Bill Gates, the second-richest man in America, was working through his second round of visits this past summer at the Federal Trade Commission, a place as creaky and antiquated as Gates’s Microsoft Corporation is sleek and dynamic.
Gates had a potent lobbying team along: Microsoft’s chief in-house counsel, William Neukom; former FTC commissioner Patricia Bailey; and partners from New York’s Sullivan & Cromwell and from Preston Thor-grimson Shidler Gates & Ellis, the big Seattle-based firm where Gates’s father is a senior partner.
But in his visits to each of the four participating com¨missioners on two clammy July days, the 38- (then 37-) year-old Microsoft chairman did much of the talking himself. And at times he got a bit carried away-just as his company sometimes does in the heat of battle against competitors. Like when he barked, "You don’t know what you’re talking about," at an FTC official in a meet¨ing in the chairman’s office.
Or when Commissioner Dennis Yao, a soft-spoken former Wharton business school professor, floated a line of hypothetical questions suggesting possible curbs on Microsoft’s growing monopoly power, including disclo¨sure to competitors and the public of technical data about its dominant personal computer (PC) operating systems, MS-DOS and Windows.
Gates was vexed. "He started by calling Yao’s ideas socialistic," recalls a source familiar with the July 15 meeting, "and as he got angrier and angrier and louder and louder, he got into calling them Communistic."
Others recall it as "a spirited exchange, but not angry," in the words of Patricia Bailey, a partner in the Washington, D.C., office of Squire, Sanders & Dempsey. By all accounts, the $7 billion man displayed a remark¨able command of the issues and made some cogent argu¨ments against the wisdom of an antitrust attack against his company. "I was impressed that he … really knows his business and is passionate about his business," recalls FTC commissioner Deborah Owen.
The Gates visits capped a week in which lobbyists for Microsoft’s largest competitors had been clogging the halls of the FTC doing what one calls "dog and pony shows" for each of the commissioners. Along with the FTC’s Bureau of Competition, the Microsoft competitors were pushing for a complaint charging Mi¨crosoft with a long list of abuses amounting to predato¨ry monopolization of key computer software markets. The crowd included lawyers for Novell, Inc., which has been conducting a bitter public feud with Microsoft; Lotus Development Corporation; Borland Internation¨al, Inc.; Sun Microsystems, Inc.; and an International Business Machines Corporation-Apple Computer, Inc., joint venture called Taligent, Inc.
The increasingly urgent alarms sounded by such companies have dramatized how much is at stake, as Mi¨crosoft bids to dominate the entire PC software market and pushes to supply (and control) the software that will set industry standards for networks of computers, tele¨phones, copiers, and other office machines; interactive television; mobile "personal communicators"; and much more, as the computer, telephone, television, and pub¨lishing markets converge.
But it was Gates whose visit had the building buzzing. Admired, respected, envied, hated, and-perhaps above all-feared more than anyone else in the computer in¨dustry, he is a business legend on the scale of Andrew Carnegie, John D. Rockefeller, and Henry Ford. And for all his vaunted churlish outbursts, he seems to have the planet’s most potent brew of whatever mixture of bril-liance and aggression it takes to build from nothing a $4-billion-a-year, world-beating, job-creating, high-tech en¨terprise that now tops rival IBM, tops Texaco Inc., and nearly doubles The Boeing Company in stock market valuation.
Gates is widely considered to be the most valuable piece of human capital the United States brings to the struggle for global economic leadership. And Microsoft, the Redmond, Washington-based powerhouse that Gates and a friend started in 1975 after dropping out of Harvard University, is widely lionized as a flagship in America’s fleet of high-tech assets.
"We are a propellant for a very fast-growing industry," says William Neukom, Microsoft’s vice-president of law and corporate affairs. "Microsoft is the paradigm of an entrepreneurial, information-age company that employs skilled workers and competes in a global economy."
That’s why Microsoft’s sound bite of first resort in trying to stave off a governmental antitrust suit is the danger that such a suit could create an opening for foreign competition by muzzling the fierce drive that has enabled Microsoft to beat the world and bring in $2 bil¨lion a year in export revenue. (Never mind that all of its top competitors in the world market are also U.S. compa¨nies with large exports.)
Sue Microsoft for monopolization? Go after Gates for being too successful? Revert to the discredited old "big-is-bad" school of antitrust? Throw a litigation monkey wrench into the engine of a computer industry that’s changing so fast that nobody (except maybe Bill Gates) knows where it’s going? You’ve got to be kidding, growl devotees of free-market purity and believers in the entrepreneur-as-hero
The FTC won’t sue. But Anne Bingaman will. Indeed, the Clinton Justice Department’s assistant attorney general for antitrust was moving toward a virtually un¨precedented takeover of the FTC’s three-year investiga¨tion even as Gates was down the street at the FTC on July 15, tossing the "Communist" epithet at a mild-man¨nered economist.
Bingaman, a hard-charging former plaintiffs antitrust lawyer, likes to take on big cases against big companies. Based on an extensive but admittedly incomplete review of the available evidence, I’m laying 4-to-l odds that she’ll pick Microsoft as her first target-and even odds that Gates will eventually cop a consent decree that bars several types of monopolistic predation while leaving Mi¨crosoft intact and potent.
Neukom disagrees. "We think that at the end of the day, when all the evidence is out there, they will conclude that we aren’t doing anything in violation of the antitrust laws," he says. "We’re trying to compete in a fair and le¨gitimate way."
So was Neukom covering his bets-just in case his team does not persuade Bingaman-when he was seen entering Commerce secretary Ronald Brown’s office at 10 A.M. on October 8, in the company of Lloyd Meeds, a former congressman who is a partner in Gates’s father’s firm’s D.C. office?
Meeds says the meeting was to discuss Bill Gates’s planned public support for President Clinton’s push to pass the North American Free Trade Agreement, and to talk about developing software for national electronic su¨perhighways. Nobody even mentioned the antitrust inves¨tigation, he says.
But perhaps it crossed somebody’s mind that when the time comes to decide whether the Clinton administration should sue Microsoft for monopolization, Brown-a Clinton confidant and former Democratic National Com¨mittee chairman-will remember how Microsoft came through for the president on NAFTA, and how helpful an unfettered Microsoft could be on the electronic super¨highway front. Can’t hurt.
"I really shouldn’t say this," Bill Gates declared at a May 1981 PC industry conference, "but in some ways it leads, in an individual product category, to a natural monopoly."
The "it" referred to by the far-seeing Gates was the computer industry’s need for standards-that is, for uniformity in the functioning of key building blocks-to foster a broad market for compatible products, just as the uniformity of electrical outlets fosters a broad market for plug-in appliances. Twelve years ago IBM and Microsoft established Microsoft’s MS-DOS "operating system" software (which controls a computer’s most basic func¨tions and with which other software must be compatible) as the PC industry standard. Since then, DOS has be¨come so dominant that most PC makers pay Microsoft li¨cense fees to install MS-DOS at the factory, and the profits keep on mounting.
The only significant direct competitor in the PC operating system market in recent years has been a program widely known as DR DOS, now owned by Novell and renamed Novell DOS. Barely alive in the marketplace, Novell’s DR DOS is the victim-in-chief (amid a dozen or so companies clamoring for protection from Microsoft) of the allegations of monopolistic predation at the heart of the case prepared by the FTC antitrust staff.
Gates has scornfully dismissed DR DOS as an "attempted clone product." But a page of history raises some question about who cloned whom. For one of DR DOS’s prime assets-one of some relevance to the antitrust analysis-is the fact that Microsoft’s original MS-DOS borrowed heavily from a program called CP/M. And it, in turn, was written by Digital Research, Inc. (DRI), which later developed CP/M into DR DOS.
CP/M was on its way to being the industry-standard operating system for desktop computers before the IBM PC put the market on the map beginning in 1981. Thentiny Microsoft, which had been plucked from obscurity in 1980 to supply computer languages for the PC, was not much interested in supplying its operating system too. As detailed in several recent books, the best being Gates, by Stephen Manes and Paul Andrews, neither Gates nor IBM foresaw then that the big money-and control over the direction of the entire PC industry- would go to the company that owned the operating sys¨tem. Besides, Microsoft didn’t have one.
So Gates sent IBM to his then-friend Gary Kildall, the head of DRI and author of CP/M. Kildall went off on a business trip and left his corporate vice-president (who was also his wife) to meet with IBM about supplying it with an operating system. She balked at IBM’s one-sided nondisclosure agreement, and there was no deal.
IBM then asked Microsoft to come up with an operating system, in a hurry. Gates turned to Seattle Comput¨ing, which had written a CP/M clone called QDOS that mimicked CP/M’s technical features so that it could use existing software applications written to run on CP/M. Gates bought QDOS, after IBM had passed up the chance, for about $75,000. Microsoft renamed it MS-DOS, upgraded it a bit, and licensed it to IBM for the PC, while wisely retaining the rights to license it to others as well.
Thus was laid the foundation stone of Bill Gates’s fame, his $7 billion fortune (30 percent of Microsoft’s $24 billion stock market valuation), and his unrivaled power over the PC industry. The explosive success of the IBM PC created a huge market for applications software (word processing, spreadsheets, graphics, and other functional programs) compatible with MS-DOS. When IBM’s high prices left an opening for clone-makers like Compaq Computer Corporation and Dell Computer Cor¨poration to seize most of the market, they needed to in¨stall MS-DOS on their IBM-compatible computers. Mi¨crosoft licensed it to them-on its terms.
MS-DOS now runs on 80-90 percent of the world’s 130 million IBM-compatible PCs-a monopoly by almost any antitrust law definition. It sold over 20 million copies in 1992 alone, for more than $750 million, ac¨cording to Dataquest, Inc., a computer industry informa¨tion service. Most of this is pure profit: the cost of licens¨ing copies is negligible, and the cost of marketing, updating, and upgrading is modest.
Microsoft’s success wasn’t all borrowed technology, luck, and IBM’s coattails, of course. It has unexcelled prowess at converting others’ ideas into first-rate products. Gates has outthought, outworked, and outsold everyone else in the industry. He has outnegotiated IBM and others again and again, beginning with his reservation of the right to sell MS-DOS to clone-makers.
He hired smart people just out of school and spurred them to put in 16-hour days at Microsoft’s campuslike complex of more than 20 buildings on 260 acres, 15 miles east of Seattle. He imbued the whole company with his own ferocious drive to beat the competition both in bringing good software to market and in selling as hard as necessary to win every contract.
Keeping The Monopoly
There’s no law against getting a monopoly, or keeping it, by competing hard and winning. But the FTC antitrust lawyers concluded that MS-DOS owes at least part of its long reign to monopolistic predation by Microsoft, aimed at killing off one potential challenger and scaring off others.
At a glance, the MS-DOS monopoly might seem an inviting target for people wanting a piece of it. Huge pots of money tend to attract competitors. And "nobody has ever pretended MS-DOS is good software," as Charles Ferguson and Charles Morris wrote in their 1993 book Computer Wars. Its arcane commands have driven users bonkers from the start; techies have long cursed its limi¨tations; and upgraded versions are still tied to the flaws of the original.
But there were big barriers to entry in the DOS market. It would take a large initial investment, long before any revenue would come in, to write a new operating system competitive with the upgraded versions of MS-DOS. To recoup this investment the new operating system would have to sell millions of copies, against an entrenched competitor prepared to cut prices to the bone to hold market share.
Any MS-DOS competitor would also need to emulate the functions and programming interfaces of MS-DOS (as DR DOS does) in order to work with the huge installed base of applications software that has been writ¨ten on the MS-DOS standard. And that could risk a law¨suit for violating Microsoft’s intellectual property rights.
The company in the best position to fend off any claims of illegal copying was DRI, whose parentage of CP/M gave it a plausible claim to being a grandfather (or at least great-uncle) of MS-DOS.
DRI introduced DR DOS in 1988, becoming Microsoft’s only direct competitor in operating systems for IBM-compatible desktop PCs. By 1991 it looked like it might be getting somewhere. The release of DR DOS 5.0 in April 1990 and of DR DOS 6.0 in late 1991 brought fa¨vorable reviews in industry trade journals and some awards. Many (though not all) technical analysts said DR DOS 5.0 was a considerable improvement on the then-current version of MS-DOS.
DR DOS got another boost when Novell, the world’s second-largest software company after Microsoft, bought DRI in October 1991, for about $125 million in stock. This put real financial and marketing clout behind the product. By early 1992, DR DOS surpassed MS-DOS in retail sales-a relatively small percentage of the overall market, to be sure, but a sign of growing acceptance.
Then the bottom dropped out. Beginning about March 1992, DR DOS retail sales began to drop, then to plunge. DRI and Novell had very little success cracking the most important channel of the operating system market-the millions of copies of MS-DOS licensed by PC makers for factory installation. By early 1993 many analysts wrote DR DOS off as dead (despite a poll in the July 1993 PC Magazine finding it "the clear winner" in customer satis¨faction over MS-DOS).
What happened? Microsoft had not been asleep, of course. DR DOS was more than a trivial threat. If it had captured even 30 percent of the DOS market, that could have cost "hundreds of millions of dollars a year out of Microsoft’s bottom line," notes David Bradford, Novell’s senior vice-president and general counsel. "They had to stop DR DOS on the beaches," one anti-Microsoft market analyst says with a bit of hyperbole, "because once DR DOS gets a beachhead, MS-DOS is dead meat."
Microsoft met the competition by pouring more resources into upgrading MS-DOS to match the new fea¨tures offered by DR DOS (after having let MS-DOS stag¨nate for years), and by cutting prices as sharply as necessary to dissuade PC makers from switching.
Microsoft sales executive Mark Ursino is quoted in Gates, saying that Gates had told him: " ‘If somebody would be dumb enough to get DR DOS instead of MS-DOS as their sole bundled operating system, [Ursino should] get that deal no matter what it took.’ "This, Ursi¨no explained, meant that "we’d pull out all the big guns and sit down and say, ‘What would it take? How do we buy you out of this deal?’ Anything went. That’s why the pricing was never published."
All perfectly legal.
How DR DOS Got Sick
But that’s not all Microsoft did, according to evidence and allegations compiled by FTC investigators. They, and Novell’s lobbyists, have told the commissioners and now the Justice Department that Microsoft also engaged in a systematic campaign of predation aimed at destroying DR DOS and the threat that it represented to the MS-DOS monopoly.
First, almost immediately after DR DOS 5.0 was introduced in April 1990, Microsoft sought to freeze sales of DR DOS by vowing to bring out its own MS-DOS 5.0, with all kinds of new features both matching and besting DR DOS 5.0. That was legal enough. But anti-Microsoft lawyers say Microsoft crossed the line by sug¨gesting it would deliver MS-DOS 5.0 in 1990; in fact it took until June 1991, and even then lacked some of the predicted new features.
Unrealistically optimistic projections of when new products will be ready are a routine business tactic for freezing competitors’ sales, of course-so much so in the software industry that there’s a word for it: "vaporware." Microsoft had done this dating back to its days as a tiny start-up.
Ordinarily such unrealistic preannouncements do not raise any antitrust issues. But some cases hold that a firm with monopoly power can not make deliberately false preannouncements to kill off competition. Novell claims Microsoft knew all along that it could not possibly meet its predicted timetable for DOS 5.0. Microsoft has said it acted in good faith and did not anticipate the delay.
The second, and most devastating, of Microsoft’s allegedly predatory tactics was use of "CPU licenses" (which competitors call "CPU lockout contracts") to pressure PC makers to deal exclusively with Microsoft and shun DR DOS. These standard-form contracts, which Microsoft has induced most PC makers to sign, re¨quire them to pay Microsoft a fee for every machine shipped that contains designated types of central processing units (CPUs), whether or not the PC makers want or use MS-DOS on every machine.
