Oral argument — US Airways, Inc., for American

I believe we're ready to hear our first case, U.S. Airways v. Sabre Holdings. May I proceed, Your Honor? May it please the Court, my name is Evan Chesler and I represent Sabre. In my time this morning, I'd like to make two basic points, if I may. The first point is that this case involves a two-sided transaction platform. And we believe that that is true as a matter of law and that it was error for that issue to have been submitted at all to the jury. I think it's important while you're doing your two parts to engage us a little bit about the so-called hypothetical verdict of the jury on the two-sided part of it. Yes, Your Honor, I absolutely will do that. The point I wanted to start with, which gets to that point, Your Honor, is that after the district court decision in the Amex case was rendered against Amex, the U.S. Air parties went to the trial court in this case, asked to drop their damages claim and to get a bench trial instead of a jury trial. And at that time, they told the district court, and I'm quoting, most importantly from our standpoint, we feel that the likelihood is that this case is going to get resolved in the Second Circuit along with the Amex case and an informed decision making, particularly at the appellate level, would benefit by a well-considered, well-thought-out district court opinion. So at that time, they said the two cases were so identical that they were going to be resolved together in this court. Then, of course, this court reversed the district court in the Amex case, and their position changed. Now, the cases are, in fact, identical. They both involve two-sided platforms with indirect network effects where incentives paid on one side affect demand on the other side. And therefore, we believe the district court should not have submitted the case to the jury at all. But having done so. The cases say that market definition is fact-intensive. It's a factual question. The question was put to the jury here, and the jury found that it was one-sided. Why shouldn't we defer to that decision? Because, Your Honor, the Supreme Court opinion has made it perfectly clear, we believe, in the Amex case, that if, in fact, the market meets the criteria which were involved in the Amex case, which are identical to this case, as this plaintiff argued before, it, in fact, was a matter of law, as the Supreme Court case made clear. If, in fact, the markets are identical, that case has already been decided. Now, I want to come to Your Honor's question about the jury form. When the judge instructed the jury, she misinstructed the jury on the test for two-sidedness. The district court instructed the jury that if a price effect happened on one side, if it affects demand on the other side of the market, then it's two-sided. That was a fundamental and harmful error, because the Supreme Court held that if a price change on one side of a platform affects demand on the other side of that platform, it's two-sided. In this case, there are multiple platforms. Assuming that we agree with you that this is a legal determination and that that was an error in the definition of how two-sidedness should be evaluated by the jury, why is it harmful? If it should not have been evaluated by the jury at all, Your Honor? No. Let's say we say that it's a legal determination, but we had this hypothetical proceeding. There was a definition given, and you're saying, I assume, that that definition was erroneous and would have had an injurious effect on the verdict, even assuming we can engage the hypothetical determination. Yes, because, Your Honor, if there had been a two-sided market, it would have fundamentally affected the antitrust injury question. Antitrust injury in a two-sided platform is a very different analysis from the injury question in a one-sided platform. Indeed, the district court instructed the jury that the harm to the plaintiff was that they were paying too high booking fees. That's a one-sided harm because it doesn't account for the incentives paid to the agents, which are part of that price. So the entire analysis that was provided to the jury was erroneous. And now to Judge Sack's question. A hypothetical question at the end of a multi-page verdict form where there are nine questions that relate to a single-sided market, and then the court says, hypothetically, if everything you've just spent all that time debating turns out to be wrong, would you come out the same way? And they check off the box that says yes. Does not cure that error. I've been trying cases for over 40 years. When a jury gets to the end of the verdict form, they see the light at the end of the tunnel. They know they're going home. And to ask them at the last minute of the form, well, hypothetically, if your entire analysis were turned upside down, would you come to the same result? And they check off a box that says yes. Doesn't cure the error that was infected into this record. Let me turn, if I may. Just before you move off this point. Yes, Your Honor. You're saying the instruction, instead of saying the price on one side of the market affects demand on the other side, I'm looking at the instruction at 685. It should have said platform. Yes, Your Honor. And I'm having difficulty really