The Benefora Social Security COLA Gap Index

COLAs are national: every beneficiary gets the same percentage wherever they live, while consumer prices move at different speeds from place to place. The index asks one narrow question: over the same July–September windows the COLAs measured, did the raises match what consumer prices did in each area? “Fell behind” and “kept ahead” on this page refer only to that comparison over Q3 2019 → Q3 2025.

• Not a cost-of-living comparison. BLS: “Can indexes for individual areas be used to compare living costs among the areas? No, an individual area index measures how much prices have changed over a specific period in that particular area; it does not show whether prices or living costs are higher or lower in that area relative to another.”
• Not a retiree price index. Local prices are measured with the CPI-U, which covers all urban consumers, retirees included. BLS publishes no local index for older Americans; its research index for Americans 62 and older (R-CPI-E) is national only.
• Not a statement about anyone's finances. The index does not measure incomes, savings, housing tenure or taxes, and does not show that retirees in any area are better or worse off. In BLS's words, “A national average reflects millions of individual price experiences; it seldom mirrors a particular consumer's experience.”
• Not a policy verdict. The index describes what happened; it takes no position on how COLAs should be calculated. On escalation generally, BLS writes: “We also recommend using national or regional indexes, due to the volatility of local indexes.”
• No timing lag in the headline. The comparison uses the same July–September windows as the COLAs. It does not capture the months between a price rise and the January payment that reflects it.

How the COLA works. SSA: “A COLA effective for December of the current year is equal to the percentage increase (if any) in the CPI-W from the average for the third quarter of the current year to the average for the third quarter of the last year in which a COLA became effective. If there is an increase, it must be rounded to the nearest tenth of one percent. If there is no increase, or if the rounded increase is zero, there is no COLA for the year.” The CPI-W is the national index; a quarter's value is the average of its three months. The COLA applies to December benefits, which are paid in January (ssa.gov). The COLAs in this edition:

• 1.3%, effective December 2020, payable January 2021
• 5.9%, effective December 2021, payable January 2022
• 8.7%, effective December 2022, payable January 2023
• 3.2%, effective December 2023, payable January 2024
• 2.5%, effective December 2024, payable January 2025
• 2.8%, effective December 2025, payable January 2026

Formula. C is the product of (1 + COLA/100) over those COLAs (1.268036). For each area, P is the average of the not-seasonally-adjusted all-items index values BLS publishes for July, August and September; L = P(2025) / P(2019). The gap is G = 100 × (C / L − 1): the percent change in the purchasing power of a benefit raised only by COLAs, measured against local prices. G is not the simple difference between the two percentage changes. For Tampa-St. Petersburg-Clearwater, FL, that difference is 26.8 − 35.15 = −8.35, while G is −6.18.

Months used per publication cycle. BLS publishes three areas monthly and the rest every other month. We use whichever July–September months BLS publishes for each area, the same months in both years, so every area's average is centred on mid-August like the COLA's third-quarter average. Nothing is interpolated and October data are never used.

• monthly (Jul, Aug, Sep): Chicago-Naperville-Elgin, IL-IN-WI; Los Angeles-Long Beach-Anaheim, CA; New York-Newark-Jersey City, NY-NJ-PA.
• bimonthly (odd months) (Jul, Sep): Tampa-St. Petersburg-Clearwater, FL; San Diego-Carlsbad, CA; Riverside-San Bernardino-Ontario, CA; Dallas-Fort Worth-Arlington, TX; Denver-Aurora-Lakewood, CO; Urban Hawaii; Boston-Cambridge-Newton, MA-NH; Minneapolis-St.Paul-Bloomington, MN-WI; Washington-Arlington-Alexandria, DC-VA-MD-WV.
• bimonthly (even months) (Aug): Miami-Fort Lauderdale-West Palm Beach, FL; Seattle-Tacoma-Bellevue WA; Phoenix-Mesa-Scottsdale, AZ; Atlanta-Sandy Springs-Roswell, GA; Detroit-Warren-Dearborn, MI; Baltimore-Columbia-Towson, MD; St. Louis, MO-IL; Philadelphia-Camden-Wilmington, PA-NJ-DE-MD; Houston-The Woodlands-Sugar Land, TX; San Francisco-Oakland-Hayward, CA; Urban Alaska.

Areas with fewer months in the average are noisier. The national, regional and division rows are monthly and use all three months.

Why CPI-U first and CPI-W as a check. The COLA uses the CPI-W, but BLS says the CPI-W population excludes “households of professional and salaried workers, part-time workers, the self-employed, and the unemployed, along with households with no one in the labor force, such as those of retirees.” (BLS). The CPI-U covers all urban consumers, retirees included, so it is the primary local measure. Local CPI-W uses the same population definition as the COLA and is shown as a robustness column; where the two put an area in different bands, the table marks it.

Why a 2019 base. Q3 2019 is the last pre-pandemic COLA measurement quarter and the end of the window measured by the COLA effective December 2019. By then all 23 areas were on the publication design BLS introduced in 2018, with no semiannual-only areas, and the Riverside index (which begins in December 2017) exists. A multi-year window also helps: BLS notes that standard errors “decrease on a relative basis … as the price change interval gets longer.”

Why bands, not precise ranks. BLS: “Each local index has a much smaller sample size than the national or regional indexes and is, therefore, subject to substantially more sampling and other measurement error. As a result, local-area indexes are more volatile than the national or regional indexes.” BLS publishes standard errors only for the nation and the four regions, and has written that “Standard errors for local metropolitan areas, such as Boston or Philadelphia, would typically be even larger.”

Illustrative dollars. The dollar column applies the COLAs to a fixed starting benefit: SSA's average retired-worker benefit for December 2019 ($1,502.85), from the SSA Annual Statistical Supplement 2025, Table 5.B8, all retired workers, average monthly benefit, December 2019. The figure is that benefit × (C − L); negative means the COLA-raised benefit ends that many dollars a month short of one that grew with local CPI-U. It describes a hypothetical continuing beneficiary. It ignores SSA's truncation of benefits to the dime and dollar, changes in Medicare Part B premiums, and taxes. SSA's published average benefit is not comparable with these figures: it changes as beneficiaries join and leave the rolls, not only through COLAs.

Data and checks. BLS states that CPI-U and CPI-W indexes are final when issued (this index uses the not-seasonally-adjusted series), so the results are reproducible. The build ran 23 automated checks (23 passed), including that each COLA matches a recomputation from SSA's CPI-W third-quarter averages and that the compounded COLAs match the national CPI-W ratio. Sources:

• BLS CPI-U all items, not seasonally adjusted (cu.data.1.AllItems, series CUUR...SA0)
• BLS CPI-W all items, not seasonally adjusted (cw.data.1.AllItems, series CWUR...SA0)
• BLS CPI area names (cu.area)
• BLS geographic revision 2018 (area definitions and publication schedule)
• BLS metro county lists (2018 definitions)
• BLS CPI questions and answers (area indexes, sampling error)
• BLS CPI variance estimates (standard errors published for the U.S. and four regions)
• BLS Beyond the Numbers (2012): Consumer Price Index data quality
• BLS notice on 2025 CPI collection reductions
• BLS FAQ on the 2025 federal government shutdown and the CPI
• BLS research index for Americans 62 and older (R-CPI-E)
• SSA cost-of-living adjustment history
• SSA CPI-W third-quarter averages
• SSA Annual Statistical Supplement 2025, Table 5.B8 (average retired-worker benefit)