Two Greek thinkers examine a pension fund, wage ladder and coins above a removable contribution ceiling.

Lifting The Payroll Cap

The Social Security payroll tax stops at a cap each year, so earnings above it pay nothing toward the program. On 23 June 2026 the Democrat Elizabeth Warren and the Republican Bernie Moreno proposed lifting the cap, and the Club for Growth leads the opposition on the right. No bill text has been published, and Congress has not taken the idea up.

78 percent of scheduled benefits can still be paid once the retirement fund's reserves run out in late 2032 ^[1]^[2]
6 percent of workers earn more than the cap in a given year ^[4]
334 billion dollars a year would come in from earnings above the cap ^[3]

The proposal

Charge the Social Security payroll tax on earnings above the taxable maximum of 184,500 dollars, which today stop being taxed altogether.

The debate

Warren and Moreno call the cap unfair, because most workers pay on all their earnings and the best-paid only on part, and want the revenue to spare retirees a cut. Opponents on the right answer that a tax without matching benefits turns an earned pension into welfare and hits small-business owners hardest. Both sides claim to defend the earned-benefit character of the program. No bill has been introduced, so nobody has voted.

“Instead of cutting benefits for the retirees who count on Social Security, we need to take bipartisan action to protect those benefits, reward work and restore fairness.”

argues for: Bernie Moreno and Elizabeth Warren, Republican, Ohio, and Democrat, Massachusetts Joint op-ed, press release, 23 June 2026
“Removing the cap severs that link and turns an insurance program into a welfare scheme, reducing a worker’s contributions to just another tax on income, with no guarantee anything comes back.”

argues against: David McIntosh and Ken Blackwell, Club for Growth Op-ed, press release, 29 June 2026

The analysis

Lifting The Payroll Cap

71% of the weight is for, 29% against.

An argument's weight combines three questions: how big the effect is, how certain it is and how much the good at stake matters. The larger the slice, the more the argument weighs.

For

Against

On the current analysis, lifting the payroll cap is a good proposal, and the reason is a single transfer.

Very likely better

Very likely better: Under the assumptions used, the proposal brings more than it costs in more than 99 of 100 model runs.

Overall effect: very large

How strongly the proposal works overall, in five steps from very small to very large, comparable across all debates. The ring is always full, even when the overall effect is small; this step says how large it is.

Like Warren and Moreno, the analysis favors lifting the cap, though less because of fairness than because of who receives the money. Nearly all of the weight is money changing hands, from earners above the cap to people drawing a pension, and the result turns on where that money does more good. A dollar does far more for a pensioner who lives mostly on Social Security than for someone earning several times the median. Count the two as equal, and the work lost at the top, the strongest argument against, would tip the balance the other way. The broken link between contributions and benefits, the opponents' main point, is the smallest item, though its true size is the hardest to know.

In detail

All arguments

Arguments for2 arguments

Pensions that are paid in full

22points

The retirement fund runs out of reserves in the last quarter of 2032, and current law then permits only 78 percent of scheduled benefits. Taxing earnings above the cap closes most of that gap.

Value 5 · Household budgetsImpact 8.1Plausibility 5.5

▸ Show reasoning & sources ▾ Hide reasoning & sources

The stream is money arriving in the households of people drawing a state pension, priced at the middle of the scale as all money is. What makes this the largest argument on this site is not the value class but where the money lands: about four in ten American pensioners take more than half their income from this one source, and one in seven takes almost all of it. The people paying it appear as the argument opposite, at the weight of the top few percent of earners, and the difference between the two weights is the whole of what this evaluation says. The part of the base that keeps moving above the cap as top pay pulls away is inside this figure rather than a separate argument: it is the same money, and it is worth what money is worth. Nothing is counted here for the security of knowing the payment will arrive, which is a separate argument. Nothing is counted for the additional benefit credit some versions of the proposal would give the people paying, which would reduce the transfer and is not in the version scored here. The value is the middle of the scale, and the entire weight of this argument sits in the distance between two income bands.

