There’s good news and bad news for anyone who felt like things were not as they appeared to be with the labor market.
The good news: You weren’t crazy. And the bad news: The labor market was performing way worse than initially reported by the Bureau of Labor Statistics (BLS).
Diving into the data. Employers added 911,000 fewer jobs between April 2024 and March 2025 than previously reported, according to a preliminary benchmark revision released by the BLS this week.
The preliminary benchmark, accounting for around 76,000 fewer jobs a month on average, is the largest downward revision since 2002. By comparison, last year’s first revision downsized job growth by 818,000 (though it was later adjusted to 598,000), and a revision in March 2009 downsized growth by 902,000.
Despite being released in September 2025, the benchmark is unrelated to the latest employment situation data from the BLS, which found that employers added only 22,000 jobs in August.
“The latest economic data—which are wholly unaffected by today’s preliminary revisions—suggest the labor market is weakening for all workers,” Elise Gould and Ben Zipperer, senior economists at the Economic Policy Institute, wrote via the nonprofit. “Job growth has been especially weak since May. Household survey rates also point to falling prime-age Black employment and higher unemployment for US-born workers.”
Leisure and hospitality saw the steepest downward revision of all major industries, adding 176,000 fewer jobs than originally reported. That was followed by professional and business services and retail trade, which saw declines of 158,000 and 126,200, respectively.
However, transportation and warehousing and utilities both had more job growth than initially reported, adding 6,600 and 3,700, respectively.
How it works. Each year, the BLS adjusts its employment numbers for a specific period, originally taken from its monthly establishment survey, using estimates from state unemployment insurance tax records as its benchmark. These records are less timely, but more comprehensive than the BLS’s survey, which has in recent years struggled to get robust response rates. The BLS is expected to release its full revision early next year.
Quick-to-read HR news & insights
From recruiting and retention to company culture and the latest in HR tech, HR Brew delivers up-to-date industry news and tips to help HR pros stay nimble in today’s fast-changing business environment.
One reason for the discrepancies between the monthly survey data and revisions may be how immigration is factored into each report, Ron Hetrick, principal economist at Lightcast, wrote in a LinkedIn post. State unemployment insurance records only include workers who are eligible for unemployment, not undocumented workers who are on employers’ payrolls and pay taxes but ineligible for unemployment benefits. Meanwhile, the monthly survey data does not ask about immigration status. And, as the Trump administration began revoking the legal statuses of thousands of immigrants, they were removed from payroll counts.
“There were more people working on payrolls and that is what the survey is measuring. Now they aren’t,” Hetrick wrote. “These people drove growth and were initially counted, and now they are leaving payrolls.”
Zoom out. Tuesday’s revisions come as the Trump administration has turned its attention to the agency’s data and practices. President Trump’s nominee to lead the BLS has advocated for releasing the jobs report on a quarterly basis to ensure what he claimed is greater accuracy. But economists reacting to Tuesday’s benchmark data noted that despite its flaws, it’s better to take the monthly report for important talent-related decision making than have to wait months to get that data.
“If I had to choose between less accurate information today to guide a decision, versus more accurate data four months from now—after the decision is made—I’d always choose the former,” Kory Kantenga, head of economics for the Americas at LinkedIn, wrote on the platform.