What is the Social Security trust fund, and what happens when it runs out?

Most people have heard some version of the same headline: Social Security is running out of money. The implication, whether stated or just implied, is that the program is going to disappear, and anyone who hasn't retired yet will be left with nothing.

That's not what's happening. The trust fundA savings account made of government bonds. It holds what was left over in the years when Social Security took in more than it paid out. is the savings account the program built up over decades of collecting more than it paid out, and the retirement half of it is projected to hit zero in the fourth quarter of 2032. When it does, the program doesn't stop. It keeps paying benefits, but only with the money coming in from current workers' payroll taxes. The Trustees' math says that will cover about 78% of what's been promised. That means an automatic cut of roughly 22% to every retirement check — unless Congress acts before then.

The question isn't whether Social Security survives. It's what kind of cut or tax increase we end up with, and when.

What's in the rest of this article

• Where Social Security came from — the Depression conditions behind the 1935 law, who actually paid for it at the start, and the changes that made it the program people know now.
• How the money actually moves — why your payroll taxA tax taken out of every paycheck, with a matching amount paid by the employer. Social Security’s share is 6.2% from the worker and 6.2% from the employer. is not sitting in an account with your name on it, and what the two trust funds really are.
• What the 2026 Trustees Report says — the date that moved, the three reasons it moved, and the numbers that are confirmed against the report itself.
• What is actually draining it — every force pulling money out or keeping it from ever coming in.
• So did the government raid it? — the straight answer, in three parts.
• Would raising the capThe yearly income ceiling above which no Social Security payroll tax is taken. In 2026 it is $184,500. Its official name is the taxable maximum. fix it? — what the government's own scorekeeper says each version is worth.
• What about just putting it in the stock market? — why the returns argument is half right, and what Chile's forty-year experiment actually produced.
• What a switch would actually take — the transition bill, and the two things the pitch leaves out.
• How Congress fixed it last time — the 1983 deal that is everybody's template, and what each side had to give up.
• What is on the table now — what the advisers and a sitting senator expect, and the bills already written.
• What it means for your own plan — whether to claim early to beat the cut, and what a planner says to stress-test instead.
• What is still unsettled — the figures nobody has confirmed yet.

Where Social Security came from

Before 1935 there was no national retirement check. If you got too old to work you lived on savings, family, charity, or a state pension. The Social Security Administration's own history says only about 5% of elderly Americans received a retirement pension in 1932, and that by 1934 more than half of the elderly did not have enough income to support themselves. Thirty states had an old-age pension law by 1935, but only about 3% of the elderly collected.

President Roosevelt signed the Social Security Act on August 14, 1935. Title Two created the federal retirement benefit. Title One sent federal money to the states to pay cash to people already old and poor. Title Nine taxed employers to fund state unemployment insurance.

Who paid depends on which part you mean. The retirement program ran on two matching taxes, starting on wages paid after the last day of 1936: 1% out of the worker's pay, and a matching 1% from the employer. No government money at all. SSA historian Larry DeWitt calls it a payroll tax "imposed equally on employers and employees (with no government contribution)." The unemployment tax was the one paid by employers alone, and only by those with eight or more workers, and it never touched retirement checks. The money for the old and poor came out of general federal revenue. So one part of the Act was funded entirely by businesses, and the retirement program was split down the middle between workers and their bosses. Coverage was narrow: a little over half the workforce. Farm workers, domestic workers and the self-employed were left out, and only the first $3,000 of a year's wages was taxed at all.

The 1935 law did not schedule monthly checks until 1942. The amendments signed on August 10, 1939 moved that up to 1940 and added benefits for a retired worker's spouse and children, plus survivors' benefits for the family of a worker who died. Ida May Fuller, a retired legal secretary from Ludlow, Vermont, received the first Social Security check ever issued, on January 31, 1940, for $22.54. She had paid $24.75 in Social Security taxes over three years of work. She lived to 100 and collected $22,888.92.

