Yesterday, I had the great pleasure to discuss ongoing antitrust evolutions with Bill Kovacic, Professor of Competition Law at GW, Non-Executive Director of the Competition and Markets Authority, and Former Chair of the Federal Trade Commission, at the 4th IBCI annual conference. We touched upon the substance of antitrust law (e.g., ex-ante rules, such as the DMA, environmental protection, etc.), the political transformation of the space, and institutional design. Bill provided us with a unique perspective on the most crucial issues of the day, and, also, concrete tips for antitrust agencies and researchers. I am most grateful for this discussion as Bill has been long-time support and role model of mine — Thibault.
A lightly edited transcript:
Welcome, everyone. Eduardo, you put a heavy burden on us by making this discussion the very last of the entire conference. Now, I’m incredibly excited to have a chance to talk to you in public, Bill. I do not need to introduce you. I will simply say that you are a Professor of Competition Law at GW and the former chair of the FTC. So, we will cover what’s happening in the academic space, but also in the space of enforcement.
My idea for the conversation is to structure the discussion around three main subjects. Subject number one, the substance of antitrust, its evolution, and where it might go. Of course, prediction is hard, especially when it’s about the future. So, please, do not use this video in just a few months from now to prove us wrong. Subject number two: I do want to discuss something which, of course, also impacts the actual substance of antitrust: populism, or I should probably say, the subject of the modern transformation of US antitrust law and policy. And then, I have some more outside-the-box questions, I suppose. This will be the third time of our conversation.
So, without further ado, I will dive into the very first part of our conversation regarding the substance of antitrust. I want to address the elephant in the room: the DMA. It seems that the logic of implementing per se / ex-ante rules is actually taking over Europe, and also the US with many acts which are currently undergoing. My question, Bill, is the following: how good of an idea are those ex-ante rules?
I must say that I am concerned about two things, both of which appear in the DMA rationale. First, the Commission says that it has enough experience with the practices listed in the DMA. If that is true, it means that those practices are therefore covered by competition law. So, the idea that the DMA is just a complement to competition law is then is proven wrong. Second, I’m also concerned that some of those practices are, indeed, not competition law. If that is true, I doubt that antitrust agencies have enough experience. What is your take on it?
Thibaut, I’m thrilled to be here with you and to have the opportunity to participate in this discussion. Many thanks to IBCI as well for the wonderful work they’ve done over the years in creating a platform for discussion and debate on the subject matter. So, I’m enormously grateful to both IBCI, and to you, Thibault.
In general terms, I’m skeptical about the use of bright-line categorical prohibitions in this area. I’m shaped in part by living in an environment — the United States — which has tended over time to reserve absolute condemnation for a relatively narrow band of behaviors, namely, so-called hardcore offenses involving direct competitors. But even there, there’s been the recognition that you’ll need a filtering mechanism that identifies when such restrictions are directly related to a pro-competitive purpose and have to be evaluated more completely, or truly can simply be condemned out of hand.
I think that the DMA is understandable for its desire to provide clarity — that’s a natural impulse in our field — in so many areas of the law where we’ve had the debate between bright-line rules and standards. I think what we see generally in the law is that bright-line rules tend to be converted into presumptions, sometimes powerful presumptions, but not conclusive presumptions. And the world of standards starts to borrow from concepts and bright-line rules to create presumptions, so there’s a convergence around rebuttable presumptions. The DMA is painting itself into a bit of a corner when relying on categorical prohibitions.
Even if the allowance of rebuttals was made extremely difficult to invoke, I’d still want the flexibility in my system to listen to the good business justification of the firm I bring forward. I say that realizing that in the course of implementation, those justifications can become so sterile, so overlooked and routinely ignored, that you could say what’s the point of having them at all? I know that happens in a variety of systems over time. But, I’d like to have enough faith in the entire process to believe that when good arguments were provided, they would be heard out, maybe not routinely embraced, but they would be heard out.
My experience at the CMA has increased my caution about categorical presumptions. I think of the work that the CMA has done in the course of market studies in its individual casework. My inference about that work — and I don’t speak for the institution, I only speak for my reflection around being a non-executive director on the board — is that presumptions will serve the useful purpose of providing clarity, but will provide a necessary opportunity to allow for justifications.
In a dynamic technical environment, it’s hard to foresee all of the circumstances in which specific justifications might be valid. I think all of the institutions involved in this process will acknowledge that the process and the welter of policy development globally are experimental. That involves testing different approaches. I’m not fond of carrying out that experiment on the basis of bright-line rules. I will mention that I have colleagues in the United Kingdom, policy observers, who are concerned that the DMU approach in the UK is too open-ended, too broad in its orientation, and that it searches for the perfectly tailored set of behavioral constraints on each platform will be burdensome to devise and difficult to operate. So that’s the warning label that would go on the DMU if you were selling the product over the counter.
At the same time, though, my anticipation would be to gravitate in the direction of presumptions that are rebuttable. So, I think if you use bright-line rules, there is a great possibility that in a short period of time, you’re going to have to revisit them and modify them to allow justifications. I don’t doubt the deep reflection that’s going on inside the European Commission and DG CONNECT and in COMP about what the right approach is. These are enormously capable people working with very hard problems. But I guess if I were writing the referee report for the DMA, I’d say, press the authors to explain why the bright lines are clearly justified based on our experience in the past.
And if we purport to exclude justifications, aren’t they going to creep into the analysis somehow? Don’t we end up with opinions at some point that look like cart bond care and say, “well, before you apply the categorical 101§1 prohibition, you have to be very careful, careful that it fits within the boundaries of the prohibition”? We can end up having classification debates, where jurists and maybe enforcement agencies say “of course it’s a bright-line rule but, ah, in this case, you didn’t fall within it”. That’s a long way of saying I like rebuttable presumptions, even if the presumptions are fairly powerful.
Companies will surely argue that their practice is not one of the practices listening to DMA. Plus, you indeed may want to allow companies to come up with good explanations under rebuttable presumptions. The DMA does not allow that, which is one of the drawbacks of ex-ante approaches.
But there is another issue: if you want to put a clear rule on a piece of paper, then you have to deal with the trade-offs. There is one that we’ve been discussing for quite some time now, i.e., the trade-off between privacy and competition law. I think there is another one that is coming our way: the trade-off between competition and security. That appears very clearly with the draft Open Apps Market Act (in the United States) that will prevent app stores from requiring app developers to use an in-app payment system controlled by the store. We have something similar in the DMA. So, how can we approach this trade-off between competition and security? Are there any rules out there that you could, you know, provide us with, or is it just something for a case-by-case basis?
You’ve pointed to another severe complication with using the categorical prohibition, either in the DMA or in some of the proposals in the United States. I don’t know how many webinars all of us have gone to now where some speaker — it might have been me at some time — said we have to take account of these related policy domains or we’re not going to get the right answer, or we have to take account of how the different forms of guidance that come from each domain might not be coherent.
Competition authorities are being told you must bring all this on board. You have to use what the Biden administration has called a whole of government approach, or when academics when reaching desperately for a cliche called holistic policymaking, which I assume means a more complete framework in which you take account of adjacent are overlapping policy domains. If I’m being taught to do that, and I candidly recognize that the impulses that come out of each of those systems, on a good day, might be congruent, on another day might be really complimentary, but on a difficult day conflict.
How are we supposed to reconcile these, except by giving first an opportunity to say, “I have a good reason for what I’m doing here, I’m responding to the policy imperative that comes out in the related domain.” Now, it would be a fair answer for a competition agency to say, “that’s a conversation worth having. We should talk about the justification, as you invited out. And by the way, we’re more impressed if it comes from contemporaneous business records, rather than the cleverness of your lawyers and economists after the fact. But we’ll also have as part of that conversation, an exploration of whether or not the approach you’ve taken is in the language of the European Commission’s jurisprudence… proportional. Did you use the right tool well-calibrated to serve the problem that’s identified?”
It seems to me that if you jump to categorical prohibitions, you’re attempting to forestall a conversation that is going to break out in one form or another, at some point. I mean, my prediction would be that if you have the categorical prohibition that takes no account of these policy complementarities, or conflicts, or uncertainties, you’re going to be coming back in a short period of time to put the DMA back on the rack, or the online Competition Act in the US back in the shop to repair it, to make modifications which probably you will have to revisit anyway because things are going to change in ways that human beings can’t anticipate.
It appears to me that one way that you account for this intersection collision parallelism of policy commands is that you have rebuttable presumptions that allow, as one line of defense, a firm to say, “I’m responding in a good-faith manner to the policy guidance that comes from this other policy domain.” And… if you’re not going to address the hierarchy of authority directly in the DMA, you have to entertain that conversation, or I would suggest you’ll build up such frustration that you’ll be bringing the DMA back in the shop in a short period of time to amend it.
And it might be that no trade-off can’t even be considered. Following the DMA discussions in European Parliament, it seems that objective justifications won’t be possible while, in fact, we hear about the necessity to consider environmental protection when it comes to competition policy. So, it will not be possible for a company to say “yes, my practice is listed in the DMA, but it is good for the environment, therefore, give me a free pass.”
This leads me to discuss another trade-off with you: sustainability vs. low incomes. If you allow cartels because they benefit the environment, it might be that the prices will go up, which will hurt low incomes because they won’t be able to access the products anymore. How to deal with that? Should we prefer one over the other? And also, can we trust antitrust agencies to make such a decision? Or… do we exclude objective justifications so we don’t even have to deal with the trade-off in the first place?
As a starting point, I would say that when competition authorities come to the point in which they’re going to make the trade-offs, this framework places an absolute premium on being honest about what they’re doing. That is explaining in a very direct and careful way, what they perceive the trade-offs to be, how they reconcile them, and what evidence was relevant in formulating the ultimate solution. If this is a math exam, I think the instructor has to be brutally insistent on saying I want to see all your work. I don’t want you to wave your hands and say “this was an inevitable result, I took everything into account and this is the answer.” I want to see how you got from the definition of the problem to the solution.
And that’s going to be awkward. I mean, my experience in writing opinions for the Federal Trade Commission when I was in a position of having to participate in the board in an adjudication function is: if you’re not entirely certain about what you’re doing or how to do it, there’s a lot of voices sits on your shoulder that says, “the solution here is to obfuscate, hides the trade-offs, hide your uncertainty, construct a straw man for your opponent and then use your awesome logic just to crush them.” And that that takes you in the direction of not being honest about what you’re doing. But that honesty is indispensable for both evaluating what the methodology is or deciding in the future if we want to make changes.
So, let’s assume that putting everything into the framework is desirable. And maybe is it mandated under the terms of the treaty in Europe. I want to see how you do it. And I do not only want to see the ultimate decision. I think back to my youth in this field in 1979, a few years ago, when I was a junior case handler working sometimes on mergers. I didn’t have a specific ideological ax to grind. I wanted management to tell me what do you expect me to do here. “And gathering facts, what facts do you want me to gather? And when I collected them, what methodology do you want me to use? And don’t just give me a hierarchy of factors or no hierarchy, and say, maximize all these variables.” I’m thinking of that young person is going to have to do that work. So, it’s incumbent on the agencies if this is all thrown in their lap, to explain clearly what they’re doing and to explain to their own people what they’re supposed to be doing.
I think that a problem we’re going to face is that the more things you bring in, article 101 is going to start to unravel. It’s really going to fray at the edges. There’s going to be extraordinary pressure on agencies, maybe when you get to 101§3, to start bringing things in and taking them seriously. And how do you ignore them, if you have strong declarations, principle declarations, that say “we’re going to save the planet and we’re going to be receptive to a wide variety of experiments that involve collaboration among rivals to achieve that”.
What person respecting the process of democratically elected officials taking decisions on those issues would say “you can’t do that”? Of course, you can. Yet, for the purpose of policy coherence, you want guidance to the implementing institutions about what to do and how to do it, and, you want them to exercise their discretion in ways that make clear what they’re doing — not be so worried about being seen as making analytical jumps of faith, or simply failing to provide a convincing basis. So we’ll have the debate and the criticism and ask do we want to keep doing this in this way.
By the way, this is going to be a universal pressure, I think, Thibault. Why did agencies retreat so much to the consumer-oriented welfare framework, which arguably involves price-quality, innovation, monopsony effects, and labor markets? One explanation I’ve heard many gave to me over the course of my practicing lifetime is that’s the only manageable way to do this job. And if you open the door more broadly, you end up with results that become impressionistic, unpredictable. By the way, if everything is relevant, why shouldn’t elected officials jumping in and saying “you’re doing it wrong”? Why should they not do that and say, “that’s what we do, we do that all the time? Will tell you what to do here.”
Several years ago, the remarkable Mario Monti was asked in a program in the UK, “why did you join the consumer welfare parade and the more economic approach? Why did you do it?” He said, “two reasons. One is to make the United States happy and get them off our back after GE Honeywell. That was one. But second, was the concern that if you did not build limits around your methodology and approach you have opened the doors to very broadly based popular sentiments that tell you to maximize everything”. He said, “this was the only way we had to cut that off.” Those pressures are always present in our decision-making in the CMA, so, we have a great deal of pressure now to open that door back up. And, I don’t say for a second that opening the doors is illegitimate by any means. But tell me how you’re going to do it.
