Confirmed: In the latest BLS release available on 1 November 2026, consumer prices rose 3.7 percent over the 12 months ending in September 2026, and the CPI-U level stood at 336.940 — about 31 percent above December 2019. The Federal Reserve targets 2 percent inflation over the longer run, measured by PCE, not a return to earlier price levels. Not confirmed as a BLS all-items figure: the claim, in First Citizens material, that prices would have been about 14 percent higher than in 2020 under steady 2 percent. It is recomputed below; it is not a BLS figure. Lower inflation does not mean lower prices explained starts with the distinction the headline hides: inflation is a speed, prices are a position.
Think of a car. Inflation is the speedometer. The price level is the odometer. If the car slows from 70 to 40 miles an hour, it still moves forward and the odometer still rises, more slowly. For it to run backwards, the car must reverse. In price terms, reverse is deflation: the index itself falling. In September 2026 prices were rising 3.7 percent a year, and even a slowdown to 2 percent would still mean rises from an already-elevated level.
The speedometer and the odometer: rate versus level
The Consumer Price Index measures, in BLS’s words, ‘the change in prices paid by consumers for goods and services.’ Each release reports a percent change — the rate — and rests on an index level tracking cumulative change since 1982–84, when the index equalled 100. September 2026 reported a 12-month change of 3.7 percent and a level of 336.940. The 3.7 percent is the speed; the 336.940 is the position. Shorter comparisons follow below, computed by dividing the later index by the earlier one.
A falling inflation rate with a rising level is the normal pattern of disinflation. BLS annual-average CPI inflation was 8.0 percent in 2022, 4.1 percent in 2023, 2.9 percent in 2024 and 2.6 percent in 2025, while the level rose every year. Each year of ‘falling inflation’ was a year prices rose. Households expecting a lower rate to restore 2021 shelf prices were expecting deflation. No major central bank targets that outcome; what they target is a slower, steadier rise.
Even a single month in which the index dips does not change the level story. In June 2026, the seasonally adjusted all-items CPI fell 0.4 percent on the month in BLS Table A, yet the 12-month rate for June was still positive, because the level remained well above June 2025. Monthly wiggles — often driven by energy — do not undo years of cumulative increase. The level is a stock accumulated over years. The monthly rate is a flow in one month.
Lower inflation does not mean lower prices explained: the price-level maths from BLS
All index values are CPI-U, U.S. city average, not seasonally adjusted, 1982–84 = 100, from BLS. The September 2026 level, 336.940, comes from the September 2026 CPI news release. The earlier levels come from the BLS historical series: December 2019, 256.974; the 2020 annual average, 258.811; December 2020, 260.474. The formula is: percent change = (later index ÷ earlier index − 1) × 100. A $100 basket is the same calculation with $100 as the starting amount.
From December 2019 to September 2026: 336.940 ÷ 256.974 = 1.3112, a rise of 31.1 percent; a $100 basket cost about $131.12. From the 2020 annual average: 336.940 ÷ 258.811 = 1.3019, a rise of 30.2 percent ($130.19). From December 2020: 336.940 ÷ 260.474 = 1.2936, a rise of 29.4 percent. The answers differ because the starting points differ. December 2019, the last pre-pandemic December, is headlined below.
Now the steady-2-percent comparison. Compounding 2 percent for six years gives 1.02⁶ = 1.1262, a rise of 12.6 percent ($112.62). Over six and three-quarter years from December 2019, 1.02^6.75 = 1.1430, a rise of 14.3 percent ($114.30). The actual basket, $131.12, is $16.82 above that counterfactual — the accumulated excess per $100 of December-2019 spending.
Where does ’14 percent’ fit? First Citizens Wealth, in its note on the September 2026 FOMC meeting, wrote that under steady 2 percent ‘prices would have been 14% higher today compared to 2020,’ with inflation above the Fed’s 2 percent target for five straight years. Its September update described about 14 percent over six years for a necessities basket — a First Citizens framing, not a BLS all-items calculation. Recomputed from BLS all-items values, 14 percent matches 6.75-year compounding from December 2019 (14.3 percent), not flat six-year compounding (12.6 percent), and not the actual rise from any 2020 starting point (29–30 percent). The difference is basket, base date and rounding. This article’s headline result is the BLS recomputation: about 31 percent from December 2019, against about 14 percent under steady 2 percent over the same period. Readers seeing ’14 percent’ elsewhere should ask: of what basket, from what date, over how many years?
A worked example: the $100 basket
Take a basket costing exactly $100 in December 2019 and assume, for illustration only, that it moved exactly with the all-items CPI-U. In September 2026 it costs $131.12. That is the level. Now apply different futures over the next year, holding the basket fixed — itself a simplification, since households substitute and the CPI basket is updated over time.
If inflation is 2 percent — the Fed’s longer-run goal — the basket costs $131.12 × 1.02 = $133.74: $2.62 more, and nothing got cheaper. At 3.7 percent, the September 2026 rate, it costs $135.97. At zero, it stays at $131.12. Only deflation brings it below $131.12: at minus 2 percent, $128.50.
| Next 12 months | Basket cost | Change from $131.12 |
|---|---|---|
| Inflation at 2% (Fed goal) | $133.74 | +$2.62 |
| Inflation at 3.7% (Sept 2026 CPI rate) | $135.97 | +$4.85 |
| Inflation at 0% | $131.12 | $0.00 |
| Deflation at −2% | $128.50 | −$2.62 |
The example shows why ‘inflation fell’ headlines can feel false at the till. If the rate falls from 3.7 percent to 2 percent, the correct headline is ‘price rises slowed.’ The till still shows $133.74, not $100 or $114.30. The shopper remembering $100 compares the level with a memory; the statistician reporting 2 percent compares the speed with a target. Policy is set on the speed. Life is lived at the level.
