Inflation Calculator: What Today’s Dollar Will Buy Later

Linda gets a raise every January. Her Social Security statement calls it a cost-of-living adjustment, and most years it lands somewhere between 2% and 3%. She has never once felt caught up. Her Medicare Part B premium and her blood-pressure prescription have both climbed faster than her raise, every single year since she retired, and for a long time she assumed she was simply managing money badly. She was not. The index used to calculate her raise was never built to track what she actually spends on — and the government has known this, and published the alternative, since 1987.

Linda is a composite character — a stand-in for a pattern that shows up across millions of Social Security households, not a real benefit record. Her numbers are invented. The index gap that keeps outpacing her is not.

What your raise is actually measuring

By law, Social Security’s annual cost-of-living adjustment is calculated from the Bureau of Labor Statistics’ Consumer Price Index for Urban Wage Earners and Clerical Workers — CPI-W — comparing the average of July, August and September each year against the same three months a year earlier. CPI-W was built to track the spending of working, urban wage earners and clerical workers, a population that by definition is employed. It covers a little under a third of the country’s population, and almost none of it is retired.

The index built for exactly this problem, sitting unused

In 1987, Congress directed the Bureau of Labor Statistics to build an experimental price index for Americans aged 62 and older — the CPI-E. It weights medical care and shelter more heavily, because that’s where older households actually spend a larger share of their budget. BLS’s own reconstructed data shows the gap plainly: from December 1982 to December 2007, the experimental CPI-E rose 126.5%, compared with 110.0% for the CPI-W actually used to calculate Social Security’s raise. Over that 25-year stretch, the index that tracks how retirees really spend ran more than 16 percentage points hotter than the index their checks are tied to.

Linda’s actual arithmetic

Linda has been retired for eleven years. If her benefit had been adjusted using the historical CPI-E/CPI-W gap rather than CPI-W alone, her check today would run meaningfully higher than it actually does — not because anyone miscalculated her COLA, but because the index behind it was never designed to capture the medical-care-heavy basket a 71-year-old actually buys. The shortfall doesn’t show up as a single missed payment. It shows up as eleven Januaries in a row where the raise never quite covered what her actual costs did, compounding quietly into a real gap between her income and her cost of living that no single year’s statement reveals.

Why medical care is where the gap concentrates

Prices for medical care and shelter increased more rapidly than overall inflation during the exact period BLS studied, and both are weighted more heavily in the CPI-E precisely because older households spend more of their budget there. A retiree who assumes the headline inflation number describes their own situation is almost always understating what their actual basket — heavier on prescriptions, premiums and housing — is doing to their spending power.

What to actually do about it

Don’t plan retirement healthcare spending using the headline CPI number your COLA is based on — run those specific costs at a rate closer to the CPI-E’s historical pace instead, since that’s the basket that actually resembles a retiree’s. Track your own Medicare Part B premium and prescription costs year over year rather than assuming the January raise covers them; the gap is exactly where a retirement budget quietly breaks. And build a specific cushion for medical inflation into any retirement projection, separate from the general cost-of-living line, since the two are not moving at the same speed.

What this does not mean

This is not a claim that Social Security’s COLA is rigged or that the calculation is done incorrectly — CPI-W is exactly what the law specifies, calculated exactly as the law describes. It’s also not a reason to expect Congress to switch to CPI-E soon: the CPI-E remains explicitly experimental, is built from a smaller sample with larger sampling error, and switching would increase program costs, which is precisely why it has stayed a research product since 1987 rather than the statutory basis for the raise. The point is narrower: know that the index setting your raise and the basket you actually buy are measurably different things, and plan your own numbers accordingly rather than assuming the January letter means you’re fully covered.

Frequently asked questions

Is Social Security’s COLA calculated incorrectly?

No. It’s calculated exactly as federal law specifies, using CPI-W data BLS publishes every month. The issue isn’t an error in the math; it’s that the statute picked an index built for a working population rather than a retired one.

Why doesn’t the government just switch to CPI-E?

The CPI-E is still labeled experimental by BLS, drawn from a smaller sample with more sampling error than CPI-W, and adopting it would raise program costs — all reasons Congress has kept it as a research index rather than the statutory basis for benefit increases, even though BLS has published it since the 1980s.

Does this mean my COLA is worthless?

No — it still adjusts benefits upward most years and is meaningfully better than no adjustment at all. The point is that it tracks a working household’s basket, not a retiree’s, so treating it as a full measure of your own cost-of-living change tends to understate what’s actually happening to your budget.

What should I actually check each year?

Compare your COLA percentage against your own Medicare Part B premium change and your specific prescription costs, not the headline inflation number, since medical care is exactly where the CPI-W/CPI-E gap has historically concentrated.

Statutory sources, all official: Social Security Administration, Cost-of-Living Adjustments and the Consumer Price Index; Bureau of Labor Statistics, Experimental CPI-E for Elderly Americans, 1982-2007; Congressional Research Service, A Hypothetical COLA Based on CPI-E. The framing of this as a compounding “gap” over a specific retiree’s timeline is Linqz’s own analysis of the published BLS comparison, not a claim made by BLS or SSA themselves.

---

Disclaimer: General information, not financial or tax advice, and Linqz is not a registered investment adviser. “Linda” is a composite character with invented finances, not a real person. COLA methodology, CPI-E’s experimental status, and Medicare premium levels are set or reviewed by federal agencies on their own schedules and were last checked against ssa.gov and bls.gov sources in July 2026 — verify current-year figures before relying on them, and consult a qualified professional about your own retirement plan.