Our own Justin Ellis attended today's hearing at the 10th Circuit Court of Appeals on Novell's appeal in Novell v. Microsoft. This is the antitrust litigation Novell brought over WordPerfect. He has a report for us. He begins with his general impressions, and then provides his notes on the arguments.
To help you follow along, here are some resources:
His general impression is that Microsoft will prevail, as the judges seemed more positive toward its arguments. But keep in mind that you can't always tell what judges are thinking from their questions.
He also felt that David Boies didn't do as well as David Tulchin, for Microsoft. But on reading the report, I thought Boies' rebuttal was strong. Mr. Tulchin is widely regarded as one of the best, though. So we'll have to wait and see. The impression I get from Justin's notes is that monopolists can be about as evil as they want to be in the 10 Circuit.
I know you join me in thanking Justin for taking time out so we could find out what happened. The 10th Circuit doesn't always make transcripts available, so this is as good as it gets. When the decision comes down, Justin's notes will make it possible for us to comprehend it. Here are Justin's notes and general impresssions:
The oral arguments for Novell v. Microsoft took place in Courtroom 2 of the Byron White U.S. Courthouse, which is the United States Court of Appeals for the Tenth Circuit located in Denver, CO. The presiding judges were Hon. Neil M. Gorusch, Hon. Paul Joseph Kelly, Jr. and Hon. Jerome A. Holmes.
David Boies from Boies, Schiller & Flexner argued for Novell, and David Tulchin from Sullivan & Cromwell argued for Microsoft. There were about 30 spectators in the courtroom, most of them (presumably) attorneys that were there to argue other cases (Novell v. Microsoft was the first argument scheduled for the day).
It seemed that the major focus throughout almost the entirety of the arguments dealt with antitrust concepts fleshed out in Verizon v. Law Offices of Curtis V. Trinko, LLP, 540 U.S. 398 (2004) and Aspen Skiing Co. v. Aspen Highlands Skiing Corp., 472 U.S. 585 (1985). Mainly that (as I understand having had only a limited amount of time to review these and other cited cases), since a company has no obligation to deal with a competitor and, since a company will of course engage in activities to the detriment of its competitors in order to compete in the marketplace, antitrust allegations based on a company's refusal to deal with another company need to be supported by evidence that the Defendant company's decision not to deal does not make sense from a business standpoint or, as they often put it, that the Defendant company is forgoing profits or advantages in the short-term in order to harm its competitor in the long term.
Two other cases were oft-cited in the discussions: Four Corners Nephrology Assocs., P.C. v. Mercy Med. Ctr. Of Durango, 582 F.3d 1216 (10th Cir. 2009), Christy Sports, LLC v. Deer Valley Resort Co., 555 F.3d 1216 (10th Cir. 2009). From what I could tell (not having read all of these opinions in great detail), they were usually discussed in the context of how how they fit in with, and differentiate from, Trinko and Aspen. [PJ: The best way to understand the cases is to read them and then to read the parties' briefs, where they will be used to establish their points.}
As for the general feel of the courtroom during the arguments, I heard it said in the media not too long ago that David Boies, in another case in which he was arguing before another court, appeared to be ill-prepared and outclassed by his rival. It pains me to say that, in this instance, I got somewhat the same feeling.
Mr. Boies started his argument off by recapping the factual history and explaining what APIs are for, information that was already covered in the parties' respective briefs (and, of course, in the record of the original case). I felt that Mr. Boies' limited amount of time would have been better spent by immediately addressing the items before the Court and delving into the legal arguments (you can even see that Judge Gorusch interrupts Mr. Boies early on with questions going more toward the heart of the matter). Mr. Tulchin, on the other hand, got right to the arguments of law and was immediately citing case law to back up his arguments.
Of the three judges on the panel, Judge Gorusch was easily the most active of the three, often interrupting with questions and talking back and forth with the attorneys and with the other judges. Second was Judge Holmes, with Judge Kelly participating the least (Judge Kelly did not ask a single question or even say a word during Microsoft's arguments). The general feeling that I got was that all three of the judges are leaning heavily toward agreeing with Microsoft's side of things, which I think shows through in the notes I took during the arguments.
What follows are notes that I took during the arguments, as best I could in trying to keep up with what was going on (there is a lot of paraphrasing mixed in with direct quotes, with little notation to distinguish the two). There will probably be a lot of spelling errors, snippets that don't make sense and the like, but I've tried to clean it up as much as possible so that it hopefully will not be too hard for everyone to follow along. Anything in brackets are my own notes or observations.
