A grocery bill in January rarely looks like one in June. Some prices follow the calendar. Seasonal adjustment is the statistical step meant to account for that in the monthly Consumer Price Index (CPI), the main federal gauge of what households pay.
What seasonal adjustment is meant to do
Seasonal adjustment is a statistical method. In general, it aims to strip out price moves that repeat at about the same time each year. What remains should make a month-over-month change easier to read. If a price rises every spring, a spring rise says little about whether this year is unusual.
Here is a made-up example, with invented numbers. Say an unadjusted index for some seasonal item rises 1.0 percent in April. Say that in a typical year the April rise is about 0.8 percent. An adjusted series would show a small gain, near 0.2 percent. The raw price still went up 1.0 percent. The adjusted figure just says most of that was the usual spring pattern. The real calculation is more involved and works from estimated seasonal factors, so treat this as an illustration of the logic only.
Where the BLS material sits
BLS publishes a set of pages on this topic. The agency's Handbook of Methods appendix on seasonal adjustment is titled as covering the methodology. Separate pages on intervention analysis in seasonal adjustment exist in a 2023 version, a 2025 version, and a 2026 version. The existence of three editions suggests the method gets revisited over time.
Other pages cover CPI calculation, the CPI news release technical note, and CPI notices. The monthly CPI news release PDF is where the figures themselves appear. Those pages are where any specific mechanic should be confirmed before you quote it.
Adjusted versus unadjusted readings
A seasonally adjusted series tries to remove calendar effects. An unadjusted series tracks the raw price level. They answer different questions. Month-to-month comparisons are where calendar patterns tend to matter most, which is why adjusted figures are commonly used for them. Over a full year, the same calendar month is compared with itself, so the seasonal pattern largely cancels without any adjustment.
The choice matters in money terms. Many contracts and benefit formulas point to a specific index, and that index is often an unadjusted one, because a payment tied to a figure that gets revised would be awkward. Cost-of-living adjustments to benefits and escalator clauses in leases are common examples. Which series applies depends on the program rules or the contract wording. Read the exact index named in the document, not the headline number from the news.
When you open the monthly release, look for the tables that label each series as seasonally adjusted or not seasonally adjusted. Check the label before comparing any two numbers. Mixing the two is an easy way to misread a change.
Intervention analysis and revisions
Intervention analysis, in general terms, is a technique for handling unusual one-time events that could distort a seasonal pattern. A sudden disruption can look like a seasonal swing to a model that has never seen it. Flagging the event lets the model treat it separately instead of folding it into the yearly pattern. The BLS pages carry the phrase in their titles, and the 2023, 2025 and 2026 pages are the place to read how the agency applies it.
Adjustment also relies on statistical models, and models are typically re-estimated as new data arrive. That can mean adjusted figures are revised after first publication. The CPI notices page and the technical note are the places to look for the current schedule and any announced changes.
How to check a monthly number yourself
- Open the latest CPI news release and find the table labeled seasonally adjusted.
- Find the matching unadjusted table and note the difference in labels.
- Read the technical note for how BLS describes the adjustment.
- Check the notices page for any announced changes to methods or factors.
- If a contract or benefit cites CPI, find the exact index and series it names.
That short routine keeps the focus on the primary documents. It also guards against treating one headline figure as final. This article is educational only. For your own situation, rely on official sources or a qualified professional.
