Opinion: Social Security's 2027 raise shows a formula no one in Washington wants to fix

On October 14, the Social Security Administration announced that benefit checks will rise 3.6 percent in January 2027, the largest cost-of-living adjustment in three years, adding about $75 a month to the average retired worker's check. For the nearly 72 million people who depend on Social Security, that should have been unambiguously good news. Instead, within a day advocacy groups were calling the raise inadequate, and the complaint is not simply political theater. The formula that sets the annual adjustment is built to lag behind the inflation it is supposed to offset, and the one bill in Congress that would narrow that lag has been introduced, reintroduced, and left to die for more than a decade, because fixing it collides with a second, harder problem: Social Security's trust fund is running out faster than official projections said a year ago.

A formula that is out of date before it is even announced

The COLA is not pegged to inflation in general. It tracks the Consumer Price Index for Urban Wage Earners and Clerical Workers, or CPI-W, a market basket built around what working-age wage earners buy rather than what retirees and people with disabilities actually spend money on. Each fall, the agency compares the average CPI-W for July through September against the average for the same three months in the year of the last adjustment; the percentage difference becomes the following January's raise. That design means the number is already stale by the time it is published. The Bureau of Labor Statistics' own September inflation report, released the same morning as the COLA announcement, showed prices up 3.7 percent over the prior 12 months on the broader CPI-U measure and 3.9 percent on CPI-W itself, both higher than the 3.6 percent figure that governs January's checks. The gap exists because the official adjustment averages three months of data rather than comparing single months, so by the time the extra $75 lands in a retiree's account, the cost of living it was meant to match has already moved past it.

What the raise will not cover

A meaningful share of the increase will never reach beneficiaries' pockets. Most Medicare enrollees have their Part B premium withheld directly from their Social Security payment, and that premium has been climbing faster than overall prices. The Centers for Medicare & Medicaid Services set the 2026 standard Part B premium at $202.90 a month, up 9.7 percent from $185 in 2025, and the Medicare trustees' own projections point toward roughly $209.50 for 2027, an increase CMS is expected to finalize this fall. A jump of that size would consume close to a tenth of the dollar value of the average retiree's entire COLA increase before a single grocery bill is paid. An AARP poll of adults 50 and older taken in September, before the official figure was released, found that most did not believe a COLA in the range ultimately announced would keep pace with their actual costs, and the group's reaction to the final number was similarly unimpressed. The Senior Citizens League, an advocacy group that tracks retiree purchasing power, was blunter. Its Shannon Benton said the increase was welcome but incomplete:

"A 3.6% COLA is certainly welcome, but it's unlikely to restore the purchasing power that seniors have already lost. It's another year, another COLA that falls short."

The group's broader complaint is about method, not just magnitude: because CPI-W is not designed around retirees, it understates the weight of medical care and housing in an older household's budget, two categories that have outpaced general inflation for years.

The fix Congress keeps shelving

There is a ready-made alternative. The Bureau of Labor Statistics has published an experimental Consumer Price Index for the Elderly, or CPI-E, for decades, weighting the basket toward the goods and services households over 62 actually buy. A freestanding CPI-E Act was first introduced in the House in 2011 and reappeared in similar form in 2019 and again in 2021, each time dying in committee without a vote. Legislation to make the index the official COLA yardstick resurfaced this year as part of the Social Security 2100 Act, H.R. 9519, introduced by Rep. John Larson in June. Its cost-of-living section would require the agency to use whichever index, CPI-W or a research version of CPI-E, produces the larger adjustment for the years 2027 through 2036. A Senate companion was referred to the Finance Committee the same month. None of these bills, across fifteen years of attempts, has ever reached a floor vote in either chamber.

The reason is not simply inertia. A richer COLA formula raises outlays precisely when the program's financing is tightening. The Social Security trustees' 2026 report moved up the projected depletion date for the retirement trust fund to the fourth quarter of 2032, a quarter earlier than the prior year's estimate, with incoming payroll tax revenue covering only about 78 percent of scheduled benefits after that point absent congressional action; the combined retirement-and-disability fund date held at the third quarter of 2034, when it could still pay roughly 83 percent of promised benefits. Switching to a higher-cost index without a financing offset would shrink that runway further, which is the honest reason both parties have let CPI-E legislation sit in committee since it was first proposed. No one in the current debate has proposed paying for a richer COLA with new revenue or offsetting cuts elsewhere, which is what an honest fix would require.

What comes next

None of this is an argument for leaving the formula alone. A measure that structurally undercounts what retirees spend on health care and shelter, two of the least substitutable costs in a household budget, is not a minor technical quirk; it is a slow, compounding transfer away from the people the program exists to protect. But it is also not a problem that can be solved by raising the index without confronting the financing question in the same breath, and lawmakers on both sides have preferred to let advocacy groups make that case publicly each October rather than answer it themselves. CMS is expected to confirm the 2027 Part B premium within weeks, which will show precisely how much of January's raise survives contact with Medicare. The next trustees report, due in the spring, will say whether the funding runway shortened again. Until Congress is willing to pair a better cost-of-living measure with a credible way to pay for it, every October's announcement will keep arriving a few months late, and a few tenths of a percentage point short, of the inflation retirees actually lived through. That is a policy choice, not an accident of arithmetic: lawmakers have had a technically workable alternative index sitting at the Bureau of Labor Statistics for more than a decade and have simply declined, year after year, to decide who pays for using it.