Chicago Fed President Austan Goolsbee on Vibes vs Data

Chicago Fed President Austan Goolsbee on Vibes vs Data

Sentiment, prices, the job market, and the AI buildout

kyla scanlon

Oct 15, 2026

148

17

14

Share

Good morning from an airplane! Yesterday I

published a piece

on how AI, oil, and Treasuries are impacting the economy. Today, an interview with Chicago Fed President Goolsbee.

This is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

A year ago, I had the honor to

interview

Chicago Fed President Austan Goolsbee. At the end of our conversation, he warned that we might be headed back into a conflict between the data and vibes. I sat down with him, a year later, in Chicago to talk about how that conflict is going, and where things are at for how people

feel

about their economic circumstances.

We recorded this around an event I moderated at the Chicago Fed with NORC at the University of Chicago, called

“Why Consumers and Economists See Different Economies.”

So this piece is the interview with President Goolsbee + some of what I heard from our panelists on stage. With Goolsbee, we get into why consumer sentiment and consumer spending are so disconnected, what he prioritizes, why people feel so bad, where the job market is at, and if the AI buildout is starting to heat things up.

This transcript has been lightly edited for clarity and flow. Below is the video version!

Recorded on

September 30th, 2026

Data vs Vibes

Kyla:

Last time that we spoke, you said that there was going to be a fight between data and vibes. Where are we in that fight now?

PG:

The family fight continues. It never stopped. The argument began at Thanksgiving and it extended to Christmas and into the New Year. I feel like mostly the data won at the Fed, in the sense that we set interest rate policy over the intervening year, or year and a half, or however long it was, based on the actual numbers. And the vibes never came back. The numbers continued to be stable and decent, and the vibes continued to be horrible. I think that’s where we are.

Kyla:

One thing that you talked about last time is that “data dogs sniff out data” and there is clearly this disconnect between sentiment and data. When you look at things like spending, do you feel like that matches up to sentiment, or are you surprised by how the consumer is acting relative to how they feel?

PG:

I’ve been a little surprised by how the consumer’s acted compared to how they feel. As you know, we’ve got some top researchers here at the Chicago Fed looking at exactly that. That relationship didn’t just break down recently. It’s been getting a little weaker, a little weaker over time, and then recently the correlation between sentiment and spending has been very, very weak.

Vibes are important for a lot of things. But if you were using the vibes as a predictive variable for consumer spending, which was historically kind of how the Fed was using it, it makes you a little less [certain]. You’re never going to hear me say throw any data away. But it moves it down on my priority list.

Chart:

Reassessing the Relationship Between Consumer Sentiment and Spending with a New Composite Index

The List

Kyla:

What’s at the top of your priority list?

PG:

Still the big ones on the job market. I like the rate-based measures, not the monthly payroll, because you’ve got immigration and you’ve got a bunch of question marks. So the unemployment rate, the vacancy rate, the layoff rate, and the hiring rate. Those are one side.

And then on the consumer, retail spending. Using the actual hard numbers to make forecasts, I think, gives you the most accurate forecasts you can get. There’s always a lot of variability when it comes to forecasts, but the incremental addition that comes from adding sentiment to the hard numbers has gone down. Right now, it doesn’t seem to be that high.

Can Bad Vibes Become Self-Fulfilling?

Kyla:

I think one of the things that people worry about is that sentiment will become predictive, right? That the economy will match the bad feelings. How do you feel about that ending?

PG:

Look, it’s dramatic. Consumer sentiment is dramatic, and it moves around a lot, in a way that some other variables, which are kind of smooth, don’t. We love a good change story. And historically, big drops in consumer sentiment were indicators that a recession is beginning. I think that’s why people are as paranoid as they are to see these poor showings in sentiment. They’re like, “Yikes, when we see numbers like that, a lot of times that means recession.”

But now we’ve been several years without the recession that was kind of forecast from that. So I think mostly that one has resolved in favor of the hard numbers.

Let’s do some more exploration of consumer sentiment.