That’s a perfect segway to the second part of our discussion, but first, a few reactions. One, I think not all of the agencies are actually living in fear of facing such a situation. Some agencies are actually looking for such cases where they can be the first to say “look at what we’ve done. This was a cartel and yet good for the environment, therefore, we gave a free pass”. I could see how the press would react to that.
Second, I had once a discussion with Philippe Aghion who told me that the great economists are the ones that will tell you what are the forces in presence and which one is taking precedence over the others. I think this is exactly what you described for enforcement: what are the trade-offs and how do you deal with them on a systematic basis. And if you can’t do that, then I see no reason why you will have indeed per se rules.
Third, to be more positive with the DMA and other Acts, they are at least forcing us to come up with a list of the objections and the trade-offs to provide something clear to the companies. Another good thing is that if you do impose too strong restrictions, it might create a desire for companies to evade those markets, and therefore to innovate somewhere else, where there is no regulation. So, it might be that regulation is pushing companies in a direction that is more disruptive.
I think you underscore what’s going to be a major obligation for the academy, for professional societies and others, and indeed for competition authorities, is you cannot escape the need to do an ongoing and careful assessment of the actual intended and unintended consequences of these policy changes. If these are experiments — how many times have you heard leading figures say “it is experimental, we have to try something”.
Indeed, there is a desire on the part of some agencies to test out what you might do in 101§3 — we’re going to have to be willing to watch carefully to keep score to see how it’s all working out. Whoever is raising their hand to get into the 101§3 sustainability analysis, I welcome their participation and I look forward to their careful explanation about what they’ve done, and how they did it. So we can look at that methodology and learn from it, discuss it… that’s wonderful, that kind of assessment will be a necessary part of a virtuous circle of experimentation.
It means lots of great papers and books coming our way. Now we’ve been talking about agencies and this will lead us to discuss the political implications of antitrust — the second part of our conversation. Here, I want to insist for everyone watching us that you’ve written a stellar article entitled “Root and Branch Reconstruction” in which you explored the modern transformation of US antitrust policy. And what you do in the paper is that you start by drawing a distinction between the traditionalists, expansionists, and transformation advocates. You say that those labels better describe and illustrate what those groups actually stand for. I very much agree with you. Can you explain for all of our listeners, who and what do they cover?
I guess the only way to begin those conversations is to spend time on definitions. What do we mean by this? I was looking for a way to describe functionally what these groups want to do, and there’s one group functionally that says, certainly in the US, but elsewhere, “the system works pretty well as it is, leaves it alone”. I call them traditionalists, but their main belief, and certainly in the US environment, is that the body of doctrine and policy has got things in pretty much the right place. Now, some would prefer to move specific elements of doctrine in one direction or another, and it’s open to a fuller debate about what to do, but the conclusion of that group — I look at my co-author Josh Wright as being representative of that view is — is living alone, work pretty well.
The second group I call expansionist. I’d say the spiritual core of the group comes from alumni in the antitrust agencies during the Clinton Administration and the Obama Biden administration where quite a few of the analysts had positions of senior leadership in these institutions. And at least indirectly, I think their view is “we should have been tougher, we should have done more.” As a consequence, they’ve said “let’s use existing tools more aggressively, and in particular, accept the greater possibility of losing cases, perhaps as a way of going to the legislature.” Who are some of these? Carl Shapiro, Fiona Scott Morton, John Sallet, Bill Baer, Jonathan Baker… these are all representative of this point of view.
Do they have exactly overlapping preferences? No. And indeed, a number of them say we need new legislative tools, especially some form of an ex-ante regulatory framework for big tech. But they are agreed emphatically that the focus of attention should be a more conventional traditional welfare analysis focused on the well being of consumers, price, quality, innovation, and the well being of workers to the extent that labor market distortions induced by monopsony behavior, distort the appropriate allocation of resources and effect incentives to provide certain inputs. That’s included in the mix. And I think it’s their belief that if you do all those things well, you’ll have as a major byproduct, very good distributional effects. But, they reject the broader egalitarian vision of competition law that was cresting in the United States in the 1960s.
This brings us to the third group, the transformation advocates, who agree with many of the policy prescriptions of the expansionists, but they love the egalitarian vision that the expansionists regard as poison. Cases that expansion is held out as being excessive in their perspective. Cases like the Brown Shoe merger decision in 1962, the Alcoa exclusionary conduct case of 1945. These are the Word of God for the transformationists, in particular the lyrical passages that talk about the protection of small business control by local communities of commerce that takes place there and the preservation of an atomized business environment that enables democracy to live.
The current chair of the FTC — Lina Khan, and a number of her supporting cohort — says “this is about democracy, and about the preservation of sound political institutions first and foremost. And then if you don’t accept that vision, you’re missing the whole point.” That separates them dramatically from the expansionists. There, has emerged a real hand to hand combat between the expansionists and the transformationalists, a harsh and bitter debate where the transformationalists describe expansion figures as being too timid, as having shown in their period of public service that the only thing they could do was do nothing, and to do it on a grand scale.
The expansionists thought they were going to have the big jobs in the Biden administration. They thought they would get the call, and they find this new group outflank them dramatically and pass them up. Now, they refer to them as naive, inexperienced in the ways of the world — that’s a direct punch to Lina Khan — and unreasonable in their belief that this broader vision can never be applied in practice. But to their great dismay, the expansionists have found that Lina Khan is the Chair of the Federal Trade Commission, Tim Wu is in the White House advising the president, and Jonathan Kanter is the soon-to-be nominee for the Department of Justice antitrust division. The traditionalists and the expansionists never saw this coming.
We’ve seen a couple of books this last few years arguing that the last 30 years have been pretty much good for nothing, that everything should be thrown under the table, that we should even go back to the beautiful days of the 60s and the 70s when everything was “perfect”. I’m not too sure as to whether you would consider that those books have been written for the most part by expansionists or those advocating for transformation. This is part one of my question. But also, if we are to answer that and to show that not everything has been bad in the last 40 to 50 years, which enforcement actions would you put under the spotlight to show that some good things have been accomplished along the way?
The approach of the transformation team is exactly as you describe it. They basically say that the world went to hell at the end of 1980 when Ronald Reagan becomes US President and then begins to embed very permissive attitudes into the enforcement program. They describe that as being a calamitously bad program and failure. Upon cross-examination, they will make minor concessions and say that okay, the Microsoft case in the late 1990s, that wasn’t entirely bad. But most of it was bad. And so the Microsoft case was a rare lucid interval for a psychologically deranged person who may be 5 or 10 minutes a day sees the world clearly but then devolves into a state of catatonia. Almost by accident that it happened, but the rest of it’s awful.
The expansion detests that description because as it writes off the Clinton administration in the Obama administration. The transformations detest Obama more than they did test Reagan and both of the Bush presidencies because Obama came with such hope, so many declarations about how he was going to renew the system, especially for competition law. “He had the chance to do it and then, in the view of the transformationalists, he failed abysmally. He raised hopes to the stratosphere and could not deliver even Silver Spring Maryland, which is on the boundary of Washington DC. He promised the universe and delivered just about nothing. And he embraced the world of Wall Street, Silicon Valley, big tech, so… he’s despised more than the Republicans because we know they’re godless anyway; do you never expect anything of them.”
I do have a bit of a dog in this fight because I was at the FTC for most of 2001 to 2011. And I like to think that my time there was not absolutely bad, purely out of fear that the US government could dump to me and ask for my salary back for those years. So, I am I’m naturally inclined not to think that 10 years of my life were a complete disaster. This narrative has precursors in the history of US antitrust law, people have made the case for basic reform by saying that everything that happened before they got there was a disaster. Robert Bork in 1978 says that “all of antitrust from 1940 up through the early 70s is bizarre. He mentioned one case that he likes neared monopolization, but he says the courts are crazy, Congress has to range and the antitrust agencies have one goal: that’s to acquire power and to use it extensively. It’s the only thing I care about. But I can save it.” Well, the transformationalists are saying the same thing now, that only they can save it. What do you do with this kind of history? “I’m here to save the leper colony but it is a leper colony.” As you say you overlook good things that took place during the period of disaster. I’ll just mention a couple.
Coming into 2001 when I was the head of legal services at the FTC, the Supreme Court had at blown a big hole in the US equivalent article 101, that section one the Sherman Act, and had raised major doubts about whether you could use the presumption for certain types of behavior that resembled a hardcore offense but was not identical to it. Coming in, we said, “we have to restore that.” We decided on a series of cases that were specifically designed to do that. That ever was pretty successful. Cases known such as Polygram: we took a hospital merger program that had failed to block a merger for about a decade, and we said let’s look at the assumptions that are guiding the courts, let’s do some empirical work to contest them. And then let’s bring our own cases to try to bring the new evidence, the new learning to bear. This was largely successful.
And again during this program of failure, three times over the past 10 years, the Federal Trade Commission has been before the Supreme Court with a competition case. And mind you this is a Supreme Court that’s severely skeptical of antitrust intervention, and the FTC won them all. One of them was called Activis, which was not a sweeping victory, but a lot better than what existed before. The other two cases were significant victories involving the role of state intervention. We won them all. It didn’t happen by accident. These were conscious policy choices
If I’ve seen that happen before, and I’m coming in with my program, I want to study all of those examples and understand “how in the face of a skeptical judiciary, did they manage to do that?” I want to study those cases carefully so that I can find the same path to the summit. But, if it’s all rubbish, nonsense, no need for that.
That leads me to the final question I want to ask you regarding the transformation of US antitrust policy, because if indeed not everything was bad, and if there are some good things justifying to conduct retrospective studies to look at what happened and how we’ve actually come to some great success, then my question is, how do we avoid the pillars that the transformation advocates are putting in place? For the enforcers listening to us, and also for the young scholars, where should they go to access great resources outside of those pillars to actually learn about what you just described? Which journals should they read? Which conferences should they go to and also avoid?
I’m gonna make a couple of pitches. I think of the program that Nicolas Petit is building at EUI — which builds upon one standing program where the spirit of the conferences and programs has been an honest debate among people who do not agree with each other. It’s an intimate setting, it’s a workshop, in many instances with 20 or so people around the table, and you don’t get in the room unless you write a paper in advance that’s circulated to everybody. The ground rule for debate is to use a phrase I’ve absorbed from Nicolas is “you get to play the ball, but you don’t get to play the player.” So if you want to make a hard tackle, do it clean, and we will applaud you for your skill and your aggressiveness. If you go for the players, you’re not going to be in the room.
No American football.
Yes, American football is a brutal ghastly sport. Nicolas is trying to create a spirit and environment that I think is enormously healthy in our field so we get to a common understanding about what actually happened, and then have a thoughtful discussion about whether we like what happened and what we’ve learned about it for the future. So if I had to pick one, I like what the EUI is done with their workshops in the past, and I applaud the spirit and approach that Nicolas has taken to address that. I think we need a lot more of that and maybe it has to be a quiet or smaller discussion where people are willing to play by these rules not a room with 300 people which is going to be a show trial. You put people on stage, they act. And they quickly feel cornered by what they’ve said before, they lash out, they fight back and the discussion degenerates into a Twitter-like name-calling contest, where “you are a prostitute and I’m a purist”, or “you’re bought by the companies and I’m a clean person.” That is so uniquely uninformative!
And by the way, I see this as the spirit of a number of programs that we see around the world. IBCI is one of them. Brazil has wonderful programs. There’s also the Centro Competencia that Julián Peña has organized and carried out in Buenos Aires over the last 20 years. It started as a breakfast meeting; breakfast with a few friends who are practitioners, competition officials and people from the bar, and the civil society, that sit down at the table, have a meal, and talk about common ideas. That’s grown. Now, it’s a big table now, but it’s still breakfast. The whole point of what Julián did is “how do we create a community in which we can get some common understanding of what happened and use that as a baseline for deciding what we got to do going ahead?” Those can be a raise of hope.
Something to do with food, I suppose, because EUI is no stranger to that. Of course, I can only emphasize that EUI Ph.D. program is a place to be for young scholars. So, have a look at their website.
I will tell young scholars, in that spirit, whenever you see an especially strong claim in the work of scholarship that is drawing upon a portrayal of the past, go back and look at source materials on your own. Go back and read the cases on your own. Go back and read the secondary sources on your own and see whether or not you come to the same conclusion. I applaud the interest in historical scholarship, it’s wonderful. But too often, it’s become another form of advocacy of brief writing. Over my lifetime, I backed off of that birds like “always”, “never,” because of being wrong often enough, but when you see strong claims and say history says X, well… there are many histories and you want to form your own opinion.
I want to discuss one history at least, the one that you encountered while being at the FTC and, currently, inside the CMA — although there is some information that you can’t share. This is part three of our discussion. I want to discuss the institutional part which is left apart very often. I think that’s a shame. My first question regards how we can actually deal with economic expertise within agencies.
