Resolve and Fortitude : Microsoft's ''SECRET POWER BROKER'' breaks his silence

by Joachim Kempin

  Last but not least, in one of my e-mails to Bill, I had mentioned an application barrier of entry for our competitors when competing with Windows. I reasoned that the Windows standard we had created luckily protected us—until now—while pointing out our luck could run out soon. For Malone, the barrier he had in mind was absolute, as his economists had told him, enabling us to defend our alleged monopoly forever. The trial was casting a shadow.

  I should mention one last item that had captured the Feds’ attention. In a separate mail to Bill back in ’94, I suggested sending IBM a hit team in case they failed to comply with one of our requests. An ill-chosen word for sure. But in any high-powered sales organization, hit team is a colloquialism implying no physical harm, meaning “Let’s be sure to send the highest-caliber, toughest negotiators possible.” For Mr. Malone, it expressed that we had power to harass and coerce customers. As long as I could remember, IBM never seemed intimidated by us. But for a government attorney like Mr. Malone, such language was not permissible and proved malignant intent or harm punishable by law.

  Naturally, his other interest regarded the by-now-famous Windows boot-sequence restrictions we had imposed on our OEM customers. Again probing for anticompetitive behavior, his main line of questioning was about the reasoning for our regulations. Why did we restrict OEMs from removing our icons from the Windows desktop? He for sure did not like my answers, but our copyright existed, and our restrictions defended it. After two days of intense cross-examination, he left—not terribly satisfied. My attorneys praised me for a reasonable performance. I had survived another day.

  The DOJ strategy for winning the trial had become crystal clear as we studied the substance and shape of arguments our witnesses had been entangled in during their depositions. In all antitrust trials, one of the key elements consists of defining and sizing the relevant market for the judge. The Feds wanted the market definition as narrow as possible, with us insisting on the opposite. A judge could not easily condemn a company of having monopoly power if her competitors had 85 percent market share. Judge Posner described the struggle to define the relevant market in his book Antitrust Law as an always present dilemma: “The importance that antitrust law attaches to defining a market is another consequence of the law’s failure to have developed an approach at once genuinely economic and operational to the problem of monopoly.” No science here as I understood his remark—the subjective beliefs of a trial judge sufficed condemning you. This aspect of paramount importance would resonate throughout the trial delineating utter consequence.

  The other obvious part of the Feds’ strategy was to discredit as many of our witnesses as possible. That tactic had priority above all on Boies’ courtroom menu. He is quoted in Fortune magazine as saying, “What you do is set it up so you get the jury [sic: in this case a judge] to think the witness is lying—before you even suggest the witness is lying.” Most trial lawyers attempt this, though Boies arrived with the reputation of having mastered it. If he accomplished that feat with our witnesses, as he had already done in Bill’s deposition, our cards would look really bad. Boies profusely understood that Jackson, when issuing his final verdict, would be disinclined to base his reasoning on witnesses he failed to believe in. If Boies, in his henchman role, could trick our witnesses into tying their own hangman’s noose, no need to escort them to the gallows himself. The judge would just sit back and watch them swing in the breeze.

  The third strategy was to convince the judge of the validity of a highly abstract and unconventional economic model. In a classic antitrust case—and in all cases to date so far—an offending monopolist, after arriving at its unique position, restricted output and raised prices above reasonable levels, maximizing profits to the detriment of consumers. Promising to never charge for IE and keeping our Windows prices pretty much flat since ’95 and unadjusted for inflation without ever restricting output could hardly qualify as monopolistic behavior.

  To overcome that hurdle, expert witness Franklin Fisher, on behalf of the DOJ’s economist Daniel Rubinfeld, planned to present an esoteric abstraction of an economic model based on path-dependent network effects. The underlying theory is derived from what mathematicians label chaos theory. Despite its name, it describes a serious statistical model used to study highly dynamic systems dependent on, or being predominantly sensitive to, initial conditions. Good examples for these are the big bang, quantum movements, and weather patterns. A few avant-garde economists, like Stanford’s econ scion Mr. G. Saloner, found this self-same statistical model helpful to describe and predict anomalistic market behaviors. Their main deduction: as in physics, history can determine the ultimate outcome of a network-driven market economy and lead to irreversible results. He therefore postulated that our success with Windows could sufficiently and scientifically be explained within the means of this model. Exact science or ambitious thinking, who knew?

  Accepting this as a highly plausible theory, sufficiently describing how we had achieved our dominance, the DOJ’s economist concluded that we could therefore never be overcome by competitive forces. The destiny of being top gun forever was for them path-dependently predetermined (by God?) and manifested by networking effects working in our favor. As dubious as this sounded to me, the Feds’ economic team was dead set on putting this freshly baked model to a legal acceptance test.

  A gamble in the face of missing traditional—Chicago School of Antitrust—evidence for classic monopolistic behavior better described as clinging to a desperate last straw. Nevertheless, the Feds planned to suggest that the large-scale Windows demand we had created rewarded us with unstoppable returns to the detriment of competitors and consumers alike. Buying into this colorful sci-fi depiction of our undiminishable dictatorship would give the judge hopefully enough reasons to topple a nasty despot and save the industry and mankind from disaster.

