The National Debt: What Both Parties Actually Did

The national debt is not a single number. It is several numbers, and choosing among them without disclosure is how most misleading political claims about fiscal policy are constructed. Before any data comparison is possible, every reader needs to understand what these measures are and which economists consider meaningful.

Deficit vs. debt: The deficit is the annual gap, federal spending minus federal revenue in one fiscal year. The national debt is the accumulated total of every past deficit minus every surplus. A president can reduce the deficit (the annual shortfall is smaller) while still adding to the debt (because any deficit, even a shrinking one, adds to the running total). These two facts are routinely conflated in political debate.

Gross federal debt vs. debt held by the public: Gross debt (about $40T as of September 2026) includes intragovernmental debt. Treasury securities held by federal trust funds like Social Security and Medicare. Debt held by the public (about $32T) is what the government owes to external creditors: the Federal Reserve, foreign governments, domestic investors. The CBPP, CBO, and virtually every economist treat debt held by the public as the more economically meaningful figure because it represents actual claims on U.S. credit markets. Intragovernmental debt is essentially one government account owing another. Politicians sometimes choose gross debt to make numbers sound larger.

Nominal debt vs. debt-to-GDP: Adding $1T to the debt in a $30T economy is less burdensome than adding $500B in a $5T economy. Debt-to-GDP is the standard measure used by CBO, the IMF, and the Federal Reserve. A president can add substantial nominal debt during rapid economic growth while actually improving the debt burden, or add modest nominal debt during a recession while worsening the ratio.

Structural vs. cyclical deficit: A cyclical deficit occurs automatically in recessions. Tax revenues fall and safety-net spending rises with no new legislation required. A structural deficit is the underlying gap that would exist at full employment. CBO estimates that automatic stabilizers alone increased deficits by roughly $1.1T during 2009–2012. Much of what looks like a president’s “spending” during a crisis is automatic stabilizer spending inherited from prior law.

Debt ceiling vs. appropriations: The debt ceiling limits Treasury’s ability to borrow to finance spending Congress has already authorized. It is not authorization to spend. Both parties have raised or suspended it dozens of times; it is a procedural mechanism, not a policy statement about who caused the debt.

The table below uses gross federal debt as a percentage of GDP at the end of each fiscal year, OMB via FRED series GFDGDPA188S. Nominal figures for debt held by the public are the fiscal-year series FYGFDPUB. Average deficits are FYFSGDA188S. Congress control is noted because the president proposes but Congress passes spending and tax law. Fiscal year 2009 sits in both the Bush and Obama windows below, and it carries TARP and the first year of the Recovery Act, so the handover year does more work in this table than any other.

The only post-1980 president to reduce the gross debt-to-GDP ratio was Clinton, down 8.9 points over eight years. But this requires immediate context: the Clinton surplus was driven by a once-in-a-generation tech boom that inflated tax revenues from capital gains, a peace dividend from reduced Cold War defense spending, and divided government that constrained new spending on both sides. The conditions were extraordinary.

Every other administration since Reagan saw the ratio rise. Biden’s 1.9 points is the smallest of the six increases, because nominal GDP grew almost as fast as nominal debt, but the ratio still rose and the deficit stayed above 5% of GDP in all four years. The nominal debt has risen under every president since Eisenhower. The relevant question is not whether a president added nominal dollars, virtually every president does, but whether they improved or worsened the debt burden relative to the economy.

The budget lag: A president’s first budget does not take effect until October 1 of their first year. The start of Fiscal Year 2. A new president’s first 8–10 months run on the prior president’s enacted appropriations. Trump’s second term is a live example: the FY2025 deficit of ~$1.8T covers a fiscal year that began October 1, 2024, before he took office January 20, 2025. Simple “President X added Y in year one” claims ignore this entirely.

Mandatory spending is inherited: Mandatory spending was about 60% of federal outlays in FY2024, and about 73% counting net interest. No new legislation is needed for Social Security, Medicare, Medicaid, or interest payments to continue. They grow automatically with demographics and existing benefit formulas. A president who wants to change mandatory spending must pass new legislation through Congress, which is politically among the hardest things to do in American government.

Inherited wars and crises: The Brown University Costs of War project estimates post-9/11 wars cost the U.S. about $8T in budgetary terms across administrations. Obama inherited both wars, paid for them through most of his term. Bush inherited the dot-com recession; Obama inherited the financial crisis; Trump inherited the end of the recovery; Biden inherited the tail of COVID. These shocks do not belong primarily to the president on whose watch they peaked.

