3. Tyler Cowen: Second-Order Economics, Talent, and Straussianism

Tyler Cowen is a professor of economics at George Mason University and director of the Mercatus Center. He needs little introduction given his widespread online footprint via his regular works such as his blog, Marginal Revolution, his podcast, Conversations with Tyler, and his column on Bloomberg. Tyler is an avid traveler, the author of books including The Great Stagnation in Talent, and a food connoisseur of many cuisines. Hope you enjoy the conversation as much as I did. Thanks for taking the time to talk today, Tyler.

Happy to be here. Hi, Ryan.

First question: should the US banking system emulate more characteristics of the Canadian banking system? Should we have larger, more diversified banks and even consider abolishing FDIC insurance?

Well, we're headed to having fewer banks and larger banks, whether we like it or not. So there's a lot of money leaving regional banks. And as you probably know, our 4 too-big-to-fail banks appear to depositors to be safer. And indeed, they are safer. So should we do that? We've painted ourselves into a corner. I don't really know what options we have now. Unlike most of the rest of the world, we do not allow U.S. banks to mix with commerce in any substantial way. I would favor changing that and letting those banks diversify more, much more. But that seems unlikely. I see zero interest of moving in that direction. So I don't think we can ever be like the Canadian banking system, which it seems to me for Canada has worked quite well. But at this point, it's just we don't have access to that space.

Are institutions like the Fed necessary for long-term financial and macro stability? Does this imply a murky future for crypto? If it cannot create a credible alternative to this institution?

Well, let me take the Fed question first. So I don't believe that a gold standard is possible. I think the 19th century gold standard worked fine, but right now the price of gold is too volatile due to China and India and a number of other factors. But I don't think we could get back what we once have. Same is true of silver. I think we're, we're stuck with a dollar standard whether we like it or not. And if we have a dollar standard, we need a Fed. So it's again a point of there's no way to get back to that earlier policy space. So, uh, yeah, I believe we should have a Fed. Now, I might prefer if our Fed were more rules-bound than is the current Fed. So Scott Sumner's idea for nominal GDP targeting, I think that would be better than what we've had for the last few decades. Uh, your second point about crypto, I don't view crypto as a replacement for the dollar. I view it as a complement to the dollar, so it will be used for other things. For instance, circa 2023, I think the AIs are going to start trading crypto with each other. They need money. Uh, we're not going to let them open up checking accounts, so crypto will be super useful for that. But I don't think human beings, you know, will quote unquote carry crypto in their wallets as a transactions medium. It's a store of value. It's a way of getting money out of nasty and evil countries. And it will be a medium of payment for AIs, which is great. But I don't think crypto needs a central bank or could even have one. You know, the rules are the rules in a well-functioning crypto system. And a central bank is just a bunch of hot air.

If you view crypto as a complement to the US banking system, more or less, uh, does that mean that you are, uh, perhaps more bearish on a CBDC type thing ever evolving in the US?

I have in the past opposed a CBDC. I worry that it puts our central bank in too much of a position to make lending decisions allocating capital. It makes our central bank too political. I'm against that. Other people have privacy concerns. Those to me potentially are another reason, but it would depend on how the CBDC were done and who exactly would have access. My worry is that in some quite long run, a CBDC might be impossible to avoid. So if cash disappears and everything becomes digital, I don't know. I don't want a CBDC. My hope is we can keep on steering away from it more or less forever. I'm not convinced we'll be able to.

Can you explain why a cashless society necessarily translates into a CBDC in the long run? Like, why can't we have private market alternatives in the long run?

Possibly we can. So you can have banks issuing inside money, right? But if you have a Fed and the Fed has a clearinghouse function, the Fed has something in there that over time might not be the same as simply electronic claims to dollars. They might become the electronic claims themselves. And then you have to ask, well, who has access to having an account at the Fed? Right now there's a, a well-defined legal distinction between banks and non-banks. But for instance, if we mix banking and commerce, as I would like to do, As banks evolve quite a bit, it may become very blurry, you know, who really has access to the Fed's payment systems. And then as more and more agents start to have direct access to the Fed, you could evolve your way into a CBDC in a way I would feel uncomfortable about. But again, I'm not entirely sure how to avoid it. Though certainly, you know, in the next 20 years, it's not a thing that's going to happen. For now, we can ignore that risk.

