Social Security is headed
for a crisis —
Congress must act now.
Without reform, the trust fund runs dry in 2033. A bipartisan solution exists. Here's my plan to preserve Social Security for the long haul.
A program that transformed American retirement — now facing its greatest test.
Social Security was established in 1935 to reduce poverty among the elderly. At that time, roughly half of elderly Americans lived in poverty. Social Security brought that number down dramatically — today, only about 6 to 8 percent of elderly Americans live in poverty. According to the 2025 Social Security Trustees Report, the program provides benefits to about 68 million people, including 54 million retired workers and dependents, 6 million survivors of deceased workers, and 8 million disabled workers and their dependents.
Just over 40 percent of Social Security beneficiaries rely on the program for more than half their income. About 14 percent rely on it for 90 percent or more of their income. In total, Social Security keeps about 20 million elderly Americans and 1 million children out of poverty (per the Brookings Institution blueprint). This is not an abstract budget line — it is a lifeline.
But the program is in trouble. Starting in 2021, changing demographics forced Social Security to begin drawing down the Old Age and Survivors Insurance (OASI) trust fund to cover benefits owed to retirees. The Trustees estimate that the OASI trust fund will be exhausted in 2033. After that point, payroll contributions from current workers will cover only about 77 percent of earned benefits — a 23 percent across-the-board cut for every recipient.
The primary problem is demographics: Americans are living longer and having fewer babies. Other structural factors compound it — coverage gaps, investment constraints, and an eroding tax base — but no reform works without confronting the demographic reality head-on.
The parties have been divided. But a bipartisan path exists.
Senate reconciliation rules bar provisions that recommend changes to Social Security, which means neither party can push through changes on a party-line vote with a simple majority. That procedural reality has forced genuine bipartisan negotiation — and deadlock.
Broadly speaking, Democrats have proposed changes that increase revenue to restore solvency, while Republicans have generally proposed benefit reductions. Neither approach alone is sufficient, or politically viable. The four structural forces below explain why any single-lever approach is bound to fail — and why a comprehensive, balanced solution is the only real path forward.
The program was designed when life expectancy was much shorter and birth rates much higher. Today's retirees live longer, drawing benefits for more years. Fewer workers are entering the workforce relative to the number of retirees. Neither of these trends is likely to reverse.
The program is built on two principles: adequacy (benefits must protect against poverty for low-wage workers) and equity (benefits must be fair relative to contributions). Any reform must preserve both — and must not crowd out people's ability to save for retirement in other ways.
In 2024, roughly 6 percent of American workers held jobs classified as "uncovered employment" — mainly state and local government employees enrolled in alternative pension plans. Bringing more workers into the system is a key part of the solution.
Since 1935, the law has directed that the trust fund be invested only in Treasury bonds. Decades of experience — and smaller-scale models like the National Railroad Retirement Investment Trust — suggest we can manage the risk of broader investment more responsibly.
In February 2025, the Brookings Institution published a blueprint for a bipartisan solution designed to meet the program's needs for the next 75 years. The proposal includes revenue increases, targeted benefit reductions, benefit improvements, expanded coverage, and restructured trust fund policies — something for every stakeholder, and designed to avoid the poison pills that have derailed past efforts. My proposal adopts almost all of what Brookings proposes, with a few deliberate adjustments.
Social Security keeps approximately 20 million elderly Americans and 1 million children above the poverty line. This is the program's core purpose — insurance against poverty, not a retirement savings vehicle — and any reform must preserve that fundamental mission.
The last major Social Security reform was in 1983, when Congress gradually raised the full retirement age from 65 to 67, brought new federal employees into the program, and extended income tax to some benefits. Those reforms secured the program at full benefits for 50 years. True reform is attainable — if Congress is willing to act.
Five pillars of reform.
Pillars one through four follow the structure and categories of the Brookings Institution blueprint. Pillar five contains my own additions and adjustments to the Brookings proposal. Click any pillar to expand the policy details.
