The U.S. Department of Labor reported this morning that the national unemployment rate fell from an estimated 7.4 percent in July to 7.3 percent in August, on a seasonally adjusted basis. At first blush, a falling unemployment rate looks like good news, but upon closer examination, the supporting survey data suggest that the labor market actually weakened in August.
The unemployment rate is calculated from a large, national, monthly survey of approximately 60,000 households. Among other things, the survey asks whether adult members of the households were participating in the labor force (meaning they were either working or available and looking for work) and whether they were employed. The unemployment rate is calculated as the ratio of people who are in the labor force but not working (the numerator) and all people who are in the labor force (the denominator).
By these definitions, the August survey indicates that nearly 200,000 fewer Americans are unemployed, which should be very good news--except that all of this decline (and then some) came from a reduction in the labor force. Overall, the government estimated that the number of Americans in the labor force fell by just over 300,000, and the number of Americans who reported that they were working fell by just over 100,000.
The percentage of non-institutionalized, adult civilians who were in the labor force fell from 63.4 percent in July to 63.2 percent in August, the lowest that figure has been since 1978. The percentage of non-institutionalized, adult, civilians who were working fell from 58.7 percent in July to 58.6 percent in August. That number is little changed since the depths of the Great Recession (the rate bottomed out at 58.2 percent in 2010).
Some of the change seems likely to be sampling error. The Department of Labor cautions that monthly changes in the survey estimates that are smaller than 400,000 aren't statistically meaningful. Thus, we can't rule out the possibility that the numbers of people in the labor force and who are employed actually grew. Indeed, preliminary estimates from the establishment survey, which counts jobs at firms rather than counts of people who report employment, indicate that the number of non-farm jobs grew by 169,000 on a seasonally adjusted basis (plus or minus 100,000 given the sampling error).
Nevertheless, we can conclude with certainty that jobs aren't being added quickly enough to make meaningful improvements in the economic lives of Americans. That's disappointing and sadly, not news.