Below, we have provided the full transcript of the interview with Prof. Daniel Sokol, Professor of Law and Business at USC Gould School of Law, recorded on August 17, 2021.
This interview was done as part of the Antitrust Brainstorming Board created by CPI with the support of the CCIA.
Thank you, Prof. Sokol, for sharing your time for this interview with CPI.
A video of the complete interview is available HERE.
Do you think the current antitrust framework works for consumers?
Daniel SOKOL:
If by the current framework, we mean a system that’s based on economic analysis, empirics, really taking very case-specific information and essentially processing it in a way that judges can understand and rule on with the singular goal, that works. In that system, do we have areas that need to change? Always. Are there some areas where we should be pushing to make sure we have greater enforcement? Of course, but this has been long-standing in antitrust and for generations courts have accepted that antitrust works very much like common law, that over time you shift the analysis and you do so in a way that makes sense, but to make sure that we all understand how this has been very much a broad non-ideological way of looking at it, probably the best formulation of what antitrust does well and frankly, development of the rule of reason was done more than any other justice by John Paul Stevens.
The most important case procedurally, I think, in the last 50 years was a 9-0 decision authored by Thurgood Marshall about the need to have antitrust injury, ie. some kind of causation from antitrust harm. Some of the most important discussions about administrability has come from justice Breyer. So this is not simply an ideological divide of Republicans are pro-business, Democrats are not, it is something long standing and the judiciary and many of the important developments happened, frankly, before, or rather contemporaneous to sort of Bork and The Antitrust Paradox. So these changes were longstanding and I think the ones that matter the most are the ones for which there’s been very strong support from justices who are political appointees of both parties.
Do you believe the vertical merger guidelines need to be changed?
SOKOL:
I wrote something on the vertical merger guidelines with a number of coauthors within the past year and a half, that’s suggested critiques of what works and what doesn’t work in a sense. Whereas I think the overall framing of there are situations where there may be some anti-competitive exclusion, certainly are the case. One issue that I had with the vertical merger guidelines is it hadn’t really incorporated well the set of trade-offs that one would make of anti-competitive effects on the one hand, and efficiencies on the other. And the idea of what efficiencies get measured, I thought was very narrow to largely very traditional elimination of double marginalization. But in fact, efficiencies exists in other forms that have very strong empirical basis in the strategy, the operations literature, the information systems literature, and here I think that the antitrust agencies need to take a closer look to see what do efficiencies really mean in practice.
Do you approve of the shift from competition towards regulation?
SOKOL:
There are essentially two questions there. One is, are essentially tech platforms public utilities? And the answer is no, if we look at traditional public utility regulation, water, traditional electricity, traditional natural gas or telecommunications, the platforms of today do not bear the same kind of dynamics, right? You have entry, you don’t have a single set of copper wires, for example, that means that others cannot enter. Just because for example, you use the Microsoft Word system doesn’t mean you might not easily switch to some other kind of program that you might use for your word processing needs. In some areas you may have more durable market power, but that’s exactly whether we would be more concerned from antitrust purposes. The broader question is, is there room for regulation in the digital space? The answer of course is yes. The question is what is it trying to solve?
Is it a competition problem? Is it a problem that is not about competition? Say regulation of political speech. Is it about privacy as a distinct issue from competition? Unless you can clearly identify what it is that you’re regulating, having regulations for the sake of regulation, it’s not really a good idea. The regulatory space has been, shall we say very mixed across a number of areas in the United States, largely with unintended consequences. The two most obvious, I would say, would be Sarbanes-Oxley that in many ways unanticipated chilled IPOs, and I think transformed businesses as we know it in ways that were unintended, it turns out, worked out very well for the big four accounting firms, but it added a lot of costs for small to medium businesses to choose the IPO route. Similarly, GDPR has been in tension with competition. We now see that empirically with the number of published or forthcoming papers in top journals.
How would you ensure antitrust is enforced vigorously if no changes are made to the current antitrust system?
SOKOL:
Enforcement, I think comes in two parts in the United States. One, is private actors and the other is the government. I think we’re seeing large companies suing large companies. I worry less about that. I want to leave aside class actions and follow on to cartels. I don’t think that’s where people say, “Oh my God, we seem to have a crisis.” The crisis seems to be, one, at least the perception is somehow we have greater concentration and that leads to single firm monopoly power. And in that setting, the question is with regard to unilateral effects, are there mergers that maybe should not have been permitted? We can have a number of retrospective studies as is always the case. In some cases, the answer is yes, there’s some mergers that probably shouldn’t have gone forward, and other ones that we say, “Oops”, but that’s on a sort of case by case basis.
The best way to solve that is to get resources to the agencies, making sure that they have sufficient economists working in the agencies to make sure that the economic analysis is rigorous, and it means getting resources to make sure you have people focusing on digital issues who actually have the appropriate backgrounds. I don’t mean futurists, with degrees in journalism or in philosophy. I actually mean people who have real data analytics skills that can help in two areas, one, substantive case analysis. And that’s great, but part of it is simply streamlining the system largely with regards to discovery. And that is something that increases cost significantly. So being able to optimize the litigation process itself and by litigation, I also mean cases that ultimately don’t get to corporate, agencies that are dealing with regards to the merger review, well worth the investment with regards specifically to case analysis. What do we do differently? Make sure you don’t just have enough people in terms of the resources, make sure that the resources are being used in ways that make sense. So a lot of times we focus on high profile deals, regardless of competitor affects, largely for political reasons. That’s not the job of an enforcement agency. The job of the enforcement agency is to focus on cases because they are as actually significant anti-competitive risks.
You have agency discretion, of course, as to particular industries to focus on, but we have situations on a regular basis where we have had success because there has been significant focus. I think for example, very successful mostly, FTC focus on challenging hospital mergers that may somehow lessen competition. And frankly, the FTC doesn’t get enough credit for that. We’ve also seen cases where it’s not only our antitrust enforcers to clear a merger, it’s also our sector regulators that clear the same merger. This could be the FCC, this could be the FAA, this could be FERC, whatever it is. The thing is, somehow antitrust gets all the blame when there’s a perception, “Oh, they were asleep at the wheel.” Well, what about the other agencies? Right? So if antitrust should get credit for doing things, then antitrust shouldn’t also get all the blame in case there in fact was, let’s say, mergers that should have been blocked.
In terms of litigation that should have happened, we all have our cases. They tend to be, frankly, on the margins, you do want some experimentation, but you also want cases that mattered, not just for that specific case, but that properly try to align the development of case law in a very rigorous way. You always look for the appropriate test cases, but you don’t bring dead-bang losers.
What are your thoughts regarding start-up acquisitions?
SOKOL:
I’ve done work, both academic work and then more recently a white paper for the National Venture Capital Association specific to this issue. I have very, very strong concerns that at a time when the United States should be promoting innovation, it will in fact create set of rules and a set of presumptions that will stymie innovation by making it more difficult for start-ups to be pardoned. Here’s the reality, and it’s not even close, a significant majority of founder exits come from acquisition. Let’s imagine it doesn’t stop it, but it significantly slows it down. This has many negative consequences. Number one, you have no way of getting liquidity. You’re starting a company largely because you want to exit. And if you say it’s more difficult to exit, okay, there are a few different things that may happen. Number one, why invest, why bother to start up a company if you know you can actually exit it?
Number two, if you are the company that’s been doing the acquisition, it’s not like you say, “Oh, every year we have a $4 billion budget for acquisitions to help out,” whether it’s corporate venture capital or actually just acquisitions. All that money is now going to be shifted a hundred percent to research and development. It’s not like the research and development that gets done in a large company is the same in a start-up, because in a startup you say “I’m going to really try very, very hard to go for broke.” Essentially, “I’m going to transform the world, and if it doesn’t work out, I’m going to crash in flames.” If you’re in a large company, I can assure you in a publicly traded company and an investor calls, a CEO is not going to say “We’re going to give a senior manager the opportunity to completely blow 100% of the budget we give them on some moonshot idea.” It doesn’t happen, and so we’re going to lose that innovation and that’s quite troubling.
