edthecynic
Censored for Cynicism
- Oct 20, 2008
- 43,044
- 6,883
- 1,830
The employment-to-working age population ratio is easily the most worthless stat for unemployment that the BLS puts out. It assumes that students over 16, stay at home spouses, the retired and the disabled want jobs.I used to teach labor economics and try to keep up with the literature. There is no single measure of employment that is best for all purposes. For most purposes (especially if you are using an employment variable in a regression equation), you get most of the explanatory power using a combination of U-6 and the employment-to-working age population ratio, both reported monthly by BLS. If you want to do better than this, you need to measure the "quality" of employment. The most common variables used are the length of the work week and median weekly earnings. More recently, adding a measure of part-time employment helps. This is all econometrics, what is statistically related to what.
If you are interested in other specific questions, other measures can be important. The dispersion index of employment changes is a good measure of how widespread throughout the economy changes are. Unemployment rates for youth may be better measures if your examining crime statistics. Ethnic breakdowns can help in explaining poverty and regional variations.
So it really boils down to why you want to measure employment. I suspect that some politically motivated "analysts" misuse statistics and are inconsistent, but the BLS and Census are pretty professional and all of the data is pretty transparent. Researchers can and do have access to the raw data and can develop statistics outside the reported series and create new measures for specific research projects. So I don't have much sympathy for critics who trash the agencies but have no alternative measures that can be subject to peer review.