Chicago Booth hosted “The Signal and the Noise: What Lies Ahead for the Global Economy” in Chicago in January as part of its 2024 lineup of Economic Outlook events. In a discussion moderated by journalist Kathleen Hays, Chicago Booth’s Randall S. Kroszner, Yueran Ma, and Raghuram G. Rajan analyzed the factors that will help shape the year ahead. The following is an edited and abridged excerpt.

Kathleen Hays: Raghu, your latest book is Monetary Policy and its Unintended Consequences, and one of the things you gently touch on is politics. There does seem to be political pressure on the US Federal Reserve, in a couple of different ways. It’s not direct pressure and it’s not strictly partisan. This seems more subtle and I think more dangerous for the Fed.

Raghuram G. Rajan: This is an election year, so it’s on everybody’s mind: What is the Fed thinking? Now, officially they’re not thinking about it, but the Fed certainly doesn’t want to engineer a recession, and obviously the Democratic members of the Fed are even less likely to want to engineer that. But how do you achieve that soft landing, and this is the problem that [Federal Reserve chair Jerome] Powell faces every time he says, “OK, well maybe we don’t have to be so tough. Maybe we will start cutting rates as inflation comes down, because we don’t want to be as constraining on the economy. We are looking at some level of real rates. As inflation comes down, we may even be willing to cut.” Markets are desperately trying to anticipate this. I mean, there’s a huge fear of missing out in the markets. Participants want to be first off the starting blocks.

And this is what you see again and again, Powell trying to reassure that the Fed is not going to overtighten, and immediately that sets off a race. And then all the Fed presidents come out and say, “No, no, no, we didn’t mean that. We meant we’ll wait and see.” And can you have it both ways?

In a sense, the Fed is also fighting the markets because the markets aren’t making the Fed’s task easier. What would be nice is if the Fed said, “Don’t worry too much. We are going to get there, but don’t react to it. Don’t react. Don’t push bond yields down. Don’t push the stock market up.” And if that happens, yes, we’ll get there; but if the markets start celebrating, it’ll take the Fed longer to get there. And that’s the problem. Any market anticipation of Fed cuts will have the very effect the Fed is trying to avoid, which is basically a far sharper slowdown than it wants.

Hays: What are some other unintended consequences of monetary policy?

Randall S. Kroszner: Financial instability is a classic one. There used to be this theory that was bandied about, a so-called separation theorem, that you’d have monetary policy do its thing to fight inflation or to make sure inflation stayed low, and then macroprudential policy and regulatory policy would do their thing to make sure that the financial system was stable. Silicon Valley Bank made it really clear you can’t separate these two issues.

When things are going along the same way they’ve always been, it’s relatively easy to manage. When something’s moving and changing very rapidly, it’s much more challenging to manage that. You need to be creative in thinking about the what-ifs, thinking about the alternatives, doing stress testing. We didn’t do that in the United States. In the United Kingdom, the Bank of England’s stress scenario that they ran last year was: What if interest rates do something crazy, like go to 5 or 6 percent? They tried to look through the consequences of that, which was a very smart thing to do. Did they get everything perfect? Of course not. No one got everything perfect, but it was useful to consider that scenario.

You’re always going to have management challenges in the markets when something is moving quickly. That’s a key lesson for supervisors to have, regardless of what the regulation is. If a market or a product grows very rapidly or changes very fast, there’s going to be a lot more uncertainty around it. There’s going to be a much bigger challenge for senior executives to manage risk. Hence, in those highly dynamic circumstances, you’re much more likely to see mistakes and mismanagement.

“VC investment has declined by 60 percent since the rate hikes started. . . . That has really affected most sectors of the startup space.”

— Yueran Ma

Hays: Yueran, you have done some innovative research on this whole question of the unintended consequences of monetary policy that gets away from the banks and gets right to innovation.

Yueran Ma: Fifty-five years ago, Milton Friedman made one of the most influential presidential speeches at the American Economic Association about monetary policy. And he was saying that monetary policy affects the real economy, but the effects should be in the near term. Monetary policy will not affect the longer-term productive capacity of the economy. This past summer, I was asked to speak at Jackson Hole, and for that I looked at the effects of monetary policy on innovation, which is one of the potential ways for monetary policy to have longer-term effects.

Innovation, like any investment, can be affected by economic conditions. When the Fed tightens, for example, demand slows down and financial conditions tighten. These things, we already know, tend to slow down innovation, and that’s actually the case in the data. When monetary policy tightens, you see a slowdown in innovation expenditures such as spending on research and development and venture-capital investment, and then later a slowdown in innovation output—for example, as measured by patents. And then that may have longer-term real effects on the economy.

