In 2001, Washington Was Worried About Paying Off the Debt Too Fast

The forecast was not a mood. It was an operating assumption.

Fiscal 2001 closed with a surplus of 128 billion dollars. It was the fourth consecutive surplus and, as it turns out, the last one anybody has seen. Debt held by the public finished that year at about 32 percent of GDP, the lowest it had been since the early 1980s.

Against that backdrop the Congressional Budget Office published a ten-year baseline in January 2001 showing 5.6 trillion dollars of cumulative surplus. Run that through the arithmetic and the publicly held debt is effectively gone before the end of the decade. The CBO's own numbers left 939 billion dollars of debt outstanding in 2011, but that residual was not a shortfall. It was paper that simply had not matured yet, and the agency assumed the government would let it run to term rather than buy it back at a premium. Alongside it sat 1.6 trillion dollars of surplus with nothing left to pay off.

That second number is the one that caused the trouble.

The peril of zero debt

On January 25, 2001, the Senate Budget Committee held a hearing titled Evolving Fiscal Challenges. The witness was Alan Greenspan. The problem he laid out was not how to retire the debt. It was what the federal government would be obliged to buy once it ran out of its own securities to retire.

His answer was that a government sitting on surpluses with no debt left would end up holding private assets, and that this would be bad for the efficiency of American capital markets. If it had to happen, he argued, those assets belonged in the Social Security trust funds rather than in ordinary on-budget accounts. By March he was telling the House Budget Committee that paying off the federal debt was in reach. On April 27 he addressed the Bond Market Association in a speech titled, without irony, The Paydown of Federal Debt.

He was not alone, and he was not a fringe voice on this. Alice Rivlin, the former budget director, was in the New York Times that same month under the headline Why Fight the Surplus, and the phrase that stuck to the whole debate afterward was the peril of zero debt. On December 28, 2000, the Clinton White House put out a release under the headline The United States on Track to Pay Off the Debt by End of the Decade. Its first bullet said that by dedicating the entire budget surplus to debt reduction, the country could eliminate its publicly held debt by fiscal 2009. That is the same year the CBO landed on four weeks later, from a different building, using different assumptions. The total national debt at that moment was about 5.6 trillion dollars, of which roughly 3.6 trillion was held by the public, which means the ten-year surplus Washington expected to collect was very close to the entire debt it had accumulated since 1789.

The Treasury acted on it, and you can still see the scar

Forecasts are cheap. Debt management schedules are not. The clearest evidence that Washington genuinely believed this is what the Treasury did to its own issuance calendar.

The 20-year bond and the 3-year note were retired during the surplus years as borrowing needs shrank. Then, on October 31, 2001, the Treasury announced it would suspend the 30-year bond entirely. The reasoning was straightforward and, on the projections available, sensible: there is no point selling paper that will still be outstanding long after the national debt has ceased to exist.

Four years and three months after the government decided it would not need thirty-year money, it went back to the market for thirty-year money. It has not stopped since.

What happened instead

The window the CBO was projecting, fiscal 2002 through 2011, did not deliver a 5.6 trillion dollar surplus. Add up what those ten years actually produced, year by year from the Treasury statements, and you get cumulative deficits of roughly 6.1 trillion dollars. Against a projected 5.6 trillion dollar surplus, that is a swing of about 11.7 trillion over ten years, on a forecast published for those exact ten years.

It got worse as it went. By early 2004, with the window less than a third elapsed, analysts were already measuring the deterioration at around 9.9 trillion using the projections available then. The final tally beat that estimate, because fiscal 2009, 2010 and 2011 had not happened yet.

Debt held by the public went from 32 percent of GDP at the end of fiscal 2001 to 98 percent by the end of fiscal 2023. It has kept climbing since.

The receipt

The useful thing about a miss this large is that it has been taken apart carefully. The Committee for a Responsible Federal Budget ran the accounting on everything enacted since the beginning of 2001 and expressed each category as the debt-to-GDP it added.

Five major tax bills between 2001 and 2017 account for 37 points. Discretionary spending increases plus Medicare expansions account for 33. The legislative response to the Great Recession and to the COVID-19 pandemic accounts for 28. Strip out any two of those three and debt-to-GDP would be roughly back at its 2001 level. Strip out all three and the debt would, in fact, be paid off.

The part that gets skipped in most versions of this story is the vote count. Of those points, 77 came from legislation with meaningful support from both parties. Highly partisan Democratic actions explain 12. Highly partisan Republican actions explain 8. Whatever this was, it was not one side doing it to the other.

In fairness to the counterargument, that split depends on how you classify two bills. The 2001 tax cuts and the 2003 Medicare Part D expansion both drew some Democratic votes but were driven by Republicans. Count those as partisan and the Republican column rises from 8 points to 20. The CRFB says so itself in a footnote, and it is a footnote worth reading before anybody quotes the 77 at you.

Viewed from the other direction, the same period saw federal spending rise from 17.7 percent of GDP to 22.7 percent while revenue fell from 18.9 percent to 16.5 percent. Roughly two thirds of the deficit growth comes from the spending side, one third from the revenue side. Both are real. Anyone telling you it was entirely one of them is selling something.

Now for the part everybody gets wrong

You will see the comparison in the form of zero versus forty trillion. It is not quite that, and the piece is stronger without the exaggeration.

The 2001 projection was about debt held by the public, which is the money the government owes to outside investors. Gross federal debt, the 40 trillion number in the headlines, also includes intragovernmental holdings, mostly the Social Security and Medicare trust funds holding special-issue Treasury paper. Nobody in 2001 projected that number going to zero, because it does not work that way.

So the apples-to-apples version is this. The measure that was supposed to hit zero went from 3.3 trillion dollars to 32.1 trillion dollars. That is roughly a tenfold increase against a forecast of complete elimination. It is a catastrophic miss on its own terms and it does not need the bigger number attached to it.

Where the odometer sits this week

Gross debt was 39.94 trillion dollars on August 11. It has been running at roughly 6.5 billion dollars a day since March, which is when it crossed 39 trillion. I keep a running table of how long each trillion has taken to add, and the entries at the top of it are measured in weeks. Forty trillion arrives this month, months earlier than forecasters expected as recently as the spring.

Part of that acceleration is the tariff picture. Following the Supreme Court ruling, the government has been refunding duties collected under emergency tariff authority, roughly 100 billion dollars of about 166 billion returned through the end of July, which pushed customs receipts into negative territory for three consecutive months. Revenue that was in the projections came back out.

Interest is now the second largest line in the federal budget. Only Social Security costs more. It exceeds defense. It exceeds Medicare. And because the low-coupon paper issued between 2015 and 2021 keeps maturing and getting refinanced at current levels, that line grows even in years when nothing else changes. The 30-year the Treasury retired in 2001 for lack of need is now being sold at north of 5 percent to fund the interest on the debt it thought it was about to eliminate.

The 41.1 trillion ceiling was set in July 2025. At the current pace, the next fight over it arrives earlier than anyone planned for. If you would rather watch the number move than read about it, the US national debt clock on this site runs live.