Some PC makers have told antitrust investigators (according to sources close to the investigation) that Mi¨crosoft pressures them to sign such contracts both by of¨fering irresistible inducements, such as steep discounts on MS-DOS, and by making implicit threats.
"If somebody doesn’t like you at Microsoft, very subtle, bad things can happen to you," contends a former Microsoft executive. For example, computer makers are dependent on getting timely shipments of technical data from Microsoft to ensure compatibility between their hardware and its operating systems. "Suppose one gets lost in the mail? Or-‘Gee, where did that Fed Ex rout¨ing number go?’ "says this source.
These CPU licenses have made it virtually impossible for DR DOS to make sales to PC makers, who are locked in a bloody price war and have no incentive to try DR DOS on even a small percentage of the machines cov¨ered by MS-DOS CPU licenses. Who needs two operat¨ing systems?
Nor is Novell the only company worried about being excluded from the operating system market by CPU licenses. IBM and Apple are also among those worried- and they have discreetly lobbied the FTC and the Justice Department (via their joint venture Taligent, represented by Edward Glynn, Jr., of D.C.’s Venable, Baetjer, Howard & Civiletti) to sue to ban Microsoft’s use of CPU licenses.
So the combined forces of Apple and IBM are running to the government for protection from monopolistic pre¨dation by Bill Gates. That’s the same Bill Gates who was mistaken for an office boy by the IBM executive who first visited Microsoft’s tiny operation in 1980. And the same antitrust-averse IBM that spent 13 years staving off a mammoth Justice Department monopolization suit.
The two computer giants launched Taligent with great fanfare two years ago to create an advanced new operat¨ing system by 1995, in the hope of breaking the Microsoft monopoly. But many industry analysts predict Microsoft will squash Taligent like a bug. "IBM and Apple have no clout in this market," says one consultant. That very ex¨pectation speaks volumes about the power and durability of the Microsoft monopoly.
Microsoft defends its CPU licenses as only one of the alternative pricing deals it offers PC makers, and as justi¨fied by the need to protect against piracy-i.e., the kind of theft-by-illegal-copying that plagues all software compa¨nies. If PC makers pay to put MS-DOS on only (say) 80 percent of their machines, Microsoft reasons, some of them might cheat by secretly making and using more copies than they paid for. Others might ship "naked" com¨puters without operating systems to discounters or cus¨tomers, who might themselves install pirated MS-DOS. Microsoft has also analogized CPU licenses to legal quan¨tity discounts.
This "piracy" rationale seems at best weak, as the FTC antitrust staff concluded. There are other, less exclusion¨ary ways of protecting against piracy. And whatever in¨cremental protection Microsoft might gain from CPU li¨censing is tiny compared with its exclusionary effects on competition. As for quantity discounts, they are typically graduated based on number of units bought and reflect the lower marginal cost of selling in bulk. But the FTC concluded that the very steep discounts that Microsoft of¨fers to induce PC makers to sign CPU licenses are based not on quantity but on buying MS-DOS for an entire line of machines, and do not correlate with any plausible mea¨sure of marginal cost.
The third of Microsoft’s allegedly predatory tactics was a campaign to impede Novell’s efforts to make DR DOS work with Microsoft’s popular Windows software, and to disparage DR DOS. The Novell operating system, Mi¨crosoft suggested, was incompatible with Windows and likely would cause problems in machines running Win¨dows.
(Windows is a "graphical user interface"-GUI, pro¨nounced "gooey"-designed by Microsoft to be used on IBM-compatible PCs with MS-DOS. It mimics the easy-to-use look and feel that has made the Apple Macintosh such a success, with on-screen pictures responding to commands from a point-and-click "mouse.")
The "Windows incompatibility" scare had a devastating impact on sales of DR DOS in 1992, Novell claims. (Others say DR DOS sales fell because of Microsoft’s up¨grade of MS-DOS.) Windows 3.0 had become the most successful product in PC software history after Gates launched it in May 1991. By 1992 it had grabbed over 90 percent of the GUI market; by August 1993 it had sold over 30 million copies. Most PC buyers want Windows, and most PC makers want to install it at the factory. They also want to be sure that it will work well with whatever operating system is installed. And any competitor that wants its software to be able to run Windows programs without problems-real or concocted-is at the mercy of Microsoft.
Gates and other Microsoft executives have assiduously raised doubts about DR DOS’s compatibility with Windows. But according to Novell, the only compatibility problems were caused by Microsoft’s unusual exclusion of Novell (alone among software companies) from its Windows 3.1 "beta" (user-testing) program; the problems were largely fixed by May 1992. Still, Microsoft conducted a continuing publicity campaign to raise what Novell says were false fears about DR DOS.
FTC investigators also concluded that in order to sabotage DR DOS, Microsoft had carefully written and hid¨den a batch of code into tens of thousands of beta copies of Windows 3.1 that were sent to expert computer users in December 1991. When someone tried to run one of these Windows 3.1 beta copies on a PC using DR DOS (or any other non-MS-DOS operating system), the screen would display the following message: "Nonfatal error detected: error 4D53 (Please contact Windows 3.1 beta support.) Press C to continue."
To expert beta-testers using DR DOS with Windows, this message would convey that they could continue using the program, but it might cause problems. The effect would be to deter some from using DR DOS further; others would call Microsoft for an explanation of the sup¨posed risks of using DR DOS.
The "apparent purpose of this code is to lay down arbitrary technical obstacles for DOS-workalike programs," Andrew Schulman, a respected programming expert and author in Cambridge, Massachusetts, writes in the forth¨coming second edition of his book, Undocumented DOS. He says this seemed "a deliberate incompatibility" that would scare Windows beta-testers "into not using DR DOS." Schulman stresses that Microsoft had carefully encrypted this code to disguise its purpose and mode of operation and evade detection, and that he and another programming expert had cracked the code only by days of reverse engineering.
In a May 1993 letter to FTC investigators who had ap¨proached him earlier, Schulman said this had been an "at¨tempt to thwart discovery, and is more the sort of thing one expects from a teenager writing a virus than from a multibillion-dollar corporation."
Schulman has often defended Microsoft against what he calls "scapegoating from the second-tier software companies such as Novell, Lotus, WordPerfect, and Bor¨land." But he adds, in an interview, that he now believes Microsoft has engaged in some questionable competitive tactics. "The best scapegoat is someone who’s a little bit guilty," Schulman notes.
Microsoft has denied ever inserting code into any of its software for the purpose of making it incompatible with a competitor’s. It has told investigators it included the "nonfatal error" message in Windows 3.1 beta copies be¨cause it needed to warn users that Microsoft had tested Windows for compatibility only with MS-DOS, and could not guarantee that DR DOS or any other operating system would not cause problems. The code underlying the warning had been encrypted, Microsoft has explained, lest companies selling MS-DOS clones rig their own software to disable it.
FTC investigators found such explanations unpersuasive. If, as Microsoft claims, Microsoft’s purpose was simply to warn that it could not guarantee that Windows would work well with any non-MS-DOS system, it had already made that clear elsewhere in the materials it pro¨vided beta-testers. Why have an on-screen "error mes¨sage" too? And what was the error? "It seems," in Schul¨man ‘s words, "that the only ‘error’ is that the user is running Windows on someone else’s version of DOS."
(During Gates’s July 16 visit to lobby FTC chairman Janet Steiger, an official recalls, an FTC lawyer tossed Gates a "gotcha" question implying that Microsoft had disparaged DR DOS to beta-testers who called in about the error message. Gates’s angry retort-"You don’t know what you’re talking about"-sent "an emotional blast through the room," says this official; then "his lawyers jumped in to change the subject and cool things down," without answering the question directly.)