On August 1, 1956 Congress added cash benefits for disabled workers aged 50 to 64, which is where the disability fund started. In 1972 it made benefit increases automatic from 1975, tied to consumer prices. Before that every raise took its own act of Congress: Fuller's check sat at $22.54 until the 1950 amendments took it to $41.30.

Social Security has paid for benefits the same way since the checks started. Fuller's came out of taxes collected from people working at the time, not out of the $24.75 she paid in. DeWitt writes that the trust funds "have never been either fully funded or on a strict pay-as-you-go basis" and have always held a partial reserve. That reserve is the trust fund.

How the money actually moves

Social Security is a pay-as-you-go system. The payroll taxes you and your employer each pay — 6.2% apiece, listed on your pay stub as FICAThe Federal Insurance Contributions Act, the law behind the payroll tax. On a pay stub, the FICA line covers Social Security and Medicare. or OASDIThe two funds together, OASI plus DI. It is the official name for what most people just call Social Security. It shows up on pay stubs. — don't sit in an account with your name on it. They flow directly to the people collecting benefits right now. When the program takes in more than it pays out, the surplus goes into what are called trust funds, which are reserves held in U.S. Treasury securities that earn interest. When it pays out more than it takes in, it draws from those reserves.

There are two trust funds, and they're in very different shape.

The big one — the one the headlines are about — is the Old Age and Survivors Insurance (OASIOld-Age and Survivors Insurance. The retirement half of Social Security. Its fund pays retirees, their husbands, wives and children, and the families of workers who died.) fund. It pays retirement checks, spousal benefits, and survivor benefits to widows, widowers, and their children.

The other is the Disability Insurance (DIDisability Insurance. The smaller half of Social Security. Its fund pays workers who become too sick or hurt to work before retirement age.) fund, which pays workers who become disabled before retirement age. This one is in genuinely good shape: the Trustees project it can pay full scheduled benefits for the whole 75-year projection window.

Together the two funds held $2.56 trillion at the end of 2025, down $160 billion over the year — that combined figure is straight out of the 2026 Trustees Report. The split between them is where you have to be careful. Erik Goodge and the retirement planners working from the report's tables put the retirement fund at about $2.34 trillion and the disability fund at about $223 billion, with the retirement side running roughly $200 billion short in 2025 while the disability side ran a surplus. Those splits are consistent with the combined number the Trustees publish, but they are the advisers' reading, not the Trustees'. Treat the $2.56 trillion as settled and the split as very likely.

The two funds are legally separate, but Congress has combined them before. If it does so again, the disability fund's savings get spent on retirement checks too, which pushes the date out to 2034 — the Trustees say so directly — at which point a bigger share of benefits would be payable. Some financial advisers suggest planning to that combined date, since moving money between the funds is relatively easy for Congress to do.

What the 2026 Trustees Report says

Every year, the Social Security Trustees release a financial audit of the program. The 2026 report moved the projected depletion of the retirement fund forward by one quarter, into the fourth quarter of 2032. One quarter sounds small, and the date is the part that gets reported. The number underneath it moved much more: the 75-year gap went from 3.82% of the nation's wages in last year's report to 4.42% in this one. A date that slips three months and a hole that grows by that much in a single year are the same news told two ways, and the second way is the bigger story. The report points to three reasons.

The first two are demographic. The assumed long-term fertility rate was lowered from 1.9 children per woman to 1.75. And both historical and projected levels of net immigration were revised downward. Fewer births and less immigration mean fewer future workers, which means less payroll tax revenue flowing into the system.

The third is the One Big Beautiful Bill, which made the lower income tax rates from the 2017 Tax Cuts and Jobs Act permanent. It also expanded the standard deduction and added a temporary additional deduction for taxpayers over 65. The effect, as financial adviser Scott Caufield explains, is that the bill reduced the taxes collected on existing Social Security benefits — and those taxes flow back into the trust fund. Less coming back in means the fund empties sooner.

