THE GLOBAL ANTITRUST ECONOMICS CONFERENCE
31 May 2019 — New York University Stern School of Business
Welcome Remarks: Luis Cabral & Lawrence White
Luis Cabral:
Good morning, New York. My name is Luis Cabral and I’m the Chair of the Economics Department here at Stern School of Business at New York University.
It gives me great pleasure to welcome you all to the Global Antitrust Conference. I’ve looked at the lineup and it’s a super-super-duper star-studded lineup, so I’m well aware of the opportunity cost of me standing here in front of you.
But I cannot resist telling you a brief anecdote that involves one of our speakers, Roger Noll, who is here in front of me, and that’s the following. Next week it will be thirty years to the day that I was defending my doctoral dissertation — I was five years old back then [Laughter] — and Roger was on the committee. I was told it was just a formality, but of course I was very nervous. I mean it was the first time I was defending a PhD thesis — it was also the last, by the way, thank God. My thesis was on regulation and antitrust.
So I go through the normal motions: I give my spiel and then the various members of the committee ask questions and so forth. So I go through it. Paul Milgrom said, “Okay.” Then Mike Riordan asks a few questions, then he’s fine. And Tim Bresnahan asks a few questions and he’s fine.
Then, finally, it comes to Roger, and there’s a pause. Remember, this is a thesis on regulation and antitrust. He asks, in a very matter-of-fact, serious manner, “So how does your work relate to macroeconomics?” — and then there’s a pause. He doesn’t remember that, but I do remember it very well, for reasons that are obvious; because at that brief moment, which was probably just a fraction of a second but it seemed like an eternity, I was actually kind of panicking because, to be honest, I did not know what the relationship of my work was to macroeconomics. But, fortunately, that was all broken by one of Roger’s unmistakable laughters.
Participant:
Guffaws. [Laughter]
Luis Cabral:
And I realized it was only a joke. It had a happy ending, so I did not have to answer that question, fortunately.
But fast-forward thirty years, and we hear antitrust is a very important and central issue in quite a variety of fields. If you can believe, in finance they are talking about joint ownership; is it an antitrust issue or 1 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
is it not? In politics, of course it has become very central to the political debate. And, even in macroeconomics, several of my colleagues here are precisely studying the decline in the investment rate in the United States and to the extent that that is the result of antitrust — or lack thereof, if you will. So, Roger, now I know the answer: it is actually quite related, and in fact it is quite important, not only for the macroeconomy but for society more generally.
A conference like this one could not be more topical, important, practical, and relevant. So it gives me great pleasure to welcome you all here.
The lineup is fantastic. I wish I could take some credit for that, but I can’t, so I’ll just bask in reflected glory for all the work that the organizers did for this conference and I will give the floor to one of the core organizers, my colleague Larry White. Thank you.
Lawrence White:
Thank you. I just want to again second and reinforce Luis’s welcome.
And let me mention that there were a number of organizers. Dan Rubinfeld, my colleague in the Law School here at NYU, was one of the important co-organizers, and so he deserves a great deal of the credit as well. As well as Nicolas Charbit, who I’m sure you’ve all met.
A couple of logistic issues that I think are important.
We are here on the second floor of this building. If you go out that exit, there’s a stairway that will take you down to the ground floor. If you go out here, you are going to discover there’s a bank of elevators immediately across. Unfortunately, those are express elevators, they don’t stop on this floor; and that’s compounded by the fact that all of our executive MBAs are here today. They do their classes on Fridays and Saturdays, and this is one of those Fridays, and that’s just going to mess up the elevators even more. So I urge you, if you want to get to the ground floor, the easiest way is just take that stairway down. If you really seek them out, there’s a set of local elevators, but then you’ll have to leave a trail of breadcrumbs behind you in order to be able to make sure you get back. So that’s just the easiest way to get down to the ground floor.
You’ve got the program in the folder in front of you and you’ve got the speaker bios. My strong feeling is when people have the bios in front of them already there’s just not a lot of extra value for an introducer to be spending a lot of time repeating what everyone can read right in front of him or her.
So, without much ado, I want to introduce Noah Phillips, who is one of the Commissioners on the U.S. Federal Trade Commission. He has been there for slightly over a year. You’ve got his bio. Commissioner Phillips.
Opening Keynote Speech: Noah Joshua Phillips
Thank you, Larry, and thanks everyone for having me here. It’s really a delight for me to be back. Just weeks after I first came to the FTC, I came up to New York to speak here at this conference, which is really one of the best conferences that goes on in antitrust every year. So it’s a pleasure and an honor to be back in the presence of people like Eleanor Fox, among others.
2 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
I’m supposed to start with the caveat, which certainly applies today, that the remarks I give are my own and don’t necessarily reflect the views of my fellow Commissioners or the Federal Trade Commission as an institution. Let me just begin with that.
Last year when I was here, I spoke about the idea referred to just a moment ago, that common ownership presented a competition problem and how the antitrust assumptions underlying the common ownership theory run up against our understanding of corporate law and policy.
Today I want to talk about another context where corporate law, governance, and antitrust come together; namely, the market for corporate control.
Considering the market for corporate control is critical right now. As politicians and the popular press focus increasingly on antitrust, they — and policymakers like me — need to take into account what experience and learning tell us about that market: we need to avoid policies that undermine its efficient functions.
The market for corporate control is essentially the competition for companies, competition to run them better. This competition comes from a number of sources, including other firms that may be current or future rivals; private equity firms; activists’ funds specialized in identifying and enhancing underperforming firms; and so on.
The market for corporate control is not about consolidation. In fact, it can be, and has in American history been, the mechanism for deconsolidation, forcing firms to focus on core strengths and shed extraneous efforts. This competition for the company can meaningfully benefit consumer welfare by ousting sub-par managers and keeping management generally on their toes, at least at public companies.
The market for corporate control can increase efficiency within a firm, making it a stronger competitor that, for instance, is more innovative, offers better prices or quality, or increases output.
The market for corporate control is a market mechanism that forces firms to compete. Given the competition — and, thus, the real consumer benefits that the market for corporate control can drive — I would like to spend some time today unpacking what it is, how it operates, and how antitrust agencies should consider its effects when making our decisions, the punchline being that policies that work with beneficial market forces work better and last longer than those that ignore them or work in opposition to them. We want to harness the market to achieve the ends of antitrust.
The notion of the market for corporate control was revolutionary when Henry Manne first outlined it in his seminal 1965 article.1 Up to that time, following Berle and Means, corporate law scholarship preoccupied itself with the assumption — rather dismal — that management had all the power and that shareholders had no meaningful way to constrain management, leaving shareholders open to self-dealing and managerial abuse, among other wrongs.
Antitrust law and scholarship for its part thought that mergers between competitors generally had no important saving grace and that their only justification, aside from increasing market power, was a sort of failing-firm defense, that the transaction would avoid a liquidation or a bankruptcy.
1 Henry G. Manne, Mergers and the Market for Corporate Control, THE JOURNAL OF POLITICAL ECONOMY, Vol. 73, No. 2. (Apr., 1965), pp. 110–120. 3 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
Manne’s groundbreaking work challenged both of these premises, introducing a new procompetitive rationale for such transactions. Drawing on the new institutional economics literature and their insight that we should not rely on a black-box theory of the firm and a plain-vanilla theory of markets, Manne questioned how transaction costs might be affecting firm organization and strategies.
His basic idea, in his own words, was “that the control of corporations may constitute a valuable asset; that this asset exists independent of any interest in either economics of scale or monopoly profits; that an active market for corporate control exists; and that a great many mergers are probably the result of the successful workings of this special market.”
To summarize, competition to run a firm, or competition for the firm, is a real phenomenon driving M&A strategy and decision-making. The competitive impulse “Hey, I can do that better” applies to management just as well as, say, product development.
Management matters. In fact, a book Gallup published earlier this month bears this out, finding that 70 percent of firm productivity depends on the quality of management. How robust are these findings? Gallup called this conclusion “the single most profound, distinct, and clarifying finding in its eighty- year history.”
In 1965, Manne was concerned that antitrust law was, but should not be, artificially limiting the market for management, i.e. the market for corporate control. In essence, the way that market control operates is that, all else equal, a relatively poorly managed firm’s underperformance is reflected in lower stock price, making the company an attractive target to those who believe they can manage it better. As Manne explained, “the potential return from the successful takeover and revitalization of a poorly run company can be enormous,” and the tremendous wealth generated by private equity firms and activist investors bears this out, at least in part.
But the point is that the value reflects tremendous gains from better management. The returns from improving a firm’s management are myriad and extends not only to the new managers and shareholders, but to consumers as well. For antitrust purposes, it is critical to recognize that many of these benefits coincide with consumer welfare-enhancing outcomes. For instance, a more efficient firm may increase its output, lower prices, offer better quality, or increase innovation, as I said before. And while we tend to think of innovation in terms of new technologies or pharmaceutical drugs, business innovation — like new management models and new pricing schemes, integration, and so forth — is, as Gallup’s recent book highlights, also essential to a thriving marketplace.
Competition to manage a firm thus operates to drive consumer welfare. Firms profit principally by providing value to consumers, by competing to offer superior products and services or better prices, or both, not principally by engaging in anticompetitive conduct, although they sometimes do that.
If the goal is competition — and it should be — then market mechanisms compelling firms to compete are not just good but essential. No amount of government intervention can match the power of the market; and, where the latter fights against the former, it will be less effective.
The welfare increases from the market for corporate control can arise not only from consolidation, as I said before, but also from deconsolidation. The theory of the market for corporate control does not depend on an increase in market share or a company’s size; it is about moving resources to their highest- valued use, and there is no ex ante rule for where that value may be located. The goal of the law, in keeping with Coase, should be to facilitate that move by reducing transaction cost, provided it is not otherwise anticompetitive.
4 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
As we’ll see, empirical as well as anecdotal evidence indicates that the consumer welfare benefits of a functioning market for corporate control can arise independently of size or market power factors. For years scholars and policymakers around the world have recognized and accepted the notion of the market for corporate control.
While Manne’s primary target in 1965 was antitrust law, corporate law has embraced his lessons more enthusiastically. Frank Easterbrook argues that the existence of a market for corporate control — I really need an acronym — should make us skeptical of incumbent managers’ resistance to changes in management, including for instance tender offers. In the United States, scholars like him and Jonathan Massey, as well as shareholder advisory services, push for governance structures that permit the market for corporate control to function and avoiding those that impede it by imposing unnecessary transaction costs.
The European Union adopted its 2004 Directive on Takeover Bids following expert input recommending that Member States adopt mechanisms “favorable to the development of an efficient market for corporate control in the EU.”
The broad acceptance of the positive impact of the market for corporate control reflects ample real- world evidence. Whether a firm’s underperformance is reflected in its stock price is, for instance, often readily observable, so we frequently see public instances in which an investor announces its belief that a firm is undervalued. Such actions can lead to changes in management, to changes in incumbent management behavior, or to other outcomes that may be procompetitive.
Facts surrounding a recent case at the Federal Trade Commission bear out a version at least of this dynamic. Last April, Genuine Parts Company announced it would spin off its office supply wholesale distribution business S.P. Richards and merge it with its direct and primary competitor in this space Essendant. Depending on how you define the market, this was a merger to monopoly.
Staples, a retailer of office products and related services, soon thereafter made an unsolicited initial proposal to Essendant. Unlike an S.P. Richards/Essendant deal, the Staples variant would not eliminate a direct competitor but, rather, would combine firms at different stages of the office supply distribution chain. Essendant declined Staples’ initial proposal, opting for the horizontal merger.
Just weeks later, however, Staples made another bid for Essendant, which Essendant’s board determined was reasonably likely to lead to a superior proposal, and further negotiations ensued. After months of back and forth, Genuine Parts declined to make any further offers and Staples emerged as the successful bidder.
As you may have read, the Commission divided on how to address the vertical merger. But I’m not aware of anyone who believes that the S.P. Richards deal, the horizontal version, was less anticompetitive. In other words, market forces helped drive the realization of an almost certainly better deal from an antitrust perspective, avoiding a deal between two primary direct competitors that almost certainly would have raised far more significant competitive concerns. The market for corporate control was allowed to act in concert with the antitrust laws and it yielded greater consumer welfare.
Empirical evidence corroborates that the market for corporate control generates value for shareholders, and not at a cost to consumers. Studies find that the evidence indicates that corporate takeovers generate positive gains; that target firms’ shareholders benefit; that on average bidding firms’ shareholders do not lose; and, further, that the gains created by corporate takeovers do not appear to come from the creation of market power — that’s a really important point for what I’m saying today.
5 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
These findings are consistent with the idea that the market for corporate control is real and not dependent on anticompetitive reductions in output or increases in market share; rather, it can be an important force for enhancing competition and consumer welfare.
All of this is not to say the market for corporate control always functions perfectly, that shareholder value always aligns with consumer welfare, or that M&A decisions are never driven by anticompetitive goals — of course that’s not true — but what this does establish is that the market for corporate control is one that we should try to foster and demonstrates the kinds of tradeoffs that come from chilling M&A activity generally.
The deconsolidating function of the market for corporate control underscores this point. While transactions consolidating high-profile companies tend to receive the lion’s share of popular ink, especially these days, deconsolidation and spinoffs, also important results of the market that I’ve been describing, tend to receive considerably less attention from the popular media.
I want to dwell a bit today on these kinds of transactions. And just to be clear, I take absolutely no view on the management or the criticisms or any of the proposals that I’ll discuss. I’m using them just to illustrate the market for corporate control in action.
This market can force deconsolidation on longstanding more traditional companies and high-tech ones alike. General Electric, which was founded as everybody knows by Thomas Edison in 1890, and which grew to become the quintessential conglomerate with everything from nuclear power plants to jet engines to Saturday Night Live, is currently in the midst of a significant restructuring.
After peaking around 2000 with the tech boom, GE’s share price tumbled in recent years. GE responded by announcing it would spin off approximately $20 billion worth of business units in order to focus on core strengths. GE has since spun off several business units — including energy, railroad, industrial engines, health care — to buyers including existing firms in the same or adjacent markets or private equity firms. When GE released its 2018 fourth quarter earnings report announcing that it sold $8 billion of assets that quarter, its shares increased as much as 18 percent.
Even Silicon Valley, often characterized today as being immune to antitrust laws or market forces, feels the effects of the market for corporate control.
Hewlett-Packard, for instance, was a pioneering Silicon Valley firm. From its start in a one-car garage in 1938, HP grew and expanded over several decades to become a global company involved in nearly every part of the tech business — PCs, printers, servers, supercomputers, software, storage, and so forth.
Beginning around the mid-2000s, analysts came to believe that HP would perform better if it split off these businesses, allowing the component parts to increase focus on their respective strengths. While HP resisted for a time, it ultimately split into HP Inc. (HPQ.N) comprised of the PC and printer businesses, and HP Enterprise (HPE) comprised of the corporate hardware and services operations.
This was not the company’s first foray into spinoffs — HP spun off what is now Agilent Technologies, its original test-and-measure business, over a decade earlier — but it was the most significant. The HPQ/HPE split became final in November 2015. By June 2017 HPE made two additional deconsolidating moves, spinning off its enterprise services to CSC and merging its software segments into Micro Focus.
By May 2018 HPE was up 42 percent and HPQ was up 67 percent, both beating the Nasdaq Composite, which was up just 31 percent. In the words of one reporter [Dana Blankenhorn, InvestorPlace May 31, 6 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
2018]: “HP’s fall from grace between 2010 and 2013 is now the stuff of business legend. A company that had been a doyen of Silicon Valley was laid low by a succession of bad managements” — “bad management,” to borrow from the reporter, that was ultimately corrected not by government action but by the market for corporate control.
With eBay, once a market-leading tech platform, two important and one so-far-still-potential spinoffs offer still more examples of the market for corporate control in action.
The first involves PayPal. PayPal IPOed in 2002 and became a wholly owned subsidiary of eBay later that year. In 2015, following significant pressure from Carl Icahn, eBay reexamined its ownership of PayPal and concluded that spinning off the PayPal Payments Division would allow both it and the eBay Commerce Division to have more focus, more flexibility, more agility, and more ability to move quickly, according to the CEO at the time.
Both Commerce and Payments faced increasing competitive pressure. For instance, in the months leading up to the 2015 announcement, Apple announced its new Apple Pay system; Google and Amazon were renewing focus on their existing payment options; and other competitors, like the now-prolific Square, were gaining popularity.
The market for corporate control was exerting pressure, which eBay management ultimately could not ignore, counseling deconsolidation that would allow the business units to play to their strengths. While the effect on eBay’s stock appears minimal as of earlier this year, PayPal has clearly benefited, increasing 134 percent since the split.
The second notable — and this one is still pending — example involving eBay is the recent call for it to divest its ticket sales business StubHub. Elliott Management argues that such a spinoff could bring in more than $16 billion. Starboard Value LP, another hedge fund, has echoed that sentiment. eBay opposes the proposal but has responded by streamlining its operations and taking steps to increase advertising revenue.
Again, I’m not weighing in on the value of any of these suggestions. I’m just illustrating how every day the market for corporate control operates.
What the efficacy of this market to force deconsolidation highlights is the role the market itself has to play in addressing the frequent concerns we hear about firm size. If that concerns you, you should want to unleash the forces that significantly restructured some of the great concerns in American history, like GE and HP; think of ITT. The market for corporate control has worked — and can work — to cut large firms down to size.
All these examples underscore the reality that competition to manage a firm is a real phenomenon underlying certain M&A strategies. This competition for the company is directly analogous to competition for the contract, which antitrust law recognizes as an important form of competition.
Firms may not compete day-to-day for the contract, but instead may compete periodically for exclusive contracts, bid for better shelf space or innovate products to win an RFP. In the same way, competition for the company does not manifest at all times, but it does nonetheless drive meaningful competition.
Competition to manage firms better in turn can yield consumer welfare benefits, as I’ve outlined before. It can thwart inefficient arrangements preventing the realization of such benefits much more effectively sometimes than government intervention, and it can do so without increasing a firm’s market power or size. 7 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
In fact, these benefits can arise even when the market for corporate control leads to a decrease in a firm’s size, as it sometimes does. That these benefits can arise even when a firm’s size diminishes underscores that beneficial buyers might be those who are not current competitors, be they potential competitors, startups, or private equity, or other buyers specializing in firm management.
The market for corporate control theory and evidence demonstrate that managing the firm has its own value, and so it is unsurprising that some firms may specialize in just this kind of management. Adopting a categorical rule that deems certain kinds of buyers bad for antitrust purposes would be itself anticompetitive. It would undermine this competition to run firms better, competition that in turn has real consumer benefits, competition that can also cut firms down to size.
In addition to squelching the consumer benefits that can arise when better managers are put in charge, there are further costs associated with limiting the pool of potential buyers:
• Doing so necessarily reduces the competition to manage firms, thereby decreasing the likelihood that better managers will oust the inferior ones;
• It also reduces the likely sale price of the firm by reducing the pool of bidders and the competitiveness of the bidders in that pool;
• It makes management more complacent, reducing their incentive to compete; and
• It decreases exit options.
The adage that barriers to exit are barriers to entry makes the general too-often-overlooked point that the harder it is to exit, the higher the cost of entry in the first place. This is a concern that should be kept front of mind in regulating our innovation-driven economy. Startups, for instance, help to drive innovation and dynamic growth, and are often cited as examples of why antitrust enforcers need to intervene to prevent mergers and acquisitions that threaten to gobble up all the startups.
But while M&A can threaten competition — and sometimes it does — it can also foster it. The vast majority of startups fail, and leading explanations include various managerial failings. Acquisition is a critical exit path for many or most that do not fail, the chilling of which will deter not only the innovation of the startups but the investment pipeline on which that innovation depends.
To preserve the innovation that has been critical to our economy’s growth over the last several decades we need to be cognizant of these kinds of potential tradeoffs and avoid chilling incentives for innovation unnecessarily or unintentionally.
Adopting dramatic legislation that would impair significantly the M&A market would threaten just such harms. Scholars have long recognized that legislation might undermine market activity that drives consumer and societal welfare. Aspects of, for example, the Williams Act and, key for antitrust purposes, the Hart-Scott-Rodino Antitrust Improvements Act raise such concerns.
Hart-Scott-Rodino is key to preventing anticompetitive mergers in their incipiency, but it balances the concerns that I have raised here today. To be most effective Hart-Scott-Rodino needs to be tailored to identifying and addressing competition issues only. Beyond that it loses its purpose and distorts the market for corporate control.
8 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
Coming back to Coase, the goal is to allow the assets to move to their highest-valued use and transaction costs slow that down or even grind it to a halt. In imposing these costs, overly aggressive legislation and regulation destroys real avenues for consumer value creation.
The idea of the market for corporate control in many ways revolutionized how we approach corporate law. It also has important implications for antitrust.
That M&A strategy is not always dictated by desires to take unfair advantage of shareholders or consumers, but by real opportunities to enhance a firm’s competitive position and societal welfare in the process, is one so appealing that in hindsight it seems it should have been articulated and accepted earlier.
Today we have the advantage of years of experience observing the market for corporate control in action. We know that it can lead firms to consolidate as well as to deconsolidate. It can be a force against concentration. We know that buyers specializing in firm management may be able to increase the efficiency of such firms in ways that benefit consumer welfare. And we know that impairing the proper functioning of this market can be anticompetitive.
Regulators, including antitrust regulators, should seek to foster its operation. By doing so we can help make markets more competitive and move resources to their highest-valued uses.
Moving forward, we should look for opportunities to unleash the market forces that spur competition and tread carefully with proposals that might inhibit it.
Thank you.
Lawrence White:
All right. We have time for a little bit of Q&A from the audience. Let me urge you if you have a question to raise your hand, briefly identify yourself, and briefly ask the question. Please, no speeches.
Question [David Sutcliffe]:
David Sutcliffe, Sports Technology. My question is, there’s a lot of talk from people like Jerry Nadler and others about breaking up Amazon, Facebook, and Google. Is the United States going to lead, or are they going to follow the Europeans while the Europeans set the standards for ethics and true competition as opposed to monopoly capitalism? It was a question.
Lawrence White:
There’s a lot loaded into that question.
Noah Joshua Phillips:
Let me say a few things. I’m not aware of European efforts to break up any of those companies. Let me start there.
I gave a speech a few weeks ago about the use of structural remedies in the context of Section 2 monopolization cases, which are the historical precedents sometimes trotted out for what you’re describing.
9 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
One of the things I said in that speech, which I believe very, very deeply, is that when you do law enforcement, including antitrust enforcement, you should target that which violates the law, not firms. Sometimes firms violate the law — sometimes big firms do; sometimes small firms do — but it is very important, for the rule of law and nothing less, that when we talk about antitrust enforcement we focus on the wrongs, not particular firms. I think that is going down a very, very scary road. That’s the first thing.
The second thing: when we do antitrust enforcement, we ought to think very carefully about what those wrongs are and how the remedies should be structured to address the wrongs. We should do that in many contexts, but especially in the context of conduct cases. What is the thing that we are objecting to, and what is the remedy that will achieve it?
There may be times when breaking up a firm will achieve that, but if the wrong you’re looking at is one thing and the remedy has nothing to do with it, I think that ought to give us a lot of pause. I think remedies ought to be structured to track particular wrongs.
So one of the things I’ll refer you to is a whole different speech that’s on my website, where one of the things I talk about is that, especially in the context of conduct cases, the general rule that we have of preferring structural remedies in the merger context to behavioral ones shouldn’t necessarily follow over in the conduct context.
In the conduct context, behavioral remedies can be more effective, and we have seen from history that sometimes structural remedies don’t really work that well.
The most famous of all, Standard Oil, didn’t really succeed in terms of any of the metrics that we use today to think about what the purpose of the antitrust laws is. If your goal was simply to break up Standard Oil, yes it succeeded at that. It created a series of local monopolies and prices went up. In fact, it was like the first job the Federal Trade Commission had to do in 1915 was study why oil prices were so high.
It’s like the Chinese curse, may you live in interesting times. The focus on antitrust today is super- exciting, it’s great. It’s for me as an antitrust enforcer invigorating to see ideas about antitrust talked about, not just in conferences like this but in op-ed pages, and by politicians.
But I do think we should not forget the importance of targeting that which is illegal, not just particular firms, and mapping remedies to wrongs, looking to address problems we see in the market with remedies that speak to them.
Lawrence White:
We’ve got time for one more question and then we’re going to have to move on.
Question [Cecile Kohrs]:
Cecile Kohrs with TIG Advisors. You were talking about the tension between innovation and competition. In pharmaceuticals now that’s a real issue. Gene therapies are brand-new. Isn’t it better for the FTC to stay out of the way of things that are way down the pipeline and let the technology evolve if there are mergers with those new technologies that haven’t been proven yet?
10 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
Noah Joshua Phillips:
My short answer is no to the question you’re asking. There is some really good academic work about what happens to drugs in development when they are bought by firms that have competing drugs. I think we need to look really carefully at mergers that involve drugs in their incipient regulatory process.
The FTC has a wonderful track record — I was talking about this in Congress a few weeks ago — over many decades of looking at anticompetitive conduct, abuse of the regulatory process that has anticompetitive effect, in the pharmaceutical industry, and I think we need to keep our focus there.
Lawrence White:
All right. It turns out we’ve got a little more time. A question over here?
Question [Jonathan Spitzer]:
My name is Jon Spitzer. I’m with CTFN. Commissioner Phillips, you mentioned the Essendant case, which obviously laid bare some intellectual division among the Commission. It strikes me that one application of what you were discussing today comes into play when a divestiture is offered as a remedy to an antitrust case. When word of divestiture gets out to the media, there is often talk of whether or not private equity buyers would be included as potential candidates and what the merits of that would be. From your remarks today it sounds like you would tend to look more favorably at these buyers as bringing in management expertise versus just financial acumen to the case. But could you talk a little bit about how that might be received by your counterparts on the Commission?
Noah Joshua Phillips:
I’ll leave my counterparts to speak for themselves. Commissioner Chopra has been very vocal about private equity, although I think if you actually read what he says and listen to what he says, it’s a little more nuanced than sometimes gets reported. But again, I want to leave him to speak for himself.
My view is that private equity is an important part of the M&A environment. When we look at private equity buyers, we should scrutinize them for the same things that we scrutinize other folks — like are you going to stay in the business and compete out? That’s an important consideration. And we can look at a private equity shop and what their plans are and what they have done in the past. These are valid considerations.
But, look, I think they need to be part of the pool. I think they are an important aspect — not the only important aspect but an important aspect — of the market for corporate control.
Lawrence White:
A last question, this time really the last question.
Question [Pallavi Guniganti]:
Pallavi Guniganti, Global Competition Review. You have spoken negatively before about conglomerate consolidation and this speech today seems to be picking up similar themes. I’m interested in how you see that difference between a large company that does many different things versus the consolidation into large companies that do one thing and dominate the market for it.
11 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
Noah Joshua Phillips:
Pallavi, thank you for the question. You know, I was looking as I was preparing for this speech for instances where you see the market for corporate control say that you’ve got a horizontally integrated firm; it really kind of does one thing but on a larger scale. I’m still looking. I didn’t mention it in my speech because I didn’t find it. That may be because that’s a more efficient business arrangement, right?
It’s possible that if you had firms merging in a way that wasn’t competitive, that the dynamic I’m describing wouldn’t operate as well because it’s going to favor efficiency, which, as I mentioned earlier in my speech, will sometimes be at tension. But my point is that it is not always in tension, and what you don’t want to do is chill M&A activity generally because there are a lot of places where the market can be harnessed to effigiate similar kinds of goals.
The other thing is that, while it’s not my view of the world, there are a lot of people right now who really are just talking about size, and they are looking at big firms that are not consolidating horizontally, they are going into new businesses. Let me take an example. Let’s say a giant online retail company buys a grocery store. This happened recently.
Lawrence White:
Not naming any names.
Noah Joshua Phillips:
However you feel about that, that’s a kind of conglomeration. We can have a discussion about logistics or whatever.
So I think this discussion remains very relevant even if its track record has been much more effective dealing with conglomerate-type arrangements rather than sort of horizontally similar ones, if that’s fair.
Lawrence White:
Please join me in thanking Commissioner Phillips.
Panel 1: Antitrust in Sports
Lawrence White:
We’ve got the next panel. Please come on up. I’m going to introduce them as well and we’re going to try to move right along here.
As you can see from your program, we are going to have a session on “Antitrust in Sports.” We have six distinguished panelists. Again, you’ve got their bios in your program and I’m not going to take the time. Content really is king here and I want us to get to the content.
The format is going to be the following. Initially I’m going to throw out a question and ask each of our panelists to address that question briefly, no more than five minutes; then there will be a second question, 12 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
again each of them addressing it for no more than five minutes. That will leave us time. We are starting ten minutes late and we will run ten minutes late because I want to give you guys a full opportunity. At the end of there we will have time for Q&A because I’m going to keep them to the five minutes.
Are you ready, gentlemen? Good.
James Keyte:
Start your engines.
