ADP Employment Change

ADP Employment Change is the headline of the ADP National Employment Report (NER), the change in US private jobs, excluding government jobs. It is published by ADP Research, a research group inside payroll company ADP, with the Stanford Digital Economy Lab, a Stanford research lab. It uses weekly payroll data for over 26 million workers at over 500000 firms, plus about 15 million monthly pay observations. It covers private jobs only, excluding government jobs, unemployed people and the jobless rate; the headline is seasonally adjusted.

It comes out monthly on the first Wednesday at 8:15 AM ET, with exact dates set by the publisher calendar. Jobs use the week including the 12th and pay uses the full month, with a weekly estimate each Tuesday as a four week average. Splits cover industry, size, region and pay in adjusted and unadjusted forms, with weekly history back to 2010 and charts online.

Each release revises the prior month, drops inactive firms and benchmarks yearly to QCEW census counts. Read the sign and size of the adjusted change, groups and pay, with no target and no forecast of government reports.

The same job change is cut by industry, employer size, place and pay, with no fixed weights. Parts may not sum to the total from rounding.

• Goods-producing industries: mining and natural resources, construction and manufacturing activity.
• Service-providing industries: trade, transportation and utilities, information, financial, professional and business, education and health, leisure and hospitality, and other services.
• Employer size: 1 to 19, 20 to 49, 50 to 249, 250 to 499, and 500 or more workers per establishment (one work location).
• Geography: Northeast, Midwest, South and West census regions (official state groupings) and their divisions.
• Pay Insights: median yearly pay change and levels for stayers and changers, by industry and size, plus state, gender and age for stayers.

In the April 2020 example (reported May 6, 2020) employment fell by 20236000 jobs, more than double all jobs lost in the Great Recession. Services lost 16007000 jobs, led by leisure and hospitality at 8607000, while goods lost 4229000 led by construction at 2477000 and manufacturing at 1674000. Large firms lost 8963000 jobs, versus 6005000 at small firms and 5269000 at medium firms. It used data only through April 12 and missed later damage; March was revised from a loss of 27000 to 149000 and April to 19557000.

1. Weekly snapshots: Take weekly payroll data; count Payroll Employment (attached to employer) apart from Paid Employment (actually paid that week).
2. Matched growth: For each establishment (one work site) present in two straight weeks, compute growth as g = (e_t - e_{t-1}) / e_{t-1}.
3. Weight to census: Weight each industry by state by size cell by QCEW jobs over ADP jobs for national fit.
4. Average and level: Combine weighted growth into one national weekly rate, cumulate into an index and scale to QCEW levels. For example, weight 2.0 on 100 workers up 3% plus weight 0.5 on 1000 workers down 1% nets about 0.14% growth.
5. Adjust and headline: Seasonally adjust the weekly series and read the week including the 12th as the monthly headline, split by industry, size and region.