Microsoft Vs. the Feds: Drawing the Battle Lines

Knowledge at Wharton: What do you think of the government’s proposal to break up Microsoft into two businesses?

Schmittlein: I expect that it’s feasible, although the fact that it’s feasible now doesn’t mean it might be as feasible or as straightforward three or five years from now, or whenever such a plan could be implemented. But I think, in very general terms, the notion that you would split certain kinds of applications software, like Office, from the operating system, Windows, makes the most sense if you were going to do a division. It may not be as feasible years from now because of the evolution of software and the integration of software, including the potential integration of certain application software with the operating system itself. This is certainly a business where a great deal can change over a several-year period. One thing that can change is Microsoft’s dominance, or perceived dominance. Another thing that can change is the company’s level of market share. The third thing that can change is the practical feasibility of splitting up the company.

My concern has always been that this litigation process has been a significant distraction for the company. That’s what I think has been the biggest negative for Microsoft out of this whole thing. In this respect, the company has already been punished. The time and attention the company has taken to respond to the litigation has taken away from other potential initiatives. In addition, Microsoft has demonstrated self-restraint during the course of this litigation. These, I think, are very likely the greatest consequences that the government would realistically hope for in this case. And they are possibly, in themselves, consistent with the magnitude of the violations that the courts have determined.

Yao: The government’s plan has the advantage that, relative to some intermediate remedies, it greatly reduces the extent to which the government would have to remain involved. That’s a big plus. A potential disadvantage is a loss of efficiencies associated with having operating system and applications development in one company. I can’t say the government’s proposal is definitely the right solution. A lot depends upon how easy it would be to monitor and enforce intermediate remedies and how effective those remedies will be.

Faulhaber: I don’t think it’s a good idea. Breaking up AT&T was a good idea because in that case you could separate out the monopoly piece in ways that were helpful to solving the monopoly problem. With Microsoft, the split is between the operating system and applications, but that doesn’t really define what’s in the operating system. In the case of the phone company the monopoly was local access. Once you defined what was long distance and local, you could focus control and regulation on that monopoly. You could tell the Baby Bells to open up their networks to make sure long-distance companies could come in. It cost billions to do it, but they mostly got it right.

With Microsoft, the question is how do you really separate applications from the operating system? You could argue there’s an existing split within Microsoft, but that’s complicated for a couple of reasons. First, there’s both server and client software, for example. Also, the idea of saying Windows is a separate entity – well, that doesn’t tell you what features you can add to Windows without running afoul of the conditions of the breakup. What is the new Microsoft company going to ship with Windows? Should it include Microsoft Outlook or should Outlook be in Office 2000? Are you going to take the browser out of Windows? Put it back in? Or suppose the new Microsoft puts into Windows some preprocessor for Internet Explorer, which makes that browser work better than Netscape’s. It could be a nightmare.

The second thing I don’t like is that setting boundaries for what kind of market Microsoft can be in requires a lot of policing. This happened in the AT&T antitrust case where a federal judge ruled the telecom industry for over 10 years. Any time a new service was introduced or relations with connecting firms changed, you had to go to the judge. That was okay in the old phone business, which was slow moving. But in the technology business, it’s crazy. Do you want a judge deciding what features are going to go into the next version of Windows?

Knowledge at Wharton: Are you saying it’s okay to have a monopoly?

Faulhaber: I think it’s very clear that Microsoft has huge market power in the operating system market. In markets like this, though, I think it’s okay to have a monopoly because in the real world customers value standards. If one company owns the standard, such as Windows, then customers will value that standard and support a monopoly. Competition that resulted in competing standards usually doesn’t work in the customer’s favor. Often in such markets you will find that one monopoly lasts a while and then somebody else comes along. In the early ’80s, WordStar software was overthrown by WordPerfect and WordPerfect was later overthrown over by Word. So monopoly isn’t necessarily bad if setting a standard is valuable for customers and a better product can oust the current monopolist.

But what gets bad was that Microsoft was erecting huge barriers to entry for the next guy taking over. Netscape threatened the Windows monopoly, so Microsoft did everything they could to kill Netscape’s browser and the Java language, the other vehicle for the destruction of Windows. In and of itself, that’s okay; but they used their market power in the OS market to monopolize the browser market and kill nascent competition in the cradle. They behaved in a predatory way. I would tend to say, yes, Microsoft was nasty and probably guilty of antitrust violations, but breaking up the company would not be my solution to this problem.

Knowledge at Wharton: So, what’s the solution?

Faulhaber: My solution would be to make the source code for Windows public. That source code is one of the most closely guarded secrets on earth. I would require Microsoft to put that code in the public domain and let anybody use it, maybe with a small fee to Microsoft. And anybody could modify the code. So, if any company uses the code for its own product, they could call it, maybe, Dell Windows, but not Microsoft Windows. Microsoft would maintain some control over brand. To enforce this, I would force all distribution of Windows to go through an independent third party. Microsoft would deliver the Windows source code to the – let’s call it the Windows Distribution Company. That company would compile the code and ship all the disks to whatever companies wanted to buy Microsoft Windows. If a company, such as Dell, wanted their own version of Windows, with their own special add-ons, they get the source code from the Windows Distribution Co. and modify it, and they choose whether or not to guarantee compatibility.

