U.S. Weekly Jobless Claims Stall at 884,000

An American flag on the fence of a farm near the city of Commerce in the State of Oklahoma
Getty Images

U.S. weekly jobless claims held steady at 884,000 last week, slightly above Wall Street expectations.

Economists had forecast initial claims to drop to around 850,000 from 881,000 originally reported last week. Last week’s figure was revised up by 3,000 to 884,000.

Last week’s figures were the lowest since the pandemic drove unemployment claims to unprecedented highs.

Claims hit a record 6.87 million for the week of March 27. Until a month ago, each subsequent week had seen claims decline. But in late July, the labor market appeared to stall and claims have since hovered around one million, a level so high it was never recorded before the pandemic struck.

Jobless claims are a proxy for layoffs and have been closely watched as a signal for how the pandemic is influencing the economy.

The Bureau of Labor Statistics said last week it was changing the method employed for adjusting initial jobless claims to account for seasonal swings in employment. This means that the figures from the past two weeks are not directly comparable to those of the prior weeks.  It is likely that the now-retired method for making seasonal adjustments had inflated the level of claims over the past month or so.

The unadjusted number of actual initial claims under state programs, totaled 857,148 in the week endingS eptember 5, an increase of 20,140—or 2.4 percent—from the previous week.

The number for seasonally adjusted insured unemployment during the week ending August 29 was 13,385,000, an increase of 93,000 from the previous week’s revised level. That is equivalent to a 9.2 insured unemployment rate, slightly higher than the week before. This number is not equivalent to the official unemployment rate released by the Labor Department each month because it counts only Americans eligible to collect unemployment.

 

 

 

COMMENTS

Please let us know if you're having issues with commenting.