David Rosenberg of Gluskin-Sheff is the latest to blast holes in Friday’s supposedly “good” jobs report. Basically, he thinks the number was a total lie, and that the household survey, which showed continued deterioration, represented the truth.
It is astounding how market commentators leap on every piece of economic data; they merely look at the headline, and then make a judgment on whether it is weak or strong. The U.S. payroll report that came out last Friday was spurious, at best. Yes, yes, the +151,000 headline was nice and well above expectations, but it was also highly concentrated in just a few service sector industries, led by waste management. For an economy gone to waste, maybe that’s totally apropos.
But let’s get real here. The raw data showed that 919,000 payrolls were somehow created in October, which therefore would have made this the second strongest October in the last 11 years — in October 2009, the tally in the raw nonfarm payroll data was 646,000 even though the economy then was accelerating at a 5% annual rate. That 919,000 not seasonally adjusted surge in October far surpassed what we saw at the peak of the cycle in 2007 (740,000 jobs) as well as the boom periods of 2006 (698,000) and 2005 (727,000). The data bear no resemblance to the reality of an economy barely growing at all in real per capita terms.
For a bond or a stock trader, it all comes down to the headline nonfarm payroll number. For a labour market analyst, what is important is the information that comes from many parts of the Household survey. Who in their right mind could ever refer to the jobs report — it is an entire report, by the way — being strong when the employment-to-population ratio (the “employment rate”) dipped two-tenths of a percentage point to 58.3% in October. The labour force plunged 254k and the participation rate fell from 64.7% in September to 64.5% — the lowest level since November 1984! How is that bullish? If not for the slide in the labour force last month, the unemployment rate would have gone back up to 10%
The level of unemployment rose 76k in October and is up now in two of the past three months. They may take issue with Mr. Market’s and Mr. Media’s response to the headline payroll figure. The Household survey, when put on the comparable footing to the payroll report (the “population and payroll concept adjusted” series), showed a 505k slide in employment last month, the steepest decline of the year. That was certainly no +151k. Not only that, but the Household survey found that 124,000 full-time jobs were lost in October, making it a five-month streak during which 1.1 million of these positions vanished, only replaced in part by 690k part-time workers.
To be sure, the payroll survey flagged upward revisions, an uptick in the workweek and a rebound in work-based pay. But the Household survey is consistent with an economy still mired in deep malaise if not contraction. So which survey is correct? Hard to say. Historically, only 5% of the time do the Household survey and Payroll survey diverge in any given month to this extent, and usually it is the former that has the story right. Time will tell.