An Unexpected Payroll Decline: How Should We Read It?

The July payroll report showed a decline of 23,000 jobs, against an expected increase of 80,000. Understanding the report requires more than its headline figure. Unemployment, labour-force participation and revisions to earlier data also matter.

Does lower unemployment mean employment improved?

The July unemployment rate was 4.1%, down 0.1 percentage points from June. Labour-force participation also fell to 61.4%, a multiyear low. People who leave the labour force are no longer counted as unemployed. A lower unemployment rate can therefore accompany fewer jobseekers, without an increase in jobs.

Read current payrolls alongside revisions

Payrolls declined by 23,000 in July, while the May and June figures were revised down by a combined 103,000. Those revisions mean that employment growth in the preceding two months was smaller than initially reported. Assessing hiring requires the revised sequence of observations, rather than initial releases alone.

What do the two surveys tell us?

The establishment survey showed a payroll decline of 23,000. The household survey showed employment falling by 87,000 and the labour force shrinking by 264,000. The first measures jobs at employers; the second measures employed people and participation in the labour market. Both offered weak readings, but they cover different populations and cannot simply be added together.

Bring the readings together

Payrolls fell by 23,000, unemployment was 4.1%, participation was 61.4%, and hourly earnings rose 3.2% year on year. Although unemployment declined, employment and participation also weakened. The report needs to be understood through these changes together.

Data: Bureau of Labor Statistics, July 2026 Employment Situation report.