Microsoft dropped its initial plan to keep an on¨screen warning in the commercially shipped version of Windows 3.1 But the disparage¨ment of DR DOS as incompatible continued. Microsoft vice-president Brad Silverberg, for ex¨ample, told PC User in a May 1993 interview that MS-DOS’s 90 percent market share was safe-partly be¨cause before trying DR DOS, PC buyers should and would ask themselves: "Why take the risk with all the compatibility problems that DR DOS has had?" And the Windows 3.1 bundled with a laptop I bought in August comes with the ominous warning that running it "on an operating system other than MS-DOS could cause unex¨pected results or poor performance."
The DR DOS-Windows compatibility scare is closely related to a broader issue: Microsoft’s use of secret fea¨tures known as "undocumented calls" in both MS-DOS and Windows to the disadvantage of competitors.
"Call" is software shorthand for "applications pro¨gramming interface": a group of technical features in operating software that enable it to respond to com¨mands sent by complementary software products. A competing software firm cannot take advantage of calls that Microsoft does not document publicly (or disclose to that firm)-not, at least, without time-consuming re¨verse engineering, and without risking that Microsoft will make its software obsolete by removing the undoc¨umented calls from the next release of MS-DOS or Windows.
Undocumented calls are common in software, and are not necessarily illegitimate. It would be difficult to docu¨ment every conceivable program interface, and wasteful to document features that have not yet been perfected or may not be retained in future versions.
Still, the Justice Department has received numerous allegations that Microsoft has secured an unfair advan¨tage by using undocumented calls to make MS-DOS and Windows run more smoothly with one another and with Microsoft applications, like Word and Excel, than with competing products.
Until the summer of 1992, Microsoft had repeatedly denied that its applications software relied on any undoc-umented calls in Windows. These denials were proven false by Schulman in his 1992 book, Undocumented Win¨dows. Microsoft was forced to admit that Word and Excel use at least 16 undocumented calls in Windows, suppos¨edly in violation of Microsoft internal policies. The com¨pany has continued to maintain that this gives its applica¨tions developers no unfair advantages. Competitors ask why, if that’s so, Microsoft relies on these undocumented calls at all; if they’re important enough to use, they’re im¨portant enough to document.
Indeed, Microsoft itself seems to concede that its own reliance on undocumented calls has the effect of excluding competition, at least in operating software. In questioning DR DOS’s compatibility with Windows, Microsoft execu¨tives, including Brad Silverberg, have said that Windows relies on so many undocumented calls in MS-DOS that DR DOS (or any other competing operating system) could not easily achieve compatibility with Windows.
That’s the one-two Microsoft punch: First it uses tin-documented calls in its monopoly operating systems that hinder competitors from making their software fully com¨patible; then it disparages the competitors’ products as in¨compatible.
Nor is DR DOS the only alleged or potential victim of such conduct. Microsoft has served notice on the software market this year that the company "plans new versions of Windows and MS-DOS that will make it "virtually im¨possible’ for Novell, Sun [Microsystems], or IBM to guarantee that their operating software will run future Windows and MS-DOS applications," as The Wall Street Journal said on March 15, 1993, quoting Michael Maples, a Microsoft executive vice-president. And in May Bill Gates implied that Microsoft would use intellec¨tual property litigation to hammer any competitors that nonetheless succeed in replicating Windows functions.
Growing The Monopoly
"If someone thinks we’re not after Lotus and alter Word¨Perfect and after Borland, they’re confused," the same Michael Maples declared at a November 1991 briefing for reporters and analysts. "My job is to get a fair share of the software applications market, and to me that’s one hundred percent."
"In the old days," observes an FTC official, "antitrust counsel would cringe at that kind of talk."
But Microsoft makes little effort to conceal its hege¨monic designs. Bill Gates is not content merely to domi¨nate the PC operating software market. He wants it all: the larger, but so far more competitive, market for major business applications like word processing, spreadsheet, and graphics programs; the market for networking soli-ware that has long been dominated by Novell-software for tying computers to fax machines, copiers, printers, and phones; the market for wireless data receivers; and on from your office to your home-software for hand¨held computers, personal digital assistants, CD-ROM disks with multimedia encyclopedias and golf games, (he coming 500-channel cable and interactive television, and whatever else lies at the end of the "electronic superhigh¨ways" that the Clinton administration is so excited about building. In Bill Gates’s vision, these will all be rolled into one big Microsoftian ball.
Since the smash success of Windows 3.0 in 1990, Mi¨crosoft has surged to or near market leadership in the major categories of business applications software for IBM-com¨patible PCs. It dominates applications for Windows, the most dynamic and profitable segment of the market.
Microsoft’s Maples is not closing in on 100 percent ye(, but he’s moving in the right direction. Meanwhile, Mi¨crosoft’s largest competitors in PC applications-Lotus, Borland, and WordPerfect Corporation-are all com¨plaining that it has used predatory tactics to beat them out.
The success of Windows benefited many smaller soil-ware companies supplying niche products. But Mi¨crosoft’s growth has also come at the expense of its biggest rivals. Borland, in particular, has been struggling to stay profitable over the long term and has seen its stock plunge with analysts doubting its viability. Lotus has had a very good year, but one industry lawyer says that "all those companies are going to have a hard lime in the long run," prospering in a Microsoft-dominated world. With $2 billion in cash and no debt, Microsoft is in an enviable position to wage price wars and outspend competitors on research, development, and promotion.
(Microsoft also poses a potentially mortal long-term threat to still-prospering Novell’s main asset: its NetWare’s 60-70 percent share of the market for networking software, which Gates has long coveted.)
The central charge by these losing competitors is that Microsoft has been using illegal arrangements unfairly to exploit the monopoly power of MS-DOS and Windows, in an attempt to monopolize the biggest applications software markets.
Competitors say Microsoft has tied sales of its applications to those of its operating systems through various package deals. They say it has also engaged in "technological tying"-updating MS-DOS and Windows in ways designed to raise the costs of applications competitors, and using undocumented calls and other devices to give Microsoft’s own applications programmers earlier and better information than it gives to its biggest competitors about how best to interlace with DOS and (especially) Windows.
Competitors also charge that Microsoft has victimized them through a process that some call "brain suck": They claim that the Microsoft operating software division, which often obtains trade secrets about other companies’ applications software in the process of working out compatibility problems, sometimes shares these secrets with Microsoft’s own applications division.
While Microsoft denies these charges, and some independent analysts are skeptical of them, it’s clear that Microsoft’s control of MS-DOS and Windows at least gives it the opportunity and incentive to provide its applications soft- ware developers with an advantage over their competitors. That’s because applications software can work well only if those designing it have intimate and timely knowledge of the technical features needed to work smoothly with Microsoft’s currentand future operating software.
For this reason, a few computer industry leaders, notably Steve Jobs, the co-founder of Apple and current head of NeXT Computer. Inc., have called for breaking Microsoft up by splitting its operating and applications software divisions into two companies. They reason that any steps short of a breakup will be ineffective at restraining Microsoft from leveraging its monopoly power. This logic had some appeal to some at the FTC, but was never seriously entertained at high levels, according to officials.
For one thing, there is nothing necessarily illegal about Microsoft gaining competitive benefits from the undoubted efficiencies of vertical integration; Microsoft opera-lions systems experts may swap ideas with applications programmers over a pizza. The legality of such conduct becomes debatable only if and when Microsoft deliberately denies to competing applications providers critical information about MS-DOS or Windows that is available to its own people.
For years, Microsoft fended off claims that it had an unfair advantage by suggesting that it had some kind of internal Chinese wall preventing any unfair leakage of information. "There is a very clean separation between our operating system business and our applications software,’" Steve Ballmer, Gates’s best friend, chief lieutenant, and fellow Microsoft stock billionaire, told Business Week in 1983. "It’s like the separation between church and state."
But the irrepressible Mike Maples, who sometimes seems almost to be trying to give the company’s antitrust lawyers heartburn, declared such statements inoperative in a late 1991 interview with InfoWorld. "There is no Chinese wall," Maples declared. "We don’t want there to be a Chinese wall, and I don’t think we’ve ever claimed that there is a Chinese wall."
About the same time, as if to underscore the "no Chinese wall" point. Gates put Maples in charge of both Microsoft’s operating systems and its applications.
Some independent analysts as well as Microsoft itself dismiss the complaints of unfair competition as the whining of losers; they contend that Gates has been winning by being smarter and getting improved products out the door faster. And Lotus, Borland, and WordPerfect have all been widely criticized in the industry for failing to react quickly enough to the success of Windows and to produce first-rate applications for it. Still, one reason they were slow was that Microsoft was encouraging them during late 1989 and 1990 to develop applications for a new operating system called OS/2-while quietly refocusing its own efforts on applications for Windows.
The Government Gets Into The Game
When the FTC’s Bureau of Competition started sniffing around in June 1990, the issue was not alleged monopolization, but possibly illegal collusion between Microsoft and IBM.
At the November 1989 Comdex, a PC industry convention, Gates and IBM’s top PC executive had announced that both companies jointly would focus their efforts on OS/2, a new PC operating system that was to replace Microsoft’s aging MS-DOS as the industry standard. An IBM-Microsoft press release said Microsoft would encourage industry acceptance of OS/2 by holding back from Windows-which (in combination with MS-DOS) would compete with OS/2-valuable features that Gates had been developing for years and putting them into OS/2.
The FTC lawyers wondered whether this was a horizontal conspiracy to limit competition in PC operating systems by strangling Windows in its crib, and consigning MS-DOS to premature obsolescence. A logical question-but one so rapidly overtaken by events as to be almost laughable in retrospect, as well as a reminder of the need to craft any antitrust decree carefully to avoid instant obsolescence amid such rapid metamorphosis.
The first version of OS/2 was destined to be a disastrous flop in the marketplace. The new and improved Windows 3.0 had begun taking the industry by storm a month before the FTC started investigating in June 1990. Far from conspiring with IBM, by late 1989 Microsoft had been moving to end the by-then-tattered relationship that Gates had ridden to industry leadership.
But the FTC investigators did not stop when their original premise collapsed. Nor should they have. They were being deluged with complaints from Microsoft competitors, customers, and others that Gates and company had committed myriad acts of predation to maintain their PC operating software monopoly and extend it to the applications software market. So the FTC lawyers re-focused their inquiry on whether Microsoft had abused its monopoly power.
One such complaint was that what Gates was really up to at the November 1989 Comdex was a "head fake"-a plot to send IBM and competing applications companies down a blind alley. While they poured time and money into developing applications for OS/2, Microsoft would (and did) race ahead with Windows and applications for it.
In Computer Wars, industry analysts Ferguson and Morris theorize that Gates may have plotted from early on to "[ensure] OS/2’s failure," by having it written in a code that Gates knew would soon be obsolete. And Microsoft never delivered on a series of pledges to support OS/2 by developing features and applications for it.
The case that Microsoft deliberately misled the rest of the computer world when it announced the OS/2 effort at the 1989 Comdex-a charge it denies-seems less than compelling. The tensions between Microsoft and IBM were already evident by then, with Gates openly resisting IBM pressure to put Windows on the back burner. It is plausible that Microsoft thereafter increasingly pulled away from OS/2 in response to its deteriorating relationship with IBM and to changes in the fast-moving software market. And even if Gates did do a head fake in November 1989, it may not have violated antitrust law, because it involved no direct use of Microsoft’s monopoly power in the MS-DOS market. (Windows did not yet have such power in the GUI market.)
Some of the other allegations of Microsoft predation and attempted monopolization seemed more solid, however, including the claim that in 1990 Microsoft encouraged competitors to continue to develop applications for OS/2 after Microsoft had privately decided to dump it and had shifted its efforts to Windows. And FTC lawyers could not but be impressed at how fast Microsoft’s market power was growing, even as their investigation unfolded.
Which was slowly. For most of the three years, Marc Schildkraut, the veteran FTC antitrust litigator in charge of the investigation, and Norris Washington, the lead full-time investigator, had only two recent law graduates helping them. "[Schildkraut] is real good, but the case has been woefully understaffed," says a software company lawyer. The FTC investigators did not lack for help from Microsoft rivals, however.
At least ten companies, including the four biggest software vendors in the world after Microsoft-Novell, Lotus, WordPerfect, and Borland-have lobbied the FTC and now the Justice Department to tackle Microsoft. Some have supplied sworn affidavits alleging various predatory acts, expert economic and technical reports, and detailed legal analyses, as well as huddled with FTC and now Justice Department investigators. From these submissions emerges a portrait of Gates and company as ruthless monopolists with a bold disregard both for antitrust law and for simple honesty and fairness in business dealings.
Novell has been the most active of Microsoft’s accusers, especially since overtures by Gates to buy Novell for more than $12 billion-a handsome premium- broke down after eight months in March 1992. Novell has since complained that the merger overtures were another Gates head fake all along, designed to slow down Novell’s push to take a chunk of the MS-DOS monopoly with its competing DR DOS operating system.
Novell convened a meeting to compare notes with lawyers for other Microsoft competitors on September 29, 1992, at Novell’s Dulles Airport office, outside Washington. Attendees included Novell’s in-house counsel, David Bradford, and its then-D. C. counsel, Sturgis Sobin of Washington’s Ablondi, Foster & Sobin; WordPerfect’s in-house counsel, Duff Thompson; Lotus’s FTC counsel, Andrew Berg of Washington’s Akin, Gump, Strauss, Hauer & Feld; and lawyers for Borland, Logitech Inc., and GO Corporation, which all had complaints of dirty tricks by Microsoft.
Novell was (and is) considering a private antitrust suit against Microsoft, but Lotus counsel Berg urged all those present to wait lest a private suit give the FTC a legitimate reason not to use its scarce resources to sue. Besides, only a governmental suit offered any realistic prospect of immediate relief via a preliminary injunction.
This meeting, and subsequent conference calls, have not led to close coordination of the anti-Microsoft lobbyists’ strategies. In some respects the agendas of the two largest Microsoft rivals have been in conflict. Novell, which sells only computer operating systems like DR DOS and NetWare (and which has its own quasi-monopoly in networking software) has urged government lawyers to bring (at least at first) only a narrow case, focused on Microsoft’s alleged monopolization of the PC operating systems market. Lotus, on the other hand- hurting from Microsoft’s surge in applications, especially spreadsheets-has urged a broader case alleging attempted monopolization of key applications markets as well, while warning that that it may be too late to salvage competition in PC operating systems.
Microsoft’s lawyers have been busy too, of course. They produced more than 500,000 pages of documents, reams of electronic data, and 20 witnesses for the FTC, while meeting with FTC staffers almost 20 times from mid-1990 through this summer, according to Microsoft chief in-house counsel Neukom, who has directed the company’s strategy.
Neukom, who started representing Microsoft in the late 1970s when he was a partner at Preston Thorgrimson Shidler Gates & Ellis, has assembled a formidable outside legal team for the antitrust investigations, including Hugh Bangasser and David Binney of Preston Thorgrimson; Donald Kaplan and James Weiss of its D.C. office; Steven Holley and Richard Urowsky of Sullivan & Cromwell; and Patricia Bailey, the former FTC commissioner, whose most conspicuous assets were a close friendship with FTC chairman Janet Steiger and close relationships with two top FTC staffers-Mary Lou Steptoe, acting director of the Bureau of Competition, and James Hamill, Chairman Steiger’s top staff person, both of whom were once attorney-advisers to Bailey.
Neukom and Microsoft outside counsel declined to comment in detail on the matter. (Caveat reader: For obvious reasons, some Microsoft accusers were far more willing to detail their evidence and arguments for me than were Microsoft’s lawyers.) Neukom explains that his department’s policy "is not to try cases in the press," and that it would be "inappropriate to comment on the merits" of the nonpublic inquiry at this point.
