Where it stands

In 2025, the government spent $7.0 trillion and took in $5.2 trillion.

The difference is borrowed. Here is the shape of the gap before you touch anything.

$1.8T borrowed in a single year — 5.9% of the entire economy, against a 50-year average of 3.8%.

Where every dollar goes · ~$7.0T

Soc. Sec. Health Other Defense Nondef. Interest
Social Security ~$1.5T Medicare + Medicaid ~$1.6T Other mandatory ~$1.0T Defense ~$0.9T Nondefense discretionary ~$0.9T Net interest ~$1.0T

Where every dollar comes from · ~$5.2T

Income tax Payroll Corp. Other
Individual income ~50% Payroll ~33% Corporate ~8% Tariffs + other ~9%

Income and payroll taxes alone are about 83% of all federal revenue.

~$1.0T
Net interest on the debt — now roughly the size of the entire defense budget, and rising fastest of all.
~100% → 120%
Debt held by the public as a share of GDP, today and on the current path by 2036.

How it works

A deficit is one year. The debt is all of them.

A deficit is the gap between what the government spends and what it takes in over a single year. The debt is every past deficit, stacked on the last. To cover a deficit, the Treasury borrows — it sells bonds to investors, pension funds, foreign governments, and ordinary savers.

Here is the trap. That borrowed money isn't free: it carries interest, paid every year, mostly by borrowing still more. Bigger debt means bigger interest; bigger interest means a bigger deficit next year. The loop tightens on itself — and interest is now the fastest-growing line in the entire budget.

Deficit Debt Interest borrow owes interest feeds next year it compounds
~$30T
Debt held by the public — what the government owes outside investors, about 100% of GDP. The number economists watch, and the one this tool uses.
~$7T
Intragovernmental debt — money the government owes itself, mostly the Social Security and Medicare trust funds.
~$37T
Gross national debt — the two together, about 123% of GDP. The figure on the debt clock.

Where the money goes

Most of the budget isn't decided each year.

The government doesn't start from a blank page. Most of what it spends is locked in — promised by laws written decades ago, or owed as interest on money already borrowed. The slice Congress actually votes on each year has been shrinking for half a century.

1965 1985 2005 2025 60% 14% 26% 2035 proj
Mandatory — Social Security, Medicare, Medicaid & more Net interest on the debt Discretionary — what Congress votes on each year

Share of total federal outlays. Mandatory spending pays out automatically under standing law; discretionary spending — including the entire military — is set annually. Approximate; 2035 is the CBO projection.

~73¢ of every dollar spent in 2025 is automatic — Social Security and Medicare benefits written into law in 1935 and 1965, plus interest on debt already borrowed. Congress votes on the other 27¢, and that slice keeps shrinking. Most of this year's spending was decided years, even generations, ago.

How we got here

Every crisis ratchets it up. Lately it never comes back down.

Debt held by the public, as a share of the economy, since World War II — with the moments that moved it. After 2000, each shock lifts the line and it no longer falls back between them.

50% 100% 150% now 106% — WWII 23% — 1974 low 2008 crisis COVID, 2020 156% by 2055 projected 1946 1980 2008 2025 2055

Debt held by the public, percent of GDP. History from OMB and Treasury; 2026 onward is the CBO projection (dashed). The U.S. is on track to break its 1946 wartime record — this time with no war to explain it.


What the big moves cost

The crises and the choices, side by side.

Ten-year effect on the deficit, as scored. Some were emergencies; some were deliberate policy. One never passed at all.

2008–2009 Financial-crisis response +$1.6T

TARP, the auto rescues, and the 2009 Recovery Act. But TARP was nearly all repaid — Treasury closed its books on it at a small profit. The lasting cost was lost tax revenue in the recession.

Treasury / CBO
2020–2021 Pandemic relief +$5.3T

Six bills, from the CARES Act to the American Rescue Plan. The 2020 deficit hit $3.1 trillion — about 15% of the economy, the most since World War II. Much was temporary; not all of it came back down.

CRFB / PGPF
2017 Tax Cuts and Jobs Act +$1.5T

Cut the corporate rate from 35% to 21% and lowered individual rates. CBO's later estimate ran closer to $1.9T, and extending the expiring pieces could add roughly $4.6T more.

