I'm not claiming this is perfect, or even the right path. What it is is a starting point. A proof of concept that SS is not unfixable and that politics is the problem, not the solution. I will likely try to pass this on to some politicians, to get it out there...someday. This is the culmination of weeks of research, and then a LONG back in forth with Claude AI - last two days using Fable 5. It is largely written by Claude, prompted over many drafts and final editing by me. I am looking for feedback, that I can keeping pushing to Claude and revising the plan. The early 2030's failure of funding, with automatic cuts, is this plans purpose. To be "out there" in the fantastically small chance that someone important finds it, and works with it.
Part 1: The Problem, Quantified
Social Security has two failure modes operating simultaneously, and conflating them is the source of most bad analysis. The first is solvency. The second is adequacy. They require different fixes, and a solution to one does nothing for the other.
Solvency. The Old-Age and Survivors Insurance trust fund is projected to deplete around 2034. The mechanism of failure is worth stating precisely, because it is automatic and statutory: at depletion, the program is legally permitted to pay out only what incoming payroll revenue covers, which is approximately 77% of scheduled benefits. That translates to an across-the-board cut of roughly 23%, applied uniformly regardless of recipient need, on the depletion date. No further legislation is required to produce that cut. Legislation is required to prevent it.
The cause is demographic, not financial mismanagement. A pay-as-you-go system is a ratio: current workers fund current beneficiaries. That ratio was approximately 5.1 workers per beneficiary in 1960. It is approximately 2.7 today. It continues to decline as longevity rises and birth rates fall. No investment strategy fixes this, because there is no investment — payroll taxes collected today are paid out today, with the surplus historically lent to the general fund in exchange for special-issue Treasury bonds. The "trust fund" is a claim on future taxpayers, not a pool of invested capital.
Adequacy. Set solvency aside and assume full benefits are paid. They are still insufficient. The average monthly benefit is approximately $1,900. Median monthly expenditure for retiree households is approximately $3,800. The program covers roughly half of what its recipients spend. The maximum possible benefit, available only to high earners who delayed claiming, is approximately $4,018 — still barely enough.
This is not a design failure in the original sense. Social Security was architected as one component of a three-part retirement structure: Social Security, employer pensions, and personal savings. The defined-benefit pension has largely vanished from the private sector. Personal savings rates are inadequate across most of the income distribution. The other two legs broke, and the remaining leg was never sized to bear the full load.
A structural footnote that matters later. Under Flemming v. Nestor (1960), Social Security benefits are not property. The Supreme Court held that a worker's contributions create no contractual or property right to benefits; Congress may alter or eliminate them by ordinary legislation. This is why the 2034 cut is legally permissible. It also means the entire system rests on political will rather than legal entitlement — a fact with direct bearing on any replacement's design.
Part 2: The Revenue Base
The Social Security payroll tax is 12.4% of wages (split nominally between employer and employee), levied only on earnings below a cap — $184,500 in 2026. Earnings above the cap are untaxed for SS purposes. Because higher earners derive a larger share of income from above-cap wages, the effective rate is regressive: it declines as income rises past the threshold.
Removing the cap — taxing all wage income at 12.4% — generates an estimated $150–200 billion annually at current wage levels, growing over time as wages rise. The Social Security Administration's own modeling indicates that cap removal with continued benefit accrual on the new earnings defers trust fund depletion from roughly 2034 to roughly 2059. Cap removal without benefit accrual on the new earnings — treating it purely as revenue — produces a larger fiscal effect.
Medicare eliminated its equivalent wage cap in 1994. The Social Security cap's persistence is a matter of legislative inertia rather than actuarial logic.
The proposal here does not use cap removal to extend the existing system. It uses it as transition financing for a structural replacement. The distinction is the entire point: cap removal applied to the current system buys roughly 25 years of solvency on an inadequate benefit. Cap removal applied as a transition bridge buys a permanent replacement that is both solvent and adequate. Same revenue, categorically different return.
Part 3: The Framework
The replacement converts retirement provision from a pay-as-you-go promise into funded, individually titled accounts, phased in over approximately 40 years. It comprises six accounts, distinguished by funding source, access rules, investment latitude, and treatment in a retirement-stage means test. The organizing principle: investment restriction tracks funding source. Tax-funded balances are restricted and locked; the worker's own discretionary contributions earn progressively more freedom; and the boundary between the two is precisely where the means test stops examining.
Account 1 — Legacy Trust Fund. The existing system, maintained for current retirees and transitional cohorts. Funded by cap removal revenue, residual payroll tax, and transition bonds. Its obligation declines over four decades as cohorts retire under the funded system instead, until it functions solely as a means-tested supplement.
