Investigations · Economics & the Record

Is Social Security Going Bankrupt?

Both sides are spinning you about the same set of numbers. The 2026 Trustees Report, the CBO, and SSA’s own actuaries — checked against the doom, the denial, and the one-weird-trick fixes.

Social Security cannot “go bankrupt” the way the doom narrative implies — it is pay-as-you-go, and even after the trust funds run dry in 2032–2034, ongoing payroll taxes still cover roughly 76–83% of scheduled benefits. But the minimizers are wrong too: the 75-year deficit just hit 4.42% of taxable payroll — the largest in decades — an automatic cut of roughly 17–24% is now scheduled-law reality arriving as soon as late 2032, and even completely eliminating the payroll-tax cap closes only about half to two-thirds of the gap. Both parties are describing a different program than the one that exists.
Q4 2032
OASI (retirement) fund reserves deplete — 78% of benefits still payable after
83%
of benefits payable after combined-fund depletion in 2034
4.42%
of payroll — the 75-year deficit, largest in decades (2026 report)
≤67%
of the gap closed by even fully scrapping the payroll-tax cap

00The narrative being tested

Social Security solvency is unusual among political narratives: both sides are materially misleading the public, in opposite directions, about the same set of numbers. This page pulls apart three claim clusters:

Three honesty guardrails

  • Vintage discipline. The most dangerous errors on this topic are stale numbers, not false ones — the 2026 Trustees Report (June 9, 2026) superseded widely-quoted 2025 figures. Every number here carries its report year.
  • The “automatic cut” is a projection convention, not a statute. The Social Security Act does not actually specify how a post-depletion shortfall would be administered. “~17–24% cut” is the projected consequence under scheduled law, and it is presented that way.
  • Two official scorers disagree at the margins. CBO runs more pessimistic than the Trustees (deeper cuts, bigger deficit). Where they differ, this page shows the range, not a cherry-picked end.

01What “bankrupt” actually means

The doom claims all rest on one image: a fund that hits zero, and checks that stop. That is not how the program works. Social Security is primarily pay-as-you-go — today’s payroll taxes fund today’s checks. The trust funds are a buffer of past surpluses, not the engine. The buffer can run dry; the engine keeps running for as long as Americans earn wages.

The current official numbers, from the 2026 Trustees Report (released June 9, 2026 — the latest): the retirement fund (OASI) depletes its reserves in the fourth quarter of 2032, one quarter earlier than projected last year, with 78% of scheduled benefits still payable from ongoing revenue. On the combined retirement-plus-disability basis, depletion comes in 2034 with 83% payable.

What actually happens when the trust funds run dry

Share of scheduled benefits still paid by ongoing payroll taxes after reserve depletion

Retirement fund (OASI), Q4 2032 78% still paid –22% Combined funds (OASDI), 2034 83% still paid –17% CBO’s OASI scenario, 2034 76% still paid –24%

Sources: 2026 Social Security Trustees Report (SSA, June 9, 2026) for the first two bars; CBO Long-Term Projections for Social Security, pub. 61492 (June 2025) for the third. Reduction shares are the projected consequence of scheduled law; the Social Security Act does not specify the administration mechanism.

CBO’s independent projections confirm the shape while running deeper on the cuts: a 24% OASI reduction in 2034 (pub. 61492), and on the combined basis 21% in 2035 rising to 26% by 2055 (pub. 61270). The honest range across both official scorers: 76–83% of benefits keep flowing; the automatic cut is roughly 17–24%.

The claim

“Social Security is going bankrupt — it won’t exist when you retire.”

The record

False. Pay-as-you-go financing means benefits cannot go to zero while wages are earned. After depletion (Q4 2032 / 2034), 76–83% of scheduled benefits continue under every official projection.

The claim

“Social Security is a Ponzi scheme.”

The record

False. Pay-as-you-go intergenerational financing is the statutory design, reported publicly in exhaustive actuarial detail every year since 1941 — the defining features of a Ponzi scheme (concealment, promised returns from nonexistent enterprise) are absent. PolitiFact reached the same verdict on Musk’s version in 2025.

The claim

“You’ll get nothing when you retire.”

False — but the cut is real

Nothing in current law produces zero. But an automatic ~17–24% reduction around 2032–2034 IS the scheduled consequence if Congress does nothing — and that deadline is now inside a decade.

02The kernel the doomers are holding: the gap is real — and just got worse

This is where the minimizers get their verdict. The 75-year actuarial deficit in the 2026 report is 4.42% of taxable payroll — up sharply from 3.82% in 2025 and 3.50% in 2024. The Committee for a Responsible Federal Budget calls it the largest shortfall in nearly half a century. CBO’s estimate runs larger still: 4.91% of payroll / 1.66% of GDP (June 2025 vintage, which predates the 2025 tax law’s effects).