Once the reserves are gone, the program can only pay benefits out of incoming payroll taxes. The Trustees put that at 78% of scheduled benefits — the 22% cut. Senator Bill Cassidy, interviewed by Erin Moriarity, cited a figure of "as much as 28.5%." He did not say where it came from, and it does not match the 22% the Trustees give for the automatic cut. It is in the report, though — as the answer to a different question. A 28.5% cut is what it would take to close the whole 75-year gap if Congress waits until 2034 to act. That is not the automatic cut at depletion; it is the size of the deliberate fix after eight more years of delay. So the number is real and the senator is quoting his own government. He just did not tell the interviewer which of the two he meant, and the difference decides whether you are hearing about something that happens to you or something somebody chooses.

The Medicare Part A hospital trust fund is on a similar timeline. Medicare comes in parts, and the difference matters here. Part A is the hospital part, and your payroll tax already paid for it — there is no monthly bill. Part B is the doctor-and-outpatient part, and you pay a monthly premium for it out of your Social Security check. Part D is the prescription drug part, and you pay a premium for that too. Only Part A has a trust fund that can run dry.

The 2026 Trustees Report puts Part A's depletion in the second quarter of 2033, with 89% of scheduled benefits payable at that point. Scott Caufield says the same thing, and argues that when Congress addresses that shortfall, changes to Part A are very likely to be paired with changes to Parts B and D — the parts retirees do pay for — meaning higher premiums or thinner coverage for outpatient care and prescriptions.

What is actually draining it

The three reasons above are the ones that moved the date this year. They are not the whole story. Here is everything that is pulling money out of the system, or keeping money from ever arriving.

The 2025 tax law — money that stopped coming back in. This one needs a step most people skip. Since 1984, if your income is above certain levels you pay income tax on part of your own Social Security check, and some of that tax money goes straight back into the trust fund. It is one of the fund's three sources of income. Be careful with the split, because it is where most write-ups get it wrong: the tax on the first half of a benefit goes to Social Security, and the tax on anything above that half goes to Medicare's hospital fund instead. The One Big Beautiful Bill, signed in July 2025, cut income taxes for people over 65 by adding an extra deduction for them, on top of making the 2017 tax rates permanent. Seniors owe less tax. That also means less of that tax flows back into the fund. The Committee for a Responsible Federal BudgetCRFB stands for the Committee for a Responsible Federal Budget, a nonprofit group that studies and comments on federal budget choices. puts the loss at roughly $30 billion a year, about $168.6 billion through 2034, and the Trustees name the law as one of the three reasons the depletion date moved. Nothing was taken out of the fund. Money that was scheduled to come in stopped coming in.

Fewer babies. The Trustees lowered their long-term assumption from 1.9 children per woman to 1.75. Every child not born is a worker not paying payroll tax in twenty years.

Less immigration. The Trustees revised both past and future net immigration down. Same effect, sooner: immigrants who work on the books pay payroll tax, and fewer of them means less money in.

The cap has been leaking since 1983, and nobody voted for it. This is the biggest one, and almost nobody talks about it. The cap is the income ceiling above which you stop paying into Social Security — $184,500 in 2026. If you earn a dollar above that, no payroll tax comes out of it, and it is not counted when your benefit is worked out either. When Congress set the modern version in 1983, it picked a level meant to cover 90% of every wage dollar in the country, and then let it rise automatically with average wages. That formula assumed pay would grow at roughly the same rate up and down the ladder. It didn't. Pay above the cap grew faster than pay below it, so year after year a bigger slice of the nation's wages sat above the ceiling, untaxed. The share covered fell from 90% to about 82.5% by 2000 and has stayed near there since. Stephen Goss, Social Security's Chief ActuaryA person whose job is to use math to estimate future money: how many workers, how many retirees, how long people live, and what that costs., told the Senate Budget Committee that this means payroll tax has come in more than 8% lower than the 1983 plan expected — and that this shrinking taxed share "explains most of" the program's higher costs and earlier depletion compared to what was projected in 1983.

