Email, response to PolitiFact Texas, Cheryl Abbot, regional economist, U.S. Bureau of Labor Statistics, Feb. 20, 2013
We don’t produce seasonally adjusted (SA) unemployment rates at the county level, so you are correct that Travis County unemployment rates are not seasonally adjusted (NSA). I don’t recommend that you compare SA and NSA series. Seasonally adjusted data allow analysts to more easily see underlying economic changes by ‘smoothing’ the data through a review of long-term historical movements. For example, as summer approaches, large numbers of teenagers tend to join the labor force and initially drive up the unemployment rate as they search for jobs. Our seasonal adjustment methodology looks at the historical norm for this type of movement and discounts it from the seasonally adjusted series. So, using a simplified example, if the NSA unemployment rate typically climbs 0.5 percent points between May and June as teenagers enter the labor force, and this year it also rose 0.5 percentage points, then the SA data might show no change at all in the unemployment rate. Again, this is to allow better observation of underlying labor force movements. Other periods that often show seasonal volatility include winter/early spring as construction employment typically begins expanding, as well as stronger retail hiring around Thanksgiving and Christmas.
The attached worksheet contains not seasonally adjusted unemployment rates from 2009 to the current for the U.S., Texas, and Travis County..... As far as the numbers you mention below for December, they are correct, but if you look at NSA data only, the U.S. was 7.6%, Texas was 6.0%, and Travis remains at 4.9%.