Face-off At The FTC
The FTC lawyers were aware of the self-interested agendas of the anti-Microsoft lawyers, but glad to make use of their time and talents. "Any lawyer inside any federal agency, if they can get outside people to work for them, why not?" explains one anti-Microsoft lawyer who billed a lot of time drafting materials and gathering evidence requested by the FTC staff.
By late 1992 the FTC’s Schildkraut, Washington, and about seven colleagues had churned out hundreds of pages of confidential findings and conclusions including allegations of a long list of predatory practices by Microsoft. But the staff was under pressure from some commissioners to keep the case narrow, and not to seek a complaint that might point toward breaking up Microsoft as a logical remedy and invite comparisons with the Justice Department’s disastrous U.S. v. IBM. So the staff put the broad case on hold; its legal premises were shakier than those of the narrow case, since Microsoft had clear monopoly power only in operating systems, and the factual disputes would be enormously complex.
Instead, the Bureau of Competition, headed by acting director Steptoe, proposed several alternative approaches. The most prominent was a suit seeking a preliminary court injunction-which would have been an unprecedented step in a monopolization case. Such an order would bar Microsoft from using blanket CPU licenses; from using exclusionary tying practices in the DOS market; and from purposely creating incompatibilities between its software and other operating systems. The staff held in reserve the idea of broadening its attack later. (The FTC’s Bureau of Economics, which rarely supports antitrust complaints, recommended against one in this case.)
While Schildkraut and other staffers spent weeks of 15-hour days preparing for a possible trial, lobbyists for both Microsoft and its rivals got in their last licks with the commissioners at the FTC. "We were all running around the halls not supposed to know who each other was and what each other was doing," one lobbyist recalls. "It was all shadow-boxing."
The commissioners met on February 5 for three hours, as a crowd of reporters and stock analysts waited and Microsoft’s stock price gyrated, swinging Gates’s net worth up and down by hundreds of millions of dollars. Then the commissioners emerged without even acknowledging the existence of the 32-month-old-but-still-officially-secret Microsoft investigation.
Word quickly leaked that they had deadlocked 2 to 2, with Chairman Janet Steiger and Dennis Yao, the commission’s only economist (and only Democrat), voting to seek a preliminary injunction against Microsoft, and Deborah Owen and Mary Azcuenaga (both lawyers) voting no. (The fifth commissioner, Roscoe Starek III, has been recused from the matter because of computer stock tied up in a family trust.)
Had all the high-paid lobbyists had any effect? Yes, in that the FTC staff might have lacked the resources to put together a respectable monopolization case without the help of the lawyers for Microsoft’s competitors. No, because in the end, the commissioners voted their philosophical predispositions. Both sides presented their cases well, by most accounts. But Steiger and Yao are antitrust liberals who voted as one would have expected; Owen is a conservative who has become an increasingly ardent champion of Microsoft since February; Azcuenaga, the most likely swing vote, is moderately conservative and so cautious that some staff lawyers from the outset had seen little chance of getting her vote.
The investigation remained open after the February 5 vote because Azcuenaga wanted more information about the claims of deliberate incompatibility. Some FTC staffers also took hope from rumors that Owen, whose term would expire in September 1994, was planning to take another job before then. That would clear the way for a 2 to 1 vote to sue Microsoft. But Owen stayed on. That seemed to spell the end of the idea of seeking a court injunction-arguably the only chance for an effective remedy. But some staffers held out a long-shot hope of persuading Azcuenaga to vote for a relatively narrow complaint to be tried by an FTC administrative law judge.
As the second FTC vote set For July 21 approached, Microsoft rivals made a renewed push, with Berg churning out more legal and factual analysis for Lotus and Novell bringing in an Arnold & Porter team including former FTC general counsel Michael Sohn, one of Washington’s leading FTC advocates, and Hadrian Katz, a computer whiz. But most FTC insiders were still betting (correctly, as it turned out) that Azcuenaga wouldn’t budge.
Meanwhile, Microsoft and its lawyers showed no interest in seeking a settlement. "Microsoft has told the FTC, ‘Go pound sand,’ " an anti-Microsoft lawyer said at the time. "That’s not just empty bravado on Bill Gates’s part, although he’s full of empty bravado. It’s a sound position."
Why should Microsoft give an inch? Another 2-to-2 vote would be a victory that it could (and later did) hail as a glorious vindication. Besides, Microsoft’s lawyers were well aware that the FTC may be institutionally incapable of litigating a big, rich company into submission, so hobbled is the agency by scant resources, by jurisdictional limitations, and by the commissioners’ chronic inability to resolve major administrative complaints in under five to seven years.
Microsoft had every reason to believe it was almost home free.
Justice Takes Over
Enter Anne Bingaman, Clinton’s new trustbuster.
The possibility that the Justice Department would sue Microsoft had barely (if at all) crossed the minds of most lawyers close to the case until sometime in July. One had actually started a rumor that Bingaman might intervene, but not (at least at first) with the idea that she would, according to a knowledgeable lawyer. Rather, it was hoped that Azcuenaga-who FTC staffers say had been angling for Clinton to appoint her as chairman, replacing Steiger-might feel pressure to vote out a complaint.
So strong was the tradition of FTC and Justice Department trustbusters slaying out of one another’s business, and of not subjecting companies to duplicative investigations, that Justice Department intervention was almost unthinkable even to some whose clients stood to benefit.
"It would just be an earthquake, a real bureaucratic shock wave through the system," one such lawyer said in mid-July. "You get one bite at the apple, not two."
But Bingaman had had months to think about getting a piece of this apple. At least one lawyer not directly involved in the Microsoft investigation had suggested to her even before her April 29 nomination that if she got the antitrust job, she might be able to make some good law (and a big splash) by jumping into the Microsoft matter. A sell-described "unabashed and enthusiastic supporter of vigorous antitrust enforcement," Bingaman has cited trust-busting New Dealer Thurman Arnold (a founding partner of Arnold & Porter) as a role model and has signaled a yen to bring the kinds of big cases against big companies that the Reaganites shunned.
In addition, "people in Silicon Valley"-including some of those seeking action against Microsoft-"know her and have known her for many years," says another lawyer (referring to her time in private practice in, among other places, the Santa Fe office of Phoenix’s Brown & Bain, which is counsel to Apple). "None of them know her in a way that would influence her decision, but the people here know that she will give this careful consideration," this lawyer explains.
Besides that, this lawyer stresses, career antitrust division lawyers have been watching Microsoft’s growing dominance with concern since 1991. At least two of them-including Richard Rosen, chief of the antitrust division’s communications and finance section-had previously worked at the FTC with Schildkraut and Steven Newborn, a senior antitrust litigator who had joined the anti-Microsoft team in April. Newborn has also been Rosen’s close friend since law school. The antitrust division lawyers could be expected to treat with respect the FTC staffers’ considered judgment that Microsoft had illegally monopolized the PC operating system market.
"There’s close connection and new convergence of thinking between the FTC staff and the Justice Department staff," the lawyer says. "They’re on the same antitrust thinking wavelength."
With the FTC apparently headed toward another 2-to-2 deadlock, the case was ripe for the plucking. Bingaman plucked it, in a series of steps that combined boldness with political finesse-and that have created a widespread expectation among lawyers working the investigation that she won’t stop short of suing or extracting a consent decree barring specified acts of predation.
If Bingaman had needed a formal invitation, it was readily supplied by Senator Howard Metzenbaum, the Ohio Democrat who heads the antitrust subcommittee of the Judiciary Committee-and whose Senate colleagues include Anne Bingaman’s husband, Jeff Bingaman, a New Mexico Democrat.
In a July 13 letter to the FTC (with a copy to Bingaman, who had okayed the letter in advance) Metzenbaum suggested that the commissioners refer the Microsoft investigation to the Justice Department, should they deadlock again in their second vote on July 21.
