 We're talking about integration, product improvement, tying one of the thorniest parts of this case, one of the least resolved issues of this case in some ways, given that the DC Circuit sent it back and the Justice Department decided not to retry it. Lots of unanswered questions left that I hope we'll have a chance to really get into and introduce the moderator of the panel and let him take over from there. Professor Andrew Chin, an associate professor at the University of North Carolina teaching intellectual property and antitrust in addition to his law degree, also as a doctorate in computer science. And during the trial of this case, he was the law clerk to Judge Kennedy in the District Court in DC and spent a lot of time helping Judge Jackson's chambers on some of the technical issues. It's written about the case, written about this particular issue, and it's a really good person to help try to guide our panel as they try to guide us through some of the issues. So with that, Andrew. Thank you very much, Phil. I'm delighted to be on this all-star panel with very impressive colleagues who I only got to observe in silence from the jury box during the trial as a law clerk succumbed onto the case. And I described this courtroom sketch as the foul ball I caught at the World Series. I'm delighted to be joined in the order they'll be speaking. Steve Holtzman, who's a partner at Boy's Children Flexner and was a senior member of the trial team for the Justice Department. Dave Heiner, who's vice president and deputy general counsel of Microsoft. And Ed Felton, professor of computer science and public affairs at Princeton University who was an instrumental part of the government's case on tying and in fact was examined by both Dave and Steve. And I guess another reason, you know, the talk of the town is the Section 2 report from the Justice Department this week and another reason why this still not fully adjudicated time claim continues to be of currency, I think, from the Microsoft case is that if the Justice Department's recommendations that the nominal per se rule against tying be finally abolished in favor of a rule of reason and if they can go further, the disproportionate standard is used to weigh a claimed pro-competitive benefit against anti-competitive harm. Those developments will have been inspired in large part, I think, by the approach the DC Circuit took in Microsoft 3. So these are just a couple of exchanges from the call a quiz involving our panelists. So one of the odd aspects of Judge Jackson's management of the case was that direct testimony was submitted in writing and so we didn't get direct examination in terms of courtroom testimony from Professor Felton until the summer during the rebuttal part of the case and so Steve was able to bring out the Justice Department's theory of Professor Felton's contribution to the case, technical contribution to the case which was a suite of programs which he argued was a proof-and-concept that demonstrated that Microsoft could deliver a version of Windows 98 from which the Internet Explorer web browser product had been removed and could be done so without disturbing the non-web browsing functionality of Windows 98. Microsoft sought to pin Ed down on what code represented the Internet browser product in Windows 98 and in particular argued that four files of called dynamic linked libraries which are used to provide platform code to a number of functionalities in Windows 98 including web browsing and the rendering engine within the browser and other fundamental aspects of the web browser. All of those are supported by common DLLs but they're also used to support other functionalities and new services in the operating system some of which are relied on by third-party applications developers and so Dave cross-examined Ed during the case in chief and they had this exchange over whether these four files constituted part of the product and Ed stuck to the definition that he gave of the web browser in Windows 98 and it's an implicit definition. It tells more what the product does, the browser product does rather than what it is. So his definition is that which allows the user to browse the web which may seem circular but as I'll have comments later on that dig a little more deeply into that definition. So Microsoft's position in its proposed findings of fact was that the software product consists of code and nothing else and why this sort of gives the foundation for what the DC circuit did and what may provide the inspiration for a move to a rule of reason is that the DC circuit took from the idea that a software product consists of code the notion that the OS and the browser product in Windows 98 are integrated not in the sense that's given in the 1994 consent decree but in the sense that two products have been made one physically by virtue of being delivered in the same set of code that's put on the CD ROMs and that this provides the pro competitive benefit that is not properly accounted for in the per se analysis of time claims and in fact would be enough to provide a justification under the rule of reason and I note that Dave has supported the DC circuit's analysis in recent symposium article in the University of Chicago overview so with that I'll leave it to Steve to discuss the Justice Department's theory of the time claim and put that in the perspective of developments in time law I actually was going to march through what the Justice Department how we approach time claims but when the Justice Department, today Justice Department issued its report on single firm conduct earlier this week it sort of changed my plan because it was an opportunity too good to pass up so we really talked about the contrast between then and now and as a way of framing things going forward so as most of you probably know this past Monday the department did issue a new very long statement on monopolization single firm conduct under the Sherman Act one thing that's interesting at the outset about the new report is that it covers tying at all and it's interesting for the maybe superficial reason that tying is ordinarily thought of not under section two of the Sherman Act but under section one not a single firm conduct but as multi firm conduct it's also interesting because in the Microsoft case tying was really an issue under both section two and section one and in the DC circuit opinion in 2001 that Andrew I think alluded to it's interesting because the DC circuit really came out different ways on the same conduct between section two and part of the section two analysis and in looking at the section one analysis so it seems clear from all these things that tying particularly when it's looked at under section one in regard to which the DC circuit held that you shouldn't apply per se rules in the circumstances of this case but should instead apply the rule of reason it seems to put people start tying them into knots contortions to figure out how we ought to deal with these thorny issues of coercion and integration and the benefits and harms that flow from them and one thing I think it's worth asking actually sort of the jurisprudential question is why it is under section one that the tying claims in this case and in others do seem to tie people and courts commentators in knots I think one answer may be because so much of the development of the law actually had been stifled by the existence of a per se rule when you have bright line rules like that the deal in absolutes it makes it difficult for courts and litigants to figure out really what are the issues that should drive the analysis under section two when the DC circuit looked at the fundamental tying claims the conduct that underlie the tying claims it was able to sort of march through it had this balancing test set up and it was able to march through really very expeditiously the pros and the cons and say this is anti competitive it's part of a section two violation but when section one came up had the per se rule the court immediately focused on really had to focus on the sole issue of whether a per se rule ought to apply and this frankly prevented the DC circuit from getting to the substantive issues sent it all back said there shouldn't be a per