That last one deserves a second look, because it answers the question most people are really asking.

So did the government raid it?

There are three parts to the answer.

Nobody took the cash out and spent it on something else. The Social Security Administration's own history office addresses this directly: the trust fund has never been moved into the government's general fund, and the way the program is financed has never changed. The raid story is false.

But the fund never held cash in the first place. It can't. Every surplus dollar is lent to the U.S. Treasury by law, and the fund gets back a special government bond that earns interest. The Treasury spends that cash on whatever else the government is doing that year. So the government does owe the fund money, and that debt is real and has been paid every time. If you call lending money to the Treasury a raid, then yes. If you mean someone drained an account, no.

What actually hurt was not a raid. It was money that never arrived. That is the cap leak above, and it is the real answer to the suspicion that the system was never built right for the way the country changed. The 1983 fix was built for a country where pay grew evenly and people had more children. We got neither. Nobody stole anything. The design just quietly stopped collecting what it was written to collect.

Two things the government did do that people file under "raid": it started taxing benefits in 1984, and raised that tax in 1993. The 1984 tax sends money into Social Security's own fund — in, not out. The 1993 increase was written to go to Medicare's hospital fund, so it never helped Social Security at all. And the 2025 tax cut shrinks the 1984 money going in.

Would raising the cap fix it?

You hear a version of this everywhere: lift the cap to $400,000, or scrap it entirely, and the problem is solved. It is the most popular answer there is, and it is not what the government's own scorekeeper says.

First, what "the gap" actually is, because every number below is a share of it.

Add up everything Social Security is scheduled to pay out over the next 75 years. Add up everything it is scheduled to collect. It comes up short. That shortfall is the gap.

The Trustees size it three ways. It is one hole measured three times, not three problems. $29.3 trillion in today's dollars. Or 4.42 cents out of every wage dollar the country will earn between now and 2100. Or about 1.5% of everything the country produces.

That is the hole. Now the useful part, which the Trustees also print and almost nobody quotes: exactly how much medicine it takes to close it. Do any one of these in 2026 and the program is solventHaving enough money to pay what you owe when the bill comes due. for the full 75 years.

• Raise the payroll tax from 12.40% to 16.65%. That is the 6.2% out of your pay and the 6.2% from your employer becoming about 8.3% each.
• Or cut every benefit, for people already retired and people not yet retired, by 25.2%.
• Or protect everyone already retired and cut only people retiring from here on, by 30.3%.
• Or any mix of the three.

Waiting makes the dose bigger, and the Trustees price that too. If nobody acts until 2034, the same list becomes a payroll tax of 17.30%, or a 28.5% cut for everybody. Four and a quarter cents on the wage dollar now. About five cents in eight years.

Everything below is a piece of that dose. When a proposal "closes 67% of the gap," it covers two-thirds, and something else still has to cover the rest.

Social Security's Office of the Chief Actuary prices these proposals one by one and publishes what share of that gap each one closes. Measured against the 2025 report, where the gap stood at 3.82% of all taxable wages:

• Tax earnings above $400,000, and pay bigger checks on the extra. This is the shape of a bill called Social Security 2100, written by Congressman John Larson: closes 58% of the gap.
• Tax earnings above $250,000, and pay bigger checks on the extra: 62%.
• Tax earnings above $250,000 and pay nothing extra in benefits: 65%.
• Scrap the cap completely and pay nothing extra in benefits: 67%. That is the best any cap change does.
• Scrap the cap completely but credit those earnings toward bigger checks: 48%.

So the answer is no. Not one version of it gets there. The most aggressive one on the list — tax every dollar anyone earns and give the high earners nothing back for it — closes about two-thirds of the gap.

