Government continues to be a drag - government employment dropped by 17,000, and this would have been worse without 22,000 workers returning to work after the Minnesota shutdown. The Verizon strike also had an effect, reducing payrolls by about 45,000. That is, without the Verizon strike, there would have been a gain in private-sector payrolls of 62,000, which isn't particularly good. Also, the employment growth for June and July were revised downward.
The payroll employment number is calculated from a survey of employers. The unemployment rate comes from a survey of households. Overall, the household survey looks a little better (though it is considered less reliable because it has a smaller sample). The unemployment rate held steady at 9.1% and the number of people employed rose by 331,000. The labor force participation rate and employment-population ratio also ticked up slightly.
Overall, this report should help convince the Federal Open Market Committee to take more steps in the direction of "easing" policy, and add urgency to further fiscal policy action to try to stimulate job growth.
On a non-seasonally adjusted basis, payrolls increased by 118,000 (i.e., August is a month that normally has a slight gain, which is removed by the seasonal adjustment). Non-seasonally adjusted unemployment was also 9.1% (but down from 9.3% in July).