How will near-term versions of AI affect real interest rates, if at all?

You know, there's something I coined, it's called Cowen's Third Law, and it says all propositions about real interest rates are wrong. So I don't know. Cowen's Third Law is itself wrong. I certainly know that in theory, as productivity goes up, real interest rates should go up. And that would be my best guess. But I'm more aware of Cowen's third law than of having faith in that prediction. I do think AI will boost productivity. I'm not sure about the rest. It could be you boost productivity, there's a lot more savings, that goes into the market for loanable funds, real interest rates don't have to go up.

You recently wrote a book with Daniel Gross called Talent. As the talent market matures and becomes a market with greater liquidity, does that necessarily imply that programs or employers seeking talent will look earlier into people's lives? And will it become truly normal to give out grants or something similar to 10-year-olds?

I don't quite think 10-year-olds, and there may be legal reasons for that, but I think 14-year-olds, and I think what you're describing is already happening. There are plenty of areas, one would be basketball, another would be chess, where you can measure talent quite early, or at least have some idea. Basketball, it's not all about height, If there's a 14-year-old who's 6'7", uh, you're going to give him a second look, right? Chess, you can just play, you have a rating. So again, I would say you're describing the status quo, but I don't think it will get back down to 10. I think parents find that too weird. 10-year-olds can't really handle the interview, but 14-year-olds can.

Okay, uh, and for 14-year-olds, how do you control for that, that kid's intrinsic motivation versus that manifestation of their parents or the guardians?

I think a lot of times if you just talk to the kid. So I had one Emergent Ventures applicant and he seemed quite smart and he went to a very good school. And I asked him, well, how did you hear about Emergent Ventures? And he said, my mother told me to apply. So that was a no. Like, he might have won otherwise.. But I just figure, you know, not self-motivated, the kid will do fine. Pushy parents, in a top school, lived in a great part of the country, but not for us.

What's the youngest age of a grant recipient of Emergent Ventures, if you don't mind me asking?

I believe it's either 14 or 15. I wouldn't swear to that, but almost certainly it was either 14 or 15.

Is there something to people who are exceptionally good at a given field, whether it be chess, a sport, etc., but not quite the best, that leads them to do kind of greater things in the long run? I believe you've attributed the fact that you weren't the best at chess to kind of take a step back and explore different avenues. I think Bill Clinton played the saxophone, Peter Thiel with chess as well, and others seem to have experienced something similar in their childhood.

I think it's very useful. Here's my hypothesis. I've never seen this tested against data, but if you do very well but not too well, like, the first thing you learn is that you can win and you get a lot of confidence. But the second thing you can learn is that you can lose and you realize you have to keep on improving. There's too many young teenagers who are never quite challenged against the very best, and they just think they're so smart. And they don't realize they've got to work harder than anyone else to really get somewhere. So I think these competitions, games, whatever, it's very useful to be very good, but also experience some real losses and just, hey, I'm not the best.

Doesn't the internet kind of provide a space where they can actually play against the best? Like in many sports or something like that, it would be like chess or video games. There's like a leaderboard. So that they actually know that they're likely not the best?

Absolutely. And maybe you can do something with or against your AI now, right? I mean, that's the internet too, but it's a little different than just the plain internet. So kids playing chess against computers, learning from computers, that's been a thing for a long time. And you just see so much great young chess talent because of this.

There was a recent interview with Sam Altman on the increasing age of successful tech founders. Does this imply a more stagnant, less dynamic tech sector?

It does, but I don't think it's a bad thing. I think it's a low-hanging fruit issue. So before Mark Zuckerberg and Facebook, yes, there was social media, MySpace, but it wasn't that good. It wasn't a huge thing. It was just like a thing. And now it's a huge thing. And you could build Facebook starting at age 18 in your Harvard college dorm. I think now the key companies, they're a lot more complex. Like, look at NVIDIA. It's not a young company, but only recently is it worth as much as it is. You can't do that at age 18. So the fact that we're seeing greater importance for mature companies is just a signal of what we've already done so far.

