Nonfarm payrolls fell by 2.8 million in January – before seasonal adjustment, that is. Adjusted, payrolls advanced 157,000, about as expected. However, annual benchmark revisions showed a more rapid pace of job growth over the last two years – a pace at odds with the Household Survey data. How might the Fed view the range of job market data?
Each year, the Bureau of Labor Statistics benchmarks the Establishment Survey data to payroll tax receipts. In September, the BLS estimated that this year’s revision would add 386,000 to the March 2012 level of payrolls (prior to seasonal adjustment). Instead, the revision was +424,000 (or +0.3%, well within the usual range for these revisions). The details showed faster growth in private-sector jobs and somewhat larger declines in government jobs (mostly state and local). As it stands now, the economy averaged a gain of 181,000 jobs per month in 2012, vs. 175,000 per month in 2011, but to be consistent with the growth in the working-age population, we need about 120,000 per month. The unemployment rate should be falling.
The unemployment rate has declined in the last few years, but that’s been due largely to a decrease in labor force participation. The unemployment rate peaked at 10.0% in October 2009 and is now down to 7.9%. However, the employment/population ratio is right where it was then. The pace of payroll growth over the last two years should have boosted this ratio. Something is amiss.
Recall that the Establishment Survey covers about 145,000 worksites and about 557,000 separate worksites, and is tied back to payroll tax receipts. The Household Survey covers about 60,000 households. Monthly changes in nonfarm payrolls are only accurate to ± 90,000 (that’s a 90% confidence interval for those of you who stayed awake during your statistics class). The monthly unemployment rate is accurate to ± 0.2%. The two surveys have a number of differences. The Household Survey includes agricultural workers, self-employed workers whose businesses are unincorporated, unpaid family members, and private household workers. The Establishment Survey does not. The Household Survey counts people on unpaid leave as employed. The Establishment Survey does not. These and other differences could account for some of the discrepancy between nonfarm payroll growth and the unemployment rate.
The Fed has an unemployment rate threshold (6.5%) for the overnight lending rate, but a qualitative threshold ( “substantial improvement” in the labor market) for its asset purchase program. It’s a judgment call, but the recent range of labor market indicators could hardly be called “substantial improvement.”
Many of the economy’s headwinds (housing, state and local government) have faded, but the risks (Europe, excessive fiscal tightening) remain tilted to the downside. The bigger test for the job market will come soon enough. Unadjusted payrolls typically pick up sharply from February to June. That pace may determine whether the Fed cuts its asset purchase program sooner rather than later – but most likely, it’ll be later.
© Raymond James