Lawrence White:
I thought maybe I ought to announce them one by one when they came in through the door with cheering and confetti being thrown. But no, we’ll restrain ourselves. [Laughter]
The first question: What are the two most important sports-related cases that are currently in the courts and why are they important?
Let’s start at that end of the table, work across to the audience’s right, and then for the second question we’ll go in the opposite direction. Roger?
Roger Noll: As unusual as this is, I’m not going to answer the question; I’m going to put a different spin on the question.
From my perspective, the implicit premise of the question is one that is based on a broader history of antitrust, which is everybody who teaches a course in the law or economics of antitrust has iconic cases that sort of represent the core debates in antitrust that you use to illustrate various principles.
But in the case of sports, I would say the question is the equivalent of saying: If you watch a massive glacier calve icebergs in a bay in Alaska, which of those icebergs is responsible for carving out the valley that extends for twenty-five miles into the mountains?
It is not possible to do that, and that’s because of the nature of antitrust in sports, which is plaintiffs almost always technically win but almost always lose in terms of the nature of the relief. So, what antitrust leads to is a sequence of small victories, as opposed to a single big thing you can point to. It’s very hard in the history of sports antitrust to say, “Here is the crucial case that caused everything to change.” That’s because what happens is small gains.
Interestingly enough, this is a feature of U.S. antitrust, but it’s not the same in Europe. What I want to simply point to is we have no problem in Europe saying, “What is a crucial antitrust decision in Europe with regard to sports?” It’s the Bosman case. And so, what’s different about Europe that would allow us without hesitation to say, “The world changed in 1996 because of the Bosman case,” but we can’t make a similar statement about the United States? The answer to that seems to me twofold.
The first reason is something I’m not going to spend any time on — I’m going to let lawyers talk about that — which is the underlying difference between having competition policy be rooted in essentially a constitution, which is the Treaty of Rome, and the way things are spelled out about the free movement of labor in Europe as part of the deal of the European Union. That is fundamentally different than antitrust law in the United States, particularly rule of reason antitrust, which is what all these sports cases are about. Rule of reason antitrust is inherently something that leads you to case-specific remedies
13 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
that have a very short-term horizon. So that’s the first reason, and I’m not going to say anything more about it.
The second reason is that in the United States there is this magical thing that otherwise-weak labor unions have. Europe has much stronger unions than the United States, but, interestingly enough, in sports there are no collective bargaining agreements in Europe, because basically what the unions do is dispute resolution. In the United States, what unions do is grant antitrust exemptions to leagues. So rules that would normally be things that would expose you to antitrust liability get shielded if you can bribe the union and can put it into a collective bargaining agreement. Therefore, this tension exists: antitrust is used as a mechanism for consolidating the strength of unions by virtue of granting antitrust exemptions. That means a series of small victories, and gradually, as time progresses, sports become more competitive, but it’s a very slow, Chinese water torture kind of process as contrasted to the way it works in Europe.
Lawrence White: Andy?
Andrew Zimbalist:
I’m going to talk about the grant-in-aid cases that have emerged over the least fifteen years in college sports, starting with White,2 moving to O’Bannon,3 and then Alston4 and Jenkins.5 I’ll assume that others are going to discuss the details of those cases. Instead, I want to discuss some meta-questions that inhere to these cases and to the issue of compensation for college athletes.
I don’t believe that antitrust law and antitrust economics have within them the intellectual tools to adequately address the complex underlying issues.
First, as you all know, in 1984 the Supreme Court in NCAA v. Board of Regents6 set out a two-part process for assessing antitrust claims in intercollegiate athletics: Step 1, decide whether the challenged rule or behavior is fundamentally commercial in nature; and, if so, Step 2, since production is inherently cooperative, to apply a rule of reason analysis, wherein the anti- and procompetitive effects are weighed, less-restrictive alternatives are put forward, and presumably the judge or the jury balances the anti- and procompetitive effects. There are no bright lines here. More often than not, no balancing takes place and the trier of fact issues a subjective opinion.
Second, pretend as we might to the contrary, the absence of objective standards means that the individual judges or groups of judges, themselves political appointees, get to decide major issues of economic and social organization based on narrow yet ambiguous economic criteria. In the case of pay-for-play for intercollegiate athletics, judges who are not students of intercollegiate athletics or college life in general are making decisions about significant structures which have a profound impact on U.S. culture and U.S. universities.
2 White v. Nat’l Collegiate Athletic Ass’n, Case No. CV 06 0999 (C.D. Cal. 2008). 3 O’Bannon v. Nat’l Collegiate Athletic Ass’n, 7 F. Supp.3d 955 (N.D. Cal. 2014), aff’d in part, vacated in part, 802 F.3d 1049 (9th Cir. 2015), cert. denied (Oct. 3, 2016). 4 Alston v. Nat’l Collegiate Athletic Ass’n, Case No. 4:14-md-02541 (N.D. Cal. 2015). 5 Jenkins v. Nat’l Collegiate Athletic Ass’n, Case No. 4:14-cv-02758-CW (N.D. Cal. 2014). 6 Nat’l Collegiate Athletic Ass’n v. Bd. of Regents of Univ. of Oklahoma, 468 U.S. 85, 98–99 (1984). 14 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
Third, these cases expend a gargantuan amount of resources and time as they work their way through the judicial process. O’Bannon was filed in 2009; five years later, Judge Claudia Wilken filed her ninety- nine-page opinion which was appealed first to a Ninth Circuit panel, then to the Ninth Circuit en banc, which would not hear the case, and then to the Supreme Court, which indicated it would not accept certiorari in October 2016 despite the fact that the Ninth Circuit rejected the NCAA’s reliance on dicta from the 1984 Board of Regents case that the eligibility rule of amateurism was noncommercial in nature. So, after nine years, 1.2 million pages of documents, seventy-six depositions, eleven amicus briefs, and well over $100 million in legal expenses, there was an outcome of profound confusion over the direction of athlete compensation in college sports. Now a similar process is being played out in the Alston/Jenkins case so far only in the Ninth Circuit but not unlikely to pop up elsewhere.
The good news is that there is an alternative. Intercollegiate athletics, particularly in the upper reaches of Division 1, impacts many fundamental issues on the role of universities in U.S. society. How these issues are resolved should involve much broader considerations than just whether pay-for-play would increase output.
The structure of intercollegiate athletics is a matter for public policy. As ossified, corrupt, and dysfunctional as national politics have become in the United States, we have little choice but to figure out how to make our political system work. Without a functioning legislative and executive branch, the Earth and its inhabitants have little future.
Allowing decisions about pay-for-play to be resolved in the court system, particularly one infested with Trump appointees, fosters the illusion that by keeping government out of the economy we are deferring to the market system to make allocative decisions. In my view, this is a counterproductive ideological fantasy.
Lawrence White:
Really?
Brad Humphreys:
Larry, I’m only going to answer half your question, so I guess I’m a little bit better than Roger in that sense, and I’m only going to talk about one case. The case I want to talk about is the current City of Oakland v. Raiders case, which is about franchise relocation.
I was just thinking about this panel. Sitting to my immediate right are the two editors of Sports, Jobs, and Taxes: The Economic Impact of Sports Teams and Stadiums, which was published twenty-two years ago, and, to my left, a major contributor to that volume. I think that Sports, Jobs, and Taxes really was an important milestone in economic research on sports leagues and, in particular, antitrust oversight or lack of oversight of sports leagues.
Since that point in time, we have advanced the research agenda on understanding the economic consequences and impact of professional sports substantially, but on the public policy side nothing has been done. We are basically where we have been forever in terms of the ability of sports franchises and leagues to control where franchises are and franchises to do what we well know they do, which is extort taxpayer money out of local governments.
I think if there is any hope whatsoever of moving forward on this important public policy issue, it might come out of a case like this. I don’t have very much confidence that anything is really going to come out of that case, but still, if there’s a hope, I think it is going to be there. 15 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
Part of the issue is — Andy alluded to this — the inability of the legislative branch of our government to do anything about this problem. The exemption that leagues have was kicked to Congress a long time ago. They have done nothing.
So I would hope that if we’re going to do something to reduce the extortion by team owners of local governments and tax dollars, this is the case that we might expect, hopefully, something would happen.
Lawrence White:
Rod?
Rodney Fork:
I have the freedom afforded to somebody who’s not an antitrust economist. I’ve never claimed to be one, I work occasionally in the area, but I’m not one. I’m an economist who’s always looking for how economics can be interesting and useful to help the courts solve problems. That’s why it’s interesting to me. So I’m going to talk about two areas instead of two particular cases because of the economics that are just interesting there.
The first one for me is the ongoing litigation that has to do with antitrust and sports federations. To me that means the string of cases that are currently up against United States Soccer. We see Relevent Sports v. U.S. Soccer in the handling of international friendlies in America. We see NSL v. USF in terms of structuring the competition of sports into first and second divisions and the possibilities for promotion and relegation, which have to happen in soccer — I mean isn’t that the definition of it? And then we see the U.S. women’s national team suing the U.S. Soccer Federation over pay.
The interesting thing that I’ve noticed in all of these — and I worked on one of the precedent cases — is that there’s an assumption that these organizations are about money, that they’re about revenues, and actually, in the case of the U.S. women’s national team, using things that are based on profit maximization, terms that have to do with the marginal revenue product of men soccer players versus women.
When I go back to the founding legislation and the mission statement of the United States Soccer Federation, I see nothing other than “grow the game, make it one of the most important sports in the world,” and that doesn’t necessarily follow that you’re going to behave in a way that has to do with paying people who generate the most revenue the most money. And there is interesting economics about that that can be brought to bear on the case.
The other one for me is whatever the next case is about sorting out the issue of media and gambling market power. Right now I see firms positioning themselves in such ways that they are going to create that market power through exclusive agreements — I know, there’s nothing illegal about exclusive agreements unless they’re detrimental to competition, okay? This is going to be especially interesting because, at least as far as my outside use in an amicus way, courts have trouble with goods that are difficult to characterize, goods that have multidimensionality, goods that have joint consumption properties, and that’s what these things are all about. You can’t talk about these without talking like that.
Courts are then led into things like, “Well, what is the unit of output?” The answer is “games.” “How many physical games are played?” — because that’s what they said in 1984. Well, 1984’s media and gambling world was decidedly different than it is now, and so there’s going to be some very interesting economics, I hope, brought to bear to try to help people sort out those issues as well.
16 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
Lawrence White:
All right, we’re moving right along. Gentlemen, I really appreciate this. Michael, your turn.
Michael Hausfeld:
I have the privilege of saying, either by freedom or disclaimer, that I’m probably the only person on the panel that knows least about sports, that is as a sport. I tended in any sports litigation to basically screw up both the rules of the game and whether there were level playing fields or not, but I had the good fortune of having a judge who was very much like me. I’d like to address, as Roger does, at least from the U.S. perspective, elements of U.S. judicial consideration of sports law in terms of what permeates some of those decisions.
The first experience that I had with sports law dealt with the Major League lockout in baseball of the teams during the World Series. I had the audacity, as one court told me, to actually challenge that lockout on antitrust grounds on behalf of the fans of the city in which the Series was supposed to be played and the businesses.
What I learned in a very lengthy opinion was the judicial attitude towards literally baseball as the country’s pastime. There was a reminiscence going back to the 1800s, and possibly even before, as to the origins of baseball, the meaning of baseball, and how important baseball was, and that baseball essentially had a judicial immunity from antitrust accountability because it clearly was not a game that was played in interstate commerce. There was a sense by the judges that they would not divorce, at least in their minds, the significance of the popularity of the sport from the accountability of the sport under the antitrust laws which would apply to any other business which was not necessarily a sport.
The second, as Roger said, with regard to the Bosman case, the nonacceptance at times or in certain areas of sport not involving the free movement of labor or the fact that there was a labor market, in particular, for example, college athletes or amateur athletes, whether they be in college or Olympic athletes. There is small recognition by the courts that amateur athletes or college athletes deserve the protections of the antitrust law because they are not looked at as a labor market.
Putting aside wage issues, there are issues of name, image, and likeness which give value to today’s athletes over and above their playing on the field. How is that going to be addressed? If you don’t consider that value of those athletes as an athlete in their economic terms, you are not going to reach decisions that make sense from a market perspective.
One of the things that the court pointed out clearly in the O’Bannon case and in the Alston case is that, for whatever reason, college athletes have to be tethered with regard to their value in terms of the educational cost to the schools. Why, when for most of those athletes, as well as other amateur athletes, that value that they have is probably the greatest value that they are going to have during their careers? And yet, that is not recognized because there is some super-nonjudicial, nonexplanatory imposition of the fact that this is not a labor market.
Lawrence White:
Okay, great.
James, you’re batting sixth, cleanup.
17 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
James Keyte:
I’m quite outnumbered as a former Skadden partner who litigated a lot of these cases with some of the esteemed economists on the table, but that’s okay because the law is on my side. [Laughter]
In answering this question I wanted to talk a little bit about the law because it really sets the framework for where economics can come into play, whether amateurism versus other types of collaborations.
The first thing to highlight is U.S. law is not a regulatory law. U.S. law is about enforcement. Absent sham, the U.S. law does not really question venture formation or collaboration. What this means, at least in the history of sports antitrust law, is you can have what are viewed as closed collaboration, closed ventures. I know for a fact that on our panel we have a few economists who would prefer the European system, where if your team is bad, you get knocked down and you have to work your way up. And some of us might prefer that as well with the Knicks and some other teams. [Laughter] But from an antitrust perspective, closed ventures are the state of play and they’re perfectly fine and embraced by the courts. That has important implications for how to assess these cases.
The second thing that’s important is the days of viewing collaborations in any context of sports as cartels is long gone. That’s decades ago. Yet, I know there are economists and some plaintiffs’ lawyers who just want to say: “Everything is a cartel. You’re competitors, you’re talking to each other, you’re agreeing to things; oh, that’s a cartel.” Maybe there are exceptions.”
The law is just the opposite. The law is long past that. The law is long past the notion that the per se rule can apply, absent some really odd situation where you’re doing something outside the scope of the venture. This rule of reason really applies to all aspects, whether it’s on the output side, whether it’s on the input side, including with players.
We had a very interesting case — I think Roger was involved — involving Judge Duffy, the Williams case, where he ordered a one-day trial, which was crazy. It was essentially about a salary cap. That ended up going up effectively to the Supreme Court with Brown, where we ended up with the issue of the non-statutory labor exemption. But the interesting part about the case is Judge Duffy found that a salary cap passed the rule of reason in the alternative below. It’s all because the rule of reason actually had been applied to these player restraints for a number of years in terms of player systems, again all stemming from the idea of competitive balance.
And then we have of course American Needle, which, on one hand, just is a great case for lawyers and economists because a single entity is 90 percent dead absent some particular circumstances. But the interesting thing about that case is not the holding, it’s the dictum, the last two pages, which basically says: Okay, we’re not addressing the merits, but if it’s essential, the quick look that we’re all familiar with from the plaintiff’s perspective can apply on the defendant’s perspective; if the conduct at issue is essential to the operations of the collaboration, maybe a judge could take a quick look at that in favor of defendants, especially if it’s essential. That’s how it dealt with that aspect of essentiality; it rejected single entity and moved it into a quick-look rule of reason, and then it said the same thing about competitive balance.
The other decision to keep track of is Dagher — joint venture, rule of reason — because of the odd way it got to the Supreme Court, which said: if it’s a core activity (there it was coordinated pricing of a joint venture), it’s coordination but not in the antitrust sense. Again, it fits into a potential quick-look for inside venture restraints. It said the ancillary restraints doctrine, a reasonably necessary test, is only for limiting a venture’s members’ outside activities.
18 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
So, in that respect, looking at the two NFL cases going on will be interesting.
The NFL Sunday Ticket case: Will it be a core activity under Dagher? Will American Needle’s defendant’s quick-look apply? And then, if you get to the rule of reason, you have interesting issues about market definition and market power; the counterfactual “what would the world look like?”
The same with the relocation case. The old Raiders case is not the relevant case. It’s the CBC and SDC case that really again almost gives a presumption that it’s okay to control that.
But I still think the thing to watch for in those cases is how American Needle and Dagher potentially will apply to those internal agreements.
Lawrence White:
Okay. Very good. As you can see, we’ve got a diversity of views and opinions here.
My second question, given that I expected that most of the audience would be on the attorney side of the house, I thought it would be interesting to have our six panelists address the following question: What are the two most important sports issues that are currently not being addressed in litigation but that may be litigated in the future?
I’m going to now go in the opposite direction. James, we just heard from you, but we’re going to hear from you again and then move down the table.
James Keyte:
Well, this is awkward because it’s going to be just more of the same.
Lawrence White:
That’s okay.
James Keyte:
It will be a little bit more of the same.
I think the important issue, first, is pretty broad, and it has been touched on already, which is the role of the court in looking at sports collaborations, again whether professional, whether amateur, whether closed ventures, whether sanctioning bodies.
There is a tendency in some courts to want to second-guess the internal decisions of otherwise legitimate collaborations. I think you see in these NCAA cases — again, they say the rule of reason applies; the Ninth Circuit says amateurism is part of the product and that that’s okay — but then they have judges kind of deciding, Well, what does that mean and how much money can be spent? and you have essentially judges running joint ventures or collaborations that are legitimate.
I was actually surprised that that couldn’t have been dealt with actually on a motion. I was also surprised — again I’m not involved in the case; I’m not practicing law — I also found it odd in the case that you’re litigating in a sense essentially the collaboration design: I have a product, which is amateur sports; I’m going to litigate through surveys whether that is generating demand.
19 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
The marketplace should decide whether that’s a good product, and I think the marketplace has decided that amateur sports, if it is a commercial product, is a pretty damn good product. So I think to be litigating whether that is a good product on the demand side seems not the role of the court. I think that has to be addressed from a lot of perspectives.
The other perspective I think really is: What is the right analytical framework for dealing with sports cases, whether they’re sanctioning bodies or closed joint ventures? This gets back to this notion of — you know, everybody knows it’s got to be rule of reason, but I think you just don’t hop right into a rule of reason market definition, market power, effects, justification, less-restrictive alternatives; throw it in the mix and see what happens.
I think the combination of what American Needle did and what Dagher did asks the courts to really think at a very preliminary stage — whether it’s on the motion, whether it’s summary judgment — to assess essentially what is inside a venture and what is outside a venture.
The Dagher case is fascinating reading. It’s a Supreme Court rule of reason joint venture case, obviously not involving sports. Everybody thought the ancillary restraints doctrine historically was — okay, you could assess an internal decision of a legitimate collaboration about whether it’s reasonably necessary. Well, the Supreme Court in Dagher said: “No, no, ancillary restraints is for the restraints on members’ outside activity — think about NASL cross-ownership rules, things like that. Internal decisions of legitimate ventures we don’t want to get into.”
That has not played itself out in the sports cases, notwithstanding a few amicus briefs and other things that I wrote while I was at Skadden. I think it will play itself out. So, if you then combine that with American Needle saying, “Hey, a lot of these restraints in sports leagues, including related to players, they talk about competitive balance — where do you put your output, where do you put your teams?” — a lot of that can be dealt with without going into the full rule of reason.
I think these are the issues that need to be decided: What are the preliminary quick-look-type of analysis that can be done on the defense side, before you get to the full rule of reason. Of course, then when you get to full rule of reason, I think — the last thing I will say, and Roger dealt with this when we were in the Laumann case — is a lot has to be done in terms of what is the counterfactual; absent the agreements you have, what will the world look like? I don’t think the NCAA did a very good job of looking at that — you know, what would output be; what would price be? — very complex modeling that can go on.
But, just also intuitively, it’s not an easy decision to say, “Oh, I don’t like that joint venture, that rule. We’ll just sort of start from scratch and see what the world looks like.” Well, often that world is going to have lower output, higher prices, maybe even worse results — in the NCAA situation for the 99 percent who aren’t going to be professional players but may not have the same experience, or even programs available. So I think counterfactual is something that needs to be explored by the cases.
Lawrence White:
Michael?
Michael Hausfeld:
I disagree that there is a concept of any quick-look rule of reason adopted in any judicial decision to date. A quick look only applies to whether or not a particular business conduct is violative per se, on its face. Rule of reason clearly is the only standard by which the judiciary judges sports antitrust conduct.
20 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
I see at least a challenge to antitrust accountability in sports coming from a non-sports case, and that is the decision by the Supreme Court in the Amex case. The Amex case threw into confusion at least from our perspective, in a rule of reason “What is the burden of the plaintiff in establishing a prima facie case which can overcome a motion to dismiss?” Normally it was if the plaintiff demonstrated the existence of a restraint; that was sufficient then to throw the burden onto the defendant to show the presence of a countervailing procompetitive justification; then that would switch back to the plaintiff to show that that procompetitive justification could be achieved by less-restrictive means; and then there was supposed to be a balance as to whether or not as a whole the restraint was more anticompetitive than it was procompetitive.
Amex changed that, and they said that the plaintiff in addition to establishing a restraint had to show the absence of a countervailing procompetitive benefit. If that is the situation on a motion to dismiss, you may see a lot of these cases in sports leagues being dismissed at the early stages because without discovery it’s very difficult to demonstrate the benefits or the lack of benefits of a procompetitive defense.
The second is the application again of the Amex concept of a two-sided market. You might say, “Well, how does a two-sided market apply to sports leagues?” That’s precisely the argument that was used in the Alston case to request Judge Wilken to dismiss the case because the plaintiff had not established that there was an equilibrium price that they could demonstrate in their case in chief without countering whatever the NCAA would establish is a countervailing procompetitive benefit so that there would have to be a weighing of those two sides of the market. I don’t know how that’s going to play out — Judge Wilken dismissed that argument almost as being spurious — whether this is something that is going to play in the sports leagues.
The other is the concept that has arisen recently of consumer welfare and antitrust policy. Is, as James said, what we’re looking at a litany of policies concerning consumer welfare that goes into the balancing of accountability for an antitrust sports violation; or is it just merely the suppression of competition or the elimination of competition by the sports league? That is a totally different emphasis. So you wouldn’t have to decide whether, for example, college sports would be less popular than it is today but whether or not there is just a conduct being engaged in in the field which is restricting or suppressing competition.
Lawrence White:
Rod?
Rodney Fort:
I’m going to go ahead and try to answer the question and then wander off again. [Laughter]
I think the next most interesting case, the one that people will focus their attention on, is what happens in the next go-round of the NCAA and the name/image/likeness problem. It’s an interesting evolution because the NCAA — remember, everybody’s acting here; they’re not a monolith that everybody thinks they are.
When it was clear that the courts were going to take away their ability to reduce player pay to below full cost of attendance, they formed the autonomy group and — gee guess what? — now they can pay full cost of attendance. Now they see the name/image/likeness right handwriting on the wall and they formed a special task force to think about name/image/likeness rights, which, if history prevails, will be an arrangement that the NCAA thinks that the courts will live with where they can still keep most of the name/image/likeness rights but give some right to the players. That will certainly be litigated. Judge 21 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
Wilken wants to review those things, but I think that it will ultimately be litigated because they won’t do what she wants.
The reason that I think that’s also interesting from my previous observations about how economists can be useful and wonder about this is there’s a group of sports economists who think antitrust approaches to sports, especially in the labor arena, are counterproductive, wielding an imprecise tool in a sort of “one size hammer fits all.” I wonder about those things too, if we want to wander away from, “Okay, well you can do antitrust or you can do something else.”
I send my undergraduate students through the following exercise. I make them go to irs.gov, get Form SS-8, the form where you have the IRS help you determine whether you are an independent contractor or an employee. I ask my students — some of whom are athletes, and I pay particular attention to those when they come back — to imagine they’re an athlete and go through and answer that form about what is required of them as they are athletes who are students. Nearly universally, the answer is, “Uh, I’m an employee. All the boxes that I check and all the answers that I give would probably lead the IRS to say I’m an employee.”
This makes me wonder about those naysayers on antitrust, about whether the right way to head to pay- for-play — and I don’t like that term because we don’t even know what it is — maybe we would have to ask the athletes how they want to get compensated, right? I mean that’s how we all do it. But, instead, we try to devise lots of very cool ways to handle that for them.
Well, if it really boiled down to that the IRS would determine that athletes are employees, then it’s off the antitrust table. I keep wondering why a smart lawyer doesn’t go in that direction. I see head-shaking, and I see them all thinking It happened that way before, and I’d love to have you fill me in.
Or, if you like the antitrust approach, you could adopt part of what went on during Board of Regents. The schools argue: “Because the NCAA is involved in antitrust, they’re asking us to break the law, almost like co-conspirators, so you need to stop the NCAA from doing that so that we are not breaking the law.”
Well, suppose I have clients who are college athletes and I bring that same idea. Right now my clients are willing to attest that they are employees; they’re willing to fill out the Form SS-8, tell the IRS they are employees; and the NCAA rules are forcing them to file false income tax returns.
I wonder why these other ways of approaching the issue are not coming forward and instead it’s the — I get the idea that you can get compensation, and sometimes nice compensation, for you and your client, and you can occasionally establish precedent through an antitrust action, but in terms of actually moving the whole issue forward, simply getting the athletes granted their right to have the same rights to their stream of value as any other undergraduate would seem to me to be the noble ambition, and there seem to be a lot of ways to do that.
James Keyte:
Rod, let me introduce you to Michael, who tried that approach and I think it didn’t prevail.
Rodney Fort:
Oh, okay.
22 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
Michael Hausfeld:
It didn’t prevail before the National Labor Relations Board either.
James Keyte:
But it’s a good instinct.
Lawrence White:
All right. We will come back to that.
James Keyte:
He litigated that.
Rodney Fort:
So it’s just my ignorance, but that’s okay.
Brad Humphreys:
Following up on my earlier comment, I think obviously the case that needs to be revisited is Federal Baseball because the exemption in there is still widely applied to franchise location issues. Again, I think that’s a big issue in sports that needs to be addressed.
But I want to follow up because I think another unexamined issue — both Michael and James talked about competitive balance and the important role that competitive balance plays in consumer welfare and in leagues’ defenses of things that they do that look anticompetitive. That all rests on an idea that sports economists call the uncertainty-of-outcome hypothesis, the idea that fans prefer games or matches with uncertain outcomes to those with certain outcomes. Over the last several years, I have somehow taken up a crusade to try to argue and persuade sports economists that that idea is wrong, that the uncertainty-of-outcome hypothesis just, first, rests on pretty shaky theoretical grounds, which I think I’ve showed; and, second, is not widely supported by empirical evidence, at least at the game level; whether it might at the season level I don’t know.
I just think it is tremendously important to get this sorted out because of the vital role that competitive balance has played in antitrust law applied to sports leagues. I think that fans like to watch games that have relatively certain outcomes. In particular, they want their team to win, and they don’t really care about seeing games that the team they’re a fan of might or might not win. I just think that doesn’t accurately describe sports fans’ preferences. If that is the case, then it throws into question a lot of existing antitrust law. I think it is largely unexamined in the literature, and untested in the courts as I understand it. So that’s what I’d like.
Lawrence White:
Andy?
23 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
Andrew Zimbalist:
I want to pick up on the proposition that Brad offered earlier about extortion of subsidies and, in particular, his hope that the situation with Oakland will produce some positive steps in the right direction.
Brad Humphreys:
I didn’t say that.
Andrew Zimbalist:
Didn’t you say that?
Brad Humphreys:
I said potentially it would.
Andrew Zimbalist:
Potentially, okay, and you hope that it does. I hope that it does, too, but I’m not optimistic.
I think if we look back at a number of instances over the last thirty or forty years, we discover that, in spite of the partial success that the Los Angeles Coliseum had back in 1982 and some cities that have threatened antitrust action that resulted in salutary outcomes for them, there has been no city that has sued on antitrust grounds and has either kept a team from leaving or brought a team to the city.
I think the kinds of explanations that one hears about why those outcomes obtain is that antitrust law is about competition, it fosters competition, it does not protect competitors. You hear that what these cities want is to join a monopoly league; they are not looking for open competition. And you hear other explanations along those lines and, as a result, that reasoning has prevailed and cities cannot protect themselves with antitrust law.
So I think here again this is an area where there has been some reliance upon and substantial expenditure of resources spent on antitrust that has not produced productive outcomes, and once again it is an area where we need some thoughtful public policy.
It has been attempted in the past. Senator Moynihan attempted to eliminate or greatly reduce the amount of public subsidies through the tax exemption for stadium financing. That was in the 1986 Tax Reform Act, trying to redefine private activity bonds. There were too many loopholes in what he proposed and it didn’t accomplish what he wanted. But that doesn’t mean that the Congress should give up. I think that at the end of the day there could be much more straightforward and direct legislation that would curtail the extortion problem that Brad was alluding to.
Lawrence White:
Roger?
24 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
Roger Noll:
I would like to go back to my glacier analogy. The valley is melting due to climate change, but there is still twenty-five miles’ worth of ice, and there’s going to be a lot more calved icebergs that are spun off.
The reason for it is endemic to the nature of antitrust and sports, which is I don’t have as negative a spin on it as Andy, but essentially his description is correct that a natural consequence of rule of reason in antitrust being that the way one thinks about the application in sports is that, because the rules keep changing and the economic circumstances keep changing, nothing is ever finished. That is to say, all the same antitrust issues that we have litigated in the past fifty years will be relitigated again numerous times in the next fifty years.
