There was a lot of commotion Wednesday about an obscure — at least to non-economists — labor data revision that, on the face of it, showed 818,000 fewer nonfarm jobs in America than previously thought. To put this annual exercise in context and allay some concerns, it’s helpful to roll back the clock for a moment to late spring of 2023.
A Bureau of Labor Statistics report had just shown that the U.S. added an extraordinary 339,000 jobs to nonfarm payrolls in May, and members of the economics and markets commentariat were hysterical. Here’s what former Treasury Secretary Lawrence Summers said in an appearance on Bloomberg Television with David Westin, about how the Federal Reserve had to tighten monetary policy promptly:
If they don’t raise rates in June, I think they have to be open to the possibility that they may have to raise rates by 50 basis points in July if the economy continues to stay way hot and if inflation figures are robust. We are again in a situation where the risks of overheating the economy are the primary risks that the Fed needs to be mindful of.
Ultimately, the Fed held rates steady in June and raised them by just 25 basis points in July. But as the record shows, the concern at the time was that the economy was too strong to bring down inflation and that the Fed’s rate increases just weren’t working as expected.
This background is important, because some folks are now suggesting that the preliminary revisions to April 2023 to March 2024 data are evidence of a looming recession (which isn’t correct) or proof of the Biden-Harris administration’s poor economic stewardship (which is silly). Trump campaign senior advisor Jason Miller wrote on X that the revisions are part of “one big continued, intentional cooking of the books to benefit Kamala Harris and Crooked Joe before her.” It’s hard for me to think that anyone could type those words with a straight face.
In actuality, the revisions are confirmation that the labor market was never quite as hot as it seemed in real-time — which helps explains why inflation could have cooled so dramatically — but not that it’s in particularly perilous territory right now. This is what “2023 Larry Summers” and others were seemingly hoping for: evidence that the labor market wasn’t overheating last year, and that monetary policy still basically works. I’m happy for “2023 Larry,” and our 2024 selves should be fine with it, too.
What does it mean for the current moment? If anything, it gives the Fed and Chair Jerome Powell a measure of comfort that they are not being hasty in contemplating interest rate cuts starting as soon as next month and a gradual normalization of policy through 2025. It may also allow Powell to tweak his messaging at this week’s Jackson Hole conference, emphasizing that the risks to employment may now be slightly greater than the risk of a reacceleration in inflation data. Minutes published Wednesday from the Fed’s July 30-31 meeting showed that a majority of the participants on the rate-setting committee are already moving in that direction. The latest revisions should more definitively tip the balance.
The closely scrutinized monthly payrolls numbers that we all report on in real time are highly imperfect (like most economic data). They’re based on a sample of employers that can’t always keep up with constant changes in the economy, including the creation of new businesses and the closing of others. Each year, the Bureau of Labor Statistics corrects for this with a benchmark revision process based on a more comprehensive measure of payrolls, the Quarterly Census of Employment and Wages. The revisions to the April 2023 to March 2024 data will be applied to next January’s jobs report, but with the latest QCEW data published Wednesday, we now have a preliminary estimate. (For additional details, check out this great tutorial on the entire process by Burning Glass Institute Director of Economic Research Guy Berger.)
The latest development was noteworthy as revisions go — the largest since 2009 — but that reflects the challenge of publishing real-time in a fast-changing post-pandemic economy, not a deep state conspiracy. It now appears that employers added around 174,000 jobs a month in the period, down from an average of 242,000. But 174,000 is still excellent, and broadly consistent with what you’d hope for if you’re trying to tame inflation without causing a downturn. One interesting wrinkle is that professional and business services payrolls appear to have actually contracted somewhat in the April 2023 to March 2024 period, instead of just flatlining. That’s consistent with the idea of a general malaise in a category that includes many white collar professions, but it won’t shock anyone that follows the data closely.
I’m not surprised about the hubbub, of course. One thing I’ve learned from the past several years is that market bears and recessionistas love highly technical statistical developments that the general public can’t follow. Obsessing about them makes bears feel smart. When they seem to convey hidden weakness in the economy — even better!
But the revisions don’t alter vanilla labor market statistics including the unemployment rate, which remains at a relatively low 4.3%. Nor do they change the fact that initial claims for unemployment insurance remain subdued. Nor do they change the fact that retail sales data points to a resilient, if slightly more cautious, consumer. The stock market even briefly rose after the release, a sign that some traders were expecting the revisions to be a bit deeper. But in a sense, it’s about correcting history, and the path forward still looks relatively encouraging, especially if you believe, as I do, that Fed rate cuts are just around the corner.
Jonathan Levin is a columnist focused on U.S. markets and economics. Previously, he worked as a Bloomberg journalist in the US, Brazil and Mexico. He is a CFA charterholder.
