Big Tech and Antitrust: A Discussion With Randal Picker

Randal Picker is the James Parker Hall Distinguished Service Professor of Law at the University of Chicago Law School, Senior Fellow at the Computation Institute of the University of Chicago Argonne National Laboratory, and affiliate faculty with the Coase-Sandor Institute for Law and Economics.

Professor Picker currently teaches classes at the Law School in Secured Transactions and Antitrust and a seminar on antitrust and intellectual property policy. In prior years, Professor Picker has taught Network Industries, Bankruptcy and Copyright; Technology, Innovation and Society; Corporate Reorganizations, Commercial Law and Civil Procedure. He has also taught seminars on Game Theory and the Law and The Legal Infrastructure of High-Tech Industries.

Scott Wallsten: Hello and welcome back to TPI’s podcast, Two Think Minimum. It’s Tuesday, July 16th, 2019. I’m Scott Wallsten, President and Senior Fellow at the Technology Policy Institute. Today we’re excited to talk with Randy Picker, who is the James Parker Hall Distinguished Service Professor of Law at the University of Chicago Law School, Senior Fellow at the Computation Institute of the University of Chicago Argon National Laboratory, and affiliate faculty with the Coast Center Institute for Law and Economics. Randy teaches classes at the law school on secure transactions and antitrust, and a seminar on antitrust and intellectual property policy. He’s also taught network industries, bankruptcy, copyright technology, innovation and society, corporate reorganizations, commercial law and civil procedure and courses and seminars on game theory and the law and the legal infrastructure of high tech industries. And that will serve as a foundation for the issues of the day, which are primarily about big tech and antitrust, which is a hot topic here in DC and elsewhere. I’m going to now hand it over to Tom Lenard, who is the TPI Senior Fellow and President Emeritus joining me and Randy, and Tom will start the interrogation.

Randy Picker: Wow. Okay. Go ahead.

Tom Lenard: Thank you Scott and thank you Randy for agreeing to participate in this interrogation. I will dive right in. So one of the things we see now is that the polarization that we see in politics generally, that’s kind of obviously has spilled over into what used to be the relatively specialized and arcane field of antitrust. Does that surprise you at all?

Randy Picker: You are right to say that some of us used to live these almost priestly like lives, and occasionally people would show up just sort of like sometimes people only go to church on say, Easter and Christmas. But now, yeah, obviously these areas are getting a lot of attention. I think what is, part of what’s driving that I think for sure is, is that fact that we have these very large tech platforms, they are not removed. So when the government chased IBM in 1969 and started chasing IBM in 1967, IBM was at that point the most valuable company in the world by capitalization. My guess is most people on the street only had a passing familiarity with IBM. The companies that we’re talking about right now are all companies that we have in our hands, multiple hours a day. And so I think that gives it a kind of public salience that it might not have had in the past.

Tom Lenard: But do you think that the experience of most people with these companies with would lead them to be critical of them?

Randy Picker: Well, that’s- so that’s a good question. So I think you’re right to say that you can look at these companies and man, this is me in the main and say these are wonders, look at what I get for from Google for free, as it were. We can talk about that characterization in just a little bit. I think as they’ve moved into business models which depend a little bit more on getting information, I think that’s given them a slightly different posture. Certainly Facebook’s roll or non-roll in the 2016 election, different posture. I think the complaints about Amazon and I gather we’re seeing more of those literally today; it appears maybe that the European Union Commission’s going to move forward, I think that’s driven by competitors and much less about the public.

Tom Lenard: A common theme generally today, and this is espoused by a variety of people, including some mainstream economists, is that the US has a market power problem, that the economic forces today mirror the industrial concentration and economic inequality of the turn of the last century. My question to you is, do you think the U.S. has a market power problem, either generally or specifically with respect to the large tech platforms?