Interesting, interesting. I think in The Great Stagnation, you reference an economist, I believe Charles Jones, who kind of suggested something similar. Where kind of all the low-hanging fruit has been picked, and that'll just lead to lower economic growth rates in the long run. And it takes a lot more education and a lot more— a lot larger clutches of people to do something truly innovative these days.

I strongly agree, but with the qualification: every now and then you refresh the low-hanging fruit with a new development. So right now I would say it's AI. There's a lot of low-hanging fruit like, oh, Use AI to edit your paper. Use AI, you know, to run your plugins once that gets better. And I think we have another 10, 20-year period of incredible low-hanging fruit. Now, might we see another wave of 20-year-old founders? I'm not sure. It could just be the big companies like OpenAI, Google just build that stuff into the core product. And maybe there's not so much room for founders and like add-ons. I'm not sure, but I think unlike Robert Gordon, I think periodically low-hanging fruit gets replenished.

Who do you think will capture the most value in AI? Will it be the infrastructure providers or the application layers that actually own distribution, or will it be obviously some mix between the two?

Users. I mean, if you look at market values for these companies, they're certainly up, but they're not up by some extraordinary amount. So first users and second hardware. NVIDIA is on the verge of being a trillion-dollar company as we're speaking. So if it turns out we need a new and special device to run our AIs on, that the iPhone is, while useful, not perfect for AIs, uh, whoever builds that new device, maybe it's Apple, but whoever that is, I think would reap high returns. Because there's much higher barriers to entry when you build a supply chain. I think a lot of AI will become commoditized and pretty cheap, and mostly you and I will gain.

What do you think a well-educated but zero marginal product worker in, in their mid-30s should do to remake themselves for the next 30 years?

You said well-educated?

Yeah, like well education or, or decently well-educated?

Well, they need to get off drugs, right? If they're zero marginal product, they're either deeply depressed or they're taking harmful drugs. Uh, they can have a positive marginal product just by avoiding self-destruction. Now, a lot of people don't do that, but I think that's the actual problem. It's just the person's discipline. I don't have any recipe unique to me for how to do that, but There's plenty of people who study it.

In your career, what was or is the ideal equilibrium with respect to fame? Do you think that you've become less weird or your public image has become less weird than it was 10 or more years ago?

I'm probably slightly more famous than is optimal. I would agree with that, but I don't think I'm so far from the optimum. Like, I don't have truly mass popularity. I don't have so much money that I don't like ever think about working again. So, uh, yeah, pretty— it's gone pretty well for me, but I should be somewhat less famous.

You spent a lot of time, uh, involved in many spheres including academia and industry. How can we best encourage academic professors to become more involved with projects outside of academia? What are the incentives or mechanisms, uh, you would employ to encourage this behavior? Assuming that you agree that this would lead to desirable outcomes?

I think if you could reconfigure everything, it would be great. But a marginal move in that direction from the status quo, I'm not sure is so useful. First, most of those people would not be very productive in these different capacities. So why do we want to put them there? Second, plenty of people now do consulting and they're quite good at consulting and they earn good money doing consulting. I don't have anything against that, but it's not obvious to me that all this huge value is being added now by having these economists do consulting. Sometimes it's for legal cases. There's good economists on both sides, like, okay, fine. I don't know. It doesn't feel to me like the direction we need to press in. So I think there's just too much consulting and not enough like getting out there and just learning things, being curious. Being what I call anthropological.

How would we encourage that sort of latter behavior and perhaps less consulting and more sort of zero-to-one entrepreneurial anthropological efforts?

I don't know that we can. It seems to me there are too many barriers to becoming an economist. The degree takes too long. Now you're supposed to get a pre-doc, you know, you're supposed to write these 90-page papers with multiple appendices. So it encourages hard workers, which is fine, but maybe discourages the curious, and it all feels like too much of a grind. And I don't know how to undo that, but just to turn the clock back a few decades, if you could, the people who were attracted to economics were more like idea-rich, more curious, more willing to consider doing different things. Now what's required of you? University committees reviewing papers for journals. Filling out forms to raise money, writing those 90-page papers. It seems very wrong to me.

Is there any way to sort of change that sort of culture, or do you think that's sort of the result of kind of bureaucracy kicking in as the institutions ossify?

