What Kevin Warsh Should Consider If He Wants to Lower Interest Rates

Kevin Warsh originally came to the Fed with a thin résumé that raised questions. This week, his first press conference as Chair is backed by a thicker résumé that raises even more questions. After leaving the Fed in 2011, Warsh built a public profile as a critic of Fed policy, calling it dangerously inflationary during a period now widely accepted as one in which the bank failed to create enough inflation—a position also  widely accepted by experts at the time. This is what I call the Past Kevin Warsh (PKW) era—a period where PKW argued for policy that was blandly too hawkish, doubling down on the dangers of inflation even as those claims became increasingly at odds with available data.

Sometime between the COVID-19 inflation wave and 2026, PKW turned into New Kevin Warsh (NKW), and NKW seems to hold almost polar opposite views on the macroeconomy. NKW argues that AI adoption could risk an upswell of productivity growth—such that not only could the Fed adopt more accommodative policy, but it may have to adopt anticipatory lower rates.

Whether Future Kevin Warsh (FKW) is more like PKW or NKW is an open question. What matters for interest rates policy is whether FKW can convince enough of his Federal Open Market Committee (FOMC) colleagues that these are his own views, and whether these views are correct. If Warsh hopes to credibly defend his aspirations to lower rates, a neutral interest rates framework—discussed further below—could offer a rationale if macro conditions change later this year.

Kevin Warsh and Neutral Interest Rates

An irony confronting the FOMC is that just because Warsh was wrong to be hawkish (i.e., particularly occupied with raising rates to control inflation) a decade ago, it wouldn’t make him wrong to be more dovish (i.e., laxer on rates) than the consensus now. Speculation about FKW’s decisions centers on AI, but that’s just one reason the Fed could be correct in setting future interest rates lower than they are today. In fact, the same dynamic that led PKW to be hawkish and incorrect could be the source of FKW being dovish and correct—conventional wisdom today is that the early 2010s were a unique time of low inflation and low interest rates that is behind us, but if inflation expectations were to return to their relatively low 2024 range, it would be foolish to dismiss a return to broadly lower interest rates.

To see why rates could fall, it’s useful to understand the neutral rate of interest, or r*.  For most of the late 20th century, the main challenge for researchers focused on Fed policy was inflation—the Fed sets interest rates in nominal terms, so economists decompose the Fed’s interest rate, the federal funds rate, as i = r + π, where i is the federal funds rate, π is the inflation rate, and r is the “real” interest rate. It’s the real rate, r,  that ultimately determines whether the Fed is stimulating the economy. The challenge researchers grappled with is that lenders’ profits on new loans, and thus how Fed policy affects macro conditions via bank lending, depend on the expected inflation rate in future years, not the actual rate last year. So if the Fed is trying to stimulate a depressed economy or slow down inflation in an overheating one, it needs to set interest rates that take expected inflation into account so that the real interest rate, r, can steer the economy.

In the 21st century, with expected inflation much lower and much more stable, the Fed should have been able to essentially set a rate of r + 2% to balance the economy. But after a very mild recession in 2001, it took the Fed five years to get rates back to the 5 percent level that had been so stable in the 1990s—and that only lasted a year. This is obviously too simple; external events change the economy all the time, and the interest rates are affected by more than just inflation or recession. To understand why the same Fed rate can be a recipe for prolonged stagnation at one moment but accelerate inflation in another, a digression on Taylor Rules is worth a moment.

The general idea of a Taylor Rule is that the Fed can set an inflation target π* and an unemployment target u*, and choose r by taking a weighted average of how far actual inflation and unemployment are from these targets. Under any Taylor Rule—there are many options to choose from—there is a real interest rate, r*,  that balances this equation when inflation and unemployment are at their target levels. This r* is called a neutral interest rate. But just because r* balances the Fed’s inflation and employment goals doesn’t mean it’s constant—if people suddenly shifted to saving much more money, r* would fall because the supply of money to lend would go up and interest rates would fall to balance the market.

In other words, the Fed is able to adjust interest rates to steer the economy, but because there are larger forces that affect interest rates, the neutral rate of interest r* varies over time and the real interest rate the Fed chooses is contractionary or stimulative relative to that neutral value. While r* is abstract and sounds disconnected from real-world policy, this has been far from true for most of the 21st century; in part because its persistent decline was underappreciated for years, in nearly every advanced economy. Inflation hawks, Past Kevin Warsh (PKW) included, failed to take account of this trend and consistently recommended policies that would have been catastrophic if implemented. And because fiscal policy failed to account for this shift and pushed these economies into counterproductive austerity, the actual level of r* became far from an academic question.

A Stronger Argument For the New Kevin Warsh View

Since the pandemic and the inflation that followed, r* has returned to the more academic topic it once was. Yet some estimates of its value suggest concerns about declining r* have been abandoned more rapidly than some experts suggest is warranted. Indeed, a new branch of research focuses less on measuring the descriptive trends affecting r* and more on identifying underlying macroeconomic trends to forecast future neutral interest rates. This research is still quite speculative, but offers a number of factors that could drive interest rates up or down over the near future, including demographics, tax policy, public borrowing, AI, and other important macroeconomic factors.

At the very least, a neutral interest rates framework is a more rigorous way to express the view that Warsh has advocated recently: If we are on the verge of an AI productivity boom that lets the Fed cut rates, the r* framework is the most consistent way to fit it into the FOMC context. Doing so might require accepting a bit of grief about having been wrong in the past and embracing the idea of falling neutral interest rates, but it is likely to be much more effective in securing the FOMC votes and backing Warsh would need to lower rates.

Of course, a more rigorous framework cannot compensate for a policy that is wrong on the merits, but if the US experiences less inflationary pressure later in the year, a neutral rates framework would offer a defensible rationale for adopting a more expansionary approach with a higher chance of success. Ultimately, Warsh’s push toward less communication and forward guidance is risky, and likely makes it harder for him to convince his colleagues and financial markets that lower rates are coming as part of a process of deliberate consideration.

If Future Kevin Warsh is going to lower rates without stoking concerns of political influence that could spike inflation, a critique of Past Kevin Warsh would be a useful first step.