  We, on the other hand, characterized customers buying and staying within the Windows environment a genuine selection process, by volition. Potentially, a natural network had been created by an avalanche of entities joining up, comparable to a phone network or the Internet. Economists understood how a popular and growing network could naturally shape the structure of a demand curve and lead to high returns. For the Feds, the extraordinary steepness of that curve in regard to Windows distorted the proper workings of the industry, unfairly tipping the playing field to the advantage of, and wrongly rewarding, a monopolist.

  Historical events contradicted this. In the early ’80s, Apple had reigned supreme with her Apple II platform, and the myriad of application programs written for it served as a barrier of entry for competitors. Attacked by the emerging DOS-powered IBM PCs, it eroded and eventually faltered. The barrier of entry was overcome by free market forces: risk-taking entrepreneurs (PC manufactures and ISVs alike) and customers opting for emerging computing platforms!

  Three years later, Apple attacked the by-then dominant IBM PC platform with her superior Mac PCs—unsuccessfully this time. The Mac went on and blossomed as a niche product until Windows 95 and its successors stole more of its limelight. As with VCR standards before, an inferior product (VHS) won over a better contender (Beta) because consumers—speaking with their pocketbooks—favored the less expensive choice. . Much like what happened to Apple’s Mac sales when consumers preferred the more affordable DOS-based IBM clones and later the ones based on Windows. Early on Steve Jobs and the ones who followed him at Apple did not get it right!

  In all these cases, the market had spoken, applying its own common sense realism—not needing government interventions. Leaving only one conclusion: as slick as the Feds’ economic model looked academically, it did not pass the reality test of the PC revolution.

  Our strategy to win the trial was a much simpler one. Based on our experience with the judge, most members of our legal team did not trust his ability to comprehend the intrinsic matters before him. Others believed he was tainted by our eye-poking behavior in round one. Both had a point.

  Our legal team
believed that all our actions were defendable and that the market we operated in was much larger than the Feds alleged. Admitting to owning a monopoly was therefore not in the cards. So let Jackson do his work, and if needed, appeal his ruling. It worked once; it should work again. Banking on winning on appeal was a delicate strategy to execute and could easily backfire. We needed to assume the judge would do everything within his discretionary power to avoid another stern rebuke from the higher court. He was an old hand and knew the procedural rules well enough to navigate them to his advantage.

  Unbeknownst to us, he had already decided to issue his finding of facts separately from his final verdict, hoping that even if his verdict got overturned, the finding of facts would survive. What he, like any other judge, would issue would not be facts in a true sense. They represented in a condensed form what Thomas Penfield Jackson, as a subjective human being, had discovered. As long as the trial record even vaguely supported his findings and bias could not be explicitly proven, it would be hard to disregard them. In talking to our attorneys, I got the impression that they knew full well the risks of betting on such a simple yet narrow path for victory and believing in the appeals process as the savior for their client. It struck me as odd, but I was in no position to recommend an alternate path.

  WITNESS PARADE

  On the pretense of shortening the trial, all witnesses had been asked to prepare written statements so testimonies could be obtained in advance. Several added videotapes explaining technical details with graphical images and compilations of event sequences. I never composed a single word of my written testimony. Our attorneys did. We reviewed the draft, and with the exception of modest changes, I approved what they had produced. My submitted videotape was a different story. The prosecutor had found mistakes in two of them, presented by MS witnesses preceding me. To avoid another disaster, we therefore remade my tape. It was comprised mostly of screenshots of Windows desktops as delivered by OEMs. The revised version showed a growing number of OEMs placing competing browsers and other competitive icons on Windows desktops. This would not sit well with the prosecution wanting to prove the opposite.

  At the opening of the trial, the government paraded her witnesses first: not a single OEM customer under them, making the Feds claim they were afraid of us. Seven of the prosecution witnesses were from competitors, two were economists, and the three others were software experts. Our legal team had a chance to cross-examine them first and to soften up their written testimony. The government used its time trying to prove MS had systematically harassed these firms so she could erect and defend a competitor eliminating barrier to entry of her alleged monopoly.

  The DOJ’s two economic experts attempted to convince the judge to accept the narrowest of market definitions. Subsequently, they tried hard to get their theory of a network-effects-driven economic world order in software markets validated by him. An attempt was made to make our zero-dollar pricing for IE look predatory. Windows pricing, on the other hand, was labeled excessive and typical of a monopolist without providing sufficient empirical data for the allegation. To prove this, the DOJ’s experts compared our unvarying Windows royalties with ever-falling PC prices. There is no positive correlation between the two. A correlation between the price of a car and the price of steel or aluminum certainly existed. Yet the price of a car and the price of gasoline, which you for sure needed in ’99 to actually drive one—did not.

