The August Jobs Report

OK, deep breath.

The job market was weaker than expected in August, as employers added no new jobs, on net (this 'net' business is important -- some jobs were created last month, some lost -- this month, creation equaled loss, so net is zero).

There were some extraneous factors, but they don't change the dismal result: the now-ended Verizon strike took 45,000 jobs out of the private sector count in August, but going the other way, 22,000 Minnesota state employees previously shut out, went back to work.

Since monthly numbers can always jump around, it's important to average over a few months to get a feel for the trend. The figure below does that for payroll jobs.

Source: BLS

You can clearly see the deceleration in job growth. From Dec 2010-May 2011, we were adding around 150,000 jobs per month, not enough job growth to push down the unemployment rate, but about enough to hold steady.* But over the past three months, we're crawling along the bottom with 35,000/month. (Add in the factors noted above -- strike, end of shutdown -- and that number goes to 43,000... no real difference.)

Other points of note:

  • Weekly hours fell slightly, another sign of weak labor demand;

  • Nominal hourly wages--before inflation--were down three cents;
  • So, with fewer hours at lower hourly wages, weekly earnings were down about3.30;
  • Over the past year, both hourly and weekly earnings are up 1.9%, considerably slower than inflation, up 3.6% (that's July2010/July1011 on inflation, since we don't have August price data yet);
  • Budget cuts led to more state and local employment layoffs again, as has been the case for 12 of the last 13 months (down 400K over that period).
  • In sum, very weak labor demand, no matter how you cut it.

    *There's a sleight of hand here--the job count that determines the unemployment rate is from a different survey, but the much larger payroll survey is more accurate, month to month.

    And there's a first dose of my colleague Chad Stone's analysis here.