October Employment Report: Headline Jobless Rate is Troubling, but it is a Lagging Indicator
Civilian Unemployment Rate: 10.2% in October vs. 9.8% in September, cycle low is 4.4% in March 2007.
Payroll Employment: -190,000 in October vs. -219,000 in September, net gain of 91,000 jobs after revisions of payroll estimates for August and September.
Hourly earnings: $18.72 in October vs. $18.67 in September, 2.4% yoy increase vs. 2.5% yoy increase in September, cycle high is 4.28% yoy increase in Dec. 2006.
Household Survey - The significant increase in the unemployment rate to 10.2% in October from 9.8% in September is a shocking headline. At the first blush, the 3.5% increase in third quarter real GDP and the weak employment report are puzzling but closer scrutiny indicates that the unemployment rate is a lagging indicator (always peaks at the end of a recession or after a recovery has commenced, see chart 1). There is an important difference to note in the timing of the peak of the jobless rate in the post-war period. The unemployment rate took longer to peak after a recovery was underway in the 1990-91 and 2001 recessions. In other words, real GDP was advancing for several quarters before the unemployment rate peaked. A similar case is projected for the current recovery. Our current forecast is for the unemployment rate to peak in mid-2010. At the same time, real GDP should continue to advance during the final months of 2009 and all of 2010.
In addition to the level of the unemployment rate, the magnitude of the change in the unemployment rate in October is bothersome. Have these sort of large monthly increases occurred during past recessions? Yes, they have (see table 2) in all post-war recessions excluding the 1990-91 and 2001 episodes. The gains in the unemployment rate were below the median (3.4 percentage points) increase during the 1990-91 and 2001 recessions and the monthly increases during these periods were relatively more subdued compared with other recessions. The current recession has set a number of records and two of which are the largest increase in the jobless rate from the recent cycle low and the largest number of monthly gains which exceed 0.4 percentage points in the post-war period (see table 2).
Establishment Survey – Nonfarm payrolls fell 190,000 in October, following upward revisions of August and September estimates which led to net addition of 91,000 jobs over the two-month period. The trough of the year-to-year change in payroll employment appears to be August 2009 (see chart 4). About 7.3 million people have lost jobs since the recession began in December 2007.
Highlights of job losses/gains in October:
Construction: -62,000
Manufacturing: -61,000
Services: -61,000
Retail trade employment: -40,000.
Professional and business services: +18,000
Temporary help: +34,000
Financial activities: -8,000.
Health care employment: +29,000
The workweek in October held stead at 33.0 hours. The index of hours worked fell 0.2% in October, which is smaller than the 0.5% drop of September. The 0.4% decline in the index of factory man-hours points to a drop in industrial production. However, industrial production advanced in September despite a decline of the man-hours index, with a repetition likely in October. The 5 cents increase in hourly earnings to $18.72 and a 0.3% gain in weekly earnings augurs positively for personal income in October. The 2.4% increase in hourly earnings from a year ago continues to stress the absence of inflationary pressures (see chart 3) in the near term.
Conclusion – The October employment report paints a picture of continued weakness in labor market conditions. However, the positive elements should be highlighted lest they are buried among the details of the negative aspects. First, nonfarm payrolls are falling at a significantly slower pace that in the early months of 2009 (see chart 4). During the three months ended October, the average job losses was 188,000 compared with 357,000 in the previous three-month period. During the six months ended April 2009, a staggering 645,000 jobs were lost each month. Non-farm payroll employment, unlike the unemployment rate, is a coincident indicator, with the decelerating trend indicating a gradual improvement in hiring conditions.
Second, the longer manufacturing work-week (+0.1 hour) and overtime (+0.2 hours) of the factory sector in October compared with the prior month are noteworthy. Third, temporary payrolls have increased for three consecutive months.
Fourth, the upward trend of the diffusion indexes (see chart 6) are noteworthy as they underscore a slow but steady progress. In sum, the grim October employment report supports expectations of a Fed on hold for several months. As The Economist magazine puts it, there are aspects of both “gloom and glimmer” in the October employment report.