To be very specific, there is a clear question right now on the table: killer acquisitions. I’ve looked at empirical work recently published, and I’m just going to give you two examples. One, entitled “The aggregate effects of acquisitions on innovation and economic growth”, shows that by banning startup acquisitions, you will actually increase economic growth. They do have empirical data to back up their story. So, here you would say “killer acquisition is a big problem, so let’s ban any sorts of acquisition.” If I look at another study, entitled “Effects of big tech acquisitions on stock, funding, and innovation,” they have analyzed over 30,000 venture capital deals in 173 different segments of the technology industry. They show that when there are acquisitions by Google, Facebook, Amazon, Apple, and Microsoft, it creates a positive impact on venture capital activity. Against this background, with empirical work going a bit all over the place, what should we do? What should antitrust agencies do? How did you deal with economic expertise in the past and what’s the best way to approach this expertise for antitrust agencies?
An enormous benefit that we had at the Federal Trade Commission, and that is one of the great strengths of the CMA, was to have a significant, endogenous technical capacity. We had lots of first-rate economists adopting quantitative methods. If you have that team within your agency, you can take opinions and positions in the debates, say what you think. And you can get a very good opinion from professionals who are working with you. In many instances, if these economists are part of a separate operational unit in your agency, they will act very much I find in the best traditions of a good academic faculty and they’ll give you a first-rate referee report on this kind of work.
In parallel with that, you run your own research, you do your own work, you invest in your own empirical work. Maybe it doesn’t retrace the steps of the other researchers completely, but you do your own work. If that’s not in your budget, if you are a small institution, you find pairs with whom you can work to do it collectively. And maybe you go to an external academic hub and you enlist their support in doing research. Another thing you can do is look at your own cases that pose comparable issues. Now, you might not have a completely representative dataset, and I’m aware of all of the complications associated with not having a representative data set.
It’s quite possible they’ve looked at your earlier decisions and said “here’s where you missed something.” When I was at the FTC, I voted to let the Google / Doubleclick merger go. Would I go back and take a look at that once again? I certainly would. Could it be the first time in my life that I’ve ever made a mistake [laugh]? It could be. I’m not going to concede that[laugh], but I’d go back and take a look. But I don’t want to discuss it in isolation. I want to bring in the people who made the decisions in the agencies. If the companies send some, that’s fine. Bring the critics, have a detailed reconstruction of the decision to see how it took place. So, an agency can have the right people, the right research agenda, and that puts you in a great position to do the assessment.
One other idea: you can convene events that bring the competing parties together and say “have it out, we’re going to give you a lot of time. It’s not going to be a five-minute summary of your paper, you can do more than that. But after you and your counterparts have laid out their point of view, we’re going to have an academic hot tub where the two of you sit here and discuss each other’s work.” A model for that, in my mind, is a conversation I saw at a dining room table in Chicago. It was at the end of a conference, this was the informal lunch get-together, and at the table were Dennis Carlton and Carl Shapiro. Dennis and Carl had both given their views about the extent of concentration trends towards concentration growth in margins. This was an earlier version of the larger debate about whether we have a big problem. That lunchtime conversation lasted about an hour. I wish I had taken notes. That would have been rude, but these were two people who were willing to play the ball the whole way. It was Dennis, who’s a thoughtful guy, it’s Carl, who was a thoughtful guy, and both of them by the way dove it points into deep technical details that escape me.
I’ve been a fugitive for mathematics since I was in high school, but I could tell from the other economists at the table that they were identifying and acknowledging methodological strengths and weaknesses in the two points of view. The result of the conversation was: “wouldn’t it be interesting to know more about XYZ?” That’s what we ought to be looking for. What you had was a completely constructive discussion, between two of the best we have in our field who put aside saving face, who said let’s not let’s just abandon those kinds of rituals and let’s just talk about the ideas right here.
I’d like to think that you could reconstruct that lunchroom talk with an audience of two, myself and one other person, on a larger scale. That’s a discussion worth having that I think really does lead you to think more. When Dennis and Carl walked away from that, you know there’s a bit of a natural rivalry, there they’ve been opponents as experts in cases, but if you talk to someone who’s acting in good faith and you walk away from that conversation, I think you walk away thinking, “God, that was really a good idea. That’s a pretty good point, I might have to think about this again.” That’s a discussion worth having, and maybe agencies, in affiliation with academic hubs, can promote that kind of conversation where you bring the people together and say, “we want you to have it out here, take your time, you got plenty of time to talk. You guys guide the conversation interactive.”