- Raise the Taxable Maximum Ceiling The taxable maximum is currently $184,500 (2026), adjusted annually by the national average wage index. Because the highest earners' incomes have grown faster than average wages for decades, the ceiling now covers only about 83 percent of all earnings — down from 90 percent in 1983. This proposal would increase the ceiling each year until it once again covers 90 percent of total earnings, then maintain that 90 percent target going forward, indexed to wages to hold it there.
- Close the Pass-Through Payroll Tax Loophole Owners of S-corporations and certain limited partners can categorize a significant share of their business income as distributions rather than wages, avoiding Social Security payroll taxes on that income. This proposal requires all active pass-through business owners who materially participate in their business to pay Social Security taxes on their full earnings up to the taxable maximum — the same obligation all other workers carry.
- Expand Benefit Calculation Years from 35 to 40 Social Security benefits are calculated using a worker's highest 35 years of earnings. This proposal would gradually increase that to 40 years, phasing in one additional year every two years beginning in 2032 and reaching 40 years by 2040. This better aligns the benefit calculation with actual lifetime earnings given longer working lives.
- Tax All Social Security Benefits of High Earners Currently, only 50 to 85 percent of Social Security benefits are subject to federal income tax, depending on income. This proposal would subject all benefits of single filers with adjusted gross income above $100,000 and joint filers above $125,000 to income tax, with those thresholds indexed annually for wage growth.
- Phase Out the Dependent Retiree Spouse Benefit A spouse can currently receive up to 50 percent of their partner's Social Security benefit even without their own earnings record. This proposal would gradually reduce and eliminate that benefit for new retiree beneficiaries beginning in 2027, phasing out over ten years. It would not affect anyone receiving the benefit before enactment, and would not apply to disabled spouses or widow(er)s.
- Eliminate Child Retiree Benefits Children of retirees currently receive Social Security benefits, a structure originally designed for children of disabled or deceased workers. This proposal would end benefits to children of retirees and the associated caretaker benefit for new beneficiaries beginning in January 2027. Everyone receiving these benefits in January 2027 would be protected. The benefit would continue for disabled children, adopted children, and children in the care of a grandparent or eligible relative.
- Five Brookings Improvements — All Adopted The Brookings blueprint proposes five targeted benefit improvements, all of which I support: increasing survivor benefits so that a widow or widower receives the higher of their own benefit or 75 percent of the couple's combined benefits; creating an Early Retirement Disability benefit for workers aged 58 and older who can no longer perform their past work but do not qualify for SSDI under current strict standards; restoring student benefits for children of disabled or deceased parents through age 25, including community college and trade school; providing a child benefit to grandparents and eligible relatives who are the primary caregivers for grandchildren; and improving benefits for disabled adult children by removing marriage as a disqualifying event.
- Expand Legal Immigration to Grow the Workforce The Brookings proposal specifies percentage increases to employment-based and family-based immigration caps — a 50 percent immediate increase to major employment caps followed by 3 percent annual growth, and 1.5 percent annual growth in other permanent categories. Based on those percentages, my calculations suggest these changes would add approximately 1.1 million new temporary workers and 1.2 million permanent residents to the U.S. economy over 10 years, with continuing growth thereafter. (These figures are my approximation from the Brookings percentages; Brookings' own SSA-modeled estimate projects a cumulative net immigration increase of 3.1 million above baseline over the same period.) Expanding the pool of workers paying into Social Security directly addresses the demographic challenge at the core of the program's funding gap.
- Achieve Universal Social Security Coverage About 6 percent of American workers — primarily 5.9 million state and local government employees — hold jobs not covered by Social Security, instead participating in alternative public pension plans. This proposal would require all newly hired state and local government employees to be covered by Social Security, with a five-year transition period beginning after enactment to allow pension plans to adjust. Low-cost federal loans would be available to help states manage near-term transition costs.
- Redirect Social Security Benefit Tax Revenue to OASDI Since 1993, revenue from taxing Social Security benefits has been directed to Medicare's Hospital Insurance trust fund. This proposal would redirect those proceeds into the OASDI trust funds as soon as doing so would not harm the Medicare fund — with the HI fund made whole through increased Medicare revenues and reduced Medicare Advantage outlays. The taxation of Social Security benefits is more directly tied to the Social Security program than to Medicare, and the revenue should follow that logic.