under-appreciating why it makes a difference, whether it says market or whether it says platform with that one word. So if you could just explain that for us. Absolutely. This was defined, the plaintiffs defined this market as a GDS market, in which only companies like my client, Sabre, were included. And excluded. There are only three. There are only three, exactly, exactly. Now, when the court said a price change on one side affects demand on the other side of the market, the burden that that placed on us in that jury instruction was to say, if we increased our incentive payments to a travel agent, that price increase would need to affect demand across all of the platforms in the market, not just our platform. That's a much greater effect on the market than the Supreme Court case mandates to determine two-sidedness. That was almost an insurmountable burden for a jury to find two-sidedness, because the market had multiple platforms. The analysis in the credit card in Amex would have been the impact on Amex and not the entire credit card. Exactly, precisely, Your Honor. That's the difference, as opposed to an impact on the Visa customers and on the MasterCard customers, Discover, et cetera. And this is because we're just talking at this point about defining the market and not on the bigger picture of the anti-competitive effects. That's right. But once you define, as the Second Circuit said in the Amex case, once you misdefine the market, the other error flows from that. That was exactly what the Second Circuit said in the Amex case. What kind of impact on the travel agent side of the platform market? Are they alleging or are you saying it turns it into a two-sided market? That when incentives are paid by out my client to agents, that encourages them to book more flights on their platform, which aids or assists the plaintiff's business. More bookings yields more customers on their planes. If they have all of their flights listed on our platform, the price change which incents the agents to do more booking affects demand on the other side of the platform. It may not affect demand on the other side of the entire market, and that was the error we submit in the court's instruction below. Just give me an example. Make it a little bit more concrete for me. You pay an agent. Let's say I'm going to St. Louis. Help me out here. I don't know why, but if you want to go to St. Louis, sure. Okay. I'll let them know how you feel. Please don't tell them. So let's say we pay the agent $5. I'm making the numbers up, $5 fee incentive. And we say, you know what, we're now going to pay you $8 if you book flights to St. Louis. And now this agent is going to be more aggressively, presumably, pursuing customers who want to go to St. Louis and booking more flights. And if US Air has a lot of planes that fly to St. Louis, they're going to have more customers. And they're going to make a lot more money on those tickets than whatever the booking fee is that they're paying my client for those tickets, because the booking fee is just a small percentage of that. But they said their harm in this case was that their booking fees went up. If I have business in Memphis, they're going to get me to fly to St. Louis? No, but if you have business in St. Louis, they may get you to fly on their airline to St. Louis as opposed to somebody else's airline to St. Louis, and that would increase their business. But a travel agent deals with all airlines, don't they? Or are you saying they just try – They deal with whatever agents are on a particular platform that they may be using to book that flight. That may be all airlines. It may not. But they all dealt with all major airlines. Well, in our case, we have a lot of the airlines. I don't know if we have every one, but we have a lot of them booked on ours. The ones that service businesses, the four – Yeah. I think I know what they are, and I think, of all things, Air Alaska is probably one of them. But that's all – I mean, they could go to any one of these three and get any one of those airlines. They could, Your Honor, but if, in fact, this airline has more seats or this airline they think is a better service, to the extent that more bookings take place as a result of that incentive, they're going to get more traffic. Even if their share stayed the same, if it's a share of a larger number of flights that are going there, they're going to have more business, and they're going to make a lot more money on that ticket that you buy than they're paying my client in a booking fee. And yet their harm is that they said the booking fees were higher than they would have been but for our policies. That's exactly what the issue was in the Amex case. The merchant said their transaction fees were higher than they would have been, and our defense was, but you're getting a lot more volume. People are spending more money on the cards, and you're making a lot more money. And this court said that that was correct and overruled the district court. Your position is that in these two-sided platforms, the only way you show a substantial economic bad effect on competition is by showing it limits output. That's my second. Thank you, Your Honor. That was the second point I want to get to. The Supreme Court was very clear in the Amex case that the principal harm, the principal antitrust injury, in any case, whether it's a two-sided platform or a one-sided platform, is restriction on output. Indeed, the Supreme Court quoted the Areta and Hovenkamp treatise and said that it is, in fact, a limitation on output that causes a price increase that is the harm that the Supreme Court needs to look at. I guess I don't understand that because I would think prices above a competitive level and effects on quality would be sort of classic antitrust injury. Your Honor, actually, Justice Kennedy said in the Brook Liggett case that increases in price in the face of increasing demand tell you nothing about whether it's anti-competitive. So you need to determine whether output is being restricted and therefore the market is being affected in an anti-competitive way, and that's why the price has gone up. As opposed to if you and I both want the same car, we may actually bid up the price of that car without any antitrust violation at all because it's increasing demand. And so what the Supreme Court was focused on is exactly that issue. Why was the price to the merchants going up? Was it a pro-competitive reason or anti-competitive? I get that an increase in price may not be enough to show that it was the result of an antitrust injury, but I'm having difficulty with the blanket statement that you could not, the only way in this two-platform market you could show that there was an anti-competitive injury is if you show some restriction on output. Yes, and we believe, Your Honor, that that is the law, that in fact markets that are characterized by increasing output are often characterized by increasing price. And it is when, and here's the quote from the Supreme Court. This is at page 2288 of 138 Supreme Court reports, and I quote, market power is the ability to raise price profitably by restricting output. They are inextricably linked. That's the test. And the problem with the U.S. Air case here is that their theory was, this is Professor Stiglitz's testimony, that the result of our full content provisions was that transactions that would otherwise have gone to other platforms, which were outside of their market definition, did not go because of our provisions. If those transactions had gone to platforms outside of their GDS market, output would have gone down, not up. And that's exactly contrary to what the antitrust laws mandate. Now, they may tell you when they get up, well, those provisions, those other platforms would have been in the market, but for these restrictions, but that runs afoul of the smuggler's notch case. Other platforms, what are you referring to? Well, for example, U.S. Air has its own platform. You could call up U.S. Air and book a flight directly. They have their own platform. But in the plaintiff's market definition, that's not in their market. So if our provisions, our full content provisions, prevented a transaction from going from Sabre to U.S. Air's own platform, which is their position, this is on page eight of their brief, then that would be taking a transaction from inside the market to outside the market and lowering output in the market. Is there any reason to think, or maybe one should find this out as a matter of evidence, but is there any reason to think that the way this is structured, that there would be anything conceivable in the light of the way things work that would suggest that when a company goes to a travel agent, the travel agent would go outside one of these three platforms? Well, their argument is that if we weren't somehow locking these people in, holding them onto our platform by the incentives we pay them, that they may well book it outside of our. Well, but maybe the theory is, and I say maybe, but maybe the theory is that if it wasn't for this, there would be other platforms just like the three that now exist. That is to say, the three that now exist, as I understand it, are rather backward, and they survive because they're the only three, and a fourth one has never arisen. And I suppose the question is why. Your Honor, again, that was exactly the argument the government made in the Amex case, that there would be more charge card platforms but for Amex's provisions. And the Second Circuit and the Supreme Court both rejected that, primarily because it was clear that output was increasing in the defined market. And in the face of increasing output, that shows a healthy competitive market, not a market burdened with unlawful market power. It's exactly the same argument, which is one of the reasons why we say this case and the Amex case are exactly what the U.S. air people told the district court when we lost in the trial court. They're the same, and they're going to get resolved together in this court. Now they're going to tell you that they're different and that that case doesn't control them. Can you just return to my colleague's original question about the hypothetical proceeding again? Let's just assume that we agree with you, that Amex says that this now becomes a legal issue, that this is a two-sided transaction platform. It would have been better if this case had been tried from the beginning with that instruction, but it was part of the mix at this trial, and the jury was instructed, given the