  When it came to our alleged monopoly power, the DOJ economic expert Franklin Fisher concluded in his direct testimony, “It is difficult to imagine that in an open society such as this one with multiple information sources, a single company could seize sufficient control of information transmission so as to constitute a threat to the underpinning of a free society. But such scenario is a realistic (and perhaps probable) outcome.” The company who had wanted every human to have a PC on her or his desk—a threat to mankind?

  What both of the Feds’ gentrified economists did not provide was sufficient empirical evidence for their broad-stroke wisdom. Bill Neukom confirmed my observation in a press conference: “This is the fundamental flaw in the government’s case: they apparently never took the time or brought in objective analysis to the question of whether there’s any consumer harm to our actions.” When Fisher was cross-examined in court and asked directly if we had harmed consumers, he had to concede: “On balance, I’d think that the answer is no, up to this point.” So why were we a danger to society? When Boies reexamined him later, he tried to change his story by saying that the consumer harm had been done by depriving them of choices through restrictive OEM contracts. I should have known!

  The Feds’ next argument centered on raising Windows prices. For six years we had kept them unchanged. But in ’98, I reduced some rewards OEMs could achieve by a small amount for the simple reason that they had made no attempt to gain them. Reducing incentives was therefore presented by Boies as a price increase. Yes, he could argue along these lines. Without boring the reader with details, rest assured that the increase did not even compensate us for increased development cost or inflation. Neither was the higher-value proposition for the improved 98 version ever considered or analyzed. Therefore, my math-driven argument that we had kept Windows prices basically flat during the time in question still stands. Mr. Boies nevertheless scored brownie points with journalists, exhorting, “It’s clear that Microsoft has been raising its prices at a time when everyone else in the computer industry has been lowering them.” His noise made great headlines without being close to economic truth or hard math. Again there was and never will be any proven correlation between software and hardware prices in any industry except a negative one.

  An executive from Apple testifying for the Feds was no economist but insisted MS had a monopoly based on his understanding of “both the industry and the technology therein.” He referenced our high market share and the symbioses between Windows and its applications. The guy was speculating wildly. Neither high market share nor the natural symbioses between any OS and its applications proved his point. All the same, his amateurish-appearing evidence was allowed to stand.

  Richard Schmalensee, our own economist, testified quite differently. He, like Fisher, got into a pickle early on when attorneys were able to show that his written testimony contradicted with what he had written in earlier print publications. He went on, however, presenting a detailed and thorough analysis, using formulas well accepted by economists to prove at no time had we ever charged a monopolistic price. Without waiting for the prosecutor, the judge freely opined that a monopolist looking for long-term gains did not necessarily have to charge a high price. Was he arguing for the prosecution? While Schmalensee did not make the greatest impression on the witness stand, the government econ experts were equally lusterless. One mistake committed by our defense team was having Richard Urowsky examine Schmalensee. Richard was the self-same advocate who had won our first case arguing our appeal, thus helping Jackson to more than one black eye. Who knows how lucid his memory was in this respect?

  In summary on the economic front, the Feds were hard at work trying to prove how we had harmed consumers. Expert commentary was abundant. A typical distillation of critiques appeared when William Koviac, a former attorney for the FTC, was interviewed by the Washington Post: “The government has not introduced that much evidence to demonstrate that consumers are suffering grievous harm today at the hand of MS. It’s one of the weakest elements of the government’s case and one of the strongest elements of the MS defense.” An interesting comment came from Mark Patterson, a Fordham law professor: “Unless it’s very clear, the courts worry that the potential of the long term harm may never materialize.” Reiterating the conclusion that courts, not possessing a certified crystal ball, normally bore a skeptical view of any company, causing long-term hurt.

  The Feds’ technical experts told the judge that they could find little if any consumer benefit in the integration of Windows and IE. Judge Posner, in his book, Antitrust Law, made a wise c
omment about such experts: “There are very few genuine neutrals in the technical field judges can rely on.” And this one was not one of them. The benefits were obvious for consumers and software programmers alike! The judge himself did not ask too many questions. He fell asleep on the bench now and again—a bit detached.

  After the Feds had presented their witnesses, the onlookers were eager to glimpse at ours next. Our legal team decided to call me as witness eleven, with just one to go after me. Playing excerpts from Bill’s videotaped deposition the day the first MS witness showed up, Boies stunned the courtroom and alarmed our legal team. At once our lawyers contested the legality of his exploit; Jackson allowed it. Bill had done a lousy job during his interrogation, fighting David Boies tooth and nail every step of the way. Not picking the winnable fights, he looked unreasonable, had numerous lapses of memory, was evasive, and stubbornly refused to explain the most obvious details. A certain degree of complicity is necessary to a witness’ credibility. Bill wanted none of that.

  For the rest of the trial, David Boies mischievously introduced supplementary snippets of Bill’s video footage to convincingly manifest that our dear leader could not be trusted. It worked. The court of public opinion was turning sharply against him and consequently all of MS when played on TV nightly. True to form, Boies had achieved what he set out to do without putting Bill on the stand.