If it’s not a great indicator of how the economy is doing or about to do, maybe it is some kind of an indicator of consumer expectations

. And that does matter a lot. We’re constantly trying to figure out forward-looking measures, and there aren’t that many that come from the actual economic decision-makers themselves. Maybe sentiment can fit in there. So we’re sniffing around in that space. I think that might be more promising than going back and refighting the question of whether sentiment is a better measure of the economy than the measures of the economy are. I think it’s not going to prove to be that.

Why People Feel So Bad

Kyla:

A lot of people have come out with research. Jared Bernstein, former chair of the Biden Council of Economic Advisers,

points

to price level as a driving factor of the poor consumer sentiment. Other people point to social media. Other people point to structural affordability: housing, childcare, eldercare. Where do you sit? Why do people feel so bad?

PG:

I think there is a component of salience.

We’ve always known that the price of gasoline is a very public price, and if you’re driving down the street and you’re seeing diesel is $6 a gallon, people are going to say they think the economy stinks. And we know that there are public prices at the grocery store that people are going to respond to. So I think part of a higher inflation environment is going to show up as poor consumer sentiment.

You’ve seen the partisanization of everything. In a way, the sentiment numbers might be telling you a little more, like they’re a replacement for political polls, now that political polls are going away. The day of the election, even purely factual questions like

“Has the unemployment rate gone down over the last year?”

completely switch based on the politics of the responder and the politics of the president. I think we’ve gone through an unusual period.

We’ve gone through some periods where real incomes went down, and whenever real income goes down, people get sour for a perfectly good reason. We’ve had episodes like that. And then we get into a thing of, well, the prices went up, but the incomes went up almost as much, and if you count the tax cuts that you got… eh.

I think when people see prices going up faster than incomes, it’s going to blow back.

Kyla:

I think a lot of people feel like they’ve never gotten a break, right? It’s just been one hit after another since COVID. What would you tell people who feel that way?

PG:

Well, they’re kind of right. If you’re filling up the tank of your car and you’re paying $70, $80, $90 dollars, you can see why people do not like that kind of environment. You were identifying the vibecession before anybody was talking about that. So I’m going to turn it on you. There are these salient factors that we’ve known about. But as the queen of the vibes, do you think that something else is going on?

Kyla:

In terms of why people feel so bad? I think it’s a lot of stuff. I think social media—

PG:

You think it matters?

Kyla:

I think it does. Think about gas prices, right? You’re on every corner and you see gas prices. Social media is that, but amplified across the entire economy.

PG:

I see. For many other prices. That’s an interesting observation, that the publicizing of the price is what makes people more attuned to the cost of living. Maybe they’re not necessarily computing a Laspeyres or Paasche price index. They’re just seeing the price of all the stuff going up. That’s probably true.

Kyla:

And I think it’s complicated even further by the introduction of AI. I just feel like people, everywhere they turn, don’t really know who’s fighting for them. And it’s hard to communicate and get that through the algorithm.

PG:

I was a professor for 30 years down at the University of Chicago, and at the end of the semester, they would hand out teaching evaluations to the MBA students. There were all these categories. Was the professor organized? Was the professor clear? Were they interesting? And if the students decided that they liked you, you would do well in every category, even if you didn’t deserve it.

The first class I taught, I was not organized. I was one week ahead of the class. I’m scrambling every time. And I had a colleague who taught the same class every year for 30 years, and he knew that at 12:20, we would be doing this problem set. Very prepared. Handed it out on the first day. So objectively, massively more organized, but he was a more grumpy guy. And so in the evaluations, I would be more organized than this guy, because they just decided. They felt better about it.

I do wonder about that sometimes in these vibe and sentiment measures. They usually forecast something like stagflation. When consumer sentiment is low, their expectations about inflation are high, and their expectations of what’s about to happen in the economy are bad. Now, that might mean that they do expect a supply shock to hit. But it might mean that they don’t like the professor and they’re going to give him a bad grade in every category. And I can’t tell.

Averages are Not the Economy

Kyla:

Joanne Hsu, who runs the UMich survey, has talked about how averages can obscure what’s happening under the hood. Especially in this current economy, where higher-income spending is driving a lot of growth, and with the stock market, where people are benefiting from asset ownership versus labor income. What do you think about that?

