No Substantive Impact From Hurricane Sandy According to BLS, U Rate Down to 7.7 Percent
- The November Payroll Employment Survey showed a gain of 146,000 jobs, beating expectations.
- The Unemployment Rate for November fell to 7.7 percent as the labor force declined.
- The Household Survey showed an increase of 191,000 jobs in November.
- Average Weekly Hours were unchanged for the month at 34.4. Hourly earnings were up 0.2 percent.
The BLS reports that payroll employment increased by 146,000 jobs in November and the unemployment rate declined from 7.9 to 7.7 percent. The decline in the unemployment rate resulted from a large 350,000 worker decline in the civilian labor force, not from large gains in the household survey of employment. Further, the BLS reports downward revisions to September and October payroll jobs that total 49,000. So the net result of the payroll data is +97,000 from where we thought we were a month ago, and that is consistent with prior expectations. According to the Bureau of Labor Statistics Hurricane Sandy “did not substantively impact the national employment and unemployment estimates for November.” This is a head scratcher. The cause of the apparent disconnect in the employment data from at least a significant short-term disruption to labor markets along the East Coast is likely two-fold. The first issue is the timing of data collection and the definitions of “employment.” In the payroll survey, the data is collected during the week of the month that contains the 12th of the month. In order for the loss of a worker to have a negative impact on the payroll numbers they would have to be off the payroll for the entire month. This requires an assumption of the duration of temporary layoffs fairly early in the month. The second issue is how the BLS accounts for companies that did not respond to their survey. If no adjustment was made for companies completely shut down by Sandy, then an over-count is likely. Finally, according to the BLS, the usual impact of severe weather is a reduction in average weekly hours rather than payroll employment. There was no reduction visible in the November hours worked data, adding weight to the notion that there is a sampling bias in the November data. The household survey of employment is also resistant to the impact of weather. Both surveys likely overstated the jobs numbers for November. The household survey is typically not revised. The payroll survey is, and we may see a downward revision to payroll jobs for November in the months ahead. Finally, the household survey sample is much smaller than the payroll survey sample and has a sampling error of about four times the payroll survey. No doubt this is more than you ever wanted to know about employment data. The take-away from today’s employment report for November is simply that there is no clear trend in the data; lots of noise, little signal. We cannot definitely say whether labor market conditions improved or deteriorated through November. My expectation is that the underlying trend on payroll job growth is in the range of 120,000 to 150,000 jobs per month for the fourth quarter. The unemployment rate would decline only slowly with that rate of payroll job growth. Upside risk to that jobs outlook comes from gains related to improving housing markets, a gradual broadening of momentum in the service sector and very accommodative monetary policy. Downside risk comes from increasingly defensive behavior on the part of many businesses in anticipation of economic drag stemming from the Fiscal Cliff.
Market Reaction: U.S. equity markets opened with gains. Treasury bond yields are up. NYMEX crude is down to $86.24/barrel. The dollar is up versus the yen and the euro.
Click here for a PDF version of the Comerica Economic Alert: