Seasonal adjustment removes the effects of recurring seasonal influences from economic series.
Seasonally adjusted data are referred to in a number of ways: adjusted, seasonally adjusted, and SA. Likewise, data that have not been seasonally adjusted are often referred to as not seasonally adjusted, unadjusted, or NSA.
Why are data seasonally adjusted?
By removing the seasonal component, we can make useful comparisons between observations. Seasonal adjustment also tends to smooth a data series out, allowing data users to see changes in trends more readily.
Are seasonally adjusted data better than data that are not seasonally adjusted?
No. Depending on the type of information that you’re looking for, one may fit the bill better than the other. For example, if you’re interested in analyzing the seasonal hiring in department stores, you will likely want to look at data that are not seasonally adjusted if you wish to compare October-to-December employment changes over recent years. By contrast, if you’re interested in understanding the most recent monthly fluctuations in the CPI, you would want to use seasonally adjusted data.
How seasonally adjusted data are calculated
There are several programs and variations that can be used to construct a seasonally adjusted series. A common program used at BLS, is X-12-ARIMA, developed by the U.S. Census Bureau.
The following links discuss seasonal adjustment methods, or ways to use seasonally adjusted data and data that are not seasonally adjusted: