A jobs report can show payroll employment rising while the unemployment rate also rises, or show little movement in both. That is not necessarily a contradiction. The figures come from separate surveys asking different questions: one counts jobs on employer payrolls, while the other classifies people by whether they are working, looking for work, or outside the labor force. In September 2026, both measures changed little, but their distinct definitions explain why they do not always tell the same short-term story. Bureau of Labor Statistics
One survey counts jobs; the other counts people
The establishment survey asks businesses and government agencies about their payroll employment, hours and earnings. Its unit is a job. A person with two payroll jobs can be counted twice, and people who are self-employed in unincorporated businesses, unpaid family workers, agricultural workers and private household workers are outside its coverage. The survey’s larger sample makes its estimate of monthly employment change more precise than the household survey’s. Bureau of Labor Statistics
The household survey instead samples households and classifies each civilian, noninstitutional person age 16 or older by labor-force status. It counts a person who did any work for pay or profit during the reference week as employed, including someone who worked part time or held a job but was temporarily absent. Its broader scope includes groups excluded from payrolls, and it can report labor-force measures by demographic characteristics. Bureau of Labor Statistics
The unemployment rate has a specific denominator
The unemployment rate is not the share of everyone without a job. To count as unemployed, a person generally must have no job, have actively looked for work in the previous four weeks and be available to take a job. People who have neither a job nor qualifying recent job search are counted outside the labor force. The rate divides the number unemployed by the labor force: employed people plus unemployed people. Bureau of Labor Statistics
That distinction can change the rate even when the number of jobs does not. If someone stops actively searching, that person leaves the unemployed count and the labor force. If someone begins searching and is available, that person can enter both the unemployed count and the labor force. Neither change, by itself, says how many positions employers added or cut.
September’s two readings were both subdued
The September 2026 report said nonfarm payroll employment increased by 29,000 and the unemployment rate was 4.2 percent; BLS described both as little changed. The report also showed 163.2 million people employed in the household survey, up 406,000 from August, while the household estimate of unemployment rose by 78,000. Those household changes were not large enough to establish a clear change in the overall picture, and the agency notes that the household survey has a larger margin of error for month-to-month change. Bureau of Labor Statistics
The payroll figure is also preliminary. September’s release revised July employment from an earlier gain to a loss and lowered August’s reported gain. That is a reminder that the employer series incorporates later business reports and seasonal recalculations. A single monthly change should be read alongside revisions and the broader trend, rather than treated as a final census of every new position. Bureau of Labor Statistics
Read them together, without forcing a match
The two surveys are most useful as complementary measures. Payrolls offer a more precise gauge of month-to-month change in covered jobs and provide industry detail. The household survey describes people’s employment status and captures a broader range of work, including self-employment. They overlap in subject, but they are not designed to produce identical totals or identical monthly movements. Bureau of Labor Statistics
A practical distinction: if payrolls rise, that points to more covered jobs at reporting establishments, not necessarily an equal rise in the number of employed people. If unemployment rises, that means more people meet the survey’s definition of jobless, active search and availability as a share of the labor force; it does not directly count layoffs. Together, the figures describe different parts of a changing labor market. Bureau of Labor Statistics
Sources & further reading
- September 2026 employment report ↗Bureau of Labor Statistics
- How the government measures unemployment ↗Bureau of Labor Statistics