PG:

It’s an interesting idea. Take it to the extreme, because sometimes the extremes are the most helpful. Let’s say one-tenth of 1% of people had all the income gains and 99.9% of people’s incomes went down. Then any survey-based measure is not going to look good, because it’s population-weighted, not income-weighted.

My only thing is, the

New York Fed

puts out some statistics about incomes and consumption that break it out by income level. There have been differences, but the K-shaped economy hasn’t been as extreme as we might have thought it would be, at least in the data. With the stock market performance of the last four or five years, and stock ownership being pretty concentrated, you would have thought this massive increase in wealth would have accumulated to a very small number of people, and so you would be seeing a pretty extreme divergence in the growth rates of consumption. At least in the New York Fed data, it’s there, but it’s like a one percentage point difference between the high-income and the lower-income groups.

So this is a fruitful area we should sniff around. Joanne Hsu, I have tons of respect for. She’s on our academic advisory panel here at the Chicago Fed. If she says it’s a thing, I bet it is a thing, and we’re going to find that out.

Chart:

New York Fed

Low Hire, Low Fire, Year Two

Kyla:

When you’re looking at the labor market, a year ago you said low hire, low fire. Is that still the case?

PG:

Yeah, it’s still the case. It’s weird to be low hiring and low firing at the same time. That’s odd. I thought that was from uncertainty. I still think that, but if we go for five years, it can’t really be uncertainty anymore. You know what I mean? Eventually they’ve got to make some decisions.

Kyla

: You’ve got to be a little certain eventually

.

PG

: In the data, it’s still true that the hiring rate is as low as it normally is in a recession, and the layoff rate is as low as it normally is in a boom. And both of them are stable. So where does that leave you? I still think it mostly points to stability in the job market, not a deteriorating job market.

If anything, when we’re out talking to businesspeople, and I just got back from downstate Illinois, they’re highlighting that maybe the job market is getting tighter, not looser. Especially for skilled labor, partly from competition from the data center build-outs, it’s getting harder for them to find people, and they’re having to pay higher wages than they had allotted even six months ago.

AI, Data Centers, and Overheating

Kyla:

The Wall Street Journal had an interesting

article

about that, how much the data centers were adding to local economies in a surprising way.

PG:

Right. I’m of two minds about it. A, the bigger the hype about the future, the more incentive they have today to build out the data centers and spend out of these valuations. You could easily overheat the economy and complicate the day job of the central bank.

Kyla:

Is that happening?

PG:

Maybe.

And then B is whether that thing jumps out of its lane. To the extent that an AI data center going up comes at the expense of something else, the overall aggregate is not an overheated economy. It involves a change in relative position, and it’s going to make everybody who’s getting displaced upset. But the overall temperature isn’t overheated. To the extent that it overheats and jumps out of just its AI and direct AI-affected lane, that’s old-fashioned demand management for the central bank. That’s not a supply shock. That’s just a demand shock.

And the theory is pretty clear on what the central bank has to do if the economy starts growing faster than potential and that’s contributing to inflation.

Kyla:

What do you make of the AI build-out as the technology has evolved, and as these companies are entering the bond market and using off-balance-sheet leases?

PG:

Look, it comes with a component of concern, anxiety, something. Is that a bubble? It depends how you define bubbles, but they’re moving into the highly speculative, counting on really substantial GDP and productivity increases. I certainly hope that’s true. But if instead of growing like [points up], it’s growing like [flatter line], that’s what we saw in 1999, 2000, 2001. That can have some pretty profound implications. The way I’d phrase it is, nobody spends a trillion and a half dollars thinking they’re going to give the product away free. So either it’s going to be really expensive, or the valuations are going to take a hit, in a fundamental way.

What He’d Want You to Hear

Kyla:

This will be my final question. What would you want someone who feels bad about the economy right now to hear from you?

PG:

Oof. Look, it’s presumptuous of me. I’m not in the PR business. I want to hear from them the things that are making them feel bad. How are they progressing? Are they stable? Are they getting worse? I’ve tried to go out throughout the Seventh District, and I’m not surprised at all that the data on sentiment are as negative as they are, because that’s what I hear when I’m out. And the overwhelming concern is less about growth and is all about prices and costs.

