Kevin Warsh spent years building a brand on one argument: the Fed kept rates too low for too long and caused the worst inflation spike in forty years. He was right. Then, sometime between that diagnosis and his January 30 nomination, he decided the cure is more of the same disease, just dressed up in different clothes. A man who blamed cheap money for breaking the economy now wants cheaper money, justified by productivity gains from artificial intelligence that have not shown up in a single inflation print. That is not intellectual evolution. That is a job application.
The Senate Banking Committee should reject this nomination. Not because Warsh lacks credentials, but because his recent pivot tells you everything about how he would run the Fed: eyes on the White House, not the data.
The AI Alibi
Warsh's argument goes like this: AI will make workers so productive that the economy can grow faster without generating price pressure, so the Fed can safely cut borrowing costs. Many Fed officials disagree. They should. The productivity data does not support the claim yet.
Total factor productivity growth in the U.S. has been sluggish for over a decade. AI investment is enormous and accelerating, yes. But productivity gains from major technological shifts historically take seven to fifteen years to show up in macroeconomic data. Electricity was invented in the 1880s; it didn't reshape factory output until the 1920s. The internet went mainstream in 1995; broad productivity gains didn't materialize until the early 2000s, and they faded within a few years.
Warsh is not saying the Fed should monitor AI productivity and adjust accordingly. He is saying AI productivity should justify rate cuts now. Before the evidence arrives. That is a conclusion looking for a reason, and the reason happens to be the one his nominator wants to hear.
To be fair, Warsh's original critique of the Fed's pandemic-era policy was sharp and largely correct. That credibility is precisely what makes his reversal dangerous. A nominee with no record would be easier to evaluate. Warsh has a record that says one thing and a mouth that now says the opposite.
The Independence Test He Already Failed
The value of a Fed chair is measured by one thing above all others: willingness to make the unpopular call. Volcker raised rates into a recession. Greenspan raised them in 1994 when markets screamed. Even Powell, for all his flaws, hiked aggressively into 2023 while the White House wanted relief.
Warsh has not even been confirmed yet and he is already aligning his stated views with presidential preference. Trump wants lower rates. Warsh now wants lower rates. The justification is AI, a theory that conveniently provides intellectual cover for doing what the boss wants.
Consider what this means for a family carrying a variable-rate mortgage or a small business owner rolling over a line of credit. If the next Fed chair cuts prematurely because he promised the White House easy money, and inflation re-accelerates, those borrowers get crushed on both ends: higher prices and, eventually, even higher rates when the Fed has to reverse course. The 2021-2022 whiplash Warsh himself criticized would repeat, except this time he would be the architect.
The confirmation itself is in trouble. Senator Tillis's refusal to vote yes until the DOJ resolves its investigation into Powell creates a narrow path that could leave the Fed leaderless past May. That procedural mess is real, but it is a sideshow. The substantive problem is bigger. Even if Warsh gets through, the market will know it confirmed a chair who pre-committed to the president's preferred policy before taking the oath.
Central bank independence is not a slogan. It is the mechanism that keeps inflation expectations anchored. Once markets believe the chair takes direction from the Oval Office, long-term rates move on political signals, not economic ones. That repricing is brutal and fast.
Warsh diagnosed what went wrong with the last era of cheap money. Now he wants to prescribe the same treatment with a different label on the bottle. The Senate should read the ingredients.
