Last year, did it seem like the Colorado unemployment rate didn’t match what was happening on the streets?
For the most part, the initial data, which is used in the monthly CDLE media blitz, told a much different story from the unheralded benchmark revisions released in March. In fact, the correlation coefficient between the initial release data and the March benchmark data is .44. This means there is a relationship between the two data sets, but it is not strong.
The initial data indicated that unemployment rate was flat at 7.8% for the first quarter, had four months of increases up to 8.3%, then five months of declines.
On the other hand, the benchmarked revisions started at 8.3% and dropped in February to 8.2%. The rate stayed at that level until June, when it declined for the last 6 months of the year. The benchmarked data ended the year at 7.5% compared to 7.6% for the initial data.
Good data is most valuable during times when the economy is the most volatile. The unemployment rate published by the Colorado Department of Labor failed to meet that critical need in 2012.
©Copyright 2011 by CBER.