Not With a Bang, but a Whimper
Microsoft did appeal—within the week, which was probably good, because no one had fully addressed an unspoken but critical question: who would determine how to divide Microsoft’s greatest asset—its people—between Winco and Appco?
If the breakup order had gone through, something along the following lines would have unfolded: Jackson would have given Microsoft about six months to craft a divestiture plan.
Six months later Microsoft would have presented its plan, and the DoJ would have opposed it. Jackson would have agreed and rejected it, and Microsoft would have appealed. A few months later, the appellate court would have refused to interfere and shipped the case back to Jackson, who would have given Microsoft detailed instructions for its next divestiture plan, ordering it to give personnel and organizational information to DoJ so that DoJ could devise an alternate plan. Then they would both have come back to court in six months to present their plans to Jackson who would... what? Meanwhile, Microsoft would have been up to its old tricks, continuing its leveraging tactics to push its dominance even further across the Internet. Bottom line: Breaking up is hard to do.The 1984 breakup of AT&T worked because, in the final analysis, AT&T agreed to work with the government to develop a divestiture plan, and because Judge Harold Greene agreed to become “the AT&T judge.” And AT&T’s business lent itself to some obvious geographical divisions. Earlier antitrust breakups, like Standard Oil, more-or-less severed the companies along divisional seams left over from earlier mergers. But Microsoft had grown as a single organic company. No one had ever broken up such a company before without the full cooperation of both management and the board. No one knew how to do it. So all in all, it may not be horrible that we never got there.
But that doesn’t mean that we shouldn’t have tried. Even more importantly, it doesn’t mean that a behavioral remedy would have been better. All of the problems that motivated the drive towards radical surgery were still in place. Without a structural remedy, Microsoft would retain most of its weapons, its incentives would be largely unaltered, and the market would remain its private playground. Klein’s DoJ, its supporters, and Jackson, had done a great job appreciating the shortcomings of behavioral remedies. They were less thorough thinking through the practical implications of structural remedies.
Microsoft’s appeal returned the case, once again, to the D.C. Circuit Court of Appeals. This time they agreed to hear the case en banc, which means that all of the Court’s judges (except for a few who recused themselves) participated in the proceeding. It took them a little more than a year, but their unanimous conclusion gave us all something to read over 2001’s Fourth of July holiday. They accepted Jackson’s findings of fact. They agreed with him about some points of law, disagreed with him about others. In yet another legal area, the application of tying law to the specific case of platform software, they decided to change the rules governing the legal analysis. Then they concluded that since they had just made up a new rule, Jackson couldn’t possibly have followed it correctly. They thus ordered a new trial on the government’s tying claim (just one of the government’s several leveraging theories). Finally, and on this point they were trivially correct, they noted that since Microsoft was now guilty of fewer violations than Jackson had thought it to be when issuing his breakup order, the court also needed to be revisit the remedy.
When the dust settled, a few things were clear. Microsoft was still an adjudicated monopolist; it had violated the laws pertaining to the maintenance of a monopoly. The breakup order was off the table, at least for a while. We were going to have another trial.
And Judge Jackson was off the case. The appellate court had been particularly harsh with Jackson, accusing him of appearing to violate the canons of judicial ethics. Among his many faux pas, he had spoken to the press in the midst of the trial. Microsoft had outlasted its second judge. The court held yet another lottery to pick a new judge. Judge Colleen Kollar-Kotelly drew the short straw.She began her tour of duty armed with the Court of Appeals’s sage, if incomprehensible, advice:
As a general matter, a district court is afforded broad discretion to enter that relief it calculates will best remedy the conduct it has found to be unlawful. This is no less true in antitrust cases. And divestiture is a common form of relief in successful antitrust prosecutions: it is indeed “the most important of antitrust remedies.”
On remand, the District Court must reconsider whether the use of the structural remedy of divestiture is appropriate with respect to Microsoft....
In devising an appropriate remedy, the District Court also should consider whether plaintiffs have established a sufficient causal connection between Microsoft’s anticompetitive conduct and its dominant position in the OS market.... Absent such causation, the antitrust defendant’s unlawful behavior should be remedied by “an injunction against continuation of that conduct.”...
While we do not undertake to dictate to the District Court the precise form that relief should take on remand, we note again that it should be tailored to fit the wrong creating the occasion for the remedy.36
In other words, the punishment (and we’re not going to tell you what it should be) should both prevent Microsoft from repeating its past behavior and be fair given the specifics of Microsoft’s actual liability.
