Fed questions the BLS jobs model

I have talked about flaws in the BLS birth death model many times. The latest was in
Employment on Pluto Rises.

The next question is whether or not we really created 92,000 private sector jobs. I doubt it. The overall number seems plausible until you look at how we got there. Closer inspection shows this is the fourth consecutive jobs report that appears like it was written about Pluto or Mars.

Once again The birth/death model numbers are other worldly.



Look at all the subprime lenders that blew up. Look at the problems in housing itself. Next take a look at the number of jobs the BLS assumes to have been created in construction.

The BLS is assuming not only that jobs were added, but that new unaccounted construction businesses were created in this environment (where business capex spending has been weak, housing has been horrid, over 60 lenders have gone out of business or stopped writing loans, and the GDP has collapsed from 3.5% to .6%).

With 142,000 construction jobs being added over the last four months by the BLS model, these reports look like they are from some other planet.

Fed Questions the Birth/Death Model

The jobs numbers are so distorted that even the Fed is now questioning the accuracy of the BLS birth death model. John Berry writing for Bloomberg is asking Why Haven't Home Construction Jobs Disappeared?

The mystery is why the collapse of the bubble in the housing market hasn't been a much greater drag on payroll employment than it has. "Something isn't right," said economist Ray Stone of Stone & McCarthy Research Associates. "We think that the BLS monthly payroll estimate is overstating the pulse of labor market conditions."

Federal Reserve officials have been waiting for months for construction employment to fall, but it hasn't. [Janet] Yellen [president of the San Francisco Federal Reserve Bank] expressed some skepticism about the accuracy of the construction jobs data, and the way in which it has depressed growth in productivity, the output per hour worked.

"Most of the recent slowdown in labor productivity growth can be accounted for by such lags in just one sector -- residential construction. Although this sector has experienced huge drops in spending, employment has been remarkably well sustained," she said.

"Going forward, as the adjustment lags work themselves out, residential construction employment may post significant declines and productivity in that sector and the economy as a whole may rebound," Yellen said.

Aside from construction, Stone also questioned why the 42,000 drop in temporary-help agency jobs since June 2006 hasn't been reflected in a pronounced slowing in overall payroll jobs gains. Historically, changes in temp services payrolls have led overall payroll trends, Stone said.

This time it's different. Temporary employment stopped growing at the beginning of last year and then began to decline. Overall payrolls growth has slowed, though not by all that much. In the last 12 months, 2 million jobs have been added. The big discrepancy, though, remains in residential construction.

Stone said he thinks the problem lies in the combination of a relatively small number of construction firms in the Bureau of Labor Statistics' sample of business establishments -- on which the payroll estimates are based -- and the extensive use of undocumented workers in the industry.

"It may be that these payrolls are declining quickly, but the decline may be most pronounced among illegal immigrants that were not officially counted in the payroll data," he said.

Stone goes on to point out that the household survey has taken a big hit in payroll growth over the past 6 months, averaging a mere 45,000 jobs while the establishment survey is averaging 145,000 jobs a month.

Given that the unemployment rate is based off the household survey, and that the unemployment rates has not risen, we have yet another discrepancy in the BLS data. The game (lie) here is to lower the participation rate. If fewer people are looking for jobs then the unemployment rate does not rise.

But I don't believe we can account for all of these discrepancies because of illegal aliens. If Stone is correct that a relatively small number of construction companies are surveyed, then one might expect those companies to be the larger ones (major home builders like Lennar, Meritage, KB Homes, etc). Now I do not doubt for a second that those companies hired illegal aliens, but what I do doubt is that they were undocumented. The difference being someone receiving pay under the table with no taxes withheld vs. someone receiving a payroll check using a bogus Social Security number (and therefore showing up on the establishment survey).

More to the point, shouldn't the slowdown in temporary workers in conjunction with the slowdown in starts and permits have affected the model at the BLS?

That last question is key to understanding where the problem with the numbers is. Somehow, for some unexplained reasons, the model at the BLS is assuming a creation of construction businesses even as temporary workers are being let go, even as homebuilding starts and permits have plunged, and even as the GDP fell to a crawl at .6%. One cannot blame illegal aliens and the establishment survey for that.

The only valid conclusion is the birth/death model used by the BLS is fatally flawed. That bogus model actually added 142,000 construction jobs over the last 4 months. To any extent the Fed believes (or did believe) the numbers from the BLS, the Fed then was making interest rate decisions based on flawed data. If one also believes the CPI is fatally flawed, then one can see the entire process used by the Fed to set interest rates is a complete joke.

But the major problem is not that the Fed is using bad data. The major problem is that we have a Fed in the first place. It's time to get rid of the Fed and to stop all of these shenanigans. The way to do that is simple. Vote for Ron Paul.

Mike Shedlock / Mish
http://globaleconomicanalysis.blogspot.com/