February lost 92,000 jobs. A strike and government cuts shaped the drop

U.S. payroll employment fell by 92,000 in February, the Labor Department reported March 6. The unemployment rate changed little at 4.4 percent. The decline grabbed attention because payrolls had increased in January, but the BLS report also identified two forces behind it: strike activity reduced health-care employment, while federal government employment continued its longer downward trend.

Why the total fell

The payroll survey measures jobs on the books at businesses and government agencies. It does not count the number of people who are employed, and it does not assign a cause to every change. When a strike stops workers from being paid during the survey period, those workers may not be counted as employed in that month's payroll data. The BLS specifically said health-care employment decreased, reflecting strike activity.

Federal employment had also continued to trend down. That is a different story from a private company laying off workers because sales fell. Government payrolls can change when agencies reduce staffing or reorganize. The February report therefore combined a temporary labor disruption with an ongoing decline in one part of public employment.

Why the unemployment rate did not jump

The unemployment rate comes from a household survey, not the payroll survey. It counts people without a job who are available and have recently looked for work. A person who is out of work but has stopped searching is generally outside the labor force and not counted as unemployed in the headline rate.

Because the two measures use different surveys, they can move differently in the same month. A decline in payroll jobs does not automatically produce a matching rise in unemployment. The BLS said the rate changed little at 4.4 percent. That does not make the payroll decline irrelevant; it means readers should not treat the rate as a complete count of everyone experiencing job loss or wanting more work.

What the report says about the wider economy

A single monthly decline is a warning signal, not a complete recession diagnosis. Payroll estimates are based on surveys and are revised as more employer responses arrive. February's report followed January's delayed release, so the publication calendar had already been disrupted by the federal funding lapse. The data still described the month of February, even though it arrived later.

The February drop also came as fuel costs and uncertainty about the war with Iran began to matter more to businesses and households. Those events could affect hiring through transportation costs, customer spending and companies' willingness to invest. The payroll report itself cannot prove that a particular geopolitical event caused a particular lost job. It records employment; explaining why requires other evidence.

What workers could take from it

For a worker deciding whether to leave a secure job, a falling payroll total is a reason to look at the details and the local market, not a forecast that every employer will stop hiring. For someone looking for work, a steady unemployment rate can still coexist with fewer openings and longer searches. The report is a national snapshot, not a promise about an individual's prospects.

Later reports revised earlier estimates and showed how much the first headline can move. The clearest reading of February is that employment weakened for that month, with strike activity and falling government payrolls part of the explanation. Whether that weakness would persist had to be answered by subsequent releases.

BLS, February 2026 Employment Situation