Interest on inherited debt: When a president inherits a large debt, they are required to pay interest on it regardless of any new policy choice. Net interest payments have doubled as a share of GDP from 1.6% in 2020 to a record ~3.2% projected for 2025. The federal government now spends more on interest than on defense or Medicare. Every dollar of interest on pre-existing debt that appears in a president’s spending total is an obligation they did not create.

The Constitution gives Congress the power of the purse. Presidents propose; Congress enacts. In multiple periods, Congress spent substantially more than presidential budgets proposed. The debt ceiling requires Congressional action. Tax legislation requires Congressional action. The president’s leverage is primarily the veto and the bully pulpit, real but indirect.

The most rigorous conservative argument does not claim Republicans always borrow less. The data does not support that. The strongest case is structural: the long-run fiscal problem is primarily a mandatory spending growth problem, not a revenue problem. Revenue as a share of GDP has remained relatively stable at 17–18% historically. Spending has grown from ~20% of GDP in the 1980s to ~23% in 2025 and is projected to reach ~27% by 2055 under current law, driven almost entirely by Social Security, Medicare, and interest, not discretionary choices.

Republicans can correctly point to the ARRA and American Rescue Plan as large discretionary spending additions during Democratic administrations, and to the ACA as an entitlement expansion. They can correctly note that the 1990s surpluses required divided government and a boom, and that Democratic unified government periods (2009–10, 2021–22) produced the two largest single-year deficits in peacetime history outside of COVID emergency spending.

On tax cuts: the strongest supply-side case is not “they pay for themselves”, that claim is false under official scoring. The strongest case is that lower marginal rates raise investment returns, which raises growth modestly, which partially offsets revenue losses. JCT’s dynamic analysis of the TCJA found the law would raise GDP by an average of about 0.7% over the budget window, cutting the deficit by $385B (a $451B revenue gain less $66B in added debt service) against JCT’s own conventional cost estimate of $1.456T. That recovers about a quarter of the static cost. The growth effect is real and it is not self-financing.

The most rigorous progressive argument focuses on the revenue side and on what CBO actually scores. The Bush 2001/2003 tax cuts added an estimated ~$1.8T to deficits over their first decade. JCT scored the TCJA at $1.456T over its first decade, and CBO’s later estimate of the conventional cost runs near $1.9T. CBO found that making the TCJA permanent without offsets would add about $37T to deficits through 2054. Output would run above baseline in the first several years, start subtracting from growth in 2028, and leave the economy smaller than baseline in 2034 and after, as interest crowding-out effects overtake the growth benefits. These are not partisan claims. They are official scoring from a nonpartisan agency.

Democrats can correctly note that Clinton’s 1993 tax increase on higher earners, passed with zero Republican votes, was followed by the only sustained period of deficit reduction and eventual surpluses in modern U.S. history. The argument is not that tax increases alone produce fiscal balance, but that the revenue side is a real lever that supply-side rhetoric consistently dismisses.

On the IMF austerity evidence: post-2010 research revised earlier views. The IMF found fiscal multipliers in recessions are higher than previously estimated, meaning sharp spending cuts during downturns can worsen debt-to-GDP by contracting the denominator (GDP) faster than the numerator (debt). The European austerity experience post-2010 is the strongest case study: countries that cut spending sharply saw prolonged stagnation that kept debt-to-GDP elevated despite nominal spending reductions.

Reinhart and Rogoff’s 2010 paper claimed countries with debt above 90% of GDP experienced sharply slower growth, including negative growth. This became a major argument for austerity policy across the developed world. In 2013, Herndon, Ash, and Pollin found the paper contained a coding error, selectively excluded data, and used unconventional weighting. When corrected, average GDP growth for high-debt countries was +2.2%, not −0.1%. Reinhart and Rogoff acknowledged the error while maintaining the broader point that very high debt correlates with slower growth.

The honest summary: there is no clean 90% cliff. High and rising debt does impose real costs: through higher interest burden, reduced fiscal space for emergencies, and potential crowding out of private investment, but the relationship is not as deterministic as the original paper suggested. Japan has carried 230%+ debt-to-GDP for decades without collapse, partly due to domestic bond holdings and domestic savings. The U.S. benefits from the dollar’s reserve currency status, which gives it more latitude than most countries. Neither of these facts means high debt is harmless.