The taxable maximum is 184,500 dollars in 2026, and about six percent of workers earn more than it in a given year. Charging the tax above it raises roughly 320 billion dollars a year before anybody changes their behaviour. Two things move that figure over the twenty years. The share of wages above the cap has grown by about a quarter of a point a year for four decades; carried on at that pace and averaged over the horizon, it adds a tenth to the base, in a range from 4 to 20 percent. And a fifth less kept on every dollar above the cap means about five percent of those earnings stop being reported — the same response, at the same elasticity of 0.25, that the argument on work and reporting is built on, so that one number governs both sides. Together that is 334 billion dollars a year. No bill text sets how the new revenue is split ^[8], and this evaluation assumes all of it is credited to the retirement fund, as the proposal's purpose implies; under the current split of 10.6 points to that fund and 1.8 to the disability fund ^[7], about 285 billion dollars would reach it and 49 billion the disability fund, a case the lower end of the range covers. The retirement fund's reserves are otherwise exhausted in the last quarter of 2032, after which current law permits only 78 percent of scheduled benefits — a cut of about 22 percent for every recipient at once. The money collected in the first six years builds reserves rather than paying anybody; but the programme already costs about 1.7 trillion dollars in 2026 and more every year, so the cut it prevents from late 2032 runs at 450 billion dollars a year or more and exceeds the yearly revenue; over the fourteen years from 2033 the cuts prevented come to about 7 trillion dollars against about 6.7 trillion collected over all twenty, so the reserves built early are paid out inside the horizon — this evaluation does not discount within it. The money therefore lands with pensioners, who carry a weight of 1.4, in a range from 1.1 to 2.0: a mix of the poorest fifth, for whom this is nearly all their income, and households around the middle for whom it is a part. The Impact is the largest figure anywhere on this site and it is simply the size of the transfer multiplied by the distance between two income bands.

▸ Show calculation ▾ Hide calculation

	Revenue from charging the tax above the cap, before anybody changes behaviour static estimate over the next ten years [3]	320 billion dollars a year, 2027 to 2036	320 billion dollars a year
×	Base that keeps moving above the cap over the twenty years Setting, range 1.04 to 1.20: the covered share has fallen from 90 to about 83 percent since the cap was designed, roughly a quarter of a point a year; carried on at that pace and averaged over the horizon, that adds a tenth to the base above the cap [4]	1.10	352 billion dollars a year
×	Earnings that stop being reported at the higher rate the same response as in the argument on work and reporting: elasticity 0.25, a fifth less kept, five percent less reported — one number for both legs and for the output lost [5]	0.95	334 billion dollars a year
÷	In euro-equivalent exchange rate used throughout this evaluation	1.16 dollars to the euro	288 billion euro-equivalent a year
×	Weight of a dollar for the households receiving a state pension Setting, range 1.1 to 2.0: four in ten pensioners take more than half their income from this source and one in seven almost all of it, alongside households around the middle for whom it is a part	1.4	403.2 billion euro-equivalent a year
÷	Normalised Impact scale of this evaluation	50 billion euro-equivalent a point	8.06

Score 8.06 Impact × 5 Value × 5.5 Plausibility ÷ 10 = 22 of 100

The revenue is arithmetic — a statutory rate applied to earnings that are already reported — and the benefit cut it prevents is what current law does automatically when a trust fund is empty, so neither end of this requires anybody to decide anything. The counterfactual is that law, unamended, which the trustees' 2026 report dates precisely. Two things hold this well below what an arithmetic certainty would otherwise carry. The first is timing: the money is collected from 2027 and the cut it prevents begins in 2032, so for the first six years of this evaluation the revenue accumulates rather than preserves anything, which is inside the twenty-year horizon but not evenly spread across it. The second is larger and is not a technical objection: Congress has never once allowed a scheduled Social Security cut to take effect, and if it would have acted by some other means, the benefit preserved here is only the difference between this fix and that one, which nobody can size. That is named and unresolved. Reverse causation does not arise. The Plausibility is a little above the middle: the arithmetic is certain and the thing it is compared against is a law nobody expects to be allowed to operate.

