Luke Froeb joins the podcast to talk about his career in economics, what it's like to be the chief economist at the FTC and DOJ antitrust division, how these agencies make decisions about merger cases, the history of the Chicago School consumer welfare standard and the types of analytical tools and modeling that underlies the approach, along with the rise of the New Brandeisians and their failures thus far.
Jon: “This is the Capitalism and Freedom in the 21st Century podcast, where we talk about economics, markets, and public policy. I'm Jon Hartley, your host. Today, I'm joined by Luke Froeb, who is a professor of management at the Vanderbilt University Owen School of Management, and a former chief economist at the Department of Justice Antitrust Division, and former chief economist at the FTC. Luke is one of the nation's foremost experts on antitrust issues, and is a leading expert on competition policy. Welcome, Luke.”
Luke: “Yeah, thank you for having me. I'm thrilled to be here.”
Jon: I'm curious, like, where did you grow up, and how did you first get interested in economics? Like, I know you did your undergrad at Stanford. Was that where you got interested in economics, and where you decided that you wanted to get a PhD in economics?”
Luke: “Yeah. I grew up in San Diego. I actually wanted to live at home and go to UCSD, and my mom kicked me out of the house and made me go to Stanford, so it was definitely my second choice.”
Jon: “You grew up in La Playa, right?”
Luke: “Yeah, yeah, 50 yards from the beach. I mean, I had no idea. When I grew up there, it was a middle-class suburb, and it's just, you go back there now, and oh my gosh.”
Jon: “It's funny. I met your parents once on a family vacation. I didn't meet you. It's just a happenstance kind of thing, but I remember I had heard a lot about you back in 2010, and since then, we've met and got to know each other a little bit. So, you grew up in La Playa, and then you went to Stanford, Northern California. Is that where you first got interested in economics?”
Luke: Yeah. I wanted to be a physicist, and so I studied. I got really excited about physics when I was in high school. I took a physics class, and I thought, this is so cool. I like the idea of modeling everything, and you could understand the entire world with physics, and I got to physics. I mean, I got to Stanford, and I started taking physics, and you have to, well over half of your required courses are math and physics, and so it didn't leave much time for much else, and I found that I was one of the smartest kids, or the kids who did the best in my high school in my classes, but when I got to Stanford, I was just an average physics student, and my roommate was taking econ, and he kind of explained to me what it was. I had no idea what it was. Anyway, so he showed me some of his problems, and I could do them, and I kind of got interested. It was the exact same tools as physics, and I loved the idea of being able to model stuff and understand stuff through models, and so I started taking econ, and I really liked it, and I got interested. At the time, I was on the left side of the political, way left side of the political spectrum. I lived in Columbia, the social change through nonviolence co-op, and I got interested in Marxist economics, and I studied really hard. I remember this one guy who was a senior at the house. I was in 10th grade, and he talked me through Das Kapital, and I was taking John Gurley, who is a really good teacher, and he was a Marxist, and he was in the econ department at Stanford, and he introduced me to, he taught, I think he taught the core microeconomics class. He was a really good teacher. It was really fun, and I thought, man, this is great, and then I took his follow-on Marxist class, and I looked back, you know, I had a copy of Das Kapital. That was the reading in the class, Economist's Manifesto and Das Kapital, and there's a couple other readings, but, and I just, I looked back and I underlined everything. I just tried really hard to understand it. I could never understand it. And it was only about a decade later, you know, when I, when I, you know, I looked back on it, I go, oh, yeah, you know, they don't, they don't consider incentives, you know, they, they just assume that, you know, everybody's going to work, you know, for the common good. And, oh, my gosh. I'll tell you a funny story. When, when I lived in Columbia, they, they modeled the house, and they still do, I think. I asked, you know, they run the house by consensus, and it's really easy to decide, you know, which third world political, you know, revolutionary we're going to lend our support to. But when it came, it was an old frat house, and there were some really nice singles, and some really terrible quads. And when it came to decide, you know, every quarter, we, we'd decide who lived in which room, and we'd make the, we'd have this consensus meeting, and, and you'd have people arguing that for the good of the house, and for the good of social change through nonviolence, that was the, that was the theme of the house. I deserve a really nice single. And, and I, this can't be right. And it just, it was kind of the first inkling that I got, you know, that, that the idea of kind of collectivism just, just doesn't hold up with individual incentives. And, and people were just naked, naked self-interest in these, in, you know, who decides who gets to live in which room. But when, when there wasn't a cost to, to lending our support to third world revolutionaries, you know, that was easy to decide. But anyway, it's kind of, it was kind of fun to get looking back on it.”
Jon: “That, that's fascinating. It's, it's funny how central planning seems to work really well, I guess, in a, in a small family where you have sort of a, you know, a dictatorial sort of parent or two. Central planning works quite well, strangely. But when you expand it to, to more people, you know, I guess to size of a house or, or a school or a country, things seem to go really south in terms of efficient allocation.”