I think this takes away the Windows monopoly. That was the source of the problem to begin with. Microsoft, of course, would fight my idea tooth and nail. Clearly, they would claim that this is their intellectual property and they deserve to exploit it. In general, this is true. But if the signal is, ‘We’re only going to let you make $100 billion from your invention and then you have to go public,’ I don’t think that will discourage too much future innovation.

Besides, there’s precedent for this. When the Justice Department and AT&T settled the first [1949] antitrust suit in 1956, a condition of settlement was that AT&T would have to give up its basic transistor patents. These key patents formed the basis of all chip advances since then, and making those patents public literally created the computer business.

Knowledge at Wharton: If Microsoft were divided, would we end up with two monopolies exercising the kind of dominance that led to the court decision against Microsoft in the first place?

Schmittlein: The answer depends on the time frame after a hypothetical breakup. In the period shortly after a break up, if it were to happen soon, you’d have two high-market-share entities that would, at least in those terms, continue to dominate their markets. They would, however, be more vulnerable to potential competition. I think the issue is not so much whether another organization could compete, but whether another organization would elect to compete.

Knowledge at Wharton: Would a breakup be hard on the economy? It’s been suggested, for example, that individuals and companies that rely on a common platform to carry out many functions could suffer confusion and incur large costs to make different operating systems compatible.

Faulhaber: That’s an old problem in the software business. That’s why we like to have standards like Windows. Having incompatible operating systems is a nightmare for any corporate information technology manager. But I think it is highly unlikely that the market will move in that direction. Compatibility and standards are far too important for customers to buy into myriad operating systems. More to the point, there are big changes on the horizon with the Internet, and many are predicting that Windows’ days are numbered as the dot-com guys figure out how to go around, under, over and through the Windows monopoly via the Internet. Bill Gates is right to say he’s running scared all the time, worrying about who’s going to take away his business.

Yao: Microsoft has argued that the split would cause some confusion in the marketplace because of the existence of multiple standards. Multiple standards pose the possibility of short-term inefficiencies. But competition over standards also stimulates innovation towards a better standard, so there’s a tradeoff. Consider a polar case to illustrate this tradeoff. Suppose you took the current operating system and froze it forever. You’d have a standard but you wouldn’t have any innovation. Now consider a situation in which two or three companies compete for a new operating system standard. There is likely to be some transition cost as the market selects the best one (assuming that we move to a single standard), but one is also likely to get a better standard in the end. That’s the essence of the tradeoff. Microsoft, of course, would argue they have incentives to improve their operating system. But the question is whether society would be better off with level-playing-field competition in operating systems versus an uneven playing field.

Schmittlein: It’s among the least of the things I would worry about. Microsoft’s competitors would have a great deal of incentive to make systems compatible, and the successor components of Microsoft would have an incentive to do that, too. The operating system people would have incentives to work with Microsoft’s application successor, and the application successor would have incentives to work with operating system providers. I don’t want to suggest that those collaborations are trivial, but it’s difficult to argue that they are insurmountable hurdles.

Knowledge at Wharton: What do you think about Microsoft’s counterproposals?

Schmittlein: It appears to me that the proposed remedies on the part of Microsoft come in two flavors. The first is essentially closing the barn door after the horses have left. The second proposes largely unenforceable norms for corporate behavior.

Faulhaber: The problem with Microsoft’s proposed restrictions is that they hinge on Microsoft obeying the letter of the agreement while figuring out a way to get around the wording of the restrictions and keep doing predatory stuff. Unfortunately, the Justice Department has been down this road before with Microsoft in the case a few years ago where the company was doing exclusionary contracts. The company signed a consent decree but then proceeded to figure out loopholes around the consent decree. The government decided it couldn’t trust Microsoft.

Yao: The effectiveness of the counterproposals depends on the implementation details. Some of the counterproposals would be problematic in terms of the ability of the government to ensure appropriate conduct. Also, the Microsoft plan appears to allow its internal application developers to have better access to relevant operating system information. That would mean that, with respect to application developments, Microsoft would have an advantage over application development by other companies.

Knowledge at Wharton: How would a breakup affect shareholders?

Schmittlein: The market’s reaction to these judicial proceedings — or should I say non-reaction — stems from two factors. One is the sense that nothing is going to happen any time soon. Second, if something were to happen soon, shareholders would presumably not be adversely affected. Microsoft has not seen its stock price drop that much, relatively speaking. The price has dropped, but the decline is not out of line with the general decline in the tech sector.

Yao: It could go either way. It’s difficult to say because we don’t know exactly what perceptions the market currently has about where Microsoft’s value lies. If the market is greatly valuing Microsoft’s ability to leverage its market power, then one might expect some decrease in the stock price. It’s conceivable as well that a breakup could lead to new companies that are more innovative than the old company, and this would lead to increases in the stock price. We saw from the AT&T breakup that the stock went up and shareholders fared well. So I don’t know how this will shake out.

Knowledge at Wharton: How do you think the Microsoft case will play out?

Faulhaber: It will take years. Judge Jackson will come down favorably for the government, but it will be appealed. Who knows what will happen in the long run? But in the short term, I don’t think a breakup ruling will do a thing. There’s so much inertia in the market, I don’t think there’s a tremendous demand for new and different operating systems. In terms of a new operating system, that’s already in the works. If it will be anything, it will be Linux, not "Son of Windows."