Account 2 — Mandatory Private Account. The direct replacement. Approximately 5 percentage points of the existing payroll tax (phased in by cohort) redirected into an individually titled account, invested through a Thrift Savings Plan–style structure: federal governance, competitively selected private fund managers, administrative cost near 4 basis points. Fully locked — accessible only at retirement age, permanent disability, or death. No loans, no hardship withdrawals. Roth tax treatment.
Account 3 — Mandatory 401(k). A new mandatory contribution layered on payroll tax. Auto-enrollment at 3% of wages, escalating 1 percentage point annually to approximately 10%, plus a required employer match near 3%. This is the one component that asks for new contributions beyond existing payroll tax; the analysis does not obscure that. Two regulated access doors — early retirement and adjudicated catastrophic hardship — distinguish it from the fully locked Account 2, with every withdrawal accounted for via the plus-up mechanism described below. Investment latitude splits by source: the employer match is restricted like Account 2; the employee portion gets an expanded menu (sector, international, REIT, infrastructure funds) at a 50-basis-point cost cap, but no individual securities, leverage, or crypto.
Account 4 — Voluntary 401(k). Discretionary contributions above the mandatory layers, traditional or Roth, under a combined annual contribution cap across Accounts 2–4 of roughly $15,000–$20,000. Broad investment freedom: individual securities, alternatives, crypto capped at 10% of balance. Entirely excluded from the means test.
Account 5 — Emergency Account. Funded by sequencing rather than new contribution: the first two escalation points destined for Account 3 route here until the account reaches a floor of $5,000 or two months' wages, then revert. Capital-preservation investments only, fully liquid, excluded from the means test. This is the structural answer to hardship that permits Accounts 2 and 3 to remain genuinely locked. The infrastructure exists — SECURE 2.0's pension-linked emergency savings accounts are a smaller-scale precedent already operating.
Account 6 — Birth Investment. A means-tested grant at birth ($10,000 below the poverty line, tapering to zero above 300% of it), invested in the restricted menu and locked for 65 years. Estimated cost approximately $45 billion annually.
The retirement-stage means test examines only Accounts 2, 3, and 6. It converts combined balances to an imputed monthly income (4% drawdown), compares to a threshold (150% of the poverty line), and provides a sliding-scale supplement from Account 1 to anyone below it. Accounts 4 and 5 are never examined. The test recalculates every five years through retirement, using trailing three-year average balances to neutralize point-in-time market timing.
Part 4: The Plus-Up Mechanism
The historical objection to personal-account systems is the tension between liquidity and integrity: if workers can access funds early, accounts are depleted and the safety net is gamed; if they cannot, the system is rigid and politically brittle. The standard resolution is prohibition, which is unpopular and leaks through hardship exceptions.
The plus-up mechanism resolves this as an accounting problem instead. Every Account 3 withdrawal is logged with date and amount. At the means test, each withdrawal is compounded forward to retirement age at the trailing 10-year Treasury rate as of the withdrawal date. The "phantom balance" — actual balance plus the compounded value of all prior withdrawals — is what the means test evaluates.
Worked example: a $50,000 hardship withdrawal at age 57, when the 10-year Treasury yields 4.5%, is treated at age 65 as $50,000 × (1.045)⁸ ≈ $71,000 still present in the account. Early retirement is handled identically — a balance drawn from at 60 is evaluated at what it would have been at 65 untouched.
The consequence: access is permitted without subsidizing it. A worker may withdraw, but cannot use withdrawal to manufacture eligibility for a larger supplement. This collapses the gaming surface to near zero and eliminates the need for Medicaid-style lookback investigations, because the ledger is lifetime and automatic. It is, to my knowledge, more sophisticated than any equivalent provision operating at scale.
Part 5: The Projections
Assumptions, stated for replication. Career length 43 years (ages 22–65). Real wage growth 1.5% annually. Account 2 at 5 points; Account 3 escalating to ~10% employee plus ~3% match. Drawdown at 4%. All figures in constant 2026 dollars. The base-case real return is 5.5% — deliberately conservative. The historical real return on a diversified, lifecycle-adjusted portfolio over a 40-plus-year horizon is closer to 6–6.5%; pure equity is closer to 7%. The 5.5% figure builds in room for the valuation-compression risk discussed in Part 7.
Median worker, $60,000 career average, base case: combined mandatory balance ≈ $1.43 million, producing ≈ $4,800/month at 4% drawdown. Against the current system's ~$1,900, that is roughly 2.5×.
Low earner, $30,000 career average, with birth investment: ≈ $1.04 million, producing ≈ $3,450/month — roughly double current Social Security and above the adequacy threshold, which substantially reduces the supplement obligation for this cohort.
Sensitivity:
| Real return | Median balance | Median monthly income | Supplement population |
|---|---|---|---|
| 4.5% (compression case) | ~$1.13M | ~$3,750 | ~30–35% |
| 5.5% (base case) | ~$1.43M | ~$4,800 | ~20–25% |
| 7.0% (historical) | ~$2.20M | ~$7,300 | ~15% |
The critical property: across every modeled return path, the median outcome exceeds current Social Security by a wide margin, and underperformance manifests as a larger fiscal supplement obligation rather than as retiree poverty. The system degrades into cost, not hardship.
The accumulation curve explains why the locks and the plus-up matter. For the median worker at base case, the balance trajectory runs roughly $120K at 35, $340K at 45, $710K at 55, $1.2M at 62, $1.43M at 65. The final decade contributes more than the first three combined, because compounding is back-loaded. A $50,000 withdrawal at 45 forgoes roughly $270,000 of terminal balance at 5.5%. This is the quantitative argument for restriction: early leakage is disproportionately destructive, and the plus-up prices exactly that.
Part 6: The Birth Investment Math
The single most counterintuitive result in the framework concerns time horizon. A worker contributes for 43 years. A birth account compounds for 65. Those 22 additional years, at the front of the curve where compounding is most powerful, dominate.
A $10,000 lump sum at birth, at 5.5% real, reaches approximately $325,000 by age 65. At the historical 7%, it exceeds $2 million. To generate the median worker's full ~$1.43M mandatory target from a single birth investment would require roughly $27,500 at 7%, or about $105,000 at 5%.
The policy implication: a relatively modest, means-tested grant at birth — costing ~$45 billion annually, against ~$150–200 billion freed by cap removal — produces six-figure balances for low-income children purely through time. This is the most cost-efficient adequacy mechanism in the framework, because it substitutes 65 years of compounding for decades of monthly supplement payments later. Front-loading the investment is dramatically cheaper than back-loading the support.
Part 7: Justifications and Known Weaknesses
Why funded beats pay-as-you-go here. A pay-as-you-go system's return is bounded by the growth rate of the wage base — population growth plus productivity growth, historically low single digits and falling as demographics deteriorate. A funded system earns the return on invested capital, historically several points higher. Over a 40-plus-year horizon, that gap compounds into the 2.5× difference shown above. The cost of capturing it is the transition: the generation that funds the switch must, for a period, support existing retirees and fund its own accounts.
The transition cost, quantified. Redirecting payroll tax to personal accounts while still paying current benefits opens a financing gap that peaks near $600–660 billion annually in years 6–15 — the window when the trust fund is depleted, demographic outlays peak, the carve-out is growing, and no cohort has yet retired on reduced traditional accrual. The gross 30-year shortfall is approximately $11–12 trillion; post-crossover surpluses (the system flips around year 30) recover roughly $3 trillion, for a net transition cost near $8–9 trillion.
Cap removal covers approximately 30% of the gross gap. The remainder is bridged by dedicated transition bonds, serviced by a flat 1% levy on all retirement withdrawals (generating ~$20–25 billion early, scaling to $40–50 billion at maturity) and retired by the structural surpluses the funded system produces after crossover. These figures are structured estimates with material error bars (±20% per period); precise figures require actuarial scoring. The shape, however, is robust: every plausible specification runs through a $500–700 billion peak in the same window.
The de-risking argument. Conventional target-date funds de-risk early — toward 50% equity by 65 — for reasons that do not apply to locked accounts: they protect against panic-selling (impossible when accounts are locked) and against provider litigation (eliminated by a statutory glide path). A funded national system carries an implicit put — the means-tested supplement plus periodic recalculation catch any individual whose balance craters — which permits a collectively more aggressive allocation than any individual could prudently hold. Delaying de-risking to age 50+ adds an estimated 0.4–0.5 points of realized return, roughly 10–15% of terminal balance, concentrated in the high-balance years where it matters most. This is the basis for expecting the real world to outperform the published 5.5%.
The valuation-compression risk. This is the most serious threat to the projections and deserves explicit statement. Mandatory savings at national scale represents price-insensitive demand for equities — tens of trillions in inflows over four decades that must be deployed regardless of valuation. Price-insensitive buying at scale plausibly expands price-to-earnings ratios without commensurate earnings growth, compressing forward returns. The mitigations are structural: only the tax-funded layers are restricted index flows, while the discretionary layers (employee Account 3, all of Account 4) flow into a broader universe including small-cap, international, and real assets; global diversification is mandated within the restricted menu; and the official projection basis is held at 5.5% precisely to remain valid under compression. The honesty problem remains: the policy alters the historical conditions that produced the historical return, which is a circularity no projection fully escapes.