So if they were going to hear something from me, it would just be: to the extent that inflation is driving that, we here at the central bank are committed. We’re going to get the inflation rate back to 2%, even though it’s been a long time that we’ve been above it. Getting price stability and getting prices under control, I do think, is a big component of why people are feeling bad.

Kyla:

Thank you.

PG:

It’s great to see you again, Kyla.

Six Takeaways

Sentiment stopped predicting spending:

Last year President Goolsbee warned we might be headed back into a conflict between the data and the vibes. For decades, the Fed watched consumer sentiment because it was a good read on consumer spending — so if people felt bad, they usually spent less. That link has faded. But that doesn’t mean that feelings don’t matter! He’s saying that for the job that the Fed used them for, the hard numbers are more useful for forecasting. And he also said that the Fed is “sniffing around” whether sentiment works as a measure of expectations instead.

Of course people feel bad:

His three reasons were salient prices (gas, groceries), partisanship, and periods where real incomes fell. He didn’t argue anyone out of feeling bad. The Michigan data backs this up — in August, 53% of all households said higher prices are a reason they’re worse off than a year ago. In 2019, that averaged 7%, and in 1980, during the Great Inflation, it was 41% (

Michigan Table 7

). CPI Inflation

is 3.7%

but the price level is about 30% above January 2020, and that’s what people are responding to. Brian Fabes’s

slide

said it perfectly: “People live in levels, not rates.” Joanne Hsu

showed

that people also report hearing far more bad news about prices than they did in the 1970s and 1980s, and walked through how the information environment changed, from evening news to cable to algorithmic feeds. When I asked her if the online world is part of this, she was careful not to draw a causal line.

One person, one vote vs. one dollar, one vote:

Hsu said this during the event and it’s such a good way to think about the disconnect. Sentiment counts each person once, whereas GDP, the CPI, and the stock market all count dollars. Goolsbee made the same point by saying that if a tiny share of people do get all the gains, any survey is going to look bad while the economy grows. And there is a little bit of a K-shape in spending —

since

the start of 2023,

real retail spending

is up about 7.6% for households over $125,000 and about 1.3% for households making under $40,000, with the lower income group back to flat by

August

.

The economy is local:

Joanne Hsu summarized this really well — people don’t judge the economy by GDP or labor force participation rates. They judge it by their friends and families. Hsu talks about how the survey hears these concerns, that a lot of people will say “I’m doing fine financially, but I’m really worried about my child who is unable to move out of my house.”

Low hire, low fire labor market:

a year ago, he described hiring as low as a recession and layoffs as low as a boom and blamed uncertainty. The

hiring rate

is 3.3% (the average in the 2001 and 2007-2009 recessions was 3.5%) and the

layoff

rate is 1.0% (the expansion average is 1.3%) and neither has really moved over the past two years. But as President Goolsbee says, after five years, this can’t just be general uncertainty anymore. That’s why Goolsbee looks at rates rather than the monthly payrolls, because payrolls can only tell you something if you know how the population is growing, and immigration has made that a question mark.

The Fed has tools:

AI is arriving at the Fed as a demand story (construction, wages, “maybe” overheating) before it arrives as a productivity story, and he’s clear that if it does spill out of its own lane and start overheating the broader economy, that’s an ordinary demand shock and the Fed knows what to do with those. He also points out that no one spends $1.5 trillion planning to give away a product for free.

The bigger picture here is that the fight between data and vibes isn’t really a fight about who is right anymore. The hard numbers are better at forecasting spending, but sentiment is still an enormous amount of information about (1) where people are relative to where they expected to be and (2) who the averages miss and (3) if people think that anyone is fighting for them. Fabes’s summary slide said it pretty plainly: “People don’t feel the economy is working. This is reality.” What we do about that is the question. Talking about sentiment and the importance of economic feelings is one step in the right direction.

This is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

Thanks for reading.

148

17

14

Share

Previous