And the magnitudes are actually quite large, even historically. Most recently we’ve seen, as an example, VC investment has declined by 60 percent since the rate hikes started. Of course, there might have been some froth as of 2021 and correction after that, but that has really affected most sectors of the startup space. And to think that has no real effects seems not so plausible.

But what’s really interesting here is that these real effects will take a while to play out. It takes about a year or two for monetary policy to be fully transmitted into innovation expenditures, another two to three years then to show up in innovation output, and then another five years to be reflected in aggregate output. It’s possible that what we’re seeing now is not just transitory or just about today; it may have longer-term effects.

Hays: What are you paying attention to with artificial intelligence, and what should we be aware of as we watch its progress, including from the perspective of its impact on the economy, the job market, and the country?

Ma: Let’s go back 150 years to the second industrial revolution, which had a profound impact on our world. At the time, what were people saying about industrialization? They were talking about technical unemployment, the impending rise of large firms that will dominate society, and the end of capitalism. Now if you look at what people say about AI, they talk about technical unemployment, the dominance of Big Tech and big firms, and the potential end of capitalism. [OpenAI CEO] Sam Altman said in an interview that AI might break capitalism.

Well, technical unemployment probably has happened less than we thought. We found new things to do. In fact, industrialization enabled the migration from manufacturing to services. Probably we’ll find more work to do for ourselves.

The dominance of large firms actually has happened. I looked at the data on the size distribution of all US companies for the past 100 years, and it turns out that the dominance of the largest firms has been a perennial trend for the past century. Fifty years ago, 50 percent of US output came from about 10,000 companies. Now 50 percent of US output comes from fewer than 2,000 companies.

Lenin was a strong believer in the inevitability of large companies. He did some calculation for the 1910s that found that the largest 1 percent of enterprises accounted for 50 percent of the output at the time. Now if you do that calculation, the largest 1 percent of enterprises account for about 80 percent of the output. We’ve come a long way.

That is a trend that dominated the past century. Some people think that AI will be able to capture more implicit knowledge and enable large-scale production to an even greater extent.

Now, come to the extension of that, the end of capitalism: Why might these things end capitalism? One concern was that the dominance of large firms would lead to the concentration of power and wealth. The Communists said that it would lead to the inevitability of planning because planning would become so efficient when you had all these industrial technologies. That has not fully materialized. There’s still a role for the invisible hand, as Hayek said.

“The Chinese engine of growth, which was real estate, is dead in the water. And they’ve tried to refloat it; it’s not happening yet.”

— Raghuram G. Rajan

Rajan: Distinguishing the hype from reality is tough in these cases. I mean, if you look at what’s actually happened, it’s not much. We’ve all tried ChatGPT and then said, “OK, interesting.” It’s going to be an aid for some time. Some companies are using it already for things like consumer services, but really for helping employees rather than actually displacing them.

Will displacement occur? Of course there will be some displacement. Every technology does that, but usually, as Yueran put it, much slower than we think. It takes time for companies to figure out how to use it well, to trust the technology, and so on.

Some jobs will be displaced, but there’ll be a whole set of new jobs created. David Autor at MIT says that 60 percent of the jobs we have today didn’t exist in the 1940s. New jobs will be created, some jobs will be lost, and existing jobs will be augmented. This is going to happen. This happened with every technology; it will happen with this. What is the pace at which it’ll happen? Maybe it’s faster with AI than with some of the other technologies, but we don’t know.

One interesting point that the Economist has made is that the first automated telephone switching system was invented in the late 19th century, and telephone operator lasted as a job into the 1980s. That’s just an example of how technology takes time. I don’t think AI will create mass displacement in the next five to 10 years. After that, we’ll see.

Hays: Looking at the global economy, Raghu, what do you see for China right now?

Rajan: The Chinese engine of growth, which was real estate, is dead in the water. And they’ve tried to refloat it; it’s not happening yet. Eventually things will start moving up again, but it’s taken a lot of time, and we saw with the real-estate sector in the US how difficult it is to get it back up once it implodes. The other engine of growth, which was substantially increasing exports and the investment to support those exports, that’s also something that they have to rethink.

They’ve also undercut some of the aspects that fed the old growth model. Our Booth colleague Chang-Tai Hsieh has this very interesting observation that what worked for China was essentially competition between the various regions—each region trying to expand growth by changing the rules of the game for their own particular region and championing local firms to grow faster. And some of that has been undercut by the increasing centralization and the anti-corruption drive, because the local government official can’t bend the rules selectively any longer. You make yourself vulnerable to being hauled up if you fall outside the good books of the central administration.