Two things make that picture worse rather than better. Those figures are scored against the 2025 report; the 2026 report widened the gap to 4.42% of wages, and the actuaries have not re-scored the options yet, so the real shares today are lower than the ones above. And catching up is harder than it sounds: back in 2024, when the cap was $168,600, Goss put the level needed to cover 90% of wages again at about $350,000 — more than double.

So what would fix the whole gap in one move? Raising the payroll tax rate itself, from 12.4% to 16.4%, closes 102% of it. That is not a tax on high earners. That is roughly two more cents out of every dollar from every worker and every employer in the country. Which is exactly why Congress has not done it, and why every serious plan is a combination rather than a single lever.

What about just putting it in the stock market?

This is the other thing you hear constantly, and the first half of it is true. The trust fund is required by law to hold government bonds. Over the last forty years the stock market returned far more than those bonds did. So if the money had been sitting in an index fund all along, wouldn't we be fine?

So take the question at its word: what if the surplus had gone into an index fund from the start, instead of into government bonds?

It would be bigger. There was a real pile to invest — the fund peaked at about $2.9 trillion in 2021 — and stocks beat those bonds over that stretch, so a fund invested that way would hold more today. Nobody can tell you exactly how much more without picking a starting year and a mix, which is how that argument usually gets fudged.

It would not be solved. That is the part the argument skips. The pile was never the program's money — it was only the leftover from the years Social Security collected more than it owed. Everything else went straight out as checks the same month it came in. The hole to fill is about $31 trillion in today's money, roughly ten times the pile at its biggest. A better return on a tenth of the problem does not fix the problem.

And it would have been selling into the crash. A trust fund is not a retirement account that can wait for a recovery. It has to pay 70 million checks on schedule. In 2009 it would have been selling stocks at the bottom to do it.

There is also the question of what to do from here, and that is where it gets stark. The surplus is being spent down now. It reaches zero in late 2032. A 10% return on nothing is nothing.

Social Security's actuaries have actually priced this, and the result needs both halves to be honest. Put 40% of the fund's reserves into stocks, phased in over fifteen years, and assume a healthy 5.8% real return every year. It does help a little over the 75 years — worth 0.49% of payroll. By the 75th year it does nothing at all: 0% of the shortfall in that year. And the actuaries refuse to print a percentage for the long-range column, because the help depends entirely on how much money is in the fund to invest. Their own words: if the program ends up purely pay-as-you-go, with reserves just above zero, then investing in stocks "would have no effect on the actuarial balance." So stocks are not nothing. They are a little, early, and then nothing — because zero times any return is still zero.

There is also a lesson from the country that actually ran the experiment. Chile replaced its state pension in 1981 with private individual accounts: 10% of every worker's wage, invested in funds. The funds performed well — the oldest one has reported roughly 10% a year in real returns. And it still did not give people enough to retire on. For Chileans who retired between 2015 and 2023, the median pension their own account paid for was 26% of their old wage for men and 11% for women. In March 2025 Chile kept the accounts and bolted on an 8.5% employer contribution and a shared insurance pillar to hold up the bottom.

The reason is the part the returns argument leaves out. Your account only grows if money goes into it, every month, for forty years. Raising kids, caring for a parent, getting laid off, working off the books — each gap is a hole in the account that no rate of return fills. A defined benefit is not better at investing. It is better at not caring whether you had a bad decade.

What a switch would actually take

Switching to private accounts runs into three problems.

Somebody still has to pay today's retirees. The moment your payroll tax goes into your own account instead, the money paying your parents' checks disappears. Not shrinks — disappears, the same month you start. Those checks are owed and they don't stop. That gap is what people mean by the transition cost, and it has to be borrowed. When President Bush proposed personal accounts in 2005, his own Office of Management and BudgetOMB stands for the Office of Management and Budget, the White House office that prepares the president’s budget and prices proposals. put the first ten years at $664 billion, or $754 billion once you count the interest on borrowing it. Estimates over longer windows and bigger carve-outs ran from about $1 trillion to more than $5 trillion. Nobody on either side disputed that the number was large; they disputed what it bought.