se rule and of course that was the end of it and as a result the law still hasn't developed under section one so that's where we are now DOJ is still struggling with them the FTC and others are still struggling with the tying issues and in fact it's now DOJ has even apparently exported this these contortions to section two because it's discussed this in the context of single firm monopolization so very briefly what in the report this week has the DOJ said about tying first of all it has very strongly reiterated the view that no per se rule should apply to tying at all moving well beyond the cases of technological integration so that's number one number two the DOJ has seemingly compartmentalized the way of thinking about tying it talks about contractual time talks about bundled ties it talks about technological time so that's number two number three as a general matter regardless of which type of tie it's talking about the department says that plaintiffs have to prove harm that is substantially disproportionate not just outweighs but substantially disproportionate to pro-competitive benefits and then with regard to technological tying it focuses on it defines technological tying as quote a tie achieved through integration of what could be viewed as two products so that's the definition and then once it has a tie which it chooses to characterize as a technological tie its commentary is that technological ties often ubiquitously in fact give consumers features they want it says that judicial intervention into product design risks chilling innovation and that as a result technological tying should be condemned only in exceptional cases where integration which again the definition of the technological tie is integration serves absolutely no purpose other than to harm competitors so it's actually to those of us who litigated the Microsoft Microsoft case it sounds fairly familiar because it was the other side's position so obviously we have a bit of an evolution between from then and now in the government's position which among others the Federal Trade Commission has been at pains this week to point out so just a very few quick comments because I think that on an issue like this which admittedly does involve very complicated questions it's best to have the discussion guided by questions and answers but going through very quickly the current DOJ's view of tying I just want to make a few quick observations as to how it contrasts with the way that we who tried the case on the government side looked at the tying claims first of all I talked about how the DOJ seems to want to compartmentalize the tying analysis and it wants to do it in two ways number one there's nothing in the DOJ's report which talks about looking at any particular element of conduct in concert with other elements of a defendant's conduct whereas to us who tried the case it was crucial that each element of the case each element of the alleged anti-competitive conduct be viewed in concert with all the others it would be fundamentally misleading we thought to consider the anti-competitive effects of any one element of conduct without considering the cumulative results of all of them DOJ now also compartmentalizes the analysis by drawing these definitional distinctions between contractual ties, bundled ties, technological ties and on this one again having thought a great deal as we led up to the trial we very much wanted to resist the pigeon-holing of the tying claims into one or another of those three in fact our view was and we presented evidence at trial that it was all three of them and that you couldn't just look and shouldn't just look at the case as a question of technological integration because you had these contractual elements terms in the licenses that Microsoft did with OEMs and of course you had this issue of the non-removability of Internet Explorer from Windows which to us very much distinguished the case from earlier cases in which you simply had a fact of integration without a coercive element so that's the compartmentalization part now the second thing I mentioned is that the DOJ now is suggesting that tying should be condemned only when the harms are substantially disproportionate their words to pro-competitive benefits and I think my only observation there this one doesn't relate so much to what we were thinking in doing the case although I will note that whether a per se rule or the rule of reason applied to the tying claims whether it's part of section two or section one in our case was not of any great moment to us because we were very fully prepared in part because of the DC circuit's earlier opinion with regard to the earlier consent decree for the fact that we very well may have to face a very high burden of proving anti-competitive effect of proving an absence of pro-competitive purpose and intent so again we didn't tend to think per se rule of reason and of course the other one of the reasons for that is that we were thinking primarily in terms of section two where we knew there would have to be some form of balancing analysis as opposed to section one we didn't really distinguish the two but the other key observation about the DOJ's substantially disproportionate test is that it does seem to go quite far beyond even the rule of reason as it has been articulated elsewhere including by the DC circuit in this case where the DC circuit is talked about you simply have to show that the anti-competitive effects outweigh not be substantially disproportionate but outweigh the benefits and then third with respect to technological tying the key point here I want to focus on is the definition again a tie as DOJ says now that is achieved through integration of what could be viewed as two products and I think this one is key because I think the way you define a technological tie may or may not may probably will predetermine the outcome if one is to intend on looking at a technological tie as simply integration then it's very likely because the Justice Department and others have observed that this is something that is very common in fact ubiquitous particularly in rapidly evolving markets it's going to be very difficult to show that there aren't benefits or that the harm substantially outweigh the benefits or so forth our view was, my view remains today that it's a misnomer to simply say technological tie equals integration it's much more important to focus in my view and our view then on the elements of coercion the tie itself is there evidence that you have forcing not only integration which sometimes very much does provide benefits but you have forcing do you have things like these contractual restrictions do you have non-removability do you have the fact that the benefits that this integration provides are equally available when you have the products delivered separately as opposed to together you know certainly we didn't go into this case and we didn't try this case thinking that we by any means wanted to chill innovation but this was precisely why we talked about things like what other operating system vendors did on the one hand Microsoft put on some very good very compelling evidence that what's the problem here all the operating system vendors in the world are doing the same kind of integration it's very clear as the web becomes more and more important more ubiquitous we all want to integrate the browsing functionality into the operating system or at least bundle it with the operating system whereas our point of view was wait a second let's really look firms without power how are they doing it well in every case and Ed demonstrated this in his testimony the other operating system vendors did bundle a bundle browsers but did so in extraordinarily modular removable way in many cases using browsers that they hadn't even developed them not many but in one or two cases the browsers they hadn't even developed themselves and so what this showed to us was there was a stunning lack of benefit and on the harm side it was very important just to finish up that the harm here you need to look at not only the tied product market not only the browser market but much more importantly the tying product market as well and other panels later today I'm sure we'll talk