At the end of every calendar year, BLS makes routine revisions to the seasonally adjusted data in the Current Population Survey (CPS), which includes the national unemployment rate. The revisions include all additional information to identify if changes in the data are one-time or seasonal temporary blips as opposed to permanent changes in the numbers. These changes are usually minor and don’t drastically change labor market trends. With this month’s release of the employment situation, BLS made that same set of routine revisions, and as expected, the changes to 2011 data were minor.
For 2011, the national unemployment rate was changed for 8 months, but these changes were never more than one-tenth of a percentage point in any month. For example, the unemployment rate for October was revised down to 9.0% from 9.1%. Also, for the 8 months of changes, there were just as many increases as there were decreases to the rates. Similarly, in the past four years, changes have been equally divided between upward and downward revisions in the unemployment rate. On net, the story is the same: since the middle of the third quarter of 2011, the rate of unemployment has been dropping every month.
These annual revisions are taken to improve the survey’s ability to adjust for seasonal trends in the labor market data, making it easier to observe the true effect of fundamental changes in the economy. The size of the revisions depends on year-to-year changes in the timing of calendar-related events that affect employment data, such as the start of the school year or annual weather patterns. Since the timing of those annual events usually doesn’t change much, the revisions tend to be small. BLS offers a more detailed explanation on their website: http://www.bls.gov/cps/cpsrs2012.pdf. Below is a table showing the revisions made to the 2011 national unemployment rates.
Seasonally adjusted unemployment rates in 2011 and changes due to revision, January – November 2011