In theory, the ADP Employment Index should provide analysts with highly valuable information. Automatic Data Processing (NYSE: ADP) is the largest firm that provides payroll services to thousands of U.S. businesses with millions of employees. As such, it has a unique view into how the private sector of the U.S. economy is performing.

And yet, its flagship research report that seeks to leverage its database to provide a picture of private sector employment seems to have never lived up to its potential. In 2010, it was flagged by two of the analysts interviewed by Time's Justin Fox for being an overrated indicator. Here were their comments:

Harm Bandholz, UniCredit Markets and Investment Banking: My least favorite economic indicator is by a wide margin the ADP employment index (hope that is prominent enough). The guys at ADP have repeatedly revised the calculation methodology so that their index has some ability to explain payroll changes ex ante. But the out of sample forecasting ability is just awful. I do not understand, why anybody should follow the index.

Kurt Karl, Swiss Re: The ADP report—too inaccurate and not early enough to be helpful.

That was in 2010. Has ADP's employment index gotten better in the 13 years since?

This being 2023, we opted to let artificial intelligence do the research legwork for us, asking Bing Chat to answer the question "What makes the ADP employment index less than useful?" Here is its response, with links to its sources:

The ADP Employment Index has faced criticism for several reasons:

1. Methodology: The ADP has been criticized for its methodology, which led to a decision to pause the release of its private-sector employment report to retool the methodology.
2. Discrepancies with Government Data: There have been instances where the ADP's job numbers were significantly different from the government's. For example, in one instance, the ADP reported that U.S. companies had added 568,000 jobs in a month, while a Labor Department report showed they'd added 317,000. This discrepancy was attributed to a miss in the leisure and hospitality sector.
3. Estimation Challenges: The current labor market is tumultuous, making it hard to distinguish between permanent and temporary layoffs. This causes estimation challenges for both ADP and the Bureau of Labor Statistics.

These factors can make the ADP Employment Index less useful for some analysts and investors.

Right off the bat, the first item of the Methodology in producing the ADP employment index looks like it was never satisfactorily addressed in between 2010, when Harm Bandholz identified it as an issue, and 30 June 2022, when Barron's reported ADP would "retool" its private-sector employment index. That by itself would be enough to raise caution flags for any analyst looking to divine useful information from the report.

In our view, ADP's discrepancies with government data is less of a concern. As ADP and the BLS draw their data from different sources, differences should be expected. What is really useful is consistent quality in the reported data. If the ADP employment index produces data of consistent quality, differences between it and official government employment data would not be an issue.

Meanwhile, estimation challenges simply comes with the territory of analyzing employment data. It's neither surprising nor disqualifying for ADP to have the same issues as the government's Bureau of Labor Statistics' data jocks do. If anything, that should be an advantage for ADP's research team, because they have access to a 100% sample of the company's private sector payroll databases for its clients. In addition, ADP offers analysts a real potential benefit in providing an independent source for employment data and trends, not the least of which has been the periodic absences of timely government-reported employment data during the various government shutdowns that have taken place since 2010.

Going back to its methodology, there's a natural question that arises from the fact of ADP's 2022 "retooling" of how it determines its employment index: is the revised version more useful? Here's how ADP described what they would be doing in the future in a 23 August 2023 press release:

ADP Research Institute (ADPRI) and the Stanford Digital Economy Lab (the Lab) have developed a new methodology for the ADP National Employment Report (NER) that will provide a more robust, high-frequency view of the labor market with a focus on both jobs and pay. Using fine-grained data, this new measure will deliver a richer labor market analysis that will help answer key economic and business questions and offer insights relevant to a broader audience.

It has been less than a year since the retooled ADP employment index has been producing reports, so the jury is still out on whether it has become a more reliable and useful indicator of the private sector employment situation in the United States. We'll have to revisit this topic at a later date.

The jury however has long since delivered its verdict for the pre-retooled version of the ADP employment index. It never lived up to its potential and was considerably less than useful.

Previously on Political Calculations

• Less Than Useful Data in the World of Big Data
• Less Than Useful Data: Consumer Confidence
• Less Than Useful Data: FHFA House Price Index
• Less Than Useful Data: Index of Leading Economic Indicators
• Less Than Useful Data: Weekly Chain Store Sales
• Less Than Useful Data: Challenger Layoff Announcements
• Less Than Useful Data: ADP Employment Index

Image credit: Wikimedia Commons ADP logo sourced from ADP's 2007 Annual Report. Public domain image.