Abstract: Retirement security requires stronger private savings and a durable Social Security system. This paper proposes eight reforms to equalize saving opportunities across employment arrangements, reduce premature disbursements from retirement accounts, support spouses and caregivers, encourage early saving, improve investment choices, and partly finance assistance through a targeted levy on large inherited Roth balances. It also calls for a Social Security agreement combining revenue increases and changes in future benefits, with protections for workers nearing retirement. The central principle is that any reduction in future public benefits must accompany meaningful improvements in households’ ability to accumulate assets, especially for younger workers.
Introduction
Retirement security presents a different political problem from health care and student debt. Congress has demonstrated that bipartisan agreement on retirement saving is possible, including through the SECURE Act of 2019 and SECURE 2.0 in 2022. Yet incremental improvements have left substantial disparities in saving opportunities, while policymakers continue to postpone decisions about Social Security’s financing.
The Bernstein version of the Powell doctrine asks five questions about major domestic economic reform: Is the problem serious enough to justify substantial change? Can the reform avoid substantial harm to major groups? Is there a feasible transition? Is the resulting system economically and politically sustainable? Is there a more efficient way to achieve the objective?
My earlier proposals on health care and student debt use these questions to identify reforms that improve security while limiting disruption, correcting incentives, and using public resources more efficiently. Retirement policy requires the same discipline, including attention to fairness across generations.
This paper proposes eight reforms that recognize the inextricable connection between private retirement saving and Social Security. The private retirement savings reforms focus on facilitating increased saving by households, which are currently having difficulty saving. The Social Security reform proposal recognizes the necessity of protecting workers nearing retirement, young adults entering the workforce and future generations which need access to the program.
Retirement saving opportunities depend too heavily on where and how people work. Workers face unequal contribution limits, employer support, fees, investment choices, and legal protections. Scattered accounts and early withdrawals erode accumulated savings, while tax deductions provide the least help to many households struggling to save. Unpaid caregiving, gaps in health insurance, and onerous student debt further limit contributions. These weaknesses make Social Security reform inseparable from improvements in private saving: any reduction in future public benefits makes it more urgent to help households accumulate retirement assets.
Reform One: Equalize Saving Opportunities for all Workers
Retirement saving rules lack parity across employment situations. In 2026, someone relying solely on a conventional IRA faces a $7,500 annual contribution limit, compared with $24,500 in employee contributions to a 401(k).
Employer contributions can raise the combined 401(k) total to $72,000, leaving room for $47,500 from the employer when the employee contributes $24,500, subject to compensation and plan rules. An eligible self-employed worker can establish a solo 401(k) and contribute as both employee and employer, reaching the same $72,000 ceiling with sufficient earnings. These figures exclude catch-up contributions.
Saving capacity therefore depends heavily on access to a workplace plan or the ability to establish a business retirement plan.
Contribution limits, Roth access, and legal protections should be identical across employment arrangements. An employee-owned portable retirement account would give workers without workplace coverage the same saving opportunities. Employers, staffing firms, and platforms could contribute directly, including matching contributions for contractors and gig workers who serve them. Automatic enrollment with an opt-out would support saving across multiple jobs and income sources.
One combined limit would govern workers’ contributions across accounts, with a uniform ceiling for total worker and employer contributions. Common rules would govern after-tax contributions and Roth conversions.
Workplace plans should also meet minimum standards: employer contributions would vest within one year, all fees would receive standardized disclosure, and savers would have access to low-cost investments and installment withdrawals.
Reform Two: Make Rollovers and Account Consolidation Easier and More Automatic
Small retirement accounts can carry disproportionate fees: GAO found that fees exceeded investment returns in most forced-transfer IRAs it examined. Scattered accounts are also easier to lose track of. In a separate study, 17 states reported receiving $35 million in unclaimed retirement savings in 2016. These transfers to state custody, commonly called escheatment, can leave owners facing another hurdle to recover their savings.
Retirement savings should follow the worker. Each saver would designate a primary retirement account to receive transfers and federal matching contributions.