It is not only a retirement program. Social Security also pays workers who become disabled before they can retire, and it pays the families of workers who die young. A twenty-six-year-old with two children who is killed in a crash has not built up an account worth anything. The current system pays his family anyway. Any switch has to say what replaces that, and it is usually the part left out of the pitch.

It does not fix the hole by itself. This is the actuaries' own conclusion, not an opinion: individual-account provisions of this kind "generally do not, in themselves, improve the solvency of the Social Security trust funds." They change who carries the risk. Closing the gap still takes the same unpopular levers — more money in, or smaller checks out.

How Congress fixed it last time

The last time Social Security faced a crisis like this was the early 1980s. The retirement fund was months from running dry — estimates had it missing checks as early as August 1983. Congress didn't act until the deadline was right in front of them.

The fix came from the Greenspan Commission, a bipartisan body chaired by Alan Greenspan. The deal was ultimately a handshake between President Ronald Reagan and Democratic House Speaker Tip O'Neill. It raised the full retirement ageThe age when you can claim your whole earned check. Claim earlier and every check is smaller for life. Claim later and it is bigger. from 65 to 67, phased in so slowly that it only fully took effect in 2026 — more than four decades later. It taxed Social Security benefits for the first time, applying to up to 50% of benefits for higher-income recipients. (In 1993, that rose to 85% for the highest earners.) It expanded who paid into the system, bringing in federal employees and nonprofit workers. And it accelerated already-scheduled payroll tax increases.

Neither party got everything it wanted. Republicans accepted tax increases. Democrats accepted a benefit cut in the form of a higher retirement age. The political pain was shared.

What is on the table now

Both the advisers and Senator Cassidy expect Congress to follow the same pattern this time — waiting until the deadline is close, likely 2032 or 2033, before acting. Cassidy says: "Most likely Congress will do nothing." He says borrowing is "easier… than to take a tough political decision."

Those two things sound like they cancel out, and they don't, so here is the piece that usually gets left out. The reason an empty fund forces a cut is not that the money runs out of the country. It is that the law only lets Social Security spend what its own fund holds. Nothing else in the budget works that way — when the government wants a fighter jet or a farm subsidy and the money isn't there, it borrows. Social Security is walled off from that, on purpose, so it can never quietly become a line item somebody has to defend every year.

Congress can take that wall down in a sentence. It could vote to let Social Security draw on general revenue, or simply hand the fund enough Treasury money to keep paying full checks, and the automatic cut would never happen. That is what "doing nothing" would most likely turn into: not 70 million people taking a 22% cut, but one more thing paid for with borrowed money, on top of a debt that crossed $40 trillion in August 2026 and now costs about a trillion dollars a year just in interest. No serious person calls that a fix. It is the path of least resistance, and it is the one Cassidy is warning about when he says borrowing is easier. Worth knowing before somebody tells you the checks are guaranteed to be cut — what is guaranteed is that the wall gets tested.

When Congress does act, Scott Caufield predicts the fix will include higher payroll taxes and probably the elimination of the wage cap without a matching benefit increase for high earners. Social Security was designed as an earned benefit: you pay in on your wages, and your benefit is tied to what you paid. Removing the cap without adjusting the benefit formula turns it into something closer to a transfer. Caufield says the political climate seems more comfortable with that framing today than it used to be.

There are at least six bills in Congress addressing Social Security, according to financial planner Erik Goodge. They fall into three groups. The first are process bills, like the Promise Act, which don't change a single tax or benefit — they hand the drafting to an outside commission and force Congress to vote on whatever comes back. Political cover for a politically painful decision. The second group raises revenue from high earners and expands benefits, lifting or removing the payroll tax cap and, in some cases, extending the tax to investment income. The third group — ideas not yet introduced as legislation — includes raising the full retirement age again, or changing how yearly raises are worked out so each one comes in a little smaller.