A good illustration of that is the sequence of antitrust cases regarding the NCAA. The NCAA did try to defend itself in Alston on the grounds that it was already decided in O’Malley, and the judge just paid no attention to that essentially, saying that things had changed; you went from one set of rules regarding scholarships to another set in terms of what the fundamental scholarship would be; you went from one set of rules about what other benefits could be provided to athletes to another set of rules, and this new set of rules is producing over $5000 per athlete per year in basketball and football; the net effect of going from the old grant-in-aid (GIA) cap to the cost-of-attendance cap was estimated ex ante to be $2500 per athlete, and it is in fact $6000 per athlete because it turns out there is no serious constraint on how a university defines its cost of attendance and, magically, a bunch of Southeastern Conference schools discovered that the cost of going to them was roughly twice as much as they had previously estimated.
All this adds up to college athletes between O’Bannon and Alston were getting paid roughly $10,000 a year more than they were prior to the O’Bannon decision. Now, Andy may think Who cares? But I care, because 70 percent of college football and basketball players in the Power Five conferences qualify for Pell grants; that is to say they come from families that are in or near poverty. That strikes me as a big victory for antitrust. Michael should be congratulated on producing roughly ten grand per athlete in basketball and football in Power Five conferences.
But it will be litigated again because, regardless of what the final court of appeals decision is in Alston, the circumstances will change again, and five to ten years from now somebody else is going to say, “Oh, these new rules are anticompetitive” and it will be relitigated again, and who knows what the outcome will be?
One final point. There’s as big, huge issue coming down the pike, which is the role of gambling in sports. I was brought up on the notion that gambling is unbelievably destructive to sport, and examples were, for example, the University of Kentucky gambling scandal in the 1950s.
Lawrence White:
How about NYU shaving points in the 1950s?
Roger Noll:
That’s true. Well, I wasn’t going to be nasty to my host, but now that you’ve pointed it out — [Laughter]
Few people know that part of the land on which the new football stadium in Los Angeles is going to be located has been deeded to an Indian tribe for the purpose of putting in a casino. The Raiders’ new stadium in Las Vegas will have casino activities inside the stadium.
25 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
We are about to go into a world in which both college and professional leagues will face a decision of trying to create a new entity that is the formally licensed gambling entity of that sport, being justified by they need to control it to avoid the unsavory aspects of gambling, but in reality to make certain that the billions of dollars of incremental revenue that are going to arise from more intense real-time betting, sitting in your seat at the football game and betting on whether the next play is going to pick up a first down — those kinds of betting activities are within five years of happening, and they are going to be attempted to be monopolized by the entities that govern professional sports.
That is going to be the next big version of antitrust case about what is legitimate within the joint-venture activities of a sports-governing organization.
Lawrence White:
All right, great. We’re going to get back on-schedule, gentlemen. Thank you. We’ve got fifteen minutes for audience Q&A. Again, the previous rules: please identify yourself briefly, please make your question a question, and make it brief.
Question [Steve Ross]:
Steve Ross from Penn State University. I’d like the panelists to comment on the keynote address. Focusing on professional sports, one of the things that the panelists did not bring out is there is no market for corporate control of clubs in professional sports.
Lawrence White:
Because they’re not publicly traded.
Questioner [Steve Ross]:
Because they’re not publicly traded. The NFL actually has a rule that prohibits a market for corporate control; based on Roger’s testimony, the First Circuit threw it out. So, if the panel could comment, drawing on Commissioner Phillips’ keynote address, whether law or public policy ought to address that as a problem.
Lawrence White:
Anybody want to answer? James?
James Keyte:
I’ll take it on. I was involved in a later case, the Sullivan case the second time around.
Questioner [Steve Ross]:
That’s what I’m talking about.
James Keyte:
I think the antitrust laws can at least be focused on that issue. I think it’s a very interesting issue of market definition. If you’re really talking about, “Okay, I have a lot of money and is professional sports capital a distinct economic market” I think to me becomes the key issue there. 26 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
And then, are you going to narrow it to those who prefer or have an interest in their billion-dollar hobby in football versus hockey? What about if it’s in Canada; can it be in Europe? It all could be done anywhere.
So I think there could be a marketplace to be assessed. I just don’t think that it raises any interesting antitrust issues because it’s so broad.
Lawrence White:
Anybody else?
Michael Hausfeld:
Dividing that into two parts, one with regard to professional sports. Taking the keynote address that there is a value to corporate control, it would seem the answer would be yes; you should see what the market would produce with regard to controlling those sports leagues.
With regard to the NCAA, I have a question for the audience: Who controls college sports? Is it each university; is it the amorphous NCAA, whatever that is; or is it the conferences? Before you can answer the question of whether there’s got to be outside control, you’ve got to decide who really is the controlling entity behind college sports. Then the answer has to be, if there is no recognition realistically in the market that college sports is a business, then the answer is that you have to open that up to competition for governance of that business.
Brad Humphreys:
Steve, I was reflecting on this very point during the keynote speech. I think it’s constructive to think about the differences in governance that we see in Europe compared to in North America in this instance. If you look at the top professional football leagues in Europe, you see a lot of different ownership forms. There are sugar daddy owners, but there are also fan-owned clubs, and a lot of Premier League teams had shares floating on stock markets for some time.
The question that comes to my mind is: How do we get from here to there? I think that some of the problems with sports that we’ve been talking about here would be addressed by the creation of a market like this. But I don’t see how we get from where we’re at here in North America now to more of a market in this. Without public policy I don’t see a legal path.
But I think it’s a really interesting question and makes a nice tie-in here.
Lawrence White:
All right.
Andrew Zimbalist:
If I could, I was also reflecting during the keynote, and my mind went immediately to Marge Schott.
James Keyte:
Why?
27 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
Andrew Zimbalist:
Then it went to Frank McCourt. I hope, James, that when you were listening you thought about James Dolan.
James Keyte:
Yes.
Andrew Zimbalist:
There’s little question but that a market for corporate ownership would be beneficial to sports, and the reason it doesn’t happen is because we have closed leagues that are very tightly controlled and that set rules about becoming a member of the club. So the positive process that was being described in the keynote is something that sports is immune to and sports fans suffer as a result.
The problem I have — and maybe this is also what Brad had in mind — is, how do we get there, or how do we get to a more open system? Who’s going to be the plaintiff and how is the court system going to react to that?
Let me just say, because I might have misled you in some of my comments about antitrust earlier, I think that antitrust is fine. I just think that it’s important to see it as a complement to, not a substitute for, public policy.
I certainly am not unhappy that college athletes can receive $10,000 more on top of their old grant-in- aid. I think that’s a productive step. I just think that there are more direct ways and more comprehensive ways of addressing those issues.
Lawrence White:
All right. Next?
Question [Thomas Lopez-Pierre]:
My name is Thomas Lopez-Pierre. I’m not an attorney. I’m a private equity guy at Harlem Real Estate Fund. My question is for Professor Roger Noll and for the rest of the panel. In your last comment you spoke about the potential future of being in a stadium and gambling on the next yardage. That just seems like crack for gamblers. Could the panel talk about the likelihood of that actually happening? I mean was that you just trying to be cute or is that the future of sports?
Roger Noll:
Of course I’m always cute. [Laughter]
Lawrence White:
You’re the cutest guy at the table. [Laughter]
28 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
Roger Noll:
We do not know what the plans are of Stan Kroenke in Los Angeles and Mark Davis in Las Vegas with regard to the details of how they would plan to integrate their gambling activities into their professional sports activities. What I said, however, is that I have discussions with people who are in the casino industry in Las Vegas who believe things like that are going to happen and the only issue is what is feasible and administrable and is not going to get you in both political and legal trouble to do.
James Keyte:
What do we know about the British betting?
Brad Humphreys:
Already going on.
James Keyte:
They bet on everything.
Brad Humphreys:
There’s already in-play betting going on.
Roger Noll:
In Europe there is in-play betting. That’s what they call it, it’s called in-play betting. I was talking about this to Andy last night. In baseball we’ve had a debate for the last twenty years about how the game is too slow and try to speed it up. Gambling is going to reverse that because the longer it takes between pitches, the more bets you can have.
James Keyte:
That’s so cynical. [Laughter]
Brad Hausfeld:
But true.
James Keyte:
In absolute terms it’s true. Whether that’s a venture goal is different.
Lawrence White:
Rod?
Rodney Fort:
My thought for you is just watch what the states do in the way that they define the way that sports betting is going to be able to happen in their states. I think you are going to see a variety of different ways. For
29 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
example, and the one that facilitates the kind of in-play bet, is whether they can use mobile phones to do it or not, right? I think that’s open to debate and a sticking point for some states about how that’s going to shake out.
But we do know that teams and leagues are already forming alliances with exactly the kind of firms that are going to be able to do exactly the world that Roger described for us. To think that’s not going to happen short of state restrictions, I think you’ll be missing the point.
Roger Noll:
And I’ll bet you ten bucks the next panel doesn’t finish on time. [Laughter]
Rodney Fort:
Can we get the odds on that?
Lawrence White:
James?
James Keyte:
One thing is this is an antitrust panel, so on this topic, setting aside the potentially addictive element of gambling and public policy issues, it’s a question of venture formation. For those reading the law, this is “three tenors” — we have a venture that starts here, now we’re going to expand our venture.
When you look at that venture from an antitrust perspective, you are looking at the same kinds of issues: closed ventures get to define what they want to do together. If you’re not eliminating competition in a meaningful way in expanding that venture to other areas, the antitrust laws aren’t going to have much to say about it independent of other public policy issues.
Rodney Fort:
My response would be short of American Needle, right? I mean I don’t know that it has been established that leagues can consolidate the exclusive contracting arrangement on gambling issues across teams.
James Keyte:
There will be litigation over that.
Rodney Fort:
Okay, good.
James Keyte:
But it will be kind of independent of these other policy issues. It will be more about can they compete against each other independently; do they need to coordinate, and to what extent — you know, the same old kind of antitrust analysis.
30 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
Rodney Fort:
We’re in agreement. I just wanted to make sure the American Needle thing got in there.
James Keyte:
Oh yes.
Andrew Zimbalist:
Can I just add that I don’t want to leave the gambling issue without tying it back to the discussion of pay-for-play in college sports? It seems to me that, whatever unpleasant things might be happening down the road in professional sports, the threat to college sports is much greater.
If one reflects back on Charles Kaminsky and the Black Sox scandal in 1919, the root of that was that players were not being paid by the White Sox as much as they were on other teams. Players felt cheated. Players did not have high incomes.
I think that college athletes who are increasingly conscious of at least the sense of exploitation and certainly are living impecunious lives are very open subjects for gamblers. I can see immense problems.
The NCAA, as far as I can tell, is going to embrace gambling. They’ll say that they want to control it and they’ll want integrity fees and they’re concerned about it. But the control of gambling and its impact on college sports I think is extremely problematic.
Lawrence White:
And, as many of us are familiar, think about eighteen-year-olds being approached by and having to make in-the-moment decisions. It makes you wonder.
Questioner [Thomas Lopez Pierre]:
Eighteen-year-olds from low-income families.
Andrew Zimbalist:
Exactly.
Lawrence White:
Which is Andy’s point.
Is there another question?
Question:
My name is Kemal.* I’m an undergrad student at St. John’s. My question pertains to what the implications of geopolitics are being involved in the daily operation of sports leagues. For example, a few months ago the current Administration handled the deal between Major League Baseball signing Cuban baseball players from the Cuban Baseball Federation. And then, a few weeks ago, the NBA
31 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
canceled the League Pass for games being broadcast to Turkey. What are the risks associated with that and how do you think that can be resolved?
Lawrence White:
So geopolitics in sports.
James Keyte:
Not much antitrust there to talk about actually in my view. But, oddly, there is a case applying a defendant’s quick-look, Michael. There was an issue of whether the NHL could coordinate on how they deal with a Russian league in getting players. The court found essentially that that is an internal league decision from an antitrust perspective. Clearly, it is relevant in a sense to geopolitics, if there were interesting geopolitics with those countries at that time, which was probably not true then but may be now.
One thing that those who practice law when they have foreign clients sometimes don’t realize is that when you’re in a courtroom with a judge dealing with antitrust laws, typically, subject to changes in consumer welfare definitions, the geopolitics and other politics, and even other important policy concerns, just fall away in the antitrust analysis. They just do.
They might influence a judge subjectively, but for the most part there is so much law that restricts the court’s thinking on what it can consider — primarily to a consumer welfare standard, not even a total welfare standard — that those tend to fall away even if they might have some other implications for what Congress may do or what other regulators may do.
Brad Humphreys:
I think that it’s going to play a huge role in domestic U.S. league outcomes because the NFL is now playing games in London every year, many games in London every year.
Lawrence White:
Mexico and Japan.
Brad Humphreys:
Mexico and Japan, and the NBA is huge in China, and there have been rumors of attempts to establish permanent franchises certainly in London. Geopolitics is going to have a big influence on that, and potentially I think the antitrust implications of that, which I don’t understand.
James Keyte:
I’m not sure there are any because they are controlling where they place their teams in a closed venture, and the U.S. laws are pretty much going to let them do that.
Michael Hausfeld:
That’s what James thinks. Some of us may differ because the U.S. laws may not apply to those geopolitical situations.
32 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
James Keyte:
Then they won’t put their ventures there.
Lawrence White:
All right. I think we have time for one more question for our panel. Any takers? [No response] In that case, please join me in thanking our panelists for a stimulating conversation.
Panel 2: Telecom Mergers
Katja Seim:
Welcome back to this next panel on telecom mergers. I want to start by introducing our panelists. We have, starting on the left and coming down the table: Dennis Carlton from Booth School at the University of Chicago; John Harkrider, who’s a Partner at Axinn; Cristina Caffarra from CRA; Tim Brennan, who is at the University of Maryland, Baltimore County (UMBC); Gail Levine, from the FTC; and, last but not least, Greg Rosston from Stanford. My name is Katja Seim. I am at the University of Pennsylvania.
Now, with that in hand, let me briefly give you a sense of where we will go with this panel. In preparation for the conference, I think we in talking about telecoms and what’s interesting about telecoms from an antitrust perspective touched on a few issues that the panel will hopefully get to. We’ll see.
If you think about telecoms in the last ten years, that’s mostly I think on the broadband side we’ve seen an explosion of video and other data-intensive content that consumers increasingly eat up. We’ve seen a lot of vertical integration, going back for example to the Comcast/NBC merger, but also maybe players like Google and Facebook increasingly investing into infrastructure. And then, lastly, I think we’ve seen an increasing role for wireless data as a broadband source and the types of consolidation issues that might bring about and how technology interacts with that.
With that in mind, we wanted to talk both about vertical issues in antitrust enforcement of mergers but then also horizontal issues as well as maybe more conglomerate mergers where you see wireless providers merging with wire-line companies. We’re going to start with the vertical side and then move on to horizontal issues afterwards.
The big vertical merger, of course, in recent years was AT&T/Time Warner, so I actually want to start with Dennis down there to give us an overview of what he thinks the lessons are that we might have learned from that merger.
Dennis Carlton:
Okay. Thank you.
As Katja said, there have been lots of changes in telecommunication in the last decade — on your cellphone you can make calls, you can play games, you can watch TV — so that just suggests that with all these economies of scope that you’re going to be seeing not just horizontal mergers but, as in the AT&T/Time Warner case, lots of vertical integration, vertical mergers. That’s really why I wanted to start out by talking about a particular case, the AT&T/Time Warner case. 33 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
I was the main economic expert for AT&T. That case was the first vertical merger case that was litigated in over forty-odd years.
The essence of that case is that AT&T owns DirecTV, a satellite company, and if you subscribe to that satellite company you can watch programming; you can watch CNN, you can watch TBS, you can watch TNT. Those are examples of content that Time Warner provides. And DirecTV competes for subscribers with lots of other players, but in particular with cable TV providers. So, for example, if you subscribe to Comcast, you can also see CNN, TBS, TNT, as well as other shows.
The government was concerned that as a result of this vertical merger AT&T in its negotiating role when it’s negotiating the price of Time Warner content to Comcast would have an incentive to raise the price to Comcast. If it raises the price to Comcast, Comcast eventually would be forced to raise its price to subscribers. Some subscribers might leave and stop watching anything. That would be bad because then you’ve lost someone who’s watching Time Warner. But other subscribers might go to DirecTV, and then you’ve not only not lost the subscriber of the Time Warner Content but you’ve gained the margin from distribution through DirecTV. That was the government’s theory in a nutshell.
I wanted to just briefly talk about three main lessons that will set the stage for the subsequent discussion. There are three lessons I take from that case. There are a lot of lessons to take. I have an article that you’re welcome to read coming out in Competition Policy International. Let me just focus on three points here.
First, the courts’ decisions emphasize to me that history is going to continue to matter in the adjudication of antitrust cases. What do I mean by that?
There were past vertical events in this industry: there was vertical integration; vertical disintegration; and, in particular, there was one example that was quite close to the vertical in the merger of AT&T/Time Warner; namely, in 2011 Comcast vertically integrated with NBC Universal under conditions pretty similar to what was going to prevail post-merger in the AT&T/Time Warner case, and you could look at what happened. The answer is you didn’t see the harm the government said we would.
History matters. What else does that mean? If you looked at the last ten years, you saw enormous change technologically. Netflix becomes an important force, spending billions on programming; no one ever thought that was going to happen.
So you see these new competitive dynamics from efficiencies of combining data with content creation. That was one of the things that AT&T/Time Warner wanted to do. AT&T had the data as to who was watching; Time Warner had the ability to create and obtain content. They couldn’t reach agreement on how to share that information. Vertical merger solves that problems. So that was the first lesson. I was surprised the government paid very little attention to history mattering, but it was an important part of my testimony and the court’s decision in concluding that the merger wouldn’t harm competition.
Second — and this applies not just to telecommunications — vertical models, which is what the government’s expert used in explaining why there would be a harm, are really complicated and they are fragile. They are commonly used now by the Department of Justice and the Federal Trade Commission in evaluation of lots of vertical mergers.
I don’t have time to go into the details. Let me just explain one aspect. They used what is called Nash- in-Nash models. What’s a Nash-in-Nash model? Just let me simplify it.
34 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
They used what’s called Nash bargaining. What’s Nash bargaining? Suppose John and I are negotiating over something. How are we going to figure out how the price gets determined? Under Nash bargaining, the basic idea is that how favored Dennis is in the transaction will depend upon what would happen if John and I walked away from the table? How bad off would Dennis be? How bad off would John be if we can’t reach agreement? That, therefore, is critical, what happens in the no-agreement point.
For the government, the no-agreement point was blackout. What was curious about that — and I strongly criticized it — is that was precisely what could not occur as a result of what AT&T had agreed to. Therefore, even if you like these models, the government had the wrong no-agreement point. That’s the first point.
Also, which I’ll focus on, as I said, these models are complicated. No one ever showed that this particular model had any predictive power. Moreover, if you look at the many elements this model depended on and you updated a few; like the profit margin, instead of using an old profit margin, you used the current profit margins. Instead of making believe there were no contracts, you took into account contracts. Then what happened? What happened is the tiny harm the government showed flipped to a positive benefit. So these models may not be very robust.
Third — my last point and then I’ll stop here – I think this is more important for the lawyers rather than the economists, but it’s a very important economic point. That is that in merger cases efficiencies matter. There was no question in the adjudication of this case that the government recognized that there were inherent efficiencies that flowed from their analysis, and that is going to be true in all of the typical vertical models that the government uses. There’s an inherent efficiency from getting rid of what’s called double marginalization, but there are other efficiencies. I already talked about the example based on Netflix. Both the district court and the court of appeals recognized these efficiencies.
I’m no lawyer, so I’ll defer to the legal scholars in the room. I think it comforts economists that court decisions, to the extent there was any doubt, will hopefully continue — and these opinions will hopefully confirm that direction — and the courts will take into account merger-specific efficiencies.
I’ll stop there and turn it over to the rest of the panel.
Katja Seim:
Thanks, Dennis, for this introduction.
Could I get some reactions from the panelists on these lessons that Dennis proposed, and maybe, if these are really the lessons, what we should do with them going forward, maybe in regard to the bargaining models that Dennis described or how we might measure efficiencies that are merger-specific?
Gregory Rosston:
I agree with a lot of what Dennis said. I also was in the background working on the merger so I have some similar opinions.
In the past, the FTC published its report and in an appendix they talked about their models for why they thought there was going to be harm from the vertical integration in the Comcast/NBC merger transaction and they put forth this model that looked at history — Dennis calls it “history matters”; I think evidence matters – and basically they said under this model you can see what happened in other vertical integration and disintegration events. DirecTV and Fox had been together; Time Warner and Time
35 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
Warner Cable had been together — and they put these into a regression model and found under a specific formulation of that econometric model that they would predict harm from the Comcast/NBC merger.
In Comcast/Time Warner Cable, Mike Topper of Cornerstone Research and I submitted something that looked at what happened since the Comcast/NBC merger and concluded that there had been no effect. The conclusion said: There’s no effect of this merger on prices because either there was competition and this was not a problem or because of regulation. We couldn’t discern because they both happened at the same time.
I think that’s true also with the AT&T/Time Warner merger, that both market forces and regulation were put in at the same time. AT&T made commitments that were similar to the conditions imposed on Comcast for NBC. I think that that is important to think about in this context: What are you testing for? Is it just that the vertical merger doesn’t have a problem or whether it’s both the antitrust side and the regulation side that might work together on this?
So that’s the first thing, that you do need to look at what has happened, if you can, if you can see it. One of the things in both the analysis that Dennis did and that we did was it was using the FTC’s model; it was not trying to come up with a new model. It was like, “Okay, you guys set the framework and you guys concluded there was harm. Let’s update it to see what has happened since then.” That was more of a legal strategy than an economic strategy. It was like “You can’t criticize us for this because this is what the rules were set out to be.”
The second is: Yes, I definitely agree with Dennis that these vertical mergers depend on inputs. I wouldn’t call it fragile because that I think has the pejorative sense that they’re not very good. I think you want to have the right input. This is again evidence. What’s the best input you can have?
This will not happen just with vertical merger analysis but also with horizontal merger simulations as well. These are also in that sense fragile in that they depend on what inputs you use. I think that’s important as well.
And then, finally, the efficiencies are also important to think about, that there are efficiencies. I have been a student of communications for a long time. Elimination of double marginalization in programming and distribution is that programmers charge a fee per subscriber, so that if you subscribe to Comcast or DirecTV or some of the new over-the-Internet video providers, your provider pays ESPN a large amount of money — I’ve seen reports of like six bucks a month from the public — whether you watch any ESPN or not; so whether you liked the previous panel or not or care about sports, your provider pays six bucks a month.
But the marginal cost of you getting that is zero. When they would internalize this by having a distributor also realize that, they might realize that the marginal cost is zero, not $6.00. They haven’t come up with ways yet of pricing to eliminate this double marginalization via contract. They haven’t done that, and that’s a question of are there alternatives that could get rid of that double marginalization problem without a vertical transaction.
I’ll let Tim go.
Tim Brennan:
Thanks, everybody.
36 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
I did not work for anybody on this matter. However, I kind of agree with Dennis, but I’ll just make a couple of hopefully brief points at the pretty high level, somewhat higher level than I think Greg was talking about.
On the Nash bargaining model stuff, people who follow this may not realize that when the Department of Justice appealed this decision there was one brief that was filed by, I think, thirty-seven economists on one side of it and another brief filed by twenty-seven economists on the other side of it. That made sixty-four. Probably some of those economists are in the room. I’m not going to get involved in the thirty-seven versus twenty-seven fight particularly.
But in terms of what this fight was about may reveal in some sense how little I really know in my gut about how Nash bargaining models work or how a lot of economic models work. Essentially, some of the story is not the way Dennis put it, but the way some of the arguments against the application of that model were essentially that: Because they’re going to cut a deal, you can’t use as a threat point no deal.
There’s a general thing in a lot of economic models about how out-of-equilibrium beliefs influence equilibrium outcomes. If one is going to make the argument that way, then the baby may be going out with the bathwater. We want to be careful about that. If there was a specific promise that AT&T made, that’s a somewhat different matter.
The other thing on how these models work, assuming one gets around the point that I just made or over the hurdle that somehow is, what’s the but-for thing here? In the FCC’s analysis of Comcast/NBCU, and I think the analysis Justice put in on this case, the but-for-the-merger thing was no problem. I don’t think that that’s right for a lot of reasons.
First, in this particular case HBO’s contracts, I believe, are per-subscriber contracts by and large, or at least a substantial component. So let’s say HBO negotiates a deal without vertical integration first with AT&T. They then are negotiating with Comcast. If that falls through, assuming that’s the relevant standard, then there are going to be more subscribers to AT&T, which means more revenue to HBO at that point. So the but-for case is not some sort of pristine bargaining setting, but HBO is already going to have an incentive, just because of the nature of standard contracting in this sector, to have more favorable bargaining terms with the second multi-view channel video provider it negotiates with given that it has negotiated with the first one already.
That just brings up an idea about: Well, why is it that in some sense the but-for thing is this ideal kind of outcome with no bias in it or nothing that we should probably be concerned about? What is the marginal effect of the merger? And, in particular, what is it about the merger that makes matters worse? Why couldn’t HBO, if it being the first contractor creates this great bargaining advantage, why couldn’t that be part of the terms for that first contract?
Now, there may be a reason for that, but that’s why I think the transaction costs specific to that question should be central in these vertical cases. If you want to hear more about that, you can come to the Bates White conference on June the 10th and I’ll talk about that more then.
Katja Seim:
A little advertisement here.
Do any of our other panelists want to add anything? Otherwise I wanted to touch a little bit on just general vertical issues.
37 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
Dennis Carlton:
I have a few quick responses.
First, Tim, there is frequently a most-favored-nations clause in these, and when you have a most- favored-nations clause that really complicates what in the world you mean by bilateral Nash bargaining because it affects everything else. That’s a problem with Nash bargaining applied here.
Second, your point about transaction costs, if I could just state it differently because I think it’s a very good point, and I made it in several articles and I made it at trial. Tim’s point was if you think the government’s position is correct that AT&T would benefit if Time Warner raised price to Comcast, they could do that by contract, and we don’t see that.
Tim Brennan:
Yes, right.
Dennis Carlton:
Some people may say it might violate the antitrust law. That’s a separate issue. But you don’t see those types of contracts.
The third thing is I wanted to just respond to something Greg said, which I agree with. There was a claim — and it came up on my cross — that the FCC had done a study of a prior vertical event. DOJ didn’t; the FCC did. It was an unpublished study and they didn’t reveal what they had done. We tried to reproduce it. If you try to reproduce it with current Kagan data, you don’t find any effect. So we’re in agreement there was no effect on those prior transactions.
Katja Seim:
I think one thing that has come out of the panel discussion so far is that what’s specific to telecom maybe is very high fixed infrastructure cost, low marginal cost, and that vertical integration might extend powers to these companies to use contracting more generally.
I want to ask Gail how she thinks the FTC might be able to more preemptively monitor behaviors such as the ones we’re concerned about here, what the FTC is doing in that space around sort of big data in terms of industries more generally, and how that might fit in with these types of merger evaluations that we are seeing.
Gail Levine:
I’m happy to talk on particularly the second half of that question, your point about big data. When people ask the Federal Trade Commission about big data, often what’s behind that are questions of the high- tech space, and we often get questions about what the Federal Trade Commission is doing in the high- tech space.
One recent development that is worth noting at a conference like this is the creation of the Federal Trade Commission’s new Technology Task Force. The Chairman launched it in February. It has launched in earnest now. We have some fifteen staffers passionate and active and engaged in their work on the Technology Task Force.
38 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
The mission of the Task Force is to make sure the agency has the intellectual knowledge and the factual basis for moving fast in the high-tech space. Technology industries are disruptive. The technology industry is innovative. The technology industry moves fast. If you are going to stay at pace with that you are going to have to have a strong factual foundation in what’s going on in that industry already. Hence, the creation of the Task Force so that we can be able to appreciate exactly what’s going on in the industry on big data and other areas like that.
Katja Seim:
Great. I think to me data and technology change is at the heart of the second type of merger that I wanted to touch on a little bit, which are more horizontal in nature. Sprint/T-Mobile is one example that maybe comes to mind here, but other mergers that aim at integrating wireless and wireline providers as well, I think, share this feature of focusing on technology.
I want to turn to Cristina and ask her to give us a little bit of an overview of how European regulators and analysts think about mergers there and what types of issues come to your mind.
Cristina Caffarra:
Okay. Let me see if I can get my slide up. I have one slide, but it is worth it.
[Slide – map of Europe] Let me start by saying that’s where we are. We are transported geographically. It will get more interesting than this.
Let me just start by saying that it is particularly timely when I understand here you are looking pretty much at cases like T-Mobile/Sprint and discussing what undertakings might be acceptable to let the deal go through. I’m not involved, so I’m looking at it as an outsider to mention that the European experience is very apt and very extensive on these things.