When employment ends, an inactive account below a statutory threshold would automatically transfer to the designated account after notice unless the worker elects otherwise. Larger transfers would require longer notice and clear information about fees, investments, withdrawal rights, and legal protections.
Current tax rules create another obstacle: pre-tax IRA balances count in the pro-rata calculation for Roth conversions, while 401(k) balances do not. Consider someone with $63,000 in a pre-tax 401(k) who contributes $7,000 after tax to an otherwise empty traditional IRA and immediately converts it to Roth. Leaving the $63,000 in the 401(k) makes the conversion tax-free, assuming no investment gains or other IRA balances. Rolling that money into an IRA first makes 90 percent of the $7,000 conversion—$6,300—taxable. The rule therefore encourages keeping savings in a 401(k), even when IRA consolidation would otherwise make sense.
Pre-tax workplace assets placed in a separately reported qualified rollover subaccount would be excluded from that calculation. They would remain taxable when withdrawn or converted, with strict reporting rules preventing ordinary pre-tax IRA contributions from being relabeled as workplace rollovers.
Reform Three: Preserve a Protected Core While Providing Limited Emergency Access
Premature withdrawals undermine retirement security, particularly among households already struggling financially. Vanguard reports that about 6 percent of eligible participants took a hardship withdrawal in 2025, compared with 4.8 percent in 2024 and 3.6 percent in 2023. Participants earning less than $100,000 were approximately 3.5 times more likely to take hardship withdrawals than higher earners. My paper, “Pre-retirement use of 401(k) funds,” published in Business Economics in 2021, found that people borrowing or taking hardship distributions from their accounts tended to have other debts and poor credit ratings. Early access can relieve immediate financial pressure while leaving vulnerable households with fewer resources for retirement.
All new retirement accounts would preserve at least half of cumulative contributions and all investment earnings until retirement. Cumulative withdrawals before retirement could not exceed half of cumulative contributions. Permitted withdrawals would carry no early-withdrawal penalty, although ordinary income taxes would apply to pre-tax amounts. The protected portion could not be withdrawn or borrowed against.
Keeping half of contributions accessible should encourage participation among households concerned about tying up their money, allowing accounts to support both emergency needs and retirement saving.
Reform Four: Expand Access to the Saver’s Match, Encourage Spousal IRAs, and Recognize Caregiving
Federal assistance should help more households build retirement savings. I would retain the Saver’s Match at 50 percent of the first $2,000 contributed—a maximum $1,000 federal contribution per eligible person—and make participation easier for nonworking spouses and caregivers.
I propose automatic spousal IRA enrollment with an opt-out and voluntary contributions from household earnings. Eligible contributions would receive the same Saver’s Match available to other savers, without a separate spousal subsidy. Spousal contribution eligibility should also extend to couples filing separately.
Congress should adopt the central provision of the bipartisan Improving Retirement Security for Family Caregivers Act, allowing qualifying unpaid caregivers to contribute to Roth IRAs despite insufficient earnings. Eligible caregivers who contribute would receive the same Saver’s Match.
Reform Five: Increase Incentives for Saving by Children and Young Adults
I enthusiastically support incentives that help children and young adults begin saving earlier. Congress should make the Trump Account program’s $1,000 federal seed contribution permanent beyond the currently eligible 2025–2028 birth cohorts and encourage employer and philanthropic contributions. Private commitments—including SpaceX President Gwynne Shotwell’s pledge of more than two million shares for eligible children and gifts pledged by Michael and Susan Dell and Ray Dalio—demonstrate the potential to supplement public support. Additional federal assistance should favor children from families with limited resources, drawing on Senator Cory Booker’s proposed income-targeted deposits. Earlier saving would complement the proposed withdrawal rules by building both accessible savings and a protected retirement balance.
Policy should also remove barriers to early saving. Ordinary IRA contributions generally require earned income, limiting access for children without earnings. FAFSA excludes retirement accounts from assets, but withdrawals can count as income and reduce aid eligibility. The provision allowing qualifying unused 529 savings to move into a beneficiary’s Roth IRA, subject to a $35,000 lifetime limit and other restrictions, is a welcome step toward connecting education and retirement saving.
