September 3, 2010
Private Payrolls Gains Pushes Against Double-Dip
Nonfarm payrolls fell 54,000 in August, about half of what was expected. The government let go of 121,000 employees, 114,000 of which were temporary census workers (close to our estimate of -100,000). The remainder were state employees, reflecting budget difficulties. Private payrolls grew by 67,000, almost exactly matching our projection of 70,000. Additionally, the prior two months were revised up by an impressive 123,000. The unemployment rate ticked up to 9.6% from 9.5%, equaling expectations. The workweek was unchanged at 34.2 hours. Although the report isn’t as bad as many had feared, private job creation averaging 78,000 a month for the past three months doesn’t move the needle on the unemployment rate, doesn’t make the public or policymakers happy, and keeps the Fed accomodative for "an extended period.
Private service-providing industries added 67,000 workers, while the goods-producing sector was flat. Health
care continued to add jobs, rising 28,000. Professional and business services rose by 20,000, nearly all of which were
temporary workers. Temporary help services has grown 22% over the past 12 months, the most since November
1984!, evidence of just-in-time labor and a cautious attitude on the part of businesses to add permanent
employees. Employment in the financial sector fell to its lowest level since 1998 -- a concrete sign of the effects
of deleveraging. Manufacturing fell by 27,000, 22,000 of which was due to vehicles, which experienced difficulties with
seasonal adjustments over the past two months. Construction surprisingly added 19,000 jobs. The diffusion index
slipped 3.7 points to 53.0, indicating that job creation remains fairly narrow. Private payrolls are 0.3% above a year
ago, and are performing almost identically to the 1991 recovery.
The average workweek for production and nonsupervisory workers increased to 33.5 from a downwardly revised 33.4,
implying little overall change. Similarly, aggregate hours worked was unchanged (but was probably slightly positive
before rounding). Although average hourly earnings rose 0.3%, the y/y change fell from 1.8% to a new low of 1.7%,
indicating falling wage pressures. As a result, aggregate payrolls increased by 0.3%, and have risen 3.2% over the
past year, the fastest pace since March 2008. Once again, we didn’t have any issues with seasonal adjustments (if
anything they are biasing payroll counts lower) or the birth/death model.
The labor force swelled by 550,000 (mostly younger workers), pushing the labor force participation rate up to 64.7%
from 64.6%. Nevertheless, labor force growth is below its level of a year ago, and well below where it should be at this
point in the economic recovery. The number of employed rose by 290,000 (effectively all part-time wage and salary
workers) and the number of unemployed increased by 261,000. Household employment continues to lag badly the
jobless recoveries of 1991 and 2001. The number of unemployed who were laid off temporarily rose by 212,000,
reflecting cost control and the uncertain business outlook. The number of self-employed fell to 8.5 million, the
fewest since at least 2000, and a departure from the behavior seen in the previous recovery.
Signs of structural unemployment were mixed. The number of those unemployed for more than six months fell by
323,000, bringing their share down to 42% of the unemployed from 45%. It is not clear whether these people found work
or dropped out of the labor force. Those unemployed between 5 and 14 weeks jumped by 575,000. So the median and
average duration of unemployment retreated rather sharply. And those working part-time for economic reasons rose by
331,000. That helped the U-6 unemployment rate, the broadest measure, rise to 16.7% from 16.5%. The unemployment
rate rose for all educational levels. Adjusting the household survey to the establishment survey resulted in a gain of
369,000. That brought the two surveys back together over the past 12 months. |