Just to give you a flavor for this — and things will start appearing magically — I will talk about a number of things.
This is just to give you a sense of the deals which have taken place horizontally, mobile-and-mobile. There has been enormous consolidation in Europe, mostly four-to-three, in wireless operators.
This has not been by any means the only type of consolidation we have seen. We have seen also cable- and-cable, a number of cable deals.
We have seen what we call hybrid deals as well, and many, many of those: many cable companies and wireless companies together at the time in which you are essentially seeing convergence and expansion of the offers of triple play, quadruple play. That has been a major driver.
There has been also consolidation which involved content in macro. As you can see, for the continent there have been a huge number of deals, mostly evaluated by the European Commission.
The reality of Europe is of course that the markets are national, and so the questions that we will touch upon later about the economies of scale, the size of markets, and investments are very central to this.
What is I think interesting is that in evaluating deals — and I will focus particularly on the horizontal piece — really what drives these deals is the kind of narrative that a few people have touched upon, which is they are essential to the telecom industry. There is a sense that voice and messaging are 39 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
completely saturated, the waves of broadband are completely saturated as well. As a business, everything of course has moved over time towards data, that’s where growth is; you need a great deal of dense fiber in the ground to be able to support the data-intensive kinds of services that consumers want.
All of this is moving forward now to further services. There are adjacent spaces that telecom companies feel they ought to be increasingly expanding into — not just data, but there are all sorts of other financial areas, and even health — in which the use and the exploitation of the data that is available is seen as a further potential for business, because in fact in the core area there is no money to be made, very little money; the margins are small and declining. So this is underlying this kind of movement towards consolidation and bringing assets and rationalizing assets.
How has it been seen by European agencies? This broader narrative about the need to consolidate in order to motivate investment has failed to act completely. We will go back to that later. Just to say that to the extent that operators have been using this kind of motivation they need to merge.
Four networks are far too many. Particularly in small countries where the geography is not supportive of four networks, we need to consolidate to motivate the investment. This has not really worked. I’ll explain later why and what underpins this kind of view on the part of regulators. Fundamentally, the analysis of these deals has focused on short-term price effects with fairly standard types of models, GUPPI models, sort of baby merger simulation, looking at very simply what the price effects are going to be. Of course you always find that there is a divergence between the parties and you always find that there is going to be a price effect that is going to grow anywhere between 8 percent and 10 percent.
The remedies in these cases have really been gradually sort of developing fundamentally. Initially, the first of these cases was the one in Austria that you can see there, Hutchison’s H3G Austria and Orange Austria. Orange 3G was basically just an access remedy. The Commission basically imposed on the parties an access obligation whereby mobile virtual network operators (MVNOs) were given access, a pay-as-you-go sort of tariff, to the network.
Then the view developed over time that these kind of access remedies were actually not very effective. This is very germane to the discussion I think you are going to have in T-Mobile/Sprint.
So the Commission and the view in Europe has very much shifted towards the position that a sort of an access remedy with a variable price per unit is effectively not going to allow for competition to develop. What you need to do as the merged entity is effectively slice off a portion of your network and commit it to a new entrant, almost a divestment of the network except you don’t really divest it, but you give 30 percent to them at a fixed up-front price, so the incentives going forward are going to be zero marginal cost and you are going to have the incentive to compete, plus some spectrum kind of given on top.
As things have gone forward, these remedies have developed more towards almost a quasi-structural type of remedy.
In the Italian case, which is the most recent one, essentially you have gone from four to three back to four, because the Commissions has mandated that a significant portion of assets were divested and given to a new entrant, Iliad, which has actually come into the market and whipped the backside of everybody.
So I think that the lesson — and I’ll conclude on this — of these mergers and remedies is that in Europe the view is that pure access remedies have failed to work because the remedies taken in the main for these deals were cable companies were looking to get access for their “triple play” offer and so they weren’t replacing the competition in wireless that was being eliminated. The Italian case is an exception 40 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
because, of course, it was taken up by a serious mobile player that is killing everybody in the Italian market by competing very hard on price.
This is a question mark: Do we want that type of competition? The telecoms are clearly complaining that that is racing to the bottom and it misses completely the big dynamic that is going on and the investment needs of these companies. So if you accept a remedy that in the end replaces competition but effectively races in terms of price competition to the bottom, everyone is fairly screaming.
I’ll stop here and then we can go to the investment piece.
Katja Seim:
Thanks, Cristina.
Before we turn to the investment piece, I wanted to get some reactions from our panel to some of the issues raised.
I think the first one that Cristina touched on, which you can see on the map here also, and may be useful in thinking through the Sprint/T-Mobile merger, is just that we tend to see consolidation by weaker players trying to maybe catch up with the larger players due to economies of scale. I want to get people’s reactions. I will turn to you, John, in terms of whether that kind of increase in scale is sufficient to outweigh possible anticompetitive effects and how we might think about that.
John Harkrider:
Sure, happy to.
I want to start actually by reacting to something that Dennis said earlier, which is this idea that you are actually seeing lots of change and lots of disruption within telecom. I actually challenge that a bit. I think that what you do see certainly is you see lots of new content providers — no question — and you also see lots of resellers and a lot of people who are coming in from related markets. But what you don’t see is a lot of new facility-based carriers, and you especially don’t see people building out, sort of de novo, lots of new dense networks.
People who have worked in telecom for a long time know that there are two really big problems in telecom generally: one being the problem of the last-mile conductivity, being able to get into the house; and then the other just sort of dense, ubiquitous networks in underserved communities and who has the incentive to build those out. I think there are significant public policy questions in both of those.
In response to your actual question, I’m a careful person, and so I think it’s very hard to make general comments about whether a certain argument is likely to work or not work in a given context. I think you need to have not only a specific case in mind — and obviously we do, at least with respect to T- Mobile/Sprint — but you also need to have access to the documents, you need to have access to the businesspeople, and access to the data. I don’t have access to any of that stuff so I want to be super- careful in what I say.
Generally, we think that competition is actually the best way to incentivize investment as opposed to consolidation. I think antitrust is generally built upon that principle. I think that it could be possible. Let me just talk about two issues that I think are germane to that merger. One is the ability and incentive to invest in 5G networks. The other one is the ability and incentive to continue to work with other MVNOs, especially disruptive ones. 41 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
I think on both of those they are going to have challenges making those arguments. I think, first, you have to look at why firms are mavericks in the first place. Is there something structurally going on that causes them to adopt a certain business model, and does a merger actually make them look a lot more like the incumbent and, therefore, less likely to engage in maverick behavior?
The second thing is that you need to look at what the premerger plans of the companies are. Is it actually realistic for us to think that either of those companies was not going to invest in 5G? That’s like saying, “I’m not going to invest in air today.” The idea that these companies would be able to survive without 5G is a really difficult thing to argue.
And then you come down to an issue, which is: Are there economic arguments, is there economic data, and are there business documents to actually support the idea that but for the merger they would not make investments and as a result of the merger they actually would make those investments?
I want to say that I don’t think any argument is off the table, and I believe very strongly in that. I think having priors is something that we all have, but I think we should fight against it and look at each case as sort of a de novo presentation of facts and evidence. But I’m skeptical that those claims can succeed.
Katja Seim:
Gail?
Gail Levine:
I want to see if you’ll give us permission to expand a little bit beyond the merger horizon here and talk about a different aspect of the telecom space, not so much the provider space but the chip space.
Katja Seim:
Sure, totally.
Gail Levine:
There has been a timely development here. About ten days ago, the Federal Trade Commission won its lawsuit against Qualcomm in a case that is important for the chip market. That’s shorthand for saying that’s what powers connectivity in all of these telecom markets. The case that the Federal Trade Commission brought and that the district court resolved in the Federal Trade Commission’s favor about ten days ago was aimed at restoring competition in that critical market.
There’s a baseband processor in every smartphone that we’re talking right here, and with the advent of the Internet of Things this is only going to become even more relevant. The case was designed to make sure that the competitive channels are clear in the emergence of these new technologies in the telecom space.
If it’s alright with you, I can give you a quick overview of what we challenged, what the court did, and then what the take-aways in that space are.
The short answer is Qualcomm, according to the Federal Trade Commission’s complaint, engaged in a variety of mutually reinforcing behaviors to improperly maintain its monopoly power in chips. We’re talking here about the “no license, no chips” policy, which I’ll come back to in a minute, incentive payments, exclusive dealings, and a refusal to license. 42 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
A word about the “no license, no chips.” Qualcomm was using the threat, we alleged, of “chip cutoff,” denying access to critical chips, to ensure that device OEMs couldn’t or wouldn’t have practical access to the FRAND determinations that would have allowed them to constrain the royalties on Qualcomm’s standard-essential patents. That meant that Qualcomm could shift price onto its royalties and away from its chips, which would then enjoy a unique and artificial pricing advantage in the market while Qualcomm received its inflated supra-FRAND royalties.
FRAND is an industry acronym for fair, reasonable, and non-discriminatory licensing royalty, and that placed Qualcomm’s chip competitors at an improper advantage, protecting Qualcomm’s chip monopoly.
What did the court do ten days ago? It found for the Federal Trade Commission across the board. It granted a succinct injunction that Qualcomm cease its anticompetitive prices, uncouple chip supply from license negotiations, and compete on the merits with a role to play at the Federal Trade Commission on regular reporting.
With all that, what are the key take-aways that I see in this case?
First, you can break the antitrust laws by unlawfully maintaining even a lawfully acquired monopoly. You can acquire a monopoly position lawfully, do innovative work and great things, but problems arise when you build a wall around it through unlawful conduct. Monopoly maintenance through anticompetitive conduct is just as unlawful and just as harmful to American consumers as monopoly acquisition through anticompetitive conduct.
The second thing — and this goes back to the question we were asking earlier — is that cases like FTC v. Qualcomm show that the agency is willing to bring monopolization cases against big tech companies. We at the FTC are vigorously enforcing the antitrust laws including the law against monopolization in high-tech markets. The Technology Task Force is all about that.
And I just want to give a shout-out here to the extraordinarily talented, effective, and diligent trial team that prosecuted the FTC’s case against Qualcomm. Qualcomm is just the latest example of the agency’s engagement in investigating and, where appropriate, bringing sound monopolization cases in the high- tech space, and we will not be resting on our laurels.
Finally — and this is sort of to state the obvious — the last take-away from the case is there’s no zone of antitrust immunity or special pleading just because contract rights or IP rights are in play. The Qualcomm case was an antitrust case; it was an antitrust case with IP rights and contract rights in play in various ways.
But the court appreciated, and we appreciate, that just because there are rights in other areas of the law — IP, contract, trademark — just because that’s in play doesn’t change the fundamentals of antitrust law. The Qualcomm decision underscores the integrity and the uniformity of our antitrust laws, and I think that’s something that we at the agency, and I think in the bar and in the industry community, should all welcome.
Katja Seim:
Thank you, Gail. Before I turn it over to the audience, just maybe one comment.
Gregory Rosston:
I’d love to respond a little bit to what John said about changes. 43 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
Twenty years ago I had a choice for my home phone service of AT&T. Twenty years ago for my cable television service I had actually a choice between nascent DirecTV satellite and DISH Network and Comcast. And for cellphone service in my house I really had a choice of AT&T and Verizon; the others didn’t reach my house and were not reliable. And then I got broadband service from Comcast and I had slow DSL service.
There has been, I think, a lot of change in twenty years. Even though the names haven’t changed, the services they provide have changed substantially. Now I can get AT&T fiber at my house to get broadband, and I can watch Netflix that I couldn’t watch before that I can watch broadband on my phone, and now I can get T-Mobile in my house or Sprint in my house. So there have been a lot of changes. These networks are much, much faster than they used to be as well. The Comcast network is much faster. The AT&T network is much faster. So while the names haven’t changed, the products have really changed a lot.
John Harkrider:
I don’t think there’s any dispute that we enjoy the benefits of competition.
Cristina Caffarra:
Can I add a point very briefly? To me this is the central point. The discussion here about mergers in telecoms is fundamentally about whether consolidation will be enabling further investment into quality. This is the argument that the industry pushes forward. As John says, it has been very difficult to make that argument persuasive. The prior is really pretty much, as economists and generally, that it is competition that spurs investment.
Just to share a quick episode, when we were doing the cases in Europe, while trying to persuade the Commission that there was a causal link between somehow consolidation and better investments, the CEO of one of these companies went on the record and said in France to a bunch of analysts, “We are being killed at the bottom by competition from one of these kind of cheaper guys. We need to invest much more to move ourselves to the top of the market to escape competition to the bottom.” It made the immediate, obvious, clear link between competition and investment.
I think the economic literature is also not remotely clear that this link exists between consolidation, more concentrated markets, and investment.
Now we are moving into a world of the cloud, 5G, Internet of Things. Investments are large and the industry is talking even more about the need to consolidate to motivate that investment. But I think that kind of narrative and that rhetoric isn’t going to play very well yet.
Dennis Carlton:
My view is that our economic models of mergers, vertical and horizontal, are focused primarily on price and have very little to do with making predictions about investment. When you’re talking about an industry that is rapidly changing and you say it’s investment that matters, that just means that you should understand that a lot of what our models are talking about may not be particularly on point to the main question. That I think is important.
The second thing I want to say is when you are looking an industry that is sort of regulated but not always in all its parts, and it is an industry with rapid technological change, that is really hard to regulate
44 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
and not really create hard problems. On the other hand, if you do nothing, it is also an industry that could get monopolized and create harm that way.
So it’s a really hard problem. It makes me worry when I hear an argument that says, “Let me extinguish competition in order that I can get market power in order that I can invest.” That seems to tip the antitrust laws upside down.
Now, maybe that means antitrust is ill-suited to this particular industry and they need to somehow have regulation. But when you have regulation in a rapidly changing industry you’ve got to understand the limitations of the regulators. The real trick is — and it’s a hard one — I want to have enough regulation that I sponsor enough competition that I’m not extinguishing it; but, at the same time, I understand that there could be public policy objectives that possibly could be inconsistent with competition. But I’m always worried that you can’t use one argument to encourage mergers and at the same time say, “They’re going to be great for society.”
Gregory Rosston:
One thing in the AT&T/T-Mobile case that AT&T said was, “If you don’t allow this merger to go through, we’ll only build out to 80 percent of the country” — this is not necessarily innovation but investment — and they said, “With the merger we’ll go to 98 percent of the country.” This was completely non-credible to me — I was working at the FCC at the time — and it was like, “No you’re not; you’re going to go to 98 percent regardless.” Nobody believed it. The merger was not approved and they ended up covering more than 98 percent of the country. It’s sort of: How do you think about competition influencing it? I agree. Competition forced them to go to 98 percent. We knew they weren’t going to stop at 80 percent because of that.
John Harkrider:
I know I’m on a panel with a lot of economists, but I think it is important sometimes just to take a step back to say — one thing lawyers are good at, I hope, is actual fact discovery. We can have all the models in the world and they all have imperfections because they are imperfectly modeling reality.
I think on both sides the discovery of what the facts are, what the intentions are, is a really important thing, sort of “but-for the deal what would we have decided to do?” I think access to and actually looking very hard at company documents and taking really hard depositions can sometimes be as effective — or hopefully maybe more effective — than abstract models or looking at natural experiments, what actually happened, which I’m a huge fan of.
Tim Brennan:
I just want to say quickly that this discussion brings up a methodological point that also Andrew brought up in connection with what to do about student athletes, which is I think in context going beyond this point that antitrust is not very good at cost/benefit analysis. On the one hand, you’ve got claims that “Maybe this merger increases the likelihood or size of 5G investment.” On the other hand, you’ve got “Maybe this merger will have anticompetitive outcomes.” The way those arguments typically get resolved is by denying one side or the other, that “this merger will have no effect on investment” or “this merger will have no effect on competition.” If it has both, it’s a problem.
I don’t envy the staff economists at Justice who I think are probably working very hard trying to figure this out, or at the FCC similarly. But there’s a lot of context where in principle “Gee, there’s efficiency
45 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
benefits; gee, there’s potential competitive harms,” and the way these things typically get resolved is by denying the existence of one side or the other.
I don’t have a good answer for that, but it’s a problem, and it’s a problem that also affects these NCAA cases, and undoubtedly a whole bunch of others.
John Harkrider:
I think, frankly, the method by which they’re denied is by using different standards, which I’m totally against.
Merger specificity on efficiencies and saying, “Well, gee, over the last twenty years we’ve actually never seen somebody try to do these sorts of contractual solutions to a problem that everybody thinks exists, but yet we simply hypothesize that if the merger doesn’t go forward, they’ll be able to solve that problem,” I think that’s a ridiculous argument.
I think that we should have similar standards for the costs and the benefits and walk into that morass, and maybe the result is you can’t do enforcement at that point or you need better tools. But I am very against what I think actually does happen, which is that one side of the equation, usually the benefits, is held to a vastly different and higher standard than the theory of anticompetitive harm, which can frequently be relatively speculative and not supported by anything.
Katja Seim:
All right. Let me turn it over to the audience. Are there any questions for our panelists? If you could speak your name and affiliation.
Question [Cecile Kohrs]:
Hi. I’m Cecile Kohrs with TIG Advisors. I wanted to ask Dennis about when you’re talking about the industries that have concerns and there are rapidly changing things, how do you deal with potential competition concerns in pharma as an economist? How can you model the likelihood of success for a drug that’s in a pipeline and how much thought has been given to that? And I’d open this to the other economists as well.
Dennis Carlton:
Well, you actually asked me an easier question than you could have. I think potential competition is really hard to evaluate in a changing industry. The Department of Justice tried to do it in a case in the 1990s when they blocked the ZF/GM transaction. I was involved in that for GM. The argument was there was an innovation market and if the transaction went through there would be fewer innovations and that there would be innovations from the separate parties.
What I did is I followed what happened after the transaction didn’t go through because of that reason, and I would call GM and I’d ask, “Well, what are the innovations? Did what the government say happen?” “Nope, nope, nope, nope.” After ten years I stopped asking. So if you go and look at who is innovating, can you predict who is going to innovate that product five years retrospectively? That’s really hard to do.
Now, in pharma I think you have a little bit easier time because there is a pipeline: you have to apply to the FDA; you can see how far along they are in trials. So I actually think pharma is one of those cases 46 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
where it’s a little easier. But in other cases where it’s a new technology that is being developed and a lot of people are involved in it, I think it’s very hard to predict. I don’t have great hopes that we are going to have very refined ability to make accurate predictions. So I think it’s dangerous.
Now that’s a little different than the argument that is being raised in some circles that when you see mergers of firms that look like they are not doing the same thing — well, maybe the databases are what you should be looking at, and maybe you should look at it as a merger of databases rather than a merger of what the databases are used for. And, even if the databases are used for very different things, that doesn’t necessarily mean there might not be a horizontal overlap. That’s a different spin on the question that might be worthy of investigating.
But I think it’s very hard to make these arguments about potential competition. It’s very easy to say, “Gee, in retrospect you shouldn’t have allowed that merger because that guy’s doing great.” Well, maybe that guy’s doing great because of the merger.
Katja Seim:
Another question?
Question:
I don’t know if anybody actually takes this seriously, but if you look at government policy, a merger predicted to raise price 6 percent for thirteen months and then be completely wiped out by competitive forces thereafter would be barred under current American antitrust policy because of the time lag when competition will finally take over.
I appreciate the brilliant theoretical insights of Professors Carlton and Brennan, but try telling somebody in Los Angeles who wants to watch the Dodgers that (a) competition will shift all the Dodger fans to the appropriate multichannel video programming distributor and (b) that a no-deal is simply not in the cards.
Given Commissioner Phillips’s talk about conduct versus structure, does the panel think that there are areas where, even if you believe on the structural level we ought to permit mergers or other structural agreements, regulators should come in to deal with the specific cases where blackouts are happening and consumers are harmed even if theory suggests that eventually the market will take care of it?
Dennis Carlton: I’m not quite sure how to respond.
The first part of your question had to do with a short-run loss versus a long-term gain. From an economic point of view, an economist would say you would have to evaluate all of those. Now, I understand maybe under the Merger Guidelines or prosecutorial discretion that might or might not occur, but from an economic point of view the answer is pretty simple: you want to look at the overall harms and benefits. I agree with you that you don’t want to have tunnel vision and not pay attention.
I think it’s a different question, the second part of your question, which says, “There’s going to be a blackout and, dammit, I want to watch the NBA playoffs” — I wanted to watch it last night but I was unfortunately on an airplane — and I would have been annoyed if they said, “Sorry, Dennis, you can’t watch the fourth quarter because we just had a dispute.” So yes, I’m upset.
Are you asking me whether I think that there should be a government body that prevents Dennis from getting upset? Maybe. [Laughter] 47 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
Questioner:
I would be a part of that.
Dennis Carlton:
But I just think you’ve got to worry a little bit that if you set up such bodies, they’re going to make lots of decisions and you might not like a lot of them. The question really is — I understand we’re not in a perfect world — do you think we are going to get closer to having a competitive world trying to perform in competition by having one that’s highly regulated?
My own view is that — I once wrote a paper with Randy Picker on this7 — look at a contest between antitrust and regulation over the last 150 years. Railroads are a good example; we’ve had both. Which does better?
That doesn’t mean you should never have regulation or not have regulation, but it does mean you have to understand that both are imperfect. Anyone who thinks competition always gives the right answer; he’s wrong. Anyone who thinks regulation always gives the right answer; he’s wrong. So I think you have to use your judgment as to where you want to put your bets.
Tim Brennan:
Two quick points. One is that if antitrust economics has evolved to the point where we can predict that a merger will raise price exactly 6 percent for exactly thirteen months, I would be astonished at the sophistication of the empirical work.
The other thing is I think that question should have been directed to the prior panel because ultimately the market power involved in watching Dodger baseball games rests with the Dodgers.
Gregory Rosston:
And, as I’ve been trying to explain to Steve for thirty-five years, it’s a public interest benefit to blackout the Dodgers. [Laughter]
Katja Seim:
Any other questions for our panel?
Question [Linda Baroli]:
Linda Baroli with Oscar Gruss. My question is for Ms. Caffarra. Given the rumors we’ve seen about Amazon being a potential buyer, what they’re saying is it’s going to be, I guess, called like a virtual sliced-off network.
Cristina Caffarra:
Yes.
7 Dennis W. Carlton & Randal C. Picker, Antitrust and Regulation, University of Chicago Law & Economics, Olin Working Paper No. 312 (October 2006). 48 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
Questioner [Linda Baroli]:
Just from what you’ve seen in Europe, do you think someone coming completely outside the industry in to buy some of the divested assets is a workable solution?
Cristina Caffarra:
I don’t know enough about what the operation being potentially divested really does. I can only say that certainly the learning in Europe has been that when you have had this type of access remedies, the takers have tended to be players who are trying to complement their offer. So they were not themselves hard players in the wireless space, they were cable companies wanting to do a full play, and therefore they sort of took it and they often bundled, but it wasn’t really effective competition that was replacing what was being lost.
To the extent that this potential solution with Amazon effectively provides Amazon — I only know what I’ve read in the papers — with the possibility to offer some sort of additional service to its customers, that’s fine. But how effective is that competition going to be I think is the question, right?
I hear John when he says we haven’t seen that much change in terms of the growth of other infrastructure-based competitors. This may well be clearly a reality in the United States. In Europe the problem we face is probably a surplus of infrastructure that we have that somehow needs to be rationalized. But the reality is that unless you have infrastructure, the economics don’t allow you to compete effectively against an infrastructure player.
As a final point, what we have tried to do in Europe is we are concatenating actually an infrastructure economics through a remedy by giving the remedy taker effectively the economics of a player. So they get access to the capacity at a fixed cost so they can compete in principle, but it doesn’t really work very well.
Katja Seim:
Another question?
Question [Harry First]:
Harry First from NYU, but down the street at the Law School. This, I guess, since I’m from the Law School, reflects the question which is directed at the economists, particularly maybe Dennis.
My question is this. I wonder whether antitrust has become way too much over-dependent on economists. I ask this from the point of view of the AT&T/Time Warner case and how Judge Leon began his opinion, which I’ll read to you: “If there ever were an antitrust case in which the parties had a dramatically different assessment of the current state of the relevant market and a fundamentally different vision of its future development, this is the one.”
So I wonder whether your wonderful work and Carl Shapiro’s wonderful work really had all that much to do with Judge Leon’s decision, and whether the Justice Department would have been better served, or at least as well served, by having a media expert come in and explain to the judge where this industry was going — not from the point of view of economic theory, but how the industry was developing. I think he just saw this as: “They’re under siege by Netflix and everybody else. Come on! Cords are being cut? Tell me where the harm is. You have a different view of the world.”
49 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
So I wonder whether from the point of view of antitrust policy enforcers are telling the story persuasively enough just by trying to rely on the models — which frankly, Dennis, you demolished; that was your effort. Are they useful enough?” Do we need different ways of approaching how to prove what is going to happen in the future with these mergers?
Dennis Carlton:
I think I can respond in a few ways. First, I thought Judge Leon’s opinion, aside from being extremely complimentary to me, was extremely thorough. He went through each one of the economic arguments and evaluated them in light of the evidence that was provided. From my point of view, it was a homerun for economics, confirmed by the industry witnesses. So on that score I don’t think there was any problem at all.
I think it would be a mistake if antitrust went away from listening to economists. Economists do have something useful to say — maybe not as useful as they would like to think all the time, but they really have improved antitrust evolution. If we had to go back to the 1970s and the 1960s and you see things like Vons Grocery, I think I’d shake my head and say, “My God, we’re going back to the Stone Age.” There’s no question that economics has had an enormous effect.
On the use of very complicated models, which I criticized in my opening remarks, I agree with you they have great dangers. But that’s why linking it to the evidence — that’s probably a better word than what I called “history”; Greg called it “evidence,” and I agree with him — those natural experiments are critical for an economist to understand the economic forces.
Judge Leon, to his credit, focused on those. He also focused on the deficiencies of the very complicated — I think as he put it “Rube Goldberg model” — that Shapiro used. I have high regard for Carl, he’s a friend of mine, but I did think that my criticism of his model destroyed what he did and the credibility of what he did. The Judge agreed with that. I think that’s based upon economics.
So my shorter, one-sentence answer to you is that I think it would be a mistake to go away from economics. Economics has a great deal to add to prevent antitrust from going off the rails. If you just have media experts who have no training in economics, that will be a throwback to the Fire Age and I don’t want to rely on people’s gut opinions as to “Is this going to be good or is this going to be bad?”
I do a lot of work on merger activity. You always go into that room and the two CEOs say, “This is going to be great for consumers. How could anybody think anything different? We’re going to get such efficiencies.” I say, “Can you prove that stuff? You know, I’m going to go to the Department of Justice. What is the basis for this?”
So if you’re going to just rely on the CEOs, you’ve got to be really careful, and I think it would be a big mistake to take economics out of the picture.
Katja Seim:
We are basically out of time, unfortunately, but before we break for lunch, I just want to give this side of the panel a quick opportunity to respond.
Gail Levine:
I want to say there’s got to be an understanding of the role that economics plays in cases like this that is somewhere between the homerun, the grand-slam approach, and the strike-out approach. Look, the right 50 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
way when you’re evaluating whether to bring a case, merger or non-merger, is you’ve got to think about your economic theory, your modeling; you’ve got to think about what your witnesses are going to say; you’ve got to think about what your doctrines are going to say. If you have a case that can be built on all three legs of that stool, you are in a much stronger place. Take economics out or rely overmuch on them and you’re in trouble.
Katja Seim:
All right. I think that’s a good way to close.
Panel 3: Pricing Issues in Pharma: Pay-for-Delay, Product Hopping ...
Robert Willig:
Ladies and gentlemen, this is the best of all worlds. This is the panel about the most excitingly litigious antitrust industry in the United States, and I understand that it is coming to be that way in Europe as well. This is not sports or telecom, sort of old stories, but this is pharma, where all the action is.
We have a great panel of wildly distinguished participants in the competition issues in this sector, both in the United States and in Europe, and you can see their names and bios on the board. Not only are they distinguished and involved, but they also are very articulate, and so an hour seems like a very short time, and I shouldn’t take up too much of it myself.
I tried to organize this panel slightly differently than the others this morning. Instead of giving everybody an equal shot at every subject, we have four different lead subjects. Product hopping, reverse payments, excessive pricing, and specialized legislation are our four principal topics. Individuals on the panel have more or less volunteered (or been volunteered) to play the number 1 role, the number 2 role, the number 3 role on each of those subjects, but they’re all entitled to jump in on top of each other — with sufficient politeness I would add, but they seem like a polite crew in addition to their distinction.
So let’s get going on that basis. I posed the questions to the panel before they volunteered to take on different roles, and I thought I would share with you abbreviated versions as we get to the questions.
Number 1 is product hopping. I love the colorfulness of the title. The question I posed is: Given a situation where there’s a patent holder who shifts its marketing to a newly patented variation on the originally patented product and persuades the doctors — and this is in the United States; maybe it’s true in Europe, too, but certainly in the United States — to make the shift just because of the application of the persuasion of the pharmaceutical company — maybe there’s research backing it; maybe there’s not — but the idea is to get the docs to switch before the old patents expire and before the generics come flooding in to the old category.
