The benefits ran out first

Nearly four in ten people who draw regular state benefits see their unemployment benefits run out in full. The US Department of Labor’s second-quarter UI Data Summary, published September 11, 2026, puts the exhaustion rate at 39.26 percent. The quarter itself counted 516,252 exhaustions and 1,125,223 first payments. NPR reported on October 11 that nearly four in ten people exhaust their unemployment benefits before they land a new job. The Labor Department’s table carries a matching figure.

The benefit is shorter than the search

The same tables put the average benefit at 15.8 weeks, at $500.61 a week. That is how long people drew benefits, not how long they searched. The Bureau of Labor Statistics’ September report, released October 2, 2026, puts the median spell of unemployment at 11.5 weeks and the mean at 24.8 weeks. The median sits under the average benefit. The mean sits nine weeks past it. NPR’s Scott Horsley draws the connection: the average worker is taking longer to find a new job, so more people exhaust their benefits before any new paycheck begins.

Where you live sets the clock

The national rate hides a wide state spread. North Carolina’s exhaustion rate was 53.42 percent, with benefits drawn for 9.5 weeks on average at $313.18 a week. Oklahoma, California, Kentucky and Florida sat at 49 percent or above. Vermont was lowest at 16 percent. Colorado, at 47 percent, paid $652.36 a week for 14.8 weeks on average. NPR describes the program limits: Colorado offers six months of benefits, North Carolina less than three, capped at $350 a week. Those limits are not the Labor Department’s measured averages. The department also notes that not all state reports have been received and that the data will reflect state revisions with each quarterly update.

What the rate measures

The 39.26 percent is a rate for people who drew regular state benefits. Not every unemployed person claims them, so it is not four in ten laid-off workers and not four in ten of the unemployed. The Labor Department divides average monthly exhaustions by average monthly first payments, with the denominator lagged six months “to allow for the normal flow of claimants through the program”. It is a rolling measure of the program, not one person’s odds.

Inside the window

When the window is this short, every application sent inside it has to go to an open job and be written for it. CoBlack keeps a search running against openings drawn from employer career pages and applicant tracking system feeds rather than public job boards, matches on capability rather than job titles or keywords, and shapes each application to the specific opening.

The benefit ends on a date the state sets. The search ends on a date nobody sets. In the Labor Department’s second-quarter 2026 tables, for nearly four in ten claimants, the first came before the second.