Evidence: Projection · plausibility at most 6 Evidence of effect: Definitional

Something to plan around

0.08points

A worker of forty today cannot say what the state pension will pay when they reach sixty-seven, because current law promises a cut in 2032 and everyone assumes it will not happen. Removing the shortfall removes the question.

Value 4 · Provision and securityImpact 0.04Plausibility 5

▸ Show reasoning & sources ▾ Hide reasoning & sources

The stream is the ability to make a plan: what to save, when to stop working, whether to buy an annuity. This site places that with ordinary usefulness and comfort rather than higher, because what is at stake is the quality of a decision rather than the money it concerns — the money itself is the argument above. It is genuinely separate from that argument in the way that insurance is separate from a payout: a person whose benefit is eventually paid in full still spent thirty years not knowing. Nothing is counted for the people already drawing a pension, for whom the horizon is short and the uncertainty small. The value sits in the lower-middle of the scale, because what is priced is the quality of a decision and not the money the decision is about.

About 70 million people are within thirty years of drawing a state pension and currently face a promise the law says cannot be kept in full. What removing that uncertainty is worth to each is set as if it were worth about 35 dollars a year, in a range from 12 to 120 — the band this site uses for a benefit that reaches a very large number of people and matters modestly to each, and roughly what an annual insurance premium against a small chance of a large loss would cost. That gives a Normalised Impact of 0.04 on the scale of this evaluation. The figure would be larger if the shortfall were closer or the cut deeper, and smaller if people simply do not believe the cut will happen, which surveys suggest most do not. It is deliberately small next to the money in the first argument because it is the same event valued twice over: once as the payment and once as knowing it will come. The Impact is about a two-hundredth of the money in this debate, which is the honest size of certainty priced separately from the thing it is certain about.

▸ Show calculation ▾ Hide calculation

	People within thirty years of drawing a state pension [1]		70 million people
×	Value of knowing what it will pay, per person per year Setting, range 10 to 100 euro-equivalent (about 12 to 120 dollars): the band this site uses for a benefit reaching a very large number of people and mattering modestly to each	30 euro-equivalent (about 35 dollars)	2.1 billion euro-equivalent a year
÷	Normalised Impact scale of this evaluation	50 billion euro-equivalent a point	0.04

Score 0.04 Impact × 4 Value × 5 Plausibility ÷ 10 = 0.1 of 100

The uncertainty is documented — the trustees publish the depletion date and the payable share every year, and it is widely reported. The counterfactual is the same workers under current law. What has never been measured is what the uncertainty costs anybody: no study puts a value on knowing what a state pension will pay, and the figure used here is this site's convention for a small benefit reaching a large population rather than a finding. The confounder that matters cuts against the argument, and it is the same one as in the first: most people do not expect the cut to happen, and a person who does not believe in a risk is not paying anything to avoid it. That is named and unresolved. Reverse causation does not arise. The Plausibility is at the middle: the uncertainty is real and documented, and what it costs the people living with it is a convention.

Evidence: Plausibility · plausibility at most 5 Evidence of effect: Definitional

Arguments against3 arguments

The six percent who pay it

5.6points

About six percent of workers earn more than 184,500 dollars, and all of the money comes from them. A person on 400,000 dollars would pay roughly 27,000 dollars a year more than today.

Value 5 · Household budgetsImpact 2.0Plausibility 5.5

▸ Show reasoning & sources ▾ Hide reasoning & sources

The stream is money leaving the households that earn above the cap, priced at the middle of the scale exactly as the pensions it funds are. This argument exists so that the first one is not counted as though the money came from nowhere. The weight is what differs and it is the whole substance of this evaluation: a dollar at the top few percent of American earnings carries 0.35 on this site's scale against 1.4 for a pensioner household, which is why the same sum weighs four times as much on the other side. Nothing is counted for the sense of unfairness at paying into a system that credits little of it back, which is a separate argument. Nothing is counted for the work and reporting response, which is the argument after this one. The value is the middle of the scale, and this argument exists so that the pensions above are not booked as if nobody funded them.

