From Jay Hancock's blog:
The January jobs report was a mixed bag. Unemployment plunged for the second month in a row, but the increase in payroll employment of 36,000 jobs was far less than the expected ~150,000. Nevertheless it's an encouraging report, especially in the context of other recently released economic indicators.
Two straight months of unemployment declines of 0.4 percent is big news and looks a lot like the Reagan economic recovery of 1983 -- read the NYT's Floyd Norris's column on this from yesterday. Normally people discount the unemployment report because it's based on an uncertain telephone survey, can jump around and depends on the technical definition of unemployment, which means actively looking for a job.
But perhaps the unemployment survey is telling us more about the recovery than the payroll survey. The payroll survey has a habit of missing new jobs in a recovery. The Labor Department can't survey employers that it doesn't know exist, and in recoveries new companies are formed. They hire people totally off the Labor Department' radar screen, at least initially.
In past recoveries the revised numbers have shown much stronger job growth than was initially reported. Already the Labor Department has been pretty consistently revising payroll employment upward in the months after the initial report comes out. It happened again in this morning's report for December and November. My bet is that the January payroll numbers will be revised upward later.
There is other good news out there. Consumer spending was decent in the fourth quarter. Industry surveys are showing positive results. Consumer confidence is up. The most recent unemployment claims were down sharply. Those data suggest that a 0.8 percent decline in unemployment over two months shows the start of a jobs recovery.