David Boies (for Novell):
- Reserves 3 minutes for rebuttal
Boies: In 1993, Microsoft decided it would publish and document certain APIs so that other developers could attach their application programs to Windows [explains that APIs are the way that developers can attach their own applications to an OS or something else]. In early 1993, key Microsoft executives recognized and warned that doing this would allow Novell and other competitors to battle "on even terms;" nevertheless, Microsoft confirmed to Novell [and presumably others] that it would make them available. For a year, Novell...
Gorusch: [interrupting] After Trinko we're stuck with a test that requires that a monopolist engages in activities that don't make sense, that forgo short term profits; what is on the record that Microsoft intended anything anti-competitive?
Boies: Plaintiff exhibit 1, 1994 email said it was a "tough decision" not to publish and document the APIs even though they are valuable and "a nice piece of work" because withdrawing would make it harder...
Gorusch: That is a nice piece of intent, but after Trinko we don't just ask about intent, with good reason, because we expect competitors to have bad intent, so what else is there?
Boies: The District court [in the original Novell v. Microsoft case] recognized that by removing the APIs Microsoft reduced the sales of Windows.
Gorusch: But you can argue that it increased sales of Word on the other side [so the net effect was a solid business decision].
Boies: The only increase comes from the anticompetitive effect; there is nothing in the record that says that they are doing it for profit.
Gorusch: [interrupting] That's intent again, I'm asking about effect [of foregone profits]. Is that email your best piece?
Boies: There are better pieces, I will address [trying to move on from the email]. They took out the APIs in 1994, put them back in '96; they were actually there all along; they told the public and the developer companies that they were not going to use the APIs, but that was not true; they continued to use them and brought them back after Novell's share had gone down [6%? Something-6%]
Kelly: Well, Novell was aware that it was subject to change, right? Didn't they determine through their own FOD[??]? If they had gotten it done in timely manner there wouldn't have been any problem...
Boies: No, they would have gotten it done timely had they been told they couldn't rely [on the APIs].
Kelly: But they were told it was subject to change.
Gorusch: [To Kelly] Subject to change but not for a bad reason.
Boies: For bad reasons, but also for bad effect; Microsoft sucked Novell in and Microsoft knew that they had to have these APIs.
Holmes: I don't understand where these lines come from; from the beginning, Novell knew it was a beta, there was documentation expressing that these APIs might not exist.
Boies: Yes and no. One of the things that Microsoft did, and that the court indicated that the jury could have found, was that when Microsoft first announced (look at JA 604-605, testimony from Microsoft witness) he says you are putting a stake in the ground, that people can rely on it and that it will be available when they say it will be available; they were telling the world they were relying on it, unless something happened that necessitated a change...
Gorusch: So then you have a breach of contract action, but with antitrust it is different, which Trinko and ?? [missed the name of the case] tell us.
Boies: But there are also antitrust violations if there is anti-competition, which here is palpable; back then you had only two word processing programs; by 1996 Novell's share was drastically reduced, all because of this action.
Holmes: We are only focused on whether the OS market was impacted, correct?
Boies: Yes, correct.
Holmes: Why are we talking about word processing?
Boies: Novell's word processor was recognized by Microsoft as being competitive to the OS itself because it allowed competitors to attach software; [lists evidence of MS executives saying that Word Perfect was, "scary, since it's just another windowing API and fairly complete”].
Holmes: By virtue of having one monopoly allows you to leverage another market; even if you take Bill Gates talking about taking a hit against their OS so that they could in the long term have better products and increase market share in applications, where is the anti-competitive behavior? Four Corners says that that kind of leveraging itself is not anti-competitive.
Boies: What they are saying is that by giving up short term profits, they are eliminating competition for the OS market, which is exactly what the 4th Circuit talks about.
[Mr. Boies' time runs out, he sits down]
David Tulchin (For Microsoft):
Tulchin: The assertion that a change to a beta OS under development could be anti-competitive is refuted by Christy Sports and by Four Corners (2009 cases in the 10th circuit). In Four Corners it is written that a monopolist has no duty to deal with its competitors or under terms and conditions that the competition finds advantageous.
Gorusch: Yes, but this court also includes Aspen Skiing; my question is the inverse of the one for Novell; why don't we have evidence that Microsoft is forgoing short term profits to kill a rival? We have a lot of intent evidence.
