The federal government’s jobless data that tracks new claims for unemployment benefits tumbled 13% last week to the lowest reading since mid-March – before schools, offices and stores closed amid the worst pandemic in more than a century.
It’s good news for a number tracked in the mortgage and housing industries as a gauge of America’s home-buyer power, but it comes with a catch: The Department of Labor changed the way it computes the seasonally adjusted number and it didn’t apply it retroactively. Eventually, it will get around to it, the Thursday report said – but not before the beginning of 2021.
“It’s hard to understand why they wouldn’t apply the new formula retroactively – it’s something you can do in five minutes in an Excel sheet,” said Dean Baker, senior economist at the Center for Economic and Policy Research. “It’s something you tell a worker, when he says he’s leaving for the day, ‘Hey, before you leave, can you just make the initial claims series consistent.’ Without that adjustment, the numbers are skewed.”
Baker’s workaround was to look at the unadjusted data, which usually isn’t the headline number from the report. It showed a 0.9% increase in new claims to 833,352 from 825,761. Compared to a year earlier, the not-seasonally adjusted number, known to economists as the NSA, was almost five times higher than the year-ago week.
Economists expected the seasonally adjusted number in the jobless claims report to improve. They didn’t expect it to drop off a cliff. A Trading Economics survey of economists showed the average forecast was for a 1.6% drop to 995,000.
Wells Fargo economists, in a report titled “Don’t Get Too Excited About the Drop in Jobless Claims,” said the number published in the report was an embellishment.
“Since the Department of Labor did not apply the new methodology to prior data yet, the drop in today’s headline number embellishes the rate of improvement and should not be viewed as a sign that the labor market’s recovery is kicking into a higher gear,” the economists said.
Goldman Sachs economists said if the government hadn’t switched to the new formula, there would have been a decrease of 17,000, not the 130,000 reported on Thursday.
“The DOL switched from a multiplicative to an additive seasonal factor in this release, but applying a multiplicative seasonal factor implies that seasonally adjusted initial claims decreased by 17,000 to 994,000,” the Goldman Sachs report said.