about that in greater detail the idea here that you have so much of the tying activity in order to protect the monopoly that Microsoft had in the operating system market so I think I'll stop there hopefully that was coherent enough but mostly wanted to sort of frame the issues again then versus now and as obvious you have a very very different world view I think that I was starting starting to write some notes about the question of well would this case have been brought today but it didn't seem worthy of discussion because not only do you have this analysis on time you also have a DOJ that has not brought any monopolization cases at all so I'll leave it with that provocative comment okay next we have Dave Heiner of Microsoft who will give some more perspectives on the meaning of integration and how the American and European antitrust agencies are evolving approaches to that phenomenon in the software industry well thanks Andrew and I want to thank Phil for inviting me and my Microsoft colleagues to this event and thank the Berkman Center I have to say I do feel a little bit like an intruder at the reunion of the winning football team from ten years ago but it is good to be here during your walk down memory lane yesterday as you went through Microsoft 1 and Microsoft 2 and so forth it reminded me of when I first joined Microsoft which was in August of 1994 in the year preceding that was the year that the DOJ had picked up the investigation from the FTC and it was quite a big deal at the time and it settled in July of 1994 and I got to work in August and my job was to be the antitrust lawyer at Microsoft the first sort of full-time dedicated antitrust person I was looking forward to working on this matter and I got to work and said congratulations Bill Newcomb on settling the case personally I'm a little disappointed I was looking forward to working on it and he said there will be other matters I said yeah turned out there was a little more than I ever had in mind excuse me what I'd like to do now is just focus a little bit on the remedy specifically as it affects the design of windows and with respect to in particular the removable that Steve alluded to and the coercion element and then compare how it came out in the US and how it came out in the European case so I think there were really two categories of considerations here that were important the first is really an end user perspective and this is the promotion of the browser on new PCs Professor Zitrin talked about this yesterday and quite accurately explained the concerns from the DOJ side the Microsoft browser was built into windows and therefore it would appear on the desktop in every copy of windows and therefore on 90% of PCs Microsoft's point was yes but there was no exclusion in the sense that there was no contract prohibiting OEMs from shipping the Netscape browser and in fact there was testimony that Sony for instance was shipping a machine with four browsers on it but the Justice Department and the states explained that well in their view actually there was a disincentive for OEMs to ship that additional browser because consumers might be confused if they want to get to the internet and there's two icons there to get to the internet which one would you click and the court agreed with that testimony and that was a critical finding in the case that this risk of consumer confusion if there were multiple browsers presented to the end user the other aspect of the case that was much discussed was the platform aspects of the browser and this was important of course because the theory of the case was that Netscape might itself one day develop into a real platform that could support running applications now Netscape never actually did the work to make that platform come to fruition as Phil was saying it was a nascent possibility but there's some irony in that Microsoft did that work and so in the 96, 97, 98 timeframe it componentized the browser meaning broke apart its functions the ability to display HTML and the ability to handle the internet protocols and made them available to the developer community and that of course is what Windows is it's a platform for running applications and so Microsoft put a lot of emphasis at trial on isn't this a benefit we're making the platform richer to enable it to be relevant in the internet age and one of the points Microsoft emphasized which was much discussed then in court at all the different levels of the proceedings was that developers want to know that if they write an application it will run on every copy of Windows on every PC and that way if an application is written it will run on an HP machine and on a Dell machine because all the code will be there and then users will know they can buy an HP machine or Dell machine and the applications will run so this is the compatibility effect which then gives rise to the network effect which was much discussed and so the question is sort of how do you address these considerations there's the risk of confusion at the promotion level and yet there seems to be at least in Microsoft's view this platform benefit and the issue actually looms even larger than it sounds like from just discussing the browser case because the Justice Department in the States were saying this is not just about the browser this is potentially about other features as well and there's quite a few features in a modern operating system you know I think it was Professor Arita probably in the 70s who observed that any product can be conceived of as sort of the sum of its parts you know a car has a tire and this slide is just showing across the rows some features of operating systems as of a couple years ago when I did the slide and the columns are four different operating systems the Mac OS and three versions of Linux and you can see there's a lot of checkboxes and it'd probably be more today if I updated it just showing as a quite a wide range of features so when we're thinking about how should we handle the browser we're potentially thinking about many many many features in the operating system so what was the solution that was arrived at well the solution worked out by the Justice Department in the States and then implemented by Professor Felton was really to address the promotion advantage while trying to preserve and in fact preserving the platform benefits and so Professor Felton built his prototype removal program to demonstrate that it was entirely possible which wasn't really disputed but it was a possible to remove the means of accessing the browser and then at that point as Professor Felton emphasized it's as good as gone from an end user perspective and an OEM could then ship a machine if the tool had been run or if Microsoft had provided that ability and install Netscape and it would appear as if Netscape were the only browser on the system no user would think otherwise at the same time all of the underlying code these are the DLLs that Professor Chin was referring to they're all still there and so they're providing the benefits to the developers and they're providing benefits to other parts of Windows so that seemed like a sensible solution in the view of the court it's the basis on which the district court ruled and on which the appellate court considered the subject and it was then the basis of the remedy so the remedy was that Microsoft had to create a tool that would enable the removable that Steve Holtzman was referring to that tool is in the system today both in Windows XP and in Windows Vista and OEMs are authorized to use it as well and so it enables the creation of a system in which it appears that the browser is not present now that was then backed up by licensing terms as well so OEMs have a set of rights authorizing them to use those tools to install any browser they want on the system to promote that browser exclusively and to set it as the default browser which is just the browser that would come up if the user clicks a link that's an internet link so the remedy was really very well tuned to the liability finding risk of confusion and the risk of confusion was eliminated by what was worked out by the settlement process and then Judge Cotelli in the litigated remedies phase so what can we