U.S. v. Microsoft: Choosing The Right Remedy

by Eric M. Bennett

> Home

> Keynote

> Manifesto

> Lead Article

> Site News

> Bookstore

> BMS Forum

> Bumper Stickers

> NT Files

> Dirty Tricks

> Feature Articles

> Departments

> Monopolist's Cookbook

> Site Credits

M

ost observers currently believe that the Department of Justice has convinced U.S. District Judge Thomas Penfield Jackson that Microsoft has violated federal antitrust laws. More important, there is a growing consensus that such a ruling by Jackson will withstand appeal. With Microsoft's defeat seemingly assured, it is now critical to determine how Microsoft should disciplined.

There can be two component parts of discipline in an antitrust case: punishment for violations of the law, and so-called "remedies" to prevent future violations from taking place. Punishments are not too difficult to set, and for a company accused of civil violations of antitrust laws, the Sherman Antitrust Act permits the judge to levy significant fines. Remedies, however, are often hotly debated, and this article is intended to serve as an introduction to the remedy debate in U.S. v. Microsoft.

It is helpful to begin by borrowing an analogy from medicine. For a doctor, it is not enough to merely recognize that a patient is ill with a particular set of symptoms. To devise an effective remedy, it is necessary to understand the underlying cause of those symptoms. Without such an understanding, a doctor may propose a cure which is ineffective or even harmful. The same is true for a judge choosing remedies in an antitrust proceeding, and so the first issue is to identify the cause of Microsoft's antitrust violations.

There are at least three factors that were involved in Microsoft's decision to violate the antitrust laws. The first is the possession of monopoly power, for without that there would have been no leverage and no violations. The second is the will of Microsoft executives to use that power. The final factor is the perception that even if Microsoft were caught violating the law, it would be able to escape significant punishment. One might argue that if any of these three factors could be eliminated, Microsoft would comply with the law.

A Cure Worse than the Disease?

Although eliminating any of the three factors could bring Microsoft back into compliance, those three factors are not equally easy to deal with. A recent proposal that attempts to deal with the third factor can probably be classified as the worst proposal made to date. It comes from industry analyst Rob Enderle in an article by Inter@activeWeek Online:

Rob Enderle, an analyst at the Giga Information Group, instead supports the creation of a watchdog team, much like the United Nations weapons inspectors that monitor Saddam Hussein's activities.

As an IBM employee in the late 1980s, Enderle was part of an internal audit team Big Blue put in place as part of the DOJ's agreement to drop its antitrust action. The team's role was to ensure IBM executives and divisions were not engaged in business practices that could land the company in court. The monitors reported to an independent entity.

"The core of the problem is Microsoft's continual misuse of its power," Enderle said. A team of monitors, he said, "allows the company to remain intact, but puts in place intelligent oversight to make sure the practices the U.S. government does not want to happen, don't. That's what the consent decree was missing: an oversight committee with teeth."

But how might that affect the naturally combative corporate culture that has helped Microsoft vault to the top of the software industry?

Mike Maples, a former senior executive with both Microsoft and IBM, thinks it would be poison.

Indeed, he said, any monitoring or oversight of Microsoft activities would likely grind down and even destroy Microsoft over time. The former Microsoft and IBM senior executive, respected for clear-headed thinking and candor, speaks from experience.

He argues that IBM was literally handcuffed during its 13-year battle with the DOJ ending in 1981. That trial, he contends, ruined Big Blue.

"Every presentation had to be reviewed by lawyers. IBMers could not review competitive products and if they did, they had to be taken off the comparable product."

"There were certain words we could not use. You don't have to dull the knife very much so that it does not cut very well."

It is probably difficult for Microsoft critics to find themselves agreeing with somebody like Mike Maples, but in this case Mr. Maples is correct. Rob Enderle needs to consider just how ineffective "monitoring" of Saddam Hussein has been. The idea of "monitoring" leaves the perpetrator's power intact. Empirically, that doesn't seem to work very well. If the only thing the power can be used for is to violate agreements or laws, failing to destroy that power is asking for trouble. The best way to prevent Hussein's repeated misuse of his power is to get rid of that power, not to "monitor" it. And the best way to prevent Microsoft's misuse of monopoly power is to get rid of the monopoly, because under U.S. law there is not really anything you can do with your monopoly power that is legal. Furthermore, proponents of a monitoring solution are failing to do anything to address two root causes of the problem: the power to violate the law still exists, and the will to exploit that power still exists. If either of those two things did not still exist, there would be no need for the monitoring.