Is break-up the best solution for the digital economy and for consumers?
SOKOL:
So there are essentially two kinds of break-up. One, is structural break-up, think of standard oil, for example. The other would be defacto break-up where you used to be able to do businesses A, B, and C, but we’re going to say, “Actually, you can only do A.” I’m going to take the second one first, so what we might call line of business restrictions empirically we know, and it’s very clear from the empirical record, Glass-Steagall was a disaster. We could look back both in the top finance and economic journals to show that the repeal of Glass-Steagall and frankly, the world before Glass-Steagall, was better than the world under Glass-Steagall from the perspective of competition. We’ve also seen essentially limits on any kind of vertical integration, such as by the ICC with regards to transport, with regards to the CBA, with regard to air transport. Also, things that really hurt the economy.
Here again, the empirical record is clear. So when people say, “I want to return to the past,” I’m thinking, have you actually looked empirically at the past? This is the bizarre world that we live in today. We are told, for example, oh, that we should constantly be thinking about passing reinvigorate antitrust. We say, “that’s great.” If we look at the nadir of antitrust, when it was too heavily enforced, we’re told now that, “Oh, it’s dragging the economy.” Why don’t we actually look at the economy, say circa 1968 to 1981, the worst period in the economic history of the United States going back a hundred years. So if people say that there’s a causal link between antitrust and the economy, we have to then ask ourselves, “Why was the period say between Reagan in 1981 to the George W. Bush presidency where antitrust opened up? Why was that the period of significant economic growth at the same time that we were seeing more consolidation?”
So some consolidation is actually good. What we’re concerned about is where there’s the unlawful exercise of monopoly power. And this gets to really the first part of your question. What does break-up mean? So when break-up has been used, it’s been used very rarely. It turns out standard oil’s break-up was a disaster for competition. The person who benefited the most John Rockefeller, because essentially what he created was a series of smaller monopolies, simply because the way that gas worked at the time, you didn’t really have the economies of scale, that’s different than the kind of break-up that we would have today. But the outcome I think, would be the same. The very simple proposition, is that the value of a platform is the platform. And we have to think, is the value of the network effects greater than any kind of anti-competitive harm that might be fixed in other ways, whether it means regulation, maybe if it means some kind of behavioral remedies, right?
Does it make sense to break up a platform essentially? And the empirical work on platforms from the people who actually study how platforms work. These are professors of information systems, professors of strategy, professors of marketing. When we look at the value creation of platforms it significantly outweighs any kind of value destruction, but herein lies a larger issue, we’re concerned, rightly so, with diversity and inclusion in the United States. We know that the poor face disparate impact, whenever prices go up, it’s a little odd to hear that we care about diversity and inclusion, and we’re saying at the same time, “We’d like to raise prices. We’d like to raise prices on basic and consumer products. We want to make the Internet not free by limiting mobile advertising.” For people who live in tony suburbs in Washington D.C. on the Maryland or Virginia side of the border, maybe it’s not going to make a difference to them, but for communities that are economically vulnerable, increasing the cost of e-commerce, increasing the costs of simply using what had been a free Internet system, because we have break-up, this has significantly negative disparate impact on such vulnerable communities.
How do you see the role of the FTC and the DOJ in ensuring competition works for consumers?
SOKOL:
There are some 200 plus competition authorities around the world, and so we have a series of natural experiments. I don’t think if anybody was starting from scratch in the United States, we would have our two agency model that we have today, but this is the world we live in. There are many things we would do differently if we could do it from scratch, but there’s a certain amount of path dependency. So, there’ve been a number of areas over the last 15 to 20 years where sometimes we’ve had significant tension between the two agencies. Does it make sense to have a singular agency? Unclear, but should they work together? Yes, it’s a little difficult if they don’t share the same goal as to how the economy should be structured. It’s a little difficult, if in fact, functionally you have different merger standards as between the two agencies, because then we begin to have a rule of law type problem.
