US jobs growth was likely far weaker than originally estimated in the year to March, closely-watched (and irritatingly delayed) data released by the BLS on Wednesday showed.

The US economy added 818,000 fewer jobs over that 12-month stretch, suggesting monthly payrolls growth was 68,000 slower than previously reported.

Economists expected a downward revision of anywhere between 300,000 and a million. 818,000 thus counts as quite large, and not just relative to estimates. As the figure below shows, it’s the biggest downward preliminary (note the emphasis) tally in a decade and a half.

The implication is that the pace of hiring from April of 2023 to March of 2024 was 174,000 on average, down from 242,000.

This revision will itself be revised early next year. This isn’t the final word on the matter. But it is the only word we have for quite a while.

The figures below, from Wells Fargo, give you some context. The chart on the left shows the final revisions. 818,000, if confirmed six months from now, would constitute an enormous drop.

Large though it is, the figure “wasn’t anything paradigm shifting,” BMO’s Ian Lyngen said of the release, on the way to reminding investors that last year’s preliminary revision showed a 306,000 decline “only to be revised again to just -187,000.”

Of course, markets can’t trade a final revision they don’t have, nor can the Fed calibrate policy to a second estimate that won’t be published for another six months. So, I’d gently suggest that the release shifts the odds incrementally towards a 50bps rate cut at the September FOMC meeting. Or, perhaps I should put it this way: In the event of a sub-100k headline NFP print for August and/or another uptick in the jobless rate, Wednesday’s big downward revision will become more relevant to the policy discussion than it will be in the presence of a robust August jobs report.

“Historically, benchmark revisions have been small. Over the past 10 years, the absolute change from the benchmark revision averaged 0.1% of total employment,” Wells Fargo’s Sarah House and Aubrey Woessner wrote, in a lengthy explainer published earlier this month. Wednesday’s tally was 0.5% of total employment. If that were to hold in the final revision, it’d be five times larger than average.

“We doubt the preliminary benchmark estimate will be a complete game-changer for FOMC members’ current perceptions of the labor market,” House and Woessner went on. “That said, a large negative revision would indicate that the strength of hiring was already fading before this past April, making risks to the full employment side of the Fed’s dual mandate more salient.”

That latter bit — the “more salient” part — is important. Recall that the July FOMC statement included a sea change in the language around the dual mandate. The Fed’s now just as concerned about downside risks to the labor market as it is upside risks to inflation. It’s safe to say subsequent developments (i.e., the triggered Sahm rule and another relatively benign CPI report) increased the Committee’s conviction in the notion that the balance of risks has shifted meaningfully and durably. That’s the context for Wednesday’s payrolls revision.

Commenting ahead of the release, JonesTrading’s Mike O’Rourke said that in his view, “the distinct and clear rise in dovish Fed commentary” this month might’ve been an attempt to get out ahead of the revision, which he suggested will eventually “prove more important than Chairman Powell’s Jackson Hole speech.”