Disinflation versus deflation — and why the Fed does not target falling prices
Disinflation is a falling inflation rate with prices still rising. Deflation is a falling price level. In disinflation, fixed debt generally becomes easier to bear in real terms; in sustained deflation, its burden rises and spending may be delayed, deepening a downturn. Those are textbook mechanisms, stated generally. The Fed targets 2 percent PCE inflation to keep a buffer above zero.
The October 2026 FOMC statement illustrates the framework. The Committee held the federal funds target range at 3.75 to 4 percent on 28 October, by a 10–2 vote, and said ‘Inflation remains elevated’ and spoke of ‘a timely return of inflation to the Committee’s 2 percent goal.’ Note: a return of the rate to 2 percent, not of prices to 2020 levels. For the full account, see the Federal Reserve October 2026 rate decision explainer. Individual prices do fall — used cars 1.5 percent, dairy 0.7 percent over the year — but a 30 percent all-items fall would require deflation the Fed seeks to avoid.
The level still matters to policy. Each year above 2 percent adds permanently to it, unless a later year of deflation removes it. A 2 percent rate for a decade adds about 22 percent (1.02¹⁰ = 1.219); 3.7 percent for a decade would add about 44 percent. Hitting the target never makes anything cheaper than today — but the target is the difference, compounded, between two very different costs of living a decade hence.
Why your basket is not the CPI basket
No household buys the CPI. The CPI-U is a national average for urban consumers, weighted by aggregate spending. In September 2026 the components diverged sharply: energy rose 20.4 percent over the year, gasoline 34.2 percent and fuel oil 74.2 percent, while food rose 2.5 percent, shelter 3.2 percent and commodities excluding food and energy only 0.7 percent. Diesel, which moves freight, set a record in the EIA weekly series in September; see the diesel prices at a record high explainer. A personal inflation rate is not published by BLS.
Weights also change over time. As spending patterns shift, BLS updates basket weights, and substitution — buying chicken when beef rises — is partly captured in the chained CPI and weight updates. The headline CPI-U is not a cost-of-living index for a fixed standard of living in the strictest sense. Treat $131.12 as what the all-items arithmetic implies, not as a claim that every household’s spending rose exactly 31.1 percent.
In summary: the level rose about 31 percent from December 2019 to September 2026; core inflation was 2.6 percent, PCE inflation 3.6 percent. Lower inflation, if it comes, will mean the level rises more slowly — not that it falls. Nor can a lower mortgage payment be read off a lower inflation print, as the Fed decision and mortgage rates explainer sets out.
Frequently asked questions
If inflation drops to 2%, will prices go down?
No. A 2% inflation rate means prices rise 2% over the year from their current level. On the worked example, a $131.12 basket becomes $133.74. Prices fall only under deflation, when the inflation rate is negative.
What is the difference between disinflation and deflation?
Disinflation is a falling inflation rate while prices still rise — for example, from 3.7% to 2%. Deflation is a falling price level — the index itself going down. The Fed targets 2% inflation (disinflation toward the goal), not deflation.
How much have prices risen since 2020?
On BLS all-items CPI-U values, September 2026 (336.940) was about 30.2% above the 2020 annual average (258.811) and about 31.1% above December 2019 (256.974). A $100 December-2019 basket cost about $131.12. The ‘about 14%’ figure is a First Citizens framing for a necessities basket under steady 2% compounding, not the BLS actual rise.
Does lower inflation help mortgage rates?
Only indirectly. The Fed does not set mortgage rates, which reflect longer-term yields and expectations. Slower inflation, sustained, can ease those expectations, but a lower print does not automatically cut any individual’s mortgage rate.
Updated: 1 November 2026. CPI values are September 2026; the October CPI is due 10 November 2026. Basket figures are arithmetic from the BLS values cited.
SOURCES
- U.S. Bureau of Labor Statistics, “Consumer Price Index — September 2026,” USDL-26-1574, 14 October 2026 — https://www.bls.gov/news.release/cpi.htm
- U.S. Bureau of Labor Statistics, Consumer Price Index for All Urban Consumers (CPI-U), U.S. city average, not seasonally adjusted, series CUUR0000SA0 — https://data.bls.gov/timeseries/CUUR0000SA0
- U.S. Bureau of Labor Statistics, “Consumer Price Index Historical Tables for U.S. City Average” — https://www.bls.gov/regions/mid-atlantic/data/consumerpriceindexhistorical_us_table.htm
- Board of Governors of the Federal Reserve System, “Federal Reserve issues FOMC statement,” 28 October 2026 — https://www.federalreserve.gov/newsevents/pressreleases/monetary20261028a.htm
- Federal Open Market Committee, “2025 Statement on Longer-Run Goals and Monetary Policy Strategy,” adopted effective 24 January 2012, as amended effective 22 August 2025 — https://www.federalreserve.gov/monetarypolicy/monetary-policy-strategy-tools-and-communications-statement-on-longer-run-goals-monetary-policy-strategy-2025.htm
- U.S. Bureau of Economic Analysis, “Personal Consumption Expenditures Price Index” (current release 29 October 2026) — https://www.bea.gov/data/personal-consumption-expenditures-price-index
- First Citizens Wealth, “Fed commits to price stability,” FOMC meeting, 16 September 2026 — https://www.firstcitizens.com/content/dam/firstcitizens/pdfs/wealth/market-outlook/2026/september-fomc-meeting.pdf
- First Citizens Bank, “Making Sense: September Market Update,” recorded 30 September 2026 — https://www.firstcitizens.com/wealth/market-outlook/2026/making-sense-september-market-update