say about this approach more generally and I think what's interesting is that the focus of the consent decree with respect to this issue that the time, the design of Windows and other respects is all focused on increasing opportunities for competitors as opposed to constraining the design of Windows so the idea was Netscape ought to be able to be installed by OEMs on PCs with no disincentive it ought to install well it ought to be able to be the default browser and the benefit here really is that consumers then get a choice of an integrated solution offered by Microsoft if they want and OEMs have that choice or a standalone solution, Netscape, Mozilla, Firefox, whatever it may be at the same time the platform benefits are fully preserved one of the benefits from a legal perspective of nine states declining to join in the settlement moving on for the additional set of remedies that all these issues were explored sort of again in the litigated remedies phase and that went up on appeal again and in that litigated remedies ruling by Judge Cotelli she really emphasized that the appellate court had emphasized the importance of the platform benefits and she entered additional factual findings saying that software developers and consumers would really be harmed if code were removed from Windows if you had a solution where you were actually removing these blocks of code that constitute the platform value of the browser so that's the general approach and in the consent decree it's backed up by a number of additional provisions relating to contracts relating to interoperability relating to OEM relationships and each and every one of those is focused on creating opportunities for other browser vendors not so much saying Microsoft should not design Windows in one way or another and then the other level of generalization is as I was alluding to before two additional features and so it applies to email and instant messaging and media players and potentially other features as well so what's happened with this since 2001 which is when the decree was voluntarily put into effect by Microsoft in 2002 it became formalized Microsoft has then taken the decree approach and generalized it further and so as we think about new features in the operating system even if they're not strictly speaking governed by the decree we say is there a way to build a feature into the operating system and at the same time ensure that competitors can offer something comparable that will work very well with the system and integrate in a nice way and so an example here might be for instance the way web search works in Internet Explorer 7 and you know it's a box that got created up there in the upper right hand corner to really promote web search and it's very easy to run a search against one search provider say Yahoo and then with one click you can run it against Google and then you can run the same search against Live Microsoft Live you can run multiple searches quickly you can change the default very easily everyone sort of plugs in in a very convenient way similarly we have a security center in Windows Vista that was a big push part of the Vista development Better Security and we really worked with Symantec and McAfee and with some regulatory agencies getting feedback through that loop on ways to design that system so that it would report is McAfee running, is it running correctly is the Symantec firewall working properly if not go get a firewall so that sort of thing we then took that approach and presented that in a speech Brad Smith gave a couple of years ago in DC as a set of principles that we would live by going forward that generalized the decree approach and extended sort of in perpetuity that this is the way we will design Windows so that computer manufacturers know software developers know they can rely upon these principles as they go about their business and this is a commitment to the industry these principles are up on our website for all to see the other aspect that comment upon is features not included in Windows you know there was some concern during the trial that Windows could sort of become the blob that takes over the earth as all interesting software and some categories would just be built in over time and the extreme version of this was often said what would prevent Microsoft from say combining office with Windows and Phil and I used to discuss that from time to time and it hasn't happened you know obviously and then some categories of software have actually come out of Windows or not been added and so for instance instant messaging software was a concern during the decree we ended up just removing that from Windows and offering it separately during the security push in connection with the design of Vista a careful thought was given to should there just be antivirus software built in out of the box you get your PC, you're safe, you're up and running and we thought about that hard and decided for a mix of reasons not to do that and so living under the rule of reason test of the DC circuit something that we're giving thought to all the time and how we design features into Windows or not into Windows and there'll be some more news about this in the coming weeks then in terms of the sort of industry reaction at point to one or two things one thing we see today for sure is a very wide range of software shipped on PCs in many many categories and this was one of the overall objectives of the case to ensure that OEMs would feel free to install any software they want, browsers or anything else and today there's so much software shipped on new PCs that it's actually created a new market opportunity for retailers to charge people a fee to take it off of the PCs which is kind of an astonishing thing Microsoft gets a royalty of something like $60, $70 for a copy of Windows and retailers are getting $30 just to go out and sort of clean up software off these machines for users who want that which is an interesting I think sort of contrast the other thing I'd point to is the growing use of web based applications and this is quite interesting I think in two respects I'm talking about things like Facebook now like in Pekasa all out on the internet but one is there's no possibility of integrating these things with Windows because they're not client code they're out there on the internet so they're inherently separate and that risk of sort of Windows incorporating them doesn't exist the other which is I think quite interesting is these applications are of course accessible via any browser running on any operating system which will tend to diminish the famous applications barrier to entry over time and of course it was actually that was the nascent threat posed by Netscape was that it would develop some platform and that was supposed to be kind of a client code platform as I understood it with Java and other technologies that would enable applications to run on multiple operating systems we're seeing that coming to fruition today in a somewhat different way which is based on internet standards and then there's kind of an irony here which is that by Microsoft enhancing Windows with support for these internet standards it really promotes the development of those applications so I wanted to just close I know I'm running a little here short on time but I just wanted to close then briefly with what happened over in Europe because that case has been quite interesting there's two aspects to the case I'll just focus on the tying aspect of the case and there the European Commission was focused not so much on the browser in this case but on Windows media player and so this is software that plays back audio and video and many of the arguments were on both sides somewhat familiar from the US case there was the promotion advantage that the commission emphasized if the media player was on every PC people would tend to use the Microsoft media player then there was the platform side and here Microsoft said you know in the 21st century you want to have audio and video in a modern operating system and applications developers want to be able to provide audio and video and so it makes a lot of sense to build that in and we pointed to the findings of Judge Cotelli saying in fact it would