The entrenchment argument. Because Flemming v. Nestor makes current benefits revocable, the current system's stability rests entirely on the political cost of cutting it. Funded accounts change the legal character: individually titled balances are Fifth Amendment property. A future legislature may halt future contributions but cannot confiscate accumulated balances without compensation. This converts each contributed dollar from a breakable promise into a constitutional claim — a stronger guarantee than the current system offers, and one that strengthens automatically as balances accumulate.
Unresolved problems, flagged rather than hidden:
- Disability and survivors. Approximately 20% of current beneficiaries receive disability or survivor benefits, funded by 1.8 of the 12.4 payroll points. Funded retirement accounts do not replicate insurance against disability at 35. This requires a separate mechanism — likely mandatory group disability and term-life coverage with a public backstop for the uninsurable — and is the largest unbuilt module.
- Non-employer workers. The framework assumes payroll infrastructure. Gig, self-employed, and uncovered workers require a parallel mandatory mechanism, probably routed through estimated tax payments. State auto-IRA programs demonstrate feasibility; enforcement against chronic under-reporters is unresolved, and that population is precisely the future supplement population.
- Geographic cost variation. A 150%-of-poverty threshold means materially different things across regions. Regional indexing introduces distortion; the trade-off is deferred.
- The early-decade window. Maximum cost, minimum accumulated balance, minimum constitutional protection. The framework is designed to traverse this window quickly via cohort phasing, but the risk during it is real and not eliminable on paper.
Part 8: Inheritance as a Structural Feature
A final consequence follows from the titled-account structure and is worth isolating, because it distinguishes funded retirement from pay-as-you-go in a way that compounds across generations.
Current Social Security is an annuity. It pays until death and then stops. A worker who contributes for 40 years and dies at 62 receives little and leaves nothing. Because life expectancy correlates positively with income, this embeds a regressive transfer: shorter-lived, lower-income populations systematically subsidize longer-lived, higher-income ones. The transfer is invisible because nothing is itemized, but it is real.
Funded, inheritable accounts invert this. The worker who dies at 62 leaves an account balance — potentially $800,000 to $1.2 million — to heirs. For the populations the annuity structure treats worst, this is among the framework's largest improvements, and it operates automatically rather than through any targeted program.
At scale, the aggregate effect is a wealth-formation mechanism without historical precedent. Where pay-as-you-go consumes 12.4% of wages and leaves no residual, a funded system passes most of the residual to the next generation. At maturity, plausibly $500–700 billion annually in unspent balances transfers to heirs — wealth that under the current structure simply does not exist. Because the mandatory system builds balances across the entire income distribution rather than only among voluntary savers, the base of this wealth formation is universal rather than concentrated. The inheritances are large enough to be consequential (six figures to roughly $1 million per heir, typically landing when heirs are in their 50s) but well short of dynastic.
A second-order effect: a bequest motive slows retirement drawdown, which softens the coordinated selling pressure a large retiring cohort would otherwise exert on asset markets, while enlarging the eventual transfer. A 10-year distribution requirement on inherited balances (mirroring current inherited-IRA law) smooths the resulting consumption wave.
Summary
The current system fails on two independent axes — it cannot fund its promises past ~2034, and its promises were never adequate. Removing the payroll tax cap generates enough revenue to address solvency, but applied to the existing structure it merely extends an inadequate benefit. Applied instead as transition financing for a funded, individually titled replacement, the same revenue purchases a system that is both solvent and adequate: a median worker retires on roughly 2.5× current benefits at conservative return assumptions, low earners are carried across the adequacy threshold primarily by a birth investment whose 65-year horizon makes it the cheapest mechanism available, and the resulting balances are constitutional property that no future legislature can revoke and that pass to the next generation as inheritable wealth.
The costs are real and named: an $8–9 trillion net transition over 30 years, an unbuilt disability module, a coverage gap for non-employer workers, and a valuation-compression risk that the conservative return assumption is chosen to absorb. The framework is constructed to degrade gracefully — every parameter has a workable range, and reductions in the carve-out or cap rate slow its maturation from 40 years toward 55–60 without changing the destination.
The arithmetic is the argument. The full proposal, with year-by-year financing tables and complete assumptions, is available for anyone inclined to check it.
No comments:
Post a Comment