Deep Dive Episode 144 – United States v. Google: Examining the Historic Antitrust Case Against Big Tech
On October 20, 2020, the Department of Justice (DOJ) filed its much-anticipated lawsuit against Google. The case is the most high-profile antitrust challenge since the Microsoft case more than 20 years ago. The Justice Department has alleged that Google monopolized the search and search advertising markets, inhibiting rivals such as Bing, DuckDuckGo, and Yahoo from succeeding and thereby ultimately harming competition and consumers. This distinguished panel debated the merits of the DOJ’s antitrust claims, discussed the potential parallels to the Microsoft action, and opined on the government’s likelihood of success at trial.
Jack Derwin: On October 30, 2020, The Federalist Society hosted a webinar event: The United States Department of Justice’s antitrust case against Google. The following is the audio of that event. We hope you enjoy.
Nick Marr: Welcome, all, to this special Federalist Society virtual panel debate. And, on today, October 30, 2020, we’re discussing United States v. Google. And we’re “Examining the Historic Antitrust Case against Big Tech.” I’m Nick Marr. I’m assistant director of Practice Groups at The Federalist Society. As always, please note that expressions of opinion on today’s call are those of our experts.
We have a distinguished panel with us here today. And I’m just going to introduce our moderator, Brianna Hills. She’s an associate at Boies Schiller Flexner LLP, and she’ll be moderating a discussion between our panelists, and, first, she’ll introduce them and a bit about the case. So, Brianna, thanks for being with us here today. The floor is yours.
Brianna S. Hills: Thanks, Nick. Like Nick mentioned, we’re here to talk about the Google complaint filed last week, which alleges that Google violated Section 2 of the Sherman Act, which many of you will know prohibits monopolizing or attempting to monopolize a market or an industry.
Basically, Google is a accused of monopolizing the search and the search advertising markets by, among other things, making deals with Apple and other original equipment manufacturers to make Google their default search engine, which the government says inhibits Google’s rivals, such as Bing, and DuckDuckGo, and Yahoo, thereby, according to the government, harming competition and consumers.
So I’m joined today, like Nick mentioned, by a fantastic panel of antitrust experts who will have a lot to say about this case. First is Geoff Manne, who is the President and Founder of the International Center for Law & Economics. He’s also a distinguished fellow at Northwestern University Center on law, business, and economics and is just an all-around noted antitrust expert on tech policy and other antitrust issues related to big tech.
Next is Professor Doug Melamed, who is a Professor of the Practice of Law at Stanford Law School. He also served, as relevant today, in the U.S. Department of Justice as Acting Assistant Attorney General in charge of the Antitrust Division, including during the DOJ’s historic case against Microsoft 20 years ago.
And finally, we have Professor Chris Sagers, who is the James A. Thomas Distinguished Professor of Law at Cleveland State University School of Law, where he specializes in antitrust. Professor Sagers has written extensively on antitrust issues and big tech, including his book United States v. Apple: A Competition in America. So thank you all for being here today. I’m looking forward to this conversation.
First, I just wanted to kick off the discussion today with Microsoft, which many commentators have mentioned the corollaries between that case and this case, both in terms of the historic nature of the two cases and on the merits. So, generally, why do you think there have been so many people comparing the two cases and what have learned from Microsoft that might be relevant to this case?
Douglas Melamed: Well, why don’t I start since you used the word Microsoft in introducing me. I think the cases are very parallel. Let me just be clear, I think the U.S. v. Google complaint is a very conservative complaint. It doesn’t break any new legal ground of any consequence. It’s a fairly traditional lawsuit challenging restriction on distribution that disadvantages rivals and insofar as it involves an existing monopoly allegedly using those restrictions to maintain its monopoly insulated from new competition.
It is very analogous to the Microsoft case, which was more complex. But the core of the Microsoft case were agreements between Microsoft, and various distribution channels, ICPs, IAPs, and principally OEMs that disadvantaged, in that case, really a complement or two or a potential entrant itself, complementor, the potential entrants, or perhaps itself a potential entrant. So I think it’s very analogous to the Microsoft case.
But I think you have to understand—and this is why I call this a conservative case—when the Microsoft complaint was filed in 1998, much of what we take for granted today was hotly contested. There was tremendous dispute around the Microsoft case as to whether network effects were of any consequence, as to whether the antitrust laws could effectively be applied where a product, in that case browsers, were given away for zero monetary price, where intellectual property, and innovation, and technology were involved with, allegedly—and certainly, Microsoft argued—a wider birth, greater freedom from antitrust restrictions, and so on. So I think the Google case is building on or relying on really a lot that was established for the first time in the Microsoft case and is itself a very conservative, in some ways almost replica, of that case.
Geoffrey A. Manne: I would suggest that — I think it’s true that the complaint is clearly intended to hue as closely as possible to Microsoft. I think it should be noted, the part of the Microsoft case that it really parallels is not the headline part of the Microsoft case. It’s the part of the case dealing with restraints on distribution of Netscape to internet access providers, independent software vendors, and the like.
We could talk about whether the other elements of the case, even though they don’t exactly mirror the structure of the issues involved here is nevertheless relevant, but that’s the most obviously relevant piece. But I think there are — I guess, talking about Microsoft, one of the first things I would suggest is maybe what’s changed since then, as Doug suggested, but some of the changes are not going to be in the DOJ’s favor.
One of the big things, I think, that’s changed is the nature of competition in these broader markets. At the time of the Microsoft case, whether they were correct or not—and some of have suggested that we weren’t—the court basically tossed out the prospect of real competition in the operating system space between Microsoft, and Linux, and Apple. Microsoft was a kind of pure dominant player there, and while that arguably could be true of Google, in say the search space, currently, it’s not true of Google, especially, with respect to the other players involved here, Microsoft and Apple, in the broader space in which they’re competing.
And I think it’s important to think about the contracts at issue here relating to search as being part of these companies’ broader strategies and broader competition largely with each other, but not exclusively, along other dimensions as well. We can talk about some of this later, but I just want to point that out. That’s particularly relevant, I think, with respect to the Android related aspects of this complaint.
I think another big difference that’s going to come into play is the fact that Microsoft was primarily, if not exclusively, about nascent competitors rather than existing competitors. Here, to the extent, we’re — to simplify, we’re talking about Google search versus Bing. Bing exists and has existed for a long time, whereas, in the Microsoft case, the argument—Netscape existed, of course—but the argument was about it competing in the middleware — competing in the operating system market, which is a nascent competition that may or may not have come about. We can talk about how, and why, and whether that’s actually relevant.
And then one final thing is just the prospect of procompetitive justification here. So the Microsoft case far from excluding the prospect of procompetitive justification makes very clear that that’s an important part of the analysis. It just didn’t find it anywhere. I think, in part, for reasons that I said at the outset, because of this broader competition, there’s going to be a lot more scope for possible procompetitive justification here, whether it’s sufficient or not, of course, is a different matter. But, in that sense, the course of litigation here isn’t likely to exactly track the Microsoft opinion.
Douglas Melamed: Geoff, can I ask you a quick question? Chris, I don’t mean to cut you off, but —
Christopher L. Sagers: It’s okay.
Douglas Melamed: — when you talk about broader competition, at the end you said, well, it goes to justification. I understand that. But are you also saying that, maybe, the government is wrong to think of specific markets in specific domains of competition such as search?
Geoffrey A. Manne: It’s certainly not doctrinally wrong, of course, and to the extent that it might be wrong in the sense that it might get us wrong answers from a social welfare perspective, yeah, I have papers, and I’ve made arguments before that—and especially in the context of these platforms—the way we do market definition can be too constraining and maybe problematic.
But, of course, it’s not doctrinally wrong, and that’s why I think the way in which that competition is going to most matter is probably going to be in the pro-competitive justification, but I certainly think that the DOJ should lay out relevant markets and will be able to make a case there that will survive — sorry. For excluding this broader competition, that will not tank any case the DOJ would bring.
Christopher L. Sagers: I’d just like to point out that somebody in the comments literally just said, “Let Professor Sagers talk.” I was trying to talk before, and I did the boomer trick. I had the mic off, and I was talking, and I’ll probably do that about five more times.
That’s pretty good stuff, and we were talking before this panel, just before we started, about how much we’re actually going to disagree with each other, but it sounds like we’re probably going to agree a lot about a lot of stuff. I definitely like the first thing that Doug started off with. Part of the apparent political backstory here is that some people think this case was a little bit rushed out of DOJ for possibly for political reasons, and that suggests to me why it feels like it’s — virtually, it’s Microsoft 2.0.
Interestingly, I’ve heard some people say that it really isn’t. Some people are claiming that there are actually major differences. I’ve also heard some smart economists saying that it shouldn’t be. The allegations shouldn’t have so closely followed Microsoft because the economics are in some important way different. I don’t know. I like the complaint.
And I agree with Doug, stuff that 20 years ago was groundbreaking doesn’t seem like it now, and primarily, because the government — though a lot of people felt like the remedy was not desirable, the government on the substance of merits won an overwhelming victory before an en banc panel of the D.C. Circuit that made important law. The Microsoft opinion is as important in monopolization law as many Supreme Court opinions. And, with that opinion in hand, it’s hard to doubt that this case, at least, has legs to stand on.
I would like to mention one thing that I don’t think Doug and Geoff have said, which is about another important similarity between the two cases that—I don’t know—I think it’s important. In my humble opinion, a truly groundbreaking idea, in the Microsoft en banc opinion, was acceptance of the idea of the government’s theory of Microsoft’s market power and, in particular, why Microsoft’s large market share in operating systems was entry protected, even though a lot of people at that time were saying that it wasn’t really competition. It was at bay only because Windows was a better system, whatever.
There was this model, that now, I think we might more familiarly call something like a platform economics idea. There’s this idea of mutual indirect network effects going on that, at the time, made Windows a very difficult product to challenge. We just didn’t have the same vocabulary for talking about it then, or, at least, we didn’t commonly. That same idea it seems, to me, runs throughout the DOJ’s complaint.
It is very important for Google to use those same techniques to foreclose distribution for other competitors to protect its position and that is because unless other firms can get in and start getting some scale and start attracting developers and hardware developers to adopt competing search products, then Google’s search position is jeopardized. And it feels, to me, like an economic theory very similar to Microsoft. So I’m glad they alleged it that way.
I had one other thing to say, which I’ve forgotten, but I think it’s like Microsoft, and it’s good that it’s like Microsoft, and I’m glad. I wish we had somebody on the panel who disagreed because I’m not exactly sure I understand the arguments I’ve heard that it’s different.
Geoffrey A. Manne: Can I ask you a question, Chris? Brianna, do you mind if we ask questions instead? I feel like I’m taking your job.
Brianna S. Hills: Go ahead. No. No. This is great. Go ahead.
Geoffrey A. Manne: But, Chris, I guess, I don’t disagree with that, but I don’t think it’s obvious how — it sounded like you were making a reference to the applications barrier to entry aspect of Microsoft.
Christopher L. Sagers: Yes. Definitely.
Geoffrey A. Manne: And I’m not sure it’s obvious how that parallel looks like here, in the sense that it seems obvious, to me, that it is not precisely the same. The barrier to entry is not developers developing for Google’s platform or Microsoft’s platform. And that’s partly a function of the fact that this case really resembles the AOL section in the Microsoft case and not the bulk of it.
So I do think there’s a case to be made here, but I’m curious how important you think it is that whether or not there is something analogous to an applications barrier to entry here and what you think that is in the case.
Christopher L. Sagers: Yeah. That’s a good question. So I have two thoughts on that. First of all, a problem, for me, in meeting the DOJ’s complaint is that I think it’s scattered. It does feel, to me, like it was a little bit rushed out. In some ways, they buried the lead in that complaint, and it’s a little hard to know exactly what it’s major themes are.
So I do think that there are scattered allegations throughout the complaint that resemble the applications barrier to entry. We’re told that in various ways, Google needs, not only to have preinstallation of its engine in products, and browsers, and so on, but it also needs to have its suite of supporting applications that all interface with its search engine, and it needs to have them in place so that everybody who wants to keep Google and not consider other alternatives.
But my second thought is I actually agree with you. To the extent that the government wants to make an argument like that, that this sort of mutual indirect network problem, right — which is how I think of the applications barrier to entry. To the extent that government wants to make that the theory of entry barrier, it’s a little bit confused by the fact that the government also makes a very different, big argument for barriers, which is the cost of developing your own web index, right—the cost of having a big web probably going, and anybody challenging Google. And, among us, I think if the government wins this case, it’s going to win it on that claim, basically.
Douglas Melamed: I don’t fully agree, Chris, about the applications barrier to entry. The applications barrier to entry was a very important part of the story in Microsoft because the story from the plaintiff’s point of view was that Microsoft wanted to prevent Netscape from having widespread distribution as a browser. Because if it’s was widely accepted as a browser, it could be a platform for other applications, and if it gave users another route to all of those applications, they wouldn’t be dependent on Microsoft.
So it was very much linking the two sides of the platform as a theory of harm. In the Google complaint, the word they used is not network affects its scale. And I think the reason for that is that while they do allege to advertising markets, the principle focus of the case is on the user side is on search, and what they’re really saying is — and this could be true with an old-fashioned Rust Belt industry.
You need scale to develop an effective product, and scale requires that you have access to a lot of users because with the scale that they’re talking about is data. And the agreement for the distributors principally making Google the default browser, they say, is a factual matter—I think this is going to be hotly disputed—prevented other search engines from developing scale.
So while I think there are clearly — an understanding of the business would clearly uncover a lot of the network effects things you’re talking about. I think the theory of the complaint is largely independent of network effects, maybe, because they seem to try so hard to ignore American Express. They didn’t want to get into platform issues. And so I think they bought this as a scale economy case rather than a network effects case.
Christopher L. Sagers: Yeah. Well, I will defer, and that may be right. I guess what I had in mind, again, was simply – I don’t know. There’s a half a dozen little allegations throughout the complaint in which the court said it was a key part of Google’s exclusionary strategy to keep rivals from getting the scale that would be necessary to make them meaningful challengers, to make sure that Google was preinstalled with a little collection of apps that only work with Google, or work best with Google, that would discourage installation of other programs. So I definitely take your point, and network economics may not really be the theory, but I see the same sort of reasoning of exclusion, as in the applications problem, even if the economics are a little bit different.
Geoffrey A. Manne: I think there has to — so I think that’s right. The element of scale here isn’t working in exactly the same way that the applications barrier to entry works, but this is a foreclosure case. You have to make the argument that competitors were unable to compete for distribution sufficient to reach scale, and the theory is not unlike the applications barrier to entry theory, and it has to with, as Doug said, obtaining sufficient data, sufficient users, sufficient scale in order to —
Christopher L. Sagers: Enough money.
Geoffrey A. Manne: — really to compete efficiently. But I think that becomes a, if not the, crux of the case is, how is the DOJ going to be able to really demonstrate that? In a sense, the allegation is that the default position that Google has, it alleges a substantial share of the distribution is what prevents Bing—let’s just use that as a stand-in for everyone else—Bing from being able to achieve a sufficient scale to get sufficient data in order to ever win the competition for the default positions, right? I think that’s a key part of the allegation here, it’s not just that Microsoft doesn’t want to or can’t pay as much as Google to be the default. That would be absurd. Microsoft’s a bigger company than Google. We all know it can pay for those position.
So I think the argument has to be, even if it could pay for those positions — well, actually that becomes a counterargument. Their argument has to be that Google’s ability to obtain those defaults puts Microsoft in a position where it can’t be competitive in the competition for these deals. But so that raises the question where I think they’re going to have a real problem, which is if those deals are so valuable for achieving scale, why, indeed, doesn’t Microsoft expect that by paying extra for those positions, it will get the scale it needs, as if that’s the only thing it needs to be able to compete with Google is the sufficient scale by buying those default positions. It could indeed do that and then win all of the competitions it wants.
Now, you might be able to make a vague raising “rivals’ costs theory” that it would have to pay too much, but it’s not clear that Google is not engaged in competitive bidding, competition for those slots, and it’s not clear that they would have to pay too much. But maybe.
Christopher L. Sagers: This is off topic a little bit, but I think we’ll get to it later. I wonder if maybe Microsoft can’t – let me just say, that is an excellent point. That’s going to be a very difficult factual problem for the government, and I hadn’t really heard anybody say it in quite those terms. The government’s own allegations are perhaps a little bit incoherent here.
I think the most interesting player in this story, that isn’t very well addressed in the complaint, is Apple. And I wonder if the reason that Microsoft isn’t paying more for this service is because Apple doesn’t want to sell it to Microsoft for I don’t know what reason.
Douglas Melamed: I don’t know that Apple’s anymore hostile to Microsoft these days than Google.
Christopher L. Sagers: Well, but, maybe, it gets something from Google that Microsoft couldn’t do.
Douglas Melamed: Well, that’s right. And that’s the key, I think, Chris. Because I think the answer to just [inaudible 25:15] is this — by the way, I think the government’s going to have some problems proving that default is a big deal. It’s not like in the old browser days and Microsoft where downloading a new operating system or a new browser actually was a big deal but leave that aside for the moment.
Assuming it was running an auction to be the default and imagining the default has some real value, there is, I think, a fairly traditional, in a way, a theory, that applies to these vertical restrictions that might explain why Google wins the auction, even though Microsoft has plenty of money, and it’s this: if they’re competing for just distribution, buying eyeballs and data, unless there are tremendous economies of scale, in the sense that the incremental value of the data means a lot more to Google than to Microsoft, you would think Microsoft would have every incentive and ability to compete against Google for those data.
But if Google is paying not just for the value of the data but for the value of kneecapping its rivals, which it can in effect recoup by preserving monopoly power in search advertising, then it has a return from the investment that is not available to Microsoft because Microsoft has no monopoly power to preserve. And that would mean that a perfectly rational Google would pay a great deal more than a perfectly rational Microsoft, and that might explain what’s going on here.
Geoffrey A. Manne: It’s possible. It’s possible. I think it’s a factual question. I think though while Microsoft doesn’t have a monopoly to preserve — to the point I was making before is that it arguably has a monopoly to try to obtain. If the idea is that if an implicit notion here is that scale is so important that you essentially end up having a market tipping to one provider or the other, obviously, it’s somewhat risky for Microsoft because it may not happen or it may not happen fast enough, but there’s still that prospect, which I’m not saying means that they’re inherently willing to pay as much as Google is if it has a monopoly to protect. But it certainly substantially decreases the difference in willingness to pay.
But I think also it’s pretty clear that the basis on which these default positions are won is not just the amount of compensation. It’s also the quality of the search product. Ultimately, Apple cares a lot about the product it’s offering to its users, and if it thinks that Microsoft is crappy compared to Google, that also would necessitate Microsoft paying more than Google. But, of course, that sounds like absolutely what we want from competition. That doesn’t sound like there’s anything wrong with that at all, unless the only reason they are so much crappier is because they couldn’t get the scale in the first place, and then it turns on how much advantage you get by being the default.
And as you intimated, Doug, I think it’s pretty hard to — it’s not impossible that the DOJ — that’s going to be the big question though, and it’s not quite clear to me — it is clear there is some advantage from that, but it’s not clear to me how big that is. And then, sorry, just to — I’ll stop talking. But I really think that you can’t talk about that without talking about the procompetitive justifications and this broader competition between these entities, especially, when you recognize that there are two elements required to achieve scale.
There is the access to the distribution channel, and there’s the size of the distribution channel itself. Well, if the size of the distribution channel itself is any way dependent on the relationship between Google and the distributors, i.e. Apple, and Android, OEMs, then we have to, at least, acknowledge the possibility that Google’s willingness to pay more for this distribution actually increases the size of the available scale, even potentially for parties that come in second place, especially, if being a default isn’t quite as much as it’s cracked up to be.
Christopher L. Sagers: We’re leaving Brianna completely out of this. Sorry about that, but I think this is pretty good.
Brianna S. Hills: No. It’s great. Keep going. It’s great.
Christopher L. Sagers: Two responses to Geoff. I think an important question is why — this is the question, I suppose: why is Google willing to pay more apparently than Microsoft, or at the very least, why does Apple want to deal with Google and not Microsoft?