be positively harmful to the industry and to consumers to remove that platform code and the commission's response was essentially funny you would bring that up because the very thing that you're saying is pro-competitive we regard as anti-competitive and the reason is that if Microsoft is free to build in this platform code to Windows it will generate a network effect of applications using that code that's why Microsoft built it in after all which will then cause more people to use the client code and it will feed on itself and over time the Microsoft media player will be the only media player with any market share now this allegation was being made right around the time or right before the emergence of iTunes as the leading platform for running playing back music on the client side and Adobe Flash as the leading platform for audio and video on the internet but anyway that was the allegation that's what the court the commission found and as for remedy what the commission said was it's okay that Microsoft built the media player into Windows what's not okay was failing to provide a choice of a version of Windows without media player and so in the commission's view the logical remedy for tying where they conceived of the product as including the platform code was untying and that is logical and so the commission ordered Microsoft to create a version of Windows XP and then Windows Vista Windows media player and Microsoft did that sales of that have been negligible to date and so you know we have a situation where liability was found failing to produce a product for which it turns out there's no demand but that was the theory of the case and that was affirmed by the court in September 2007 the court not really taking account under their jurisprudence of the sales because it sort of came after the commission decision and they were just looking at the commission decision now just to wrap up then a comparison of the two cases in the US side of course it's a rule of reason test as the court of appeals ruled in large part focusing on the platform benefits that judge Jackson hadn't really given much weight to on the European side it's more akin to a per se approach this is somewhat controversial to say that commission itself might disagree they would say they looked at the range of factors the court ultimately said that in looking at the foreclosure side was there foreclosure of other media players that was kind of extra credit for the commission they didn't really need to do that and the real focus was just is there separate demand for media players and if so it's a tie and that really looks like a per se test to me in the US as I was saying the addition of the platform benefits as the court of appeals looked at and judge Cotelli seems to be a pro competitive conduct in the European Union that's anti competitive it seems and finally as a matter of remedy on the US side it's really focused on promoting opportunities for competitors not so much limiting the design of windows and on the European Union side it's more focused on taking code out of windows which is a concern to us as you then apply it to more than one feature thank you Dave finally we have Ed Felton who will speak to the government's theory of the case 10 years later and I'll get his slides up thanks that's what you should see 11 years ago if someone had suggested to me that I would willingly set foot in a law school on a Saturday morning I would not have believed them and if they had suggested that I was going to be speaking in that law school on a Saturday morning I certainly wouldn't have believed that but all of that started to change in October 1997 one day I was sitting in a crummy hotel room in Sunnyvale California and I dialed up to the internet and picked up my email and I got this this is from an attorney at the Department of Justice we're seeking a technical consultant for an investigation we're currently undertaking in the software industry I thought that must be Microsoft I'd appreciate it if you'd call me to discuss this matter etc etc I thought about it for a while and I called her and one thing led to another and here I am in a law school on a Saturday morning but I don't want to spend my time on nostalgia I prefer to spend it on hindsight so let me look what I want to do here is talk about the government's theory of the case the basic argument that the government made about the about the technology in the market and to look back on that with the benefit of knowing what's happened in the last 10 years in that market okay so very quickly let me review the government's theory first of all the government argued that Microsoft had monopoly power in the market for PC operating systems second that Netscape and Java threatened to reduce the threatened to reduce that monopoly power and third that Microsoft acted any competitively to protect that monopoly power today I'm going to leave aside number three the question of the of the challenged acts of Microsoft and just to talk about the structure of the market what was going to happen what the government said was going to happen had Microsoft not acted and again compare it to what we have seen happen okay so let's break this down a little bit more if you look under Microsoft having monopoly power they're basically two aspects to that if again we're leaving aside evidence from what Microsoft did the first one is high market share and of course high market share with the application barrier to entry making it hard for for new operating system products to come into the market the application's barrier to entry again to review pretty quickly here we have windows and an application that runs on top of windows by runs on top of windows what we really mean at the technical level is that this application A has to talk to windows to ask it to do things so the application can run and windows like every operating system has its own quirky and difficult language that you have to speak in order to get it to do things and the application developer would have to devote a fair amount of work to making their application speak the language that windows understood and this was the case for all the other applications as well every application had to invest a significant amount of development time in getting their application to work with windows just as they would with any other operating system okay so along comes a new operating system which is going to enter the market and the developer of application A if they want to get their application to run on windows I'm sorry on the new operating system if they want to port it across is going to have to get their application to speak the language of this new operating system so let's make an analogy say that windows speaks English and the new operating system speaks Spanish so we'll dub it Nuevo OS and if you want your application to run on Nuevo OS you have to teach it to speak Spanish or at least you need to give it enough phrases that it can ask the operating system to do what it needs to do and that's a bunch of work for the application developer it's costly if you want other applications to do the same you have to teach each of them to speak Spanish and you can't just get them all in a classroom and teach them you have to do the work individually for each application you need lots of applications to make the Nuevo OS actually viable in the market it's expensive and that poses a significant barrier to entry that's essentially the government's theory about a barrier to entry now when a browser comes along Netscape and Java what they do is change this story a little bit so what they do in particular is they kind of jack up these applications and they slide underneath it a new thing which the government dubbed in their theory of the case middleware so you can imagine that this brown thing is a browser and the browser speaks English to talk to windows and it exposes a different kind of interface up at the top the applications can speak let's say Esperanto to the browser the browser will translate that into English and talk to windows alright now if Nuevo OS comes along if you want to port all of this stuff to run on top of Nuevo OS you need to take the browser and teach it to speak Spanish