Modern Monetary Theory (MMT): The strongest MMT position is that a sovereign currency-issuing government cannot face a solvency crisis in the same way a household or currency-user can. Spending and tax policy can proceed without being constrained by revenue in the short run; inflation is the operative limit. The mainstream counterargument: real resources are limited, and sustained deficit spending in a near-full-employment economy produces inflation (as the 2021–2023 experience illustrated). The Fed itself does not analyze U.S. debt as irrelevant despite dollar reserve status. MMT’s framework is coherent but does not eliminate real constraints.

CBO projection: CBO’s February 2026 budget outlook projects the deficit rising from $1.9T in FY2026 to $3.1T in 2036, with debt held by the public going from 101% of GDP in 2026 to 120% in 2036 under current law, and 175% by 2056 in the extended projection. The dominant drivers are major health care programs, Social Security, and net interest. This is not a partisan claim. It is CBO’s nonpartisan projection of what current law produces.

Unfunded liabilities dwarf the reported debt. The 2026 Social Security Trustees report puts the 75-year open-group unfunded obligation for OASDI at $29.3T in present value, up from $25.1T in the 2025 report. Medicare’s comparable 75-year unfunded obligation runs near $53T. Treasury’s FY2024 Financial Report puts the 75-year gap between projected non-interest spending and receipts at $72.7T in present value. None of this appears in the ~$40T gross debt figure most politicians cite.

Who holds the debt matters. As of 2025: domestic private investors hold ~50%, the Federal Reserve holds ~13% (down from COVID-era peaks as QT proceeds), foreign investors hold ~24% (Japan ~$1.1T, UK ~$809B, China ~$756B. China is not the dominant holder the political narrative implies). About two-thirds of the debt is held by U.S. entities. The “we owe it all to China” framing is a significant oversimplification.

Off-budget war spending. For years after 9/11, Iraq and Afghanistan costs were funded through “emergency supplemental appropriations” outside normal budget caps. This made annual headline deficits appear smaller than they were and understated the fiscal cost of the wars during the Bush administration.

The Federal Reserve’s role. Quantitative easing involved the Fed purchasing trillions in Treasury securities, effectively recycling debt and keeping interest rates suppressed through 2022. This helped finance large deficits at low cost. Interest payments on Fed-held Treasuries flow back to Treasury, partially offsetting their burden. The unwinding of QE (quantitative tightening) since 2022 has shifted Treasury bonds back to private markets and contributed to rising interest costs.

Federal assets. The Treasury’s FY2024 financial report shows federal assets of ~$5.7T against liabilities of ~$45.5T, for a negative net position of ~$39.9T. This is different from the “debt to the penny” figure and is another reason simplistic slogans are inadequate.

Both parties have added substantially to the debt. This is true in nominal dollars and, except for Clinton, in debt-to-GDP terms. No post-Reagan president left the country in a better fiscal position by the debt-to-GDP measure except Clinton, and Clinton required divided government, a historic tech boom, and a peace dividend that cannot simply be replicated by ideology.

Republicans have more consistently driven debt increases through tax cuts. The Reagan 1981 cuts, Bush 2001/2003 cuts, and Trump 2017 TCJA are all scored by JCT/CBO as adding trillions in deficits without self-financing. These are the most cleanly attributable partisan fiscal choices in the record. “Tax cuts pay for themselves” has been the dominant Republican fiscal claim for 45 years and has never been confirmed by official nonpartisan scoring.

Democrats have more consistently driven debt increases through spending. ARRA, ACA, and the American Rescue Plan are the most clearly attributable Democratic fiscal expansions in the recent record. Emergency spending is not inherently irresponsible (counter-cyclical stimulus in deep recessions has strong economic support) but the scale of Biden-era deficits even after COVID recovery represents structural spending above revenue that official projections do not show correcting.

The structural problem transcends both parties. Social Security and Medicare costs grow with an aging population regardless of any presidential ideology. Interest on accumulated debt grows with the debt stock. Mandatory spending and net interest together were about 73% of federal outlays in FY2024, and CBO projects both to grow. CBO’s projection of 120% debt-to-GDP by 2036, and 175% by 2056 under current law, reflects demographic reality, not partisan choices, though partisan choices on taxes and spending determine how much worse it gets.

Claims that should simply stop being made:

“[Party] alone caused the debt.” Both parties contributed through different mechanisms over 45 years.

“Tax cuts always pay for themselves.” Official scoring has never supported this. The strongest dynamic estimates recover about a quarter of the static cost.

“Obama doubled the debt.” As a causal attribution, this is inaccurate. As raw arithmetic in nominal gross dollars starting from a crisis baseline, it is true. The distinction matters.

“Spending alone is the problem.” Revenue choices also matter. Both sides of the ledger built this deficit.