COLA vs. Inflation: Does the Raise Actually Keep Up?

COLA vs. CPI-U, 1976–2024

■ COLA (SSA determination year)   ■ CPI-U annual % change (BLS)

Chart reads: the COLA determined in a year (paid from the following January) against that calendar year's CPI-U annual-average change. The 2009 annotation is the zero floor at work: CPI-U fell 0.4% but the COLA was 0.0%, not negative — the one structural feature that always favors beneficiaries.

The full data table, 1975–2024

Sources: BLS CPI-U (series CUUR0000SA0, annual average of monthly indexes) via the BLS Public Data API; COLA from the SSA's official series. The gap column is a comparison, not the COLA formula — the actual inputs are third-quarter CPI-W averages, worked through with real numbers on the how COLA is calculated page.

Why the two lines legitimately differ

1. A different index: CPI-W, not CPI-U

The COLA tracks the Consumer Price Index for Urban Wage Earners and Clerical Workers — the spending of a working population, roughly 29% of Americans — not the all-urban CPI-U most headlines quote. The two rarely move identically. The long-running third candidate is the CPI-E, BLS's experimental index for people 62+, which weights medical care and shelter more heavily and which BLS has long characterized as rising about 0.2 points a year faster than CPI-W since the early 1980s; bills to adopt it (and, from the other direction, to adopt the slower chained CPI) surface in nearly every Congress. The index choice, compounded over decades, matters more than any single year's number — run the arithmetic on the cumulative calculator.

2. A different window: the third quarter, with a ratchet

The COLA compares only July–September averages, year over year, and the comparison base is the last third quarter for which a COLA was paid. After the 2009 price decline, that ratchet produced 0.0% in both 2010 and 2011 even as annual-average prices recovered — the single biggest COLA-versus-inflation grievance of the modern era, and a deliberate (if debatable) feature of the statute.

3. The zero floor

When prices fall, the COLA is 0.0%, never negative — benefits are never cut by the formula. In 2009 that floor was worth +0.4 points against CPI-U. Over a full career this asymmetry quietly adds up in beneficiaries' favor.

What advocacy groups estimate

This box is not official data. The Senior Citizens League (TSCL), a nonpartisan seniors advocacy group, publishes an annual "Social Security Loss of Buying Power" study. Its method differs from everything above: it tracks price changes in a basket of goods and services retirees buy most (heavily weighted toward food, housing and healthcare costs since 2010) and compares them with the COLAs actually paid over the same span. On that methodology its studies have concluded that benefits have lost substantial ground since 2010 — a conclusion driven by the basket choice and the 2010 start date (immediately after the 5.8% 2008 determination, before the zero years). The figure changes with each edition; consult the current study for its latest number. This site quotes TSCL's COLA forecasts on the prediction track record page — a different, checkable output — but does not blend advocacy estimates into its official tables. TSCL's headline buying-power estimate itself (13.7% since 2016) is presented with its methodology and the dispute around it on the Social Security purchasing power page.