I have two thoughts. This goes against my grain as the liberal that wants in. I trust plaintiffs to win, but I wonder if Google’s willing to pay more because it has a better product, but it has a better product not because nobody else could duplicate it, but just because it has substantial first-mover advantage. This is an important theme in the government’s complaint, as I understand it. They’re basic claim is, at this point, no one could catch Google without an investment that no one else is willing to make, aside from the installation and disclosure, I think. But if that’s true, even I have to acknowledge, that poses a serious remedies issue, if it’s true. But whether it’s true or not, it does seem to —
Geoffrey A. Manne: That would be a liability question.
Christopher L. Sagers: Well —
Geoffrey A. Manne: Why would there be liability for that? [inaudible 31:14].
Christopher L. Sagers: Well, I don’t know. That’s a different question. Should we just not have antitrust for people who manage to get monopolies through skilled foresight and industry? I don’t know. But that aside, there is this question that I think is unresolved: why is Google apparently willing to pay more? I wonder though, too, if Google is willing to pay more, as Doug said, because it does in fact have an anticompetitive advantage of some kind—not putting words in Doug’s mouth, but that seemed to be a suggestion. And that being the case, why Apple is particularly interested in dealing with Google?
And the thing that makes me most interested, in Apple, is that Google is paying Apple a really huge amount of money. Google is paying an amount of money to Apple that is the equivalent of about a third, I believe, of Alphabet Incorporated’s entire profits every year now. What are they paying that for? Maybe, the government’s right that preinstallation is so valuable as an exclusionary mechanism, and it’s really worth a third of the corporate family’s profits.
I bet it’s not though, and I just wonder if the reason Google’s willing to pay this is because Apple is the only plausible search competitor. And I’m not the first person to suggest this. Other people have said this. Is Apple being paid to stay out of search? And I understand why the government didn’t allege that because finding evidence of such an agreement that could survive Twiqbal seems pretty unlikely. But I just wonder if that’s really what’s going on.
Douglas Melamed: Yeah. I think that what Google’s going to have to do, in effect, to defend itself is something like this: they’re going to say, “Look, Microsoft uses Edge as the default browser on Windows.” By the way that calls in a little bit into question the idea that quality — I don’t mean browser. I mean search engine. It uses Bing. It calls a little bit into question the idea that the distributors really value the quality of the search engine. Although, Microsoft obviously has complicated motives along the lines, I think, that Geoff’s talking about.
But I’ll start with that, and I’ll say, “But look what happens.” A huge percentage of Windows users change their search engine. That proves being a default isn’t a big deal. And that’s going to be at hotly contested point of contention. The government says, over and over in the complaint, “Nobody switches away from Google when Google’s the default.” I don’t think that’s enough for the government because the explanation for that might be Google’s a better search engine. But what the government has to show is some other benchmark to show that being a default really makes a big difference.
But Google’s going to come in, and say, “It doesn’t make a big difference.” And you look at the Windows story, and that explains it. Then, they’re going to say, “Look, what that means is that all we get by being a default is those first few weeks, or whatever it is it takes, before the users switch default.” But they’re going to say, “That’s really worth a great deal because during those few weeks, those users, over all the distribution channels we’ve tied up this way, are going to make X number of gazillion queries, and that’s going to be a lot of data, and we value data at X, and that’s what we’re buying” if the numbers add up.
If the numbers don’t add up—and Apple’s obviously going to be the poster child that I would think for this—the inference that that’s not what Google’s doing—that Google’s buying something else, whether it’s preference vis-à-vis other existing search engines, as plead in the complaint, or whether it is to induce or maybe even reach an unwritten agreement with Apple, that Apple will stay out of the search business—could obviously point toward a big victory for the government.
Geoffrey A. Manne: I think one potential problem with that is — well, sorry, maybe not potential problem. But what do companies’, like this, marketing budgets look like? If you view the default placement as essentially a form of marketing—certainly not exclusive to that, but importantly that—I don’t know that the payments they’re making are wildly out of line with what one would pay for that. Maybe, they are, and to your point, I’m sure there are other reasons for it, but I don’t think your list of possible explanations for why they would pay for default is exhaustive, and I think marketing, like benefits, is an important one.
Christopher L. Sagers: It’s a good point. I think they pay a lot for marketing. And even if they pay substantially more for preinstallation, it may only be because it’s a more effective kind of marketing for this product.
Geoffrey A. Manne: And I think it also does go to, again, this larger point of the relationship between these big platforms. Maybe, this doesn’t apply as much on Apple, but certainly with respect to Google on Android, and Microsoft, and Bing on Microsoft devices. No, I think it’s an important part of their — there are a lot of motivations behind why you would do that, even if it’s not the optimal thing, even if it’s not the statically revenue maximizing decision to make because you think it helps to promote the completeness of your platform offerings.
The idea that you can get everything from Microsoft, or get everything from Google, may be important to them, and it may also be important and seemingly potentially procompetitive that these different avenues, into these companies, are avenues into other products that the companies are offering.
I think Chris was intimating that there was a potentially anticompetitive story behind that. but I think there’s a potentially procompetitive story behind that as well. Again, maybe, that’s a harder to case to make when it comes to being the default on Apple devices, but I think it’s certainly part of the explanation for why Google wants to be the default on Android and Microsoft, and Bing is the default on Microsoft. Oh, we’re done.