but once you've done that you can just drop the applications right in place because they speak Esperanto to the browser just like before and so in order to make everything work on Nuevo OS here you need to teach just one program the browser to speak Spanish that's a lot cheaper it's a lot more reasonable and easier and you can imagine that a new operating system entrant might be able to afford it therefore the arrival of this middleware the government argued would reduce the barrier to entry and that's what Microsoft was afraid of now this is the story the government told at the trial and it was a story that was adopted by the courts essentially so I want to look at this story and actually evaluate it against what we know well there's a key technical contention that underlies all of this and that is the contention that serious applications could be run in browsers which which we've learned something about in the years since is it the case that serious applications could be written to run on top of browsers this way or to put it another way is it the case that an application can can live a healthy prosperous life while talking to the world through an Esperanto interpreter well let's look at what happened and I'm going to divide the years since 1998 into two periods the first period 1998 through 2003 and the second one the first five years and the second five years in the first five years here's my capsule version of the history of those years in the browser market first Netscape withered away and internet explorer came to seriously dominate the browser market second I think the pace of browser innovation slowed during this period browsers continued to advance but we didn't see the rate of progress that we had seen during the height of the browser war in the late 90s and there was talk at the time about whether the browser was now a mature product that wasn't going to advance anymore or not advance at the same rate or what but I think it's pretty clear that the pace of innovation did slow down and significantly during this period serious applications really did not run in browsers you had a few toy applications maybe a few things that were suggestive that they might someday but we did not see the we did not see Esperanto speaking applications in any significant numbers during this five year period right now the second five year period 2003 to 2008 we saw firefox rise from the ashes of Netscape firefox started to become somewhat more popular reached double digits in market share and was seen as a as a real competitor to internet explore having feature parity or in some people's views maybe a slight advantage at least at some points the pace of innovation did seem to quicken we saw new features coming in we saw we saw investment of bi open source developers and companies in firefox we saw microsoft pushing more resources into browser development and we've seen even in the last year I think an increase in the pace of investment and innovation in the browser market again we're now starting to see some serious applications running browsers many people I know including me at times spend a lot of more time inside the web browser and we're seeing a lot more of the things that people do being delivered as web based applications more or less as predicted and of course as Dave mentioned we're also seeing the rise of a new class of applications which really make sense only in browsers running Facebook on your own computer not connected to the internet not that interesting but we're seeing the rise of these social applications which which really make more sense to run in a kind of browser setting so not only applications that are taught to teach Esperanto but applications that sort of were born speaking Esperanto okay so there's an obvious contrast here in the first period the first five years no serious applications in the browser in the second five years we do see serious applications in browsers so what changed something apparently changed around 2003 which caused serious applications in browsers to become a reality and I want to put forth two possible explanations for what it is that changed around 2003 two theories of what it was the first one is the competition theory the competition theory says that the key enabler for the arrival of these web based applications was browser competition as browsers competed as they got faster as they took on more features it became more viable to make these applications and developers seeing a more competitive browser market may have had more trust that the browser standards would be more stable thus protecting their investment in developing these apps so if you accept this theory then in the first period 1998 to 2003 there was less competition and innovation but later on we saw the rise of firefox driving competition and this is according to this theory is what led to the rise of web based applications the other theory is the technology theory that says the thing that happened around 2003 was a set of changes in technology that in fact web based applications were never really realistic in 1998 and it was only things that happened around 2003 that made them realistic so what were those things that might have happened I think there are three plausible candidates the first one is faster computers Moore's law made computers faster running an application inside a browser does very a performance penalty and arguably you needed a faster computer in order to make all of this work so maybe faster computers made it possible another possibility is the pen is broadband penetration arguably you may have needed a faster network to make these applications work these applications are always chattering to some back end server that holds your data and unless that can happen quickly arguably you can't make these applications work well and so it may have been that you needed to have a certain number of homes and offices with broadband internet in order to make web based applications a viable business model the third candidate here has to do with programming tools and know how the argument here is that in 1998 we just didn't know how to write these applications we didn't have the tools we didn't have the know how and it was only the birth of it was only the maturation the methodologies like javascript and it was only the growth of programming methodologies like the so called Ajax methods that really made it viable to write these applications and in other words that the programming community needed to spend five years noodling around before they figured out how to actually make a good application alright so why does it matter which these theories is right well if the competition theory is right then the reduced competition that we saw in this first five year period actually has held back progress and we are not as far advanced in the technology as we would have been had there been more vigorous competition during that period on the other hand if the technology theory is right then maybe the impact of competition in the period right after the case was brought mattered less maybe it really was a case of needing to wait for the technology to wake up or needing to wait for technologists to noodle around that's not to say that the that the lawsuit mattered not at all in the structure of the market or the behavior of companies but if this technology theory is right then it at least had less impact on the web that we see today this is why you have to prepare witnesses right right okay so which theory is right I don't know this as I prepared for this talk and I talked with this about this stuff with my colleagues and students we had quite a few interesting lunchtime debates about which theory was right so I'm not going to tell you which one I think is right instead I want to just keep telling you history and give you an optimistic history for the next five years and I'm feeling like an optimist this morning so this will be the optimistic version but plausible the first part of the optimistic history the competition continues to heat up in the browser world between internet explorer firefox and now google's chrome browser which I think is a pretty interesting competitor in this space increase browser competition second that microsoft