JCT / CBO
2022 Inflation Reduction Act +$0.9–2.0T*

Scored as a $238B deficit reduction at passage — but its energy subsidies are uncapped. Independent estimates now put them at $0.9–2.0T over ten years, and up to $4.7T by 2050 — far above the ~$370B first scored, likely flipping the law into a net cost.

CBO / Cato / Goldman Sachs

*Scored as a $238B deficit reduction in 2022; later independent estimates of its uncapped energy credits run far higher (Cato, Goldman Sachs, Penn Wharton). Figures are 10-year effects using different scoring conventions, so they are not strictly additive.


The limits of taxation

You can't simply tax your way out, either.

If spending is hard to cut, the obvious answer is to raise more. But the revenue side has its own ceilings — one historical, one economic.

The ceiling that won't move

Higher rates, stubbornly similar revenue

Revenue Tax rate → peak (debated) where the U.S. has sat ~17-19% of GDP

For fifty years, federal revenue has held within a narrow band — roughly 16 to 19% of GDP — even as the top income-tax rate swung from 91% in the 1950s to 28% in the late 1980s and partway back. Raising rates raises some revenue, but avoidance, slower growth, and a shifting tax base keep the total remarkably stable.

Closing a gap near 6% of GDP on taxes alone would mean revenue above 23% — a level the country has never sustained.

Illustrative. Economists disagree on the curve's exact shape and where the peak sits; most place the U.S. below the revenue-maximizing rate, so higher rates would still raise revenue — just less than a straight line predicts.

The growth it costs

Taxing capital isn't free money

Most of what the wealthy pay comes from capital — business profits, dividends, capital gains. Tax those harder and, in the standard model, some of the money that would have been reinvested in growing a company — new plants, hiring, research — goes to the Treasury instead. Less reinvestment means slower growth, which means a smaller economy to tax later. Part of what a rate hike raises is clawed back by the growth it discourages.

Tax profits & capital moreLess of each dollar a company earns stays in the business.
Less reinvestmentFewer new plants, less hiring, less research and expansion.
Slower growthA smaller economy, and softer wage gains over time.
A smaller tax baseSome of the revenue the rate hike raised is given back.

How large this effect is may be the most contested question in public finance. The U.S. taxes capital lightly by international standards, and the 1990s paired higher top rates with strong growth — so the brake is real, but its size is genuinely debated. Companies don't always reinvest what they keep, either: after the 2017 corporate cut, much went to stock buybacks rather than wages.


Interactive

Now try to close it yourself.

Every lever is a real policy with a sourced price tag. Spending can go up as well as down. Watch the deficit, the debt in 2036, and the years the trust funds run dry — and notice which levers move which numbers. Nothing here is rigged toward a conclusion.

Revenue

Raise more money — from payroll, income, corporations, or consumption.

Payroll tax rate+0.0 pt
+1 pt ≈ +$1.3T / 10 yrsCBO

Income tax surcharge+0.0%
+1% ≈ +$1.2T / 10 yrsCBO

Corporate tax rate21%
+1 pt ≈ +$130B / 10 yrsest

Payroll tax cap
$176,100 in 2025 · covers ~82% of wagesCBO
Itemized deductions
left as-is todayCBO
5% value-added tax (VAT)
a new national consumption taxCBO
Social Security

The only levers here that move the trust-fund dates. Some can make the gap worse.

Full retirement age
67 → 70 ≈ +$150B & closes 35% of the SS gapCBO/CRFB

Cost-of-living adjustment
CPI-W · today's methodmixed
Benefits for top earners
full scheduled benefits for allest
Spending

Cut it, or grow it. None of these touch Social Security's solvency.

Defense spendingno change
cut up to $959B or add up to $500BCBO

Non-defense discretionaryno change
1% cut ≈ -$95B / 10 yrsderived

New programs & initiatives$0
new spending Congress might addillustrative

Medicaid structure
open-ended federal match (current)CBO

The hard part

Pull every lever to its limit and the gap can close — the arithmetic was never the obstacle. The trouble is that almost every number that moves it sits on top of a promise: to people who retired counting on it, to the sick, to a generation that hasn't voted yet. The math is the easy part. What we owe one another is the hard part — and that is where the real argument starts.

Figures blend FY2025 actuals, the CBO Feb 2026 baseline, and the 2026 Trustees Reports. Lever estimates are 10-year scores and will not sum exactly, because real options interact. An honest illustration, not a budget score.