Luke: “I'd like to show my students at Vanderbilt. I show them some, there's a great, Tyler Cowen and Alex Tabarrok at the, at George Mason run this kind of online university called Marginal Revolution. And they have these great series of videos, and one of them is on private property rights. And they talk about, you know, the, the birth of modern China, which was, you know, in Zhao gang Village when these villagers got fed up with, with collective, ownership and nobody, nobody's working very hard because you've got all these people in the village and you only get, a fraction of, if you produce an extra bushel of wheat, you only get a fraction of that. And so, you had people, people free riding and they divided up the collective land into private property and signed everybody a parcel of land and they called it the Household Responsibility System. And, and output went up by a factor of six when you have the incentive aligning effects of private property. And, and yeah, I just, I'm, I've showed that to all my students. But the same thing happened. Same thing happened in, in Vietnam. Same thing happened in, with the pilgrims. When John Bradford came over on, on the Mayflower, they had collective land ownership. Nobody was working. And then, you know, he signed private property and, and that, that we were, we all learned the mythology of, of Thanksgiving, you know. The reality of Thanksgiving was that, that it was private property that we all, all would be thankful for. So, I'd make my kids watch that every Thanksgiving and then we'd go around the table, what are you thankful for? And everybody says private property.”
Jon: “Oh, that's too funny. That's, that's too funny. So, I think, you know, you decided to get a, a PhD in economics while you were at Stanford and you got into the University of Wisconsin, you did your PhD there. I'm curious, how did you first get interested in antitrust and, and competition policy?”
Luke: Yeah, so I wanted to get out of, I mean, I got, I could have gone to Stanford. In retrospect, I, you know, anybody who's listening to this, who's going to be a key time grad, so go to the best school you can get, get into. And so, I probably should have gone to Stanford, but, but I wanted to get out of California. I'd spent my whole life in California. And so, I went to Wisconsin just because the people were so nice there when I visited. And they're really good at econometrics. And that’s where I took the first year of econometrics when I was at Stanford. I studied econometrics at Wisconsin. I got a job down in Tulane. That was kind of fun, but I gave a paper at the Justice Department. I wrote my thesis when I was at Wisconsin. I was a time series macro-econometrician, but I wrote my thesis on time series of profits and concentration, which was, I don't know, it was an esoteric topic, but it was using my methodology of time series econometrics. I gave a paper at the Justice Department. They offered me a job, and I was the data guy at Justice for 8 or 10 years. That's basically how I, you know. Then I learned, at the Justice Department, I learned everything I know about antitrust and competition economics.”
Jon: “That's fantastic. So, you grew up as a staffer. You spent many years as a staffer before coming back to academia.”
Luke: “About 8 or 10 years as an econometrician at the antitrust division of the Department of Justice. Then I came back to, then I got a job off, decided to go on. My wife and I were living in, my late wife and I were living in Washington, D.C., and we were the first white heterosexuals in our neighborhood in Northeast D.C. All the white people, after the 68 riots, after MLK was assassinated, all the white people left everywhere in D.C. and all moved to Northwest. The block breakers were the gay men in the early 80s. When they died of AIDS, then the white heterosexuals moved in. We were the first white heterosexuals on our block in Northeast D.C. There were all these kids there on our block. We were a mile away from the best museums in the world, best free museums in the world, and none of them had ever been. My wife and I, every Saturday morning, we'd take all our kids in our neighborhood and take them to go see all the free stuff around. The most fun we had was taking them to the Tank Museum in Aberdeen, Maryland. There is a 100-acre big plot of land out in the midst of Maryland, and it's just filled with every tank that's ever been produced. It's wild. You take your kids there. They can play on them and play on stuff. It was really fun to see all the old tanks.”
Jon: “Wow. It's so fascinating just how much history you can go see around D.C., and it's amazing how much D.C. has changed just over the decades going from it's become an extremely wealthy place in part because of, I think, some of these defense buildups during both the Reagan and the Bush era kind of went from being what some people would call a CD, very dangerous kind of city to a very nice and affluent one. I'm curious. So, you went back to the FCC and became their chief economist, and you became the chief economist at the Department of Justice and Antitrust Division more recently. What do these positions do? I guess the U.S. is unique in that it's got two antitrust authorities. That’s got the DOJ and the FTC in different industries, but I'm curious, what does the Chief Economist do, what do these economists working at the FTC and the DOJ Antitrust Division, what do they do?”
