Federal spending, by the numbers

You are right that the debt matters. The question is who actually fixes it.

A common belief, held across kitchen tables for decades, is that Republicans are the careful ones with the budget and Democrats are the spenders. This is a sourced look at whether that brand holds up, what the 2017 and 2025 tax laws actually cost, what DOGE actually saved, and what the official scorekeepers say is really driving the debt. It is not an argument that the other party budgets better. It is narrower, and harder to wave off.

How to read this

The instinct is correct. Hold that thought.

If you hate waste and worry about the debt, your instinct is right, and the numbers back it up. The honest disagreement is not about whether the problem is real. It is about the story we tell for who causes it and who fixes it. So a few ground rules.

Every load-bearing number here is attributed to a nonpartisan scorekeeper: the Congressional Budget Office (CBO), the Treasury, the Government Accountability Office (GAO), the bipartisan Committee for a Responsible Federal Budget (CRFB), and the credit-rating agencies. Where a claim is a projection or where credible estimates disagree, it says so, using these labels:

Documented on the official record.   Estimate a strong figure, but an estimate.   Projection a forecast that assumes no policy change.   Contested credible sources disagree.

And one promise: this page concedes the strongest version of the other view out loud, in its own section, because an argument that cannot survive the obvious objections is not worth making. There is no partisan color-coding here on purpose. It is meant to read like a ledger, not a team jersey.

The scoreboard

First, the problem is real and it is getting worse.

Start where everyone agrees. In the 2025 fiscal year, for the first time ever, the government spent more than $1 trillion just on interest on the debt, more than the entire defense budget for the second year running. That is money buying nothing: no roads, no troops, no benefits, only the cost of past borrowing.

$1.0 trillion

Net interest on the federal debt in FY2025 (about 3.2% of the economy), up roughly 8% from $949 billion the year before. Interest first passed total defense spending in FY2024. (CBO, Monthly Budget Review for FY2025) Documented

$1.8 trillion

The FY2025 deficit, about 5.9% of GDP. The 50-year average is roughly 3.8%. We are running a war-or-recession-sized deficit in a year with neither. (CBO) Documented

So this is not a manufactured worry. It is a structural one. The rest of this page is about why it keeps happening no matter who is in charge, and why the usual explanation, that one party is disciplined and the other is not, does not match the record.

The brand test

Does "Republicans budget better" survive the scoreboard?

The cleanest measure of fiscal discipline is debt held by the public as a share of the economy. If a president leaves it lower than they found it, they shrank the burden. By that measure, here is every modern presidency.

Reagan (R)
26 to 41%
Bush 41 (R)
41 to 48%
Clinton (D)
48 to 32%
Bush 43 (R)
32 to 52%
Obama (D)
52 to 76%
Trump (1st) (R)
76 to ~98%
Biden (D)
~98 to ~99%

Debt held by the public, share of GDP, start to end of term, rounded. Green is the one term it fell. Source: OMB Historical Tables and CBO. Documented

Read it honestly and the partisan story falls apart. The debt-to-GDP ratio rose under Reagan, both Bushes, and the first Trump term, and it also rose under Obama and Biden. It fell under exactly one president in the last forty years: Bill Clinton. The accurate headline is not "Democrats budget better." It is that no president since Clinton has lowered the debt burden, of either party.

What this does not prove Congress writes the budget, not the president, so pinning these numbers on one person is already a simplification. Reagan faced a Democratic House for all eight years. And the worst single years here were emergencies, not philosophy: the 2009 crisis and the 2020 pandemic. These are not cyclically adjusted, so a fair comparison narrows the gap between the parties further. That is the point. The gap is not what the brand claims.

And the lone exception proves how hard it is. Clinton's surpluses came from a specific, unrepeatable mix: a 1993 tax increase, the bipartisan 1997 balanced-budget deal, a Republican Congress pushing spending restraint, a tech-boom surge in capital-gains revenue, and a post-Cold-War defense drawdown. (Cato Institute) It was divided government and good luck, not one party's virtue. Contested

The theory

The idea underneath the brand is quietly dead.

The intellectual case for "tax cuts are fiscally responsible" is a theory called "starve the beast": cut taxes, and you force Congress to cut spending later because the money is not there. It is the hinge the whole brand swings on. The trouble is that it does not work, and the people who say so are conservatives.

William Niskanen, longtime chairman of the libertarian Cato Institute and a Reagan economic adviser, studied the record and concluded starve-the-beast was a failure: cutting taxes did not reduce spending, and may have made it easier by lowering the perceived price of government. (Cato Journal, 2006) Economists Christina and David Romer found no spending slowdown after four major postwar tax cuts. (Romer and Romer, NBER, 2007) Documented

The pattern held every time it was tried. Spending did not fall after the Reagan cuts, after the Bush cuts, or after the 2017 cuts. This is the single strongest finding on this page, and notice what it does and does not say. It does not say either party is virtuous. It says the specific mechanism many fiscal conservatives trust to turn tax cuts into discipline has never actually delivered the discipline.

Trump, first term

The boom-year deficit, separated from the pandemic.

