Stocks mostly moved lower on Friday despite a massive, unexpected, and somewhat curious drop in the unemployment rate on a mediocre increase in payroll jobs. The difference was in the household survey of unemployment, on which the unemployment rate is calculated. It surged by a massive 873,000 last month (led by college-age, part-time jobs) vs. a 368,000 drop in August.
That’s the largest one-month gain in 29 years (excluding annual Census population adjustments). And it’s the largest statistical outlier in the data since 2003. Political operatives on the right are crying foul. And after an initial ramp in the first few hours of trading, Wall Street started doubting the numbers and selling into the rise given all the other evidence (factory orders, durable goods order, low CEO confidence and weak GDP growth) that the economy has hit another soft spot.
There are some irregularities. Superficially, there are things like the way the number of people that gained part-time employment for economic reasons (+582,000) was precisely two-thirds of the +873,000 household survey increase. That’s 66.666666% out to as many decimal places as you’d care to calculate.
That seems a little too perfect.
The broader U6 measure of unemployment was also unchanged, at 14.7%, because of the rise of part-time workers that would rather have a full-time job. The gap between the U6 and the standard U4 measure remains large.
The report also relied heavily on gains in the educational sector, which compensated for weaker-than-expected private sector gains, as well as part-time positions. And the household survey, the source of the drop in the unemployment rate, is well known for being very volatile and prone to seasonal adjustment errors. For instance, there has been a surge in part-time positions in September for each of the last three years, which suggests the BLS’ models need tweaking.
Analysts at Merrill Lynch aren’t convinced the gains are sustainable given the economic drag likely from the “fiscal cliff” of tax hikes and spending cuts worth around 5% of GDP unless Congress acts before year end. Already, political uncertainty has CEOs nervous, withholding capital investments and new orders.
That will drag on employment in the months to come. Unless payroll gains increase markedly, the drop in the unemployment rate won’t be maintained — the point made in the Merrill Lynch chart above.
Payroll gains of around 100k just won’t cut it. According to Gluskin Sheff economist David Rosenberg, given that we’re in the 39th month of an economic expansion, typical monthly payroll gains would be around +220,000. Also, 47% of private sector firms either cut their staffing requirements or held them steady, up from 42% just three months ago. Fewer than 40% of manufacturers are hiring.
No surprise then that over half the payroll gain was concentrated in two sectors: Education/health and leisure/hospitality, which combined represented just a 25% share of overall employment.
Also, weekly wage gains were offset by higher costs at the gas pump and at the grocery store. Thus, despite the drop in the unemployment rate, Rosenberg expects Q3 GDP growth to still clock in around 1.5%. This just isn’t fast enough.
Philippa Dunne of the Liscio Report notes that we’ve regained close to half — 48% — of the jobs lost between December 2007 and February 2010. At the current pace of job creation, it’ll take 40 months to recoup all the losses.
The proof is in the pudding. If the job gains translate into higher spending and income data, and stronger September retail sales numbers, than the critics and skeptics will be proved wrong. At this point, with CEO confidence dropping hard, I don’t think that’s likely.