January 2010 employment data

The official unemployment rate declined in January. It's still too early to tell for sure, but so far it looks like we're seeing the declining unemployment that occurs after a typical recession, rather than the "jobless recovery" of the previous two recessions. (The typical pattern after a recession is for the unemployment rate to decline. That pattern was broken by the previous two recessions when the unemployment rate rose for quite a while after the recession had ended.)


Payroll numbers suggest we lost 20,000 jobs in January. That's far better than the pace a year ago, but we need 100,000-200,000 job gains just to keep up with population growth. This graph shows the month-over-month change in payrolls as measured by the U.S. Bureau of Labor Statistics:


Here is the month-over-month change in payrolls as measured by Automatic Data Processing, a private payroll-processing company. I'm starting to favor the ADP data over the BLS data due to lower month-to-month volatility. As a reminder for conspiracy theorists, the ADP data does not come from the government.


Some economists consider aggregate weekly hours worked as the best measure of both unemployment and underemployment. For those of you who favor the U6 measure of the unemployment rate over the official rate because it takes underemployment into account, you should love the Aggregate Weekly Hours Index. Here we see that the pace of average weekly hours worked is showing continuing improvement, but it's still below zero:


Finally, some economists like to look at weekly initial unemployment claims because it is updated more frequently than the monthly statistics above. This graph shows the year-over-year percentage change. Notice it's currently below zero, which is a good sign.


For those of you on the east coast, enjoy the snow this weekend.