Luke: “So the Chief Economist has remarkably little power, you know, and part of it is just the federal and the civil service system. You can't fire, hire, promote, reward, punish, you can't do anything to align the incentives of the individuals with the goals of, you know, whatever your goals are. And the staff is largely, you know, they're very suspicious of the political appointees. And the advantage I had, at least the first time I went around, you know, I knew most of the staffers from my time, or I knew a lot of the staffers. And I was, you know, I was a staffer like that, and we'd share, I'd go over there and eat lunch. One of my best friends worked at the FTC, so I had the advantage of that. And so, I didn't, I was able to, we'd have these Monday morning staff meetings, and the bulk of our work was just merger work. So there, you get a every merger has to file a Hart-Scott-Rodino, if there's, if they're big enough, now I think it's about 55 million. You have to file a form with the federal government called the Hart-Scott-Rodino form, and it goes to both agencies, and they decide whether they want to investigate it. If they both want to investigate it, then they have to decide whom, who gets to do it. And there's, there's a little bit of a kind of historical difference, but basically if there's a big merger, and they want to investigate it, there'll be a fight over it. But, that's, and the chief economists, so that's what I did as chief economist. The staff economists, they're organized as functional organizations, and I've actually written a paper about this, and why, I mean, some agencies sprinkle the economists along with the attorneys in the agencies. But Justice and the FTC have this functional organization where you, where the economists do their own work, and they analyze a merger and send it up the flagpole, and they're just, the economist's criteria is, hey, is this merger good or bad? And the attorneys have a different focus. Their focus is, hey, can we win this one in court? And so, they're always at odds, or, if they're doing their jobs, they should be at odds with each other. They should not be cooperating with each other. And, and it was my job at the top of the food chain when the recommendations came up from the economist, and the recommendation came up from the attorney, it would be up to the senior staff members to go ahead and decide, you know, to duke it out, and at the staff meeting, decide, I would, and I would represent the econ point of view, and the, the attorneys would represent the staff point of view. Sometimes I disagreed with the staff, and sometimes the, you know, the lead attorneys, or the, not the lead attorneys, but the attorney managers, they disagreed with their staff too, so, and then the, it would ultimately go to the chairman or to the assistant attorney general, if we're at the Justice Department. So, so anyway, yeah, that was, that was my job as the chief economist.”
Jon: That's absolutely fascinating. So, I want to talk about the history of antitrust for a moment. So, in the late 19th century, early 20th century, we start with all the big trust-busting era, you know, big monopolies, you know, Taye Roosevelt, Sherman Acts, Satan Act, you know, Lewis Brandy's sort of the first era of antitrust policy or any sort of concept of antitrust policy. This is also after the first century of kind of seeing some meaningful amount of economic growth in sort of the 1800s. You get this sort of response to very big firms, the so-called trust, the big banks, the big oil companies, the Standard Oils, and so forth. And then sort of a few decades after the war in the sort of 1970s, 1980s, we see ORC, the Chicago School, the Consumer Welfare Standard, which is very dominant from that period of time in antitrust policymaking and how the judiciary approaches mergers and how the FTC and DOJ approach mergers. And now we have in the sort of late 2010s the emergence of the so-called new Brandeisians. And I'm curious, what do you think about this trajectory, and who do you think has it right here in these sorts of opposing schools of thought when it comes to antitrust policy?”
Luke: “Well, you know, so what I've seen in my lifetime, so I got there at the tail end of – there was actually a Reagan-appointed Stanford professor, and he had a very kind of economics approach to antitrust. When I got to the Justice Department, it was the tail end of Reagan, all through Bush, and the beginning of Clinton. That was my tenure there. And the Reagan revolution – and so I'm kind of passing over the early part, but this is what I've seen. And I don't study history. I kind of have the Joey Ramone view of history that, you know, that's not where I want to be. So, I don't care about history because that's not where I want to be, rock, rock, rock and roll high school. So, I actually don't know that much about the early history of antitrust, but I do know the history of antitrust beginning with the tail end of the Reagan revolution. And economists were firmly in control of antitrust policy, and the attorneys, you know, their memos would come up and the economists' memo would come up, and they often complained that it was harder to get a merger challenge out of the Justice Department than it was to win one in court. And that was largely right, I mean, mainly because the courts were, this was the time of Von's supermarket, you know, when we had the way they analyze mergers, they delineate a market, and if the shares are big, they block it. If it's a four-firm going down to a three firm, it's a four to three, three to two, or two to one merger, they're going to challenge it. But this is the time when they were challenging, mergers that were 20 to 19 mergers. You know, it was just nutty. There was no rationale. There's no economic rationale for it. And the – basically, starting in the 1968 guidelines, you know, they tried to write down some rationale and say, hey, we're concerned with market power, and so – and by the time that got to the 80s, and I don't know the history of it before I got there, but it was – the economists were in charge, and we'd say, look, is this merger going to raise – is this merger going to raise price? And that was the – that was our main concern. And we would build models and build empirical models and – or, you know, kind of look at cross-sectionally if there were markets across, you know, like supermarkets, and we see a three-to-two, we compare a three-firm market with a two-firm market, and is the price higher in the two-firm market, and that would be an empirical model of merger, or we see consolidation in the industry, and there was some data that we could look at, we asked whether price went up. By and large, most of our work was building models. We'd model how firms compete, and then we'd use the model to forecast or simulate the loss of competition following merger, and basically what the major models say is that, four-to-three, three-to-two, and two-to-one mergers are bad, and that's basically the mergers we block, but, we'd