Monday, October 05, 2026

What Would It Take to Fix Social Security? Part 3

 

by Alvin Blackshear  |  Historian & Researcher

The first article in this series measured a problem, a 75-year shortfall of 4.42 percent of taxable payroll. The second article laid out ten tools and showed that each one hands the cost to someone different. This final article assembles the tools into packages and lets the arithmetic say what each package asks of whom.

Defining solved

The usual yardstick is a 75-year actuarial balance of zero. The Social Security actuaries draw a finer line. Sustainable solvency requires that trust fund assets stay positive throughout the period and stop falling as a share of annual cost by its end. A package that merely touches zero could still leave reserves sliding toward depletion just beyond the horizon, which would simply hand the problem to the next generation.

A caution about what follows. These packages are illustrative. Each stacks stand-alone scores from the Chief Actuary's 2026 provisions booklet, and the actuaries warn that provisions overlap, so real combined results would differ. None is presented as a verified solution, and each would need formal scoring. All three aim at roughly the same financial target so the comparison stays fair, and each discloses its tax increases, its benefit reductions and its one enhancement, a stronger minimum benefit.

Package

Components and stand-alone score

Sum

A Revenue-heavy

Eliminate taxable maximum (about 50%), raise rate 2.4 points by 2053 (40%), chained CPI-W COLA (15%), minimum benefit (minus 3%)

about 102%

B Benefit-heavy

Retirement age to 69 (25%), price indexing at 40th percentile (33%), COLA cut half point (24%), rate up one point by 2037 (20%), minimum benefit (minus 3%)

about 99%

C Mixed

Taxable maximum to 90% (24%), rate up one point (20%), price indexing at median (27%), retirement age to 68 (13%), chained CPI-W COLA (15%), minimum benefit (minus 3%)

about 96%

What the packages reveal

Package A shows how much revenue the problem demands. The Committee for a Responsible Federal Budget puts full elimination of the taxable maximum near half the shortfall, so even a package leaning on high earners needs a sustained rate increase on everyone else, plus a small benefit adjustment. Package B makes visible what the phrase entitlement reform tends to hide, which is that later retirement, slower initial benefits and a smaller cost-of-living adjustment must all stack together to reach the same destination. Package C spreads the load across several groups without claiming to be the moderate choice. It lands a few points short, a useful reminder that real packages need fine tuning.

A fourth variant protects everyone already retired or close to it. The CRFB finds that changes aimed only at new beneficiaries would have required a 30 percent cut if enacted today, compared with 25 percent for all beneficiaries, and by 2034 would fall short even if new benefits were eliminated entirely. Sparing the old shifts the burden to the young, in steeper form.

Three people, one reform

Consider a 70-year-old, a 50-year-old and a 25-year-old. The 70-year-old faces only the cost-of-living change, since the retirement-age and formula provisions apply to later cohorts and payroll rates apply to workers. The 50-year-old, who reaches 62 in 2038, pays higher rates for years, meets the new retirement age and formula, and then absorbs the COLA. The 25-year-old experiences every provision for decades. Employers, meanwhile, share every rate increase, since the payroll tax is split evenly. Identical legislation, three different reforms. The same holds across earnings, since high earners bear the tax side while lower earners are shielded by the price-indexing design and the minimum benefit.

The clock is ticking

Timing is arithmetic, not politics. The CRFB estimates that acting today requires a 4.25 point payroll tax increase or a 25 percent benefit cut, while waiting until 2034 raises the required change by about 15 percent. Fewer people left to share an adjustment, and less time to phase it in, mean a larger adjustment for those who remain. Each year of delay shrinks the group that can be asked to contribute and the window in which anyone can plan around the change.

The question that remains

My first article found a financing problem. The second article showed that many tools towards a solution exist. This final article shows that sufficient combinations can be built, yet each allocates costs and protections differently. The unresolved question is how Americans want those costs divided among workers, employers, retirees, higher and lower earners, and generations not yet born.

Where the PROMISE Act fits

That is the question the PROMISE Act, S. 4979, is designed to force. Senator Dick Durbin introduced it on July 14 with bipartisan cosponsors, among them Delaware's Senator Chris Coons. The bill chooses nothing from this series. It would direct the Social Security Advisory Board to develop recommendations and legislative language capable of paying full scheduled benefits for at least 50 years, then move the result through Congress on an expedited track. Supporters, including the Committee for a Responsible Federal Budget, see a way past years of delay. AARP opposes the bill, citing limited scrutiny, amendment and lack of debate. Notably, its 50-year horizon is shorter than the 75-year window commonly used. Delaware readers might put the series' questions to Senator Coons directly. Who pays, who is protected, and when?

Sources

SSA Office of the Chief Actuary, Summary of Provisions (2026 Trustees basis)
https://www.ssa.gov/OACT/solvency/provisions/summary.html

CRFB, Analysis of the 2026 Trustees Report (June 9)
https://www.crfb.org/papers/analysis-2026-social-security-trustees-report

S. 4979 (govinfo.gov)
https://www.govinfo.gov/app/details/CRI-2026/CRI-2026-PROTECTING-RETIREMENT-OPPORTUNIT-CA7282

AARP policy statements
https://www.aarp.org/social-security/aarp-policy

https://www.aarp.org/social-security/trust-fund-report-2026
https://action.aarp.org/node/21685

No comments: