The consensus view on ADP private payrolls for August is that 38,000 is a weak number, below the Econoday forecast of +55,000, and a signal of labour market softening. That framing is not wrong. But the more uncomfortable question is whether ADP’s figure, or indeed the Bureau of Labor Statistics’ competing count, tells us anything reliable at all about the actual state of hiring in the United States.
The August ADP Number and What Sits Beneath It
Private employers added 38,000 jobs in August, according to ADP Media Center, the slowest pace of job creation since January. The sector breakdown is where the picture gets granular. Fox Business reports that education and health services led all categories, adding 45,000 positions, while manufacturing shed 17,000 jobs and professional and business services lost 16,000. Information also contracted. Construction, leisure and hospitality contributed positively, but could not offset the losses in goods-producing and knowledge-economy sectors.
By employer size, the distribution is lopsided in a way that deserves attention. Firms with 500 or more employees accounted for 34,000 of the total gain. Small businesses in the 1-to-19 bracket added 20,000, but the 20-to-49 cohort subtracted 17,000. Medium-sized employers, taken together, were flat. Large employers are doing the heavy lifting, and the smallest businesses are running in opposite directions depending on how you slice them. That is not a picture of broad-based recovery.
On wages, ADP reports base pay rose 3.2% year-over-year and gross pay was up 4.7%. The job-changer premium remains intact: base pay for job-stayers increased 3.0% year-over-year, while job-changers saw base pay rise 4.7%. Dr. Nela Richardson, Chief Economist at ADP, offered this: ‘Pay can tell us a lot about today’s choppy hiring. To understand hiring patterns, you have to look deeply into where pay growth is accelerating, where it’s slowing, and for whom. Once-predictable wage growth has been overtaken by the complexities of demographic change, persistent inflation, and AI’s effects on jobs.’
The Real Problem: ADP Private Payrolls August and the Reliability Gap
Here is where the consensus coverage loses the thread. Most outlets will present Friday’s Bureau of Labor Statistics nonfarm payrolls print as a verdict on whether ADP got it right or wrong. That framing assumes at least one of these series is close to the truth. The year-over-year comparison suggests neither deserves that status.
In July 2026, ADP reported a year-over-year gain in private employment of 1.593 million. The BLS reported 635,000 for the same period. The gap between them: 958,000 jobs. Both series undergo significant revisions, which means their month-to-month divergence is noise, but a gap of that magnitude on a year-over-year basis is not a rounding error. It is a structural problem in measurement.
The Quarterly Census of Employment and Wages (QCEW) offers a more grounded reference point, though it arrives with a lag. The latest available QCEW data runs through March 2026. Employment in March 2025 stood at 154,686,000; in March 2026 it was 154,772,000. The year-over-year gain on that basis: 86,000. Compare that to ADP’s private-sector year-over-year figure for the same March 2026 period of 708,000, or the BLS private figure of 461,000 and the BLS nonfarm total of 211,000. The QCEW, drawn from state unemployment insurance records and covering virtually all employers, is the only series in that comparison with a genuine claim to accuracy. ADP’s figure is not in the same postcode.
The consensus may be overweighting both ADP and BLS as directional indicators when the QCEW, slow as it is, suggests both are running materially hotter than reality. Investors and analysts who treat Friday’s BLS print as a reliable data point are working with a series that the QCEW has, on year-over-year terms, comprehensively undercut.
On the Federal Reserve’s September meeting: unless Friday’s BLS report delivers a genuinely poor print and the next CPI report comes in tame, a rate hike on 16 September looks the more likely outcome, given the direction of recent bond market moves and signals from Fed officials. The ADP private payrolls August figure is too soft to read as inflationary pressure, but it is also too unreliable, given the QCEW gap, to function as a clear argument for a pause. The Fed will be looking at a jobs market it cannot fully see, using numbers it has good reason to question.