On July 15 Bingaman endorsed Metzenbaum’s logic in an interview with me for Legal Times: "Speaking totally in the abstract, I think it would make sense to distinguish between cases in which the FTC has affirmatively declined to act based on a decision that there is no actionable conduct…and a 2-to-2 split in which…you have a decision by default in effect."
To FTC staffers anxious to see somebody take on Microsoft, recalls one of them, Bingaman’s quote read like a signal: "Send it to me, and I’ll move." The FTC didn’t quite send it to her; she moved anyway.
The commissioners deadlocked 2 to 2 again on July 21, as expected, amid some of the most vitriolic exchanges among commissioners in recent memory. But Janet Steiger and Dennis Yao blocked an Owen motion to close the investigation, provoking angry tirades by Owen at Steiger-and making it easier for Bingaman to move in.
Metzenbaum issued a press release the same day, calling for Bingaman to pick up the investigation; the senior Republican on the Judiciary Committee, Orrin Hatch of Utah-home of both Novell and WordPerfect-added a bipartisan note by offering a similar suggestion in a letter to the FTC.
A week later, Bingaman’s antitrust division stuck its toe into the investigation by filing a formal request with the FTC Bureau of Competition for access to its mountains of documents about Microsoft. Bingaman made courtesy calls to the four commissioners to explain why she wanted a peek. Deborah Owen greeted her with an earful of angrily pungent prose, according to sources claiming secondhand knowledge of the conversation.
Bingaman won’t discuss the call, except to say, "I like Deborah. She’s an intelligent and committed person. We talk to each other straight from the shoulder… I think it’s good when people put their cards on the table."
Says Owen: "I expressed my views fully and forcefully." Owen also asserted in confidential memos to her colleagues that Bingaman’s "efforts to second-guess the commission’s decision should be resisted at all costs, as a matter of institutional integrity," and were tantamount to "political interference in the independent functioning of this agency." Both memos (like many an FTC document) ended up being leaked to the biweekly newsletter FT: WATCH.
On July 29, with only Owen opposed, the FTC opened its files and Bingaman’s troops started combing through them. They found plenty to pique their interest. After three weeks, on August 20, the antitrust division jumped into the Microsoft matter with both feet by formally seeking (and getting) FTC clearance to open its own investigation. "If you’re Anne Bingaman, and you’re as politically sensitive as she is, you don’t even take that step unless you intend to follow through," says one lawyer close to the case. Hours later, the FTC finally voted to close its probe.
Bingaman’s senior staff met with Microsoft and its major competitors to discuss the issues in late September, holding separate sessions of as long as four hours with lawyers for Novell, Lotus, WordPerfect, and Taligent, among others. Most (and perhaps all) of the anti-Microsoft lawyers came away fairly confident that Justice would bring a case-maybe a far broader case than the proposed complaints on which the FTC had deadlocked.
"I would be astonished [if Bingaman dropped the matter]," says one anti-Microsoft lawyer. "They spent a lot of time talking to us about relief… I think Microsoft’s better shot at a slap on the wrist was to take a mild consent decree at the FTC, because they could have gotten a better deal."
Bingaman is not altogether a free agent, however. She is almost certain to get input from other parts of the administration, such as Vice-President Al Gore’s technology gurus, the Commerce Department, and trade officials; some are likely to be receptive to Microsoft’s arguments that it should be left alone for the good of the country -and of the image the Clinton administration seeks as champion of competitiveness and high-tech entrepreneurship.
Microsoft’s major competitors in the world market- all of which are also U.S. companies and large exporters-are preparing to stress that they, not Japan Inc., will be the immediate beneficiaries of any restraints on Microsoft, and that the U.S. industry will continue to beat the world in the long run only if Microsoft is prevented from destroying the profits, the competitive vigor, and the research budgets of its U.S. rivals. "What’s going on here," says Gary Reback of Palo Alto’s Wilson, Sonsini, Goodrich & Rosati, an architect of Novell’s legal strategy, "will determine the future not only of the software industry but of antitrust enforcement for this administration."
That’s not to say that Bingaman would lightly launch the litigation equivalent of a land war against Microsoft. Microsoft has a lot more money to spend on lawyers than does the antitrust division. "The greatest impediment is resources," says Reback. "Anne Bingaman has got to make a decision about how much of her budget and her staff she wants to put on this thing."
Rewriting The Rules
Even for those who believe, as I do, that Microsoft has probably committed acts of monopolistic predation, the conclusion that the Justice Department should sue is hardly foregone.
It takes more to justify wheeling out the heavy machinery of a monopolization suit than just a few slips across the indistinct line that separates hard competition from illegal predation-a line that only the most successful firms are barred from crossing. And most industry analysts say that whatever predation Microsoft has committed has been only a secondary factor in its success.
Almost any company aggressively entrepreneurial enough to gain monopoly power legally, as Microsoft did, could plausibly be accused of some acts of monopolistic predation simply for continuing to compete as hard as before.
The visionary Bill Gates who got his start in 1975 by working frantically to produce an excellent Basic program that he had already promoted as "available" is inseparable from the bare-knuckled Bill Gates who froze sales of DR DOS 5.0 in 1990 by preannouncing delivery of MS-DOS 5.0 on a timetable that Microsoft didn’t even come close to meeting.
Besides, any anticompetitive excesses by Gates arguably won’t do much harm in the long run because nature abhors a monopoly, and Microsoft’s may give way to market forces and technological change-as did IBM’s-before a big monopolization suit could be concluded. Lots of big and little companies are plotting jointly and severally to turn the tables on Microsoft. Some are hoping to bring in new technologies that could make existing software obsolete. Others are ganging up into collaborative groups to challenge Microsoft’s dominance.
Nor is there strong evidence that Microsoft’s monopoly has done much damage to consumer welfare so far. Most consumers like Microsoft’s software and don’t mind its prices, which amount to a small percentage of a new computer system’s cost. Microsoft keeps the prices low enough to generate huge volumes, albeit high enough to generate huge profit margins. Its size and vertical integration enable it to offer a suite of complementary products that work seamlessly together. Monopoly has its charms-one-stop shopping.
Indeed, most of the loudest complaining comes not from Microsoft’s consumers but from its competitors-some of which have been plausibly accused of sins similar to those Microsoft is said to have committed. Since antitrust law is supposed to be about protecting competition, not competitors-let alone quasi-monopolists like Novell-those competitors’ obviously self-interested complaints should be taken with a grain of salt.
Above all, even some who deplore Microsoft’s abuse of its power fear-with some reason-that a big antitrust suit could make things worse by smothering the industry leader with the dead hand of regulation.
Here’s a computer industry expert with no stake in the investigation who has great respect for Microsoft’s brains but an especially dim view of its character: "Microsoft is an extremely powerful and dangerous, unethical, immoral, predatory, dishonest company. . . . Those guys have enormous power in the industry and use it brutally…And in negotiating, they just lie all the time – all the time."
So Justice should sue, right? Wrong. "The cure might be worse than the disease," this expert says, likening litigation to a lottery and the FTC and Justice Department to "the Marx Brothers and the Three Stooges."
Overhanging the Microsoft investigation is the specter of U.S. v. IBM-the litigation quagmire that wasted untold resources, bogged down the antitrust division, and sapped IBM’s competitive vitality from 1969 until 1982, when Reagan-appointed assistant attorney general William Baxter said the case was "without merit" and put it out of its misery. "Nobody wants to relive that," says a government lawyer close to the case.
Why Bingaman Should Sue
These points might be persuasive if an IBM-style mega-suit seeking to break up Microsoft were the only alternative to leaving it alone to practice predation in peace. But the alternatives are not so stark as that.
The same qualities that make the Bill Gateses of the world such powerful generators of economic growth also make them dangerous once they have achieved monopoly power. And so the case law on monopolization clings to two principles that coexist in some tension. The first is that antitrust law should not penalize success by making it a violation to win or keep monopoly power through skill, foresight, and industry. The second is that monopoly power is so dangerous and susceptible of abuse that at some point it must be checked.
