Social Security Trust Fund: Full Benefits Through 2034, Then 81%

The combined Social Security trust fund — the reserve that backs U.S. retirement and disability benefits — is being drawn down. On the Social Security Administration’s 2025 best-estimate (“intermediate”) projection, the combined fund is sufficient to pay scheduled benefits in full through 2034. When the reserve is depleted, the program does not stop: continuing payroll-tax income would still cover 81% of scheduled benefits, declining to 72% by 2099, unless Congress changes the law. The reserve depletion date is the most-cited number in the debate, but it sits inside a range — 2032 in the high-cost scenario, 2051 in the low-cost scenario, with 95% of the SSA’s stochastic simulations falling between 2032 and 2039.

This page places that drawdown on a long timeline and documents what “depletion” does and does not mean, with the underlying dataset available for download. The narrower point established by the data is that the common framing — that Social Security is “going bankrupt” or “won’t be there” — does not match the Trustees’ own arithmetic: the system faces a real, sizable shortfall that grows the longer action is deferred, but not a fall to zero.

What the trust fund actually is

The common framing of Social Security’s finances is that the program is “going bankrupt” or “won’t be there” for younger workers. That framing is widely held, and it is not baseless: the system has run a cash-flow gap since 2021, when total cost first exceeded total income, and its cost has exceeded non-interest income every year since 2010. The reserve that bridges that gap is finite.

What the reserve is, precisely, is an accumulated balance held in special-issue U.S. Treasury securities. It was built up deliberately. Following the 1983 reforms, payroll-tax income was set above current benefit cost so that the baby-boom generation would pre-fund part of its own retirement; the combined reserve rose for a quarter-century, peaking in 2008 at 358% of one year’s cost — about 3.6 years of benefits held in reserve. Since then the ratio has fallen every year, to 169% (about 1.7 years) at the start of 2025. The fund is now being drawn down, as designed, to help pay the benefits of the generation that pre-funded it.

Measuring the reserve in years of benefits rather than dollars makes the trajectory legible: a reserve worth three-plus years of benefits in 2008, worth under two years today, and — on the best estimate — worth essentially nothing by 2034. At the end of 2024 the combined reserve stood at about $2.72 trillion, down from a peak near $2.91 trillion in 2021.

When the reserve runs dry — and how uncertain that date is

On the Trustees’ intermediate (best-estimate) assumptions, the combined OASI and DI reserve declines to about 9% of one year’s cost at the start of 2034 and is depleted during that year. The single date — 2034 — is the figure that dominates headlines, but it is a point estimate inside a distribution. The Trustees publish three scenarios: depletion in 2034 under the intermediate case, 2032 under the high-cost case, and 2051 under the low-cost case. Their stochastic model, which runs 5,000 simulations, places the depletion year between 2032 and 2039 in 95% of runs.

The chart above shows this as a fan: a single historical line through 2024, then three diverging paths. The width of the fan is the honest measure of how much is unknown — and it is the reason a precise single-year claim is more accurately read as a central estimate than as a certainty. The date has also moved between reports: last year’s report put combined depletion at 2035; this year’s moved it one year earlier, to 2034, largely because of the items documented in the methodology below.

What “depletion” does — and does not — mean

Depletion of the reserve does not mean benefits stop. Payroll taxes continue to flow in, and under current law that continuing income would cover 81% of scheduled benefits at the combined-fund depletion in 2034, a share that declines to 72% by 2099. Stated as a cut, depletion implies an automatic reduction of about 19% in 2034, growing toward 28% by century’s end — absent congressional action.

Three qualifications belong here, and they cut in both directions.

First, the “81%” combined figure assumes a legal change that has not happened. By law, OASI (retirement and survivors) and DI (disability) are separate funds, and money cannot move between them without legislation. The retirement fund alone is projected to deplete a year earlier — in 2033 — at which point OASI income would cover 77% of scheduled retirement benefits, not 81%. The combined 81% figure presupposes that Congress would authorize cross-fund transfers. For a retired worker on the average benefit of about $2,080 a month in early 2026, the OASI-only path to 77% implies a reduction on the order of $480 a month if nothing is changed before 2033.