The same 334 billion dollars a year, seen from the side that pays it — including the tenth that the base adds as top pay keeps pulling away, and net of the five percent that stops being reported. It falls on about six percent of workers — those earning above 184,500 dollars in a given year — although nearly a fifth of workers cross the cap at some point in a career, so the group is less fixed than the annual figure suggests. A person earning 400,000 dollars would pay roughly 27,000 dollars a year more, counting both halves of the tax, since a levy on wages is borne by wages whichever side writes the cheque. The weight applied is 0.35, in a range from 0.2 to 0.5, which sits between the figures this site uses for the top tenth and the top one percent, because earnings above 184,500 dollars span both. That gives a Normalised Impact of 2.02 on the scale of this evaluation. Little or no additional benefit is credited for the earnings taxed, which is what makes this a transfer rather than a contribution. The Impact is a quarter of the argument opposite, built on exactly the same money.

▸ Show calculation ▾ Hide calculation

	Revenue from charging the tax above the cap [3]	the same money, seen from the side that pays it: 334 billion dollars a year averaged over the horizon, at 1.16 dollars to the euro	288 billion euro-equivalent a year
×	Weight of a dollar for the six percent who earn above the cap Setting, range 0.2 to 0.5: between this site's figures for the top tenth and the top one percent, because earnings above 184,500 dollars span both	0.35	100.8 billion euro-equivalent a year
÷	Normalised Impact scale of this evaluation	50 billion euro-equivalent a point	2.02

Score 2.02 Impact × 5 Value × 5.5 Plausibility ÷ 10 = 5.6 of 100

The quantity is the revenue figure read from the other side, so it carries the same confidence and the same caveats. The counterfactual is the cap as it stands. That a transfer has two ends is an identity — if the fund receives 334 billion dollars, somebody pays it, and a statutory rate on reported earnings is handed over in every case — but how much is handed over depends on how far earnings above the cap move away from the tax, which is the same question the argument opposite turns on, so the two legs carry the same plausibility for the same sum; what is certain is that the payment happens, what is judged is its size. What is estimated is the weight, which is this site's band structure applied to a group that spans from comfortable professionals to the very rich, and the split matters — at 0.2, the figure this site uses for the top one percent, this argument falls by more than a third. The confounder is the same one as opposite and runs the same way: if behaviour erodes the base more than assumed, both this argument and the one it mirrors shrink together. Reverse causation does not arise. The Plausibility is a little above the middle, matching the argument it mirrors: the sum is fixed by the same arithmetic and the weight applied to it is a band rather than a measurement.

Evidence: Projection · plausibility at most 6 Evidence of effect: Definitional

Work and reporting that stop

3.2points

Adding 12.4 points to the marginal rate on earnings above 184,500 dollars takes the combined rate past half in several states. Some of the response is less work and some is income moved into forms the tax does not reach.

Value 6 · OutputImpact 1.0Plausibility 5.5

▸ Show reasoning & sources ▾ Hide reasoning & sources

The stream is output that is not produced: hours not worked, a practice not expanded, a partner who retires two years earlier. This site places that with the working order of economic systems and prosperity, because what is lost is production rather than anybody's income in particular. Only the part of the response that is real is counted here. The rest — income shifted from wages into corporate or capital form, or into deferred compensation — is not a loss to anybody, it is money the tax does not collect, and it is already inside the revenue figure used on both sides. Nothing is counted for the effort spent on the shifting itself, which is small and real and not estimated. The value sits in the middle-upper part of the scale, and only the part of the response that is genuinely less production is counted.

Earnings above the cap come to about 2,580 billion dollars a year at today's distribution of pay, which is what a static revenue estimate of 320 billion dollars at a 12.4 percent rate implies, and about 2,840 billion averaged over the horizon as the base keeps moving above the cap. The marginal rate on those earnings rises by 12.4 points, taking the amount kept from about 60 cents in the dollar to about 48, a fifth less. At an elasticity of taxable income of 0.25 — the central figure in a literature whose estimates run from 0.1 to 0.6 for high earners — that is a five percent fall in reported earnings above the cap, or about 142 billion dollars, and it is the same five percent by which the revenue on both sides of this evaluation is reduced. Forty percent of that is treated as real output foregone rather than income relabelled, in a range from 20 to 70 percent, giving about 57 billion dollars a year at the ordinary weight for production. The free time gained in its place is not counted, and it is left open rather than estimated. The 40 percent real share mixes fewer hours, less effort within the same hours and earlier retirement, which no study of top earners separates, and only fewer hours and earlier retirement give back time that could be counted at the rate this site uses for private time. Dividing 57 billion dollars by an hourly wage would turn output into hours only if all of it were hours. The figure therefore overstates the net loss by an amount that cannot yet be sized. The reason the real share is set below half is that the response of top earners to tax changes is dominated by timing and form in every study that can separate them. The Impact is half the money the six percent hand over, which is a large behavioural cost by the standards of this site and a middling one by the standards of this literature.