The first Trump term added a headline figure often cited as more than $8 trillion in debt. But that number is mostly the pandemic, and most of the pandemic spending was bipartisan: the CARES Act passed the Senate 96 to 0. Around $3.6 trillion was COVID relief and roughly $3 trillion was pre-existing baseline. Lumping it all together is the kind of shortcut this page is trying to avoid.

The revealing part is what happened before COVID. The deficit rose every single year of the boom, while unemployment fell to 3.5% in September 2019, the lowest since 1969. Deficits are supposed to shrink in good times. This one grew.

FY2017: $665B → FY2018: $779B → FY2019: $984B. The deficit reached 4.6% of GDP during the strongest labor market in half a century. (CRFB; BLS) Documented
The clearest policy choice was the 2017 Tax Cuts and Jobs Act. It was scored at about $1.5 trillion over ten years, and it did not pay for itself: even counting growth, analysts found only a fraction offset, leaving a roughly $1 trillion hole. (CBO, 2017; Tax Policy Center) Documented

Be fair about the spending side too: the 2018 and 2019 budget deals that lifted spending were bipartisan, demanded by Democrats as the price of Republican defense increases. The deficit growing in a boom is a shared failure. It is just not the failure the brand predicts.

The 2025 law

The "One Big Beautiful Bill," described before it is judged.

The marquee fiscal act of the second term is the One Big Beautiful Bill Act, signed July 4, 2025, on near party-line votes (51 to 50 in the Senate, 215 to 214 in the House). It is worth describing accurately first, because the honest version is more persuasive than the cartoon.

It extends and expands the 2017 tax cuts, adds new breaks (no tax on tips, on overtime, on some Social Security income), raises defense and border spending, and pays for part of that with cuts to Medicaid, to SNAP, and to clean-energy credits. CBO's conventional score:

+$3.4 trillion

Added to deficits over 2025 to 2034, or about $4.1 trillion once you count the $718 billion in extra interest. The most costly reconciliation law in recent memory. (CBO, Effects of P.L. 119-21) Documented

Credit where it is due Two things cut in the law's favor, and leaving them out would be dishonest. First, about $3.9 trillion of the gross cost is simply extending tax rates Americans had already paid under for eight years, not a brand-new giveaway. Second, the law paired its cuts with roughly $2.5 trillion in real, scored offsets, including about $1.1 trillion in Medicaid savings. That is more spending restraint than recent Democratic laws even attempted. Republicans did cut a major entitlement.

But the offsets do not erase the cost, because the tax cuts are bigger than the offsets, which is why the net is still up $3.4 trillion. And the savings come from a particular place. CBO's own all-in distributional analysis:

The lowest-income tenth of households loses about $1,600 a year (down 3.9% of income), while the top tenth gains about $12,000 (up 2.3%), on average over 2026 to 2034. This is CBO's table, not an advocacy group's. (CBO, Distributional Effects of P.L. 119-21) Projection
About 10 million more people uninsured by 2034 from the law's own provisions (the larger "15 million" figure circulating also folds in a separate lapse of ACA premium subsidies that this law did not enact). (CBO via Georgetown CCF) Projection

So the fair read is not "they were reckless spenders." It is that they chose deep cuts to programs for lower-income people and still increased the deficit by trillions, because the tax cuts at the top outweighed the savings. The discipline was real, and it was selective.

DOGE

The waste is real. The savings were not what you were told.

The Department of Government Efficiency was the part of this most likely to give a waste-hater hope, so it deserves a straight answer. Start by validating the premise: federal waste is real and large. GAO estimates roughly $186 billion a year in improper payments and a far larger range in fraud. (GAO) Wanting that cleaned up is not naive. It is correct.

But the results did not match the promise:

The savings target collapsed from $2 trillion to $1 trillion to about $150 billion a year as the effort ran into reality. (Fortune) Documented
The public "wall of receipts" was riddled with errors, most famously an $8 billion cancellation that was actually $8 million. Independent reviews could verify only around $2 billion in real contract cancellations early on. (NPR) Estimate
Meanwhile the churn had a price. One analysis put the cost of mass firings, rehirings, lost productivity, and lost IRS enforcement revenue at up to $135 billion in FY2025, more than the effort plausibly saved. (CBS News) Estimate
The fair version DOGE is partly being judged on a yardstick it could never move, and that is worth saying. About two-thirds of the budget is mandatory spending and interest (Social Security, Medicare, debt service) that no executive task force can touch. Elon Musk even publicly opposed the deficit-increasing 2025 bill. So the honest critique is not "DOGE caused the debt." It is that DOGE overpromised by an order of magnitude, and the reason it could not help is the same reason the real problem is structural.
Tariffs

A real revenue source, but a second-best one, and partly struck down.

Tariffs were sold as a way to fund the government and make other countries pay. Credit the true part: they did raise real money, fast, without needing 60 Senate votes.