opportunity to address in the hypothetical, assume that this is a two-sided platform. Do you still find a substantial anti-competitive effect? And there was some evidence to support that in the record. So what was wrong with that proceeding? Because the but for the hypothetical question that was asked at the end of the verdict form, the entire structure of the verdict form and the entire presentation of the case was as a one-sided market. They had a Nobel Prize laureate testify at length to this jury that this was, in fact, a one-sided market. He made all sorts of arguments about the maturity of the market, and it couldn't be two-sided. It couldn't have indirect network effects. So to try to tease all of that out now, Your Honor, to say what would the jury have decided had they been properly instructed, I would submit is impossible. They were presented. Do we have any case law as to this sort of a hypothetical? The closest, and, Your Honor, the answer is I think your law clerks were doing a good job and not finding very much. I'm sure they were. The closest we found is a case called Romer v. Baldwin. It's a Third Circuit case. Just let me take a moment on that case. It was a car accident case, a negligence case, in which the jury found that there was no negligence. They were asked a hypothetical question. Suppose you had found that there was negligence, what damages would you award? And they gave a damages amount. And the question on retrial became, well, is that damages amount binding? Was that a real damages amount? And what the Third Circuit said was you ask a hypothetical question like that where the focus of the case is in one direction, and then you say to the jury, but if it were different, what would you decide? And the court said it's too, there's too much danger that the jury will not make, will not take the question seriously. That's exactly the problem here. As I said before, I've dealt with a lot of juries over a lot of years, and they work very hard. And when they're done, they're done. And when you ask them nine questions, and the second question of the nine is, is this a one-sided market or two, and they check off it's a one-sided market, and they then go through seven other questions all in that context. And then you say at the end, by the way, hypothetically, if it were a two-sided market, would you come at it differently? The way is not exactly the words that they're using. No, it's not. It's not. And, in fact, the court, I have the words. Yes, please. The court said the following questions are hypothetical questions, hypothetical in italics. You should not consider the following questions until you have answered the preceding questions. So you're almost done. And the court goes on to say you should be very serious in considering these questions. You should answer the following questions with the same diligence and seriousness as your prior answers, because these answers may be very important if there's a change in the applicable law. Well, the court in the Romer case also said to the jury, you know, be serious about this. I want your answer. And the Third Circuit said, sorry. It just doesn't work that way in the real world. That's not the way juries act. And when you plug on a hypothetical question at the end of that process, you run a real due process risk that your jury has not given adequate consideration to that question. And we submit that that's exactly what happened here. When you bring in someone who's won the Nobel Prize in economics and says this is not a two-sided market for the following reasons, and then you ask the jury a hypothetical question, what if it were, and they award the same amount of damages and say, yes, I'd come out the same way in the last two questions of what is then an 11 or 12 question form, you haven't fixed the problem. Thank you, Your Honor. Thank you, Your Honor. Anton Matlitsky for U.S. Airways. I just wanted to start first on the output point. I think we can get through that quickly. Here's the quote from Amex. It says, the court will not infer competitive injury from price and output data. Absence of some evidence that tends to prove that output was restricted or prices were above a competitive level. The problem in Amex was that the offer of proof was that prices were rising. And if all you're showing is that prices are rising, you have to show a reduction in output. But here we weren't. Our proof wasn't that prices were rising. It's that they were super competitive. We had expert testimony about what the prices actually are, what the prices would be in a competitive market, including taking out incentives. Doesn't your adversary have a very good point in the fact that this jury awarded the same amount of damages on two theories suggests that the jury wasn't really engaging with the concept of a two-sided platform? No. I think it actually demonstrates the precise opposite. There was – everybody agreed that if you're calculating net price in a two-sided market, the way you do it is you figure out the price, you subtract the cost, including travel agent incentives and a reasonable profit. There was a dispute between the parties about how you calculate net price in a one-sided market. We argued that if the market was one-sided, you don't take into account travel agent incentives because our view was that they're basically akin to bribes in a one-sided market. Their expert testified over numerous pages of the transcript – this is 638 to 39 – that that's wrong, that even in a one-sided market, travel agent incentives – Numerous pages is two pages? Well, that's – well, they're the little four – you know, so it's like eight pages. That's a lot of pages. In any event, he testified that in a one-sided market, you should still take into account travel agent incentives when you're calculating net price because it's an actual cost to So the jury resolved that question in their favor. And so their best evidence that the jury was confused when it was answering the two-sided market question is that they agreed with them about the way to calculate two-sided harm. Now, in their brief, they say that that would be – that that finding would be contrary to the instructions given to the jury by the district court, and that's incorrect. The instruction they cite is the instruction that says super competitive pricing simply means setting prices higher than competitive levels within the relevant market, viewing the net price affecting customers on both sides of the platform if the relevant market is two-sided, and viewing the net price on the airline side of the market if the relevant market is one-sided. They say that would have precluded the jury from finding that net price in a one-sided market should take into account travel agent incentives. But that's wrong because the whole question is how you calculate net price, right? So anyway, so that – I don't think that argument demonstrates any kind of confusion. And again, I think it demonstrates the exact opposite, that the jury was really paying attention. But on the output point of it – You argued earlier in the case that this case was just like Amex. And if you – if we just step back and look at the two cases, you know, from a big-picture point of view, they do seem awfully similar. The credit card case involves these nondiscrimination provisions in the agreements. This case has the full content. So how are the cases different? Well, the markets – the markets are very different. And the whole principle, as Your Honor said before, is that antitrust cases in general and market definition in particular is extremely fact-intensive. It turns on the particular facts of the case. In Amex, they had a trial. And what the court held in Amex, this court and the Supreme Court – How are the facts meaningfully different in this case? Well, so for purposes of market definition, for example, the facts are different because the main difference is that the market in this case does not exhibit two-sided interdependence or indirect network effects. Here's the rule of law in Amex. A market should be – this is from the Supreme Court case. A market should be treated as one-sided when the impacts of indirect network effects and relative pricing in that market are minor. Not on that platform, but in that market. That's the mirror image of the charge that the district court gave in this case. She told the jury that a market is two-sided if there is interdependency between both sides of the market. I think their position is if a district court had read the line from Amex to the jury because it says market instead of platform, this court would have to reverse, which seems to me that it can't be right. So the jury answered that question based on the evidence in this case. And what it found, what we presented is, unlike in Amex, when you lower the price – My understanding is the argument is that in defining the market, you look at the interdependency within the platform. Just as with the American Express, it was whether it was affecting both sides of American Express, the merchants and the customers, to define the market. And then once you've defined the market, then you look at whether there are any competitive effects. Right. And so here, to apply it the same way, you would look at the platform. Right. That's their argument. But our submission is that argument is wrong. First of all, I just read the rule of law. I understand that, but why? Well, it's wrong. It's wrong. First of all, I mean, Amex just – I just read you the rule of law, and it says we're talking about interdependence on the market, on the market side. It's true that in the Amex case, the Court used market and platform interchangeably, and it talked about interdependence on the platform because – I think the problem is the language is used a little bit loosely, and both sides confine passages that support their position. Well, there's a good reason for why it was used loosely in that case, because it didn't make a difference. The way the indirect network effects in that case worked is where when you lower the price on the credit card side of the market by giving, you know, Starwood points or something like that, people spent more money. So there was increased output on the merchant side of the platform, but also necessarily on the market side. But the reason what – in principle, that you care about the market interdependency rather than the platform. The argument here is that if you increase the fees to the airlines, you'll have more revenue to use to pay more incentives to the travel