To make Kollar-Kotelly’s assignment even tougher, this time Microsoft had done more than simply outlast a judge. Microsoft had outlasted an entire administration. The market cops who moved to Redmond during the Bush pere administration had begun an investigation that lingered through the entire Clinton administration and fell into the lap of the Bush fils administration, who most definitely did not want it.
Nevertheless, Attorney General John Ashcroft appointed a strong team to take over. Charles James, the man who inherited Joel Klein’s mantle of Assistant Attorney General for Antitrust, was a respected antitrust lawyer. He replaced David Boies—who had used the break in the Microsoft trial to argue on behalf of the Gore campaign both in Florida and in front of the Supreme Court—with Phil Beck, an equally talented litigator who had argued on behalf of the Bush campaign in Florida. Seems only fair. But on September 6, 2001, before Beck could begin litigation, the new DoJ announced that it would neither seek a structural remedy nor pursue the tying claim.Judge Kollar-Kotelly had an almost impossible task. She was charged with crafting a remedy that was fair, proportional, and related to the maintenance of monopoly violations for which Microsoft had actually been ruled liable, but that also prevented Microsoft from repeating behavior that combined those violations with various leveraging actions that were no longer on the table. She was in a no-win situation with a defendant who was somehow poison—the Court of Appeals had already slapped around two of her senior colleagues, Judges Sporkin and Jackson, for trying to curb Microsoft’s monopolistic excesses.
The DoJ announcement on September 6, 2001, was also the last interesting movement on this case. More will occur again in the future, but likely in a different political climate, within a different general framework, and with a different cast of characters. But this next round may be a few years off. In September 2001, Microsoft and the DoJ still had a clock to run out, even though we all knew that the game was over. Sure, Judge Kollar-Kotelly wouldn’t issue her final ruling for almost fourteen months. And while newspaper stories continued and events kept unfolding, none of that mattered. The game was over; all that remained was a long, boring, and anticlimactic final act. Microsoft clearly was going to get away with a slap on the wrist.
The only question was how hard a slap.But even an anticlimax deserves to be told. The DoJ, Microsoft, and half of the states quickly agreed upon a behavioral consent order. When it came time for the Tunney Act hearing, they all asked Judge Kollar- Kotelly to sign the order. Nine other states sought tougher behavioral restrictions; they wanted to save the market from Windows 2000 and Windows XP, products that Microsoft hadn’t even launched until after the trial. This rare, bitterly contested Tunney Act hearing proved, once again, that nothing involving Microsoft ever unfolds as expected.
Judge Kollar-Kotelly held a long hearing, took months to deliberate, and on November 1, 2002, issued a lengthy, detailed opinion—basically rubber stamping the consent order and telling the remaining states to go away. Most of them did, though Massachusetts and West Virginia fought on. Anticlimactic. This case had finished back on September 6, 2001; the last fourteen months were just filler.
Once the Bush DoJ had dropped the possibility of corporate capital punishment, all that was left was a behavioral remedy. But behavioral remedies couldn’t possibly be effective. They couldn’t truly punish Microsoft, they couldn’t prevent it from repeating its behavior, and they most certainly couldn’t restore competition to any of the markets that Microsoft had destroyed. And as to their deterrent effect, suppose that the young Bill Gates back in the early 1980s had a dream showing how Microsoft’s future would unfold—its emergence as the platform software monopolist followed by its antitrust conviction and the terms of the settlement agreement. When he awoke, would he be likely to let things unfold as they did, or reform his ways and change the future? Answer: He’d be out of his mind to change a damn thing.
And so, though Judge Kollar-Kotelly’s order is unlikely to have much of an impact on the information sector, it probably represented the most prudent course of action she could have taken given the many constraints she faced.
She accepted the deal that Microsoft had cut with the DoJ subject to only a very few, very minor changes and wrote a lengthy, detailed explanation describing the relationship between the violations and the penalties. She thus met one of the Court of Appeals’s instructions: the remedy was proportional to the specific narrow violations that the government had proved. Since it’s not clear that she could have hit both instructions simultaneously, doing a good job on one front can’t really be all that bad.Or can it? After all, not only is her order shorn of deterrent effect, but the opposite is true. In the future, anyone who can emulate Microsoft’s behavior, will. The Bush DoJ showed Internet investors that while they may have been mistaken about the Internet barrier to entry, they were dead-on right in their quest for the next Microsoft. If you can find it, invest. Heavily. Because the next Microsoft will leverage its monopoly successfully, it will transfer rents from consumers to shareholders, and it will get away with a slap on the wrist. Therein lies the ultimate message of the Microsoft trial to the public. Thus educated, we must wonder about the future of the information sector, soon to be a wholly owned subsidiary of Microsoft. Yes, this is the way that the trial ends. Not with a bang, but a whimper.