Trump is mad about inaccurate jobs reports. One of the top reasons the Bureau of Labor Statistics’ jobs reports have been more error-prone in recent years is that the response rate to one of its most important surveys fell off a cliff during Covid and has not recovered. The Federal Economic Statistics Advisory Committee (FESAC) was a team of unpaid statistical experts that was working to solve the issue of low response rates, among other things. FESAC was disbanded in February by Secretary of Commerce Howard Lutnick.
To create the jobs report, the BLS uses two surveys: the household survey and the establishment survey. The household survey goes to employees, and the establishment survey goes to employers. The response rate for the establishment survey in 2019 was around 60 percent. It fell off a cliff during Covid and today sits at 43 percent.
The pre-Covid response rate wasn’t great; the current one is abysmal. It would be hard enough to tell how many people are employed in this gigantic country of ours if every business that was sent the survey returned it. With less than half doing so, it’s that much harder.
If you want the jobs report to improve, the survey response rate needs to improve. Declining response rates have been a problem across statistical agencies, and it was something FESAC was working on.
FESAC was originally established in 1999. It was an advisory group on technical methodology for the BLS, the Bureau of Economic Analysis, and the Census Bureau. And the first half of its last meeting before Lutnick disbanded it was all about increasing survey response rates.
On December 13, 2024, FESAC heard from David Freeman of the U.K. Office of National Statistics. The Brits moved to an online-first survey in 2023. It got them a higher response rate and larger sample size at a similar cost to the old method.
Then, they heard from Vincent Dale of Statistics Canada. He talked about how Canada has been trying to increase response rates by contacting people through multiple methods of communication.
Next, Frank Schüller and Jörg Enderer of the Federal Statistical Office of Germany discussed recent improvements to response rates in their country. They found improvements from simplifying the wording of questions.
Two FESAC members then presented on the U.S. challenges and suggested various ways of improving response rates. One suggestion was increasing the sample size, but that would also substantially increase the cost. They considered whether monetary incentives or changing the survey lengths would be a meaningful improvement.
In other words, this advisory group was doing exactly what one should want the government to do to improve its service to the public: Asking questions about what works based on experience and trying to formulate a better process. It’s very technical and boring to outsiders, but FESAC was addressing the exact issue that has been causing the jobs report and other government surveys to decline in quality in recent years.
Lutnick terminated FESAC effective February 28. The email announcing it said that “the Secretary of Commerce has determined that the purposes for which FESAC was established has been fulfilled.”
This wasn’t a cost-saving measure because the 15 FESAC members were all unpaid. (They could be reimbursed for travel expenses for two meetings per year.) They had full-time jobs elsewhere, in academia, at think tanks, and in the private sector. They were volunteering their expertise to help improve the data that the government and private sector rely on to make decisions.
Was FESAC going to completely solve the long-running problem of lower response rates? No. But having it is better than not having it, and at little to no cost to the taxpayer, Lutnick did not need to disband it. Now, the Trump administration is mad about the exact thing FESAC was working on during its final meeting.