September Employment Report

According to the BLS, in September the economy lost 95,000 jobs, and the unemployment rate held steady at 9.6%.  Neither of those headline numbers is pleasant, but the first is more discouraging than the second - I suppose the choice of which one to emphasize depends on whether one sees the glass 10% full or 90% empty.

In this case, the reason for the discrepancy is the fact that the jobs number comes from a survey of businesses while the unemployment rate is calculated from a separate survey of households.  (Another potential explanation would be a decline in labor-force participation, but in September it remained unchanged.)  According to the household survey, the number of people employed actually increased by 141,000 (the payroll number from the establishment survey is usually preferred because it has a larger sample).

The decline in payrolls of 95,000 is the sum of an increase in 64,000 in private-sector jobs and a loss of 159,000 government jobs.  Federal government employment fell by 76,000 (77,000 temporary census jobs ended), state government employment fell by 7,000 and local government employment fell by 76,000 (of which 49,800 were in education).

So, ironically enough, the headlines from the last BLS report before the election will probably give a little more momentum to those who think "government spending" is the problem, even though the report shows that cuts in government spending are already acting as a drag on the economy.

One potential bit of good news from the bad news is that the gloomy report provides another piece of evidence to support the case within the Fed for more aggressive quantitative easing ("asset purchases").

The headline numbers are all seasonally adjusted to remove regular fluctuations that occur during the year.  On a non-seasonally adjusted basis, the unemployment rate fell from 9.5% to 9.2% (i.e., the seasonal adjustment factor raises the unemployment rate in September).  Also, non-seasonally adjusted payrolls rose 428,000 - including an increase of 865,800 in the "local government-education" category, so I think the correct interpretation of the government job loss is that fewer teachers and education workers were hired this year than usual.  Also, non-seasonally adjusted private payrolls fell by 412,000, so the seasonally-adjusted gain of 64,000 means that the private sector did not shed as many jobs as usual for this time of year.

See also: Economix, Free Exchange, Calculated Risk and RTE's round up of reactions.