01

What matters

This explainer compares ADP’s September 30, 2026 release with BLS’s October 2, 2026 release. ADP reported private employment growth of 90,000; BLS reported total nonfarm growth of 29,000 and unemployment of 4.2%, describing both BLS measures as little changed. These are differently defined measures, not interchangeable readings of one statistic.

02

Verified facts

  • ADP reported on September 30, 2026 that private-sector employment increased by 90,000 jobs in September.
  • ADP describes its National Employment Report as based on anonymized weekly payroll data from more than 26 million U.S. private-sector employees.
  • Beginning with its September release, ADP said its report incorporates first-quarter Quarterly Census of Employment and Wages (QCEW) data issued by BLS on August 28, 2026.
  • On October 2, 2026, BLS reported September total nonfarm payroll growth of 29,000 and unemployment of 4.2%, both little changed. Its establishment survey measures nonfarm employment; its household survey measures labor-force status, including unemployment.

03

Analysis

Bounded analysis: The first distinction is coverage: ADP’s private-payroll estimate is not BLS’s total nonfarm estimate. The second is measurement: BLS’s unemployment rate comes from its household survey, not its establishment payroll survey. The supplied headline comparison therefore cannot isolate a government-employment change by subtraction. ADP’s payroll-based estimate is distinct, while its stated incorporation of BLS QCEW data establishes a methodological connection. That disclosure does not specify benchmarking, calibration, weighting or adjustment mechanics. Different scope and measurement explain why the figures are not directly interchangeable; these excerpts do not explain the specific numerical difference.

04 · VERXERA view

Editorial opinion

VERXERA opinion: The evidence-led conclusion is that these releases should be read as distinct employment measures, not competing answers to an identical question. ADP’s 90,000 private-sector estimate and BLS’s 29,000 total nonfarm estimate differ in coverage, while BLS’s 4.2% unemployment rate belongs to a separate household-survey measure. Connecting those facts matters because an apparent headline clash can obscure what each number actually describes. ADP’s disclosure that it incorporates BLS QCEW data also prevents an overstatement of complete data independence. The strongest limitation is that the excerpts do not identify the causes of the numerical difference or supply full methods and revision details. Deeper source review is worthwhile for explaining the gap, rather than merely recognizing that the headlines are not like-for-like.

This is VERXERA’s editorial interpretation, separated from the facts and analysis above. The reader keeps the final judgment.

05

Counterevidence and limits

The two primary producers support distinct attributed facts; they do not mutually corroborate the same estimate. Their numerical difference alone establishes neither fraud nor a contradiction proving either estimate wrong, and it provides no basis here for a Federal Reserve forecast.

06

Known uncertainties

The excerpts do not establish the specific reasons for the difference or provide the full methodology and revision details needed to explain it. The precise mechanics of ADP’s QCEW incorporation remain unspecified.