agents, and so, therefore, that's your interdependency. Right, and the difference between this case and Amex is that you raise the price on the airline side, you pay higher incentives to the travel agents, and then there is no benefit at all coming back to the airline side. There might be a benefit on the platform, but that doesn't matter to the airlines because all the airlines are on every platform. with no benefit at all to the airlines. And the reason that – Why isn't there a benefit to the airline if there are more travel agents using the platform? Because potentially there are more customers, more folks who want to fly using Sabre or using agents that go through Sabre. Because that is what – we presented the following facts to the jury. First of all, there is full participation already in the market. Every major company is already using travel agents, and travel agents are already using all the GDSs. Plus, there is inelastic – That's looking within the market, and that's the point is – Well, right, but – Determine interdependency, you want to look at the platform. Right, but that – we think that's wrong. And let me just explain why, as a matter of law, that that's wrong. The reason that this two-sided market analysis matters, why we do it, is because if somebody on side A of a real interconnected two-sided market complains that a price is too high, right, you can't tell whether that price is really too high or is super competitive if what's going on is a raise in the price on side A is used to lower the price on side B, and then side A, a benefit in output comes back to side A, because then you have to look at both sides of the market to judge whether the benefit in output outweighs the increase in price. But in this case, what's going on is you lower the price on the travel agent side of the market, and there is no benefit in output to the airline side of the market. So there's absolutely no reason to look at the travel agent side. But if you disagree with that, Your Honor, I just want to get to this hypothetical or alternative verdict here, because there is absolutely nothing in principle to preclude this kind of instruction. We cited a case, a fairly recent case from the Federal Circuit, which upheld this kind of instruction. It's nothing like the Romer case. And let me just take a few seconds, if I can, to explain why. First of all, I think it's important to understand when you're deciding whether there's an abuse of discretion here what the district court was the situation the district court was faced with. The Amex decision had just come down. In the middle of the trial, there was an en banc petition filed, and it was supported by a bunch of amicus briefs. And the district court understood correctly, as it turns out, that either en banc or cert was likely or was fairly likely to be granted. And so the district court was faced with the prospect of giving an Amex instruction without this charge, without the alternative charge. The jury finding a two-sided market and ruling that we did not show harm to a two-sided market, then all of a sudden the Supreme Court grants cert, and Justice Breyer gets another vote. And so there's no more – there's no such thing as two-sided market analysis. All of a sudden, we have an unanswerable appeal, and we have to do this whole eight-week trial all over again. There's absolutely no reason to do that, because unlike in the Romer case and unlike my colleague's – contrary to my colleague's presentation, this case was not presented to the jury as a one-sided market. In terms of it was – The case was largely tried on your side as a one-sided market. No. No? I mean, did the parties during the course of the trial argue, if it turns out that this is a two-sided market, this is what you should find? Absolutely. The focus of the trial was that these provisions were anti-competitive. They caused super-competitive pricing. They caused decrease in quality. They prevented innovators from coming into the market. Was the hypothetical – In a one-sided or two-sided market. Was the hypothetical to the jury – did the parties know that that was the plan at the start of the trial? No, not at the start of the trial, but – Towards the end of the trial. Towards the end of the trial. But the way the evidence was presented to the jury was not, this case is about whether this is a one-sided market. It was, these provisions are anti-competitive. We think it's a one-sided market, and they're anti-competitive. In a one-sided – they cause harm to a one-sided market. But if you think that it's a two-sided market, then it causes harm to a two-sided market. That's why we had testimony about taking out the incentives. Were these two markets part of the opening statements? Definitely part of the closing statements. I'm sure they were part of the opening statements, but I can't remember. But there is absolutely no ambiguity that that's the way that the case was tried to the jury. That's unlike in Rome. It's like in the EMI case where, actually, the theories in that case were tried in the alternative, and they were mutually exclusive. There was an invalidity, a patent invalidity claim where they first argued that the claim was impossible, and then they argued that it was inherent in nature. So in the Logan