When it comes to a 2024 economic outlook, growth is the name of the game.
That, according to Glenn Hubbard, Dean Emeritus and Russell L. Carson Professor of Finance and Economics at Columbia Business School, is the key to getting out of a future economic slump.
At the 16th Annual Columbia Business School Real Estate Symposium, Hubbard shared that while he can’t predict with certainty that a global recession will occur in the near future, many signs point to the likelihood of an early 1990s-like slump, with real estate in particular facing a restructuring and recapitalization of properties and locations.
But similar to the period following the information technology boom of the 1990s, productivity growth is on the horizon in the 2020s, according to Hubbard. A new technology is set to disrupt and revolutionize the growth prospects of a myriad of industries: Generative AI.
Watch the video above or read the transcript below to understand the reasoning behind Hubbard’s 2024 economic outlook and why GenAI-fueled productivity growth is set to take center stage.
Glenn Hubbard: Real estate to my mind as a non-expert is of course about location, location, location, and that's why you're here today. New York is a global capital not only of the ideas and money of real estate, but of global services businesses, Columbia Business School and the Millstein Center in particular play a big role and then all of you as a network, so thank you.
I'm going to do two things with you today and take your questions as well, but one is a short run of where we are, but I want to talk about a very big question that I think will dominate discussions over the next decade of where we're going to be, not only for the economy as a whole, but in real estate in particular.
In thinking about where the world economy is now, I think it's fair to say many economists have been surprised at how robust the underlying real economy is, at least in the United States, slower obviously in the Euro-zone, but robust in Japan by that country's growth standards, and while China is slowing, still somewhat promising growth. At the same time inflation is beginning to recede from, at least by recent years, scary levels. Although as I'll comment later, it is not in the United States at 2%, nor is it going to be 2% anytime soon, without action that remains fairly restrictive by monetary policy, obviously with effects for the economy and for real estate.
Will we have a recession? Honest answer, of course I don't know. I think so though. It's very hard to imagine the stance of policy, this restrictive without at least a mild recession. That's an economy-wide statement. For real estate in particular, and I'll come back to this toward the end of my remarks, I expect the kind of recession to look for is more like the early 1990s, not as much of an overall general crisis in the economy, but a restructuring and recapitalization of properties and locations.
Now, one of the things about the crazy upside down nature that David talked about with the economy is it changes the way that we as business people have to look at planning. As you know, most corporate plans in any industry tend to follow some kind of expected value, and the CEO lays that out to the board and then the board and he or she agree on it and march on. I expect we're not seeing as much of that right now, at least I hope not.
We're going back to what would've been more common when I was younger, which is old-fashioned scenario planning, meaning I may not know much about unconditional probabilities of various things happening, but I may know a lot about conditional probabilities, meaning if this happens, then this. I think more businesses are planning in that way for geopolitics or financial markets and as I’ll cut to in a few minutes, domestic politics as well.
In that kind of environment, I think we're moving from a discussion in business rooms of simple mastering cycles more to what I would like to think of as cycling masters. What are the big things that are coming our way? Those big things, I'll mention them in turn today, have to do with the economy and technology and politics, but I want to start out with what I think is the big thing and that is growth. There are two ways of thinking about this. One is Robert Lucas, the late Nobel Prize winner in economics who used to say it's not surprising that economists get caught focusing so much on economic growth, because once you start to think about it, you can't really think about anything else, in Lucas's words. And it's easy to see why, because small changes that are structural and long-lasting, be they good news or bad news, really spell our future in living standards.
The other growth thinker is Willie Sutton, the bank robber who said it's because it's where the money is. So if you're trying to figure out the future, it's really all about growth. Now growth is all about something else to a first approximation and that's about productivity. It's not literally everything, but boy, it is almost everything, particularly in a world where at least in most advanced economies, demographics are not favorable tailwinds to growth. Now for productivity as this big thing, if you look at the United States in the post World War II period, you see several sub periods, and I say this as an economist looking in a rear-view mirror.
