Last week’s employment numbers

As with the GDP numbers last week, the employment numbers were very dubious indeed. The headline number was still a loss of 20,000 jobs. However, the detail point to likely revisions much beyond that number when more exact data have been compiled in the future.

Four Sectors
From Mish Shedlock’s blog, one can see 4 sectors of the U.S. economy of note: Construction, Manufacturing, Retail and Service jobs. The first three categories lost jobs, while the 4th category gained jobs to help produce the -20,000 jobs.

Construction: -61,000
Manufacturing: -46,000
Retail: -27,000
Service: +90,000

In essence, higher paying jobs were lost and lower paying service jobs were the replacement.

Birth/Death Model
The four sectors do not point to any revisions down the line, but rather demonstrate that higher paying jobs are being lost and outsourced while lower paying service jobs are the likely replacement. However, the detail that is most likely to require a revision of -20,000 to a lower figure is the birth/death model.

The Bureau of Labor Statistics, which produces the Employment report cannot know how many jobs are created or lost when new businesses form (business birth) or go out (business death) of existence. As a result, it has developed a model called the birth/death model as a plug for those jobs. Last month this plug was very large: +267,000 jobs created. The problem with the model is that it is not calibrated to swings in the business cycle. Simply put, the model overestimates business failure and underestimates new businesses during upswings, as it underestimates business failure and overestimates new businesses during downswings.

For example, within this +267,000 jobs is a figure of +45,000 construction jobs. How could the construction business be hemorrhaging jobs at a -61,000, yet the Birth/Death model fills a plug for a net +45,000 new jobs? The two data points illustrate clearly that far fewer than net +267,000 jobs were created by business birth and death.

Conclusion
Obviously, one could see that the unemployment number was -287,000 not including the birth/death model. Given that we are likely in recession with the manufacturing sector collapsing and the private sector having lost jobs 5 months on the trot, major revisions to the final number should be expected. Yet, Wall Street loved this number, just as it loved the +0.6% GDP number. Both numbers are pure fiction as we are in the beginning of a major recession.