Well, I think like you said, it is just bureaucratization. Now, you can debureaucratize structures, but it's very, very hard. My hope is higher education just sees some new entrants with different models, maybe like Minerva, University of Austin, something with GPT and AI. I'm hopeful, but that, all that is going to take a while. So in terms of the short run, I'm not hopeful at all. I'm pretty convinced it's just going to get worse.

Is Marginal Revolution exploring any issuing or like leveraging AI in different ways?

Oh, well, I personally am trying to leverage AI in different ways. I mean, you could say that's marginal revolution, part of it. So Alex Tabarrok and I wrote a paper, How to Use GPT Models to Both Learn and Study Economics. I did a podcast with GPT-4 doing Jonathan Swift. I'd like to think more about how to build out AI products for education. I'm quite sure it can be done. I know it can be done. But it needs more structure somehow. I think one thing, you know, you asked about the blog. One thing I try to do to improve my own profession is just if someone does something interesting, to reward it and praise it and give it publicity. So I'm not so keen on attacking the stuff I don't like. I'd rather spend my time and effort praising the things I do like, and I hope that's some tiny bit of an incentive for people to be more interesting.

What is the rate-limiting factor as to why there are not more Mercatus Centers in the world? Is it funding, talent evaluation, culture, location? What are the characteristics of the Mercatus Center that makes it generate ambitious, unique talent?

I think having a very small number of people who really want to do it and who are willing to make it a huge thing in their lives, that's just quite scarce. I mean, I would be one of that small number of people in the case of Mercatus. Along with the founder of Mercatus, a fellow named Richard Fink. But people feel they have better things to do. Like, I don't do any consulting, so I don't earn income from consulting. I'm fine with that. But if you say to a smart economist, well, you can like try your version of Mercatus starting from scratch, or you can earn like $500 to $1,000 an hour doing consulting, what are they going to do? I don't know, but I think I do know.

What does Tyler Cowen look like in 20 years? Will you ever truly retire or will you always keep producing some outputs, whatever form that may take at the time?

Uh, short of senility, I will never retire. At some point I'll die, but I hope to be working on my penultimate day. Maybe the last day I'll take off and eat some, like eat too much chocolate ice cream or something.

If you're up for it, I would love to play a game of X or Y where I say two things and you choose between the two. Some dichotomies will be contextualized and others will be open for interpretation. Of course, feel free to pass on any. Peloton or biking outside?

Well, biking outside to me seems quite dangerous. The FDA would never approve it if it were put up. You know, is it safe and effective? It's effective, but it's not safe. So I love the inside bike in our house. When the weather is bad, I use it, you know, every single day and it works for me. So I'm going to say Peloton.

Traditional farming methods or lab-grown meat as a future of food production?

I don't think people will ever want to eat lab-grown meat. I mean, I hope they will. Even if it tastes good, you know, it's hard to scale. The current efforts I don't think really have succeeded. And I think we are programmed to want to eat actual animals, even though that's a bad thing for animal welfare. So I would be short lab-grown meat.

Does that imply that you never think that they will taste identical or humans won't be able to tell the difference?

Well, it depends what you count as lab-grown meat. There's plenty of non-meat foods right now that can taste better than meat, so I wouldn't rule it out. Or lab-grown meat, but I don't think it will be a meat substitute any more than say really good sushi is a meat substitute. So I just think we actually want real animal meat. It's kind of sick, isn't it? Uh, but that's my prediction.

Urban sprawl or compact city designs as the ideal urban development?

I don't think there is an ideal, you know, you want both and to let people choose. My favorite city is Los Angeles, which I think answers that question for you. But the people who prefer Copenhagen, I don't think they're wrong. That's great too.

West Coast or East Coast Straussians? Who do you think Strauss himself would identify with?

I'm not sure that any of those people are Straussians. You know, they call themselves Straussian. Maybe that's like their actual Straussian move. So Strauss to me is hard to figure out. I'm not sure there's a there there. My loyalty to Strauss is at a pretty general level that doesn't necessarily map that tightly onto what Leo Strauss the human being might have thought. And that's simply the notion that to understand writings, you need to take the incentives of the writer into account and realize there's some kind of implicit or explicit censorship in virtually all eras. That's what I mean by Straussian. Uh, beyond that, I'm not sure Straussianism is that that valuable. The historical pessimism, the obsession with elites, the ancient world. I don't know, I don't really feel loyalty to those things. And as different groups of Straussians out there promote their whatever, I'm like, eh, that's not my thing. Let's have some economic growth, get on board with that, you people.