People like me or economists generally tend to be pretty good at looking back and telling you about productivity turning points, we're not really so good at calling them. Paul Krugman, Nobel laureate in economics, famously wrote a book called The Age of Diminished Expectations on the very eve of a productivity boom, and he's a very smart guy, so this is hard. And here are the periods; post-war, the war to the early 1970s — really the golden age when people wish we had an era of productivity growth, that's the period they're longing for.
The second period going from the 1970s through the 1980s, early 1990s is a period of quite sluggish productivity growth and sluggish growth in real wages as a result. Early '90s for at least a decade was another productivity boom. The one recruitment didn't forecast and that boom came really from the ICT revolution, completely changed business. We then go into a period of a structural productivity drop that starts in the early 2010s and we are still in. What is changing of course is the advent of something new. When economists use words like something new, what they usually mean is an emerging general purpose technology, something that is revolutionary in and of itself, and completely changes the growth prospects of industries and techniques around it. And that of course is artificial intelligence and in particular generative AI. Generative AI, I would argue, is both similar to and different from general purpose technologies that we know and love.
It's similar in that we're already seeing ways in which it's going to completely transform not just technology, but almost anything. Dare I say, even economics professors may get disrupted by this one. This is very big, and is different in a very important way both economically in business but also for politics and it has to do with speed. Traditionally, general purpose technologies took a while to permeate through the economy. Bob Solo, a prominent economist Nobel Prize winner once famously said in the 1980s, "You can see IT gains in productivity everywhere except in the productivity statistics." And the reason for that is it takes a while for businesses, for organizations, typically to adapt to the new technology. The impact, for example, that old technology, electrification, has on productivity doesn't happen until the 1920s or 1930s and I'm sure you know that's not the dawn of electricity. It took a long time.
Likewise, mainframe computing and the internet to other GPTs took a long time because the organization of the office, the organization of knowledge work took time. This is the sense in which the views of generative AI are that it's very different. The speed will be much faster. That means that the productivity gains may be much faster, but that means so too will be the dislocation. Let me try to use an example to set the idea. If I were to compare the number of days to get to a hundred million users across recent technology platforms, it took Netflix 3000 days to hit that mark. It took Twitter, X, Facebook, Spotify, all about 1500 days to hit that mark, it took Instagram 900, ChatGPT, I want to say an afternoon. I mean it's not quite that fast, but it's a lot closer to that than any of the others.
What's more important of course, is not just the number of users, but the tremendous business use cases. If you look at studies that practitioners are doing of the likely potential cumulative growth effects, recent studies by McKinsey and Goldman Sachs have basically argued we could see a half a percentage point increase in productivity growth. That is absolutely huge, with cumulative changes of between $3 and $7 trillion over a decade depending on the model. Underneath that, and I want to say this as a foreshadowing of something I'll come back to in the political economy of technology going forward, is the job market, so those statements about productivity growth or statements for the economy as a whole, they're statements of potential growth if nothing tries to slow down the generative AI technology. But there are a couple of different effects and ways in which economists think about how AI affects labor.
One is the classic effect, go all the way back to Adam Smith. If I have a technological innovation that's going to raise productivity, it's going to raise labor demand and raise wages, and that would be the good news story for AI and it is certainly a plausible story. The wind and the sails of that story would be that AI does a lot in developing complementary technologies.
Let's say I have a relatively low-skilled job in retail or low end of services, it still may help me be more productive in dealing with customers. At the very high end, it might help me as a junior architect or a lawyer or a banker or dare I say real estate professional or economist, do his or her job. Those are all good stories. Eric Brynjolfsson at Stanford is probably the leading arguer of those very positive stories. Of course, it is possible that AI does something else.
The origins of thinking about AI are actually old. They go back to work in at least the 1950s by Herbert Simon and others. And the goal of the original discussions of artificial intelligence was nothing like I just described. It was actually quite different. It was to substitute for human general intelligence. One of the fears about AI and jobs, which has attracted the interests of politicians as well as business people, is it going to be the case that this good news story that I led with happens — productivity, labor demand, wages, everything's going up — or is it something darker? On that side, another very prominent economist, Daron Acemoğlu sits. I would argue with you that the limits on growth from new technologies over the next decade don't come from the familiar places, and the familiar places to look for a problem or two. Hiccups in the science, that we're not going to get it right.
I'm not an engineer. Actually, I am an engineer but of so long ago, my views are irrelevant, but science is not going to be the problem here. Everyone I speak to who is actually current here believes that the best things are still ahead, nor do I think organizations are going to fail to be nimble in capturing and using AI the way they might have been in some of the slower moving general purpose technologies of days gone by. I'm really worried about two things.
One is some under the hood economic forces that I'll mention and the other is old-fashioned politics, political economy, and I'll go over each of those in turn. One factor that can act to slow change in technology basically is a factor familiar, especially in a room like this, which are shifts in capital markets and interest rates. The 12 years of global low interest rates that probably dominate the careers of many people in this room abruptly ended in 2022, and moved toward what I think of at least by the standards of the recent past as a new monetary order.
I I believe that the Federal Reserve intends to deliver a 2% inflation target. Inflation is not going to settle at 2% without high rates for a while. So I would say, and of course you can discount this, because I'm not the one playing with real money, but I would say to folks on Wall Street, who are expecting great cuts anytime soon, keep expecting. It's not going to happen anytime soon. It will happen of course, but policy would have to remain restrictive to get to 2%. Now if your view is that the Fed doesn’t mean that and is willing to accept the high twos or the low threes instead, that's a different argument, but I actually don't think that is true. And even if it were true, it wouldn't affect as much predictions for nominal interest rates, because it would still mean the inflation premium would be high.
Obviously in that new monetary order of headwind for technology, the rate effect is significant. If you think about technology investments and valuations, those like real estate are the long-dated cashflow assets and if you walk into a room full of MBA students and say the real interest rates changed in one direction or another, the price of what kind of asset moves the most fast, it's always an asset with a long life. And so the new monetary order is definitely going to be a partial headwind. I think what is going to be a plus though, is dynamism with the new technology. One of the downsides of the ultra low interest rate environment may well have been a misallocation of capital that will not happen with interest rates at more conventional levels. A second headwind to worry about in the evolution of the technology is really a potential new era for fiscal constraints.
Now on the one hand you might say, "Well, what's he smoking? What's the fiscal constraint?" We have one political party in the United States that says we won't raise taxes at all. Another party says we won't cut spending at all. The two parties are roughly 50/50 in the Congress and so what's going to happen? Well, I think what's going to happen is math, and so if one accepts the prospect that defense spending needs to rise, it's spending on things complimentary to this new technology boom, IE, basic and applied research and support for adaptation to the policy needs to rise in the presence of already structural deficits, we may well have some fiscal headwinds.
We're already seeing as a result of current fiscal policy and the path that we're choosing to run higher rates, interest rates, the effects of that on the federal budget. This is the first year in which interest spending actually exceeds national defense, which ought to scare everyone in this room. And we're beginning to in the political process over focus on things that aren't important, like little tiny elements of discretionary spending or saying let's tax two or three billionaires as opposed to more fundamental reform. That failure to act can disrupt the technological innovation both through pricing and valuation, but also through enormous uncertainty about business investment.
The next factor I'd at least like to highlight or mention to you is about geopolitical and then importantly, because we can control it more, domestic political dysfunction. In geopolitics, a big wind is oil and energy. I happen to sit on the board of Total. It is a very large integrated oil company and every discussion at Total is very much centered on what is the next 10 or 20 years going to look like for the pricing of production of oil with the geopolitical consequences. We're at a turning point where we're wondering is a society not just about the price vector for oil, but the use of oil in a fuel mix. So the amount of geopolitical uncertainty from oil is probably unprecedented. The second that relates to geopolitics as opposed to events at home are defense and fiscal strains. As I mentioned, people who focus on the defense budget have argued we probably need to spend at least in this country about a full percentage point of GDP more on defense than we do today.