The 75-year deficit is growing, not shrinking

Long-range actuarial deficit, percent of taxable payroll, by Trustees Report year

3.50% 2024 report 3.82% 2025 report 4.42% 2026 report

Sources: SSA press releases for the 2025 and 2026 Trustees Reports; CRS IF13045 for the 2024 figure. Column heights scaled at 36 px per percentage point.

Why it worsened is the uncomfortable part for both parties: mostly legislation and demographics, not abstract drift. The 2024→2025 jump was driven by the Social Security Fairness Act’s WEP/GPO repeal (+0.14 points — a bipartisan benefit expansion with no offsetting revenue), delayed fertility recovery (+0.11), and lower earnings-growth assumptions (+0.12). The 2025→2026 jump reflects lower fertility and immigration assumptions plus the 2025 tax law. Congress’s most recent actions added to the shortfall.

“Lawmakers have many options for changes that would reduce or eliminate the long-term financing shortfalls. … Taking action sooner rather than later will allow consideration of a broader range of solutions and provide more time to phase in changes.” — Social Security Trustees, annual report summary (verified verbatim)
22.4% → 25.8%
The price of waiting, per the 2025 report (CRS IF13045, Table 3): closing the 75-year gap immediately took a 3.65-point payroll-tax increase or a 22.4% benefit cut; waiting until depletion raises that to 4.27 points or 25.8%. Under the worse 2026 baseline these grow to roughly 4.25 points / ~25% now, ~4.9 points / ~29% delayed (per CRFB’s analysis of the 2026 report).
The claim

“The program is fine — nothing needs to change.”

The record

False. The 75-year deficit is 4.42% of payroll — the largest in decades and growing — and recent legislation made it worse, not better. The automatic-cut deadline is inside ten years.

03The payroll-tax cap — and what the Musk comparison really shows

Social Security tax (6.2% from the worker, 6.2% from the employer) applies only to wages up to the taxable maximum: $184,500 in 2026 ($176,100 in 2025 — verified directly against SSA). Maximum worker-side contribution in 2026: $11,439. Every wage dollar above the line — and all non-wage income — pays nothing into the program.

The cap was designed to cover 90% of national covered earnings, and it hit that mark in 1983. It has been quietly eroding ever since — not because anyone voted to shrink it, but because earnings above the cap have grown faster than everyone else’s:

The quietly eroding cap

Share of national covered earnings below the taxable maximum

90% 86% 83% the 90% design benchmark (1983) 90% — 1983 ~86% — 2010 ~83% — 2020 1983 2010 2020

Sources: SSA Policy Brief 2011-02 (90% in 1983; ~86% in 2010); CBO Budget Options 58630 (~83% in 2020). X-axis positioned to scale (1983–2020). About 6% of workers exceed the cap in a given year; 20–25% will at some point in their careers (SSA).

The Sanders/Musk claim, checked

Sanders: “Elon Musk, a trillionaire, pays the same amount into Social Security as someone making $184,500.” PolitiFact (June 17, 2026): Mostly True — and the caveat cuts in a direction most readers won’t expect. The claim correctly describes the cap. But Musk likely pays less than a $184,500 wage-earner, not the same: he takes no Tesla salary and a $54,080 SpaceX salary, and his wealth growth is stock appreciation and investment income — which pays zero Social Security tax, cap or no cap.

The deeper point almost nobody makes: the cap is only half the story. Social Security taxes only wages, and the income of the very richest Americans mostly isn’t wages. Scrapping the cap would touch high salaries — surgeons, executives, athletes — not billionaires’ wealth.

The claim

“Musk pays the same into Social Security as someone making $184,500.”

Mostly true — arguably understated

The cap makes it true for wages — but he likely pays less, because investment income and stock gains pay no Social Security tax at all. The wages-only tax base, not the cap, is the deeper reason billionaires barely contribute.

04Scoring the fixes: what “scrap the cap” actually buys

SSA’s Office of the Chief Actuary publicly scores every serious proposal. Here are the cap provisions, verified from OACT’s own summary (scored against the 2025 report baseline of 3.82% of payroll; the right column rescales the arithmetic against the worse 2026 baseline):

How much of the 75-year gap each cap fix closes

SSA Office of the Chief Actuary provision scores, 2025 Trustees baseline (deficit = 3.82% of payroll)

100% = the whole gap. Nothing on this chart reaches it. Eliminate cap, no benefit credit 67% Tax above $250k, no credit 65% Tax above $250k, with credit 62% Tax above $400k, no credit 60% Tax above $400k, with credit 58% Eliminate cap, full benefit credit 48%

Source: SSA OACT, Provisions Affecting Payroll Taxes, summary (provisions E2.1, E2.5, E2.14, E2.17, E2.13, E2.2), scored against the 2025 Trustees baseline. Bar scale: 4.16 px per percentage point; the dashed line at 100% marks the full 3.82%-of-payroll gap. Against the 2026 baseline (4.42%) each share shrinks by roughly a seventh — e.g., full elimination falls from 67% to ~58% (arithmetic rescale, flagged as such).