have to support that, and it would obviously depend on how sensitive demand was to price. If demand is very sensitive to price, and you try to raise price, you lose a lot of quantity, then, you know, you're making more on the stuff you sell, but you're losing a lot of sales, and that would make the merger unprofitable, and so that's the kind of stuff that, and we estimated demand, and we simulated, had some, you know, built models, and what I'm known for is building these very tractable models that are actually used by the Justice Department. They're online if you want to try them out at competitiontoolbox.com, but that actually allow the government to figure out, hey, how much competition is going to be lost by merger. Now, to relate that back to your question, you know, about the new Brandeisians, that intellectual tradition is still going strong, and the Biden appointees who are trying to turn back the clock to the battle, to what I consider the battle days of antitrust, the 60s, before we had this economic rationale for them, they're getting, you know, they're running into resistance from staff, they're getting mugged by the, they say the conservative is just a liberal who's been mugged by reality. They're getting mugged by reality. Lena Kahn came in wanting to block all mergers, and she's getting, you know, she's losing everything in court, and it's, you could say, well, it's the Reagan judge or the Trump judges, they had a lot of appointees, and, but by and large, it's the development of the common law, and the common law has evolved and again, this is, this is not my area of expertise, I'm just relying on basically my co-author, this guy, Greg Worden, he, he'd spent his whole career at the, at the Justice Department, 40, 40 some odd years, and he's written probably 300 articles on antitrust, and, and written a book about the history of antitrust, and he told me that, you know, that, that the courts have moved, and it's very hard to move back, you can't just come in and say, oh, yeah, we're going to, we're going to turn back, we can bring cases, but ultimately, we've got to win them in court, and they're not very successful at kind of turning the law around and that says it should be, these are enforcement agencies, they're not, they're not there to make law, even though, kind of, that's probably, if you ask Chairman Kahn, that's probably what she'd say she wants to do, you know, like, she wouldn't say that in public.”
Jon: “So, it sounds like, in terms of, like, criteria, so, in this, I think, pre-Lena Kahn, you know, consumer welfare era. DOJ, FTC in terms of how courts thought about this thing, in terms of the mergers that aren't permissible, these are sort of horizontal mergers that are sort of 3-to-1, 2-to-1, 4-to-1 type arrangements. I recall there was sort of like maybe a change that like vertical mergers are thought of very differently, if there's maybe some sort of synergies and that they're vertical mergers being, they're not competing directly against each other, instead they're sort of the different, the two merging firms are at different points in the...”
Luke: “Like the way I would phrase it, look, is when you combine substitutes, so I'm worried that if I raise price, I'm going to lose money to you. I don't, that's what I care about, but if I buy you, that changes my profit calculus. I'm no longer worried that I'm going to lose consumers to you and, or dollars to you, and so if I buy you, that changes my profit calculus and I'm more willing to change, to raise price. That exact same logic, if you use that exact same logic on a vertical merger, if you have a vertical merger between two different stages that's in the same vertical supply chain, they're complementary products. And if you apply that exact same logic to complementary products, then a merger would reduce price. You know, it's like, why do supermarkets own their own parking lots? Well, if they didn't, each of them would try to raise price to capture a share of the shopping dollar, because you need to consume both, the grocery store and the parking lot in order to buy food. And when they compete, they raise price, and if you let them merge, then they lower price. And so it's the exact same logic that leads us to challenge horizontal mergers, would lead us to kind of let vertical mergers go through. Now, having said that, there's a little bit of complicated, the one part of the story that I didn't get was, well, what happens to rivals, non-merging rivals? And there is this effect called raising rivals' costs that complicates the vertical analysis. But the, you know, the first order merger effect is the exact opposite of, in verticals, the exact opposite of horizontal.”
Jon: So, by rivals, do you mean things like collusion and price fixing? Like, I remember...”
Luke: “Oh, no, no. I mean... It's still different. So, when I was at the Justice Department, we brought a vertical merger case, and, you know, I'm not allowed to divulge the internal deliberations of the Justice Department, but, you know, given what I just said, you can probably figure out whether I thought that was a good case or not. We blocked AT&T Time Warner, and Time Warner had content, AT&T had distribution, they owned DirecTV, they were everywhere. And our theory of the case, and I supported it, when I was at the Justice Department. It was really, really fun to watch the first vertical, litigated vertical merger case in over 40 years. And our theory of the case was that the merge firm, AT&T Time Warner, would raise the price of Time Warner content to non-merging firms like Comcast, and that would be an anti-competitive effect. The pro-competitive effect, of course, as I said before, that when you combine complementary products, there's an intent to lower price, and sometimes that's called the elimination of double marginalization, it's more nuanced than that, but that's the offsetting effect in figuring out how to balance the rating rivals cost against the vertical coordination or the vertical alignment of incentives, which is pro-competitive, it’s very nuance very difficult and those cases are a mess. You know that I [Inaudible] the judge who decide that, they did their best but I mean, this is really nuanced kind of economics, you know, this PhD kind of style, nuanced stuff. And, you know, I've built these vertical merger models, and just explaining them to somebody, with non-merging, right? If you don't have non-merging rivals, they're pretty easy. But if you have the non-merging rivals, then you need game theory, and again, I build those models. My closest colleague is a mathematician who's interested in, we build these game theoretic models for competition and simulate the loss of mergers. And I can tell you from working with these models that the vertical mergers, the simple logic of vertical mergers, they're almost always good. It's really hard to find a vertical merger where the effect on non-merging rivals offsets the benefits to consumers of the vertical coordination between the merging products.”
Jon: “Fascinating. I'm curious. I'm going to press you a little bit on these models.”
Luke: “Okay.”
Jon: “I'm fascinated by this topic in general, and it is a very, unlike other areas of economics, I feel, like, you know, applied microeconomics, think about tax policy. You have these sorts of natural experiments where you can look at a tax change, you have, say, a treated state that, you know, lowered tax rates and a controlled state that didn't, and you sort of look at the differences, on various outcomes, like, you know, revenues or growth or something like that, and you measure those to your treatment effect. And you can do that in a lot of places, but competition is just inherently a very difficult and sort of very model-driven type of economics. Industrial organization, even though we call it, you know, the new empirical industrial organization or the new empirical IO, sort of revolution of the 80s and 90s, it still relies on, even though it's empirical, it still relies on a lot of model assumptions. So, like, I'm curious, you know, when you think about, measuring consumer welfare, you know, this is the whole idea that you can reduce deadweight loss and increase, both, you know, consumer surplus and producer surplus by making sure that we have competition and that we don't have monopolistic firms. That's sort of the theory. But, I'm curious, like, do you use, in terms of your actual metrics and data, do you use things like, say, Hirschman indices to measure the impact on concentration in an industry? Sometimes it's, like, vague what your denominator is or how big is your industry. You know, for example, you know, if you look at something like Google or Google Search, Google, dominates, say, 90-some-odd percent of the search, online search category. But if, you know, you were to say that the real industry that Google is in is marketing and marketing revenues or something like that, it's probably the very small share of sort of marketing industry revenues. But I'm curious, are there other things like...”
Luke: “I've got to say something about the Google search case, so I'm not working on this case, so I can freely opine about it, but it's got a huge problem. So first of all, they have to delineate a market narrowly enough so that they can say to a judge, this merger is going to eliminate competition in this market. And they've delineated what they call a general search market. So, Google, Bing, and Yahoo. DuckDuckGo is really tiny, but those three. And Google has probably 80, high 80s, 90% of the general search market. And one of the things that the repudiated merger guidelines by the Biden administration says, we're concerned about market power. And you think, okay, first of all, how would you look at market power in, how would you exercise market power in search? Well, you'd raise the price of advertising, which gets you right back to, okay, is this a market for advertising? And if they raise the price, is it a market for targeted advertising? And so, they've defined this general search model, which is very narrow. But here's the huge problem. This is with monopolization cases. So, the antitrust laws, there's three parts of the antitrust laws. One is horizontal collusion. You don't rig bids, allocate customers, or agree not to compete with your rivals. But if you do, don't use the phone, because it's so unbelievable. Whenever we prosecute these criminal cases, people call each other up before auctions. First thing we do is subpoena the phone records, and they always kind of put them in a prisoner's dilemma, and they always kind of fess up. But the other part is horizontal mergers, price fixing, those are kind of loss of horizontal competition. That's fairly easy to understand. The monopolization cases are much, much harder to understand. So first of all, you've got this paradox that we're only going after really successful firms, so Google's a really successful firm. How did they get that way? They innovated more than anybody else. I mean, I remember when I was back at the FTC, Google came in, gave a presentation to us, and they talked about how they developed their search tools. And so, what they do is they find something wrong with their search tool. An example they gave us was, you know, when you search for Essex, they came up with porn sites. And they say, okay, something's wrong with our algorithm. And it's a very complicated problem, but they've simplified it down to, you know, when you put a query in, they put it in their own language, and when they turn that language into a search, they have this reduced set of instructions that give you an answer in less than a half second. I mean, if they can't do that, people won't use it. And so, they've got this pretty complicated system. And so, when they find a problem like Essex, they play around with the code, and they figure out what the problem is. And then they A-B test it. So, then they take a bunch of queries, and they send them, you know, they're coming to Google Search, and they send them to the new and upgraded Google Search. And they send some to the old search, and then they show a bunch of people on Amazon Turk. You know what Amazon Turk is? So, they have paid people who look at the two searches and say, well, this one's better or this one's not better. And they won't adopt an innovation unless if it lowers the quality of the search. And, you know, they've been doing that for 20 years. And it's no wonder they have the best search engine in the world. And because they've been working hard on it than anybody else. And the fact that they're big, I don't know, it's just there's so many problems with that thing. And here's the killer, but here's the killer. So, if you look at all the big monopolization cases that occur once a generation, you look at the IBM case, you look at the Microsoft case, you look at the Google case, you know, those are all monopolization cases. There's no relief. What are you going to do? I mean, and what are we going to do to Google? We say, hey, you can't have such a good search engine. You've got to give some share to somebody else. Or in Europe, they gave people a choice. You know, they say, hey, you can't be the default engine on Android. And so, they gave consumers a choice. And what did consumers choose? They all chose Google because everybody knows it's better. So now you're the antitrust authorities. What the F is Justice Farmer going to do if they win this case? And there's no relief. If you look ahead, reason back, Chapter 5 of my textbook, you know, what everybody should do is look at, okay, what are we going to do if we win this case? Well, they don't know what they're going to do. There's no relief. And if there is no solution, there is no problem. And that's the real problem with these cases. There was no solution in the IBM case. There was no solution in the Microsoft case, and there certainly is no solution in the Google case and it’s not going to make everything, it’s not going to harm consumers. [Inaudible 0:37:56] If I were a conspiracy theorist, oh, they're very clever, they get rid of this consumer welfare standard so they can break up Google and not be accused of harming consumers. We don’t care about that, I don’t know.”
Jon: “So I want to get more a little bit to the details of how exactly measure the impacts on consumer welfare and the types of models. I want to push you a little bit on the types of models that you're using. Because, you know, I.O., Empirical I.O. relies on models, model functions, unlike the applied micro, where, you know, you have sort of like a treatment and control group, and the treated area receives the tax cut and the control group doesn't. You can look at various outcomes like output or you can look at things like tax revenues and so forth. In I.O. and antitrust sort of economics, you're making some, you're relying on some sort of assumption. What does the demand curve look like? You know, what does, how do you measure the amount of consumer welfare that's gained or lost? You know, the whole theory kind of hinges on, you know, that a monopolist has a demand curve that looks a certain way. A monopolist has, you know, demand curves that look a certain way in a perfectly competitive market. You have demand curves that look in some way. You have to measure demand curves somehow, and you have to measure, cost curves somehow. I'm curious, you know, do antitrust authorities, how do they, like, measure markups when you have difficulties measuring things like marginal cost? You know, it's not easy to attribute marginal cost. Certainly, if you're working with a multinational company, things can be a little bit difficult. How do you measure opportunity cost? Do you assume things like Cornell competition? You know, that has a big term. Do you think like BLP, you know, Barry Levinson?”
Luke: “That's a great question. That's a great question. So, when you're doing model-based inference, you can use the model to, you basically fit the model to what you can observe, which is the current data. And typically, you don't have very good data. You don't have the time or luxury to run a complete generalized demand system. You know, you're lucky if you get an aggregate elasticity of some shares, and that's about it. But in a lot of kind of effort to, you know, on the demand side, saying, okay, let's try to develop these really cool conceptual, these flexible functional forms and we'll develop these really interesting ways to estimate them. There just isn't time in a merger investigation to do that, and it doesn't matter. You know, the times when we've been able to get a sophisticated demand model like BLP, and estimate it, and then simulate the effects of the merger, it just didn't matter that much. And one of the things you also asked about was you relative just a simple functional form like the logit model, which has, you know, just two parameters. And so, you give me some prices, shares, and aggregate elasticity of demand, and I can simulate a merger. And if you tell me that, you know, these two products are closer than others, put them in the nest, and you give me some, you know, accounting data, and I can calculate, you know, price-cost margins, and I can fit the model to those data. Or, you know, if I don't have those data, I can assume a competitive interaction, write down all the first-order conditions that define the Nash equilibrium, and then use those to back out the margin of the unobserved marginal cost. And then so in the first, that's called calibration. It's not really estimation, but you calibrate the model to the shares and the prices and the elasticities. And, you know, in any margins you can observe, you know, it's maybe four or five pieces. The shares and the prices, say it's a five-firm industry, so that's ten pieces of information. And then you get an elasticity, and that's another piece of information. And you can use all that information and back out the parameters of the model. And then you say, okay, well, what about if these two firms are closer substitutes than each other? Okay, we'll put them in a nest that adds another parameter. You give me a margin, and I can back out that nest parameter.”
Jon: “Fascinating. I'm sure there's quite a little bit of art as well.”
Luke: “Modeling is an art, yeah. Yeah, it's a real art, and you've got to be flexible enough. But one of the kinds of the really cool things about consulting, I'm going to plug in for consulting to all the grad students or any grad students that are out there. When you're an economist, an academic economist, you choose the questions you answer, and you choose questions that you can answer precisely. And so, the questions that we can answer precisely are really, really, really narrow ones, and nobody gives a damn about them. But the ones that people care about are the, you know, kind of the less precise, like is this merger going to raise prices? And typically, there's not enough information to run. There's no good natural experiments, either cross-sectionally or with time series data, or the data are not available, or it's in lousy condition. And so, you really have to go to a model-based counterfactual. And so, you say, okay, the observed world is what I can observe, and the counterfactual is the post-merger world that I can observe. And once I calibrate the model, you know, to the observed world, I can use it to forecast or simulate the unobserved counterfactual, which is the post-merger world. And, you know, sometimes the counterfactuals are different. We observe the monopolized world. We observe the monopolized world or we observe the collusion world and we want to know what would have happened if these guys were competing. Well, you know, we can calibrate a model to a collusive observed equilibrium and then back out, then use those parameters to figure out what the equilibrium would look like if they were all competing. And, they're basically Nash equilibrium models and we've developed these models that work for bidding, for bargaining, for price setting, for quantity setting, for advertising, for capacity constraints, you name it. I mean, that's what we do. You know, we build models, tractable models that we can calibrate to just a few pieces of information for all different kinds of industries. And that's been my professional life. It's really fun, it's really interesting, and it's fascinating.”
Jon: “Well, it's fascinating, you know, this whole approach, I think, to using models to think about, you know, how competition impacts consumer welfare output. The whole idea, it all goes back to that, you know, theory about if you have a monopolist compared to the perfectly competitive case, you have less Q or less output. And this is sort of everything in between that, trying to think about how mergers impact things like output and consumer welfare. I'm curious, I want to just ask one sort of question here on big tech and sort of national security, which I think is a little bit different, I think a bit of a newer front in these discussions. And it's different, I think, from just the simple, say, Lena Kahn approach of, you know, everything that's big is bad, or, for example, the approach of, you know, they've gone after the Microsoft acquisition of Activision, Blizzard. I think the big tech sort of national security thing is a little bit different. And I think that's one area that certainly irks some Republicans and quite a few Democrats as well, which is, even though tech firms provide lots of consumer surplus, you have all these apps like Gmail and Facebook, they give away for free. And there's some caveat that they're selling your data and there's been some property rights issues there. They do sort of have a tremendous amount of political power. Twitter censored the [Inaudible 0:46:32] near-post laptop stories or false news for the 2020 election. It turns out, you know, that story was actually true. You know, Apple is now canceling Jon Stewart's show, apparently, because it is talking negatively about the Communist Party of China. There's some argument, which is really a political argument, not an economic one, that, you know, these companies may be influenced or captured by foreign enemies. In that sense, are Google and Facebook, you know, too large and too powerful in a political sense? You know, should political national security considerations, matter when it comes to antitrust regulation? Or is this really more in the domain of something like CFIUS rather than, say, the Department of Justice and the FTC?”
Luke: “Yeah, I mean, I think that there are valid privacy concerns. And the tech giants are doing their best to address those privacy concerns. Because obviously the GDPR has been, oh yeah, one thing you did not ask me about is the difference between China, the US, and the EU. And oh my gosh, we'll talk about that in a sec. But the GDPR has almost killed innovation in Europe. What is just amazing to me is the lack of innovation in Europe. I mean, if you look, there's a lot of metrics that you could use. You can use the number of unicorns. There's actually a guy at Stanford, Ilya, I forget, he's got a Russian name, ST, last name. But he collects data on unicorns. He'd be a great guy to interview on your show. The number of unicorns in Europe is pitiful. I mean, they really undercook their coverage. Or relative to the US, you look at the age of companies on, say, the CAC 40. What's the age of the top 10 firms in Europe versus the top 10 firms in the United States? They're all over 100 years old in Europe. In the United States, they're 25 years old. That's the average age of the top 10 firms. And I think that just goes to the dynamism of our economy and the huge advantage we have over Europe. And the question that I kind of, you know, against that backdrop, to get back to your question, is like, whom do you want, deciding what the future looks like? Do you want these firms that are beholden to their customers and that are constrained by competition? Do you want them deciding, or do you want some government official deciding, like Lena Kahn or me? Do you want us deciding what should go down? And I think the answer is absolutely clear. And if we want innovation, which I think, you know, in your grad school classes, you learn about the solo growth models and technological growth. And this is growth that can't be explained by growth in labor. You know, think about a production function. You've got output as a function of capital and labor. You know, and then you try to fit that model to the data. And you see there's a huge component that's unexplained by either capital or labor. They call that, you know, total factor productivity. And you'll learn about that, you know. That is the technological innovation that drives our standard of living. I mean, we have grown, you know, and what makes us the richest country in the world is that we started growing a lot faster than anybody else a long time ago. China has been unbelievable since 1978 when they got private property and got this version of capitalism. They've grown really, really fast, and their standard of living has gone up really fast. But it's this incentive alignment kind of letting the markets do their work. And President Xi, I think, has probably killed the goose that laid the golden egg by attacking all the growth firms in China. But I mean, the counterfactual to that thing is, OK, yeah, you might be right that there is some threat to national security. But I mean, you can always – I mean, but what's the counterfactual? So, do you want the government controlling these firms? I mean, the cost, I think, would just be enormous. And always think about, well, what's the alternative? Well, so we got a bunch of government bureaucrats deciding what these firms can do. And there's always a balance. And when we go from administration to administration, it's a little movement left, right. But basically, in antitrust, we agree on what the goal is. The Biden administration has been an outlier in that. They want to turn back the clock. And the institutions of government are not letting them. The courts are not letting them. And Congress isn't letting them. And that's the beauty of our checks and balances. You can't just come in here with an overzealous – what I would call a zealous or – actually, I've never heard that word without overzealous over in front of it. But a zealous enforcer who thinks that way too many mergers and that's hurt, harmed American consumers. I think that's just a radical misreading of what's happened. And we are so frigging lucky to live in the United States where we have this relatively open, innovative economy because that's what's making us rich. And let's not lose sight of that. And all you have to do is look at places like Europe where they – if you want to innovate in Europe, first of all, they can't tolerate the kind of inequality that comes from innovation. They couldn't tolerate Jeff Bezos in Europe. So, Jeff Bezos wouldn't be able to innovate in Europe, not to mention the regulatory miasma that makes it hard to move assets to higher-valued uses. It's just – it's too difficult. And I think when I talk to my colleagues in Europe, that's what they're concerned about. You know, the, the regulatory overlay that's killing growth. I mean, they just don't see the same trade-offs that we do. I mean, when you, you talk about, you know, your question kind of implying, hey, we need more government and, and I'm immediately thinking about the trade-offs and, and I'm thinking, oh my gosh, what, what, what are we, what would be given up with, with more government control? And they just don't think about trade-offs that way. And, and President Xi, I don't think does either. And I think that, and you haven't even mentioned the really big elephant in the, in the room, which is, which is declining birth rates. I mean, you know, China's dead, Europe's dead, and we would be dead, but for the immigration. And you know, we've got lots of immigration, so we still have a decent growth pyramid. You know, China's growth pyramid is, is starting to get inverted as is Europe's. And you know, the ironic, your, your unintended consequence of the illegal immigration for letting, a lot of obviously it is the right way to do that. We want, you know, security checks, make sure we're not letting in terrorists, but the benefit of letting a lot of kind of young, healthy people in the United States, you know, in the prime of their life, that's going to save, save our aging population, saving us from the kind of demographic death that is China and the EU anyway.”
Jon: “And not so much, I mean, legal migrants too, you know, certainly migrants from India in particular, tend to be much more innovative in terms of their patenting output than, than say the, the average, you know, American who's, who's born here. So, it's fascinating. And there's been, I mean, there's been a lot of work that's been done on, on the connection between immigration and innovation by folks like Bill Kerr and others. And I fully agree with you that, you know, I think these antitrust regulations closely aligned with economic growth challenges and innovation. And it seems, I totally agree with your analysis, but so far it seems like, you know, the new brand usually in a con more that has not been successful so far and that they've just continued to lose cases that they've brought that certainly the, the judiciary, the FTC DOJ staff, and much of Congress is very much opposed to their attempt to overturn the, the consumer welfare standard.”
Luke: Oh, so I've been thinking about this, so I've been actually modeling the, the antitrust process and kind of looking at what, what Chair Khan is doing. And they, they have made the hard, there's filing, there's filing fees that you have to file before you, you even think, and they've made that, made that thing really long, onerous and kind of open in it. And it's really hard to gather the information. They've quadrupled the filing fees. And if you think about that as a tax on mergers, kind of think about there's good mergers and there's bad mergers. Well, we're deterring good and bad mergers, so if you're, if you think that most of the mergers are bad, you're thinking, hey, let's tax the hell out of the mergers. But if you think that, hey, there's that most of the mergers, you know, 98% of the mergers that come through the Justice Department never even get a second request. So, most of those are good. We're deterring a lot of good mergers, you know, called those type one errors and instead of filtering and, you think about, I want, I don't want a tax that, that deters all merger. I want a filter that distinguishes the good from the bad. And what this agency has done is they've dramatically raised the cost, they've dramatically increased the uncertainty of the process, and the cost and the uncertainty are deterring all mergers. I think that's, you know, you think about the, you know, what's the wealth-creating engine of capitalism? It's the movement of assets to higher-valued uses. And if our biggest and most valuable assets are corporations, and if we can't move those to higher-valued uses, we are going to, you know, kill the goose that laid the golden egg.”
Jon: “Absolutely. I fully agree with that. This has been such an interesting conversation. Luke, it's been a real honor to have you on. Thank you so much for joining us today.”
Luke: Oh, I enjoyed it, and thank you very much.”
Jon: “Today, our guest was Luke Froeb, who is a professor of management at the Vanderbilt University Owen School of Management, and the former chief economist at the Department of Justice Antitrust Division, and the former chief economist at the FDC. This is the Capitalism and Freedom in the 21st Century podcast, where we talk about economics, markets, and public policy. I'm Jon Hartley, your host. Thanks so much for joining us.”
