than "the US is headed for a cliff," less reassuring than "there's no real problem" - closer to how the data actually reads.
What this means for you
This is a slow-motion event, not an acute crisis. It won't play out next week. But under current law the direction is clear, the tools governments tend to use are historically familiar, and you can position for them without panicking.
If you hold cash savings
Ongoing inflation plus the kind of quiet regulatory pressure described in the paths above means money sitting in a savings account at 0.5% while prices rise 3-4% loses buying power every year. That has been true for twenty-plus years and will likely be true for another decade. Keep three to six months of expenses as emergency cash in a high-yield savings account, but don't park long-term savings there.
If you're thinking about retirement
If you're under 55, plan as if Social Security benefits could be about a fifth lower than currently scheduled. That is the automatic cut under current law when the main trust fund runs out in late 2032 (78% of benefits payable), unless Congress acts - and how it will act is not settled.
If you're already retired or close to it, your biggest long-term risk is inflation, not market swings - steady-income investments (government bonds, bank savings products) lose buying power when inflation runs above their interest rate. Either way: take every employer-retirement match you can, and keep meaningful stock exposure even late in retirement. Historically, a diversified stock portfolio preserves buying power across decades where bonds don't.
If you own a home
A paid-off home is one of the strongest protections against inflation an ordinary person can own. A low-rate fixed mortgage is even better - you're effectively being paid over time, because inflation erodes the real value of the monthly payment you locked in years ago. Don't refinance a sub-4% mortgage just because rates wobble.
If you're thinking about your career
Decades of fiscal stress have historically favored defense, finance, physical assets (real estate, commodities, energy), and the sectors the government chooses to subsidize for national-security reasons (semiconductors, critical minerals, domestic manufacturing). Sectors that depend on what people spend by choice - restaurants, travel, entertainment, non-essential retail - get squeezed first in any downturn.
What to do, and what to avoid
Don't panic-sell stocks for cash. Don't trust any politician claiming a clean solution - there isn't one, and anyone saying otherwise is selling something. Don't go all-in on gold or crypto as if a collapse is coming next week; slow erosion, not sudden crash, is the realistic risk. Don't take on more debt than you could still afford to pay if interest rates rose another two percentage points.
The one concrete positive move: if you have meaningful cash beyond your emergency fund, spread it beyond cash - a fund that tracks the overall stock market, some international exposure (in case the dollar weakens), a small position in real estate or real-estate investment funds, perhaps a modest allocation to gold. This isn't financial advice. In the 1970s gold and housing outpaced inflation while cash and bonds fell behind; stocks lagged it that decade too, but have beaten it over longer spans.
The mechanics behind this
The paths above depend on mechanisms you may want to understand first: how money is actually created, why prices form, what governments can and can't force people to do. These fundamentals make the debt conversation much more legible:
