Snow job: Economists’ estimates that snow would cost 100K or more jobs miss mark

By Christopher S. Rugaber, AP
Friday, March 5, 2010

Snow job: Estimates of weather’s impact miss mark

WASHINGTON — Armed with spreadsheets, mathematical formulas and historical comparisons, legions of economists sought to forecast the impact of last month’s snowstorms on the government’s employment report.

They mostly got it wrong.

Macroeconomic Advisers, a leading forecasting firm, calculated the storms would cost 150,000 to 220,000 jobs.

IHS Global Insight pegged the impact at 100,000 to 150,000.

Economists at JPMorgan Chase & Co. went with 100,000.

The Labor Department won’t quantify the impact of the storms. But most economists agree it was less than predicted — costing perhaps 50,000 jobs or less.

The storms occurred the same week that the government surveyed businesses about their payrolls for February. Under the government’s calculations, employees who didn’t work at all that week — and weren’t paid — weren’t included in its estimate of the company payrolls.

Most of the job losses were expected to show up in the construction, retail, and restaurant industries. That’s because the bad weather likely shut down construction projects. It also made it harder for millions of Americans to shop or go out to eat.

Yet job losses in those sectors weren’t that much worse than in previous months.

Overall, the department said the unemployment rate was unchanged at 9.7 percent in February. And a net total of 36,000 jobs were cut.

How did the estimates miss the mark so much?

One likely explanation: Economists underestimated how many more people are now salaried workers paid biweekly, rather than hourly workers paid weekly.

To throw off the government’s numbers, the snow would have had to prevent an employee from working at all during the pay period covered by the government’s survey. That’s less likely if your pay period is two weeks rather than one.

Those most likely to fall off the government’s payrolls are hourly workers: construction workers, retail clerks, waiters. If they don’t show up on any workday, they’re generally not paid.

But there are likely fewer of those workers than there used to be. Industries that are more likely to pay people weekly — construction and manufacturing, for example — have been hammered by the recession and have shed jobs.

Economists based their estimates on the snow’s impact in part by looking at the fallout from a huge blizzard that hit the East Coast in January 1996. That storm knocked total jobs down by about 200,000 compared to the preceding months.

But the workforce has likely changed in the subsequent 14 years.

“As we move toward a more service-oriented economy, you’re likely to have more biweekly pay periods,” said Michael Feroli, an economist at JPMorgan Chase.

To be fair, the consensus estimate by economists was that employers would cut 50,000 jobs, including the snow’s impact. That’s not that far from the actual loss of 36,000.

“Nobody knows” what the true impact of the snow was, said Nigel Gault, chief U.S. economist at IHS Global Insight. “And nobody will ever know.”

will not be displayed