Randy Picker: Yeah, so look, when I talk to my students, I teach a class at our business school, Booth, sort of a general law class for MBAs and I say we’re going to do a day on the antitrust. You should understand antitrust. That’s a problem you want to have, right? That tells us that your business has succeeded in a huge way such that people think you’re worthy of a particular kind of attention. So when you ask the question, does the United States have an antitrust problem, I think you want to come at that from a couple of angles. I think we should think it’s a remarkable success story that we have these leading firms primarily in the United States. So obviously China, I mean, 10 of the largest firms by market capitalization I think include Tencent and Alibaba. So China’s moving up as well.  But the fact that these companies are enormously successful companies based in United States, that’s something we should at some level celebrate, and then figure out what to do about it, if anything. Now the question is whether there’s a problem. I think that harder and I do not think- I think these companies are interestingly different. And so I think to try to talk about them in some sort of a grouped way is sloppy and maybe confusing. So I’ll try to resist doing that.

Scott Wallsten: It makes it even more sloppy and confusing to try to figure out how people’s concerns over privacy became part of the antitrust discussion too. Why, I mean, how do the fears that people have, setting aside whether there are harms or not, credible harms or not, how did those end up being put into the antitrust discussion? Why do people think that antitrust is a tool for dealing with those issues?

Randy Picker: Sure. I want to say a couple things on that. I do think as we’re thinking about, about discussing- I’m going to do what I said I wouldn’t do- I think as we’re talking about discussing these entities, I do think antitrust is only one regulatory tool and I think we’ve got a broader set of regulatory tools that we probably want to- we’re going to think about this seriously bring to bear. So I taught my Platforms and Network Industries course in the spring here at University of Chicago Law School, and that’s of course that starts with the Telegraph and the railroad and ends with Google and Facebook and Amazon. I think that’s the natural arc there. Antitrust is a fault based system in the United States and in truth in the European Union as well. We have to find you did something wrong.

The different regulatory approach is to say, look, it’s not as if these companies have done something wrong. We’re just not necessarily where we think we want to be from a competition standpoint and we’re going to impose utilities style regulation that asks you to answer your privacy questions since that’s where you went in specific. Look, I think we can talk about a variety of characteristics associated with good. Price, actual cash price, is only one of them. When I talk about how a company like Google would exercise market power against consumers, I say, look, this is a very old fashion business. It’s a media business. We’ve been dealing with this in the United States since the onset of radio. We had free advertising supported companies back then. And the way a company with market power in that space exercised it is you impose more ads than you would see in a competitive framework. Similar fashion- look, I think data is, is very hard and I think it’s not particularly salient, not particularly transparent. You could ask the question, how much data transfer would we see in a competitive outcome and how does that compare to what we’re seeing with when someone has market power. That strikes me as a perfectly sensible antitrust question to ask and try to answer.

Tom Lenard: Sticking on the subject of data, of course, one of the major arguments of critics of these companies relates to data and the fact that these companies have amassed large amounts of data. And the argument is that this gives them a, this gives them an advantage and it’s essentially a barrier to entry. What’s your view of that argument?

Randy Picker: I’ll say, gosh, I’m not sure. I could easily imagine it does give them advantages.

Tom Lenard: I guess the question is does it give them an advantage that’s an antitrust problem?

Randy Picker: Good. Yes, exactly. I want to separate those two. Those advantages might be things which are economies of scale and scope and the like. And your point there, Tom, is that’s not an antitrust issue. Again, you could decide from sort of a pure regulatory standpoint to do something about it, but it’s not obvious that that’s an antitrust issue.

Tom Lenard: And also just there’s a learning by doing aspect of this, which is also a bit of a competitive advantage.

Randy Picker: That’s what partially gets you to the regulated industries question, right? So if you looked at Google and said, well, that’s sort of a learning by doing business. That is, you observe people run searches and you see what they like and don’t like, and you adjust the search results based upon that. That’s a business where you’re improving your product through the use of data- that’s super valuable. But then also goes, oh, well that sounds like one where you could have a natural monopoly fairly quickly. And when we have natural monopolies, we don’t look at those through the lens of antitrust, we look at those through a different set of regulatory tools.

Tom Lenard: I want to get to this issue of regulation a little bit on because I think it’s very important and more people are starting to talk about it all the time that even a strengthened antitrust regime is not sufficient and we need something more. But before we get to that, if you look at the performance of the antitrust agencies, the Department of Justice and the Federal Trade Commission, not just in this administration but really over the last several decades when digital companies have been important, what do you think the record is? Do you think they’ve performed generally pretty well, not too well over enforced, under enforced? What do you think?

Randy Picker: That’s a good question. I should know the answer to that look. I mean, I guess when I think about the way in which antitrust is dealt with technology companies, I start earlier. I start with really AT&T and IBM. It seems to me those need to be- the Law Review at Chicago held a symposium on these issues and I have a paper discussing some of these issues that I need to turn in the next draft in the next five days. I think when you look at that record ideation-

Tom Lenard: So you should have your answer.

Randy Picker: Well I hit my 10,000 words, I didn’t have to have an answer. I think that the record on AT&T and IBM is a mixed record, and a complicated record. AT&T invents the future in 1947. What that means is right after World War II, they invent the transistor and we’ve just figured out how to build these digital computers, ENIAC. ENIAC went live in 1946 but that was all based on vacuum tube. AT&T figures out how to do transistors in 1947 and the whole digital world and technical world that we live in today basically started at, I think it’s fair to say, at exactly that point. The government sues AT&T for antitrust in 1949. I think as best as I can tell that lawsuit had nothing to do with transistors at all. It was related to the fact that AT&T was a very big entity. It was hard to figure out how to regulate it and at the same time have Western Electric selling at all this unregulated equipment at what prices. And eventually we get the 1956 final judgment. And the 1956 final judgment puts in place two really important rules. One is that AT&T basically has to give away a bunch of its patents, including the transistor patents, and AT&T gets blocked from going into computers. Those were the two results. And I think it’s fair to say- I’m in the middle of a book on all of this, which is why I know a lot about it- I think it’s fair to say that the development of technology with regard to semiconductors really is powerfully shaped at that moment, though again, it’s very complicated. AT&T had military contracts, which almost certainly would’ve required it to engage in a certain amount of second sourcing, and they start doing voluntary licensing in the middle of the antitrust lawsuit even before the ‘56 final judgement.

So what did antitrust do? They’re hard to say. We chase IBM in 1969. That was the third time chasing IBM. The government chased IBM in 1932 with regard to punch cards. Complete success in winning that lawsuit goes to the Supreme Court when again, the government is triumphed, except for the fact that they bring another lawsuit for punch cards in the 1950s because the market hasn’t changed, notwithstanding their antitrust victories. And they enter into a settlement with IBM on the day after the AT&T settlement. IBM releases the system 360 in 1964, which is this huge successful innovation in mainframe computers. Great, enormous important computer platform, and then get sued in 1969- government’s investigating them in 1967, that suit goes on for 13 years. Bork, as you probably know in his book, The Antitrust Paradox, calls it the department’s Vietnam.

But even that’s I think a little complicated because one of the core things in that lawsuit was that the government was unhappy with how IBM was bundling together hardware, software, and services. And six months into the lawsuit, IBM says, okay, we’re going to change how we do that. And so they start to separate those out. I don’t want to say that’s necessarily the point at which the software market really accelerates, but I think there’s some way in which that’s true and that suit is dismissed. As you undoubtedly know, in 1982 they announced the breakup of AT&T because there was another lawsuit against AT&T. They announced that break up and that they’re dismissing the IBM suit on the same day in January of 1982. But at the same time, IBM, I think it’s fair to say as best I can tell, when they designed the IBM PC, which they released in ‘81 I think they made that more open and they relied on more external components because of the presence of a lawsuit. Okay. Is that success or failure? I’ll stop. I don’t know.

Scott Wallsten: If we can say one thing from all of that, it’s these huge investigations take a long time, a decade or more. Do you think there’s the kind of will for the government to do that again? Do they want to get into another Vietnam? And the second thing also is even in those cases, they were at least able to define a relevant antitrust market. And we don’t see that here.

Randy Picker: I was reading a decision in the European Union about Intel the other day. It’s from 2017. They started that investigation in 2000. 17 years is fast, right? We’ve completely revolutionized, the technology arm processors are in all of these devices, so the world has changed. 17 years, and they said they weren’t done. So I’m completely with you in saying that those cases take a timeframe that doesn’t begin to match the underlying industry. And it’s sort of maybe interesting and fun to litigate over history, and here I genuinely mean litigate over history, but that’s probably not what we expect of our public policy people.

Tom Lenard: So you are a resident at the University of Chicago where you spent most of your professional career and obviously, as you know, the University of Chicago and the Chicago School have been very influential in antitrust. And now there is a reaction to that sometimes called, obviously not the most complimentary phrase, hipster antitrust.

Randy Picker: Sorry, I laughed. I knew what you were going to say. Go ahead.

Tom Lenard: Proponents of that view, there are a number of things that they believe, but one of the things they believe is that the consumer welfare standard, which has essentially been the guiding criterion for antitrust for decades, basically should be abandoned in terms of something different and probably something broader. What’s your view of that?

Randy Picker: A couple thoughts on that. I’ll do footnotes first. So with the Chicago Law Review Conference I mentioned, Bill Kovacic, who I’m sure you guys know, was there and gave a paper and he said, you guys are getting your history wrong. There’s a parallel Harvard School, Areeda and Turner and more that the world should see as influential as the Chicago School. If you want to blame Chicago, blame Harvard, you want to celebrate Chicago, celebrate Harvard. I don’t know, but I completely hear the point. So I just want to make that point the start. Okay, second point then. So let’s talk about consumer welfare standards.

Tom Lenard: But the preponderance of the hipster school are not into such nuances.

Randy Picker: And part of that is, and we had this conversation at the conference, is that Bob Bork is a fabulous villain, right? He looks like he should be in a Bond movie and should have a white cat in his lap. That’s what he looks like. Having him as your villain rather than then Philip Areeda, who seems like a nice guy, that’s a much better rhetorical move. I don’t know if it’s about nuance versus strategy. Easier to beat up on the Chicago School than to go after Phil Areeda, I think. I’m not sure what’s fair on that, but in terms of the consumer welfare standard itself, I don’t know. I feel sometimes like a heretic when I say this, but I teach antitrust each year and I teach cases, I don’t know how many cases is that standard determinative- and Bork, you’ve got the book’s published in ‘78 and he does slightly revise it, it’s not a revised edition, but he puts a new forward on and a new conclusion on in maybe ‘93. He says, look, the consumer welfare standard means sort of using economics and antitrust. And if that’s what it means, then I go, yeah, it triumphed, and yes, that’s what we do. And maybe, you know, the neo-Brandeisians, if that’s what they are, maybe they want to abandon economics. I don’t know.

Tom Lenard: I think they probably do, yeah.

Randy Picker: Then, I go case after case and if what we think we should be doing is completely broad public policy about the structure of production, I go I’m not sure Article III judges are well situated to do that. That seems to me exactly the kind of thing where we should either let democratic values succeed or fail, and that means Congress. That’s just me.

Scott Wallsten: So if you don’t want to go that far and we do want to say that economics should still be a part of it, because of course as an economist I can’t be that open minded, if it’s not the consumer welfare standard and you still want to retain economics as the basis of decision making, how does the consumer welfare standard compare to, say, a more European style of looking at the state of competition?

Randy Picker: I suspect every day in class, people joke about playing Bingo in classes, what are the magic terms people use? And I certainly use a lot of economics in class, no question about that, that is the language, but my guess is day by day I’m talking about the ways in which competition has been distorted. I have this vision when I teach of sort of what we think competitive outcome looks like, what we think someone who wins in the marketplace. It goes back to where we started in this conversation. I think Google wins in the marketplace and we can talk about how that’s evolved, but they’re enormous, great initial success I think is completely legitimate. That was the premise of the, not the second Microsoft case, but the first Microsoft case, the case relating to DAS licensing. The government said they’ve got this position legitimately. How they actually got that position is a much more interesting story, but in any event, it’s not as if you’re saying, well, what does this mean directly for consumers that often, I think. Other than saying, how is competition been distorted where we think the assumption is, is getting competition, right benefits consumers.

Tom Lenard: There’s a couple more topics I would like to see if we can cover in the short amount of time that we have. One of them is merger policy. Now one of the complaints by the critics of antitrust now is that merger policy has been too lax. In particular, they point to things like the Facebook acquisition of Instagram and say well, even though Instagram was not in the same market as Facebook, it could’ve been, it was a potential competitor and therefore acquisitions like that should be barred. Is that something that can be done consistent with antitrust the way it is now?

Randy Picker: I don’t know how, I mean- I gave a talk in Paris in March and talked about Facebook. I know what I said then, and what I said then is, look, those mergers were reviewed in the United States and were reviewed in the European Union. The WhatsApp EC ruling is an extensive ruling. It’s not as if the antitrust regulators were not paying attention. They absolutely were. I think what you’ve just said, Tom matches their characterization of it. They’re sort of in different markets. What I said in Paris was, look what maybe that misses is that I think in these dynamic markets, the hardest thing to do is to get this powerful group of people dealing with each other, going together to have a platform and to have all of the kind of dynamic network effects associated with that, network externalities, we can all throw out the buzzwords associated with that. Instagram had that, WhatsApp had that. That in some sense is the hard thing in that those three companies, Facebook, Instagram, and WhatsApp, were in those businesses, even if the particular manifestations weren’t necessarily directly competitive. What’s easy for these platforms- we’re seeing it right now. This isn’t easy, but man, look at the scale they can do. As Facebook’s obviously said, oh we’re going to do cryptocurrency and we can do that at a scale of 2 billion. And obviously there’s huge pushback on that right now. But their ability to add that functionality to the underlying platform, that’s something they can do relatively easily once they’ve got these sort of size and network externality dynamics going. So maybe if you looked at Instagram and WhatsApp and that framework, you would have seen those a little differently.

The only thing else though, I want to say, because I don’t hear this being talked about in these discussions, you can certainly have anti-competitive mergers, no question about that. But the other thing you see in is, you are getting assets, to hands that can use those assets more valuably. Instagram had succeeded with regard to I think this dynamics that I’m describing, but figuring out how to monetize that, they had not done that. Facebook was able to make that a more valuable asset faster. And that’s something that we as a society should want. So I do think that’s on the other side of that.

Scott Wallsten: It seems to me that all of these discussions are very Western focused and ignore the question of China and the Chinese companies. For every one of these US companies that people worry about, there’s a Chinese analog, there’s Baidu and Alibaba and Tencent and WeChat. And if scale is crucial to these businesses, isn’t antitrust enforcement against the Western companies basically making it easier for all the Chinese companies to dominate these various markets because the Chinese government isn’t about to break them up?

Randy Picker: I don’t know. I was at a conference at Stanford in June, which was a for Stanford-Peking University conference. I spend a certain amount of time talking to Tencent people there. I do not have a sense of the lines we’re going to draw U.S. separating Chinese firms and Chinese markets from U.S. firms and US markets. Not when I say I don’t know what we’re going to do there, I don’t feel embarrassed. That’s one of the great issues of the day right now. The huge fight we’re seeing about Huawei is a great example of that. I don’t know that the antitrust angle on that is the top level issue. It seems to me we need to resolve almost the trade question first. If we’re going to bar these companies from our markets, then, we’re in some sense removing potential competitors. We’re maybe feeding chunks of the world, China in particular, to those firms. But it’s not obvious that US antitrust policy can control whether or not US firms have access to Chinese markets.

Scott Wallsten: No, it doesn’t, but it seems like it also comes into play though, and how you define the market and I think you’re saying that they’re in the same product market, but maybe through other policies we’re keeping them a separate geographic market.

Randy Picker: Yeah. I think what I tried to say, that was better. I think.

Tom Lenard: So why don’t we close with a little discussion about the regulation issue, because there is this increasing number of people and they’re not necessarily fringe people. There’s this United Kingdom digital competition report, and there are others. I think there may be one out of the Stigler Center at the University of Chicago.

Randy Picker: That was Fiona Scott Morton who shared that committee. I was at the Stigler Center, I commented on the report. That was a lively discussion too.

Tom Lenard: They’re saying even a stepped up antitrust regime is just not sufficient for these companies. We need to have some sort of regulation, they don’t use the word regulation for much, they tried to avoid it, but basically what they’re saying is you need to have a regulatory regime that’s going to actively promote competition on top of it. And a lot of it has to do with the data issue and some of it just has to do with this, what I would call a kind of an open access issue. If you look at the UK report, the three main things that they recommend, and I’d just like your thoughts on them, are some sort of open access on a nondiscriminatory basis- that’s kind of a sina qua non of public utility, common carrier regulation to major platforms, and some sort of data mobility requirements, and also again, a common carrier type thing, giving potential competitors access to privately held data. Those are, those are pretty dramatic proposals.

Randy Picker: That’s fair. I think so. I talked about one of those, the data portability one, at a Stigler Center conference. And what I said on that was, I said that remedy doesn’t interest me in the slightest. What I think is much more interesting, but we’ve shown resistance to it, I don’t know what I called it. I called it maybe coordinated data portability or something like that. And here’s what I meant by that. Everyone’s got this idea of telephone number portability in mind. That’s the model for this idea and the idea that I can, and I’ve done this switch from Verizon, to AT&T or something like that and not lose my phone number. That’s the vision there. That’s not what’s at stake with a company like Facebook. All of my friends and I need to move en masse. Having Randy exit and going somewhere else if all my friends aren’t somewhere else doesn’t accomplish anything. The question is, how do you accomplish the kind of coordinated data portability? And my claim was, we’ve seen actually two examples of that. And in both times we’ve seen that are reasonably prominent examples. One older, one more recent. The antitrust people have resistance. So my first example was sort of in the early days of the digital world where EBay was the leading firm, there was a company called reverseauction.com and they basically said, we’re going to have a different auction model, but EBay’s ahead of us. They coordinate the buyers and sellers. All of these sellers have built up these reputations there. We’re going to copy all of those reputations. We’ll populate reverseauctions.com, you can come as a seller to reverse auction and basically you’ll bring your bring your reputation with you and we’re going to do that for everybody.

So there will be liquidity in a robust marketplace from the get go. And that decision 3-2, the FTC said, no you can’t do that, you, reverseauction.com. I assume they said it was an unfair act or deceptive practice. So you’re not allowed to do that. And then the second example of that is Yelp has this business model where it’s very much about getting consumers to deal with each other and you get these reviews and that’s got some attractive dynamics associated with it, but maybe Yelp’s not the best search engine. So maybe it would be good to take the reviews on Yelp, move those en masse somewhere and to match it with the different search engine. And that’s exactly what Google was doing when Google was scraping Yelp. And so there’s a way in which you might say, well, consumers can’t move unless they can move together, too high transaction costs. My God, Google has appeared and it’s going to make that possible. And the EC was very unhappy with that. So the data portability one, I am not sure. Individual data portability is interesting and I think coordinated data portability is a possible solution there we’ve resisted. On pure data access, so the Energy Policy Act of 1992 effective, we said we’re going to treat the electricity transmission grid as a natural monopoly, we’re going to have competition with regard to generation and now we create this elaborate connectivity regime. The 1996 Telecommunications Act does exactly the same thing with regard to local telephone grids. We ended up in a huge fight, multiple Supreme Court decisions over that. I think actually running those regimes is not easy. That’s not very interesting to say, but I think that’s the reality. Plus, here’s the other wrinkle on this goes back to something we’ve talked about a little bit earlier. The other wrinkle on it is if you are a privacy person, you’re going to react very negatively to that. Oh good, so they’ve got all this data about me and now they’re going to hand it to everyone else. What kind of solution is that? That’s the problem. That’s the solution.

Tom Lenard: There is a little irony in the fact that in privacy discussion, when any of these companies is quote sharing data with a third party, it’s terrible. And then you move over to the competition, and they say you need to share your data. Well, this has been extremely interesting to me. Scott, do you have any final words?

Scott Wallsten: No, I think that’s a good place to leave it. This was a fascinating discussion. Thanks so much, Randy.

Randy Picker: Sure. Happy to do it. See ya.

Scott, Tom: Bye.