Fed Official: Inflation Hasn't Left, and Cuts Can't Bet on ‘Transitory’
Goolsbee is worried the economy is overheating: the disinflation so far has come mostly from supply healing, and services still show no progress; cutting below neutral has to wait for evidence rather than run ahead of it.
The video won't play here. Listen to the audio instead:
The argument · tap a timestamp to hear it
With inflation stuck above 3%, the policy rate isn't tight
Asked whether current policy is restrictive, Goolsbee refused to answer directly. What matters, he said, is the real rate — the nominal rate minus either expected inflation or actual inflation. The long-run landing spot he gives is a 3% nominal rate, 2% inflation, a 1% real rate. If inflation is running above 3%, the real rate is far lower than it looks on the surface. He is willing to wait, but he is nervous about how inflation has behaved over the past six months. That opening sets up the judgment that runs through the whole episode: you cannot talk about how restrictive policy is in isolation from inflation.
— Austan GoolsbeeSaying ‘three more months’ every quarter is not transitory
Goolsbee breaks current inflation into pieces. Tariffs, and a war pushing up oil prices, are one-off shocks to the price level, and in theory they fade — but he points to the COVID experience as a caution: a supply shock large enough can drag on far longer than the initial forecast. What genuinely worries him is services inflation, which is caused neither by tariffs nor by oil, and therefore belongs to a deeper problem. His test is this: you cannot say every quarter that ‘it'll be fine in another three months’; he wants to see evidence that it is actually fading.
— Austan GoolsbeeBidding away electricians isn't overheating; wage spillover is
On the AI data center build-out, Goolsbee says what he hears from Midwestern businesses is complaints: the data center buyers are buying up all the land and pushing prices up, and you cannot find electricians or HVAC workers. But he stresses that competition for resources at the sector level is not the same as the whole economy overheating. Only when AI investment pushes wages and prices beyond its own lane — lifting indicators beyond national unemployment and GDP — does it amount to excess aggregate demand in the traditional sense. If it ever got to that point, then the current policy setting would not be restrictive enough.
— Austan GoolsbeeTwo-thirds of this inflation cycle came from supply, not demand
Looking back at the aggressive hikes of 2022-23, Goolsbee offers a split: roughly two-thirds of both the run-up and the decline in this inflation cycle came from supply, one-third from demand. He thinks the Fed's contribution was keeping ‘the other shoe’ from dropping: TIPS inflation compensation held steady at 2.3% even as CPI approached 10%, which shows the 2% target anchor was working. He concedes the Fed was slow off the mark, but he insists you cannot have it both ways — you cannot blame the run-up on stimulus policy and then credit the decline entirely to supply chains healing.
— Austan GoolsbeeA 5.25% long-end yield is not a panic about U.S. credit
On rising long-dated Treasury yields, Goolsbee says it takes time to tell the drivers apart: it could be inflation expectations, it could be the market expecting the Fed to hold rates high, or it could be increased supply of issuance. He does not accept the reading that ‘the market is panicking about U.S. credit’, because if default were the real worry, a 5.25% yield is merely a normal level by historical standards. He quotes Volcker: the Fed acts and the market reacts, and the order cannot be reversed. You can gather information from market signals, but the market should not be the one telling the Fed what to do.
— Austan GoolsbeeThe dot plot isn't a reaction function; its medians aren't one person's
Goolsbee separates forward guidance from the reaction function. He objects to forward guidance of the ‘we promise to hike or cut at some future meeting’ variety, which he thinks ties your hands and adds volatility. But the market does need to understand how the Fed reads economic data, and that is the reaction function. The problem with the SEP and the dot plot is that the median inflation projection and the median rate projection do not necessarily come from the same person, so the dot plot cannot really serve as a reaction function. He says he has opposed the SEP for years, because members write down forecasts and then get proven wrong, which damages credibility.
— Austan GoolsbeeWithout evidence inflation is falling, he won't cut ahead of it
He lays out his own reaction function: he is paying close attention to the inflation side. If he sees inflation genuinely coming down, headed back toward 2%, he is willing to return to the ‘3-2-1’ path; if inflation — services inflation especially — is still rising with no progress, he gets nervous. He explains his dissent late last year: the government was shut down and data was missing, and he was not comfortable front-loading a cut in the absence of evidence, unwilling to act on the assumption that inflation would go away by itself. That explains why he differs from the current majority of the committee.
— Austan GoolsbeePowell only moved a few seats over, but the FOMC has shifted gears
Asked how an FOMC chaired by Warsh differs from the Powell era, Goolsbee says it feels very different personally. The new chair's bearing at press conferences, his worldview, his willingness to re-examine the communication tools — all of it is different; Powell is still on the committee, he has just moved a few seats over. He will not comment on the chair's personal reaction function, saying only that everything will have to wait for the verbatim transcripts four years and nine months from now. For listeners, this says the Fed is internally in a period of shifting gears, and the market's guessing about the reaction function will continue.
— Austan GoolsbeeIn their own words · checked verbatim
I always called our star our Sasquatch and you know somehow it feels with this as a backdrop because you can never see it until after it was left
Austan Goolsbee3:06
It has to get out of its lane of just direct competition and drive up wages, drive up prices outside of just its lane.
Austan Goolsbee11:17
I wasn't there when it went up, so you can't blame me.
Austan Goolsbee18:28
Paul Volcker used to tell me, our job is to act and the market's job is to react and let's not get the order mixed up.
Austan Goolsbee22:36
The reason I dissented in the last meeting of the year when they cut rates was I'm not comfortable front loading the rate cuts, counting on this inflation to be transitory and go away.
Austan Goolsbee33:54
If inflation is going to be persistent, then it's going to force action by the Fed or by any center. If it's going away on its own, or if you even feel like it could go away on its own, then. It puts us in a very different circumstance.
Austan Goolsbee41:11
Figures
| Goolsbee's long-run neutral path | 3% nominal rate, 2% inflation, 1% real rate | 2:03 |
| Productivity growth over the past six months | six consecutive poor months | 15:24 |
| Attribution of the inflation run-up and decline | roughly two-thirds from supply, one-third from demand | 18:28 |
| Historical relationship between CPI and PCE | CPI of 2.3% corresponds to roughly 2.0% PCE | 18:28 |
| TIPS inflation compensation at the inflation peak | steady at 2.3% (CPI basis) | 19:30 |
| What the Taylor rule recommended at the zero bound | negative 6% | 44:11 |
Glossary
- R-star / neutral rate
- The real neutral rate that in theory measures whether monetary policy is restrictive; it cannot be observed directly.
- SEP / Summary of Economic Projections
- The quarterly compilation of Fed members' projections for inflation, unemployment and interest rates; the dot plot is one part of it.
- Forward guidance
- Central bank communication that commits in advance to a future rate path; Goolsbee thinks it was necessary in the zero-rate era and should be used less now.
- Reaction function
- The rule for how a central bank responds to economic data; Goolsbee stresses it is not the same thing as a commitment to a specific rate.
- Taylor rule
- A rule of thumb that maps the inflation gap and the output gap into a recommended interest rate; at the zero bound it once produced a negative number.
How to listen
Traders who follow the Fed's policy path, macro hedge fund researchers, and private-market investors trying to judge whether AI capex creates overheating risk.
The theoretical discussion of R-star at the top can be fast-forwarded; start where he breaks inflation apart and gets to the AI spillover.