Inflation calculator: U.S. and 23 metro areas, month by month

One national figure, 23 local ones

When the news says inflation is 3.7 %, it quotes the U.S. city average of the CPI-U, an index that covers about 90 % of the population. The BLS builds it from prices collected in 75 urban areas, and it publishes some of those areas on their own: 23 metropolitan areas, the 4 census regions and the 9 divisions, and two size classes for the regions taken together. Each local index has its own base of 100 in 1982-84, which means it tells you how prices moved in that place, never whether the place is cheaper or dearer than another one.

The calculator uses the index of the area you pick. A rent of $1,500 signed in Boston in September 2019 and one signed in Houston that month do not need the same raise today to keep up with local prices: $1,931 in the first index, $1,851 in the second, using the Houston annual average because that index is not priced in September.

Twenty of the metro indexes are published every other month. Boston, for example, is priced in odd months and Philadelphia in even months; only New York, Chicago, Los Angeles come out every month. When you choose a month that an area does not publish, the calculator keeps that month in the list, labels it, and computes with the nearest earlier month that was published, naming it in the result. It never invents a value between two published ones.

What this site does that the BLS calculator does not

This site does better than the first Google result (the calculator of the BLS) because it computes the value of a dollar, a salary or a pension to the month in each of the 23 metro areas, 4 regions and 2 size classes published by the BLS, and by spending group (shelter, food, energy, medical care and more), where the official tool gives only the national average. It also explains the October 2025 gap instead of refusing the calculation in silence: the missing month is never filled, and the nearest published month is offered and named. Every page of the CPI by city section carries the full table of its area, with the latest month and its date.

October 2025: the month that does not exist

Federal funding lapsed from October 1, 2025 to November 12, 2025. The BLS could not collect October prices and could not collect them afterwards, so none of the indexes this site uses exists for October 2025: no area, no spending group, none of the three measures (CPI-U, CPI-W, chained CPI-U). In the BLS tables the cell shows a dash with the footnote "Data unavailable due to the 2025 lapse in appropriations". The 2025 annual average was computed from the 11 months that exist (5 for the areas priced in even months). One consequence will show up soon: in October 2026 there will be no 12-month change to report, because its base month is missing. The October 2025 page gives the full account, from the September index released late on October 24, 2025 to the rents carried forward until April 2026.

Pay, pensions and Social Security

A cost-of-living raise is only as good as the index behind it. Social Security uses the CPI-W, the index for wage earners and clerical workers, averaged over July, August and September: 328.746 in 2026 against 317.265 in 2025, which gives the 3.6 % increase the SSA announced on October 14, 2026 for benefits paid from January 2027, after 2.8 % in 2026. The Social Security COLA page lists every adjustment since 1975 and rebuilds those since 1984 from the index, and the salary calculator tells you what raise your own pay needs in your own city.

What this site does not do

It does not compare the cost of living between two cities: no CPI can, and the BLS says so. It does not forecast inflation, and a 2027 figure appears only once an official body has published it. It does not know your own basket: the published averages weight rent, gasoline and medical care by what all urban households spend, and your own inflation can differ (the BLS explains why). Nothing on this site is personal financial advice. The method page lists the series, the rules for missing months and the tests.