▸ Show calculation ▾ Hide calculation

	Earnings above the cap, averaged over the horizon [3][4]	2,225 billion euro-equivalent at today's distribution of pay (about 2,580 billion dollars, from 320 billion dollars of static revenue at 12.4 percent, at 1.16 dollars to the euro), × 1.10 for the base that keeps moving above the cap	2,448 billion euro-equivalent a year
×	Fall in reported earnings above the cap the amount kept falls from about 60 to 48 cents in the dollar, a fifth, at an elasticity of 0.25 from a literature ranging from 0.1 to 0.6 [5][6]	5 %	122.4 billion euro-equivalent a year
×	Share that is real output rather than income relabelled Setting, range 20 to 70 percent: the response of top earners is dominated by timing and form in every study that can separate them	40 %	49 billion euro-equivalent a year
÷	Normalised Impact scale of this evaluation	50 billion euro-equivalent a point	0.98

Score 0.98 Impact × 6 Value × 5.5 Plausibility ÷ 10 = 3.2 of 100

The counterfactual is the same earners at the current marginal rate, which is what the tax reform literature constructs from the sharp rate changes of 1986 and 1993 and, more recently, the 2013 increase in top rates ^[5]^[6]. Those designs are quasi-experimental — difference in differences around a legislated threshold, and bunching at kinks — and they are among the better-identified estimates in public finance. The confounder that matters is that a legislated rate change is announced in advance, so what looks like a behavioural response is partly income shifted between years, and the studies that can separate short-run timing from long-run supply find much smaller long-run effects — after 2013, reported top incomes fell sharply for a year and then resumed their trend, a short-run elasticity above one and a medium-run one that is small ^[6]. That is named, unresolved, and the reason only 40 percent of the response is treated as real. Reverse causation does not arise from a legislated rate. The elasticity itself is contested by a factor of six across the literature, which is carried in the range rather than in the plausibility. The Plausibility is a little above the middle: the response is measured on good designs and how much of it is real production rather than relabelled income is not settled.

Evidence: Converging studies · plausibility at most 6 Evidence of effect: Quasi-experimental · at most 8

Open: A study that splits the response of top earners to the 2013 rate increase into hours worked, retirement timing and reported income would size the free time gained and allow it to be booked against this output.

A contribution that buys nothing

0.03points

Social Security has survived ninety years by being an earned benefit rather than welfare: what you pay in determines what you get out. Taxing earnings that are credited with little or nothing breaks that link.

Value 6 · Trust in the systemImpact 0.01Plausibility 5

▸ Show reasoning & sources ▾ Hide reasoning & sources

The stream is the character of the programme in the minds of the people who pay for it: a contributory pension in which a contribution buys a claim, against a tax that funds somebody else's payment. This site places that with confidence in public institutions, because what is at stake is whether a programme keeps the political protection that has carried it through ninety years and several attempts to cut it. What is priced is the change in the relationship, spread across the people paying in, and not any particular political outcome, which nobody can forecast. Nothing is counted for the versions of the proposal that credit some benefit on the earnings taxed, which would reduce this and also reduce the revenue. The value sits in the middle-upper part of the scale, and what is priced is a change in what the programme is rather than any political consequence of it.