Everything has been overturned in China, and they’re looking for a growth model again. One of the reasons I think China’s much more willing to talk to the US today is simply because they don’t want to disrupt the one thing that’s working, which is exports to the West. And those exports are still important, even though everybody’s saying “China plus one,” and everybody’s looking for an alternative to their China supplier. Often that alternative happens to be a supplier in Mexico who’s importing from all those suppliers in China. It’s one step removed, but those supply chains are holding up. The last thing China wants is a total disruption, which is why they won’t do anything on Taiwan unless pushed. There is some time on that. But of course, these things sometimes don’t go along script.

“I think people are working a lot and not reporting it. They say they’re working 8 hours, but they’re really working 9 or 10 . . . because they’re answering those emails at 2 a.m. or on weekends.”

— Randall S. Kroszner

Kroszner: If we look at continental Europe, it’s facing some real growth challenges. The vice president of the European Central Bank just today gave a speech saying recession is looking pretty likely, but they have to hang tough because inflation is still high. The ECB will have to continue to keep rates where they are for a while to ensure that inflation and inflation expectations don’t bounce back up. Also, being right on the edge of the Russia-Ukraine conflict has more direct effects than in the US, and uncertainty will affect willingness to invest.

In addition, Europe has just passed legislation to regulate AI in the European Union. They are one of the first jurisdictions to act, but they face many challenges in implementation—for example, defining what we mean by artificial intelligence. Because many of the things that we’ve used in the past are forms of AI. Is this going to apply to all of that? By acting early, before we really understand the potential—both positive and negative—of AI, they may limit its useful development and application in the EU and may not see the type of productivity growth that we may see here. Potentially, they’re shooting themselves in the foot because they’re saying, “Well, there’s this big scary thing that’s out there, and we don’t quite know what it’s going to be, so we’ve got to stop it.” That’s a problem.

And obviously the people who are more pessimistic on this say, “If you wait, it’s too late.” But if you try to stop innovation too early, that’s also a problem. European regulators have tended to have a culture and approach that has not been as friendly to innovation as in the US and that is one of the reasons you don’t see very large tech firms in Europe.

Hays: What do you think is driving the recent increase in productivity, particularly in the US? Can these higher levels of productivity be sustained without greater increases in wage inflation later this decade as the baby boomers retire?

Kroszner: It’s a key question globally because ultimately economic growth is the number of hours worked and the output per hour, productivity. China is facing a short-term problem in the financial markets and in the property markets, but also they’re facing the long-term structural problem of aging very, very rapidly. The fraction of the working-age population is declining faster than Japan’s did after its peak.

In the short run, it’s very difficult to know exactly what’s driving productivity growth. We had this paradox for a long time that it seemed that we had a lot of innovation, we had the internet, and productivity growth was very low. Subsequently, the pay-off in terms of productivity growth did appear. Curiously, we are observing high productivity growth in the US in recent quarters and that’s a bit of a conundrum.

Some of it could have to do with the difficulty of measuring the number of hours worked when people are working from home. People are getting used to working from home, and we’re getting a little bit better data on that. But also I think people are working a lot and not reporting it. They say they’re working 8 hours, but they’re really working 9 or 10 and maybe not fully reporting that because they’re answering those emails at 2 a.m. or on weekends.

Ma: Measurement of productivity is not simple at all. When AI or new technologies that are different from traditional physical investment come around, firms may be investing in these new technologies, but not in a physical way, and therefore it’s not recognized as investment. It goes into expenses such as labor and so on. And if it goes to expenses, it tends to decrease the measured near-term productivity and output. But then in the future, these investments may pay off and increase future output, and therefore there may be difficulty in measuring productivity.

Productivity is not just a simple function of technology. It also depends on institutions. One specific example is some interesting recent work [Chicago Booth’s] Austan Goolsbee and Chad Syverson have been involved in documenting that productivity growth in construction in the US has been sluggish for several decades. And if you look at construction costs in the US, they’re much higher than construction costs in comparable advanced economies. Why is productivity growth so sluggish in construction? Some attribute it to overregulation that limits the size of construction firms and limits their ability to use advanced technologies.

The other more general example is democracy. You can take North Korea as an extreme case: modern technologies exist, but with their institutions, you’re not going to get much productivity growth. As we know, institutions are quite important.

Randall S. Kroszner is the Norman R. Bobins Professor of Economics, Yueran Ma is associate professor of finance and a Fama Faculty Fellow, and Raghuram G. Rajan is the Katherine Dusak Miller Distinguished Service Professor of Finance at Chicago Booth.

More from Chicago Booth Review

More from Chicago Booth

Your Privacy
We want to demonstrate our commitment to your privacy. Please review Chicago Booth's privacy notice, which provides information explaining how and why we collect particular information when you visit our website.