In this situation, the question posed is: What additional elements should there be that would constitute an antitrust violation? There are a whole bunch of cases claiming antitrust violation. What should those additional elements be? Our first speaker to that subject is George Rozanski.
George, the floor is yours. Under five minutes, please.
51 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
George Rozanski:
Thank you, Bobby.
At the end of the last episode, our heroes, the economists, were finding themselves in deep yogurt for relying on theory to the exclusion of the facts. I don’t want to make the same mistake here, so let me start by presenting a few facts and then trying to frame the issues for the further discussion.
Just to remind everybody what’s at stake here, both as we are thinking about product hopping and as we think about reverse payments, generally there are substantial savings that are available from generics. When the first generic enters a market, it is typically priced at a 10–15 percent discount below the brand. It has a six-month period of exclusivity and then very quickly, certainly in the larger markets, we tend to see a number of generics follow that first entry.
After we see four or six generic competitors in the market, we observe that they are quickly capturing the majority share of sales, selling at a deep discount, and the cost to consumers plummets. Across all markets, typically a year after that first generic entry, generics account for 90 percent of volume; they sell at an 85 percent discount; and the cost to consumers has fallen by between 75–80 percent of what it would otherwise be. These are numbers from the FTC’s 2010 Pay-for-Delay Staff Study.8
Across all products this can amount to a lot of money. The Government Accountability Office (GAO) estimated that generics save consumers on the order of $100 billion a year. They estimated savings in excess of $1 trillion in the first decade of this century.
The concern with product hopping, as Bobby has outlined, is that incremental innovations, which might have little or no inherent value, can have the effect of limiting this competition from generics, and the mechanism to focus on relates to automatic substitution. In the United States pharmacists may — and in most states they must — dispense a generic substitute if the generic is AB-rated, meaning that it is bioequivalent and it is of the same form and strength as the branded product.
The facts are a little bit different in the case of biologics. Biologics are large, complex molecules. They are manufactured in a living system, not made by chemical synthesis. Currently they account for about 40 percent of drug spend, and they’re growing. In the case of biologics, it is often said that the product is the process, that the manufacturer tweaks the process. If two different manufacturers try to produce the same molecule, they are not going to be quite the same. It is very difficult to test and know that patients in fact will respond to like drugs, which are referred to as biosimilars, in the same way.
States are now beginning to require substitution by pharmacists in the case of biosimilars that are also interchangeables. At present in the United States there are nineteen biosimilars approved, but none of them are interchangeable.
It’s perhaps instructive as we think about whether product hopping is a problem that at present there is less threat of competition from generics faced by biologics but we still see significant incremental innovation by the manufacturers of biologics. What do we mean by “incremental innovation?” The literature usually talks about three main types:
8 Pay-For-Delay: How Drug Company Pay-Offs Cost Consumers Billions: A Federal Trade Study (January 2010). 52 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
• The first would be a new formulation, for example, an extended-release tablet; injection rather than intravenous administration. These are things that can lead to better adherence, so consumers are more likely to take their medicine and benefit from it; and better absorption, so the drug is more efficacious.
• It can be a new combination of a drug with another active ingredient, reducing the number of pills that an individual has to take, again leading to better adherence.
• Or it can be a new indication, a drug that is approved to use for a different condition than it was initially approved for, and this can greatly expand the use of a drug.
In some cases these improvements are patentable and, even if they’re not, the firm might get some period of marketing exclusivity. In the case of small-molecule drugs, not the biologics I was talking about, there is a three-year period of exclusivity in the case of new forms or indications, seven years if the indication is to treat a rare disease or an orphan drug, and these are all patentable.
We know that generally innovations can be beneficial. The question is: How are we going to know in the case of the incremental innovations that are the focus of antitrust concern here whether they are beneficial or not? The reason it’s difficult is because we lack the usual market tests that economists would rely on to determine whether there is value in an innovation, and that’s because of what is sometimes referred to as a price disconnect.
In the usual case, if consumers are willing to pay more for a new drug, you can infer that it adds value. In this case, the doctors that are choosing the drug, that are prescribing it, are not the ones who are paying for it; they don’t bear the cost. You don’t know if they are making the same tradeoff that consumers would make or if they are simply responding to the last promotion they heard from a drug manufacturer.
Turning again to some empirics, a study by Ernie Berndt and some of his colleagues at MIT and Harvard over the period 1990–2003 looking at 1200 new drug applications: 34 percent of them were new molecular entities, so new drugs; 66 percent were not, they were examples of these incremental innovations.
They did a closed study of three classes of drugs — antihypertension, anti-ulcer, antidepressants — and for these twenty-nine drugs that they looked at in these three classes in terms of the incremental improvements, they counted 135 approvals for new indications, forty-eight for new dosages, eighteen for a new combination of drugs, fifty-one for new formulations. For two classes in particular, they found that 70 percent of the use of the drugs was for new indications. These findings certainly suggest substantial expansion of sales from these incremental innovations, meaning that there are significant benefits that needed to be accounted for; it’s not just crowding out of generics.
Robert Willig:
Speaking of crowding out, George, I’m sorry time is so limited. Thirty seconds.
George Rozanski:
I was actually going to turn it over to Jack now, who I know has worked on some of the cases, and he can maybe bring some color from these cases.
I think that the questions are:
53 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
What’s a sensible approach? It has been suggested that maybe any innovation is per se legal, but that puts the burden on somebody — maybe the courts, maybe an enforcement agency — to determine whether there is any value to the innovation. That seems like a difficult thing.
You could look at the nature of the conduct. There are some differences in conduct that have been alleged between what is referred to as a “hard switch” or a “soft switch.”
There’s also a test in the economic literature called the “no economic sense” test, which says: Does this conduct make sense but for the effect that it might have to limit competition? Again, I think that’s a difficult test, but it might have some merit.
I guess, stepping away from antitrust entirely, you might say this is not an area where antitrust needs to be involved. Perhaps just changing the regulations, changing the laws to be more careful about when we grant exclusivity when we grant patent extensions, would be another way to address concerns.
Jack Pace:
George makes me feel guilty about not using facts or numbers, so I dug up something —
George Rozanski:
Pure theory; that’s what lawyers do, right?
Jack Pace:
Before I get to pure theory, two other facts or sets of numbers to be a little quantitative to start.
Henry Grabowski a few years ago did some work on this, and some of you have probably read it. He and some others found that out of every 5000 molecules tested, only 250 drugs will ever enter preclinical testing and only one will be approved by the Food and Drug Administration.
Fact number 2: Only three out of ten marketed drugs produce revenues that will ever match or exceed average R&D costs.
I guess the maybe less precise metric I might use is the number of diseases is a lot and the number of cures is a lot less, and so I think when we work on these cases you have to be aware that you are making rules, and when you make rules that might have the effect of making it harder to innovate. You have to be very, very nervous.
So the question on the table, I think, was: What are some of the things you think about? What are some of the types of conduct you might look for in evaluating whether your hypothetical, Bobby, is something that would raise antitrust alarms?
One of the first things I think you have to look at is other choices on the market. I think very often it is a big mistake to separate out the market-power inquiry from the conduct in a product-hopping case. They are very often seen as two separate questions that you do in a certain order, but they really need to be looked at together.
In a product-hopping case, what we are trying to figure out is the level of coercion, and coercion is a funny thing to test in an antitrust case. That comes from Berkey Photo in the Second Circuit that used a coercion standard in a case about new product design changes. That cited to some tie-in cases, and there 54 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
are tie-in cases that have used the concept of coercion for a long time. The Second Circuit in the Namenda case applied a coercion standard: it’s anticompetitive if it’s sufficiently coercive.
Now again, coercion is a funny thing to think about in antitrust because the anticompetitive harm doesn’t come from the typical thing that would be an anticompetitive harm, restriction of supply. Coercion is potentially anticompetitive because it actually increases demand for the product through maybe a limitation of choice or something else. But that’s what you are looking for.
So if you are asking the question — and I think the Namenda case is a really good example here, because in Namenda the discussion of whether the conduct was coercive was completely wrapped in with the product at issue; that was a case in which the defendants were trying to switch from a twice-a-day Alzheimer’s drug to a once-a-day Alzheimer’s drug — the defendants indeed had — and I say this from personal experience, having been in court saying this very thing — thought we had pretty good arguments that, especially in Alzheimer’s treatment, reducing the number of pills that you take per day is a pretty good thing.
Robert Willig:
I don’t remember if that’s true or not. [Laughter]
Jack Pace:
Take Namenda. If something is going to be a material improvement, you would think that reducing pill burden for Alzheimer’s patients might be. But there was something, of course, special about that drug and that treatment area. There was a concern that if everyone got switched to the new version, the once- a-day, and then the generics came out — and the generics, of course, at first could only sell the twice- a-day because that’s what they were gearing up to sell — that that would be cheaper, that’s a good thing, and no one is going to switch back to the generic of the old version.
There was a real concern that that sort of reverse commuting wouldn’t be a real choice because with Alzheimer’s patients there might be real difficulty in changing routines too much, and so there was a real sense that the switch from one to the other kind of locks you in in a way that maybe wouldn’t happen with other drugs.
So I would say that you really do need to look at other choices as one big factor I would look at.
The other one I’ll say — and this also comes from the Namenda case — is you need to look at, they say, so-called “additional conduct.” In other words, launching a new version and withdrawing an old version is not enough; you need something additional. In the cases, that can be anything from asking the FDA to withdraw your application so that no one else can use it as a reference drug in the future; doing a big recall not for safety or customer confusion reasons, but a recall for other purposes, that might qualify.
The last thing — and maybe we’ll get to this in response to another question — is this is where the bill that’s being kicked around in the Senate, the Cornyn-Blumenthal bill, raises some questions. That bill would prohibit not only the hard switch that Namenda was about but also the soft switch. According to that bill, anytime you launch a new version, like eight years before patent expiry, you are presumptively breaking the antitrust laws, and you have to make a big showing to rebut that, and that doesn’t have anything to do with coercion at all.
Those to me are the biggest factors that we see.
55 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
Robert Willig:
Right. Thank you.
Paul, the floor is yours.
Paul Csiszár:
Thank you very much. Good afternoon, ladies and gentlemen. I was delighted to accept this panel invitation because you call it “product hopping.” In the last couple of years I was much more polite — maybe European habits — we called it “evergreening.” It’s a much more elegant way.
I will start with that, that evergreening is omnipresent, it is in all sectors, everywhere, and every businessman thinks if you have a successful product you always want to evergreen. So it is nothing unique. Big pharma tries to do as the patent cliff approaches and the barbarians are banging at the gate.
The real question is: What conduct should society and the courts allow and where would a line be not to be crossed? This is a difficult question. We just heard Jack saying Namenda is also, at least partially, about additional conduct, not just switching hard or soft. I fully agree with that.
Interestingly, there is very clear jurisprudence in Europe on that. Our highest courts, like in the United States, don’t every year speak about pharma cases or pronounce themselves on pharma cases. We had a case from 2012, AstraZeneca. At that time, nobody called it “product hopping” or “evergreening,” but basically what happened was a very similar scenario: you have a blockbuster drug; the patent cliff is approaching; you have to switch to the new product.
And there was an additional conduct. The additional conduct that took place in the 1990s was withdrawal of the marketing authorization of the older drug. At that time, the regulatory regime in Europe, in most Member States or in all Member States probably, was that once you withdrew your marketing authorization, the generic couldn’t rely on the scientific data which is in that file. Later the legislation was changed, so now a pharmaceutical company has to come up with another additional conduct, but at that time it was the withdrawal of the marketing authorization. The Court concluded that there was no business reason whatsoever, not any purpose, for withdrawing the marketing authorization but for stifling, making the generic entry much more difficult.
I would like to read you — I hate when people read something from case law, but I think the text is perfect because the Court also says that dominant companies have the right to fend off competition from generics and they can do a lot of good things. People sometimes overlook that AstraZeneca stands for the originator as well. Here is the operative paragraph — it’s a long decision, but a relatively short paragraph. The Court said:
The preparation by a firm, even in a dominant position, of a strategy whose object it is to minimise the erosion of its sales and to enable it to deal with competition from generic products is legitimate and is part of the normal competitive process, provided that the conduct envisaged does not depart from practices coming within the scope of competition on the merits, which is such as to benefit consumers.
This famous phrase, “competition on the merits,” is a little hint which should benefit consumers ultimately. In that particular case, the highest court in Europe found that the deregistration went too far.
56 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
This was the additional conduct which did not constitute competition on the merits because it had no business purpose other than delaying or preventing generic entry.
If we think about what the ultimate goal of the pharmaceutical company is when they engage in product hopping, it is basically to extinguish the demand for the original version. If the demand stays strong for the original version, then the erosion takes place both in market shares and in the price, maybe even for the newer product, depending on the added value, which we don’t know exactly how much.
So one should look at it in a holistic manner, looking at what is the regulatory landscape. In the United States you have the AB classification, which is a key issue for generic entry. In Europe, in AstraZeneca, historically it was the marketing authorization staying there or not; tomorrow there will be something else.
So we have to look at the toolbox of the originator and see if each and every conduct has a purpose of bettering the quality of the drug, something that you bring to the consumer, or there was an effort by the company which had no legitimate business purpose — maybe I shouldn’t use “legitimate” because that’s conclusive for lawyers — there is no purpose other than creating an extra barrier.
Perhaps I’ll just stop here and I’ll be happy to answer your questions.
Robert Willig:
There are so many questions in this area, and there are even more questions in the next area, which we’re going to move on to right now, and that is reverse payment. Some people call that “pay-for-delay” — that would be the plaintiffs and the FTC — but the fact is this is reverse payment.
The situation here is a sizable reverse payment, meaning to the patent challenger from the patent holder, who in return gets a license to enter, say, halfway to the time of patent expiration or thereabouts, but who gives up the ongoing patent litigation, the litigation that’s challenging the validity of the applicability of the patent. That is the fact pattern on the face of it.
The question for antitrust policy, for litigation in the courts, and for the agencies is: What else should be necessary on top of those facts to find an antitrust violation?
Jack, you’re the first one. After Jack come Paul and Ingrid.
Jack Pace:
Thanks, Bobby. I agree with the premise built into the question that your facts alone aren’t enough.
Robert Willig:
That’s my view.
Jack Pace:
I see you didn’t ask that question. But if you had, the answer would be no, and the answer would be no not just because that’s a spin on the case law, that’s what the Supreme Court said in FTC v. Actavis. The FTC proposed a quick-look; they proposed a test where if you just made an initial showing of some sort of any form of compensation, the burden would shift to the defendant, and the Supreme Court rejected that. 57 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
Paul, I guess, gives us permission to read from cases, but we all carry around FTC v. Actavis next to the Constitution in our bags.
In explaining why the answer was no, why quick-look was not the right way to go, the Supreme Court said that is because “the likelihood of a reverse payment bringing about anticompetitive effects depends upon its size, its scale in relation to the payor’s anticipated future litigation costs, its independence from other services for which it might represent payment, and the lack of any other convincing justification.”
So, in a way, those are all the other things that I think you’d have to show. I guess I’ll focus on just one for a minute here.
I think in looking at what the so-called “payment” is, it is really important to ask the question of whether there was any sacrifice on the part of the patentee in making the payment: Did the brand company in these hypotheticals give up anything, sacrifice anything?
Again, what we are asking about in these cases is even though you do not, according to the Supreme Court, need to re-litigate the patent case within the antitrust case, this whole thing — whether there’s a delay or not; whether entry is ten years early or ten years late in your hypothetical — depends on what we are going to infer about the validity and enforceability of the patent. If we assume the patent — and I think the Constitution tells us we should do this — is valid, that is an early entry or deal. If we assume the patent is garbage, then that is delayed entry. Which assumption are we going to make?
Much of the litigation is to try to find out what inference we should draw. Therefore, if the brand was completely confident in the validity of the patent, then there really isn’t a lot of reason to make a big payment; but if the brand is giving something up, is sacrificing, then maybe that does tell you something about what they think about the validity of the patent.
Sacrifice by the patentee, to me, is one of the most important things. For example, if there are product deals that are part of a package of agreements going on between the two companies on or around the time of the litigation settlement, then you would want to know, of course, fair value — the Supreme Court asked that: Is the brand making a sacrifice in that way?
Or let’s say it’s a product deal, a license to sell a product that the brand was never going to sell. The brand has rights to something, the world has changed, and now they sell five things in the treatment area already, and product number six is something they really were never even going to market, but it might be valuable to the generic, so let’s give them a license. In that situation, the brand really isn’t sacrificing anything and I wouldn’t think that would be evidence of anticompetitive effects.
The last thing I’ll note just very quickly is one part of these sorts of cases is what the but-for world would look like: In evaluating anticompetitive effects, how important is it whether in the absence of the agreement there would have been earlier generic entry?
That is clearly critical, and I say that’s clearly critical because one I think overlooked fact is that in the United States there have been four trials in reverse-payment cases go through trial to a verdict or a judgment. At the trial court level, the defendants won all four of those cases, and causation was important in all four of those cases. Two were FTC trials.
• My colleagues tried the K-Dur case. They won at the trial level; got reversed in the Commission and then got it overturned in the administrative law judge’s decision; the defense victory reinstated by the Eleventh Circuit.
58 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
• In the recent Impax case, defendants won before the ALJ and got reversed by the Commission. It has not yet been reversed on appeal, but we’ll see.
• The other two were Hytrin and Nexium, both defense victories at trial on causation.
So clearly, causation in the but-for world, which should be welcome to the economists on this panel, is an important thing to ask.
Robert Willig:
Thank you very much.
Paul, would you add to that or from the European perspective or your own?
Paul Csiszár:
Yes, I would like to comment. I have to tread carefully here because we have also cases on appeal.
We have only one judgment. We have two pay-for-delay cases in the patent context on appeal. In the first instance, the General Court approved our first case, Lundbeck, and under that holding I would say that with your fact pattern, what you proposed, it is possible that we would conclude that this conduct is actionable.
Robert Willig:
Without anything extra? Just the money?
Paul Csiszár:
We have to be very careful with hypotheticals.
First of all, I fully agree with Jack that the size of the payment should be very crucial. But very likely with the mere fact that there is an earlier entry, you still don’t overcome from a defense point of view the counterfactual in my view: What would have happened in the absence of the sizable payment? It could have been they never settled; it could have been they settled for even an earlier entry, which is good for society; or — I don’t know what could be the other one — they just litigate it forever probably. I covered that.
Robert Willig:
A “Turducken.”
Paul Csiszár:
Right.
Participant:
Entry at risk is the other one.
Paul Csiszár:
59 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
The entry at risk, exactly. Maybe there would have been an entrant.
In that context, for us, we see that from a potential competitor’s point of view the generic received value — and in Europe the value can be in-kind, not just cash — a substantial value which incentivized it to accept either a later earlier entry or no entry at all.
So we don’t get out of the fundamental conundrum that between potential competitors there was a payment from the incumbent to maintain the market share.
I will just stop there. If I’m challenged on that, maybe I can answer.
Robert Willig: Thank you.
Ingrid, challenge him.
Ingrid Vandenborre:
I’ll keep it short.
Robert Willig:
Oh, challenge him.
Ingrid Vandenborre:
It is interesting because one of the cases is currently on appeal — they are both on appeal — before the Court of Justice, the highest court. The Lundbeck judgment is under appeal now. We had the hearing before the Court on the Lundbeck decision.
Maybe just to set the framework a bit in the European side of things, the General Court decided that it doesn’t matter whether your settlement agreement was in or outside the scope of the patent. One solution would have been “it’s outside the scope of the patent; you have a sham situation; your payment is indicative of a sham.”
The Court clearly said, “We disagree with the Commission on some of the points, some of the agreements being out of scope; but, even if they are in scope, we think there is an issue here that violates the competition laws.”
There was no clarity on what the value of the payment is and how indicative that payment needs to be — should it be at the level of cost, beyond cost, profits — or anything on that. So there is some back- and-forth. How I read the judgement is it is what value you actually have to demonstrate.
It is interesting that at the hearing — I’ll be very short — the Court of Justice asked a couple of questions that were indicative to me of where maybe some of their thinking is going. We’ll have to see where the judgment lands.
On the one hand, they were asking about specific provisions in the agreement that may be indicative of some parts of the arrangements being out of scope in terms of marketing licensing and some of the arrangements being provided. They were clearly asking about those parts of the settlement agreements, not so much about the value, and not so much contesting the lack of a counterfactual. There was no demonstration that absent these settlement agreements there would have been earlier entry, which the 60 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
parties tried to contest before the Court to say, “There’s actually no demonstration because if you litigate forever, you litigate forever, and compared to that this is early entry.”
The second thing the Court asked about was — and this is, I think, a difficult question — how do you assess the value of the patent? Are the Commission or other agencies able to assess the value of the patent? Are they able to identify whether the effect would have been to create an extensive delay of entry or whether the patent was not sufficiently valuable and we were better off litigating in court? The agency’s position typically is there is a duty to take out bad patents, and so litigation is a way of competing, and by settling you take away that possibility.
I think the Court was asking: How do we identify? Should we have experts? The Court asked in the Q&A: Should there be an expert? Should the Commission have appointed an expert to identify patent value? Do we need something more than an assessment based on contemporaneous records as to whether the patent value was significant or dubious, and what is the threshold that we want to put there?
Robert Willig:
I just feel so unsettled about this whole area. The policy is such a mush in the United States, and it sounds like it’s getting that way — or has been that way — in Europe as well. Maybe we should come back to this.
But let’s push on to the question of excessive pricing and, more specifically, about the use of antitrust law or antitrust principles in Europe and here to directly control pharmaceutical pricing.
Ingrid Vandenborre:
I’ll state a disclaimer. I should have done one earlier. I represent one of the generics in the Lundbeck proceeding and I represent Aspen before the Commission in the Italian court.
Maybe some background on excessive pricing from the European aspect. I think the European framework is very specific. It’s very different from the U.S. framework, where this is all falling under state laws.
In Europe the basic framework is set up by a judgment by the Court of Justice, United Brands, that says: when you assess whether a price is “excessive” or “unfair” in the language of Article 102 (your Section 2 equivalent), you look at whether there is a price-cost comparison that suggests that the price is substantially in excess of cost, unreasonably in excess of cost.
Then, if you determine that there is an excess when you look at cost, you still have to determine in the second part — call it “limb two” — of the analysis whether it is excessive also compared to reasonable benchmarks, whether that may be pricing in other countries, similar products; but somehow there needs to be a benchmark analysis.
There have been a couple of cases recently in Europe.
The Pfizer/Flynn judgment you may have heard of. The Competition and Markets Authority (CMA) decided that the agreement between Pfizer and Flynn allowed Flynn to price excessively when they licensed the product from Pfizer and were able to market phenytoin under a nonbranded but higher- priced level in the United Kingdom. Even though it was comparable to the price level of another formulation, it was substantially higher than the price Pfizer had originally put out before the license agreement. 61 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
The United Kingdom had compared the price pre and after the license agreement and said: “There are no real comparables otherwise to really compare with. We think therefore the increase is material. Our benchmark analysis was completed on that basis, and the price is substantially departed from cost.” On that basis, Flynn was determined to have excessively priced.
The Tribunal disagreed and said: “You have to look at more benchmarks; you cannot just look at the historic price and identify that the increase is problematic.”
The Italian Aspen case started off similarly but was decided differently by the appeals courts. The case is currently under Council of State review in Italy.
Also there the Italian Authority had looked at the original price. Aspen bought a total of six oncology drugs from Glaxo. Those oncology drugs were long off-patent — the U.K. product for Pfizer was also off-patent — and had determined, after seeing prices decline since the acquisition and costs increase because of the active pharmaceutical ingredient supply, that it needed to reprice the product. At the same time they sought uniform pricing in Europe because there were substantial parallel trades on very low- volume oncology drugs. They increased prices substantially for the drugs across Europe, and some of the Member States reacted.
Italy brought a case and found the price increase to be excessive in comparison to the original price. Again, there was a view that there were no benchmarks because the market was the molecule and there had not yet been other same-molecule products on the market in Italy at the time. They looked at the forecasted revenues that Glaxo had at the time and said, “The original price was not a loss-making price, that’s still a relevant benchmark, and we will decide on that basis.”
That’s the framework. There are some other cases. There is a Danish case which is more of a parallel trader seeking opportunity for profit, but I think it’s slightly different.
The European Commission is investigating the Aspen case, so it’s an ongoing matter. I won’t comment on the case as well, but I think maybe just to posit some key questions that I think are just questions to reflect on.
In the analysis of what is an excessive price, of course both the first and the second limbs of the analysis, both the price-cost and the benchmarking, matter. At least the Tribunal in the United Kingdom confirmed and agreed that you need to look at benchmarks. There have also been Commission cases in the past where they looked at excessive pricing in areas other than life sciences — for example, public utilities — where you do look at what is the value to the consumer.
So you cannot just look at cost-based pricing; you look also at benchmarks and overall reasonableness. Typically, benchmarks are the easiest way to do that, but even if you don’t find very close benchmarks for comparison, you do look at value, something other than cost.
That is currently the big debate in Europe: Do you look at something other than cost? How do you look at valuation, and does that differ obviously for patented and nonpatented drugs?
In the patented world, the Dutch Authority put out a paper that talks about the quality valuation of innovative pharmaceuticals and attached a figure to quality-per-life extension and to evaluate products in that way. So we’ll see progress maybe.
62 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
You may have read in the press some niche products, biologics, or even small molecules, that are for smaller patient populations have difficulty getting through a quality-based assessment because sometimes R&D is just very expensive. So it’s a big analysis.
Some of the questions are: Do we price a product or do we price a portfolio? Do we price in hindsight? Does the supplier need to have known that he was pricing excessively, or do we assess that after the fact? Does it matter whether there are barriers to entry, that the market is dynamic, or do we look at a specific point in time aside from market development?
Some of the literature on the topic has suggested that at least there should be a couple of screens that limit how competition law should look at excessive pricing.
• First, you should assume there would be significant barriers to entry even if you observe no barriers, because if there are no barriers to entry, the market will evolve; and normally, if you price excessively, entrants will come in and seek to get their share of profits, and so prices will be market-competitive as a result. So the existence of barriers to entry seems to be an important screen for review.
• The second is whether the supplier has some kind of exclusivity that is insulating that supplier from competition. Exclusive grants from the government, a monopoly, a patent — something like that would also be a relevant parameter or a screen to assess whether pricing would be excessive.
• Third is what would the remedy be. That’s the conundrum — and I’ll stop with this — that is faced in Europe, where of course Member States are close, side-by-side, with their brethren at the regulatory authority to be able to converse with and coordinate with them and say, “What do we think should be the right price so you can regulate?” because there are regulated prices in Europe for the vast majority of the Member States.
At the European level, the solution becomes more difficult because you don’t want to enter into price regulation necessarily. There is a question and debate on competency of how you find a remedy without having a regulated price. The question was also: Do we want to get into the chapter of excessive pricing if there is no good way to find a solution when the sector is already regulated?
I wanted to put some questions out. I’m sure Paul’s going to want to give, and should give maybe, some reactions to those.
Robert Willig:
David first. David has been so patient. David, the microphone is yours.
David Gilo:
Okay, thank you. Should I start with excessive pricing, following Ingrid?
Robert Willig:
If you would like to, but then you’re right that you have the lead on the next subject, which is specialized legislation, and in view of time, if you’d like to meld them, feel free.
63 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
David Gilo:
Okay. Since Ingrid was talking about excessive pricing, I’ll start from that and maybe connect it with product hopping and reverse payments as they arise in the United States, because these are practices that testify that the competitive process doesn’t always work well.
That’s why it is curious, in my view, that the United States hasn’t come up with a solution like the European Union, simply to prohibit the excessive price. After all, the aim of antitrust ultimately is to prevent deadweight loss and transfer of value from consumers, and the competitive process is only a tool in order to achieve this.
So, if the competitive process doesn’t work, even though the legislator really exerted efforts in trying to upgrade it with generics, etc., but then it has turned into a sort of cat-and-mouse situation — where the legislator tries to upgrade the generic competition, but then the companies are smart, they come up with a way to hop to another product or reverse payments, whatever — then it would be a good idea to attack the problem at its core and simply prohibit the excessive price. True, it’s not an easy case to bring, but these other cases regarding exclusion and competitive processes are also very complicated to bring.
A few words about excessive pricing and then I’ll go to the fourth question that Bobby asked.
First of all, one question is: Do you want to enforce this prohibition also during the patent term? It sounds provocative, but I would say yes, because if the intention of patent law is to allow recoupment of investment and the excessive-pricing methodology takes this into account and allows recoupment of all investments, all risks, etc., then there is no reason, if done carefully, not to enforce it also during the patent term.
After the patent expires, then for sure the branded firm doesn’t have any expectation for supracompetitive profits. Because competition is what we expect to happen, then it would be surprising to hold that here we need to allow the branded firm after the patent expires to make supracompetitive profits.