Second, “the trust fund pays benefits through 2034” understates the financing question. The reserve is held in U.S. Treasury securities — that is, it is intra-governmental debt. Redeeming it to pay benefits requires the Treasury to raise the cash, through taxes, borrowing from the public, or spending reductions elsewhere. The reserve is a real legal claim that genuinely backs benefits through depletion; it is not, however, a pool of cash sitting apart from the rest of the federal budget. Both readings — “the fund is real” and “redeeming it has a general-budget cost” — are correct, and the gap between them is where much of the public disagreement lives.

Third, a 19–23% cut is materially serious even though it is not zero. Social Security is the majority of income for a large share of beneficiaries, and for that group an automatic one-fifth reduction would be consequential. “Not bankruptcy” is an accurate correction to the headline; it is not a statement that the shortfall is small.

Why the reserve is draining

The drawdown is demographic before it is anything else. The number of covered workers supporting each beneficiary held between 3.2 and 3.4 from 1974 to 2008; it has fallen since, to about 2.7 in recent years, and is projected to reach 2.3 by 2040 as the baby-boom generation retires and lower-birth-rate cohorts replace it at working ages. A pay-as-you-go system whose worker-to-beneficiary ratio falls runs a structural gap unless its tax rate or benefit formula adjusts.

Two further factors moved this year’s projection. The Trustees pushed back the year at which the long-run fertility rate is assumed to recover to 1.9 children per woman — from 2040 to 2050 — and lowered the assumed share of labor compensation in GDP. Separately, the Social Security Fairness Act, enacted January 5, 2025, repealed the Windfall Elimination Provision and the Government Pension Offset, raising benefits for some workers with non-covered pensions and adding to program cost. Together with the one-year advance of the valuation period, these changes raised the 75-year actuarial deficit from 3.50% of taxable payroll in the 2024 report to 3.82% in 2025.

What closing the gap would take

The Trustees quantify the size of the shortfall in two standardized ways, both descriptive rather than prescriptive. To keep the combined fund solvent across the full 75-year window starting in 2025, revenue would have to rise by the equivalent of an immediate, permanent payroll-tax increase of 3.65 percentage points (from 12.4% to 16.05%), or scheduled benefits would have to fall by the equivalent of an immediate, permanent 22.4% reduction applied to all current and future beneficiaries — or some combination. The arithmetic of deferral is the part most relevant to the depletion date: if action is delayed until the reserve is depleted in 2034, the same 75-year solvency would require a 4.27-point tax increase or a 26% benefit reduction — larger changes concentrated on fewer years and fewer cohorts.

These figures are the Trustees’ own illustrations of magnitude. They are not a menu of recommendations, and the historical record offers no single template: past solvency packages (1977, 1983) combined tax increases, coverage expansions, a higher retirement age, and benefit-formula changes phased in over years.

Key Dates to Watch

Counter-Arguments and Limitations

The date is a projection, not a measurement. Depletion in 2034 is the intermediate estimate; the official range is 2032 to 2051 and the stochastic 95% interval is 2032 to 2039. Economic growth, immigration, fertility, mortality, and disability incidence all feed the result, and a sustained surprise in any of them would move the date materially.

“Combined” is a legal fiction used for illustration. The OASI and DI funds are separate. Presenting them combined is standard SSA practice for showing the system’s overall status, but the operative dates for retirement specifically are the OASI numbers (2033, 77%).

The trust-fund accounting is genuinely contested in interpretation, not in arithmetic. The reserve balance and the redemption mechanics are not in dispute; what people disagree about is whether describing benefits as “funded through 2034” adequately conveys that redemption draws on the unified federal budget. This page documents both the balance and the mechanics so the reader can weigh the framing.

Projection changes are incremental, and the framing is robust to them. A 20% swing in the deficit or a few-year move in the date would not change the structural conclusion — a sizable, deferrable shortfall short of zero. A change large enough to overturn that conclusion (for example, restoring permanent solvency without legislation) is not present in any of the three official scenarios.