Reform Six: Partly Finance Retirement Assistance Through a Targeted Roth Estate Levy
Motivated by proposals to tax exceptionally large Roth IRAs such as Peter Thiel’s, I propose a modest levy on large Roth balances after both the retiree and any surviving spouse have died. The tax would apply only to aggregate Roth retirement assets above an inflation-indexed $1 million threshold, at a rate no higher than 7 percent—and potentially substantially lower.
Roth accounts would remain a great deal for retirement savers: tax-free growth and qualified withdrawals throughout both spouses’ lives, followed by a low tax applying only to the portion of a large remaining balance above the threshold. Accounts below that threshold would owe nothing under this proposal. Receipts would help finance broader retirement saving opportunities, with the final rate determined by revenue estimates and the cost of the assistance.
Reform Seven: Improve Investment Options, Including Access to Series I Bonds and Individual Bonds
Every qualified retirement plan should offer low-cost diversified stock funds and practical options for preserving principal and managing inflation risk. Congress should permit retirement accounts to hold Series I savings bonds, which combine federal principal protection with inflation-adjusted interest. Workers should also have access to individual Treasury securities, including TIPS, and diversified investment-grade corporate bonds—options already available in many brokerage IRAs but often missing from workplace plans.
Individual bonds allow savers to arrange maturities around anticipated retirement expenses. A Treasury held to maturity provides a known principal payment; corporate bonds carry default risk. Conventional bond funds provide convenient diversification but generally lack a maturity date at which a specified principal amount is returned. Workers should have access to both approaches.
The administration’s August 2025 executive order instead promoted alternative assets, including private credit, in workplace retirement plans. Why prioritize investments with difficult valuations, substantial fees, and withdrawal restrictions when basic bond options remain unavailable to many workers? My proposal would exclude private-credit funds and comparable illiquid, difficult-to-value investments from qualified plans and portable accounts, while requiring diversified, low-cost, transparent, and liquid default investments.
Reform Eight: Restore Social Security’s Finances While Protecting Future Generations
Social Security reform involves both an ideological choice and an immediate financing constraint. The 2026 trustees’ report projects that the retirement and survivors trust fund will exhaust its reserves in late 2032, when continuing income would cover only 78 percent of scheduled benefits. Congress must decide how to close that gap without imposing abrupt losses or unfair burdens across generations.
Andrew Biggs emphasizes the economic case for benefit restraint: lower future benefits can encourage additional work and private saving, increasing investment and economic output. On the left, the Social Security Expansion Act sponsored by Bernie Sanders and Elizabeth Warren would raise taxes on high incomes to finance both solvency and expanded benefits. These approaches reflect different judgments about public benefits, private assets, and taxation.
I believe a durable agreement will require both revenue increases and changes in future benefits. Economic growth matters, but it does not make benefit reductions harmless. The Penn Wharton Budget Model’s comparison of seven reform packages illustrates the tradeoff: packages with larger benefit reductions produce stronger modeled long-term growth, but also larger losses for older workers and people approaching retirement. Higher taxes can preserve benefits while burdening younger generations and reducing their capacity to accumulate assets.
Timing limits the available choices. Raising the retirement age only for very young workers would do little to address the early-2030s financing shortfall. Applying it quickly enough to generate substantial near-term savings would impose losses on people with limited time to adjust. I would not support changing the retirement age for people already in their forties or fifties. Any increase should apply to substantially younger cohorts, with decades of notice and protections for those unable to work longer.
My concern extends beyond preserving current benefits. A durable settlement should help future retirees build private assets alongside Social Security and avoid repeatedly asking younger generations to pay more. Congress should assess any agreement for both lasting solvency and its effects on younger workers’ ability to save.
Conclusion
Social Security and private retirement savings must be reformed together. Any reduction in future public benefits must be accompanied by meaningful improvements in households’ ability to accumulate assets, including relief from health insurance gaps and onerous student debt that impede saving. The objective is to protect current retirees while giving younger generations sufficient resources and time to prepare.
Most of this agenda can be enacted through the tax reconciliation process but Social Security reform will require a separate agreement under current law.