Opinion: The Federal Reserve Raised Rates on a Labor Market That Was Already Gone

On September 16, the Federal Open Market Committee raised its benchmark interest rate for the first time in three years, lifting the federal funds target range a quarter point to 3.75 to 4.00 percent. The vote was 12 to 0, with new Chair Kevin Warsh joining every other member of the committee. Three weeks later, on October 7, the Fed released the meeting's minutes. By then, a very different labor market had shown up in the official statistics, and the gap between what the committee believed in mid-September and what the government's own data said two weeks after is now the more important story than the rate move itself.

A decision built on a labor market that was about to disappear

The committee's public rationale was straightforward. Inflation remained uncomfortably high, the Fed said, while the labor market gave it room to act. According to the minutes of that meeting, the Fed's staff estimated headline inflation at 3.8 percent and core inflation at 3.4 percent for August, with price pressure attributed in part to "ongoing geopolitical developments," which had pushed up crude oil and refined fuel prices and disrupted Middle East shipping. Against that backdrop, participants judged the labor market to be near full employment and expected unemployment to hold "close to current levels." All twelve of them signed on to the hike as insurance against inflation staying elevated longer than planned, and most said another quarter-point increase would likely be warranted by year's end.

The committee's statement put the labor-market judgment in plain terms:

"Job gains have kept pace with the workforce, and the unemployment rate has changed little."

That sentence is the crux of the problem. It was an accurate summary of the data the Fed had in hand on September 16. It stopped being an accurate summary of the economy roughly two weeks later.

The jobs the Fed was counting on did not show up

On October 2, the Bureau of Labor Statistics released the September employment report, and it undercut the premise behind the hike. Employers added just 29,000 jobs, far short of the roughly 90,000 forecasters had expected, and the unemployment rate climbed to 4.2 percent. The revisions were the more telling part: July's initial gain of 21,000 jobs was revised down to a loss of 10,000, and August's 162,000 was cut to 133,000. Combined, that erased 60,000 previously reported jobs from the two months the Fed's September decision leaned on most heavily. The deceleration tracked in the government's own nonfarm payroll series and unemployment-rate data is not a single noisy data point; it is three consecutive months pointing the same direction, each one worse than first reported.

• September payrolls: +29,000, versus a roughly 90,000 forecast
• Unemployment rate: 4.2 percent, up from 4.1 percent
• July payrolls: revised from +21,000 to -10,000
• August payrolls: revised from +162,000 to +133,000
• Combined July-August revision: -60,000 jobs

The minutes the Fed published on October 7 describe a committee operating on the earlier, rosier picture. That is not a flaw unique to this meeting; minutes are always a historical record of a decision already made. But the three-week publication lag means the document the public finally gets to read is, by construction, out of date on arrival whenever the intervening weeks bring a meaningful data revision, as they just did.

The case for patience, and where it runs out

There is a real defense of what the committee did. Monetary policy should not whipsaw in response to one month's payroll print, which is subject to revision and sampling noise; the Fed's own framework is explicitly built to look through short-term wiggles toward the underlying trend. Inflation near 3.8 percent is also not a trivial reading, and a central bank that eases prematurely while prices run hot risks a worse outcome than one that holds rates too high for a few extra months. Energy costs tied to the disruption of Middle East oil shipments complicate the picture further, since that kind of shock can look transitory right up until it is not.

There is also a political complication that cuts against the simplest reading of this episode, that a Fed shaded toward the White House delivered a favor. Warsh's own path to the chairmanship was the most narrowly contested confirmation of a Fed chair in the institution's history, a 54-45 Senate vote with only one member of the opposing party crossing over. The administration that installed him wanted lower rates, not higher ones, and made that preference public in the run-up to the meeting, with the Fed's own rate increase drawing open criticism from the administration that had just fought to install him. A unanimous committee, including the chair widely assumed to be the president's man, voting for the opposite of what the White House wanted is evidence that the institution's independence is intact. That is a point worth conceding to the Fed's defenders.

But institutional independence and analytical accuracy are not the same thing, and this episode shows why both matter. The committee was not pressured into a bad call; it made a defensible call on a reasonable reading of imperfect information, and the information turned out to be wrong in a way that was already large and already trending before the ink on the statement dried. A Fed that resists political pressure while still building policy on a labor-market narrative the data abandons within weeks has solved only half the credibility problem.

What should follow from here

The practical test comes at the Fed's next meeting, on October 28. The minutes record that most participants expected another quarter-point increase would likely be appropriate before year-end; that expectation was formed on a reading of the labor market that the government's own numbers have since repudiated. Mechanically following through on September's signal, absent some fresh justification grounded in the data that has actually arrived since, would repeat the same mistake in sharper form. The committee does not need to abandon its inflation vigilance to say plainly that the employment side of its mandate now points the other way, and that the December path sketched in September deserves to be reopened rather than defended on inertia.

There is a broader institutional lesson too. A three-week gap between a rate decision and the release of the reasoning behind it was tolerable when the underlying data moved slowly and revisions were modest. It is less tolerable when payroll revisions of the size seen in July and August have become routine, and when a single week's data release can reframe the prior month's policy decision. The Fed does not control how fast the Bureau of Labor Statistics can process its surveys, but it can be more candid, in real time, about how much weight its decisions are putting on numbers it knows are provisional. The independence Kevin Warsh demonstrated by defying the White House's preference is worth little to the public if the committee will not show equal independence from its own stale assumptions when the facts move. The next meeting is the chance to show both.