Case

How Much Does the U.S. Pay in Interest on the National Debt?

About $970 billion in the last full fiscal year, more than it spent on national defense and about 18 cents of every dollar of federal revenue. That bill is on course to more than double within a decade. Here is where the number comes from, why it is rising, whether it matters, and what has and has not worked to contain it.

Updated September 13, 2026 · 5 min read · An independent project; the site’s own view is marked where it appears.

The short answer

Last year, about 18 cents of every dollar the federal government collected went to interest before a single program was funded.¹ In dollars, that was $970 billion in the fiscal year that ended September 30, 2025.¹ It was more than the country spent on national defense.¹ And it is growing fast: the Congressional Budget Office expects the interest bill to reach $2.1 trillion a year by 2036, more than double.⁵ ⁶

Two words get confused. The deficit is one year's gap between what the government spends and what it collects. The debt is the running total of all those gaps. The debt clock shows just over $40 trillion.² About a fifth of that is money one arm of the government owes another, mostly the Social Security and Medicare trust funds, which is why economists watch the roughly $32 trillion held by the public instead.² Interest is paid on all of it, but the figure that matters for the budget is what the Treasury pays to outside lenders.⁴

Why the bill is rising

Two things multiplied. First, there is more debt: what the public holds is now about the size of the whole economy, a level last approached at the end of the Second World War.¹ Second, it costs more to carry: the average rate the Treasury pays has more than doubled since 2021, from under 1.5 percent to about 3.5 percent.³ Debt borrowed cheaply in the 2010s is rolling over at today's rates, so the bill climbs even in a year when borrowing does not.

Underneath both is a gap that does not close on its own. The government is borrowing about $1.9 trillion this year, and on current law that rises to about $3.1 trillion a year within a decade.⁵ Strip out interest and the gap actually shrinks a little over that period; interest is what drives the total up.⁵ On that path the debt passes the World War II record by 2030.⁶ Interest last took this large a share of the economy in 1991, when rates were far higher and the debt far smaller.¹

Does it matter?

Economists disagree about where the danger line is, and it is worth being honest about that. In a widely cited 2019 lecture, Olivier Blanchard argued that when safe interest rates stay below the economy's growth rate, as they had for most of modern history, debt can be rolled over without raising taxes and its costs are smaller than commonly assumed; he also warned of the case where investors come to see the debt as risky and demand more, which raises the cost of everything.⁹ Since 2022 rates have risen and that comfortable gap has narrowed. Moody's, the last of the three major agencies to do so, stripped the United States of its top credit rating on May 16, 2025.⁸

What is not in dispute is the arithmetic already in place. Interest is the one line in the budget that cannot be cut by vote; it is owed. At 18 cents of every revenue dollar and rising, it is money that funds neither defense, nor Medicare, nor a tax cut, and every point that rates rise adds to it automatically. The question is not whether the country can borrow, which it plainly can, but how much of each future budget will be spoken for before Congress meets.

What has been tried

Congress has tried to bind itself before. The 1985 Gramm-Rudman-Hollings law set fixed annual deficit targets enforced by automatic across-the-board cuts; the Supreme Court struck down its enforcement mechanism the next year, 7 to 2, because the official who triggered the cuts answered to Congress rather than the president, and the law contained no deficit reduction of its own.¹⁰ ¹¹ Its 1990 replacement swapped targets for spending caps and pay-as-you-go rules, which held through the surplus years and lapsed in 2002.¹⁰ The 2011 Budget Control Act set caps through 2021; Congress raised them in 2012, 2013, 2015, 2018, and 2019.¹² Switzerland's debt brake, approved by 85 percent of voters in 2001 and in force since 2003, has lasted because it bends with the business cycle and its exception requires a supermajority of both chambers.¹³

Where Common Sense Again comes in

The debt cannot grow faster than the country does

A family or a business that borrows faster than its income grows eventually loses control of its own budget, and so does a country. The common-sense rule is the one every household already knows: debt may not grow faster than what you earn, with real room for genuine emergencies. Congress has never kept a rule like that for long because breaking it has never cost anyone a vote.