Ingrid mentioned the Pfizer case. I have a lot of criticism about the Competition Appeal Tribunal (CAT) decision in that case because the European case law says that the Commission has a lot of leeway on choosing its methodology. So it doesn’t have to do it in two prongs; it doesn’t need to do above a price- cost test if it has a good price-cost test.
Also, the CAT demanded that the antitrust agency in the United Kingdom check, for example, other allegedly high prices in the market. Teva is charging allegedly a very high price for the tablets with the same active ingredient, and if that is excessive too, there is no reason to be lenient toward Pfizer if Teva is charging an excessive price too, etc. So I have a lot of thoughts and criticism about that.
But in order to save time — and maybe we’ll have more time for this later — I’ll touch on Bobby’s other question, which is alternative tools to tackle the problem.
One tool, as we said, is simply prohibit the excessive price. I’m not the first one to say this. Harry First here published a paper recently about this.
But also, a thought that might be naïve but something to consider with regard to product hopping is if a branded firm received a patent for a certain drug, then maybe the Patent Office shouldn’t issue a new patent for a new drug that is too similar to the previous drug. That would be in the discretion of the
64 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
Patent Office, and it would prevent some of the distortions of evergreening that are not justified that were discussed by Paul, etc.
Another thing with regard to reverse payments is that, if I am not mistaken, most of the procompetitive claims about reverse payments refer to the length of litigation, like Ingrid mentioned. So the question is: Why not force the courts to litigate these matters more quickly? I imagine you have such a mechanism for failing firms, for example. If a failing firm is trying to merge and the antitrust agency is challenging that, I imagine you have some sort of hasty way to litigate it in order not to cause too much bleeding by the failing firm. So why not here? If most of the damage is from the length of litigation, why not just shorten it?
Another thing to consider, of course, is to give the doctor more incentives to think about the price and not only about the medical issues; give the doctor a certain budget; or have the government intervene more with regard to which prescriptions are given.
All that being said, if you have specific legislation enforced by a specific regulator, I think that might be inferior to antitrust law in the sense that it is easier to circumvent specific legislation than it is to circumvent antitrust law, which is more on the substance and less specific.
Also, I think antitrust agencies are less prone to regulatory capture, to too much bargaining power on the part of the firms. The reason that the Pfizer case is not a price-control case but rather an antitrust case was that Pfizer and Flynn came up with a way to find a loophole in the price control. This would not happen with antitrust because antitrust is more flexible and antitrust authorities are more independent and stronger than industry regulators.
Robert Willig:
Thank you.
Paul, you wanted to comment on both of those subjects, so please do, but we’re almost out of time.
Paul Csiszár:
Ingrid did a very good job of listing all the formidable difficulties why agencies shouldn’t, and cannot really, make sense out of excessive-price cases. Surprisingly — it’s my personal view I should have said — I tend to agree with her, except for the conclusion that it’s not a reason not to attempt to sort these things out.
First, at the outset, I would just say, as she correctly said, in the European law an excessive-price possibility for unilateral conduct has been in the books for decades also in Member States. Nevertheless, competition authorities, both on the national level and at the EU level, hardly ever use it. They are very, very reluctant, for the very good reason that they don’t want to become price regulators. That’s true for this sector as well.
I’m not talking about the Aspen case, but there are many, many pharmaceutical prices which are high. Our in-boxes are way too limited. This is a slippery slope for price regulations.
Having said that, having talked about all the formidable difficulties, what do we face? We face a massive market and regulatory failure, at least when it comes to generic drugs. It’s the same in the United States. Probably to a lesser extent it’s true in Europe.
65 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
I like to go to these conferences with a screenshot of Mr. Shkreli’s face, Turing Pharmaceuticals, when he says, “Eh, if they want to live, they can pay.” This is what we are facing. You know that his company is still selling Daraprim as of last year for $750 a pill, after two-and-a-half or three years of the scandal.
If anybody would like to preach to me about the invisible hand, the market forces, I’m telling you the invisible hand is really stuck in the pocket. There is no reason. This has become ideology: “The market will take care of it. Entry will come soon.”
Baloney! Google it. What’s the price of Daraprim today? It’s still, three years later, $750. Moreover, if you check, Daraprim is being sold in eighty other countries around the world for between thirty cents and eighty cents a pill.
So, arbitrage doesn’t work. Regulatory solutions don’t work. The invisible hand doesn’t work. It is a mess. It is a major mess.
I have never done the math, because I’m not an economist, but if you look into many of these generic drugs, single molecules, they are manufactured in circumstances which the Soviet Union could manufacture drugs in the 1950s, seventy years ago. It’s basic stuff. It’s almost like utilities.
It’s one thing that society says, “Oh, you want to pay for this bottle of water one or two dollars or three dollars,” and it’s another thing to give tap water to the population at this price, or possibly at this price, and not for the price of the cost-plus utility. What is the cost-plus? That’s a big question. There are formidable difficulties there.
I don’t see why society should go around and waste tremendous money for healthcare — I don’t want to get into how much the Americans pay for healthcare compared to the Europeans, the Germans. They pay double. Maybe your healthcare is twice as good or whatever. But I think everybody agrees that the resources are finite, limited, for healthcare.
So, having the luxury that a drug which is sold in eighty countries around the world for thirty cents — the patent expired in 1963 — three years later is sold for $750 per pill — it’s a big mess. It’s a regulatory failure. It’s society’s failure. It’s a market failure.
Robert Willig:
Thank you. Great example.
Paul Csiszár:
A last point because I would like to make this point. Why antitrust? Why don’t you leave it to the regulators?
Unfortunately, others have heard it on the panel — I always use this example — you wake up in the middle of the night, you look out, your neighbor’s house is on fire. You rush out. You ask your neighbor, “What’s going on? Are the firefighters on their way?” He’s going to say, “Yes, yes, I called them. They’re coming.” What do you do? Do you go back and sleep? No, you still grab a bucket and you start to do something.
You look at me as the neighbor. I know the regulator is coming. They will fix Daraprim and everything else at some point in the future. There is a fire out there, and I have a legislative mandate which says
66 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
sometimes when you have unfair prices or high prices you should do something about it. When the legislator says that, the question is not whether but under what circumstances.
Robert Willig:
I thought you were going to say, “The house is on fire, don’t worry, the market will fix this.” [Laughter]
Paul Csiszár:
Not in this case.
Robert Willig:
Who wants to reply first? Ingrid, George, you’re both jumping out of your chairs.
George Rozanski:
I was going to say, thinking about whether we want to have a policy that would attack excessive prices, I think it’s useful to remind ourselves that we should be concerned, both antitrust and policy, not just about price competition but about maintaining incentives for innovation. Innovation is the engine of progress. It’s what drives welfare mainly. If you reduce returns to investment, even if a drug is off- patent, then there is going to be less investment.
I think it is useful to again look at the evidence and ask, maybe not whether particular prices are excessive, but whether drug prices on average appear to be excessive. There has been a lot of competition introduced into the drug industry in recent years, both from generics and also from payers and buyers getting cleverer about the way they design formularies to put drugs that are therapeutic substitutes into competition with one another.
There’s a recent study, again by Ernie Berndt at MIT in 2015, looking at the returns to pharma R&D over the last two decades. Taking into account revenues from worldwide sales, costs of R&D, costs for production and sales, their finding — this was in Health Affairs — is that returns to investment in R&D for the most recent cohort of drugs they could study — these are drugs introduced between 2005–2009 — were roughly equivalent to the cost of capital, there was no excess profit that they found, and this was the lowest that returns to pharmaceutical R&D have been in the last two decades.
So I think we can take our time and make sure we get the policy right; not just respond to Martin Shkreli, but think about those innovations themselves.
Robert Willig:
Ingrid?
Ingrid Vandenborre:
I clearly agree with that. Maybe I also would expand it to non-innovative drugs. Just food for thought.
The World Health Organization Fair Pricing Forum has talked about and reflected on — it was at a committee meeting, I think, sometime last year — what do we think should be fair pricing and how do we provide both sustainable pricing and affordable pricing, affordable for consumers and healthcare budgets and sustainable for producers? 67 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
They actually reported that there’s a fair amount of antibiotics and oncology drugs all off-patents, but they are having difficulty ensuring suppliers still wanting to make those products because price levels are very, very low and sometimes production costs are not disappearing just because your patent is gone. You still have to label, update your dossier, make sure your product gets to each country. You have API that’s sometimes expensive. The observation of the Fair Pricing Forum was that they are seeing increasing shortages of mainly oncology drugs and antibiotics appearing because of the fact that price pressure is becoming significant.
So barriers to entry are zero, but no one’s entering — not because the market’s malfunctioning. The only malfunction there maybe is because the price level may just not be enough to say, “Well, here’s where I as a supplier will invest and take this product.”
Some companies are responding to the development in these cases to say, “Well, I’m not getting any other drug in my portfolio where I’m going to anticipate maybe needing to revisit the product price because I’m not going to get it changed, I’m not going to get any support for continuing this drug, so I’m not getting my hands wet. Someone else can supply it.”
Maybe there needs to be at some point a solution both on the regulatory side but also on legislation related to maybe orphan drugs, but obviously not only for orphan drugs, that will recognize and say, “Here’s a list of the products that we want to make sure are on the market and what we think should be the sustainable price for that” in a coordinated fashion with the industry, with the sector, because I think there is a shared interest.
I think companies do not want to hear that patients are not getting their products. Companies do want to make sure that that’s the case, but they also want to make sure that there is sustainable pricing, and even for non-innovative pharma.
Robert Willig:
Thank you.
Okay, audience, it’s your turn. There are a lot of points of view here that are ripe for challenge. Sir?
Question [John Spitzer]:
You didn’t address merger control specifically, but when you bring up the house-on-fire analogy, is there a bleed-across from the Commission’s — and I’m speaking about the European Commission here — view on the way they hope to bring all these issues into merger review cases?
Robert Willig:
It’s the European panelists you’re asking.
Paul Csiszár:
Yes. Clearly, if it’s a four-to-three or a three-to-two merger, and obviously the fact that there is no entry — these drugs are absolutely inelastic and basically are not responsive to price increases — of course, any fact pattern in a potential merger which would lead toward that situation or make that situation more likely definitely would be an issue for us.
68 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
I would like to respond to George and Ingrid if you want to pose a question about would be my reply to their interventions.
Robert Willig:
No, no, no. Next question.
Question [Stephen Ross]:
This is actually a question for George and Bobby. Is there a way that economists with sufficient information through discovery could model what a patent would be worth if the patentee’s claims of validity were true? And then, is there a way to draw inferences from that, from the reverse payments, as a way of determining what inferences can be drawn by the level of payment? Is there a way that economic evidence can actually create inferences that courts guided by attorneys can play Sherlock and make inferences based on the actual deal in a reverse-payments case?
Robert Willig:
You’re the panelist, George.
George Rozanski:
And you’re the reverse-payments expert.
I would note, first, that there are some theoretical arguments out there that even if a patent is very strong, you could still see a large payment if there is a big prize out there that the patent holder is trying to protect. So I’m not sure that the strength of the patent really comes into it.
Robert Willig:
Narrowly, if you told me the value in the event the patent was found to be valid and the value of the litigation where the answer’s going to be “You don’t win, plaintiff,” then one could start to adduce the odds that each side perceives of winning and losing. What a court would do with those particular indicators is really way up in the air. We’re back to the “Turducken” of the patent case inside of an antitrust case that the Supreme Court wisely foundered on but didn’t really take us very far beyond there, I’m afraid. That’s my best answer.
David Gilo:
Can I add a short comment about that?
Robert Willig:
Oh, good. Are you an economist? Oh, yes, you are.
David Gilo:
First of all, as Jack said, the Supreme Court has taken patent strength off the table actually, so you are not allowed to litigate that anymore. A few explanations that I saw in the literature weren’t convincing to me.
69 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
Basically, the way I read the Supreme Court in Actavis is that the only thing you can actually claim is how to justify the size of the payment. The things that I saw were:
(1) Risk aversion of the patentee. But it’s a little hard to assume that such a large multinational public corporation in such a risky market is risk-averse, so that wasn’t convincing to me.
(2) Another explanation I saw in an article by Padilla is with regard to asymmetric information about how pessimistic or optimistic the parties are. For example, if the patentee thought that the generic is optimistic but actually he turns out to be pessimistic, that could explain the size of the payment.
But I think this sort of analysis is not very viable in court because how are you going to prove what the patentee thought about how pessimistic or optimistic the generic was? Moreover, if you can’t prove it and the asymmetric information goes the other way, then the analysis would also probably go the other way.
George Rozanski:
Well, I thought the factor that the Court was really pointing to for determining whether the size of the payment is out of line is the avoided litigation costs.
Getting back to your last point, I think there is some concern that has been expressed in the past about if you limit the role of cash in a settlement, that could reduce the possibility that parties would reach an efficient settlement and could limit challenges to patents that we think are procompetitive. I think Judge Posner put that out there at some point, and I think, Bobby, maybe you picked it up in some theoretical work that you’ve done subsequently.
Robert Willig:
I like to think when the Supreme Court said the “costs of litigation” that they didn’t just mean the costs of hiring experts and counsel, but of course one of the major costs of litigation in patent disputes, especially when there’s so much value on the table, is the cost of the uncertainty that attends the litigation process and what that does to the internal decision-making of the parties to the litigation in the first place. No court has explicitly adopted that, but that hasn’t stopped me from trying give speeches about it.
More questions?
Question [Cristina Caffara]:
It’s a question for George, but I’m happy for Paul to jump in. George, you put out there the sort of standard economic argument that you hear trotted out every single time in these cases, which is: the portfolio investment we need to fund, we need to support all these risky investments — the same kind of stuff that we hear in the media industry sometimes. We’re all familiar with that argument, and we understand it, but what is the limited principle to that argument? I’ve heard every single economist in London showing up at the CAT with various degrees of shamelessness making that argument every single time in a generic sense.
When you’re looking at something like hydrocortisone, which has been there forever and ever, and has been bought by some people, and is being jacked up in price by 3000 percent, and you’re the CMA, you’re Andrea Coscelli, and you are in a government which is in trouble, and you have the health sector which is on its knees, and you say, “Oh, I’m going to turn my back because the regulator is coming,” 70 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
and the economists are showing up and saying, “It’s the big investment theory,” I feel a bit perturbed by that.
Robert Willig:
What’s the question, ma’am?
Questioner [Cristina Caffara]:
If he would agree in principle on this point I asked.
George Rozanski:
I think I would always err on the side of making sure that we maintain investment incentives.
Another argument to consider — and maybe this gets back to the question that was raised earlier about mergers — is there’s a very provocative argument out there by some guys at Northwestern, Besanko and Dranove and Garthwaite. Their argument is that you consider a firm that has a portfolio of drugs: if they have a broader portfolio of drugs, then they are more likely to internalize the cost; if they increase prices, they are going to induce regulation, or maybe they are going to drive consumers away from being insured, and it’s only because consumers are insured that these high prices are sustainable.
In fact, they take that to the data and they show that drug companies now are accounting for, on average, a smaller share of leading drugs than they used to in the past, and they are pointing to this as a reason why we might be seeing higher prices. It’s essentially a Cournot complements effect. I think that doesn’t necessarily clash with merger enforcement in the United States, where the FTC defines product markets for drugs very narrowly and only challenges those mergers that involve substitutes.
On the other hand, getting back to investment, if you were concerned about maintaining competition in innovation and you thought that the basis for that investment is more broadly defined, then you would be concerned about a merger of large pharmaceutical companies because you don’t know where they could be investing their dollars next.
Robert Willig:
Speak, Paul.
Paul Csiszár:
Very quickly, if there is any truth in it that there is a correlation of the Shkrelis and the other opportunistic jack-up prices funneling back into R&D, there is a much simpler way to deal with this: raise the patent exclusivity period. It’s very simple: make the case that we are struggling. Big pharma spends 17–18 percent on R&D and next to it spends 22 percent on advertising and marketing for prescription drugs. That’s crazy to me.
How do I know if they make more money on these product-hopping and other situations that the 17 percent will go up to 20 percent and not the 22–25 percent? How do I know that? The simplest way is you grant more patent rights for that. I’m sorry. It’s rudimentary almost.
71 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
George Rozanski:
As I said, the best and most recent evidence seems to be that currently we are not observing excessive profits for R&D investment. So maybe we are in the right place and we don’t need to do anything big and bold here.
Robert Willig:
Closing statements.
David Gilo:
Can I make a quick point about the notion that the regulator is coming? Sometimes he never comes, and sometimes when he comes he’s too weak. In some of these excessive-pricing European cases it was actually more like a negotiation with a government, because the government didn’t want a lifesaving drug to leave the country and these companies always threaten, “If you don’t give us what we want, we’re going to leave.” I think antitrust law has a better infrastructure to fight that sort of thing, and also fight loopholes, etc.
Robert Willig:
Thank you.
Jack, you’ve been quiet for a long time.
Jack Pace: You gave me the first two subjects.
Robert Willig:
Right.
Jack Pace:
I’ll say just simply — it may be a slight tweak to what David said about the relevance of the patent in a case like this — I think there is a big question out there: What do you do? If the patent is so important because essentially its validity tells you whether the deal represents a delay or not but the Supreme Court has told us you can’t re-litigate the patent case, what do you do?
My answer wouldn’t be that patent strength is off the table or is irrelevant, but that you do have to ask that question in a way that just doesn’t — unless the parties are going to just waive privilege and produce all their patent analysis, there is a way — and we are trying it in some cases right now, and so maybe at the next conference I can report back on how it’s going — through objective evidence, outside experts who didn’t work on anything coming in and testifying about just objectively what they think the likelihood of success was in a patent case, and then you could look at that and maybe answer the question about what is the worth of the patent.
You could look at that and come up with “if the parties had a 30 percent chance of winning the case, what payment would be large in that scenario?” That’s a question that you could ask. There hasn’t been — Wellbutrin in Philadelphia is probably the one case that really got into it — but that’s an area of development.
72 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
Robert Willig:
With great economics, I would say.
It’s 3:00. Larry, are you dismissing us?
Lawrence White:
Yes. Thank you.
Panel 4: In-House Counsel Session: Platforms
Harry First:
I always like to start my classes by saying, “So here are the answers to the exam,” but for this group it’s maybe not quite the same thing.
Welcome to the final panel. The panel is called “In-House Counsel Session: Platforms.” Dennis Carlton actually in answer to a comment that I posed to him said, “My God! What do we want to do, go back to the 1960s?” And I thought, Well, okay. Maybe not quite to that extent.
I’ll start with a cultural reference to the 1960s. How many of you have ever seen the movie The Graduate?
Larry Phillips:
“Plastics, Harry, plastics.” [Laughter]
Harry First:
I wanted to get — I don’t want to say younger people’s views, but just make sure — and that’s exactly where I was going, Larry. A good one. There is a scene where there is a graduation party and Dustin Hoffman is being told by one of his father’s friends where the future lies. As Larry said, he says, “Plastics.” I was thinking that if you replayed that today and it was a law school graduate who you were talking to, you would say “Platforms” (same p-l-). Okay. So that’s my transition for that.
Platforms are in many ways at the center of our antitrust discussion today both in the legal sense and in a policy sense. We are going to try to explore some of those issues. I’ll set it up this way, just to give you a little run-through about why platforms have become so important. In some ways, what we are going to talk about I think also picks up threads of what people have said on all the panels today, and we’ll see how that goes.
First of all, we are going to talk about two recent Supreme Court decisions. We don’t get all that many antitrust decisions from the Supreme Court, but in 2018–2019 we have Ohio v. American Express on two-sided platforms and, more recently, Apple v. Pepper that involves private litigation for monopoly overcharges. Legally, these are high on the agenda.
73 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
More recently — and perhaps a little less well-known now, but I think it will be relevant — is a complaint that the Federal Trade Commission recently filed against a company called Surescripts, which has what they call an “e-prescribing platform” where they are able to link up providers and pharmaceutical companies for verifying prescription drugs. This is the Commission’s effort to take some of where the Court was going in the American Express case and apply it to another case, here a monopolization case, looking at high-loyalty payments for Surescripts to maintain its monopoly in some market. So, this is a very live issue for antitrust today.
Platform acquisitions have been big, and we are going to talk a little bit about those as well. I tried to write down all the platform acquisitions I could think of and looked at the major platforms that have been the focus of attention — the “fearsome five” as some have called them. If you look at acquisitions by Amazon, you have Zappos, Fabric, CDNow, Audible, Goodreads. Facebook acquired WhatsApp and Instagram. Google acquired Nest and DoubleClick; YouTube acquired Waze and ADMA. Microsoft acquired LinkedIn. A lot of acquisitions in this area. How do we think about those? These are all high- tech industries, acquisitions of companies at various stages in their development. So how are we going to think about those?
And, actually enforcement — as I said, in addition to cases that have reached the Supreme Court, there is something which I like to call RoW (rest of the world), which is very active in this area, many say way more active than in the United States, whether appropriately or not.
We have the European Commission cases against Google, three cases now, with fines that are in excess in the three of cases of over $9 billion, which I guess Google at some point will have to start paying attention to. Nine billion is a lot of money.
The European Commission is looking into Apple’s treatment of competitors in the App Store and opened up an investigation into Amazon’s use of transaction data. There is both enforcement and decisions by the German Bundeskartellamt under German law: against Facebook in February of 2019 for abuse of dominance in the way Facebook combines and uses data; opened up in November 2018 proceedings against Amazon to examine practices it has been using with regard to other competitors on its platform.
Italy has investigations. Turkey has four cases. There are cases in Japan. It is a worldwide phenomenon of looking at the practices of these companies. So how we are thinking about platforms and the issues they raise are important worldwide really.
There is another aspect that I hope that we will get into, which is not just what is going on under current law, but what changes people are arguing for, whether in the way we think about antitrust doctrine today or in terms of legislation. These are well-known things.
There is a call to break up Facebook. My argument is you could break it up into Face and Book. Why not? That argument was made by Chris Hughes, one of the founders of Facebook, who has a little more cred maybe than I do. There is also Elizabeth Warren’s proposal. We are going to get into and talk about some of these. And, even in terms of doctrine, Judge Diane Wood, Chief Judge of the Seventh Circuit, gave a talk a couple of weeks ago where she urged looking back at some antitrust doctrines that have been left for dead by the Supreme Court and said, “Well, maybe they should be livened up” — essential facilities, for example; leverage doctrines that actually are being used in Europe.
So it is a busy time and, as Commissioner Phillips says, there is an invigorating focus on antitrust and this is a great time for the antitrust area.
74 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
A last thing before turning to the panel. This is an article last Wednesday’s New York Times [May 21, 2019] by a tech columnist whose name is Kara Swisher, a very thoughtful tech columnist. The headline is “Taming the Apex Predators of Tech.” You really don’t have to read the rest of the article. One thing that she says is, “To rein in monopolies, maybe we need to rethink what a monopoly is.”
We do live in interesting times.
This panel, which I have to say is a great panel because I’ve actually talked to them, was figuring out what they were going to say just five minutes ago, so it is going to be right up to date.
I am going to do a little introduction. I know Larry said at the start that all of the information is in your materials and you can just read it. But, like the Mueller Report, you probably haven’t. [Laughter]
Boris Bershteyn, who is a partner in Skadden’s New York office and represents various financial institutions, will talk a little bit about two-sided market issues. He also has government experience: Acting Administrator of the Office of Information and Regulatory Affairs in the Obama Administration and General Counsel of the Office of Management and Budget. He clerked for Justice Souter and Judge Cabranes. But two important things about Boris.
First, the most important thing that you have to know in his résumé is that he is Chair of the Advisory Board of the Institute for Policy Integrity at NYU Law School — remember, you are sitting in the Business School, so that’s a little plug for the Law School. This is really a great group which is run by our former dean, Ricky Revesz, involving looking at problems of environmental issues. He gave me a fun fact, which is that Greg Rosston, sitting right there, who is also a panelist, had supervised his undergraduate thesis on predatory pricing. So there’s the generations. But what I thought was really interesting about that is that you wrote an undergraduate thesis on predatory pricing.
Boris Bershteyn:
I think that says more about what passes for a fun fact about me. [Laughter]
Harry First:
I was going to say that.
Sam Knox is next. She is Associate General Counsel for Competition at Facebook and she is going to tell us why Facebook should still be one word and not two. She formerly practiced at Davis Polk in the Antitrust and Competition Group and is active in the ABA Section of Antitrust Law.
Next is Martin d’Halluin, Vice-President and Associate General Counsel for Antitrust at News Corp, a company that is dependent on platforms. He had been an antitrust partner at Morgan, Lewis prior to that. LLM degree from Harvard — I don’t understand that — and a law degree from the University of Paris Sorbonne, so he is admitted to practice both in France and here.
Finally, David Higbee is a partner in the antitrust practice and group leader and head of the Washington, D.C., office of Shearman & Sterling. He was formerly Deputy Assistant Attorney General and Chief of Staff in the Antitrust Division of the Department of Justice.
So we’ve got people with both really great practice experience and some government experience as well.
75 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
What I’ve asked the panelists to do is we are going to start by going to go down the table and hear from each of them what they see as the critical issues in this area that they see either from their practice or as policy matters; once we do that, we are going to pick up on some of the issues that they have raised and have some discussion; and then we will ask you to have some discussion and that will be the rest of the panel.
Boris?
Boris Bershteyn:
I will start with a conversation about the two Supreme Court cases that Harry mentioned, American Express and Apple. I practice both as an antitrust lawyer and a general appellate lawyer, and so this is kind of at the crossroads of what I do.
My perspective on the two cases is that the kinds of reactions we have seen in the popular press to the two cases has been largely overblown. I think the two reactions I’ve heard are either that each of these cases represents some sort of a doctrinal sea change, or a change with significant practical effect across industries, and that the two cases are sort of two sides of the same coin. I think both of those are a bit misguided. Let me start by describing the two cases very briefly.
In the American Express case, the Supreme Court analyzed the markets for payment cards, in particular the types of credit card markets that American Express competes in. What the Supreme Court held, in a very nutshell, is that the appropriate way to analyze those markets from an antitrust perspective is to consider both sides of that two-sided platform market, the consumer side and the merchant side. The intuition behind that, of course, is that no cardholder will want to carry a card that no merchant would accept and no merchant will want to accept a card that no cardholder carries, and a credit card transaction — that market really operates on a transaction-by-transaction basis — has a cardholder and a merchant on the respective sides of it. So, in that case, the antitrust effects of anti-steering rules of American Express cannot simply be analyzed by looking at their effects on merchants without taking proper account of their effects on consumers.
The Apple case concerned a very different question, and I’ll make a case throughout my remarks that it is different in many respects. The question is on the application of the bar under U.S. federal law against private damages claims by indirect purchasers. The question arose: What happens to a claim against Apple by a user of an Apple iPhone — and I venture a lot of us here in the room have used one of those — who purchases an app? Is that user a direct purchaser of the app from Apple; or is that user a direct purchaser from the app developer and has no standing to sue Apple because of the indirect purchaser bar? The Supreme Court reasoned that a claim can be instituted against Apple because there is in effect no intermediary between the plaintiff consumer (the user of the iPhone) and the defendant (Apple).
I will submit that neither of these cases is an “end of the world” case, but in both cases we have heard some dramatic reactions from those who are on the short end of the decision. But that is a pattern that we see with almost every antitrust case the Supreme Court decides — Texaco v. Dagher a little over a decade ago, which was a case about what’s a single entity; Bell Atlantic v. Twombly, which was a case about what it takes to plead a conspiracy — there is a cycle of grief that the antitrust community goes through where first everybody, to use a technical term, “freaks out,” and then it turns out that the effect as the decision gets absorbed by the lower courts becomes much more muted than anybody initially anticipated.
76 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
Antitrust law, unlike bankruptcy law or tax law, is a common law kind of doctrinal area, and it rarely changes abruptly because it isn’t a code-based or a statutory kind of field. So all the changes in course, even from the Supreme Court, become muted as they become absorbed by the judicial system.
And I think the same is going to be true here. The American Express case surely has a lot of insight for other payment card cases and may be dispositive of some of those cases, but is every platform a two- sided platform in the same way that the payment systems are? I think that becomes a much more difficult question as we move away from those that are transaction-by-transaction based in the same way that a credit card might be. Similarly, the type of arrangement that Apple has with its iPhone users is not precisely replicable by the other platforms. And the case itself is a case that largely rests on the particular facts and is not a dramatic departure from what doctrine has been on the indirect purchaser bar.
The other reaction that I have seen in the press about the two cases is, “What is the Supreme Court up to when it decides those two cases? Is it ‘one hand giveth, one taketh away’?” Thinking about the intent of the Supreme Court across two terms is a little bit like thinking about the intent of Congress: It’s hard enough to divine what the intent was of every particular member of Congress; thinking about what the intent was of the entire body is almost impossible.
Similarly with the Supreme Court. We have kind of a phenomenon here where over two years we have had ten Justices, because Justice Kennedy has been replaced by Justice Kavanaugh, and each of the two cases was decided by a 5–4 majority. There were ten Justices in that majority and not a single Justice was in a majority twice. So thinking about what they might have intended across the two cases is a little bit counterfactual.
