Why an AI productivity boom could justify higher rates
Add Axios as your preferred source to
see more of our stories on Google.
Illustration: Brendan Lynch/Axios
Is today more like 1995, when a technology-driven productivity surge was underway but not yet fully acknowledged, or like 1999, when that boom was well-known, much-celebrated, and reflected in asset prices?
- The answer might determine what the Federal Reserve ought to do about interest rates in 2026.
The big picture: That's a new argument that Chicago Fed president Austan Goolsbee made this week, that the proper Fed response to a productivity surge depends on whether it's a surprise or widely known and expected to continue.
- If the latter — if this is 1999 — then the Fed should be on high alert for inflation and be more open to raising interest rates.
- If it's like 1995, there is more scope for lower rates as the economy enjoys the disinflationary impact of a supply-side boom.
Between the lines: Kevin Warsh, the Fed chair designee, has argued that the AI boom's supply-side benefits justify keeping interest rates low, much as Alan Greenspan did in the 1990s information technology boom.
- Goolsbee's argument is a preview of the pushback that Warsh will receive on the policy committee he'll soon lead as Fed chair — particularly given the resemblances between 2026 and 1999.
- If anything, the potential for AI to drive a productivity surge is more widely taken for granted among business decision-makers — and reflected in investments and asset prices — than the 1990s tech boom ever was.
Flashback: In the mid-1990s, Greenspan resisted calls for higher interest rates because he detected — long before it was conventional wisdom — that companies were seeing rapidly improving productivity that hadn't fully filtered through the overall data.
- That implied that the Fed need not pump the brakes on the economy with higher interest rates just because growth was robust and the labor market strong. Improvements in the economy's supply potential meant that growth was non-inflationary.
- By the late 1990s, however, the dot-com boom was near its peak, the productivity boom was well-established in the macroeconomic data, and companies were betting on endless, non-inflationary growth.
- As a result, business investment boomed and the stock market soared. Against that backdrop, the Greenspan Fed raised interest rates in 1999.
Zoom in: If investors and business leaders believe that a productivity boom is imminent or already underway, they bid up asset prices and ramp up investment spending. That creates inflationary pressures in the here and now.
- When a data center project bids up demand for HVAC contractors or construction equipment, or an investor seeing gains on their Nvidia stock buys a new boat, expectations of future productivity are being pulled forward in the form of greater demand today.
- Goolsbee's argument is that the Fed, in that scenario, needs to counteract that impulse with higher rates.
- That theory suggests higher productivity growth means a higher neutral interest rate. Goolsbee is essentially describing the mechanism through which that becomes true.
What they're saying: "If people start changing their behavior today in the expectation that there will be large increases in productivity ... you have to start thinking about the possibility that it overheats things in the short run," Goolsbee told reporters following a panel discussion at the Milken Institute Global Conference.
- "If you start to see a lot of that counting-the-chickens type of behavior ... if the central bank doesn't actually have higher rates, inflation can rebound, and instead of inflation going down, inflation would go up and rates would have to be even higher."
