A tale of two employment datasets

Maybe the 2020 recession wasn't as bad as we thought. That's the message being sent by the keenly awaited new ADP National Employment Report.

Why it matters: The monthly employment report from the Bureau of Labor Statistics is the gold standard measure used to understand the job market — alone in hitting the sweet spot between timeliness and accuracy. Now, payroll services provider ADP hopes its new monthly report will rival the BLS release on both fronts.

• The new data underscores economists' previous warnings that official BLS data at the height of the pandemic was less than reliable. But there's no way of telling whether ADP's numbers are any better than the ones coming from the U.S. government.

How it works: Instead of using survey data to measure employment — as the BLS does — ADP uses actual corporate payroll data that is less susceptible to the kind of survey errors that grew enormously during the pandemic.

The big picture: The BLS showed a massive drop in employment in early 2020, and then an ultra-fast recovery. ADP shows a much smaller fall — and then no real recovery until well into 2021.

• The most recent datapoint for the ADP series, at 121 million private-sector workers, is almost exactly where the series stood at the end of February 2020, before the broad pandemic-related layoffs. That's in line with what we're seeing from the BLS report.

The bottom line: The new report will be of particular interest to corporate economists and regional Federal Reserve banks, because of the way it breaks its data down by region and industry.

• It will be of much less interest to economists at banks and investment firms, however, who historically have been mostly interested in how useful the ADP report is in guiding expectations for the BLS report that comes out two days later. In that respect, ADP is being unhelpful by design.

Go deeper on the new ADP report: Payrolls slowdown may have arrived