Senator Cassidy says he is working with Senators Tim Kaine and Dick Durbin on a different kind of proposal: a $1.5 trillion investment fund, held apart from the trust fund, that would compound over 60 to 70 years. Notice what that is. It is the stock-market idea from earlier, done the only way it can actually work — borrow the pile first, then invest it, instead of hoping a fund that is emptying can earn its way out. By his "conservative estimates," it would eventually offset up to 65% of the program's shortfall, which he puts at $27 trillion in today's money. The Committee for a Responsible Federal Budget puts the 75-year shortfall at $31 trillion on the same present-valueWhat a future amount of money is worth in today’s dollars. It lets you add up money spread over 75 years using one yardstick. basis, so treat Cassidy's $27 trillion as his own accounting, not an agreed figure. The fund wouldn't touch the immediate gap — the money needs decades to grow — but it would make the remaining hole smaller.

Cassidy also warns of a worst case: if Congress simply covers the shortfall out of the general fund without fixing the underlying problem, the national debt balloons. He attributes to the Congressional Budget Office the view that the impact would be severe, and predicts long-term interest rates would rise, making mortgages unaffordable. That attribution is his, not something the CBO said here.

What it means for your own plan

The most immediate question for anyone near retirement is whether to claim Social Security early to "lock in" a full check before a cut lands. Erik Goodge, a chartered financial analyst and certified financial planner, has run the numbers — and his answer is no.

A 22% cut does change the break-even math. The break-even age is the point at which delaying your benefit has paid for itself — when the larger checks you get by waiting catch up to the checks you skipped. Goodge's analysis shows a cut pushes that age later, especially for people making single-year delay decisions close to 2033. For someone turning 70 in 2033 and deciding whether to delay from 69 to 70, the break-even age stretches to nearly 89.

But Goodge explicitly advises against claiming early to get ahead of the cut. Break-even age alone is a poor basis for the decision — not least because nobody knows how long they'll live. His recommendation: build a stress test into your retirement plan that assumes you get only 80% of your projected benefit, and assume higher Medicare costs on top. If the plan still works under those assumptions, you're in good shape. If it doesn't, that gap is something to deal with now, not in 2032.

Scott Caufield says much the same: don't ignore it, don't overreact to it. Build a plan that survives lower benefits, higher Medicare costs and future tax changes.

What is still unsettled

Six things in this article are not settled:

• The fund-by-fund split. The $2.56 trillion combined total is the Trustees'. The $2.34 trillion for the retirement fund and the $223 billion for the disability fund, and the 2025 flows behind them, are the advisers' reading of the report's tables.
• Cassidy's 28.5% cut. It is on the record from him, it does not match the Trustees' 22% automatic cut, and he did not show his work. It does match the Trustees' figure for a deliberate fix passed in 2034. Which one he meant is still his to say.
• Cassidy's $27 trillion shortfall. The Committee for a Responsible Federal Budget says $31 trillion. Same idea, different accounting.
• What the 2025 tax law costs the fund. The $30 billion a year and $168.6 billion through 2034 come from the Committee for a Responsible Federal Budget. The Trustees name the law as a reason the date moved but do not publish their own dollar figure for it.
• Chile's fund returns. The roughly 10% a year real return for the oldest Chilean fund is a widely repeated figure from secondary write-ups, not a number I pulled from the Chilean regulator. The replacement rates and the March 2025 reform terms are firmer.
• The wage-cap scores. Those percentages are the Chief Actuary's, measured against the 2025 report. The 2026 report widened the gap and the options have not been re-scored against it yet, so every one of those shares is now somewhat generous.

As long as workers pay payroll taxes, the checks keep going out. What nobody knows yet is how big they will be, and who pays to keep them where they are.