Common Misinterpretations

“Social Security is going bankrupt.” The program cannot “go bankrupt” in the corporate sense: payroll taxes continue regardless of the reserve, and they would fund 81% of scheduled benefits (combined) or 77% (OASI alone) at depletion. The accurate statement is a reserve depletion and an automatic benefit reduction, not a fall to zero.

“It won’t be there for me.” On current law and current projections, a worker retiring after 2034 would receive a reduced benefit, not no benefit. The size of the reduction depends on whether and how Congress acts before then.

“The trust fund is just IOUs / was raided.” The reserve is invested, by law, in interest-bearing U.S. Treasury securities — the same instrument held by private and foreign investors. It was not diverted to other spending in a way that removed the legal claim; it earns interest and is redeemed to pay benefits. The interpretive point above — that redemption draws on the general budget — is the substantive version of this concern.

“The 2034 date keeps slipping, so it’s not credible.” The combined date has in fact moved both later and earlier across reports (it was 2035 last year, 2034 this year). Movement reflects updated economics and law, not a lack of credibility; the multi-decade direction — a declining reserve ratio since 2008 — has been consistent.

Methodology and Sources

Frequently Asked Questions

Is Social Security going bankrupt?

No, in the sense that the program cannot run out of all funding while payroll taxes continue. On the Trustees’ best estimate, the combined trust fund reserve is depleted in 2034; at that point continuing payroll-tax income would cover 81% of scheduled benefits (77% for the retirement fund alone in 2033), declining to 72% by 2099, unless Congress acts. The accurate description is a reserve depletion and an automatic benefit reduction, not a bankruptcy.

Will I get nothing when the trust fund runs out?

On current law and current projections, a person receiving benefits after depletion would get a reduced benefit — on the order of 81% of the scheduled amount (combined) or 77% (retirement fund alone) — not zero. The size of any reduction depends on whether Congress changes the law before then.

Why do different sources give different dates (2033, 2034, 2035)?

The dates refer to different things. 2034 is the combined OASI+DI fund on the SSA’s intermediate assumptions; 2033 is the retirement fund (OASI) alone; 2035 was the combined date in the 2024 report. The Congressional Budget Office, using its own assumptions, has at times projected the retirement fund’s exhaustion a year or so earlier. All sit inside the official range of 2032 to 2051.

What is the difference between OASI and DI?

OASI is the retirement and survivors fund; DI is the disability fund. They are legally separate. OASI is the fund under pressure (depletion projected 2033, 77% payable). DI is not projected to deplete within the 75-year window; its reserve ratio is projected to rise. The widely cited combined “OASDI” date assumes the two could be merged, which would require legislation.

Is the trust fund “real,” or just IOUs?

Both descriptions point at the same fact. The reserve is held in interest-bearing special-issue U.S. Treasury securities — a real legal claim that earns interest and is redeemed to pay benefits. Because those securities are obligations of the federal government, redeeming them to pay benefits requires the Treasury to raise the cash from taxes, borrowing, or other spending. The reserve genuinely backs benefits through depletion; redeeming it is not free to the broader budget.

What would it take to close the gap?

The Trustees illustrate the 75-year shortfall as the equivalent of an immediate, permanent payroll-tax increase of 3.65 percentage points, or an immediate, permanent 22.4% reduction in scheduled benefits, or a combination. Delaying action until depletion in 2034 raises those figures to 4.27 points or 26%. These are descriptive measures of the deficit’s size, not recommendations; actual policy choices would combine measures and phase them in.

Conclusion

Social Security’s combined trust fund is being drawn down from a reserve worth about 3.6 years of benefits at its 2008 peak to, on the best estimate, essentially nothing by 2034. The reserve depletion is real, and the resulting automatic shortfall — about 19% combined, 23% for the retirement fund alone — is sizable and grows the longer action is deferred. It is also not zero: payroll taxes would continue to fund the large majority of scheduled benefits under current law.

The gap between “going bankrupt” and “a deferrable shortfall short of zero” is the substance of the policy debate, and the dataset behind this page is meant to let readers locate themselves in it precisely rather than rhetorically. This page documents the Trustees’ arithmetic and the range around it; it does not forecast which way Congress will act, and it will be updated as new Trustees and CBO figures are published.