Tulchin: Two reasons. Judge Motz [the District judge] said in his opinion there is no evidence that Microsoft forewent any short term profits; there would have to be some evidence that the balance between fewer sales in Windows and more in applications tips in that direction.
Gorusch: But there is some evidence [Mr. Frankenberg said] that some people would leave Windows because Novell was “sticky” [people would find Novell's products so useful that they would move to another OS rather than stay with Windows].
Tulchin: He said that Windows sales would have been higher, but Novell offered no evidence or testimony [to support that assertion].
Gorusch: What if we aggregate Microsoft's profits in those two markets?
Tulchin: It's common sense that if you're talking about foregoing short term profits (as in Aspen, which is a very different case), you need some evidence to support that (even if just from an expert); Knowell [Spelling? Presumably an expert witness in the case before the District Court] testified for days but not on this subject. [PJ: Perhaps it was Mr. Harral, whose testimony you can find here.] Also, neither Gates nor anyone at Microsoft could have made a calculation that Microsoft was foregoing short term profits (as they were able to do in Aspen) because Microsoft didn't even know that Novell was going to use the namespace API; on the contrary, Mr. Scruss' [sp?] testimony said that his survey of important ISDs said that Novell did NOT intend to use or rely on these API's; as Judge Motz said, [after this testimony] Novell came forward with no evidence to rebut the proposition that Microsoft could have no idea that removing the APIs would delay Novell.
Gorusch: What is interesting is that most cases look at business justification; except here that we do not have a business justification.
Tulchin: Except that it's an assumption the entire argument is based on.
Gorusch: I agree, but do we have any case law that addresses the absence of a business justification?
Tulchin: No, the decision in Four Corners stated that the key fact in Aspen was that the Defendant terminated a profitable relationship by denying to a rival terms that were available to all other comers; here there wasn't…
Gorusch: [interrupting] Let's not go down that road, which is not very interesting for you; when Gates pulled the rug, were there non-rivals in that first group of participants or were they all rivals?
Tulchin: The survey that Brad Scruss took in September was to figure out who among the ISVs were going to use the namespace extensions; Microsoft wanted to know whether removal would have adverse impact on rivals before Gates made the decision, because MS wanted people to come out with applications on time; Novell was one of the companies to say unambiguously that they weren't using them, so Microsoft could not have induced them to use it. When the beta was released, it was pursuant to license agreements saying that the betas can and might change, and there were features and functionality that do not represent commitments by Microsoft. [PJ: Brad Struss reported what he reported, but as you can see in Novell's reply brief, that's not the same as Novell actually saying it. And the judge ruled for Microsoft as a matter of law, and if there are disputed facts, that's not supposed to happen. What Novell is asking for is a trial before a jury. Here's what they wrote in the brief: " Instead, Microsoft relies on testimony and emails from Microsoft witness Brad Struss that, at most, created a disputed issue of fact and had little to no probative value even apart from the standard of review and contrary evidence detailed above. The email stating that Novell was purportedly "OK" with Gates' decision followed Microsoft's use of the deceptive script, which was part of the same email chain. See Opening Br. 38 n.5. Struss' testimony also was inherently suspect because he could not say whether Novell was one of the ISVs he later identified in an email as "actively developing" using the namespace extension APIs. See id. at 38 (citing JA-14298 (Struss))." And Robert Frankenberg testified that he was not aware of anyone having that conversation with Novell.]
Holmes: That sort of disclaimer is not an aberration, that is industry practice, right?
Tulchin: Yes, it is.
Holmes: So the notion that there would be reliance doesn't hold water.
Tulchin: Correct, even Novell's witness said that it was industry practice that things could or might change.
Holmes: [Four Corners and leveraging a monopoly]: Even if you take the Gates memo on face value (that in the short run Gates was willing to delay releasing the APIs *because* he wanted to be able to increase profitability in the application market), this is someone structuring as something advantageous in the long run; when I look at Four Corners, there is nothing inherently anti-competitive about that.
Tulchin: I agree; Novell's position is that, once a beta is distributed, notwithstanding industry practice that betas can and do change, that the software is locked in forever into the design of that beta, and that no changes can be made; the beta was distributed in June, 1994 to literally tens of thousands of ISVs; Novell is saying that any one of those ISVs could dictate design decisions based on what was released.
Gorusch: You are arguing a breach of contract case also, which we're not here to decide; let me ask about Quattro Pro; I see Gibb's testimony that Quattro is ready to go… [unintelligible] [PJ: You can read Gary Gibb's testimony here.