Brianna S. Hills: So I think we covered a lot in that discussion. One of the things that I think is still left open is how competition or how consumers are harmed. Is it because there’s less innovation? Is it because the consumers can’t choose what search is their defaults on the product they buy and weighing those against, as Geoff says, the potential procompetitive justifications? Google has come out and said, for example, that its agreements are similar to a soda brand that purchases the best space at a supermarket, and that this is not a new thing that they’re doing. So how do we think about the consumer harm and the competitive harm, and will the court need to move beyond our traditional formulations of that in this case?
Geoffrey A. Manne: Chris, you want [inaudible 39:25]?
Christopher L. Sagers: Yeah. Sure. I’ll say a couple of things. This is like the question for the liberal, I guess. First of all, I love Google’s PR department. They are good for a laugh. This isn’t like a soda company paying for space on a shelf because no soda company has 90 percent of soda, and none of them are paying a third of their annual profits to get shelf space, so far as I’m aware. We can in fact draw distinctions in American law.
As for consumer harm, I think it’s a mistake to think of the only consumers here being the people who make searches. There’s arguably harm to us in the form of lost innovation, or something like that, or possibly we’re being inadequately compensated for the value of our attention and our data. But there’s a different class of consumers here, which are the advertisers, and there’s a case to be made that they’re being abused, so that’s pretty straightforward to me. I don’t think we need a new law.
Douglas Melamed: I agree with that, but I’ve got a couple of additional thoughts. It does happen every day that arrangements are made for distribution that prefer one product over another, and the slotting allowances and grocery stores you referred to in the question, Brianna, are an example of that.
And yet no one would say it was consumer harm because a root beer is not on the top shelf of the soda shelf in the store, even though I might prefer root beer to cola—which by the way I do—and that’s because there’s no market power. We think the markets working in their reason of why people are paying for promotion and on and on and on.
The problem here is market power, but once you have market power—this, to me, has always been a puzzle—the insistence that the plaintiff prove how consumers were harmed. You don’t have to prove how consumers were harmed. All you have to show is that market power was perpetuated or increased compared to the but-for world. We infer harm from market power. That’s the bedrock of antitrust.
And we say because if you have market power—that the little simple diagrams we look at would suggest you’re going to increase prices, but that’s really a metaphor for other ways you might exercise market power, reducing product quality, reducing investment innovation, whatever it might be—the plaintiff shouldn’t have to prove that.
And Microsoft, above all, stands for that because Microsoft said, “You’re raising barriers to entry in the desktop operating system market, and what’s the harm? Some unknown, conceivably to develop in the future, potential competitor would be harmed by your conduct. So there was no harm in the past in Microsoft, and there shouldn’t have to be in any antitrust case.
Geoffrey A. Manne: I’m going to disagree with that as you might expect, Doug. This issue was, I guess, most recently aired in the FTC v. Qualcomm case, but it predates that. This question of whether Microsoft stands for the proposition that you can infer harm in circumstances that Doug describes or whether, as I believe, it says, in certain circumstances, you can infer causation, but not that you can infer harm.
There’s a lot of language in Microsoft that makes it seem, to me, pretty clear this is exactly what they mean. And I was ready for this, so I copied some of these down. When it talks about the AOL deals, and then moves onto the consideration of possible procompetitive justifications, it says, “Plaintiffs having demonstrated a harm to competition, the burden falls on Microsoft to defend its deals,” unless they demonstrated it just for the hell of it, that they indeed demonstrated a harm to competition.
It later says, “When it dismisses the idea that the deals with ICPs, internet content providers, were sufficient to do this—because plaintiffs failed to demonstrate that Microsoft deals with ICPs have substantial effect on competition—they haven’t proved a violation.” They address this in the discussion of remedies too, that because they had [inaudible 43:57] when suggesting that very strong injunctive relief is inappropriate, where liability is based on inference, it makes very clear what kind of inference it’s talking about.
As noted above, we have found a causal connection between Microsoft’s exclusionary conduct in its continuing position in the OS market only through inference. It doesn’t say, “We found harm only through inference.” It makes very clear that it thinks it found the causal connection only through inference.
I do understand that it can sometimes be quite difficult to separate the causation analysis and the effects analysis, and there are circumstances even in this case where I think the two are significantly overlapped, but I don’t think it’s the case that the plaintiff doesn’t have any burden to demonstrate harm to actual competition. I don’t think it’s sufficient under Microsoft to show market power.
Douglas Melamed: Well, I think those are two different statements. I agree with the first not the second. Let me make three quick points. First, Microsoft didn’t say it would harm the consumers. There was harm to competition. What does harm to competition mean?
Geoffrey A. Manne: [crosstalk 45:10] the same thing. It says [crosstalk 45:12].
Douglas Melamed: Well, no. They are the same thing. Right. Because you can infer harm to consumers from harming the competition if by harm the competition, you mean materially reducing the ability of rivals to discipline your behavior, i.e. the creation of market power. If you don’t have that, then you have no harm no foul under the antitrust laws. So, yes, harm the competition in the antitrust sense is sufficient to infer harm to consumers, creation of market power, which I think is the same thing as harm to competition is sufficient to infer harm to consumers.