Provision (OACT code)75-yr balance improvement% of gap closed (2025 baseline)vs. 2026 baseline
Eliminate the cap, no benefit credit (E2.1)+2.5567%~58%
Eliminate the cap, with benefit credit (E2.2)+1.8548%~42%
Tax earnings above $250k, no credit (E2.5)+2.5065%~57%
Tax above $250k, with credit (E2.14)+2.3862%~54%
Tax above $400k, no credit (E2.17)+2.3160%~52%
Tax above $400k, with credit — Social Security 2100-style (E2.13)+2.2158%~50%

Two verdicts fall straight out of this table. First: “scrap the cap and it’s fixed forever” is false. Even full elimination with no benefit credit — a break from the program’s always-linked taxes and benefits — closes about two-thirds of the gap at the old baseline, less under the new one. Realistic credited versions close roughly 42–62%. Second: the doom framing hides that cap changes are the single largest levers on the board. CBO’s scoring: raising the cap to re-cover 90% of earnings raises ~$670 billion over ten years and buys ~4 extra years of solvency; taxing earnings above $250k raises ~$1.2 trillion and buys ~13 years (CBO option 58630, 2022 scoring vintage — both baselines have since moved). “Nothing can be done” is as false as “one weird trick fixes it.”

The claim

“Scrap the cap and Social Security is fixed forever.”

The record

False. Full elimination closes ~48–67% of the 75-year gap depending on benefit credit — the biggest single lever available, and still not a full fix.

The claim

“Fixing it means destroying it — benefits must be slashed or the program privatized.”

The record

False. The Trustees’ own all-in benchmark: an immediate ~4.25-point payroll-tax increase, a ~25% benefit reduction, or any blend of smaller measures — with sooner action buying smaller, more gradual changes. The menu is public and scored.

05The kernels of truth — all four, stated plainly

06The argument, assembled

07Sources

Method & sourcing standard. Research run July 10–11, 2026 through a five-angle deep-research process (Trustees/depletion; CBO cross-check; cap mechanics; OACT fix scoring; both-directions spin). 25 top claims each faced three independent adversarial verification votes: 14 confirmed, 11 refuted — most refuted claims were accurate 2025-report figures presented as current, killed for staleness after the June 9, 2026 report superseded them. Cap and fix-scoring figures were then verified by direct fetch of the primary sources (SSA contribution-base page, OACT provisions summary, SSA Policy Brief 2011-02, CBO option 58630, PolitiFact). Vintage discipline: every figure carries its report year; OACT scores carry their 2025 baseline with the 2026 rescale flagged as arithmetic, not an OACT publication; CBO’s 4.91% predates the 2025 tax law. Ideologically aligned organizations (CBPP left; CRFB deficit-hawk) were used for discovery or corroboration only. Open questions flagged, not hidden: how a post-depletion shortfall would legally be administered (the Act is silent); OACT’s rescoring against the 2026 baseline; CBO’s post-2025-tax-law update.
Trustees Reports & SSA (depletion, deficit, cap)
  • 2026 Social Security Trustees Report — highlights & summary — ssa.gov · summary
  • SSA press release, June 9, 2026 (2034/83%; Q4 2032/78%; 4.42%) — ssa.gov
  • SSA press release, June 18, 2025 (2025-report reference figures) — ssa.gov
  • SSA, Contribution & Benefit Base ($184,500 / $176,100; fetched live 2026-07-10) — ssa.gov
  • SSA Policy Brief 2011-02 (90% in 1983 → ~86% in 2010; ~6% of workers above cap; 20–25% career incidence) — ssa.gov
  • SSA OACT, Provisions Affecting Payroll Taxes + summary PDF (E2.x scores) — ssa.gov · summary PDF
CBO & CRS (independent scoring)
  • CBO pub. 61492, Long-Term Projections for Social Security (June 2025; 4.91% of payroll; 24% OASI cut) — cbo.gov
  • CBO pub. 61270, Long-Term Budget Outlook (March 2025; 21%→26% combined-basis cuts) — cbo.gov
  • CBO Budget Options 58630 (cap-raise scoring: ~$670B / ~$1.2T over 10 years; ~83% of earnings under cap in 2020) — cbo.gov
  • CRS IF13045, Social Security: What Happens When the Trust Funds Run Out? (Table 3 delay costs) — congress.gov
Fact-checks & corroboration
  • PolitiFact (June 17, 2026), Sanders/Musk cap claim — Mostly True — politifact.com
  • PolitiFact (March 19, 2025), Musk “Ponzi scheme” claim — politifact.com
  • CRFB, Analysis of the 2026 Trustees Report (deficit-hawk-leaning; used for 2026 delayed-action figures pending primary verification) — crfb.org
  • CBPP on the 2025 Trustees Report (left-leaning; corroboration only) — cbpp.org

Related on this site: Are “Illegals” Using Social Security? The Money Runs the Other Way — the companion investigation on who actually pays into the trust fund.