as a result is pours even more resources into browser and web technologies that and so we see significant advancements in internet explorer as well as in the other browsers third that because of this investment on all sides we see some fundamental advances in browser engineering and we see some changes in the way that browsers work which lead to a different kind of web than we see today but of course but of course the question that follows is which of the competitors comes up with with that with those advances and how easily can they be adopted by the other competitors because this is the future I'm not going to even try to predict and assuming this optimistic future where we see more competition and increased investment and fundamental advances consumers win so the short version of this future history is that Microsoft gets a do-over on the browser war and really does it right this time thanks okay, thank you Ed in setting up my at least initial question to the panel I want to return to this question of defining a software product and in particular without taking a position on whether tying analysis should should move from a nominal per se approach to the rule of reason approach to suggest by the DC circuit and the Justice Department just to suggest that Microsoft shouldn't be the thin end of the wedge in making that change and in fact suggesting that a more explicit recognition of the legal significance of Ed Felton's technical contributions to the case would have really put the the analysis on point with Jefferson Parish and addresses the idea that there are pro competitive benefits to expanding the platform to the extent that there are benefits that those do not come out of benefits of the tying conduct so Microsoft's position again in proposed findings of fact were that the software product consists of code and nothing else but with respect to our Microsoft colleagues I strongly suspect that Microsoft wouldn't take the same litigation position in the copyright infringement suit i.e. that someone who purchased a Microsoft software product had purchased the code instead I suggest in the articles that I have at registration and if we're out of them it's at my site andrewchin.com a software product can be defined explicitly as inspired by Ed's implicit definition at trial as a set of legal rights as specified in the license accompanying the copy of the software and technological capabilities that enable the use of the accompanying software as specified in accompanying documentation related to the accompanying software but it doesn't include the software itself and just to nail down this intuition when you buy a region 2 DVD you're not buying the movie it's a digital right to watch the movie on a DVD player in the appropriate region and that's backed up by digital rights management so you're limited technologically to use of those players it does again and again does not include the movie itself and so using the implicit formulation that Ed used at trial to define the product as that which allows the user to watch the movie on a player in Europe so Ed's demonstration his prototype removal program had the effect of demonstrating that Windows is an OS product plus a browser product which can be described implicitly as that which allows the user to install and run Windows apps plus that which allows the user to browse the web and then by inference the browser product consists of the browsing rights that Microsoft might draft into a separate add-on license plus a technological capabilities that would be delivered through an inverse prototype removal program that replaced the personalized Windows 98 with the one marked by Microsoft thereby creating that which allows the user to browse the web so the tie in essence is then a contractual tie not a classic technological tie it's a mass market license reasonably understood to bundle the OS product with the browser product backed up by DRM defaults and so the claim of new APIs delivered by the platform are really not an additional benefit that's delivered by the tie and in fact Microsoft 3 is an entire land point with Jefferson Parish if we have this intuition in mind Jefferson Parish you're not buying the operating theater you're buying surgical and anesthesiological services which arguably are separately demanded and it's the same operating theater that facilitates the provision of those services in the same way that the same body of code the same DLLs facilitate provision of the OS and the browser functionalities but the patient is buying the services not the operating theater and the user is buying the functionalities not the code with that first I guess I'd like to give Dave an opportunity to respond to that but in the interest of time I guess we should open up to the general questions I guess I would just say that you know Microsoft's view if that's the way regulatory agencies want to look at the issue which essentially is how it was looked at I think in the Department of Justice case against Microsoft that's fine and that's the remedy that's in place and that's what we're living with it's very much the focus on the end user is the user able to browse the web yes or no Microsoft ought to let users remove that and Microsoft does that today and we're extending that approach to multiple other features you know looking back on it the whole debate about what is the browser, what is the code we spent a lot of time on that at the trial and a fair bit of time on that with the European Commission and there's sort of no answer to the question I mean Professor Chin was proposing an approach just now and I think it's fine other people might have other thoughts about that you know the fact of the matter is that people come to work they build this operating system called Windows it consists of code and so it seems like that's part of the product the code interacts with other code in ways or various dependencies the codes functions are exposed to developers they can call into it and you know you can take a sort of carving knife and say well this is the media player or that's the media player and there's no real answer to it as Phil said to me at dinner last night it's almost like an existential question you know what's the browser what's the media player and the only real question is what's the best design from the perspective of enabling removability or what should get built in in the first place from a competitive perspective and we have the answer in the US which is focus on the end user and risk of confusion Europe seems to be different yes Dave Heimer has told excuse me told us that he doesn't think you can integrate web 2.0 applications with the operating system and Ed Felton today at least is very optimistic about consumers winning from these new kinds of applications but what if your web surfing on your iPhone and could you both comment on tethering in this context because you're not really surfing the web you're surfing the portal that your device operating system owner lets you well the iPhone is actually an interesting example here as you point out if you want to use Facebook on the iPhone for example there are two different ways you can do it you can open up your web browser and go to facebook.com and then you get essentially the experience you get from any browser only in a smaller window or you can open up the Facebook app that ships for the iPhone and that indeed has to be approved by Apple and and will only do what Apple allows it to do and I think it's an interesting question whether there's an inherent advantage to the built-in app I in fact find I do use it but if Apple insists on too many limitations on its use people can always shift over to the web case it's rather like this this is I think related to the question I raised earlier about how much an application has to give up in order to run through a browser as opposed to running directly on the client system and I do think there is something especially on a small device like that which technologically is back maybe where PCs were in 2003 thank you I'm at the hardware business supply now I think by this because the fact that the most part of innovation is the fundamental new discovery of new concepts of the the fact that