Obsessiveness or competitiveness?

Well, I'm very obsessive and not actually that competitive. I think that's quite a strong dual feature of my personality. So like, there's not some person out there where I know how many Twitter followers they have or how many readers they have, and I compare it to me. The notion of people wanting to be like the richest person in the world or in Pittsburgh, that feels very foreign to me. But I'm very obsessive in the sense I keep on going, keep on working on it, work on it lots, just keep on doubling down. Really for my whole life. So of course, you know, I'm going to go with obsessiveness. But I'm glad a lot of people are the other way, right? Because insofar as they're competitive, they're more consumer-oriented, and I'm a consumer. But I feel they're tricking themselves. Like, who cares if you beat that guy?

I think you, uh, maybe it was Daniel himself, but you described this dichotomy between you, you and Daniel, and you were more on the obsessive side and he was more on the competitive side?

He's also obsessive, to be clear, but I'm very skewed toward obsessive and not competitive, and he's strongly both.

Would you say that, that tech founders should in general be more obsessive or be more competitive and what it like naturally selects out for?

I think it already selects pretty well for obsessiveness and competitiveness. So that— those filters to me seem to be working pretty well. I don't know that I could really improve them at the margin. Maybe something where women would have better opportunities, but that's a somewhat separate question.

Uh, stamina or intelligence?

Well, they're complements, but I think stamina is what is underrated and intelligence is what is overrated. So I'll say stamina.

Effective altruism or local community engagement, which one is more underrated globally?

Well, so many people still don't even know what effective altruism is. And I think at the margin, effective altruism is quite positive, even though I don't agree with all of it. Local community engagement, a lot of that's negative. It's like NIMBY and, oh, they can't build the new apartments here, or I don't know, I would think Possibly half of local community engagement's negative. So I'll say effective altruism.

Traditional or self-publishing as a method of higher returns over the next 50 years?

I think it will always be traditional. I don't quite understand why, but there's something about outside checks on an author that's useful. And most books lose money anyway, and the books that make real money make so much So, oh, J.K. Rowling had to give whatever percentage she gave to her publisher. I don't think it matters that much to her. So the publisher makes it better. So I think that's an eternal model and we've had it since the beginning of publishing. I would be long that.

Do you believe that in your personal case that your books are of higher quality or just different than your other content you produce online?

Well, I think at the moment my podcasts are my best content. Over a longer time stretch, I think my blog is my best content. But again, they're all compliments, so it's hard to say.

India or Nigeria in terms of relative productivity on a per capita basis, and what country you're broadly more bullish on?

Well, you can look at the numbers. India is clearly doing better than Nigeria, and India is growing at 6% a year. Nigerian growth, true growth, is hard to measure because so much of measured growth is like an oil price thing that's then removed from the country due to corruption. Uh, but it's possible Nigeria is only growing, say, at 1%. So you've got to give that one to India across the entire board.

Final question: LA or New York City as a place with the best ethnic food in the US?

Well, it depends how you measure it. If you could have a Star Trek transporter and take yourself to Queens whenever you wanted, it might be a tie. But as it stands, if you live in Los Angeles and have a car, I think you're better off eating than if you live in New York and have to deal with the subway. So I'll say LA, but there is some counterfactual where what New York has is as good. But like, who's going to take the subway to Queens and back, right?

True, true. Favorite restaurant in LA or top 5?

Oh, there are Thai restaurants in Northeastern Hollywood, and which is the best, you know, in a given year changes a bit, but I have a friend there, Dan Gould. I write Dan, he tells me which are the best, I go to them, they knock my socks off, I love it. They're just phenomenal. Northeastern Hollywood Thai food. Ćitláad is one of the famous ones that has stayed good for quite a while, but there's others that come and go a bit more.

Tyler, thanks for your time. It was a pleasure.

Thank you very much.