It is also easy to gloss over how different the doctrinal settings of the two cases are. They are really about two very different questions. In American Express the Court was asking: How does the antitrust law think about what market the players here are competing in? It’s an inquiry designed to replicate the economics of the real world.
That is not really what the Apple case was about. The Apple case was about a highly formalistic rule against claims by indirect purchasers. That rule, by definition, is not intended to replicate the real world because it is a formal legal rule. What we see in the five-Justice majority and the four-Justice dissent in that case is both sides pointing at each other and saying, “Oh, you were just engaging in a formalism, who is a direct purchaser and who is an indirect purchaser,” to which my reaction is: Well, of course it’s a formalistic rule, and the fact that Justices split 5–4 as to how to apply the formalism is further testimony to its formalism.
So I see the two cases as having no overarching design and no particular intent or theme to them. But from the perspective of a client — this is supposed to be an in-house panel — I see where they lead platforms as very different. I think what American Express emphasizes is how important it is in a litigation setting — and I’m just a litigator — to persuade the finder of fact about how your business actually works and what the true incentives are, as American Express was successfully able to do; but it is really about persuasion; it’s not about planning. Whereas the bar on indirect purchasers, as I said, is a formalism, and what lawyers do is try to build structures around formalisms. There is probably a lot of careful planning that platforms can undertake to try to shield themselves as we learn more about how courts are going to apply the indirect purchaser bar to platforms.
Samantha Knox:
Picking up on the newspaper article that Harry referenced, I just want to state the obvious. Antitrust has been making a lot of headlines lately. It has been on the front pages of papers, coming up in my dinner 77 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
table conversations — maybe yours, too — and it has become a litmus test for presidential candidates in a way that I think few of us would have expected just a few years ago.
I think the current debate around antitrust is arising out of a broader discussion and a broader concern around concentration, income, wealth, and equality. These are important issues that warrant discussion, but it is far from clear to me that these concerns warrant the type of scrutiny that we have seen being focused on tech and, in particular, digital platforms.
Digital platforms account for only a small percentage of the U.S. economy, and changing antitrust enforcement in this area with specific rules for digital platforms is not going to have a significant impact on those inequality concerns or on some of the other policy and substantive concerns that are being raised under the guise of antitrust right now. However, many of the changes that are now being discussed in the antitrust realm would have a far broader impact on many U.S. companies and I think could significantly hinder economic growth.
So what do U.S. companies and their lawyers need to know about this right now? I thought I would share a few observations from my experience on the frontlines with regulators in the business over the last year and a half. There are three points I want to make here.
• First, some are proposing fundamental changes in the antitrust law that would shift the focus of antitrust law from conduct to market share. What’s wrong with this? First, in the United States, in Europe, and elsewhere, market share, even a monopoly market share, is not illegal. What is illegal is anticompetitive conduct that is causally related to an increase in market power. Market share is not grounds for imposing sanctions on, much less breaking up, a company.
So I would pose: Shouldn’t we keep our enforcement efforts and our rules focused on punishing illegal conduct and not success? In fact, isn’t that the premise of the antitrust regime in the United States, that we want to encourage companies to compete vigorously, and that companies that compete through proper means, through lawful means, should be entitled to success and, more to the point, shouldn’t be punished for merely being successful.
• Point number two: Some are considering and proposing sector-specific regulations. However, antitrust is, and has long been, a law of general application. It is a law that has worked over the last hundred years in the United States across industries, no matter what the industry or the particular company is.
Notwithstanding that, in recent months in the United States and elsewhere we have seen some proposals for sector-specific regulation within the field of antitrust — not talking about regulation writ broad, but talking about special rules for certain sectors or certain companies in antitrust law. A few examples here: special merger notification requirements for digital platforms; special presumptions that would only apply to mergers in the tech sector. Those are just two of the specific examples being floated.
I think these proposals show a somewhat surprising lack of confidence in our existing laws and our enforcers, and it’s not clear to me why we need a special set of rules just for digital platforms. Just in case you’re thinking, Wait, what about network effects? What about winner-take-all markets? What about tipping points? I would just say that these same concerns have previously been raised about other sectors; they have been raised about wireless, broadband, cable, hardware, software, among other industries.
I would venture that when you are looking at apps, in particular, and free user-facing services — that’s what the biggest hot tech platforms today, like Facebook, provide — there are a couple of things that stand out here that distinguish Facebook from some of these other industries. 78 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
The barriers to entry right now for creating and launching a new app are much lower; there are low, if not no, user costs for switching. The evidence for both of these things is found in the fact that we do see people multihoming; we see people using multiple apps to share photos, multiple messaging apps to connect with their friends. You don’t have to take my word for it; that is in the stats.
In this environment, I would suggest that where you’ve got a highly dynamic industry — where there are low barriers to entry, there is constant entry, frequent new entrants, and low barriers to switching — might it not be better to let competition work and to have some second thoughts about regulation that might actually serve to hinder competition?
• Third point: There are some regulators and some politicians who seem to believe that mergers themselves are inherently anticompetitive. I just want to pause for a second to let that sink in. This has taken a few different forms.
There have been proposals in the United States to lower the standard for what would be considered an anticompetitive merger, lower the standard for harm to competition; or to shift the burden of proof entirely, to have to prove that a merger is procompetitive and otherwise it could not go through. There have been commentators, and even regulators, who have pointed to the number of acquisitions that a particular company has done or the total dollar value of acquisitions that they’ve done as evidence of some harm to competition in and of itself.
If our competition policy were to become driven by this kind of inherently suspicious view of mergers, what would be the economic consequences of that? In Silicon Valley mergers are one of the primary forms of exit for startups and for their investors. I can also tell you that, based on my experience in- house, mergers sometimes yield efficiencies that cannot be gained through a partnership or a joint venture; and sometimes it’s not possible for the company to build the product itself.
I do want to be clear. Mergers that would substantially lessen the competition, even if they brought efficiencies, those mergers should be blocked. I’m not saying that we should not have merger enforcement. But I do think that proposals that are based on some sort of inherent suspicion of mergers or a view that mergers are in and of themselves anticompetitive, whether this is on an economy-wide basis or only for certain industries, is something that should concern both Wall Street and Main Street.
Martin d’Halluin:
Hi. Good afternoon. As an introduction, I wanted to tell an anecdote. Two years ago when I left Morgan, Lewis to become in-house counsel at News Corp, I had a feeling that what News Corp wanted me to do, among other things, was to see if there were antitrust issues between the platforms and News Corp as a content provider. Honestly, I was not sure I would take the job. I told them: “Wait. I mean they’re very powerful, I work for some platforms, I know them, people love them, and they have a lot of procompetitive aspects. I’m not sure there is a case; and, if there is no case, what am I supposed to do, come for six months and then go back to private practice?” [Laughter]
I share this anecdote because I’m here two years later, and I still have a job, and I try to make these arguments for News Corp and help them. But clearly, a lot of people outside are making the arguments, content providers and others, so much so that now we have a panel where we see platforms being on the defensive side and trying to explain, “Well, we’re small; we don’t need intervention,” and many things that we would never have heard two years ago.
79 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
I do not want today to focus on market power, whether these platforms are big and powerful. To me it is pretty obvious as a content provider that they are big and powerful and unavoidable, and I will quote a few facts.
I also want to share before I start that News Corp owns some platforms in the real estate business, some real estate digital platforms. Our subsidiary Move, Inc. is the operator of realtor.com®, Doorsteps®, and moving.com. It’s not that big in New York, you may not know it, but it’s bigger in the West and in the Midwest. It is also a very big platform in Australia.
News Corp owns a lot of newspapers — “news media” as we like to call it, to not remind everybody that it used to be paper — and a book publisher, HarperCollins. So you can see where I’m going with the issues with platforms. There’s news and books.
What I want to discuss in my short presentation is the relations between platforms and content providers. When I thought about this panel, I remembered the French book called Platform by Michel Houellebecq. It was published fifteen years ago. You may know it or not. You may like it if you like French novel writers. He is pretty famous in France. That being said, I read the book twice and I never understood what he meant by “Platform” as the title of this book about a guy who is like losing himself in Paris and Thailand.
It’s kind of the same thing here. The word “platform” in antitrust has a lot of meanings, and it is very not clear — and I’m with you, Boris — what exactly we mean. As a content provider, there are three things that I wanted to raise.
• First, there are a lot of conflicts. The question is often: Are you in a horizontal relation with platforms or are you in a vertical relation with platforms? Well, it’s complicated. It’s like my status on Facebook a long time ago. [Laughter]
News providers are on both sides, really. We rely on platforms for distribution of content. I want to be clear here. There is a lot of direct traffic to news websites, but really, when you talk about news readers who sour it every day, the people who used to buy their paper at the kiosk, this is happening only indirectly and this is happening from people who go through Google and Facebook. People usually do not go to your website because they just want to try it. They only discover your website, discover the news, by going indirectly. So there is a big distinction between the two traffic sources.
But then there is also big competition. Obviously, our newspaper/news media is ad-funded still a lot. Clearly, the platforms that at first were distributing news articles were competing for the same ad dollars, and competing very, very, very well, so much so that there is less and less ad money for the news media, which is causing a lot of trouble, because then you just have to raise a pay war, hoping that you are going to have enough subscribers, or you just basically fire journalists, which is happening at many publishers. That’s one thing I wanted to raise.
• The second one is a legal fiction, and it’s about the Apple case. What I thought was interesting in the Apple case is that the Supreme Court said that you need to look at the economic reality of what is happening. If you read — and the Supreme Court didn’t talk about it — the Apple terms, Apple is an “agent” under its contract with the user. I looked at it today. It says, “Apple acts as an agent for App Providers in providing the App Store and is not a party to the sales contract or user agreement between you and the App Provider.”
So the problem is: What does being an “agent for an application sold on the App Store” mean? How can you be just an agent when you own the device and the App Store? It could be — and it wouldn’t be the 80 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
first time — a legal fiction because realistically for the consumer he is buying an app from Apple, but Apple made the terms so much so that the relationship is supposedly between the developer and the user. I thought it was interesting that the Supreme Court was able to look past that fiction.
The next thing I want to raise is the Supreme Court decision in Ohio v. Amex. What I thought was very interesting in that decision was, not only the decision itself, but the way the Supreme Court explained you have to look at transactional platforms. This doesn’t mean all platforms.
What did it say? From the opinion of the Supreme Court: “The key feature of transaction platforms is that they cannot make a sale to one side of the platform without simultaneously making a sale to the other.” Well, it excludes a lot of so-called “platforms” from that definition. Specifically, the opinion excluded newspapers from transaction platforms, and that’s interesting because newspapers have for a long time been seen as the old platforms. The key seems to be that basically more readers brings more ad dollars, but more ad dollars is not going to bring more readers, and you therefore don’t have perfect indirect network effects; therefore, they are not covered by the Amex v. Ohio decision.
• The last thing I want to raise, which really, I think, is not well-understood — because it is very complex because of the scale and speed and the use of consumer and pricing data by the platform — is the different levels of integration. For this last example I’m going to use Amazon.
Amazon is a lot of things. Amazon is a publisher, a manufacturer, a retailer, a provider of distribution services, a provider of storage services, a provider of shipping services, a provider of marketplace services. When you buy a book, physical or e-book, on Amazon, you may be buying one or multiple services from Amazon, and the same for the book publishers. The problem is not so much that they are different services, but that they are constantly integrated, so you buy packages of services together. From the publisher perspective, when you sell an e-book, Amazon is a so-called support agent, but it also owns a device, and therefore you are in a relation with them that is not that clear.
If you sell a print book, there are many, many, many different layers there. You can sell a used book; you can decide your book is going to be stored at Amazon’s warehouse; you can decide that you are just going to buy a marketplace service; you can decide that you are going to be the seller or you can decide that Amazon is going to be the seller; and there may be others.
What I’m trying to say is that we keep talking about direct network effects and indirect network effects and they work very, very differently. When Amazon is a marketplace for third-party sellers, indirect network effects are pretty big the more sellers and the more consumers are going to buy; when Amazon is selling its own products, there may be not so much indirect network effects; but the two things are combined on the same page and it is very hard to see what effect is from what and is guiding to which consequences.
That’s what I have for this panel.
David Higbee:
The benefit and the burden of going last on the panel for your introductory remarks is that some of the things that you were going to say may have already been said. In the interest of time, I will try to comment on these issues that have already been raised and maybe get into some other issues for the panel.
I was interested, Boris, in your description of the two cases, and I agree with you that they are not necessarily watershed cases and that unique factual distinctions may limit them. Certainly, Joe Simons 81 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
and Makan Delrahim said when the American Express decision came out that there would be limited application.
But I am interested in discussing at an appropriate time in our panel whether the factual differences and the context in which they arise justify what appears to be a difference in how the Court looked at the markets, in the sense that I think in American Express the Court described the market as “two inextricably linked markets or transactions” and it defined that collective as “one market for the enabling of transactions.” It basically compressed the whole market into one.
In Apple v. Pepper the Court did not look at the whole chain as one market, but it looked only at the platform-to-consumer side of the market in deciding the Illinois Brick question, and not looking at the app-developer-to-the-platform-to-the-consumer. It’s interesting. I don’t know, but I’d be interested in your thoughts on that.
The other thing that Sam said that was very interesting to me, and with which I agree, is caught up in this tension of the dialogue that we are hearing these days, two simultaneous messages that I think are in tension with each other. One is that the large platforms are violating the antitrust laws. The other is that the antitrust laws need to be changed to address what the platforms are doing. I think the second argument undercuts the first argument because if there need to be changes to the antitrust laws to address them, then by definition the current antitrust laws do not reach the conduct.
That tension comes from, I think, something about the conduct — in part, maybe it’s the climate and the other related issues that Sam talked about that are attracting the attention. When economic or content elements — purity and integrity, economic disparity; privacy issues I think is a big driver — are not being addressed adequately, people are grasping for some tool and saying, “Well, they’re big, so let’s use antitrust as a tool,” as opposed to asking, “What regulatory approach, if at all? What dialogue should we engage in in order to address those concerns?” Anyway, that tension is something that’s of interest, and it reveals a little bit about whether the conduct does violate the antitrust laws if it can’t be reached and hasn’t to this date.
I have the same comment as Sam made. It goes to the question of the sense that antitrust historically hasn’t been a status offense — meaning “big is bad” or large market shares alone if they were acquired through superior business acumen, first-mover advantage, or significant investment — but antitrust enforcement has been about conduct.
I have heard too much criticism, without defining specific antitrust markets, about the size of the companies and the power that they wield, and not enough about allegations of specific conduct that violates the antitrust laws, in part because of a lot of the dialogue that we read about most often is more in the political forums than in the academic forums. Maybe you guys are hearing it more than I do because I attend fewer academic forums — I’m practicing too much. But that’s something that I think needs to be sussed out, specific allegations of conduct.
And the private enforcement; where’s the private enforcement?
I think that if privacy concerns and other concerns are driving a search for a tool or new tools to reach some of these concerns when specific antitrust conduct hasn’t been articulated clearly, the question is: Should it be done? Should an expansion or modification of the antitrust laws be done judicially, through a ten-to-twenty-year patchwork of court decisions that over time result in a moving of the goalposts, or should it be done through legislative dialogue and action?
Those are some of the things that have been on my mind as we have talked about this. 82 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
Harry First:
I like the moving of the goalposts because that comes back to the first panel on sports. This just shows everything goes together.
As I remember, on that panel someone said the courts are so slow, and I think Roger made the point that we have to redo everything every fifteen years; and they’re terrible — now they’re filled with Trump appointees, my God! — so we have to go to legislation. Then we come to this panel that says, “No, it’s a common law enterprise. Legislation? No, we don’t want legislation.”
I have a couple of questions for the panel. I’m trying to think about it this way. First of all, there is a lot of talk in American Express that there are ways to limit it, to distinguish it. Curiously, the Court came up with the idea of a “transaction platform” that it made up from one line of an economists’ article. Why did it do that? Was it right or wrong?
So I want a vote. How many people think that the American Express case was correctly decided? [Show of hands] Oh, that’s really interesting. I’ve heard this case.
But Martin says “no.” Go ahead.
Martin d’Halluin:
I have a caveat. I think economically it makes a lot of sense. It’s just that it is putting the burden of proof where it is going to be very, very hard to get cases against the platforms. That is my caveat.
Harry First:
That’s the second point of it. What I think is the true oddity between the two cases changed by one vote, which is that American Express in the end — and Michael Hausfeld made this point on our first panel — makes it really hard for plaintiffs to win because of the burden that is given the way the Court wrote the opinion. I mean it is true, because of the two-sided platform, but putting the complete burden of proving effect, in effect negating the justification, makes it really hard for plaintiffs to win.
Apple v. Pepper, on the other hand, makes it easier for plaintiffs to win by allowing these plaintiffs to recover. So I wonder if people think that Apple v. Pepper actually flipped it the other way. The swing difference actually is — and this is curious; I hadn’t thought about this — Kavanaugh replacing Kennedy, and Kavanaugh clerked for Kennedy. Justice Kennedy was in the majority in American Express and Kavanaugh now writes the majority.
Is Apple v. Pepper a more plaintiff-friendly rule than if they had decided it the other way then with Justice Gorsuch? I wonder if anyone has any views on that.
Boris Bershteyn:
I’m happy to start.
Harry First:
Despite your view of liking American Express, which is a horrible opinion, go ahead.
83 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
Boris Bershteyn:
In the interest of full disclosure, I represent defendants in other payment platform cases, so there are lots of reasons to like the American Express decision. And I have litigated against Michael Hausfeld a lot, so if he thinks it makes his job harder, from his lips to God’s ears. [Laughter]
But putting all that aside, I am going to push more on the other premise, which is that the Apple decision makes it easier for plaintiffs. I’m not sure that that’s true, and that’s for two reasons. One is that ultimately in each of these litigation settings somebody is a direct seller and somebody is an indirect seller. The question in Apple is really more about the choice of defendant than about whether a claim exists or not. In any particular litigation setting one defendant may be more attractive than another, but I don’t see any reason to presume that the Apple decision necessarily makes plaintiffs’ job more difficult.
The other litigation reality in the world of my practice is that when there is a federal law bar against claims by indirect purchasers, those indirect purchasers have relief under the state antitrust laws. In a technology platform setting — and sometimes in a financial products setting, where a lot of my practice is — it is hard to understand who an indirect purchaser is when you are talking about flow of financial instruments, for example.
But I sometimes practice in a very real bricks-and-mortar industry. There is a broiler chicken litigation in which I have a client, and you can understand in that case who a direct purchaser of chicken might be and who might be an indirect purchaser of chicken under some circumstances, although there are complexities there as well. In that case, as in many cases of its type, we are facing claims by multiple classes of indirect purchasers under state laws that allow indirect purchaser claims.
So, as far as plaintiff/defendant divide, I don’t really see Apple as a particular watershed.
David Higbee:
My comment is similar. I think of Apple as an Illinois Brick decision, and Justice Kavanaugh, if the case were to come back at a different stage, might see it differently on the substance.
Harry First:
What led me to ask that — actually there are two parts. One is whether the actual plaintiffs in that case are the more likely plaintiffs — this has been one of the problems with Illinois Brick — that in saying it’s only the direct purchaser and then the direct purchaser is in a business relationship with a price fixer, let’s say, they may lack incentives to sue, although there is the state part for the indirect purchasers. People are dealing with this every day. It’s not clear that anyone wants Illinois Brick reversed because everyone is dealing with it.
But in some ways the plaintiff in Apple v. Pepper, I think, was one who was more likely to sue — in fact, not only more likely to sue; they did sue. The developers didn’t sue. So if they are the ones who are getting harmed by it, they are not bringing suit because they are in a business relationship with Apple.
But the second part is this intriguing aspect in the Court’s opinion where the Court says, “Oh, actually developers can sue too because they may be being abused by the monopsonist.” I thought that’s sort of an interesting window opening up. Is the Court really serious about that? We’ll see that in the future.
84 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
But I agree that on the American Express case there is going to be a lot of gain about limiting it to whatever transactional platforms there are.
Again, in the sports panel someone referred to an NCAA case where the claim was being made by the defense that education is a multisided platform. I guess you have students, you have donors, you’ve got parents, you’ve got people who will bribe you to get in — I mean who knows? You’ve got to keep everyone happy. So the judge said, “Give me a break. Both of the economists can’t testify to this.”
But we’ll see how all of that — it’s a good thing there are no economists on this panel who could actually talk about what Jean-Charles Rochet and Jean Tirole intended when they wrote about what two-sided platforms were. It was not what Justice Thomas wrote, I don’t think.
We are of course running out of time, we are always running out of time, and there’s too much to talk about, but I do want to talk a little bit about the other side, not just where the law is but where it might go, and some of the proposals that are getting a lot of traction.
It’s true, and David made an interesting point, that people are arguing, “Well, I’ve got some far-out theory that can take care of everything, but we know the courts are never going to buy it. So, on the other hand, if you don’t like that, I’ve got a bill for you.” There are academic proposals as well as actual legislative proposals from Senator Warren to break up Facebook or Google or to actually restructure them.
I’m going to ask Martin for some sense of this from a jurisdiction not yet heard from, which is Australia, and News Corp’s view of breaking up Google as told to the Australian Competition and Consumer Commission (ACCC_. I think people would be interested in hearing a little bit about this. And remember, this is Rupert Murdoch, Wall Street Journal.
Martin d’Halluin:
You noticed I have an Australian accent, right? That’s why you asked me. [Laughter]
Harry First:
I knew my Australian was really good. Go ahead.
Martin d’Halluin:
First, I’d like to answer very quickly to your remark that we should see in the private litigation what are the conducts. I will say two things on the private litigation. I think the Apple v. Pepper case is interesting. One of the things we didn’t discuss is there was a question during the argument when Justice Alito asked, “How many app developers have sued Apple?” The Apple lawyer said, “Zero.” That was pretty much like if you want enforcement, you need to allow cases, and the developers are not going to sue Apple. So that’s one thing.
On the conduct, thinking about Ohio v. Amex, where I explained that I disagree, it’s interesting to me that we now have to look at effects on both sides but we are still only looking at conducts on one side. In a lot of these markets the participants on the platforms control both sides. Are we going to look at the conducts together in the aggregate or at the effects on both sides? It is very hard under current definitions of conducts, which are very well known by the platforms, to find a conduct on a market that has two sides that fall under the current definitions.
85 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
In Australia, for those of you who don’t know, the ACCC started an investigation on journalism and platforms. It has been going on for a year and a half now and the final report should be provided to the government at the end of June.
News Corp was very involved, being a big publisher in Australia. There were multiple phases. We made a submission in April 2018 in which we explained a lot of issues, a lot of conducts. It’s a public document. Then there was a preliminary report, and then even more companies and individuals reacted to that report. A lot of American companies and American practitioners honestly said, “This is crazy.”
I think it’s important to understand the Australian context, where it has been said many, many times that it is not only about antitrust but also about consumer protection.
Now, about the breakup of search, we explained in our submission March 2019 to ACCC two things, one thing that I truly endorse and one thing that I don’t think we have the antitrust conduct to fully endorse yet, which is a breakup of search. I don’t think the breakup of search, the way we explained it, is fully endorsed under our current antitrust laws.
The one thing that for me was very important was how do you deal with the ad tech integration. There is power in the entire open Web and it is currently being owned and controlled by Google. I would argue that there are really no more options for publishers to sell advertising online — meaning who is the provider of these tools — and I think that it should be investigated, and I think that’s what we tried to say in this ACCC submission.
Harry First:
So the remedy in the end would be to separate out this ad tech?
Martin d’Halluin:
Well, ad tech is very complicated. It’s much more complicated than my Facebook status.
Harry First:
You’d have to explain, yes.
Martin d’Halluin:
I think the first step is to try to understand what is going on. Once you understand what is going on, you have multiple layers and you could separate out certain layers that are currently being monopolized by Google.
Samantha Knox:
News Corp is also present in that ad tech stack too, right, and you are competing with Google?
Martin d’Halluin:
Not on the publisher’s side, no, we’re not. On the publisher’s side we’re not.
86 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
Samantha Knox:
But you provide network services and are competing for ad revenue with the platforms?
Martin d’Halluin:
We do compete for ad revenues. We try. By ad network I’m not sure what you’re referring to. We have a company that is on the buy side and that is trying to buy and sell advertising for advertisers, but it’s built on the ad tech stack owned by other operators. But I like that you defend Google.
Samantha Knox:
I think for context, for those who haven’t been following the Australian inquiry as closely, the recommendation that’s on the table from the ACCC is for direct regulatory oversight of news rankings on digital platforms that exceed a certain revenue threshold and ad rankings on platforms that exceed a certain revenue threshold.
Harry First:
Do you want to explain what that means, ad ranking and news ranking?
Samantha Knox:
Basically, how news is displayed on Facebook or Google or other platforms that hit that revenue threshold — it would probably capture some news websites as well — and, same thing, how advertisements are delivered and ranked.
I think the first question that we have to ask when we’re faced with a proposal that is as intrusive and kind of unprecedented as that is: Is this likely to solve the concerns that were set out at the beginning of the sector inquiry, which were concerns around the sustainability of journalism, business models for publishers, and ensuring that we all have access to high-quality journalism, which is something that I think we all have an interest in? That’s certainly something that Facebook cares about very much.
The other thing that I do want to come back to is something that has been referred to a few times today, which is this question of whether publishers are dependent on platforms like Facebook or Google or others. Martin, as you said, and I agree with you, the most important source of audience access, audience traffic, and opportunity for monetization are the publishers’ own owned and operated websites and apps.
Martin d’Halluin:
That’s not what I said.
Samantha Knox:
So you disagree with that? You don’t think that most of your traffic comes from direct traffic?
Martin d’Halluin:
No. I explained why direct traffic is very different from indirect traffic. They serve different purposes. People do not come directly to discover the content; they go through platforms these days to find news. If they hear something is happening, then go and Google it or they see it on Facebook and then they 87 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
click on the article. The people who go directly to the publishers’ sites are their loyal consumers, their loyal readers. Both sources of traffic serve very different purposes.
Samantha Knox:
The ACCC had a market definition that sounded a lot like that. The ACCC proposed a relevant market that was defined just around referral traffic from digital platforms to news publishers. Now, antitrust law 1.01 says that a relevant market has to include all acceptable substitutes. So whether we are talking about the ability to reach an audience, the ability to distribute your news, or the ability to monetize that news content, you can’t limit it just to referral traffic from third-party sites; you’ve got to also include all the avenues through which those readers are accessing your content. What the third-party market data shows is that, by far and away, the way that most people access news is through going to apps and websites directly. So I think we do agree on that.
Martin d’Halluin:
No, we don’t. But it’s fine. [Laughter]
Harry First:
This is good. I like this. Go ahead.
Samantha Knox:
My point is just that Facebook invests a lot and tries very hard to be a valuable partner to news publishers. Our goal is to provide value. Our goal is to give you tools that you can use to monetize. And look, it’s a free avenue for distributing your content. But Facebook is not an essential channel of access to readers. It is not a gateway to the Internet. It is not the primary place that people go to get news.
With all of that in mind, I think these are things we have to look at when we are asking whether that type of regulatory intervention is warranted and, again, likely to solve any of these larger concerns around the sustainability of journalism.
The last point that I’ll make here is that I think we should all be concerned about the future of journalism. The challenges, however, that publishers are facing started long before Google and Facebook. Publishers had declines in subscriptions and ad revenues dating back many years, starting as early as the 1990s, well before Facebook was founded and long before Facebook ever sold its first ad.
Harry First:
Well, I’m glad to know that most people aren’t getting their news through Facebook because I was worried about that. I don’t, but I don’t know.
The second thing — this is just more a teaser; I want to open it up to some questions — if under American Express American Express is a two-sided market, the platform is a market, is Facebook a market because it’s a platform? Is Amazon a market? Is Google a market? Is there a different sort of growth pattern out of American Express to draw on this idea that platforms are markets in themselves? If people do a lot of general search for products and they don’t actually start on Google, they start on Amazon, does that make it into a market? This is just a teaser for where maybe American Express might go.
88 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
With that, I’d like to open it up to questions from the floor. We have seven minutes and we’ll give you the world. Okay, start over there.
Question [Pallavi Guniganti]:
I would like to ask two questions. The first is one of the things that puzzled me a little bit with the Apple v. Pepper case was what seemed to be this idea that Apple wasn’t providing the services to the developers; it was providing service direct to the consumers, to app users. Just basically, could someone explain that to me, how it is that Apple isn’t actually providing services to developers?
The second question is with regard to the idea of the transaction platform. It seems that what they were trying to get at in Amex was the idea that if you’re actually paying money through it — if it’s an Uber or if it’s Amazon, those are things where you pay money on it — it is therefore similar to Amex and therefore it fits under this definition and everything else they are pushing off to another day.
But I guess that still leaves open the question that it seems like a lot of these platforms are trying to get more into an e-commerce market, people are trying to have more transactions happen over Facebook and other platforms, so I’m wondering about the sustainability of the definition of transaction platform.