There's absolutely no room in the budget environment to do that. Raising big questions for business and growth about well, is that going to be financed by tax increases on say capital, or is it going to be financed by some other spending cut? Another possibility in this environment of course is an increase in the equity risk premium. So for a while we had a period in which not only were safe interest rates very, very low, but the applied equity risk premium was also very low, and hence the value of equity is very high. Now we're already seeing the unwinding of the safe discount factor, but we're also starting to see increases in equity risk premium with valuation effects. A lot of that is coming from this geopolitical uncertainty. For domestic politics, I think we have at least as large a problem. Part of that is the fiscal one that I mentioned, the inability to address or deal with basically full employment deficits.
We have an economy that's roughly growing at potential, even though it's slowing, then we have enormous budget deficits of $2 trillion. But I would argue that the bigger effect and where it's going to intersect directly with AI and growth over the next decade is through pressures on social stability. Think about technology and technological change over the past several decades. If you were to start your mind's eye in the early 1970s and remember I suggested that was a period in which the Gaussian productivity days were ending. We did start to see in the 1970s the emergence of very significant technological change affecting industry, and also an effect of globalization. For most people in the news media or in politics, globalization trade, that's the big story. That's not how economists see technological advances. The big story: globalization is a factor, but it is not the biggest factor.
But over the period of time from the 1970s to today, we've basically seen a slow-motion disruption. When I mentioned growth, imagine that I could describe to you I have a coin and of course coins have two sides and the head side of the coin is growth. The thing we want to talk about, tail side is disruption. The truth is there is no modern economic model of growth that isn't centered on disruption. No economist believes that growth happens with just simple gradual progress. It's jerky, discreet changes in possibilities like generative AI, but they bring with them disruption. Now, if you go to the political economy of change, much of what we've seen in the rise of populism in industrial economies has been a reaction. Chances are — and I can tell by looking at you that Econ was your favorite subject in college. You have that look, and I'm sure you remember at Econ 101, I'm sure was your favorite course — the professor talked about technological change benefiting the economy on average.
He or she probably told you that free trade or globalization benefited the economy on average. Then you probably went back to sleep or looked at your phone or something and then he or she probably also told you, but not everybody. There's going to be distributional consequences. Some people like all of us really won big in that average refloating of the boats from technological advance and globalization. Others did not. The failure of our political process to really deal with that fundamental Econ 101 question has left smoldering sectors in the economy and that was with technological advances that were slow motion changes over decades. As I argued with you before, one of the key things about AI that's different or at least expected to be different is how quickly it is going to change things.
This could well be the mother of all political economy problems. What to do? Let me first mention it in Econ speak and then mention it as if one were talking to somebody who was a doer in politics. So if you think about an Econ speak, for God's sakes, we need the golden goose. So we need to support innovation. So for an economy like this, that would mean massive support for basic research. Your favorite course, Econ 101, professor probably told you we're not going to get enough basic research if we wait for business people to do that, because they can't appropriate all its value. We could be doing a lot to support basic and applied research. We could be doing a lot to boost investment. We could try to raise national saving by getting our fiscal house in order, and importantly for the political economy of the situation, we need to encourage participation of more Americans in the life of our economy.
In terms of the way one might speak politically about it, I don't know about you, but I don't much like words like “America First,” but how about America the best? How about an economy that devotes the resources to be at the frontier? Instead of eat the rich, which doesn't sound so good to me, how about restoring the American dream and opportunity? And instead of words about industrial policy from various bureaucrats, how about thinking about a can-do and let ‘er rip economy with technology? For business leaders, this going back to the future, it's really going back to scenario planning. It's adding a twist of thinking about the social support for business that we all took for granted, it may not be there and staying focused on that high productivity and high growth future.
For real estate, obviously the current headwinds are obvious to everyone in this room, there's also a lot of opportunity in the redesign of housing, of logistics, of healthcare with complimentary skill investments in AI with the reorganization of the office. The complicated part will be on the financial side of this, how one works through a period in which the existing capital stock of real estate may have been partly in the wrong place and partly overvalued.
Working through that, through the new world is actually an interesting challenge. One for the private sector, but also for the public sector. Part of it is — David and I were speaking about this before we started — do you want to slow change because the private sector will get this done, or do you want a faster change maybe with more public sector involvement? But there's not a right or wrong answer, but I think you're going to be talking about it today. So let me stop there and thank you.