SUMMARY OF PROF. FISHERS TESTIMONY By: Greg Harris

Prof. Fisher is an economist from M.I.T. who serves as the DOJs final witness. His testimony is presented as a unifying narrative for the government, drawing upon the information provided by earlier DOJ witnesses. Fishers direct testimony is deceptively succinct, weighing in at a mere twelve pages. His cross-examination, re-direct and re-cross, however, span the greater part of six days in court. In its totality, Fishers testimony forms the core of the Department of Justices case.

Even when viewed from Microsofts standpoint, Prof. Fishers testimony is clearly helpful to the government case. He accomplishes the goal of creating a framework of analysis that explains all of Microsofts actions as separate elements of a single scheme--the continued domination of the OS market. All of Microsofts alleged actions, including anti-competitive contracting, predatory pricing, and attempts at market division are to be viewed as means of supporting this central plan. None of this is new, but Fisher skillfully weaves of these separate threads into an coherent tapestry of monopoly maintenance. This is the DOJs grand unification theory.

Here are the elements of his theory: 1) Microsoft Windows holds dominance over the Intel-compatible OS market and is maintained by network effects; 2) MS has foreseen a challenge to this dominance from platform-independant browsers; 3) Microsoft has taken anti-competitive actions to prevent browsers from disrupting its OS dominance; 4) These actions will harm consumer welfare by stifling innovation (or, at least, non-Microsoft innovation) and allowing Microsoft to retain power over pricing. I will leave it to the DOJ summary to expand upon these points, but the central issue is that although there are natural network effects in the OS market, these create barriers to entry that invite abuse by MS.

Prof. Fisher is essential to the emergence of a clear-cut government case against Microsoft. No other witness has endeavored to put the totality of Microsofts actions under one framework. It is, therefore, not surprising that Microsofts attorneys are eager to take their time in exposing every weakness in his analysis. Some of Microsofts points are nit-picking and do little to chip away at Fishers credibility, but others hit closer to the mark. Much of Microsofts effort is aimed at severing Fishers economic expertise, which is unquestioned, from his understanding of the particulars of this case, and the markets involved.

Here are Microsofts main lines of attack:

1) AdKnowledge Data  AdKnowledge was hired by DOJ, and Fisher used biased DOJ data. 5PM9  AdKnowledges Data was not corroborated by other firms, and uses a type of counting (Hit Data) that is both inaccurate and unique to AdKnowledge. 5PM10,14  Fisher claims that the AdKnowledge date produces the same conclusion (anticompetitive actions and effects) as the data used by MS. 5PM12  Fisher claims that he made only minor corrections to testimony (commas and such), MS says its much more! (Fisher changed his opinion on what the AdKnowledge data actually measures) 5PM15  Microsoft implies that Fisher framed the time limits of his study to exclude the early market, which is distorts IEs history. 5PM17  Hammers Fisher for his lack of knowledge re: the NCompass Browser, which uses IE tech, but allows alternate branding and never looks like MS, nor is it integrated with Windows. AdKnowledge data counts NCompass as IE. 5PM19-20  AdKnowledge is very rough, very inaccurate data. Overcounts on rotating advertisement sites, undercounts on ad-free sites. 5PM22  AdKnowledge discounted the number of Netscape browsers because it excludes corporate intranet users, which were a large focus of Netscapes business plan. Fisher has no real reply to this. 5PM30-1  Fishers data discounts AOL users who stay in the AOL-only shallow end of the net. What browser are they using? 5PM32-3  Fisher is forced to admit that AOL users, with their browsers, are strongly undercounted by the AdKnowledge data in the early years. 5PM36  Also problems with undercounting AOL due to caching, which shields AOL data from AdKnowledge. 5PM39  Rough measurements of browser use by an ISPs IE parity. Again, this goes to the bad data thread. 5PM50 2) Fishers Knowledge Beyond Economics  Fisher has never before testified or published in regarding operating systems, web browsing, java, multimedia or other related areas. 5PM80-3  Microsoft strongly insinuates that Fisher spent far too little time going over the record (49 hours billed) to have read all the evidence that he claims to have read, even if he argues that I read extremely fast. 5PM87-89  Fisher did not consult with or speak to any software market insiders, ISPs, ICPs, application developers or OS vendors. 5PM89-90  Fisher has a 30 year relationship with Mr. Boies, and a long relationship with Cravath (MS says hes known as their in-house economist). 5PM91-2  In a music licensing case in which Fisher testified, his scant knowledge of and research into the industry lead the court to publicly disregard his economic analysis. This was very embarrassing for Fisher, even though it occurred in his youth. 5PM94-5 Fisher was forced to criticize the judge. 5PM96  In Kodak v. Polaroid, the judge threw Fishers expert opinions out (along with all other expert opinions) as biased and unreliable. Here Fisher claims that the court in this case was misled by a, shall I say, intellectually quite dishonest witness who managed to produce a lot of confusion. 5PM97 3) Competition and Barriers to Entry:  Fisher could not identify a single entrant into the browser market who was deterred by IE. 5PM79 (Note: This is a weasel question--its just too hard to identify who didnt show up!)  But MS does damage Fishers analysis by pointing out his near-absolute ignorance of non-PC web-browsing technology, which is a crucial aspect of the Java plan to unseat Microsoft and Windows. 6AM10-3  MS raised the specter of the Netscape-AOL-Sun deal. Fisher had no strong reaction, and could not adequately account for the effects of the deal upon his analysis. 6AM14-6  Also, MS questions whether Fisher has adequately accounted for the competitive effects of network computers (NCs). 6AM19-21  Fisher claims that AOL will not use Netscapes browser, and will treat the choice between Navigator and IE as if it were between two outside vendors. Thus, he argues, MS will still be able to pressure AOL. This seems unlikely, and he is likely minimizing the synergistic effects of the merger. 6AM22-3  MS introduced an old writing of Fishers which defined barrier to entry as a two-pronged test: first, the potential competitor must be deterred from entry; second, the deterred competition would have to have been in the public interest. 6AM51-2 Fisher disavowed this view.  Linux was raised as a potential competitor with a growing number of software developers choosing to support it (Corel). 6AM59-61 4) Integration and Tying:  MS tried to pick at Fishers lack of qualification to define what should be in an OS, and what should be in a browser. Fisher aptly deferred to the expert opinions of Felton, etc. 6AM6  Fisher created a paradigm for tying that proved difficult for MS to assail. He distinguished the question of what can be done to sever IE from Win98 from the question of how MS could have designed Win98 differently--keeping all of the functionality, but with more severability from IE. 6AM9  MSs main strategy at this point was to question Fishers ability to determine whether MS really had such a choice in design. Who is he, after all, to claim that it could have been done differently and still offer all of the functionality? Who is he to conclude that MS factored anticompetitive motives into its decision? From an economist with no computer or mind-reading expertise, this is pure speculation.  Again, this speculation over whether Microsoft could create a seamless integration in experience, while keeping the products separable, appears at 7AM47-8, when Fisher claims that it doesnt seem to me that that would be a real problem for Microsoft to get around. Fisher then admits in a hypothetical that if there really are advantages to integrating IE and Windows the way it has, then that might explain its choice, and could bring ancillary pricing benefits. 7AM49  Fisher also admits that the tying of Windows and IE was desired by many consumers, though he again distinguishes between integrated experience and code integration. This hearkens back to the better parts of MSs cross of Soyring, where OS/2s integrated web browser was shown to be a huge potential selling point. 7AM69-71  One final note is that Fisher claims that combining two products so as to produce colorable efficiencies and to make it difficult to separate the products cannot constitute a defense to a tying claim, or else tying will never be found in the software market. DT158. This begs the question, so what? Maybe tying isnt appropriate in the case of software integration which produces colorable efficiencies! 5) Monopoly Pricing  MS seeks to blunt the impact of Fishers statement that, in pricing Win98, MS didnt consider it necessary to examine the pricing of any other OS maker. MS notes that the pricing of Win95 took this into account, and that the pricing of Win98 was largely determined by the pricing of Win95. Thus, by a sort of transitive effect, the pricing of Win98 takes into account the pricing of other OSs. Hmm. 6AM78-9  MS seeks to exploit Fishers ignorance of the software market by once again making the everyone else does it claim. Specifically, it uses him to introduce again Barksdales recollection of Larry Ellisons use of the cut off their air supply term. 7AM62-3 This serves to blunt MSs use of the term.  Free distribution of software is also a common practice, and Microsoft seeks to blend its treatment of IE in with Adobes distribution of Acrobat and other similar instances of seemingly irrational pricing behaviors. 7AM42

On a personal level:

Having twice(!) been reluctantly assigned to argue Microsofts position, I feel justified in stepping back for a moment and adding my personal reaction to Prof. Fishers testimony. In my opinion, it is the best the DOJ has to offer -- it is a coherent and plausible story in and of itself, but when combined with the testimony of others (Tevanian, Soyring, Warren-Boulton) and the internal correspondence of Microsofts officers, the testimony paints a damning picture of what really happened and why. What surprised me upon reading the testimony is how closely it dovetailed with Prof. Lessigs overview of the past on our first day of class. The pattern we established then seems clearer after Prof. Fis

-- Anonymous, January 21, 1999