The average retired worker on Social Security currently collects about $1,976 a month. With the 2.8 percent cost-of-living adjustment now in effect, that check includes roughly $56 more than it did before. But before the money hits a bank account, a higher Medicare Part B premium claims nearly a third of it, leaving retirees to stretch what remains across groceries, utilities, prescriptions, and every other bill that has gotten more expensive over the past year.

The gap between the official raise and the actual cost of being retired in America is not new. The 2026 numbers, though, make it especially hard to ignore.

The raise, by the numbers

The Social Security Administration confirmed the 2.8 percent increase last fall. By law, the annual adjustment is pegged to the Consumer Price Index for Urban Wage Earners and Clerical Workers (CPI-W), comparing third-quarter data from one year to the next. The Bureau of Labor Statistics published the underlying inflation figures in its regular CPI release, and those numbers drove the calculation.

According to the SSA’s 2026 benefit fact sheet, the typical retired worker’s monthly check rose by about $56. That follows a 2.5 percent COLA in 2025, a 3.2 percent adjustment in 2024, and the unusually large 8.7 percent bump in 2023 that was fueled by post-pandemic inflation. The pattern is straightforward: as headline inflation cools, the annual raise shrinks. But many of the costs that dominate a retiree’s budget have not cooled at the same pace.

Medicare takes its cut first

The Centers for Medicare and Medicaid Services set the standard Medicare Part B premium at $202.90 per month for 2026, an increase of $17.90 over the prior year. The annual Part B deductible also rose to $283. Because Part B premiums are typically deducted directly from Social Security payments, that $17.90 monthly hike disappears before a retiree ever sees the deposit.

That leaves roughly $38 of the monthly raise after Medicare’s cut. Put another way, the net gain works out to about $1.25 a day. That has to cover every other price increase a retiree faces: food, housing, transportation, and out-of-pocket medical costs that Part B does not touch. Retirees who also carry Part D prescription drug coverage or a Medigap supplemental policy face additional premium increases that further erode the practical value of the COLA.

Some relief exists for lower-income beneficiaries. Roughly one in four Medicare enrollees qualify for assistance through programs like Medicare Savings Programs or Extra Help, which can cover some or all premium costs. Eligibility details are available through the Medicare.gov website or by contacting local State Health Insurance Assistance Programs (SHIPs).

A formula built for workers, applied to retirees

At the heart of the mismatch is the index itself. CPI-W tracks the purchasing patterns of urban wage earners and clerical workers, a population that skews younger and spends proportionally less on healthcare and more on commuting, clothing, and education than a typical 70-year-old.

The Bureau of Labor Statistics maintains an experimental index called R-CPI-E designed to measure inflation as experienced by Americans aged 62 and older. It gives greater weight to medical care and housing, two categories where prices have consistently outpaced overall inflation. Historically, R-CPI-E has tended to run roughly 0.2 to 0.3 percentage points higher than CPI-W on an annual basis. Over a 20-year retirement, that seemingly small difference compounds into thousands of dollars in lost purchasing power.

BLS itself acknowledges that CPI-W may not reflect seniors’ spending patterns, but R-CPI-E remains a research series, not an official benchmark. No federal agency has published a direct comparison of what the 2026 COLA would have been under R-CPI-E, so the precise gap this year is difficult to quantify.

Why the formula hasn’t changed

Advocacy groups and some members of Congress have periodically pushed to switch the COLA calculation to an elderly-specific index. The case is intuitive: if the adjustment is meant to protect retirees’ purchasing power, it should reflect what retirees actually buy. Opponents counter that experimental indexes carry less statistical rigor, could introduce volatility into benefit calculations, and would increase long-term costs for a program already facing a projected trust fund shortfall. The Social Security 2025 Trustees Report projects that the Old-Age and Survivors Insurance trust fund will be unable to pay full benefits sometime in the mid-2030s without congressional action.

To date, no legislation mandating a formula change has advanced to enactment. CPI-W remains the governing standard, and barring congressional action, it will determine the 2027 COLA as well.

What retirees are actually paying

Beyond the Medicare premium, the expenses that weigh most heavily on older adults have been rising faster than broad inflation measures suggest. Medical care services, including doctor visits, hospital stays, and dental work, have climbed at a pace above overall CPI in recent months, according to BLS data. Housing costs, whether rent or property taxes and insurance for homeowners, have remained stubbornly elevated in many parts of the country. Grocery prices, while no longer spiking at pandemic-era rates, have not retreated to pre-2021 levels.

Meanwhile, uncertainty around trade policy and tariffs has raised questions about whether consumer prices could accelerate again later in the year. If they do, the 2027 COLA calculation, which will be based on third-quarter 2026 inflation data, could reflect that shift. But any resulting increase would not reach retirees until well after the fact, creating a lag that leaves seniors absorbing higher costs in real time while waiting for the formula to catch up.

A raise that doesn’t keep pace

The verified numbers for 2026 are not in dispute: a 2.8 percent COLA, an average monthly increase of about $56, and a Medicare Part B premium that immediately reclaims a significant share. For the roughly 68 million Americans receiving Social Security benefits, the question is not whether the raise is real. It is whether a formula designed decades ago for a different population can meaningfully protect the purchasing power of people who depend on it most.