case, what happened was, as my colleague said, there was a finding of no liability, and then the jury was asked to decide whether there was damages. Nobody obviously tried that case as, we think there's liability, but if you think there's no liability, we think damages are $30 million. So the jury would have been obviously confused as to why they're being asked about damages. The second point is the instruction in that case was incredibly confusing because at one point the district court judge told the jury that if there's no liability, there's no damages. Then he tells the jury, if you answer no to liability, then the defendants are entitled to the judgment. But still, it's helpful to us in our process if we know your view on the damages. So they're on their face being asked to answer a completely moot question, whereas in this case ---- The jury began by saying that the district court gave an erroneous instruction on exactly what a two-sided platform is or a two-sided market is, and that that would have confused the jury and tainted its two-sided analysis. Well, we think the instruction is not erroneous because it's just the mirror image of what the Supreme Court expressly said in Amex. But the whole point of the alternative verdict is that the instruction doesn't matter. If the jury found that there was a one-sided market on an erroneous instruction, then it was asked to find, as the parties argued, if you think the market is two-sided, how do you find it? And the only difference, everybody agreed, a potential difference between a one-sided market and a two-sided market is whether you include travel agent incentives in the cost when you're trying to figure out whether the prices are super competitive. The other offers of proof, there's no difference between a one-sided market and a two-sided market, like decrease in quality, right? I mean, a decrease in quality is a decrease in quality. And the third point I just want to make on the alternative verdict is, again, there's no evidence, evidence of confusion at all here. The only evidence they present is the fact, as we talked about before, that the damages numbers are the same, but there's a good reason that the damages numbers are the same, and the reason is that they – that the jury agreed with them. Can you – you talked about this before, but if you could return for a moment to my colleague's questioning about just what Justice Thomas was saying, because I am having a difficult time not understanding that if it's a transaction platform, two-sided transaction platform, it just now – it's just now legal error unless you instruct the jury to take into account both sides of the platform. Right. I mean, we think – that's wrong, I think, for the following reasons. So the rule of law stated in Amex is, as I said, the market should be treated as one-sided when the impacts of indirect network effects in that market are minor. Now, the discussion of the transactional platform, the simultaneous pricing on both sides, was intended to show that in that market, in comparison to the newspaper market, the indirect network effects were more pronounced. That's the words in the Supreme Court opinion. But that doesn't mean that every transactional platform is going to have pronounced indirect network effects. And this market, unlike the credit card market, has aspects to it that make the network effects not pronounced, and that is precisely the question that the district court posed to the jury and the question that the jury answered. Now, of course, we don't think any of this matters because the jury also was asked whether there was harm to a two-sided market. So if they're right that that was the proper question, the jury answered it in our favor. And did you want to take just two minutes to address the cross-appeal? Sure. Yeah, so we have two issues on the cross-appeal. The first is we think the district court erred in applying the statute of limitations, not allowing us to demonstrate harm within the limitations period caused by the 2006 agreement, rather limiting it to the 2011 agreement. There's a continuing harm doctrine that comes from Hanover Shoe, and it just applies here. The other cross-appeal is on market definition, the question of whether you can legally have a single-brand market. The district court held that as a matter of law, you can't. That's incorrect. You can. Usually it's going to be very difficult to plead that more than one brand in a market are not interchangeable. But here we did. We pleaded why. There was basically zero cross-elasticity of demand between the Sabre platform and other platforms, including the fact that the prices, that the cost, that the platforms themselves were extremely sticky because it's exorbitantly expensive to switch. We showed other indicia, including the fact that DOJ and DOT believed that Sabre was its own platform and the like. Thank you, Your Honor. Could you, maybe without time, this is a matter of pure curiosity and really can't have any effect on what's happening, but I'm curious, was there any lawsuit brought by another airline along, and I don't care how it came out, I just want to know if another one was brought along these lines? I don't know, Your Honor, as I stand here. I don't want to misinform the court. Let