Post by unlawflcombatnt on Jun 1, 2007 16:56:20 GMT -6
Today's report of an increase in Payroll Employment of +157K was greeted with great optimism by market optimists, and much skepticism by others, including economists Nouriel Roubini and Dean Baker. Manufacturing employment declined for the 11th straight month. May's Manufacturing employment declined another -19K. March's Manufacturing decline was revised downward from -18K to -23K.
May's Construction employment change is listed as 0. Both Roubini and Baker dispute this "0" change number, given the overall decline in Housing Starts, down 30% from it's peak. Roubini puts Construction job loss at a minimum of 50K per month. April's Construction job decline was revised downward even further, from -11K down to -22K. It seems likely that the "0-growth" number given for Construction should actually be negative.
Both Baker and Roubini dispute the additional jobs added to the NonFarm Payroll Employment report from the "imputed" jobs number. Baker discusses some of the distortion from the "imputed" jobs addition from the BLS in the following passage:
"The most obvious cause of a problem is the Bureau of Labor Statistic's (BLS) imputation for job growth in new firms that are not included in the sample. This imputation is based on past growth and therefore will miss turning points in the economy, understating job growth during an upturn and overstating it in a downturn.
So, what the's story this time? BLS imputed 529,000 jobs into the establishment data in the first five months of 2006. It imputed 591,000 jobs, or 62,000 more in the first five months of 2007. The economy grew at a 5.6 percent rate in the first quarter of 2006. It grew at a 0.6 percent rate in the first quarter of 2007. It looks to me like they've overstated job growth in new firms this time around. Of course, the good news is that productivity growth is slightly better than the data indicate...."
In general, the "imputed" jobs added to the total (from the business birth/death model), overstates job growth when the economy is slowing down. As Baker notes, there were 529K "imputed" jobs added to the total in the 1st 5 months of 2006, when GDP growth was 5.6%. Yet with a GDP growth now at only 0.6%, an even larger 591K jobs were added. This number is guaranteed to be revised downward in the future, since it is based on historical averages from time frames where the economy was growing faster.
In summary, the jobs number is almost certainly overstated, and probably a lot. A +0.6% GDP growth and a jobs creation number of 157,000 are simply not compatible. And it seems likely that it is the jobs number is more likely to be revised downward, than it is for the GDP to be revised upward.
May's Construction employment change is listed as 0. Both Roubini and Baker dispute this "0" change number, given the overall decline in Housing Starts, down 30% from it's peak. Roubini puts Construction job loss at a minimum of 50K per month. April's Construction job decline was revised downward even further, from -11K down to -22K. It seems likely that the "0-growth" number given for Construction should actually be negative.
Both Baker and Roubini dispute the additional jobs added to the NonFarm Payroll Employment report from the "imputed" jobs number. Baker discusses some of the distortion from the "imputed" jobs addition from the BLS in the following passage:
"The most obvious cause of a problem is the Bureau of Labor Statistic's (BLS) imputation for job growth in new firms that are not included in the sample. This imputation is based on past growth and therefore will miss turning points in the economy, understating job growth during an upturn and overstating it in a downturn.
So, what the's story this time? BLS imputed 529,000 jobs into the establishment data in the first five months of 2006. It imputed 591,000 jobs, or 62,000 more in the first five months of 2007. The economy grew at a 5.6 percent rate in the first quarter of 2006. It grew at a 0.6 percent rate in the first quarter of 2007. It looks to me like they've overstated job growth in new firms this time around. Of course, the good news is that productivity growth is slightly better than the data indicate...."
In general, the "imputed" jobs added to the total (from the business birth/death model), overstates job growth when the economy is slowing down. As Baker notes, there were 529K "imputed" jobs added to the total in the 1st 5 months of 2006, when GDP growth was 5.6%. Yet with a GDP growth now at only 0.6%, an even larger 591K jobs were added. This number is guaranteed to be revised downward in the future, since it is based on historical averages from time frames where the economy was growing faster.
In summary, the jobs number is almost certainly overstated, and probably a lot. A +0.6% GDP growth and a jobs creation number of 157,000 are simply not compatible. And it seems likely that it is the jobs number is more likely to be revised downward, than it is for the GDP to be revised upward.