- No Increase to the Payroll Tax Rate The Brookings blueprint includes a 0.2 percentage point payroll tax increase — 0.1 percent each from employer and employee — as the final element to achieve full actuarial balance. I oppose it. Social Security is insurance against poverty, not a retirement savings program, and the payroll tax must not grow to the point that it crowds out workers' ability to save, invest, or pay down a mortgage. Once raised, payroll tax rates do not come back down.
- Adjust the Benefit Formula Rather Than the Retirement Age The Brookings proposal raises the full retirement age for workers in the top two income quintiles on the grounds that higher earners as a group live longer. I reject that reasoning. It attributes a characteristic of a population — higher average longevity — to the individuals within it. Any given worker in the top two quintiles may or may not live longer than average; their earnings bracket tells us nothing reliable about their personal longevity. Penalizing their retirement age based on a group statistical pattern is a category error. Instead, I support one of two Congressional Budget Office options that achieve a similar fiscal result more equitably. The first would add a new bend point in the benefit formula for earners above the 70th percentile, reducing the rate at which additional earnings translate to additional benefits at high income levels. The second would make the overall benefit structure more progressive by adjusting the percentages applied at existing bend points. Both options reduce benefit growth for higher earners without making assumptions about any individual's life expectancy.
- Minimum Benefit and Caregiver Relief — If Funding Allows Two further improvements not in the Brookings blueprint deserve serious consideration, subject to whether the overall reform package remains in actuarial balance: a meaningful minimum benefit for workers who contributed over a long career but whose lifetime earnings leave them at or near the poverty line; and relief for caregivers — predominantly women — who left the workforce for years to care for children or elderly relatives. Under current rules, those years are assigned zero earnings value in the benefit calculation, unfairly reducing benefits for a group that provided enormous social value.
- Reform — Not Repeal — the Windfall Elimination Provision and Government Pension Offset The WEP and GPO, enacted in 1983, were intended to correct a structural inequity: because non-covered earnings appear as zeros in Social Security records, workers who split careers between covered and non-covered employment were being treated by the benefit formula as if they were long-term low-wage earners, generating an unintended windfall. The Social Security Fairness Act of 2025 repealed both provisions entirely — eliminating the underpayments, but also opening the door to overpayments for workers whose non-covered pension income is now unchecked against their Social Security benefit calculation. The original formulas were a blunt instrument, designed with the data available in 1983. SSA now has decades of non-covered earnings records. We are capable of designing a proportional formula that gives every worker exactly the benefits they earned — no more, and no less. I support restoring a reformed WEP and GPO built on that principle.
- Repeal the Retirement Earnings Test The Retirement Earnings Test reduces Social Security benefits for recipients who continue working before reaching full retirement age and earn above a threshold. It creates confusion, discourages continued workforce participation, and sends a counterproductive signal at a moment when we need more workers in the labor force, not fewer. Repealing it removes a disincentive to work without affecting the ultimate benefit calculation.
- Invest Half the OASDI Trust Fund in a Total Stock Market Index Current law requires the Social Security trust fund to be invested exclusively in Treasury securities — a restriction written in 1935 during the Great Depression. Nearly a century of market history, and the demonstrated success of public retirement funds like the National Railroad Retirement Investment Trust that have invested in equities for decades, give reasonable grounds for confidence that the risk is manageable. I propose investing half the OASDI Trust Fund in a broadly diversified total stock market index fund. Even if the trust fund is near depletion when this is enacted, rebuilding it over time with market-rate returns will extend the program's long-term solvency more than Treasury-only investment can.
The difference between action and inaction.
The 2025 Trustees Report projects OASI trust fund exhaustion in 2033. Without reform, 23 percent of earned benefits are cut automatically — affecting 68 million Americans overnight. With the reforms above, the program can be stabilized for the next 75 years.
Americans are living longer and having fewer babies. That is the reality we face. With determined, bipartisan action — and without waiting any longer — we can preserve Social Security for everyone who has earned it, and for every generation that follows.