The January Jobs Are Statistical Artifacts

Paul Craig Roberts
Activist Post

Last Friday the US Bureau of Labor Statistics reported that in the first month of this new year 243,000 jobs were created and the unemployment rate (U.3) fell to 8.3 percent. This good news is a mirage. It is due to faulty seasonal adjustments and to the BLS birth/death model. In a prolonged downturn, seasonal adjustments and the birth/death model produce nonexistent employment.

The unadjusted data show a rise in the unemployment rate. The birth/death model, which estimates the net effect of jobs lost from business failures and jobs created by new start-ups was designed for a normal growing economy, not for a prolonged downturn four years old. Statistician John Williams (shadowstats.com) reports that the BLS adds 48,000 new jobs per month to the payroll employment report based on the birth/death model even though the economy has not come out of the deep recession. In other words, over the course of a year, the birth/death model adds about 580,000 jobs to the reported jobs numbers. End of year benchmark revisions quietly take the nonexistent jobs out of the totals, but these revisions do not receive headlines and pass largely unnoticed.

The reported January jobs gains are contradicted by other official reports. For example, the January payroll jobs report shows 50,000 new jobs in manufacturing, but according to the recently released 4th quarter GDP, 81% of the reported growth consisted of undesired inventory accumulation. Normally, companies produce for sales not for inventories. Why would manufacturers be hiring people to produce goods for undesired inventories?

Most of the new reported January jobs are in services. The January jobs report has 24,500 new jobs in wholesale and retail trade and 13,100 in transportation and warehousing. However the data shows that inflation-corrected real retail sales are down. Why does it take more people to sell fewer goods?

The other remaining sizable components of the January jobs number are: professional and technical services (30,000), administrative and waste services (36,700), health care and social assistance (29,700), and leisure and hospitality (44,000) of which the largest component is food services and drinking places (32,800).

The leisure, waitresses and bartender employment numbers seen high for January. Perhaps it was an excellent ski month in the US. However, accommodation (hotels) does not support this conclusion as accommodation lost 3,900 jobs.

The BLS reports 21,000 new jobs in construction. However, the housing report says that housing starts dropped more than forecast in December, falling 4.1 percent. Why does it take more construction workers to produce fewer houses? Building permits, a proxy for future construction, were little changed.

As the adjusted data produce phantom jobs and employment, the BLS should headline the raw unadjusted data. With so many discouraged workers unable to find jobs, dropping discouraged workers out of the measure of unemployment seriously understates the true magnitude of the unemployment problem. If Americans were aware of the double-digit unemployment rate, would they be as tolerant of Washington’s multi-trillion dollar wars? Would Obama be facing a tougher re-election campaign? Would Republicans be pushing to reduce the federal budget deficit at the expense of the social safety net?

The phony data serve many interests, but not those of the American people.

This article first appeared at Paul Craig Roberts’ new website Institute For Political Economy. Paul Craig Roberts was Assistant Secretary of the Treasury for Economic Policy and associate editor of the Wall Street Journal. He is the author of several books including The Tyranny of Good Intentions and How the Economy Was Lost.