The law has devised no better resolution of these tensions than to subject monopolists to some rules that don’t apply to everyone else-rules against practices that make economic sense only on the assumption that they will drive competitors or potential competitors out of the market.
It’s no violation of antitrust law, for example, for a single firm to hurt a competitor’s sales by using false preannouncements, or exclusionary contracts, or misleading disparagement, or deliberate incompatibility-unless that firm has (or at least is on its way to having) monopoly power. To a Bill Gates, that may seem like changing the rules in the middle of the game just because he’s so far ahead. But it’s the law.
Microsoft’s increasing dominance of the industry is more than a little bit worrisome. Its combination of monopoly power and predatory tactics are a far more serious long-run threat than is suspected by satisfied consumers of Microsoft’s high-quality software products.
The larger and more secure from competition its domain becomes, the more a monopolist deadens innovation. If that old rule of thumb seems not to fit a young, vigorous, innovative company like Microsoft, eventually it will-especially if Gates succeeds in destroying his major U.S. competitors. Microsoft had already let innovation stagnate in its wholly owned MS-DOS monopoly market before it had to meet the new competition offered by DR DOS 5.0 in 1990.
In addition, it is far from clear that Microsoft’s monopoly power will be eroded by market forces or technological change anytime soon. Some of Microsoft’s major competitors are scrambling so desperately to stay profitable (and in Borland’s case, to stay in business) that their research and development budgets have been bled white. Indeed, it is at least as likely that Microsoft will move to dominate new realms like software for interactive television. The MS-DOS monopoly has proven durable in spite of the product’s limitations and years of stagnation. It takes a lot more than new, improved technology to displace industry-standard operating software, because of the increasing complexity of software, the cost of writing it, and the enormous cost of replacing or switching from the huge installed base of applications written specifically for MS-DOS and Windows.
Microsoft’s dominance has reached the point at which it is not only a few competitors, but competition itself- at least among U.S. companies-that may be wiped out in the biggest desktop software markets.
"Microsoft is a wonderful, upbeat, aggressive, innovative company," InfoWorld publisher Bob Metcalfe said this summer to National Public Radio, "[but] if we keep going the way we are, in ten years there won’t be anybody left except Microsoft."
Nor are competitors and FTC antitrust lawyers the only people who complain about Microsoft’s abuse of its power. So do programming experts like Andrew Schulman. So do a gaggle of smaller companies that say Microsoft enticed them to reveal their secrets by dangling the prospect of collaborative projects, and then used what it had learned to compete against them.
And so do many PC makers. Most voice their complaints privately, for fear of Microsoft retaliation. "Every-one in this industry has to do business with Microsoft, and everyone to some extent exists at the mercy of Microsoft," notes an anti-Microsoft lawyer. Still, some PC makers told FTC investigators that Microsoft uses its monopoly power to keep jacking up the price of MS-DOS, to negotiate extremely one-sided contracts, and to tie the PC makers so tightly to Microsoft as to limit their options to buy software from others.
HOW TO LEVEL THE PLAYING FIELD
It would probably be unwise, however, for the Justice Department to bring an IBM-style mega-suit seeking to break up Microsoft.
It’s not that a breakup would necessarily be a bad idea. Indeed, splitting Microsoft’s applications division from its monopoly operating systems might be the most efficient way to create a more level playing field for competing applications companies. But that’s a Utopian solution because an antitrust suit could not achieve it in a timely fashion or at reasonable cost.
Microsoft would understandably fight any such mega-suit with all of its enormous resources. The case would probably bog down into wasteful trench warfare, with an uncertain resolution too far into the future to inspire confidence that the final decree would make any sense when it came down.
Bingaman "can’t afford to litigate an IBM-style case," says an anti-Microsoft lawyer-not if she wants to get to the rest of her ambitious agenda with a complement of staff lawyers about half the size it was in 1980. An IBM-style case against an entrepreneurial company like Microsoft might also be a political negative for the Clinton administration, with the litigation grinding on through the 1996 election and no light at the end of the tunnel.
For these reasons, a suit seeking a breakup of Microsoft isn’t in the cards. The real question for the Justice Department is the wisdom of bringing a suit to bar specified forms of monopolistic predation. The answer hinges on whether the Justice Department can craft a proposed remedy that would effectively curb the most damaging abuses of Microsoft’s monopoly power without hobbling its efficiency and competitive vigor or saddling it with an intrusive regulatory regime to oversee compliance.
I think it can. The remedies the FTC staff proposed would be a good start. Since the core problem is Microsoft’s abuse of its MS-DOS and Windows monopoly power, the most potent forms of that abuse should be barred. And since litigating the case would be so wasteful, the Justice Department’s proposed remedies should be sufficiently surgical and unintrusive to tempt a reasonable defendant in Microsoft’s position to make a deal.
Specifically, the Justice Department should seek to:
1. Ban Microsoft from selling its software under CPU licenses or other contracts that serve little economic purpose other than excluding competition;
2. Prohibit Microsoft from tying sales of its operating software to those of its other products, except where it can demonstrate efficiencies of integration;
3. Prohibit Microsoft from deliberately making its operating systems incompatible with any other company’s products, or seeking to hurt competitors’ sales by alleging compatibility problems without reasonable grounds for believing that they exist;
4. Require Microsoft to document publicly all internal communications about programming interfaces in its monopoly operating systems that are used by its own applications and other complementary software, within a reasonable time. (In the case of written communications, this time should be no later than one day after they reach developers of Microsoft’s complementary software; for nowritten communications, say, three months before the f complementary software is allowed to be marketed.)
Why Gates Should Settle
But why should Gates swallow such a settlement-especially when he could have had a better deal at the FTC? He’s used to fighting and winning, not compromising. And Microsoft has been talking a very tough game so far, both publicly and in its lawyers’ discussions with antitrust investigators. Gates has said that nothing uncovered in more than three years of investigation caused him "the slightest concern."
He should be concerned. His company has some serious legal vulnerabilities. The Justice Department could make a very strong case that Microsoft has engaged in monopolistic predation, at least in operating software. While Microsoft would no doubt mount a strong defense, could easily absorb the staggering legal fees, and might well end up winning-especially in the Reaganized federal appeals courts-it would also risk losing, big-time.
If Gates passed up a reasonable consent decree deal, the Justice Department would presumably seek a far stronger remedy in court. Gates could look forward to many long days in depositions. As market developments and discovery progressed-"God knows what a guy like Gates has in his files," notes an anti-Microsoft lawyer-such a lawsuit could take on a momentum of its own, and the government might broaden its attack.
A final judgment against Microsoft could expose it not only to a possibly onerous decree, but also to billions of dollars of treble damages liability in piggyback suits by competitors like Novell, which has suggested (rather hyperbolically) that Microsoft has caused it alone $3 billion in damages. While some competitors will no doubt sue in any event, they would pose a far greater threat to Microsoft if they could hitch a ride on a Justice Department suit-at least one leading to a judicial finding of liability.
Rather than risk such consequences, Gates should be able to cut a deal that would not make an immediate dent in Microsoft’s market leadership, would not hobble its efficiency, and would not hurt its ability to engage in hard, but fair, competition.
In the final analysis, it’s a game of chicken. Bingaman has gone out on a limb politically; she’ll lose face unless she comes back with something more than a transparent slap on the wrist. But she can ill afford a litigation quagmire. And Gates would be foolish to take a litigation risk.
Is Gates-the competitive warrior-the world-champion fighter-the arrogant author of screaming tirades- temperamentally capable of settling? Perhaps not. But some FTC staffers think he is. They would have bet money that Gates would have settled in February had the FTC voted out a case.
"Microsoft is very good at playing chicken," one staffer says. "But I think if the FTC had acted, they would have made a deal." Maybe they will yet.