The monitoring solution thus boils down to a game between the monitoring committee and the people they are monitoring. The monitors try to prevent violations, and their wily opponents try to use their power in new ways that the monitoring committee is not allowed to challenge. This consumes resources and creates ill will. This policy has obviously been unsuccessful in Iraq, and history suggests it will fail with Microsoft as well.

In 1995, the Justice Department and Microsoft signed a consent decree, which included a promise from Microsoft not to engage in certain activities. The Justice Department was the monitoring committee. As soon as the agreement was signed, a battle began between the monitors and monitored. The DOJ claimed the agreement was a significant victory that would lead to increased consumer choice. Microsoft said the agreement would not affect it in any way. So far, Microsoft has violated the agreement at least twice.

The first violation, the packaging of Internet Explorer with the operating system, led to the current lawsuit. The second violation breaks with the spirit of the agreement, although perhaps not the letter of the agreement. Microsoft used to force computer vendors to buy a copy of Windows for every computer they sold, even if the vendor was loading a competing operating system (rather than Windows) on a small number of its computers. The consent decree banned this specific activity, but Microsoft has found new ways to achieve the same effect. Today, Microsoft offers "rebates" to vendors if the vendors agree not to promote competing operating systems. Microsoft has also been accused of requiring vendors who want to purchase Windows for use with a certain class of computers (laptops, for example) to purchase a copy of Windows for every computer they sell in that class. Technically, the agreement doesn't cover this, even though it achieves the same anticompetitive effect as the original practice.

So experience shows that Microsoft will do its best to exploit loopholes in its agreements with a monitoring group. The only way to fix this is to give the monitoring group real teeth so that it can properly regulate Microsoft's power. Unfortunately, it is very difficult to do this correctly. It's easy to end up with a document that gives the monitoring group too little power or too much power. To return to the earlier medical analogy, the doctor needs an effective cure for a patient's ailment, but if the doctor's treatment is too aggressive, the patient may die from the treatment instead of the disease.

Gary Reback, an antitrust lawyer who has worked for several Microsoft competitors and who has repeatedly spoken out against the company in public, agrees that simply imposing restrictions on Microsoft's behavior and monitoring for compliance will not work. And given that Microsoft supporters like Mike Maples also oppose this remedy, it's amazing that it would receive any serious consideration at all.

Fix the Root of the Problem

The monitoring proposal leaves the patient weakly infected with the disease, but lets the doctor step in if the patient's symptoms return. For diseases where there is no known cure, this is the only real option. But if a cure does exist, the doctor's best choice is to use that cure. In the case of Microsoft's antitrust violations, the cure is removing the first of the three factors that led to the violations. That factor is Microsoft's monopoly power. Get rid of the monopoly and there will no longer be any need to spend time watching for violations.

Breaking Microsoft into multiple companies can remove the monopoly power. Many people find the idea of breaking up Microsoft to be a radical and frightening application of government power. But what are the alternatives? The second factor cannot easily be dealt with; a court order cannot change Microsoft's corporate culture. A monitoring committee that attacks the third factor leads to the problems of over/underregulation, repeated lawsuits, and perpetual government involvement in the software industry. A monitoring solution seems less offensive to many at first, but it leads to more government intervention, not less; if government intervention is what people truly fear, then a break-up seems the better choice. There are better ways of dealing with the third factor, but even if Microsoft comes to fear punishment, Microsoft will have to have some way of knowing what it can add to the operating system without being challenged by the Justice Department. Under such a scenario it is likely that the government could have a long-term influence on Microsoft's product design, a consequence that nobody finds desirable. And if Microsoft ever sense that the enforcement bodies are becoming lax, it is likely to strike again.

