First reaction to employment numbers

My first reaction to the BLS employment report is, I really don't know what's going on. I know that the economy is adding jobs. I suspect that employment growth is trending upward: that's what the GDP data, the ISM manufacturing and non-manufacturing data, the data on initial claims for unemployment compensation, the ADP employment survey, etc. are all saying. But the BLS numbers are telling a different story. The highlights for January:

- Payroll employment rose by 36,000 (+50,000 private sector, -14,000 government)
- Unemployment rate fell from 9.4% to 9.0%
- Number employed according to the household survey rose by 117,000
- Number unemployed fell by 622,000
- Meaning that the labor force fell by 505,000
- Yet the employment-population ratio rose from 58.3 to 58.4 percent.

Interpretation of these numbers is complicated by two factors. First, bad weather in January may have screwed up the numbers. Evidence for this is that the big job-losing sectors were construction and transportation. Second, the BLS did its "rebenchmarking" of the data which caused revisions for all payroll employment data from April 2009 to the present.

How did the employment-population ratio rise while employment was basically flat? Well, the BLS reports that the civilian noninstitutionalized population fell by 185,000. The culprit: women, who declined by 282,000. Now I did not notice 282,000 women dropping dead or being incarcerated in January. Nor do I believe that that many women retired of a sudden. So how did this happen? I don't know, but I do know that a drop in the noninstitutionalized population occurs every January (take a look at the data here). And since the labor force numbers are scaled to the census population estimates, this phenomenon ripples through the household data, or so I would think. But I've never seen any kind of explanation for this.

Calculated Risk usually does a good analysis of the data, so I'd check there later in the day.