This site proposes a rule with the Swiss lesson built in. Outside defined emergencies, federal debt held by the public may not grow faster than the economy, judged over a rolling five-year period so a recession does not force cuts at the worst moment. CBO keeps the official score and GAO audits the process. Break the rule, and Congress must vote on a correction, on the record, by name; there are no automatic across-the-board cuts, which is what failed in 1986. Real emergencies get real borrowing, but the authority expires after two years unless three-fifths of each chamber extends it, and the excess returns to the normal rule on a published ten-year path, so an emergency ceiling cannot become the new floor.¹⁴

The objections that stand: Congress has found a way around every prior rule, usually through the word "emergency," and the three-fifths requirement is untested against that habit. A single debt rule treats borrowing for a bridge the same as borrowing for a tax cut, which the golden-rule tradition in public finance says is a mistake, and this proposal has no separate investment test. And if growth stays above interest rates for long stretches, as Blanchard described, the rule may bind when it need not. The proposal's answer is that a rule keyed to growth flexes with exactly that variable, and that the alternative on display since 2002 is no rule at all.

Read the repair and its evidence →

Go further

The interest bill is what a system that rewards keeping every fight alive costs when the bill finally comes due: no vote created it, and no vote can cancel it. The three-minute case is about the rules that produce that pattern, and the three repairs that would change it.

Read the three-minute case →

The repair in full → the debt rule, the emergency exception, and the tax and retirement changes beside it.
Will Social Security run out? → the other unfunded promise on the same books.
Why does the government keep shutting down? → the annual deadline that produces stopgaps instead of budgets.

Share this caseEmail Post on X

Sources

Primary records first. Where a source has a known lean or a limit, the note says so. Records marked with an S-number also appear in the site’s source register.