Microsoft didn’t say you had to prove harm to consumers. By the way, I think Microsoft was wrong, simply wrong, when it said that the plaintiffs there had proven harm the competition in an antitrust sense, and here’s why I say that. Look it, the court held that the government had failed to prove harm to competition, injury to competition, creation of market power, call it what you will, in the browser market, and that’s why it threw out the tie-in claim.
The only harm that was shown, leaving aside Java — the only harm that was shown in the Microsoft case was to Netscape. That harm had to take place in the browser market, and there wasn’t creation of market power in the browser market. What there was, was sufficient harm to Netscape to prevent it from achieving sufficient scale to be an entry threat into the operating system market. That was the market in which competition was harmed, but it wasn’t harmed by changing competition. It was harmed by raising entry barriers by keeping out a potential entrant.
So I do think that notwithstanding the language in Microsoft, they actually did not find harm to competition other than reducing the likelihood of new entry into the operating system market, and having said that, I agree, that’s why the remedy was edentulous—I think the D.C. Circuit said—and why I think the government’s going to have difficulty getting a huge remedy in this case. But the magnitude of harm and whether you’ve actually shown consumers were hurt may go to remedy. I don’t think it goes to liability.
Geoffrey A. Manne: But are you suggesting that it’s sufficient to have market power and not to constrain the competition for distribution, but to win the competition for distribution, and that simply competing in that instance, as a entity with market power, is sufficient to create liability. Because I think that’s consistent with what you said, and yet I don’t [inaudible 47:45].
Douglas Melamed: I’m not saying that simply competing is sufficient. There are two elements to an antitrust violation: the creation of market power, the likelihood of doing so, compared to the but-for world, and engaging in anticompetitive conduct.
So if all Google is doing is engaging in procompetitive conduct, participating in auctions for preferred distribution services, I think the government’s going to lose this case. What the government has to show is that they are either tying up too much or paying too much because they are in effect using the prospect of preserving the monopoly profits as the lever to obtain this preference over 80 percent of the search queries.
Geoffrey A. Manne: That might be right if they’re paying too much, but you also said tying up much, and there, that’s exactly where I think we disagree. If they are tying up too much of the market because they are out competing equally efficient competitors for these distribution channels, you shouldn’t be able to infer anticompetitive conduct from the fact that they are tying up too much of the distribution channel for competitors to achieve scale. That would seem to be absolutely the opposite of what antitrust should do. [crosstalk 49:00].
Douglas Melamed: Yeah. You put your finger on what I think of is really one of the really interesting tough, unresolved, and mostly unaddressed questions in antitrust law. I actually think it’s a matter of theory that the plaintiff should have to prove that the defendant is paying more than the legitimate rewards to preferred distribution. In that sense, it’s paying a premium to protect monopoly power. In that sense, I agree with what you said. It’s a matter of theory.
Here’s the problem, and I’m going to do it with talking about predatory pricing, in case, just because it’s something to make the point. We talk about predatory pricing cases in terms of standing for the power of position, that if you’re more efficient than the rival, it’s okay to drive them out of business by charging a low price provided that it’s above cost. Above cost being the measure of relative efficiency.
The problem with that — and I don’t know what to do about it. The problem with that is that if the rival is less efficient simply because it hasn’t had an opportunity to obtain scale economies and, in fact, has a more efficient production model—such that if it achieved scale, it would be the more efficient competitor—then the law disadvantages rivals.
I think we have mirror image problem here if we went full-bore with the theory I’m talking about in terms of looking at the excessive price. Having said that, point two, there’s a lot of law that says if you tie up too much distribution, we’re not going to ask what the price is. We’re going to call it an illegal restraint of trade—that Section 1 law in exclusive dealing, and I think by analogy, the courts would probably and maybe correctly, in the shortness of life, apply it here rather than going full-bore on my theory about excessive price.
Geoffrey A. Manne: There’s also some contrary laws. As you said, this is a really messy area of law. There’s also some contrary law. There are examples, especially, in tie-in cases, where courts are looking pretty clearly at a but-for world and limiting the ability to demonstrate harm to, essentially, but-for exclusion, which I think is really crucial here.
And you’re right. It’s not obvious that the courts will apply that approach. I think theoretically it’s clearly a mistake not to, but I also think there’s law that supports that approach as well. In reference to predatory pricing, I would make reference to tying, but they’re all the same thing at the end of the day anyway. And this does relate to—as Brianna mentioned—the procompetitive justification or—I don’t know. And, maybe, it should come in earlier in the case, particularly, my point about the but-for world and how much is actually being foreclosed.
If, indeed, consumers would choose Google anyway because it’s a superior product — we can even say if, indeed, they would do that, even if Bing had sufficient scale, as if that was the only thing that was needed to make them potentially as good a product as Google, then it’s pretty hard to say that there is any harm here.
At the very least, you have to acknowledge that the amount of distribution that’s being foreclosed is either or both limited in scope or time. In other words, if Bing was somehow just granted the default status in all its distribution channels, and, within a day, everybody switched back to Google, sure, there may be some foreclosure — that may show that there would have been some foreclosure of that one day’s worth of search, but that’s when we’re usually talking about three-year contracts and the like, where one day is probably not going to register.
