Dossier mode
The Interest Trap
The same investigation, restaged one beat at a time. Drive it with the arrow keys, space, or autoplay. Nothing is cut from the piece — long runs are split across frames. Read the full investigation or open the The Austerity Myth hub.
The Interest Trap.
Interest on the national debt just passed the entire Pentagon budget — the fastest-growing line in the federal ledger. The alarm is real. So is the arithmetic way out. Both parties have decided not to take it.
The deficit-interest spiral is real and Treasury itself calls the path 'unsustainable' — but the shortfall is substantially a revenue choice, a concrete fix exists (let the 2017 cuts expire; revert to Clinton-era rates), and both parties refuse it while pushing the cost onto the programs ordinary people depend on.
Graded against CBO and Treasury, not a napkin. The milestones (interest > defense) are FACT; the scary figures (6% → $2.4T, 'bigger than Social Security') are corrected to today's ~$970B at a ~3.6% rate and reframed as the mid-2030s trajectory; the Clinton-rates fix is real but, told honestly, is a middle-class tax increase too.
Interest on the debt just passed the Pentagon — the first time in modern U.S. history.
In fiscal year 2025, net interest on the federal debt reached about $970 billion, overtaking national defense spending (about $917 billion) for the first time in modern history. Interest is now the fastest-growing part of the budget — money spent before a single program dollar. Gross debt stood at roughly $37.85 trillion in late 2025, climbing about $2.2 trillion a year, so the reader's '$40 trillion' is rounding up by about a year.
'10-year at 6% → $2.4 trillion a year' is a forecast, not today's bill.
The $2.4 trillion applies 6 percent to the whole $40 trillion debt. But the government pays about 3.6 percent on average, because most of the debt is locked in at older, cheaper rates and rolls over only gradually — so today's actual net-interest bill is about $970 billion, not $2.4 trillion. Where the napkin is right is the direction: CBO projects net interest roughly doubling to about $2.1 trillion by the mid-2030s. Not wrong — early.
'Bigger than the entire Social Security disbursement' — not yet; it just passed defense.
As a present-tense claim it doesn't hold, and we correct it even though it points the right way. Social Security outlays run about $1.5 trillion and remain the single largest line in the budget; net interest, at about $970 billion, is below that. What interest passed is defense. 'Bigger than Social Security' is true only of the high-rate scenario or the later-2030s trajectory. Accurate version kept: interest now exceeds the Pentagon and is closing on everything else.
The shortfall is substantially a revenue choice, not only runaway spending.
Federal revenue has run around 17 percent of GDP while outlays run several points higher, and independent analysts attribute a large share of the long-run rise in debt-to-GDP to the unpaid-for Bush-era and 2017 tax cuts. Honest counterweight: an aging population drives Social Security, Medicare, and interest upward regardless of tax policy, so it is not purely a revenue story. But the repeated choice to cut taxes at the top and then call the gap a spending emergency is a choice.
For broader context: the top tax rate was 91% under Eisenhower — the rich paid more then, but nowhere near 91%.
The baseline the modern debate forgot. The top marginal rate peaked at 94 percent in 1944 and sat at 91 percent through the Eisenhower years (92 percent in 1952–53); today it's 37 percent — cut by more than half. The honest fence: almost no one paid 91 percent on all their income. It applied only above a very high threshold, amid abundant shelters, so the top 1 percent's effective income-tax rate in the 1950s was near 17 percent (Tax Foundation), and the effective rate on top-bracket income ran about 49 percent (Tax Policy Center) vs ~35 percent today. The rich paid materially more then — just not 91 percent. And revenue stayed near 17 percent of GDP across every regime, so the realistic lever today is reversing the Bush/Trump cuts, not restoring 91 percent.
Just letting the 2017 Trump tax cuts expire would save about $4 trillion over a decade.
The concrete, scoreable piece of the exit — and it requires no new taxes, only letting an existing law lapse as written. CBO and the Peterson Foundation put the 10-year cost of extending the expiring TCJA provisions at roughly $4.0 trillion (Tax Foundation ~$4.5T including related provisions; JCT ~$3.4T for the individual provisions alone). Before anyone debates Clinton-era rates, not renewing the Trump cuts closes a $4 trillion hole. Congress chose to extend them instead.
'Clinton rates plus adjustments → $10–12 trillion' is reachable — but it's a middle-class tax increase too.
The menu stacks: let the Trump cuts expire (~$4T), restore the 35 percent Clinton-era corporate rate from today's 21 percent, tax capital gains closer to ordinary income, revive a stronger estate tax. Over a decade, $10–12 trillion is a defensible high-end figure. Two caveats keep it from being a free lunch: full Clinton-era individual rates hit lower income thresholds, so this raises middle-class taxes too; and 'a few adjustments' is doing heavy lifting. The number is real; who pays it is why it never moves.
'No Republican is going to do that. Trump isn't going to do that' — correct, and the refusal reaches across the aisle.
On the GOP half it's the record: the 2017 cuts passed party-line, 2025 extended and expanded them, Trump campaigns on further cuts. The fuller truth is sharper: Democrats pledged not to raise taxes on anyone under $400,000, which rules out full Clinton-era reversion from their side too (it reaches below $400k). Neither party will do the thing that closes the gap on revenue — leaving the easier lever, cutting the programs people depend on.
The line we hold.
- FACT: net interest ~$970B in FY2025, passing national defense (~$917B) for the first time in modern history; gross debt ~$37.85T; Treasury's FY2025 report calls current policy 'unsustainable'; letting the 2017 cuts expire saves ~$4T over a decade.
- CORRECTED: the '6% → $2.4T' figure is a scenario, not today's bill — the effective rate is ~3.6% and today's interest is ~$970B. Interest is NOT yet bigger than Social Security (~$1.5T); that's the trajectory (CBO: ~$2.1T interest by the mid-2030s).
- PROBABLY TRUE (with caveat): full Clinton-era reversion plus base-broadening can reach ~$10–12T over a decade — but full Clinton individual rates raise middle-class taxes too, not only the rich. We don't launder it as a costless, rich-only fix.
- FACT (position): both parties refuse the revenue fix — Republicans by cutting further, Democrats by pledging no increases under $400,000 — which is why the pressure falls on the safety net.
Why it matters.
The interest is real, it's rising, and it has to be paid. There are only two sides to that ledger — collect more or spend less — and both parties have ruled out the revenue side, so the pressure flows to the only valve left open: the programs ordinary people paid into. The deficit built by tax cuts for the top becomes the argument for cutting Medicaid, food assistance, and eventually Social Security for everyone else. A large and growing share of that interest is, in effect, a transfer to the bondholders — the wealthy and foreign governments who had the spare cash to lend the government the taxes it chose not to collect. The debt isn't a looming bankruptcy; it's a slow squeeze, and the squeeze is the point. That inversion — spending is a luxury, low taxes at the top are a law of nature — is the austerity myth itself.
Help us fill these lines.
- OpenA full, line-by-line revenue score of a complete Clinton-era reversion plus the specific base-broadening moves, separating what the rich pay from what the middle class pays.
- OpenHow fast the average interest rate on the debt actually rises as cheap older bonds roll over into higher-rate new ones — the real driver of whether CBO's mid-2030s projection holds.
- OpenWhich programs are first in line when interest crowds the budget, and whether any revenue measure — even a narrow one on the very top — ever reaches a floor vote.