Customs duties roughly doubled to about $195 billion in FY2025, the highest share of federal revenue since before 1980. (Treasury Combined Statement FY2025) Documented

But "other countries pay" is not how tariffs work. They are a tax on imports, and the cost lands mostly at home:

A New York Federal Reserve analysis found roughly 90% of the tariff cost fell on US importers and consumers, not foreign exporters. Like any tax on consumption, it hits lower-income households hardest. (NY Fed, Liberty Street Economics) Estimate
And the legal foundation cracked: in February 2026 the Supreme Court struck down the IEEPA tariffs 6 to 3, cutting projected new tariff revenue from around $3 trillion to roughly $900 billion over a decade (refunds are still being litigated). Other tariffs survived; the defect was the emergency-powers vehicle, not tariffs as such. (SCOTUSblog; CRFB) Documented

Even at their peak, tariffs covered only about an eighth of what the 2025 tax cuts cost. A real revenue line, but a regressive, legally fragile one, not the pillar it was billed as.

Interest and credit

The downgrade everyone noticed, in context.

In May 2025, Moody's stripped the United States of its top credit rating, the last of the three major agencies to do so. The country had held Moody's highest rating since 1917.

Moody's cut the US from Aaa to Aa1 and named both drivers: rising interest costs and the roughly $4 trillion cost of extending the 2017 tax cuts. (Peterson Foundation; CNBC) Documented
Do not overstate this All three downgrades (S&P in 2011, Fitch in 2023, Moody's in 2025) blamed bipartisan dysfunction, not one party. Markets shrugged off all three. The dollar's reserve-currency status gives the US real room that other countries do not have, and much of the recent interest jump is the Federal Reserve raising rates to fight inflation, not new borrowing. The case here is about the trajectory, not an imminent crisis. And notice: by naming both the tax cuts and the spending, Moody's indicted neither a pure "spending" nor a pure "tax-cut" story. It indicted the refusal to deal with either.
The other side, in full

The strongest version of the case against this page.

If your parents pushed back, here is what they would say, stated as well as it can be, because each point has real merit.

"Congress writes the budget, not the president." True, and it is the central weakness of any presidential scorecard. The lone surplus came under divided government. The right frame really is "neither party balances and Congress drives spending," which is exactly where this page lands.
"The worst deficits were emergencies." True. The 2009 crisis and the 2020 pandemic response were bipartisan (CARES passed 96 to 0). Strip the shocks and the partisan gap narrows. What survives the filter is the boom-year deficit and the dead starve-the-beast theory.
"Republicans did cut spending in 2025." True, and meaningfully: roughly $2.5 trillion in offsets, including a major Medicaid cut, more than recent Democratic laws attempted. The rebuttal is only that the tax cuts were larger, so the net was still up $3.4 trillion, and the cuts fell hardest on the lowest incomes.
"The scary deficit numbers assume a tax hike as the baseline." Partly fair. About $3.9 trillion of the 2025 law's gross cost is just keeping existing rates. The real problem is inconsistency in the scoring, not that the figure is invented.
"DOGE was judged unfairly." Partly fair. It was aimed at a sliver of the budget. But it still overpromised by more than tenfold, and that gap is the story.

Conceding all of this does not weaken the conclusion. It is what makes the conclusion trustworthy, because it is what is left standing after the best objections.

What it comes down to

The debt is real. The villain is math, not a single party.

Here is the part that should matter most to anyone who hates waste, because it is the most honest and the least partisan. The debt problem is overwhelmingly structural, and it was set in motion long before 2025.

CBO's projection before the 2025 law already had debt reaching 156% of the economy by 2055, blowing past the World War II record around 2029. The law pushes that to roughly 176%. In other words, about four-fifths of the coming rise exists with or without it. (CBO Long-Term Budget Outlook) The engines are Social Security, Medicare, and the compounding interest on past borrowing, driven by an aging population. Social Security's main trust fund is now projected to hit automatic benefit cuts of about 22% around 2032, no matter what happens to tax rates. (CRFB)

These are the programs neither party will touch, because touching them is how you lose an election. That, not the usual story about one disciplined party and one spendthrift party, is the real reason the debt keeps climbing.

So keep the instinct. The waste is real, the debt is real, and being angry about it is justified. Just point that anger at the thing actually driving it. The evidence does not say Democrats budget better. It says no one since Clinton has, the theory that tax cuts impose discipline is dead by conservatives' own accounting, and the fix is the entitlement-and-interest math that both parties keep dodging. A brand that asks you to trust one team to fix it is the one claim the numbers will not support.

Sources

Where this comes from.

Load-bearing numbers come from nonpartisan scorekeepers. Links go to the specific report behind each claim.

The scoreboard and the historical record

Starve the beast and the 2017 tax cuts

The 2025 law (One Big Beautiful Bill Act)

DOGE, tariffs, and the credit downgrade

The long-term path

This page assesses federal spending and the national debt using public data from the Congressional Budget Office, the Treasury, the GAO, the Committee for a Responsible Federal Budget, and the major credit-rating agencies, as of mid-2026. Figures attributed as projections assume no future policy change and will move as policy and the economy do. Historical debt-to-GDP figures are rounded. Where credible sources disagree, the page says so. Corrections welcome.