That is very concrete advice on the edge of the last two questions that I want to ask. The first of those two is the following. When discussing expertise, nowadays, we do have to discuss computer science expertise. How do you integrate computer science expertise within the agency? To acquire the expertise, of course, is a necessary step, but then the question is how do you work with those computer scientists? Do they intervene just when deciding the case or should they intervene all along?
I am aware that some small agencies might be listening to us, and I know they have some very concrete concerns about how to acquire the expertise — should they succeed in convincing the political power to give them the budget to acquire this expertise. Do you have any thoughts on that?
The conceptual argument for acquiring the expertise is compelling. The practicalities of doing it, as our colleagues and different agencies would say, are much more difficult. The conceptual argument is: “if you want me you want me to look at big information services platforms where the business models are complex, they change all the time, their ripples through the economy are myriad, and extraordinary, if you want me to do that, I’ve got to understand what I’m looking at”. You can have a lawyer and economist look at that all the livelong day, but the lawyer and economist will not understand necessarily what’s going on. It’s malpractice to have someone you broke your foot and you go into the hospital and they say, “our best pediatrician is going to fix it for you.” You say: “pediatrician, uh, orthopedist maybe?,” and they say: “Nah, we don’t have any of those, but it’s a great pediatrician. He tends to work on much smaller ankles than yours, but, uh, they read books on anatomy, they know what’s there.” So, the conceptual case is compelling.
For a bigger agency, I would predict that within five years, everybody has a significant team. Not a technologist, but a team. The CMA decided five years ago to build this. They’ve got a team of 40. Not just scientists and engineers. They have economists with high quantitative skills analytic. It could have flopped. It could have been a disaster where people sit in their rooms and nobody visits them.
Early on, they explain to others, “these are things we can help you do. We can help you frame a study, and the data you’re getting from parties as part of the merger review production… we actually know what’s going on in there. Do you want us to show you?.” “You’re getting hammered by these expert reports. Would you like us to help you tell what’s going on there, whether the assumptions are right?.” They had the right person in charge, they had the right team, and in their early efforts, they showed, wow, it looks a lot better that way. Now there’s a line outside the building to involve them in cases, the formulation of market studies, and, crucially in the development of the DMU. They’re 40 people now.
What do I do about a smaller agency? I think of collaboration with other agencies. I think of collaborations with university departments where I might not have somebody full-time, but I can get them part-time to help out. I want those kinds of resources available. If I had to ask one agency and they’ll help with this, “how did you do it? How did you make it stick?” CMA will tell you.
Very important! Here, I do have to mention that we’ve put together a group of scholars at the Codex Center at Stanford University, so, agencies are aware that they can always come to us, we’re more than happy to help.
That leads me to the very final question. Last summer, Nicolas and I organized the Antitrust Jukebox #1. We’ve invited all of our speakers that day to make three concrete reform proposals. Because of the fast pace evolution of antitrust, if I was to ask you what is the number one reform that you would implement, which one would it?
I’d bolster the program of evaluation and analysis of outcomes. And if I really begged you, I’d say, “I’d like to increase the human capital to do all of this well.” But I would dearly love to have a broader program of evaluation and assessment, and to have that joined up with the work of other institutions around the world so that you have the benefit of a broad pool of observations with better data and record-keeping that tells me what’s going on, how many abusive dominance cases were there in the last year, what kind, against whom, how are those going, where can I look to find that out? In short, a program that takes your prior assumptions in bringing the case and matches them against what actually happened to ask why did they converge — pure luck? — why did they diverge, and what does that tell us? The greater historical awareness and knowledge that would come from doing X post-assessment as a discipline to tell you what you want to do in the future, the better. That’s number one.
Beautiful! On that positive and very practical note, it is time for us to say goodbye to everyone listening to u. I just want to say a few things. First of all, Bill, thank you so very much for such a stimulating conversation. Also, on a more personal level, thank you very much for your support over the years. It means a lot. Thank you also to IBCI. It has been a fantastic conference. As I said two days ago, IBCI number four was more promising than IBCI number three, and I think the fifth will be even better. Please make sure to come back next year. Bill thank you so very much and enjoy the rest of the day.
See you very soon. The pleasure is all on my side. Thank you, it’s been great fun.