About 170 million people pay this tax in a year. The change in what the programme is, for each of them, is set as if it were worth about 3.50 dollars a year, in a range from about 1 to 12 — the band this site uses for a broad and lightly felt change in a public arrangement. That gives a Normalised Impact of 0.01 on the scale of this evaluation. The figure is small in absolute terms and it is the argument most likely to be badly sized in either direction: if breaking the contributory link is what eventually costs Social Security its political protection, the cost is orders of magnitude larger than this, and if voters never notice, it is zero. Nothing here forecasts which. What can be said is that every serious attempt to convert the programme into means-tested welfare has failed precisely because contributors saw it as theirs. The Impact is the smallest in this evaluation by a factor of eighty, and it is the one whose true size is least knowable.

▸ Show calculation ▾ Hide calculation

	People paying the tax in a year [1]		170 million people
×	Value of the change in what the programme is, per person per year Setting, range 1 to 10 euro-equivalent (about 1 to 12 dollars): the band this site uses for a broad and lightly felt change in a public arrangement, with no anchor behind it	3 euro-equivalent (about 3.50 dollars)	510 million euro-equivalent a year
÷	Normalised Impact scale of this evaluation	50 billion euro-equivalent a point	0.01

Score 0.01 Impact × 6 Value × 5 Plausibility ÷ 10 = 0.0 of 100

The change itself is definitional: earnings would be taxed without a matching benefit credit, which is a different arrangement from the one that exists. The counterfactual is the contributory system as it stands. What is a valuation rather than a finding is the figure per person, and this site has no anchor at all for what a change in the character of a public programme is worth — the figure is the general band for a broad light change, applied by judgement. The confounder that matters is that most people do not know the benefit formula well enough for the link to be doing the work attributed to it, in which case breaking it costs nothing; that is named and unresolved. Reverse causation does not arise. Nothing has measured this and nothing is likely to. The Plausibility is at the middle: the change is certain and its value is a convention with no anchor behind it.

Evidence: Plausibility · plausibility at most 5 Evidence of effect: Definitional

How robust the result is

Under the assumptions used, the proposal brings more than it costs in more than 99 of 100 model runs.

gains more than it costscosts more than it gains

Each dot stands for 200 of the 20,000 model runs.

The result rests almost entirely on how much more a dollar is worth to a household drawing a state pension than to somebody earning three hundred thousand. The pensioner's dollar is counted at 1.4 times a dollar at median income, the dollar of the six percent above the cap at 0.35. Treat the two as equal and the measure comes out behind, because the tax also discourages some work and reporting at the top; at the weights used it lies ahead in all runs. The comparison throughout is with current law, which pays 78 percent of scheduled benefits from late 2032. Congress has never once allowed that cut to happen, so the real alternative may be a different fix rather than no fix. No bill text exists; the analysis credits the new revenue wholly to the retirement fund rather than splitting it with the disability fund as the current rate is.

Show the distribution of the model runs

20,000 simulated futures

Very likely better · 100 %

In more than 99 of 100 simulated futures the proposal brings more than it costs.

No effect is known exactly; every argument has a range. The calculation draws a possible size for each argument from its range 20,000 times and counts how often the benefits exceed the costs.

Net effect in points per future. Left of zero the drawbacks outweigh the benefits, right of zero the benefits win. Eight in ten futures fall between +9.0 and +18 points.

Scale and assessment

Scale of this evaluation: Normalised Impact, unitless and calibrated to this topic. For comparison, one point here corresponds to about 50 billion euro-equivalent a year.

Size class:very large

Effect very large (+670) · 84 % of the effect is positive

Areas show individual argument weights. The positive share of the effect nets out money transfers between the two sides; the probability of a positive net effect is determined separately.

How we score →

Review and sources

AI evaluation · not yet reviewed by a human

This evaluation was produced and sourced by an AI model; a human review is still pending. Figures and conclusions may still change. The review log is at the foot of the page.How review works →

Last reviewed by Claude Opus 5.5 · October 29, 2026 · 7× AI, not yet reviewed by a human