Breaking up Microsoft avoids long-term meddling by the government, which is probably one reason it's the preferred remedy of conservative heavyweight Robert Bork. But the breakup has to be done correctly to be effective. Some analysts have proposed splitting Microsoft into divisions based on product lines; for example, an operating system company, a development tools company, and an office suite company. But then the government has to decide which products belong in the operating system group. And since this solution leaves the Windows monopoly intact, there is still the problem of the DOJ having to keep watch over the OS company to make sure it does not bolt irrelevant products onto the OS. Furthermore, some would argue that Microsoft has a monopoly in office suites as well as operating systems; a breakup along product lines could thus result in two monopolists rather than one. Since this so-called solution leaves the monopoly power intact, preventing further abuse of the monopoly power would necessitate behavioral restraints on the new companies. And that takes us right back to the problem of monitoring again.

But if the court splits Microsoft into several companies that all have rights to Microsoft's current product line, most of these problems go away. The "Baby Bills" could all negotiate using Gates's current hard-ball tactics if they like; those tactics wouldn't be illegal since none of the companies would have a monopoly. On the flip side, since there would be no monopoly, it wouldn't be possible for them to impose the current draconian and exclusionary contract terms on their customers. The Baby Bills could integrate anything into the OS, and if PC makers don't like it, the PC makers will have the choice of going to different OS supplier. This solution places all the "regulation" of the Baby Bills in the hands of the free market, which would do a much more effective job of regulating than the government. This solution puts no long-term restrictions on what the Baby Bills can do, so the government will not have to butt in again. Most people agree that preventing repeated government intervention is highly desirable, so this solution deserves to be more popular than it is. To give credit where credit is due, Oracle's Larry Ellison was the first person to propose this plan publicly.

The only potential problem would be collusion among the split-up companies (price fixing, for example). But any collusion would be more straightforward to deal with than the murky issues ("What belongs in the OS?" "Is X or Y or Z an unfair contractual term?") that would almost certainly arise with other solutions that leave the monopoly intact. There are already laws on the books to cover anticompetitive collusion. They are time-tested and not subject to the problem of loopholes that would likely plague any consent decrees or court orders written today.

A Less Intrusive Solution?

Despite the many advantages of a break-up, there are still those who would prefer solutions that seem less intrusive. One proposal would be to force Microsoft to set specific and public pricing for Windows--pricing based solely on the volumes of Windows sold. Microsoft would also have to deal equally with all customers. For example, if Compaq and Gateway both wanted to purchase 1,000,000 copies of Windows 98, it would be guaranteed that they both pay the same price, even if Gateway is shipping computers with IE and Compaq is using Navigator. This proposal does not directly address the issue of tying. Microsoft would be able to force OEMs to take copies of Windows that had IE in them, but the OEMs would have to be free to modify the default installation in any way they see fit. Consumers who are concerned about OEM-modified installations of Windows not working correctly would certainly be able to purchase systems from vendors who have chosen to retain the default Microsoft setup.

This is not an ideal solution, but it would be reasonable it if it weren't for the fact that it would be moderately easy to evade. To continue with the above example, Microsoft could give Gateway a huge discount on Office because Gateway didn't remove IE from Windows, while Compaq did. This strategy would not work if Microsoft didn't have a monopoly on Office, since Compaq would balk at the higher Office price and switch to products from Corel or Lotus. The existence of a monopoly in office suites is more debatable than the existence of a monopoly in OSes, but the court would be foolish to implement this solution without first determining to its satisfaction that there is not an office suite monopoly (or, if there is such a monopoly, forcing Microsoft to price its suites based solely on volume as well).

There is cause for concern here, because the DOJ--unlike the states--has shown no apparent interest in pursuing the question of office suites. And if the DOJ neutralizes one of Microsoft's monopolies but leaves another intact, Microsoft will still be able to do all the things the DOJ is trying to prevent it from doing. Microsoft will simply use the second monopoly instead of the first.

In the current climate, where office suites have received no scrutiny, breaking Microsoft into several equal parts is the wisest practical remedy.

This article is copyrighted 1999 by Eric M. Bennett, and it may be reproduced free of charge if the copyright notice is preserved.