June 2, 2006 (Jobs Byte)
Job Growth Weakens in May
June 2, 2006
By Dean Baker
The BLS may be overstating jobs in new firms, which happens at turning points.
The economy added just 75,000 jobs in May according to the Bureau of Labor Statistics (BLS) establishment survey. With the April increase revised down slightly to 126,000, these figures make the two month average just over 100,000. The unemployment rate edged down to 4.6 percent according to the household survey.
The May job growth was far weaker than had been expected. The weakness was widespread, with no sector showing especially strong growth. The retail sector, which lost 27,100 jobs in May, stands out as being especially weak. This sector also lost 43,500 jobs in April, bringing the two-month decline to 70,600 jobs. This decline is concentrated in general merchandise stores, which shed 44,100 jobs since March, but no component of retail has shown much job growth over this period.
The motion picture sector also took an unusually large hit, losing 11,100 jobs – 2.9 percent of total employment – in May. Employment in this sector is 7,800 below its year ago level. The temporary help sector lost 2,800 jobs. Employment in the sector is 18,000 below the November level.
Manufacturing employment fell by 14,000 in May after rising by 19,000 in April. This was driven primarily by a one-month jump in the auto sector. Pulling out this blip, manufacturing employment has been virtually flat since January. However, there has been an interesting shift in manufacturing employment over the last year. The number of production employees, who are generally thought to be less-skilled, has risen by 123,000 over the last year, while the number of non-production employees has fallen by 144,000.
Evidence of the housing slowdown appears to be showing up in the employment data. Employment in residential construction fell slightly in May and has been virtually flat since January. Similarly, employment in real estate appears to be leveling off, with the May figure just 9,000 above the January level. By comparison, the sector had added 46,000 jobs over the eight months prior to January.
There were few notable changes in the data from the household survey, with most of the evidence consistent with a relatively healthy labor market. The percentage of the unemployed who voluntarily quit their jobs, a measure of workers’ confidence in the state of the labor market, rose to 12.5 percent in May, the highest level since August of 2001. This is a picture that would be consistent with rising wage growth. The May data showed the average hourly wage rising by just 1 cent. This is almost certainly an anomaly. The April report showed an unusually large 10 cent increase in hourly wages. The April wage numbers almost certainly overstated wage growth, which is the cause of the weak rise reported for May.
This can be seen clearly in the sector wage data. The most glaring anomaly is in retail, which reported a rise of 9 cents in the average hourly wage in April and a decline of 6 cents in May. It is implausible that retail wages actually fell by 6 cents in May. It is far more plausible that the April figure was erroneous, distorting wage growth data for the month. Taking three month averages, nominal wages have grown at a 4.3 percent annual rate in the March-May period, which is roughly even with the rate of inflation, and more than a full percentage point faster than their growth rate at this time in 2005.
While the labor market may currently be strong enough to produce healthy wage growth (if gas prices stop rising or fall back), the weak job growth numbers give grounds for questioning future strength. It is clearly only a matter of time before the buildup of housing inventory leads to further cutbacks in construction and real estate, which will put a serious damper on what already appears to be weak demand growth. The current data may also overstate job growth somewhat. In the last two months, there were 482,000 jobs added to the data to account for new firms not captured by the survey. This compares to 397,000 in April-May of 2005. If the economy has hit a turning point, the BLS model is overstating jobs in new firms and therefore overstating employment growth.
Dean Baker is co-director of the Center for Economic and Policy Research in Washington, DC
CEPR’s Jobs Byte is published each month upon release of the Bureau of Labor Statistics’ employment report.