Employers added 29,000 nonfarm payroll jobs in September, and the unemployment rate was 4.2%, the Labor Department said in a report released October 2. Both measures changed little over the month, as did employment in major industries overall. The report offers a snapshot of a labor market with modest net hiring, not a complete account of every person seeking work or every job transaction. Its two headline numbers come from separate surveys, and its payroll estimates can change as more records arrive.
What the headline counts
The establishment survey counts jobs on nonfarm payrolls at businesses and government agencies. In September, total payroll employment rose by 29,000 after the August estimate was revised to a gain of 133,000 and July to a loss of 10,000 Bureau of Labor Statistics. Those revisions left the two earlier months together 60,000 below their previously reported level Bureau of Labor Statistics. The release said the September change was little changed, following an average monthly gain of 45,000 over the prior 12 months Bureau of Labor Statistics.
The monthly net change is not the number of people hired during September. It reflects the balance of employment changes at surveyed establishments, estimated from a sample. BLS revises its initial estimates twice as additional employer reports come in and seasonal factors are recalculated Bureau of Labor Statistics. A preliminary number is useful, but later releases can redraw the recent path.
Why unemployment comes from another survey
The 4.2% unemployment rate comes from the household survey, which asks people about their work and job-search activity. Someone generally counts as unemployed if they have no job, actively looked in the prior four weeks and were available to work; a person with a job is counted as employed, including someone temporarily absent from work Bureau of Labor Statistics. The rate is the share of the labor force that is unemployed, not the share of all residents without a job Bureau of Labor Statistics.
The establishment survey measures jobs, hours and earnings by industry; the household survey measures people’s labor-force status and demographic characteristics Bureau of Labor Statistics. A person may hold more than one payroll job, while the household survey covers self-employed workers with unincorporated businesses, agricultural workers and other groups excluded from the payroll count Bureau of Labor Statistics. As a result, the two measures need not move in lockstep.
Small monthly changes need perspective
BLS describes the September payroll movement as little changed across major industries. Health care continued to add jobs, up 17,000, while employment in construction and manufacturing also changed little by the report’s description Bureau of Labor Statistics. Health care’s gain was slower than its average monthly increase over the prior year, and financial activities employment was little changed, down 7,000 Bureau of Labor Statistics. The mix suggests that the net figure masks sector differences rather than showing a uniform hiring freeze.
The establishment survey has a smaller margin of error for monthly employment changes than the household survey because its sample is larger. BLS says a payroll change of about 122,000 is needed to be statistically significant; for household employment, the corresponding threshold is about 650,000 Bureau of Labor Statistics. That does not make a 29,000 estimate meaningless. It means the number is a noisy estimate, best interpreted alongside revisions and a run of monthly results.
What the rate leaves outside the frame
A steady unemployment rate does not mean every person wanting work is counted as unemployed. People who want a job but have not actively searched in the prior four weeks, or are unavailable to start, are outside the labor force under the standard measure Bureau of Labor Statistics. In September, BLS counted 5.8 million people outside the labor force who said they wanted a job; they were not classified as unemployed under the survey’s criteria Bureau of Labor Statistics. The measure therefore has a precise definition, not an all-purpose count of unmet work needs.
The report’s revisions are also a reminder to distinguish an initial estimate from a settled historical record. September payrolls were preliminary, and the October employment report is scheduled for November 6, after the November 1 source update Bureau of Labor Statistics. No October payroll result is available here. For now, the evidence supports a narrow reading: September brought a modest estimated payroll increase, a 4.2% unemployment rate and substantial downward revisions to the combined July-August count Bureau of Labor Statistics.
Sources & further reading
- September 2026 employment report ↗Bureau of Labor Statistics
- How the government measures unemployment ↗Bureau of Labor Statistics