Tulchin: Novell relied in summation and its brief today; there are only 2-3 out of 90 pages in its briefs as to whether Quattro Pro caused the delay, but in 231 (joint appendix 9147) [not sure these references are correct], the document shows a co-complete date for Quattro Pro (actually for Perfect Storm, the suite) of October 31, 1995; the same document has a release to manufacturing date…
Gorusch: [Interrupting] I know you have evidence, but we have a Rule 50 to ask if any reasonable jury could [could find for the Appellant]; isn't Frakenberg's testimony enough?
Tulchin: No, when he saw all the programmers were leaving, he said clearly that Quattro Pro was not then complete.
Gorusch: But he comes back on redirect and kind of backtracks; he says that Gibbs would know better.
Tulchin: He does, but the law in this circuit (Herrara v. Lufkin Industries, Inc., 474 F.3d 675 (10th Cir. 2007)) was not literally that there was no evidence supporting the non-moving party, but whether there was “more than a scintilla” of evidence a jury could have found; here it is clear that Gibb's testimony is a mere scintilla. Take the testimony of Nolan Larson; he went to Scotts Valley CA and said it was a train wreck, that the product wasn't even close to being ready yet.
Gorusch: Back to the Trinko problem; it talks about a willingness to accept short term losses; could Gates' email show a willingness to accept the loss, even if Microsoft didn't incur it?
Tulchin: I don't think it can be read that way; his email makes no reference one way or the other to the impact of the OS market on the changes; it would be peculiar for him to consider that delay of Novell in coming out with applications could adversely affect Microsoft's position in the OS market, which was only going to be stronger. In the 4th paragraph of the email “making it harder for Novell” [maybe not a direct quote] could be read as evidence of impacting the applications market, but there are no actual claims as to the impact of the applications market.
[Mr. Tulchin's time is up and he sits down]
David Boies comes forward for rebuttal:
Boies: [says he wants to address a few points] To address Quattro, it's not just Mr. Gibb's testimony; evidence shows code complete of August 28, 1995, long before the programmers left; the reason it was delayed, indicated also in the testimony, was because of the delay in the shared code as a result of the extensions [he appears to mean, as stated in more detail shortly, that Quattro testing was waiting on common code, which was eventually delayed because of the API switch].
Boies: Quattro was already done, as indicated by this document and Gibbs' testimony; beta testing hadn't happened because it was to be tested with WordPerfect, which was delayed because of the libraries. To address another point, counsel also said Microsoft didn't know Novell was using the namespace APIs. In the Joint Appendix 3691 is an October document that explicitly says that Novell is using the API. Third, this is not a question of things being changed, where they simply changed the beta, this is a situation where they did NOT change the beta; they left it where it was but lied about what they did; they said they were taking the APIs out, but they left them in there, they used them themselves, and then in 1996 after July, after Novell and WordPerfect were destroyed, they came back and said "they're here again."
Holmes: But that just goes to whether MS was acting in a good way, but the fact is still that the beta can change, not whether they can lie about it or not.
Boies: But there is no expectation that the beta would change falsely.
Holmes: A lie is not the same as anti-competitive conduct.
Boies: It can be; for example the 4th Circuit in 2007 talked about withholding information; that has nothing to do with lost profit, but the circuit held it was anticompetitive; D.C. said you don't have to lose profits in order... [did not hear the case names here, if he stated them]
Gorusch: It's different when you're relying on a monopolist's property for your own business; that distinguishes it; generally each entity is free to use its own property [how it chooses]. Why didn't Mr. Gates actually do WordPerfect a favor by excluding them from Win 95 if Perfect Office Suite was so revolutionary and can be used on other OS, why wouldn't that bring Microsoft's share down?
Boies: Because in order to develop wide enough [stammering] usage to be a competitor in the OS market the APIs have to be widely distributed.
Gorusch: Isn't that a railroad, have to be on monopolist's property in order to compete with him?
Boies: With middleware, if you're going to compete…
Gorusch: I'm not talking about middleware theory, I'm talking about franchise theory.
Boies: If someone like Microsoft has 95% of the market and you are shut out…
Gorusch: It's an essential facility?
Boies: It's not an essential facility but you have to have access it in order to succeed; it's essential in the sense that no access will keep you from wide distribution. You can't build up that franchise with the other ten percent of the marketshare.
[Mr. Boies' time is up]