Lies, damned lies, and statistics

Mish points out the obvious:

To be sure, the drop in the unemployment rate [from 9.5 to 9.4 percent] was a surprise, but it was all due to the slide in the labour force — the employment-to-population ratio gives a more accurate picture of the slack in the labour market and the hidden secret in today’s report was that this metric slid to a 25-year low of 59.4% from 59.5% in June and 61.0% at the turn of the year.

Got that? Translated into non-econospeak, that means the decline in the unemployment rate isn’t because more people are working, but because more people have given up looking for work. So, they’re no longer counted as part of the labor force and therefore there are both fewer workers and less unemployment. I’ve long felt that employment-to-population ratio is a much better statistic to measure jobs because it removes one subjective variable from the equation. Calculated javascript:void(0)Risk has a nice chart up which demonstrates both the significance of women entering the work force – although not the full extent of it since it would have to go back to 19731950 for that – as well as the scale of rising not-employment.

Even these numbers appear to considerably overstate the case. The “civilian noninstitutional population” is reported by the BLS at 235.87 million, considerably less than the U.S. Census-reported population estimate of 306.29 million. I think it’s reasonable to consider soldiers employed; they’re arguably more productive than most government employees, so adding the 1,455k members of the military to the 140,041k civilian employed means that 46.2 percent of the population is employed, not 59.4 percent as reported in the July 2009 employment report.