Job Gains in Texas and Losses in Caifornia and Florida

A Tale of Two Surveys

by Elaine Schwartz    •    Oct 6, 2012    •    245 Views

Everyone was surprised that the US unemployment rate for September dipped to 7.8%. Interpreting the number as evidence of an accelerated recovery, some people were delighted. Others advised caution, saying the statistical source of the unemployment rate is not always dependable.

To complicate matters further, the monthly employment report from the Bureau of Labor Statistics also includes job creation numbers that, at 114,000 new jobs for September, were below the number that would keep pace with population growth.

So, one number great and the other not. Why?

The numbers differ because the unemployment rate and job creation numbers are based on 2 entirely separate sources. Here are some of the differences:

1) The unemployment rate is based on the household survey (technically known as CPS, Current Population Survey), a canvas of just 60,000 households. The sample, based on home visits by Census Bureau workers, comes from the entire civilian non-institutional population aged 16 and over and even includes unpaid household workers.

2) The job creation numbers come from the establishments survey (technically known as Current Employment Statistics survey), collected from approximately 486,000 business and government establishments. Their sample, collected from submitted data, looks at nonfarm wage and salary jobs.

You can look here for more about how the 2 surveys differ.

Sources and Resources: The BLS link is the best for firsthand survey information and Fox has a good interview. If you watch the video, please note they got the labels backwards on their graph but the content is good. Also, NPR Planet Money has a good summary of the two surveys.

« »