ADP Employment Change

ADP, one of the largest payroll processors in the United States, publishes a monthly estimate of private-sector job growth based on anonymised payroll records from its client businesses. The methodology, rebuilt in partnership with the Stanford Digital Economy Lab, counts workers who were paid during the month rather than modelling toward the government's number. Government jobs are not covered at all. The report also breaks employment down by firm size, industry and region, and includes pay growth for people who stayed in their jobs and those who changed.

Its timing makes it tempting as a preview of the BLS payrolls release. In practice the two have often diverged, sometimes by a wide margin, because they measure different things with different samples and seasonal treatment. Markets therefore tend to react to ADP most when it lands far from expectations or confirms a trend already visible elsewhere. When it does move markets, the effect on Treasury yields and the dollar has usually been modest and sometimes reversed once the official data arrive.

The pay insight data, derived from the same records, has become arguably more useful than the headline: it tracks actual pay of matched individuals and is less distorted by changes in the mix of jobs than average hourly earnings. Revisions to the prior month are published with each release.

It is best read as one input among several, beside weekly claims, the ISM employment sub-indexes and the Conference Board's labour-differential measure.