You get one kind of justice in front of one agency and a different form of justice in front of the other agency and in industries where you can have a clearance battle as to which of the two it goes to, there’s a certain amount of unfairness to that. Much of the time the agencies have worked very well together in dealing with these issues, but we always have to be cognizant of the fact that there is this tension. Sometimes, the tension is also bled into international matters, becomes very unfortunate when we get a series of mixed messages and people cherry-pick what they want to hear.
Do I think that both agencies over the years have done a good job? Yes. Are there mistakes that both agencies have made? Yes. Do I think that we should abolish one or the other agency? No, but that is largely based on the fact that the agencies have worked within the system and have don so to try to very carefully balance addressing when to intervene with regards to potential or actual anti-competitive effects and how might you encourage innovation, lower prices, higher non-price issues, such as quality through sometimes not intervening, and it’s not easy.
So whoever thinks that there should be easy solutions, bright-line rules. Well, I think either doesn’t understand how a complex system works or overestimates their ability to solve complicated problems with so-called easy solutions.
How would you reconcile competition and competitiveness? Should antitrust reforms take into account the potential impact on proposed changes vis-à-vis China?
SOKOL:
Here we have another natural experiment, we have seen literally the Chinese stock market loses $1 trillion in value, largely because they’ve been aggressive the way that populous groups on both left and the right would like us to be aggressive in terms of competition. Now, in the Chinese case, it’s very clear that the purpose for that is because it’s about maintaining the power of the communist party.
Thankfully, we live in a democratic system where that’s really not what government seeks to do. The question is though, what happens if we hurt innovation? So the obvious place to me, where you see significant innovation in China, is in the development of AI. And the question you ask yourself is if in the back of our minds, we think of antitrust enforcement as part of a broader strategic sort of issue, if we take some of the critiques, the agents of change for antitrust, that they want to have a broader societal impact, part of the societal impact presumably is country-competitiveness. And the question you ask yourself is if by making the country less competitive as a result of antitrust, if that say increases the power of China with regards to let’s say the IP to AI and broader data analytics, is that a good thing for the United States? No.
If you say however, antitrust should stay in the antitrust lane and antitrust can deal with specific interest issues, there are different policy instruments that can work in a complementary fashion. The answer is, it doesn’t really matter from an antitrust perspective, we have trade policy, we have other types of innovation policy that emanates from the U.S. PTO, that emanates from the Council of Scientific Advisors, et cetera, that we get from the executive branch better coordination. So the idea that you might have better coordination on innovation is a great thing. Of course, the devil is in the details. How do we promote innovation policy broadly? There are many federal agencies, all dealing with innovation, antitrust is just one piece. So it can’t be the antitrust is everything, but of course only when there’s bad stuff, or that antitrust is nothing. Antitrust plays a role, but it’s certainly not the only role, but to the extent that we are concerned that if we make it difficult, if we destroy the digital economy in the United States, the answer is not we’ll have five different e-commerce companies or social media companies, all of which are robust.
We’ll have a series of smaller companies that are probably highly differentiated, which is good, but we’ll probably still have a very concentrated number of Chinese-based companies. And to the extent that people are concerned, whether on economic or democratic or both grounds, this is a serious worry and should be a serious worry for the broader tech policy community.
Any final comments you would like to make?
SOKOL:
Yes, I have three brief comments. First, clearly articulate what’s the problem. Two, identify is antitrust the solution in part, or in whole to this problem? Three, rule of law matters. And by that, I mean, there needs to have procedural fairness in the outcomes, but also in the process. One reason that we have various, let’s call them speed bumps, whether through regulatory adjudication or in fact in cases, is precisely because we want to make sure that parties have due process. This is enshrined in The Constitution and actually forms the basis, I think, for a legal regime to have procedural fairness through transparency and through due process that can’t be lost. The final point I’d say is, when say the regulation of tech, not all tech is the same. We need to understand different business models, different competitive effects. When do we have lock-in? When does lock-in simply not exist? These are complicated questions. So, the better way of doing it is study, based on the study synthesize, then enforce or regulate, but starting with “I’m going to regulate.” I think is exactly the opposite approach in the process. Thank you.