Harry First:
Was there a question?
Questioner [Pallavi Guniganti]:
The question was asking in what sense if Apple isn’t providing services to developers how would this apply also to all of these other platforms that seem to be supplying services to people who want to sell new content through them; and then, second, similarly, the question of what constitutes a transaction platform?
Harry First:
Anyone want to respond? Go ahead, Martin.
Martin d’Halluin:
We have apps on the App Store like Facebook. I think it’s both. They do provide services to developers. That’s a reality. But I don’t understand why that would change the relation between consumers, iPhone users, and Apple. When they buy an app from Apple, as I said, the contract terms said that Apple is an “agent of the developer.” But the economic reality — and I think that’s what the Supreme Court was trying to look at — is that these users contract directly with Apple, they have a direct relation with Apple, and therefore they should have a claim.
One thing that is interesting and that was not discussed in the decision is that, more often than not, it’s how do you allocate the increased price and is it passed on to the indirect purchaser or not. In the context of digital, the marginal costs are so low that if you are going to be charging 30 percent more, basically that goes into a no-marginal-cost service provided by Apple.
There are some examples. Some companies — not News Corp — decided that if people want to buy a subscription on Apple, they are just going to pay directly. Spotify, when it was selling subscriptions on Apple, I think I recall, was just charging 30 percent more on Apple versus outside of Apple. That is just 89 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
showing you that the cost was completely passed on to the end-user just because there is really no marginal cost from Apple.
David Higbee:
I think they are providing services to the app developers. The developers are paying them 30 percent of their revenue for providing the platform and an audience of consumers. I think that reflects the value they are providing.
Questioner [Pallavi Guniganti]:
I just want to clarify that what I was seeing in the opinion about that was where Kavanaugh was saying that the developers could bring a monopsony case, which seemed to imply that he didn’t think that they were buying services; he thought they were selling.
Harry First:
We have time for two more questions, I think. Professor Fox, go ahead. I use that formal designation of Professor Fox.
Question [Eleanor Fox]:
I’ll call you by your first name. Thank you, Harry. [Laughter] NYU Law School, Eleanor Fox.
I want to ask a question that goes to your conversation about the alleged disconnect between the claim against bigness — “break them up because they’re big” — and the claims that would relate more to antitrust, which is “We want to control anticompetitive conduct.”
I want to propose that conduct is really at the heart of it. As Judge Wyzanski said years ago, “When firms get a lot of power, they usually use it.” We have seen through time that firms that have a lot of power and that have leverage to prevent competitors from competing against them do tend to use it.
There are a couple of incidents that have been publicized on Facebook — and they’re just allegation — for example, the Vine incident, where it is alleged that Facebook is using its power as gatekeeper and platform runner to disable those who use its platform from competing on neutral terms with Facebook. This is the kind of conduct that I think is coming out in the press and in the public dialogue, that gatekeepers who run the platform and compete against those who are on it are actually using conduct to disable them and, whether or not they are, that there should be an antitrust principle of neutral treatment.
Do you regard that as a possible link between claims of bigness, they ought to be broken up — and maybe these are not fully informed ideas about what a breakup will do and what it might cost — and the idea that there really is conduct at the heart of it that people are concerned about and should we be concerned about misuse of power to disable competitors?
Samantha Knox:
I think you and I agree that if there is conduct that is anticompetitive, exclusionary conduct that violates the antitrust laws, and, again as I said at the outset, is related to an increase in acquisition or maintenance of monopoly power, yes, that’s something that’s antitrust enforcers should go after.
90 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
I’m not going to comment on the specific references you made to allegations that I’m not familiar with with regard to Facebook, but we don’t disagree that antitrust enforcers should enforce the antitrust laws. Now, I don’t think that is the same as saying there ought to be platform neutrality. That’s a big difference, and I know that is a proposal that is on the table in Europe.
What does “neutrality” mean? “Fairness” is an empty vessel for whatever the particular decision maker would like it to mean. Some of these proposals are to have, for example, neutrality requirements or line- of-business restrictions on platform within the antitrust space. Why can’t the conduct that we are concerned about be reached under current existing antitrust laws? So far, I have not heard any specific conduct allegations that are not reachable under existing antitrust laws and, coming back to my earlier point, it is not clear why we would need special rules for digital platforms.
I also think on this kind of neutrality principle a lot of that discussion is following in the wake of the European Commission’s Google Shopping decision. I’ll just say decisions like that have created a lot of uncertainty, and it’s not clear what the guiding or limiting principles are there, and I’m not sure that that serves anyone well. Again, the idea that we need to have a fairness or a neutrality requirement instead of looking at exclusionary or anticompetitive conduct, I’m not sure that that’s the right way to go.
Harry First:
A final question because we’re running into Hal Varian’s concluding speech.
Question [Stephen Libowsky]:
Hi. Steve Libowsky from Dentons. It’s not technically an antitrust question, but I do think it affects the competitive nature. I’d love to hear the panelists’ comments on whether we have now reached a point where the immunity in the 1996 Communications Decency Act Section 230 has reached a time that it ought to be repealed and put digital platforms on the same footing as other businesses.
Harry First:
Do you just want to say what that immunity is for those who don’t know?
Questioner [Stephen Libowsky]:
Section 230 of the 1996 Communications Act essentially gives a pass to anything that an Internet service provider or website puts on its platform as basically on just the public forum, sort of the old public square where you tape up or nail up a piece of paper — “Hey, I’m not doing anything; I’m just putting it out there for people to look at” — and you make judgments on whether or not they are monkeying with it or not. It’s essentially a 1996 law that was meant to help foster the Internet.
Harry First:
So it’s here. Do you want to be neutral? Any reactions to Section 230?
Samantha Knox:
I’m not a Section 230 expert and I’m going to leave the debate over that to folks far more knowledgeable than I, but I do think there is one important point that is worth making here. A lot of the content that is shared on platforms like Facebook is user-generated. So while we are, yes, responsible for everything that is on our platform, and that is a responsibility we take seriously, I think everyone has seen what 91 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
happens when a mistake is made and that does not go well. That reflects very poorly on our brand and our company. Let me just underscore that that is a responsibility that we are taking very seriously, so our interests are aligned in making sure that there is not bad content on our platforms that is being shared and that our platforms are not being misused.
But there are important differences between platforms like Facebook and newspapers, where the editor has complete control over everything that’s published in the newspaper. I think the differences between these business models are important questions to keep in mind when we are looking at this.
Questioner [Stephen Libowsky]:
Just a quick follow-up. For newspapers, both online and on paper, the rules are different.
Samantha Knox:
Right, but even a newspaper website online has complete editorial control over what is published on that, which articles appear, which op-eds appear.
Questioner [Stephen Libowsky]:
That’s my precise point. They have an immunity if they put it online and they don’t if they publish it on paper.
Samantha Knox:
I’m not sure that that’s correct.
Martin d’Halluin:
I’m not sure that’s correct.
Harry First:
None of us are sure. On that sureness point, thanks to the panel.
Closing Keynote Speech: Hal Varian “The Seven Deadly Sins of Tech?”
Lawrence White:
We are now going to clear the platform here, to use a phrase, and you are going to see that Harry First and I are from different schools — one, different schools at NYU, the Business School and the Law School, but also different schools of introduction of speakers. Hal is someone who is —
Harry First:
That’s Hal. 92 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
Lawrence White:
That’s all you get, Hal. [Laughter]
Hal Varian:
If you don’t know who I am, why are you here? [Laughter]
Lawrence White:
Exactly right. Hal is the Chief Economist at Google, and he knows he is standing between us and the liquid refreshments. Hal, take it away.
Hal Varian:
Yes, I do. Thank you very much for that non-introduction. [Laughter] When I was struggling to find a title for this talk, I came up with this one, but I figured I should look up the real seven deadly sins. They are pride, greed, lust, envy, gluttony, wrath, and sloth. I figured, Wow, talking about them in New York is like taking coals to Newcastle. These are the views of the author and don’t necessarily represent the views of the employer or any other party. But of course, I don’t believe that there is that much sin in tech. What I want to look at is what are referred to as potential problems in my view are not nearly the problems that they are purported to be. Here are the alleged seven deadly sins: excessive concentration; too little competition; not enough innovation; too many acquisitions; too little entry; high level of lock-in; and of course, some sort of fixed costs that serve as a barrier to entry. That is a list of things that tech has been accused of. I’m going to go through these one by one.
Concentration If you look at concentration, David Autor and colleagues have written a very nice article looking at two interpretations of concentration. One is: are you seeing “superstar firms with higher productivity increasingly capture a larger slice of the market?” With capitalism that’s what markets are supposed to do — the most productive, most efficient firms capture share from the larger slice of the market — and you would see in that world, of course, increasing productivity correlated with this high market share. And indeed, that’s what you do see, as Autor describes.
93 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
On the other hand, an alternative description would be they “arise from anticompetitive forces whereby dominant firms are able to prevent actual and potential rivals from entering and expanding,” and of course that would be a problem under antitrust laws if that second one was their issue. What they conclude from their empirical examination is that industries that became more concentrated were those in which productivity increased the most. Now, I want to emphasize that’s correlation, that’s not necessarily drawing a causal picture, but there have been related findings by Ganapati and Bessen and the Organisation for Economic Cooperation and Development and many other people who are looking at this phenomenon. So I think we are understanding more and more the relationship between productivity, competition, and dominance, but we haven’t gotten a definitive conclusion that has come out of the economic literature at this point. Competition People say, “Where is the competition in search.” I say, “Well, follow the money.” General-purpose search, what Google does and Bing does, is a really tough business because you have to build a big infrastructure to answer 100 percent of the searches but it only makes money from 6 percent of them because only 6 percent of the clicks actually come from ad clicks or commercial clicks. So there is intense competition for those commercial clicks. Look at the general search engine business — AOL, Ask Jeeves, Yahoo!, Inktomi, Excite, Lycos. Those companies were all around twenty years ago, but they have all disappeared because it is pretty hard to pull that off. The competition is currently very intense for these commercial clicks. Look at Amazon; they want people to come directly to Amazon, not come to it from clicking on an ad from Google or Bing. Or eBay, same thing; or Yelp!, Travelocity, Expedia, Orbitz, Trip Advisor, on and on and on — they are after those commercial clicks. They want people to use their systems to bring people to their websites. They invest in building a brand in order to do exactly that. But nobody really cares about the non-commercial clicks — that is, from an antitrust point of view — so you look at book search or scholar search or patent search or encyclopedia search. All of these services are provided for free. Would you say Wikipedia is dominant? I wouldn’t, but some people might because they serve a large fraction of the searches having to do with encyclopedia-like topics. Why is the competition intense for these commercial clicks? Well, just what I said, they want users to go directly to them. Now 54 percent of shopping journeys in the United States start on Amazon, so that’s a good thing for Amazon. How did they get into that position? They invested very heavily in building a brand and a reputation for high-quality service. You advertise now, you advertise early on, because you don’t want to advertise as much in the future. Look at Amazon. You might be old enough to remember, when I look around the room, that Amazon advertised very heavily in brand advertising to get its name out there to let people know what they’re doing. The same thing happened with Netflix. Netflix is a common brand today, but that didn’t just happen; they advertised very heavily to let their brand get out. Same with Groupon and MoneySuperMarket in the United Kingdom and many, many other companies of that sort. 94 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
Now what has happened with apps is the world has changed quite a bit because people who invest in apps that go only to a single destination — for example, Yelp gets 35 million monthly unique visitors just by its mobile app — so it is much more like the offline economy where you say, “Well, I’m just going to do all my travel through Travelocity; I’m going to download their app and then take it from there.” It’s much more of a commitment to use a single provider than the multiple providers you get from using the Web. Google’s view has always been if we provide really good answers for “ancient history” or “who was Thucydides?” or something like that, people are going to use us to do commercial searches as well so we get that 6 percent of the clicks and use that money to fund everything else.
And in fact, when you look cross the Big Five here — Amazon, Apple, Google, Facebook, and Microsoft — they compete intensely with each other across a variety of domains — advertising, platforms (we’ve heard a lot about that), artificial intelligence, browsers, cloud services, digital assistants, e-books, email, games, on and on and on — and you see very heavy competition. Just as an example, look at the Amazon Echo. When it came out, people said, “What is this thing?” They said, “Oh, it’s a smart speaker; you can talk to it; it answers back.” Nine months later Google’s announcement came out about our smart home speaker and a few months after that it was available for purchase. And now, as you know, there is quite a bit of competition going on in these two different devices. Innovation
95 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
Here’s a chart from Bloomberg on research and development spending by billions. I’ve colored the GAFAM, as Europeans call it, (Google, Amazon, Facebook, Apple, and Microsoft) in red; green is auto; light blue is pharma; and then purple is tech other than the Big Five. You can see there is a very substantial expenditure in R&D. There’s a little bit of a controversy here between whether Alphabet or Amazon should be in the top spot because on our Form 10-K we split out R&D as a separate category and Amazon doesn’t. But in any event, I think we can agree they are certainly the top two.
In fact, if you consider these other tech firms — like Cisco, Oracle, Huawei, Intel, etc. — then tech has even a bigger impact on R&D. So there’s a huge amount of R&D that is going on, even completely trumping the pharma companies, which is kind of surprising. Another surprising thing, to me at least, was how much the auto companies are investing in R&D because they face this existential threat of electric cars on the one hand and then autonomous vehicles becoming real. Acquisitions We’ve heard a lot about acquisitions. The term of art I’m told is “Hoover-ing up,” which always strikes me as funny because I grew up about twenty miles from Canton, Ohio, which was the corporate home of Hoover, and we never said “Hoover-ing up.” I don’t know why they do.
96 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
The median number of hires per acquisitions by Google is six, very, very small; 25 percent of the hires had three or fewer employers; 75 percent had eighteen or fewer. They are what we call “acqui-hires.” Acqui-hires is where you are acquiring a company not so much because of its product but because of its team. Maybe it relates to a product they’ve got; maybe you want to build a new product. I’m not saying the product is totally irrelevant, but the real reason for acquiring these small companies is because you want their employees. It’s a heck of a lot easier, believe it or not, to acquire a firm of five engineers than it is to hire five engineers. This acqui-hires link resolves to the Wikipedia article on acqui-hires, which also I’ll say features Facebook quite prominently because Facebook is engaged in exactly this kind of strategy, as is every large firm in Silicon Valley. The competition is really intense, and many of those acquisitions are really just hires.
There are five times as many exits by acquisitions than by IPOs. We think of the IPO as being the goal. No, that’s wrong. It’s an unrealistic goal in most cases. What happens is the goal generally is to be an acquisition. If you get lucky, it can be an IPO; but that’s the exception, not the norm. If you go ask the entrepreneurs — this is from a survey that Silicon Valley Bank puts out — “What is the realistic long-term goal for your company?” half of them say it’s to be acquired and only 18 percent say it’s going to be an IPO. Some of them think they might stay private, some of them think they might get money from other sources, but generally the norm is an acquisition because venture capital overall might have a 70 percent success rate but only 20 percent of the companies that are out there have successful IPOs. It’s very unusual really. And the same thing is true — in fact even more true — in Europe. Entry Where is the kill zone? I think the kill zone was mentioned once today, maybe a couple of times. This is an area that’s not worth operating or investing in since defeat is guaranteed. Let’s take an area like artificial intelligence, a hot topic of the day. Google’s doing it; Apple’s doing it; Amazon; Microsoft; Facebook. China is doing artificial intelligence, very gung-ho. In fact every country in Europe seems to have announced a white paper on artificial intelligence and why this is going to be a great investment in the future. 97 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
So if you’re a young startup, you’d be crazy to go into artificial intelligence, right, because everybody else is doing it and defeat is guaranteed?
Well, no. That’s exactly wrong. In fact, 630 small venture-funded companies are describing their strategy for AI. Obviously, they aren’t all going to have an IPO and become the AI company. The likely outcome for many of these is in fact to be acquired either for their product or for their talent because both are potentially much in demand. So, the kill zone, where is it?
In fact, when you look at financing of U.S. startups — this is a nice plot from VentureSource, which is the Dow Jones database of venture-funded companies — the dark blue line is the initial round of funding and the light blue line is subsequent rounds of funding (those are measured in dollar terms) and the black line is in fact the number of firms funded. You can see the big 2000 dot-com bubble is quite apparent on the graph, you look at the 2009 recession and you see a little tiny drop there, but in general there has been a monotone increase in all three dimensions.
98 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
In fact, if you look at Europe, it’s pretty much the same thing. You can see on this graph and on the previous graph that 2018 is the first year we see that total funding exceeded the amount during the dot-com bubble, which burst in late 2000. So there’s plenty of startup activity here in both places. Lock-in Let’s look at lock-in. What do switching costs look like, let’s say, for Google? In June 2011 we announced Google Takeout, which allows you to take out your data. You can download it to your desktop computer; you can download it to Gdrive, our cloud-based file system; you can download it to OneDrive, which is what Microsoft uses; you can send it to Dropbox as well. You can take your data out whenever you feel like it. If you look at this chart, it shows you Blogger, Buzz, Contacts, Drive, Latitude, Pages, Picasa Web Albums. Anything you want you can download, including your web history. There are more than fifty of these different databases that you can download. And Facebook has something analogous to what I just described. What we ran into with Google Takeout is we provided all the capability to take out but nobody really provided the capability to take in because it was a big hassle; and there weren’t that many people that used Takeout to begin with, but we felt it was important to maintain nevertheless. So in 2018 we started Data Transfer Project that will allow you to seamlessly transfer your data between Google, Facebook, Microsoft, Twitter — those are the founding partners — and anybody else. It is an open interface and it is all done through application programming interfaces (APIs). All you have to do is just pick a common protocol to download the data in and then you can upload it and convert it to whatever protocol is used by the recipient party. That is available now.
99 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
Also, if we look at research, I don’t have enough room to really list all of the things that Google has donated for research purposes, but I highlighted a couple. OpenImage is 9.5 million labeled images with, I think, 1000 potential objects per page. These images were labeled by people from the Philippines or people from India or people from the Mechanical Turk, wherever, a variety of places. It costs real money to get those images, and that was donated to the artificial intelligence community for research. YouTube Video: Even though machines can perform phenomenally well at image recognition, they are still pretty bad at verbs. That is, you can find the nouns, what objects are in the picture, but finding the verbs is still quite hard because you see a picture: Are they dancing? Are they fighting? Are they exercising? Are they marching? All of these different things. We can usually tell pretty quickly what these people are doing or what this animal is doing, but it is a challenge for AI. We compiled 8 million labeled videos from YouTube to work on that particular challenge. Fixed Costs In 1999 if you wanted to build a web application, you would have needed a data center, hardware, software, networking. You would have had to go out and build all these things or find them and lease them. The software tended to be proprietary, internal to the data center operator, and needed to be redeveloped for anybody who wanted to use it. We had, for example, Map/Reduce, which is our way of managing huge volumes of data as a database. Hadoop is something that Yahoo! put together because they saw that Google had it and they said, “We need that, too,” so they built an instance of it based on the research reports that were published by the Map/Reduce people. BigTable, BigQuery, lots of different things like that — there was a lot of duplication of effort. But now all these tools are open sourced. It’s phenomenal. Everything you need is open sourced and is standardized. It is available at the data centers built by Google, Amazon, Microsoft, and so on, and everybody has access to technology that only a few companies could afford a decade ago. I mentioned image recognition. You can now go to cloud.google.com/images and you drop a picture in and it will tell you the items that it recognizes in that picture. This was actually impossible five years ago, to do at scale at least, and only the richest companies could afford it. Now it costs a tenth of a cent per image to do that. If you go look at the prices — I have another slide that has the prices for Google, prices for Amazon, prices for Microsoft — they have all been forced down to that tenth of a cent, so anybody who wants to deal with images can now get that basic building block immediately for a very, very low price. In fact, anybody can offer a new search engine. Now that’s a provocative statement, but let me show you. Let’s look at the data. Look at DuckDuckGo and Qwant — maybe not household words. DuckDuckGo uses 400 sources on the Web and syndicates their search results from Bing and Yahoo! When you type a query into it, it goes off to Bing, it goes off to Yahoo!, a few other places, depending on the nature of the query, and comes back with the answers. It’s got 45 million searches per month and has been profitable since 2014.
100 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
There is now a kind of wholesaler/retailer split in the industry. You’ve got the wholesalers who are doing this and then you can syndicate it to the retailers. The retailers use the search capability there and then they can add other bells and whistles to make it more attractive. In fact, you can do this. Anybody in the audience can put a Google search box on their homepage and Google gives you 51 percent of any revenue that is generated by the people who go to their results via that search page. That’s kind of interesting. We have been doing that since the beginning, since basically 2002. Qwant does the same thing. It has 50 million unique queries per month, bigger than the population of California. It is now the official search engine for the French National Assembly and the French Army. In fact, they demand that you use Qwant in those places. What has happened is the fixed costs — all that building the data centers and populating the machines and designing the software and getting the networks in place — that’s gone. That’s no longer a fixed cost. That’s a variable cost because all of these capabilities are sold on a per-use basis. So a very small company can start out with an idea, run it in the data centers, use that capability from any of those companies, or all of them for that matter, and then eliminate their fixed cost. Those are fixed costs on the supply side, on the production side. But you still have the fixed costs in terms of marketing. I said anybody can build a search engine, but the problem is getting them to come. It’s like Shakespeare: Glendower: I can call spirits from the vasty deep. Hotspur: Why, so can I, or so can any man; But will they come when you do call for them? [Henry IV, Part 1, Act 3 Scene 1] Good line for customers. [Laughter] Now what happens is to attract customers you have to differentiate your product in some way to make people come to your product rather than these other products that are out there, including the products provided by the wholesalers. In the case of DuckDuckGo and Qwant, they both emphasize very heavily the privacy aspects. There are no logs kept; there’s no information that’s stored about the users; they use Tor for some encrypted communication, or at least Qwant does; and they’ve got other things of this sort so that a particular audience finds this an attractive package. Maybe the search isn’t quite as good, but they are willing to sacrifice that for having this other capability. That’s how it goes in today’s world. And it’s not just the production side, the data center side, the cloud computing. You can start up a startup and outsource almost all of your business processes. This is what’s really phenomenal. You can fund your project on Kickstarter, hire employees using LinkedIn, get your office space from WeWork; cloud computing; use open-source software; use GitHub, Skype, Hangouts, Google Docs — all these things you had to get a license for before are now essentially free to the users. You can even use customer support by ZenDesk. You say, “I’m a small company and I can’t afford to set up a whole department for customer support.” Well, you outsource it. ZenDesk started in Copenhagen and has offices in San Francisco and Manila because it has all those call centers and support centers in Manila. 101 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
So, what was again complicated and costly to do twenty years ago is now something you can do with six employees or five employees or even two employees. Almost every small company South of Market (SoMa), which is where they are all clustered, basically utilizes these and other services. Oh, and they get their lunches catered too. So, same thing. Okay. We’ve got two minutes until drink time and we’ve got some time for questions. Thank you.
Lawrence White:
I can’t resist again the two schools. There’s the Hal Varian school, which is it used to be “the marketing, my expense thing,” but there’s also the Kevin Costner school “Build it and they will come.”
Hal Varian:
Good luck with that.
Lawrence White:
We do have a little bit of time for questions. Again, identify yourself briefly, brief question, and please no speeches. Andy?
Question [Andrew Zimbalist]:
Andy Zimbalist, Smith College. You put up a slide where you showed a large share of acquisitions were motivated by the desire to acquire talent. I wonder, first of all, whether the noncompete prohibition in California is behind that; and, when companies are acquiring the talent, are they acquiring the talent for the human capital independent of the proprietary information that those people embody, or is proprietary information one of the things that they’re after?
Hal Varian:
As you know, it depends on where they acquire them from. If we acquire a person from Microsoft, then they can dictate some terms about what that person can work on. If it goes in the other direction, we don’t have the ability to do that.
The question of what you are acquiring — the answer is, it’s going to depend. If you are doing something where you really need depth of knowledge in some specific area, you are going to acquire people in that area.
One example is voice recognition. We had nothing in voice recognition at one point. We saw mobile was coming. We went out and acquired the best people in voice recognition in the world and then they got together and tried to figure out how to get data. What they did is they created this thing called GOOG-411, which is where you said, “I want the number of Joe’s Pizza on University Avenue,” and they would say, “Do you mean John’s Pizza on University Court?” “No, I mean Joe’s Pizza.” After you do that, you’ve generated the training data for these people to apply their expertise. That was started from scratch.
The same thing with autonomous vehicles. Same thing with maps. In lots of cases you’re doing it where you want some skill in a given area. In other cases you just want a software engineer who could plug into a variety of areas. So there isn’t a single answer.
102 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
Question [Philippe Burke]:
Philippe Burke with Apache Capital. Thanks for a great presentation. I think most of us don’t so much associate the value of research with the value of the commercial place before you could then package the information and sell it to advertisers. Could you give us an idea of the relative revenue between these two activities at Google or at other firms to your knowledge? Is that a fair comment to make, first of all?
Hal Varian:
I’m not going to give you the fish; I’m going to teach you how to fish. Look at our Form 10-K, which shows there are two different sources of advertising revenue. One of them has to do with Google properties, and that would be Google Search, and the other has to do with the third-party properties, external properties, the assets, and so on.
What happens there is of that ad revenue in the second category only 30 percent of it goes to Google; 70 percent of it goes to the publisher who is publishing that content, which includes newspapers and everybody else who uses those third-party ads on the Web.
The first part is where the bulk of the money comes from — that is, from the search advertising — and that does not use personal information to any extent. What it uses is the advertiser chooses a keyword like “frying pan,” the user chooses a query like “frying pan,” and if the query matches the keyword, then the advertiser is eligibly shown as an ad. That’s what it is driven by. The prime thing is the keywords chosen by the advertiser and the queries chosen by the user. There is a little bit of personalization that goes on. People can do demographic filtering because they are advertising cosmetics. The whole question is, does the query meet the keyword?
That’s the primary source, and that is by far the largest source of income. As I say, look at the 10-K. You can figure it out.
Question:
Search ads, as you mentioned, is the largest source of revenue right now. Last year Amazon launched a similar product on their platform. So I was just wondering, what is Google’s plan in order to mitigate the movement from Google to Amazon right now?
Hal Varian:
Let me say the reason why search advertising is so useful is you are showing people the ad exactly when they are issuing the search. You want to get a new bicycle helmet, you go to Google, you say “bicycle helmet,” and you’ll get a list of organic results. You’ll also get a list of ads. If you click on one of those ads, then in fact some revenue goes from the advertiser to Google and the user goes to the bicycle shop where they can buy it. It’s like double-entry bookkeeping: we measure the clicks we send to you; you measure the clicks you receive from us.
That is what everybody wants. They want you to go directly to them and they’ll fulfill your needs, whether it’s Travelocity or whether it’s Amazon or whether it’s eBay or any of these other people. Indeed, they get a lot of direct navigation just because they have invested in building up those brands. The direct navigation they get is significantly larger than the ad clicks they get in all of these cases.
103 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.
What we are going to do about it is just what we are going to do about all those places, try to provide a better service than our competitor is providing in terms of getting people to where they really want to be.
Lawrence White:
Okay, last question.
Question:
I think your counterargument basically is you said that your contract is quite clear. I think it is fair to say your platform is quite open. But one thing you are doing is the Data Transfer Project. The outcome is still not clear from the consumer’s perspective, whether it is really transferrable or not; maybe the usability of the consumer is not clear. That is one point.
Another thing is, even if your platform is quite open and you provide accessibility for your competitors, you have a large data set that the other competitors can have access to, but I think you would have the right or power to design the platform itself so maybe you can utilize it more than your competitors. If you have a huge advantage of that, how do you answer or respond to criticisms about Google’s network effects?
Hal Varian: Well, if you ask do we think we know more about our platform than our competitors do, yes I think we do, and I think they even know about their platform more than we do. So that’s not too surprising.
What was the first part of the question?
Questioner:
About Data Transfer Project.
Hal Varian:
Oh, yes, the Data Transfer Project, right. Well, the trouble with the Data Transfer Project that we are going to see down the road is running into the privacy issue because I can send my data to someone else, but can I send your data to someone else? I have a lot of emails from you, for example, or maybe I have photographs of you, and you’re perfectly fine with it sitting on Google Photos, but for some reason you don’t want it to go to some different provider. So there is going to be a whole tension here developing over data transfer for competitive reasons and then data transfer running into privacy problems. Next year that can be a topic of discussion.
Lawrence White:
There we go. All right. Please join me in thanking Hal.
Ladies and gentlemen, I know it has been a long day, but I hope you will agree it has been a fruitful day. There are still a number of the participants in the audience here and I hope everyone will join me in thanking them for their participation in this conference.
As I mentioned before, we have liquid refreshments and other kinds of refreshments on the fifth floor. Please join us up on the fifth floor. 104 The Global Antitrust Economics Conference 31 May 2019 – New York University Stern School of Business Verbatim transcript of oral presentations provided by Concurrences without prior vetting by the speakers.