But yet I do think that that’s kind of this weird position that the DOJ’s case is in that they have to be able to argue that that period of time between when — for those customers who would switch, between starting up their phone for the first time and realizing Bing is the default, and that they’d rather have Google, and switching over to it, somehow has to be enough to facilitate Microsoft, Bing being as efficient a competitor, and not just as efficient.
I think it also has to demonstrate that somehow—and I don’t think it could do this—that that’s enough to give it the ability to have the same level of quality as Google such that it would be chosen by OEMs to be the default. And the weird thing, of course, is that the more people who would switch over to Google, the vastly less likely that is, and also there’s a problem demonstrating that scale alone, data alone, is sufficient to be able to enable that.
Christopher L. Sagers: If I could, I’d like to add one thing, and, finally, at the end of an hour discussion, I’m going to say something that’s going to really annoy a Federalist Society audience. I think that what — at the heart of this discussion—which by the way has been masterful on both sides. I think very thoughtful. And Doug made my heart sing when he said that we presume injury from market power gotten by exclusion. I love it.
But at the heart of this is what feels, to me, like a practical policy problem, and like antitrust lawyers that I know, never want to talk about this because it’s more fun to talk about economics and abstract theory. But there is a practical problem which is the ideological differences among us will remain ideological, unfortunately, because they impose empirical questions that can’t really be answered.
And I think you both are begging two empirical questions of a kind that I hear constantly that, unfortunately, we can’t answer, which is, Geoff, you, it feels to me, want a much more in-depth examination of exactly what the harms are before the government interferes on the basis of a strong presumption that interference itself is dangerous and costly and should be avoided in the absence of great certainty.
And that all feels all like science, and care, and everything, except that, to me, it doesn’t feel that way because I think there’s a different empirical question, which is that I think if we hamstring the government on a presumption that government interference is very dangerous and costly, we just invited different cost, which is it is possible that if we don’t do stuff, all of this conduct that will go unaddressed really is harmful.
And that’s why I agree with Doug, at least, as the law still stands. Even at this late date, the law still stands in antitrust that, at least, the government doesn’t really have to prove injury to the degree that you state, and I think it’s a good thing. And I’d just like to say, I actually think there are quite a lot of examples, other than ones that Doug mentioned, in which the government can win a case on existing law without proving what would be very difficult questions of harm if we had to show that ultimately consumers are harmed in some way.
It is still the law, for example, that monopsony is illegal, and price fixing agreements that stop no-poaching against — in employment, those things are illegal even though the harm to end-use consumers would be difficult to show. When abuses are caused with market power, and it can be shown that it was done in ways other than competition on the merits, we presume that there’s injury, and it’s illegal.
Geoffrey A. Manne: I agree with that, Chris, surprisingly, in the sense that I agree that the law is actually far more tilted toward plaintiffs than a lot of people are willing to admit because precisely of these sorts of presumptions. The reason I think that there’s — and you’re right that it operates in a lot of circumstances, especially, in antitrust because proof is often so difficult.
But here, I think, in particular, we’re going to have a dispute over whether there is foreclosure of the ability to compete effectively or viably for these distribution outlets. And if that competition—which some people refer to in Apple’s case as an auction. I have no idea where that comes from. My understanding is that’s not at all true that Apple doesn’t run an auction for this. They have commercial revenue-sharing agreements with companies—in fact, both Bing and Google, but I digress.
If those deals are operated in a situation in which there is no constraint on Bing’s ability to compete for those distribution deals, I don’t think that’s a circumstance in which a court is likely to infer from a fact of exclusion that there is anticompetitive harm, and nor do I think it would appropriate to do so.
So while I agree with you that in other circumstances, it’s not going to be required that you precisely show the harm to consumers if you could show, for example, that competition was rigged, I do think the courts would infer consumer harm from that. But I think it really matters how you get there, and, in this case, I think that’s going to be a problem for the — it may not be a problem. It’s a matter of evidence. I don’t know what the evidence looks like, but it’s potentially a problem for the DOJ.
Douglas Melamed: For whatever it’s worth, I think the idea of Apple running an auction, at least the way I intended to use it, is more of a metaphor, but implicitly, Apple is saying, “I haven’t asked yet, how can I maximize value from it?” They don’t have to announce for having an auction “Send me your bids by noon on Friday.”
Christopher L. Sagers: I hear people saying it a lot for some reason, and I’m not sure where it came from, but I don’t think it’s — I don’t know that it’s not true, but I don’t think it’s actually accurate. As a metaphor, I totally understand. I agree.
Brianna S. Hills: Great. I think that’s a good place to stop. We’ll turn it back over to Nick.
Nick Marr: Great. Thank you all for joining us. I want to thank you specifically for the benefit of your valuable time and expertise today, to our audience for calling in and watching us lively debate. I want to remind you all, keep an eye on your emails and on our website for announcements about upcoming calls, both Teleforum calls and Zoom debates like this one.
President and Founder
International Center for Law & Economics
Professor of the Practice of Law
Stanford Law School
James A. Thomas Distinguished Professor of Law and Faculty Director, Cleveland-Marshall Solo Practice Incubator
Cleveland-Marshall College of Law, Cleveland State University
Boies Schiller Flexner LLP
Federalist Society’s Corporations, Securities, & Antitrust Practice Group