Seasonal Adjustment

Many economic series follow a reliable rhythm tied to the time of year. Retail sales spike every December. Construction slows in winter. Unemployment claims jump in January when holiday hiring ends. Seasonal adjustment strips those expected swings out of the raw number, leaving a figure that economists can compare month-to-month without the calendar noise getting in the way.

The adjusted figure is sometimes labeled SA or SAAR (Seasonally Adjusted Annual Rate — meaning the monthly or quarterly number is also scaled up as if it ran for a full year). When you see jobs data reported as "the economy added X thousand positions," that figure is almost always seasonally adjusted. The raw, unadjusted print is usually available separately but rarely headlines.

Suppose retail sales normally rise 15% every December. If this December they rise only 10%, the seasonally adjusted figure would actually show a decline relative to trend — even though the raw number went up. That is why adjusted and unadjusted prints can move in opposite directions.

Seasonal adjustment factors are recalculated periodically, which means past data can be revised when the model is updated — connecting this concept directly to data revisions. Misreading an unadjusted series as if it were adjusted is one of the most common errors when interpreting economic indicators. See also how to read percentage moves for more on interpreting these figures correctly.