1.PRIMARYOffice of Management and Budget, Historical Tables (Fiscal Year 2027 budget edition). Table 3.1: net interest outlays $970,065 million in FY2025 (13.8 percent of outlays; 3.2 percent of GDP), national defense $916,140 million; net interest was 3.2 percent of GDP in 1991 and 1.8 percent in 2019. Table 1.1: FY2025 receipts $5,236,421 million, outlays $7,011,105 million, deficit $1,774,684 million. Table 1.2: FY2025 GDP $30,322.8 billion; receipts 17.3 percent and outlays 23.1 percent of GDP. Table 7.1: debt held by the public $30,167,224 million at the end of FY2025, 99.5 percent of GDP; 106.1 percent in 1946. whitehouse.gov →
The 18.5-cents figure is $970,065 million divided by $5,236,421 million; it is this site's arithmetic on OMB's two numbers, not a figure OMB publishes.
2.PRIMARYU.S. Treasury, Fiscal Data, “Debt to the Penny,” record date September 10, 2026: total public debt outstanding $40,047,726,949,770; debt held by the public $32,363,418,108,364; intragovernmental holdings $7,684,308,841,406. fiscaldata.treasury.gov →
3.PRIMARYU.S. Treasury, Fiscal Data, “Average Interest Rates on U.S. Treasury Securities,” Total Marketable: 1.458 percent for August 31, 2021 and 3.475 percent for August 31, 2026, read from the dataset's API. fiscaldata.treasury.gov →
4.PRIMARYU.S. Treasury, Fiscal Data, “Interest Expense on the Debt Outstanding,” fiscal-year-to-date through August 31, 2026: $1,267.8 billion across all expense categories, read from the dataset's API. This is gross interest expense, including interest credited to government trust funds, and is not comparable to net interest outlays. fiscaldata.treasury.gov →
5.PRIMARYCongressional Budget Office, The Budget and Economic Outlook: 2026 to 2036, February 11, 2026 (summary): the deficit is $1.9 trillion in fiscal 2026, 5.8 percent of GDP, and grows to $3.1 trillion and 6.7 percent by 2036; "Rising net interest costs drive much of that increase"; the primary deficit is 2.6 percent of GDP this year and 2.1 percent in 2036; federal debt rises to 120 percent of GDP in 2036. Projections reflect laws in place as of January 14, 2026. cbo.gov →
6.RESEARCHCommittee for a Responsible Federal Budget, “CBO's February 2026 Budget and Economic Outlook,” restating CBO's tables: net interest grows from a record 3.2 percent of GDP ($970 billion) in 2025 to 4.6 percent ($2.1 trillion) by 2036; debt held by the public rises from about 100 percent of GDP to 108 percent by 2030, passing the 106 percent record set after World War II, and to 120 percent by 2036. crfb.org →
CRFB is a deficit-focused advocacy organization; used here for CBO figures that appear in CBO's data tables rather than its summary text.
7.RESEARCHCommittee for a Responsible Federal Budget, “CBO Estimates $2.0 Trillion Deficit for First 11 Months of FY 2026,” September 9, 2026, citing CBO's Monthly Budget Review: $2.0 trillion borrowed October 2025 through August 2026, including $168 billion in August. crfb.org →
8.PRIMARYMoody's Ratings, U.S. sovereign rating action, May 16, 2025: long-term issuer and senior unsecured ratings downgraded to Aa1 from Aaa, outlook changed to stable from negative. Standard & Poor's (2011) and Fitch (2023) had already moved the United States below their top ratings. moodys.com →
9.RESEARCHOlivier Blanchard, “Public Debt and Low Interest Rates,” American Economic Review 109, no. 4 (April 2019): 1197–1229. The lecture argues that when safe interest rates are expected to remain below growth rates, public debt may have no fiscal cost and smaller welfare costs than commonly assumed, and it discusses the case against high debt, including the risk that investors come to see the debt as risky. aeaweb.org →
10.PRIMARYCongressional Research Service, R41901, Statutory Budget Controls in Effect Between 1985 and 2002, July 1, 2011: Gramm-Rudman-Hollings set annual deficit limits enforced by sequester and "did not include legislation that reduced the deficit"; the Budget Enforcement Act of 1990 replaced deficit targets with pay-as-you-go procedures and discretionary spending limits. everycrsreport.com →
11.PRIMARYBowsher v. Synar, 478 U.S. 714 (July 7, 1986), 7–2: assigning the Gramm-Rudman-Hollings budget-reduction function to the Comptroller General, an officer removable by Congress, violated the separation of powers. law.cornell.edu →
12.PRIMARYCongressional Research Service, R44874, The Budget Control Act: Frequently Asked Questions, updated October 1, 2019: the discretionary limits were changed by the American Taxpayer Relief Act of 2012 and the Bipartisan Budget Acts of 2013, 2015, 2018, and 2019, increasing deficits in each year from FY2013 to FY2021. everycrsreport.com →
13.PRIMARYSwiss Federal Finance Administration, “Debt brake”: the constitutional article was accepted by 85 percent of voters on December 2, 2001 and applied from the 2003 budget; the annual ceiling for ordinary expenditure is linked to ordinary receipts adjusted by a cyclical factor; extraordinary expenditure in exceptional situations requires a qualified majority of both chambers of Parliament. efv.admin.ch →
14.PRIMARYCommon Sense Again, Repair 03, “The debt cannot grow faster than the country does”: the rolling five-year test, CBO scoring and GAO audit, the recorded correction vote, and the emergency authority that expires after two years unless three-fifths of each chamber extends it, with a published ten-year path back. →

Published September 13, 2026. Last substantive update September 13, 2026. Load-bearing facts last re-verified against primary records September 13, 2026. This page is updated in place when the subject changes; it is not republished as a new article.