Inflation Calculator

What the CPI measures — and what it does not

The Consumer Price Index is a price index, not a cost-of-living index and not a receipt for your household. Each month the Bureau of Labor Statistics prices a fixed market basket of goods and services — food, housing, transport, medical care, recreation and the rest — in urban areas across the country, weighting each category by how much the average urban household spends on it. The result is expressed relative to the 1982–84 average, which is set to 100. An index of 331.655 means that basket costs about 3.3 times what it did in 1982–84.

Because both years are expressed on the same scale, converting is just a ratio: value_to = value_from × (CPI_to ÷ CPI_from). No compounding is involved — compounding only enters when you turn the ratio into an average annual rate, which this page reports separately.

Your inflation rate is not the CPI

The index tracks the average urban household's basket. Unless you are the average urban household, your personal inflation rate differs from the headline — and it can differ a lot. If rent, childcare, or a chronic medical condition dominate your budget, you have been running hotter than CPI-U. If you own your home outright and drive little, or you buy electronics and clothing, you may have run cooler. The index is also a fixed-weight measure: it compares the price of the same basket over time, so it does not fully capture the way shoppers switch to chicken when beef gets expensive. That substitution is real, and it is a known limitation, acknowledged by the BLS, not a conspiracy.

CPI-U and CPI-W are not the same number

The figures on this page are CPI-U, the index for All Urban Consumers — roughly 93% of the U.S. population. There is a second index, CPI-W, for Urban Wage Earners and Clerical Workers, a narrower group. CPI-W matters because Social Security's annual cost-of-living adjustment is based on it, as are many union contracts and some tax provisions. The two usually move within a fraction of a percentage point of each other, because they measure overlapping baskets, but they diverge in years when the spending patterns of wage earners and clerical workers differ from the broader population. When someone says "inflation was 3%", the fair follow-up is which index, and over what period. The Social Security claiming calculator deals with the CPI-W consequence directly.

Why one "average inflation rate" is misleading

A single average compresses a century of very different decades, and the decade you pick does most of the work:

• 1970 to 1980 — the index went from 38.8 to 82.4. That is 112.4% total inflation, an average of 7.8% a year, with a 13.5% jump in 1980 alone. Anyone forecasting from that decade would have predicted the 1980s wildly wrong.
• 2010 to 2020 — the index went from 218.056 to 258.811, or 18.7% total: an average of just 1.73% a year. Anyone forecasting from that decade would have predicted 2021 and 2022 — up 4.7% and 8.0% — just as wrongly.
• 1913 to 2026 as a whole — an average of 3.16% a year. Useful as a long-run anchor, useless as a prediction for any particular decade.

There have been deflationary years too: the index fell 10.5% in 1921 and 0.4% in 2009. Prices falling is not the normal state of things, but it happens, and a calculator that only counts upward movement is telling a simpler story than the data does.

What this calculator cannot tell you

A headline index is an average across thousands of items, and the spread around that average is wide. Several categories have run well ahead of it over long stretches: medical care, college tuition, and rent in supply-constrained cities. Others have run behind or fallen in nominal terms: consumer electronics, clothing, and toys. So "inflation" as one number can be wrong about your life in both directions at once — you may be paying less for a television and much more for a doctor.

There are three further limits worth stating plainly. First, the index is national: the BLS publishes separate indexes for major metro areas, and they diverge, sometimes by more than a percentage point in a year. Second, the conversion is nominal — it says nothing about your income, taxes or investment returns. If your wages grew faster than CPI, you are ahead in real terms, and this page cannot see that. Third, a future projection is an assumption wearing a number's clothing: nobody knows the next thirty years of inflation, including everyone who will tell you that they do.

Frequently asked questions

How do I calculate what a past amount of money is worth today?

Divide the CPI index for the later year by the index for the earlier year, then multiply your amount by that ratio. $100 in 1995 becomes 100 × (331.655 ÷ 152.4) = about $217.62 in 2026. Both figures come from the same series, so only the ratio matters.

Which inflation index does this calculator use?

CPI-U, All items, U.S. city average, not seasonally adjusted — the BLS series CUUR0000SA0, with 1982–84 set to 100. The annual figures are annual averages, retrieved on 11 September 2026.

Is 2026 a complete year of data?

No. The value used here, 331.655, is the average of the eight months the BLS had published at the time of retrieval (January through August 2026). Final 2026 results will shift as the remaining months arrive, so treat any figure ending in 2026 as provisional.

What inflation rate should I use for a future projection?

There is no correct answer, which is the honest position. The Federal Reserve aims for 2% a year on the PCE index; CPI-U has averaged about 3.2% a year since 1913 and about 2.5% a year since the mid-1990s. A reasonable habit is to test a range — 2%, 3% and 4% — and see how much the answer moves, rather than trusting a single number.

Why is my personal inflation rate different from the CPI?

The index measures the average spending basket of urban households, not yours. Housing, medical care, childcare and college weigh differently in different lives, and the index assumes a fixed basket rather than the substitutions shoppers actually make. If rent and healthcare dominate your budget, your personal rate has probably run higher than CPI-U.

Can I use this to compare wages or salaries across years?

Yes, with care. Converting a 1995 salary into 2026 dollars shows whether pay kept pace with prices, which is what a real wage measure does. It does not tell you whether your standard of living improved, because taxes, benefits, household size and location all change. Convert the old and the new salary to the same year before comparing them.