U.S. consumer prices were 22.0% higher in October 2026 than in October 2021, according to a November 19 Bureau of Labor Statistics recap. That is a cumulative five-year comparison, not a 22% annual inflation rate or a new November 20 monthly release.
The nonseasonally adjusted Consumer Price Index for All Urban Consumers showed food prices up 24.0% over those five years and shelter up 27.8%. Those category figures describe the stated index periods; they do not establish what happened to an individual member's bills or income.
Keep the comparison period attached to the number
For credit union finance and member-service leaders, the distinction matters when explaining household cost pressure. Slower growth in an index does not mean earlier price increases have been reversed. Conversely, the accumulated increase does not tell an institution how much a household's budget changed this month.
An illustrative index-equivalent basket costing $100 in October 2021 would correspond to $122 in October 2026 at a cumulative 22% increase. That arithmetic is not a quoted grocery bill, a forecast or evidence that the same percentage increase occurred in each intervening year. Applying it uniformly to every member would add an assumption the national series does not supply.
The underlying October CPI release, published November 10, has a separate limitation: BLS could not publish 12-month changes for most October 2026 series because October 2025 survey data were unavailable following a lapse in appropriations. The five-year comparison uses October 2021, a different starting point; it should not be substituted for those missing annual rates.
A useful local discussion therefore keeps national price context separate from observed income, expenses and payment performance. The recap can frame a question about affordability, but it does not measure a credit union's delinquency, demonstrate a particular member's hardship or establish a lending outcome. Keep the dates and denominator visible before turning a striking number into an operating conclusion.