me make just several points. First, contrary to what counsel just suggested, there was indeed evidence here of network effects. And, in fact, Professors Stiglitz testified, and this is at page 1582 of the trial transcript, which is A533 in the record. Question, and U.S. Airways gets value from Sabre by having more travel agents on Sabre? Answer, well, Sabre is providing a service connecting the two. Yes. Question, is the answer yes? Answer, I mean more value in the sense it's connecting, making more connections. Yes. Now, why is that important evidence of indirect network effects? Because the record here also showed, and this is A1419, that only about 50% of all of the reservation transactions are actually done on GDS platforms, which means that half of all the people who travel in America every day are traveling based on tickets that are reserved in some other way, and that's an enormous opportunity for network effects. Because as you incent the travel agents with those incentives, half of the people who are traveling are not already on those platforms. If they book your trip. There's an assumption there that you're dealing with the same kind of customers, that business customers and me are the same, whereas, in fact, they're not. Business customers are going to go to a travel agent because they need the travel agent to do all kinds of things and take care of everything. I, on the other hand, have a wife, you know. Your Honor, it is true that more business travel takes place on platforms, but not all business travel. A large percentage doesn't, and a very large percentage of non-business travel. Do GDS platforms provide their booking information to Expedia and these other sites? Is there anything in the record about that? I don't know the answer to that, Your Honor. But my point is that in a market where a large percentage of the transactions are not taking place within the platforms that the plaintiff defines, and every time a new transaction takes place that's brought from outside in, that's an increase in output and, therefore, an increase in the business available to USAir and all the airlines. Of course there are network effects. My client wouldn't be spending all of that money incenting travel agents just for the heck of it. They're doing it because it incents them to book more flights. And the more flights they book, the more booking charges we can levy on the airlines and use that money to fund more incentives. That's the way a two-sided platform works. It would otherwise be a mindless exercise to be paying these incentives for nothing. In terms of whether this was tried as a one-sided case or a two-sided case, let me just point the Court to the instruction that the Court gave. This is at A688. On the antitrust injury in this case, here's what the District Court said. Quote, That is a one-sided injury. The booking fees are on the one side of the market that deals with the airlines. It takes no account of the benefits that the airlines get as a result of the incentives paid to the agents in the form of more booking. Wouldn't that be just a question of the proof at trial? I mean, it's going to say I was injured because I paid too much. My fees were too high. And you'll argue in summation, well, you know, they weren't too high. No, Your Honor. Because of the effects on the other side. No, if this were presented as a two-sided case, the Court could never have instructed the jury that that was the injury. It would have had to say that the injury was that the net price, the overall price to the market, to the agents and to the airlines, has increased because of a restriction of output. Saying that the injury was just the booking fees is exactly what the government argued in the Amex case. The merchant fees are higher because of our nondiscrimination provisions. And I said to the District Court on the first day of trial in my opening statement, if the only thing that this case is about is whether the merchants are paying more in exchange fees, in transaction fees, because of our provisions, we can stop now. They probably are. But you're ignoring the other side of the market. It's what I said in the Supreme Court when I said to one of the justices that she was clapping with one hand because she was only looking at the fees to the merchants and not to the benefits to the consumers. The District Court here instructed the jury on exactly the problem that arose in the Amex case. It was a one-sided injury, and then at the end they were asked hypothetically, suppose if it was a two-sided market, would your injury conclusion be different? And they said no, it wouldn't be. That wasn't enough to fix it. My last point, if this Court has spent this much time trying to figure out the difference between one-sided and two-sided platforms, do you honestly believe that those jurors were able to sort all of this confusion out and come to a proper verdict when it was presented in this form? We make that kind of assumption all the time, Tom. Well, Your Honor, I ask respectfully that in this case you make the assumption that it was darn near impossible. This verdict must be overturned. We have a reversible error here, and we respectfully request that this Court find so. Thank you. Thank you both. Very well argued. Very helpful.