October 29, 2026AI reviewClaude Opus 5.5record updated
Page texts for the new page layout (draft G) with quotes from primary sources; numbers unchanged.
October 21, 2026AI reviewClaude Fable 5.1record updated
Money amounts in argument texts stated in dollars per style-en.md; scores, chains and balance unchanged
October 20, 2026AI reviewClaude Fable 5.1record updated
Summary and scale note rewritten to the style of 20 Oct 2026; numbers unchanged
October 14, 2026AI reviewClaude Opus 5.5record updated
Englische Debatten, Befund SO-19: volle Gutschrift an den Rentenfonds als Ausgestaltungsannahme benannt (geltende Aufteilung 10,6/1,8 Pkt. in der Spanne), zurückgewonnene Freizeit bei con-2 begründet offen gelassen.
September 14, 2026AI reviewClaude Opus 5re-scored
Gesamtprüfung 08.09. umgesetzt: eine Einnahmekette für Paar und con-2 (320 $ × 1,10 Drift × 0,95 = 288 Mrd €), pro-3 (unpaarer Transfer) in das Paar gefaltet, Quelle Saez 2013 ergänzt; r 0,73 → 0,72, Kategorie bleibt besser.
September 6, 2026AI reviewClaude Opus 5record updated
i_spanne an allen 6, normalisierung an 4 (Vorsorge-Sicherheit und Systemvertrauen haben keinen Anker), Transferpaar pro-1 ⇄ con-1 verdrahtet (250 Mrd EUR, P 5,5 beidseitig). Kernpunkt: con-1 stand auf definitorisch, pro-1 auf mechanistisch — die Zahlerseite trat in jeder Ziehung ein, die Empfaengerseite nur in 55 %. con-1 auf mechanistisch angeglichen, P unveraendert unter dem Deckel: P(D>0) 0,55 → 0,77, Kategorie bleibt Besser. Der Angleich zeigt einen Widerspruch: der Brief sagt unsichereres Bein, die offene Modellfrage schlaegt das sicherere vor — hier entscheidet das eine Kategorie.
September 6, 2026AI reviewClaude Opus 5First evaluation
First evaluation: the largest transfer scored on this site, with the behavioural response booked separately rather than deducted from the revenue twice.

Evaluations are produced with AI support and reviewed on a schedule for new developments; human passes are marked separately.How we review →

Show all 8 sourcesHide sources

Social Security Administration: Annual report of the Board of Trustees, and the taxable maximum. ssa.gov
Congressional Research Service: Social Security: selected findings of the 2026 annual report. congress.gov
Tax Foundation: The Save Social Security payroll tax cap proposal: details and analysis. taxfoundation.org
Social Security Administration, Office of Retirement and Disability Policy: Population profile: taxable maximum earners. ssa.gov
Saez, Slemrod and Giertz, Journal of Economic Literature: The Elasticity of Taxable Income with Respect to Marginal Tax Rates: A Critical Review. aeaweb.org
Saez, Emmanuel (NBER Working Paper 22798; Tax Policy and the Economy 31, 2017): Taxing the Rich More: Preliminary Evidence from the 2013 Tax Increase — short-run elasticity of reported top incomes above one, driven by retiming; only a small medium-run response 2011 to 2015. nber.org
Social Security Administration, Office of the Chief Actuary: Social Security and Medicare tax rates: allocation between the retirement (OASI) and disability (DI) funds, 10.6 and 1.8 percent combined since 2019. ssa.gov
Senator Elizabeth Warren: Warren, Moreno pen NYT op-ed: our bipartisan plan to save Social Security (23 June 2026). warren.senate.gov

More debates in this area

Very likely worse · 0 %

Size: large

Keep the Tips Deduction

Make the deductions for tipped and overtime income permanent instead of letting them lapse after 2028.

Very likely better · 100 %

Size: medium

California Billionaire Tax

Charge everyone who lived in California on 1 January 2026 with a net worth above 1 billion dollars a one-time tax of 5 percent, and spend nine tenths of it on health care.

Very likely better · 100 %

Size: large

A Fully Refundable Credit

Pay the full child tax credit to every eligible child, including those whose parents earn too little to owe income tax.

Very likely worse · 0 %

Size: large

The Forced-Labour Tariffs

An additional duty of 10 or 12.5 percent on imports from sixty economies, set by whether each bans, or has committed to ban, imports made with forced labour.

All debates in Taxes, Tariffs and the Federal Budget →