Jerome Powell: The Grand Speculator
The Fed chair once warned against using speculative forecasts to drive policy. Now he’s doing exactly that.
Federal Reserve Chair Jerome Powell made a quiet but extraordinary admission on Tuesday: if the Fed were following the actual data, it would be cutting interest rates. But it isn’t—because the Fed expects President Trump’s tariffs to raise inflation, and it’s choosing to act on that forecast instead.
“If you just look at the basic data and don’t look at the forecast, you would say that we would’ve continued cutting,” Powell told lawmakers. “The difference, of course, is at this time all forecasters are expecting pretty soon that some significant inflation will show up from tariffs. And we can’t just ignore that.”
That’s a remarkable departure from the Fed’s longstanding mantra of data-dependence. It also reveals the extent to which the central bank is allowing anti-tariff bias—and speculative inflation models—to override clear economic signals pointing toward looser policy.
The Data Say Go
The data are, in Powell’s own words, favorable to a resumption of rate cuts. Inflation has come down meaningfully. We don’t yet have the personal consumption expenditure index reading for May, but Harvard economist Jason Furman’s calculation based on CPI is that the three-month annualized rate is around 0.6 percent for headline inflation and 1.4 percent for annualized inflation. The year-over-year figure is two percent for headline, exactly at the Fed’s target, and 2.5 percent for core inflation.
Yet Powell is blocking further cuts because of what inflation might do, not what it’s doing. Specifically, he and his colleagues fear that Trump’s April 2 “Liberation Day” tariff announcement will eventually lift consumer prices—even though so far, that impact is not visible in the data. What’s more, history—from the famous Smoot-Hawley tariffs to Trump’s first term tariffs—indicate that tariffs are not inflationary. If anything, they may be a deflationary drag on growth, as economic theory would suggest a tax hike would be (making it all the more important to couple tariff hikes with tax cuts).
The next Fed meeting is in late July. Between now and then, the Fed will only get one more month of inflation data with the May PCE inflation and the June CPI and PPI. It seems highly unlikely that if Powell is not ready to cut based on the last four or five months of data, including at least two months of post-Liberation Day inflation reports, one more month is not going to move the chains.
Forecast-Dependence, Not Data-Dependence
This shift is all the more striking because Powell once warned against this exact approach.
In a 2018 speech at Jackson Hole, Powell cautioned that the Fed must not overreact to speculative or poorly grounded forecasts:
“Common sense suggests we should be very cautious in interpreting and relying too heavily on forecasts of rare events. The models used to produce such forecasts are necessarily based on limited historical data and are subject to significant uncertainty.”
Yet that’s precisely what Powell is doing now. There is no modern precedent for the kind of inflation he fears from tariffs. No historical model offers much guidance. And even Powell acknowledged Tuesday that “the effects on inflation could be short lived.” But rather than wait and see, he’s decided to let the forecast drive policy—and override the real-time data in front of him.
A Familiar Mistake: Transitory Then, Inflationary Now
In some ways, Powell’s decision to ignore current data in favor of tariff-driven inflation forecasts echoes a costly Fed error from the recent past. In 2021, the Fed insisted that inflation was “transitory,” even as prices surged month after month. Officials were guided not by what the data showed, but by what their models predicted—that supply chain pressures would ease and inflation would naturally subside. It didn’t. And the Fed was forced to scramble, hiking rates aggressively in 2022 and 2023 to restore credibility.
Now, the dynamic is reversed. Instead of ignoring inflation, the Fed is ignoring disinflation. Powell acknowledges that inflation is falling and the economy justifies cuts—but he’s pausing anyway, this time because models say Trump’s tariffs might reverse the trend. Once again, the Fed is trusting the forecast over the facts. And once again, it risks acting too late.
The Real Risk: A Fed That Doesn’t Trust the Economy
The irony here is that Powell claims the economy is strong enough to withstand high interest rates, but his actions suggest the Fed no longer trusts its own framework. The Fed’s core mission—price stability and maximum employment—is being reinterpreted as a mission to preempt possible tariff-driven inflation that may never arrive.
The risk, of course, is that by waiting for confirmation that things are not getting worse, the Fed may allow things to get worse. Credit-sensitive sectors—especially housing—are weakening. Homeowners are locked in place by high mortgage rates. And the longer the Fed waits, the more those pressures mount.
There’s no question that monetary policy must be forward-looking. But there’s a difference between anticipating future developments based on trends and refusing to act until models stop flashing yellow.
It’s also hard not to see this as evidence of a Fed ideologically opposed to Trump’s trade policy. Powell claims the Fed isn’t taking sides on trade policy, but the evidence for that is shaky. While it was comfortable predicting that a policy it seemed to like—Biden’s massive deficit spending on Green New Deal projects and the pursuit of progressive policy goals—was transitory, it is now convinced a policy it doesn’t like—tariffs—must have negative consequences.
Powell didn’t just say the Fed might cut later. He said the Fed should already have cut—if it weren’t for forecasts. That’s not caution. That’s paralysis.
And it’s a far cry from the Powell of 2018, who said central bankers must be humble in the face of uncertainty and guided above all by what the economy is actually doing.
Now, with the models in charge, the Fed is holding the line—not because of the data, but in spite of it.
Read the full article here