Seasonal adjustment
Also called seasonally adjusted, SA
A statistical adjustment that removes the influence of predictable patterns within a year, so month-to-month or quarter-to-quarter changes reflect underlying movements. The BLS uses the X-13ARIMA-SEATS method for the CPI.

- Level
- Intermediate
- Reading
- 1 min
- Sources
- 2Statistics
- Updated
On this page· 5 sections
A raw monthly change often reports the calendar more than the economy. Seasonal adjustment removes the influence of predictable patterns over a year, so that periods, including months and quarters, can be compared more consistently[1].
How it works
Data that have not gone through the process are labeled “not seasonally adjusted”[1].
The Consumer Price Index shows both versions. The BLS publishes unadjusted data and seasonally adjusted data, which are computed using seasonal factors derived with the X-13ARIMA-SEATS method[2].
The factors are recalculated once a year, with the release of the January CPI in February, and reflect price movements of the previous calendar year. The new factors revise the previous five years of seasonally adjusted data, and older indexes are considered final[2].
Example
These numbers are made up. Say a store’s sales index is 100 in November and 132 in December, and the usual seasonal factor is 1.00 for November and 1.20 for December.
- Unadjusted change: 132 ÷ 100 = 1.32, or +32%.
- Adjusted December: 132 ÷ 1.20 = 110.
- Adjusted November: 100 ÷ 1.00 = 100.
- Adjusted change: 110 ÷ 100 = 1.10, or +10%.
After adjustment, 22 of the 32 points of the December rise are explained by the usual December pattern. The remaining 10% is the underlying movement.
Watch out
The adjusted history can change. When the factors are recalculated, years of already published adjusted data are revised[2].
Adjusted and unadjusted numbers are different series. Putting one beside the other compares figures treated in different ways, so check which one a headline uses[1][2].
Seasonal adjustment is an estimate, not an observation. It rests on a statistical method applied to past patterns[2].
Sources
- StatisticsU.S. Bureau of Labor StatisticsGlossaryChecked November 9, 2026
- StatisticsU.S. Bureau of Labor StatisticsSeasonal Adjustment in the CPIChecked November 9, 2026
