Why Sometimes You Just Can’t Trust Payroll Estimates
Counting who is working and who is not working in a vast country like the United States is surely a daunting task, especially when rocked by a pandemic. Yet the employment report is a key element of economic planning.
Unfortunately, Wall Street houses are often wrong and the official government figure may be drastically revised over the following months. That’s the warning from economist Gary Shilling, who warns people to take jobs data “with a big lump of salt”.
The US Bureau of Labor Statistics (BLS) released the Nonfarm Payrolls Survey this morning and said the US added 467,000 jobs in January, a huge jump from the previous month. The unemployment rate was 4%.
Prior to this, forecasts for the past month were everywhere and no one expected such a big increase. Bloomberg’s survey of Wall Street economists said employers would add 150,000 jobs in January. Citigroup was less optimistic but still positive, predicting a leaner increase of 70,000 jobs. Goldman Sachs expected a payroll cut of 250,000. The sourest estimate was that of the PNC – a whopping 400,000 job cuts.
The only company with a decent, if flawed, view of the employment picture is payroll processor ADP, which has posted a loss of 301,000 since December. But the ADP study, released Wednesday, is less comprehensive than that of the BLS, as it only covers private sector employment.
“Government figures, subject to significant revisions over the following months, tend to miss important turning points,” Shilling, chairman of his eponymous investment research and management firm, wrote in his latest monthly bulletin. . “The payroll numbers forecast is likely to be off target and reflect the innate optimism bias of Wall Street economists.”
For November, to take an example, the economists of a the wall street journal the survey predicted an increase of 573,000 jobs from October, seasonally adjusted, Shilling said. Then the first estimate from the BLS was an increase of 210,000, only 37% more than the Newspaperthe estimate of.
Shilling pointed to the weakness of the BLS’s own system for guessing the payroll number. The agency’s initial estimate of salaried employment (the one we received today) is based primarily on two factors: the data it obtains from employers and, since it is incomplete so soon after the end of the month previous (January, in this case), this renders an extrapolation from the trends of the previous months.
The problem, he explained, is that this approach fails to detect whether a sudden change in the trend line has occurred, such as that resulting from a sudden spike in virus cases from a new variant.
Shilling added that “if the prior trend in payroll employment was up, but has just fallen, subsequent revisions using more new data and less of the past trend will be negative, and vice versa.”
Related stories:
OK, why aren’t these jobs filled? Hint: Silver
Special report: Green jobs have a bright future…but fossil fuel work will persist
Yay, UI claims plunge, but market unimpressed
Tags: ADP, Bloomberg, BLS, Citigroup, Gary Shilling, Goldman Sachs, jobs report, pandemic, PNC, Wall Street Journal