when we innovate a product with a fundamental concept and we are expanding our horizons by combining all the ideas and ways and this innovation happens to be a tiny issue, right? I have been to a lot of conferences at hardware business school where the people that participate are not so much companies that are investigated but more companies that no longer exist and so we have the old WANGs and visuals that come along in new organizations that are not innovated not that they will be innovated in time so Marco, can you push the red button more so it's a balance on the one hand of course you need to do the right things at the same time you do want to respond to what's going on in the environment around you and that's what we teach our students and maybe we should stop teaching that but hopefully not a couple of things that are important about this base I also wanted to address a little bit of also what's happened in the industry in the last few years because we've looked at that a lot I've studied and researched it and a lot of things have happened it's interesting that on the one hand we talk about Microsoft sort of not being able to do certain things and then we complain about Microsoft not being innovative we also have to understand it's an organization that if you imagine yourself running the browser project in Microsoft in 2002-2003 you're probably on a certain amount of pressure is to things in a certain way it's not such an easy thing to actually get right but the reality is that the investment in the technologies was quite large and expanded a lot over the years and from an internal perspective the innovation was probably fairly significant the major development I think that the professor felt and talked about is a really important one which is that with web tool technologies we have a whole new generation of plugins to the browser they're really where the action is right now in terms of writing these applications so we have things like Java now but also Flash player and things like that that have a penetration in the market which is in the high 90s and they're basically installed in every PC so we can run applications like that across the board and it's been around for a while even in 2001-2002 applications were certainly browser based so that's been there for a few years I think it's an interesting environment because if we actually look at the concept of the application barrier entry I don't think what we see today is what we might have seen 10 years ago it's a very different environment most applications are operating system agnostic users spend 55% of their time I think on average on the internet instead of on the computer itself and most of the stuff that we have even on the client is now available on multiple platforms so if you actually look at what people have at the run I think there's been huge changes in the last 10 years and it's one of the things that's useful to actually take a look at so it's good to look at the history I don't think it's so much for me I don't know you have thoughts about it I think that the one thing I would say about it is to go back to Ed's comments about the question of what's changed what changed in this 2003 time frame technology with competition now of course if Ed were the witness as I was joking about before you know probably would have urged well Ed is it really an either or question no it's absolutely not an either or question and that would help refine that testimony and make it more I'm sorry we should have spent more time on this I guess not that we worked very much before you testified actually it was all totally rehearsed and you know just you know unprepared but I mean I think the harder question really is in the broader sense you know the competitive landscape and particularly the competitive landscape in the context of this case and what it tried to achieve how much of a difference did it make and how much of the landscape did it open up and I confess I have not actually spent a lot of time thinking about that very much and maybe I should have but I think in part it's because it's a very very difficult question to answer you have all these different factors that you have to control for and you can't control for them with respect to how software developers are thinking about Microsoft and how they're thinking about the industry and what they can do but you know I think that it would be wrong given the many the many lessons we've had about the importance of competition in many industries to simply write off the idea that competition competition really matters Andrew maybe one last question and then we should probably break just so we don't get too far off time I had a question for Steve and the question is what do you think of DOJ's decision not to pursue the Section 1 tying claim on remand I think that you know it was a case that based on the evidence that we had all put in this expert testimony and fact evidence and documents that it's a case that was very very very winnable under the standards that the DC circuit articulated in fact I think pretty much all the evidence was already there to win it which is one reason why I didn't mind the I was going for the government to that point but didn't mind the DC circuit opinion didn't think it was clearly wrong but all the elements were in place so I think we're regretted that the DC circuit couldn't see maybe it was in the way the case was presented at the court of appeals I don't know about the guy who was arguing the case but we had gone to great lengths to try to satisfy all those burdens as I said before under the rule of reason as well as under our per se analysis I mean I want to revisit by the way and just a very very quick last comment I don't really care very much whether it was rule of reason or per se it's actually not entirely right I mean we put the section one case in part because we said hey we got this per se rule let's use it but improving the case is what I meant it wasn't something that held us back by any means David would you like I don't usually comment take the role of the justice department but for this group let me just it would be fun for me to kind of role play what the justice department said at the time which was that they could get all the relief they wanted under the section two theory where liability was affirmed without having to retry the section one and the point there was that the theory under section two and section one was what professor chin was just talking about and professor felton which was this focus on the end user and they wanted relief which would say Microsoft auto enable computer manufacturers to ship machines and appearing to the end user and users ought to be able to do that and that was in fact put in place the other thing where I just would maybe disagree with Steve a little bit in terms of were all the elements lined up to win on a time claim on remand even if there were a reason to try it the appellate court had said that the district court had failed and a proper factual findings to establish that there even was a separate market for browsers and that the agency could not put that in place on that point under a rule of reason test it seemed like it was going to be pretty hard to establish harm in a market that hadn't been established except you know and I think come back Phil I know you want to wrap up but even under the DOJ's new policy statement there is a recognition that tying can have significant anti-competitive effects back into the tying product market and although DC circuit I'm not sure was ready to acknowledge that I think that is one thing that was a little flawed from my perspective in the DC circuit opinion but I think that had that case been retried obviously trying to pass through this browser space even if you didn't define it as a market and get back to a very clear anti-competitive harm in the tying product market is something that was a very solid theory and in fact one that animated this case as distinguished for example potentially from the kinds of cases that the EU has brought I thought the interesting question would have been that's absolutely the fascinating question Andrew our panel thank you very much please join me in thanking them let's