A number of structural remedies have been proposed as possible solutions to the Microsoft antitrust activities.

Behavioral remedies are meaningless.

Before I discuss the structural remedies I want to explain why behavioral remedies are not considered here. In a nutshell, Microsoft has proven that it will not do anything voluntarily. They already agreed to one consent decree and then proceeded to flatly ignore it. Yes, they said they were just "integrating", but that claim is little more than deceit. Judge Jackson himself held hearings to discuss holding Microsoft in contempt of court for how they abided by that agreement. So, Microsoft needs to be restructured to remove any ability to violate antitrust laws.

What will it take?

What kind of restructuring will it take? How do you do it? How complete must it be? What will the stockholders get? What will the employees do? How will the customers be affected? How will future software development proceed? Who will do the necessary research and development for the technology to advance?

Well. The first idea that needs to be dispensed with is that somehow Microsoft Corporation must be kept intact in order for the technology to advance. This is pure garbage. Very little original technology has ever been developed by Microsoft Corporation in the past. It just so happens that the truly innovative and ingenious technologies do not originate with the big power hungry corporations. Microsoft is no exception. At least not within their main product forces or dominate marketplaces. You only need recall the development of windowing at Xerox to realize that great ideas can and will surface from just about any source. Xerox did not even know what to do with the technology. It took Apple to steal the ideas and later for Microsoft to steal them again before the benefit of the ideas were fully realized. Keeping Microsoft intact might enable them to steal more technology but that is no reason to do that. The development of Java is another example of how truly innovative technology can and does have origins that have little to do with the power of a large corporation. It is true that Java was developed at Sun, a very powerful corporation itself. But, the Java project was almost an afterthought. It was a very small project at its inception. And, it was not developed as any defensive or offensive weapon to protect Sun's existing product line or business.

We just do not need large dominant corporations to foster computer software technology. Linux is another recent example. Linux was developed initially by an individual for heaven sake.

If you were looking for a good reason to break up Microsoft, you could argue that they should be broken up because they have not developed any worthwhile technology. They even had to buy DOS from someone else.

But, those are not the reason to restructure Microsoft.

The only legal reasons for restructuring Microsoft relate directly to its monopoly power and the impact that power has on future product development, implementation of advanced technologies and consumer choice. In the recent past, Microsoft has successfully rubbed out a number of significant and superior technologies (and lots more are on the bubble). Technologies that offer real benefits to consumers. Examples are disc compression (across platforms), networking (across platforms) and most recently Java and internet browsers offering cross platform benefits. Microsoft clearly suppresses any and all technology that they can not restrict to their monopoly products, their consumer and enterprise operating systems. That might make sense to protect their monopoly but it harms consumers enormously. The key to any remedy is the complete removal of monopoly power.

But, how do you remove the raw monopoly power that is so harmful to the industry?

Splitting Microsoft into three companies.

Well, you can split Microsoft into three parts as some have suggested. Each part can own a complete copy of the code base for all products. And, for a while at least, these three new companies can compete against each other for the business that used to be Microsoft's. But, that would only put an oligopoly in the place of a monopoly. An oligopoly can just as easily conspire to hold off competitors to its "oligopoly". And, they would be expected to do so.

Where would independent developers stand? Right were they stand now, out in the cold. Any company offering new and advanced technology would be required to get the "threesome" to adopt its technology or suffer the real possibility that none of them would do so. How would a product like Java fair? Would one company adopt Sun's version yet another decide to fracture it like Microsoft has done? They might. Or, they might all go along with the Sun version. Or, if they do not think independently they may all go along with the current Microsoft plan to fracture the Java marketplace and attempt to make Java a proprietary technology of the "oligopoly".

The real problem with the oligopoly idea is that it would still need the same oversight and regulation that Microsoft needs. An oligopoly solves little.

Splitting Microsoft 6 or 8 ways?

But, what if you do not split it 3 ways. What if you split it 6 or 8 ways? Six or eight companies are much less likely to illegally conspire to block an upcoming competitor like Netscape was to Microsoft. But, do we need 6 or 8 little Microsofts? Could they survive? My guess is that they would not. After a short while those less well managed or aggressive would go out of business for one reason or another. This is why I have suggested that the split not create separate and independent companies but rather divisions within existing major players in the industry. You can create the separate companies if you wish. And, then let HP, IBM and other OEMs buy one of them each. I assume they would be interested. Go ahead, ask them.

If the 6 or 8 small companies were owned by the other major players, they would all survive the long haul. Surviving the long haul is important. If you are going to restructure Microsoft, you do want the restructuring to stick. You want the natural market forces to keep competition in play and not have the biggest and nastiest of the group end up with all the marbles after 5 years. Splitting the operating system technology in this way would remove any monopoly power. Actually, it would just disappear. None of the new owners would have it. Hp and IBM could include their versions with the systems they sell and could offer packages to smaller OEMs on a competitive basis. The important concept is that no one company can call the shots.

Who will do the research and development?

But, what happens to the technology? Who continues the research and development? Does the industry really want 6 or 8 incompatible versions? Answering the last question first, the industry does not want 6 or 8 incompatible versions. Consumers might want 2 or 3 versions to pick from but clearly would be confused with 6 or 8. Now, as long as applications ran on all versions, having several choices would not be bad. It may be necessary for the group of 6 or 8 to form a voluntary standards body to assure that all versions are compatible for the purpose of applications. However, all versions would not be required to offer networking services, integrated browsers, disc compression or even a full range of system utilities. Individual consumers just do not need the same utilities and services that small or medium sized businesses might need or want. You can look at the range of packages offered with Linux to realize that one box does not need to fit all customers. Perhaps a consumer box could include the Solitaire game but omit networking. While the business version could include networking but omit the damn games.

Microsoft has argued in court that all consumers must buy the same box. But, that argument is pure garbage. Microsoft itself offers two boxes (Windows98 and NT 4). Actually, three if you count NT Server. The only reason Microsoft makes that pitch is to force the sale of unwanted products.

But, how is the standard compatible operating system to be maintained and developed over time? Well, it is not an impossible task at all. Microsoft and others are suggesting that the Java standard be placed into a standards body and not be controlled by Sun. Well. If that suggestion has any merit at all, then "Windows" can also be placed into a standards body. The 6 or 8 companies can participate in that body. They could even contribute funds to allow the body itself to further develop some technology as necessary. But, keep in mind that operating systems do not need continuous and ongoing research and development. OSF failed for Unix. But, Unix versions did not need to be 100% compatible in order to have a viable market. Larger systems just do not use mass distributed shrink-wrapped applications. The consumer OS marketplace is very different. Not one of the 6 or 8 players would want to violate the standard set and approved by the group.

But, what about applications? Would applications be stymied because the operating system technology is stagnant? Applications rely upon APIs, right? And, right now they all come from the operating system. Well. APIs do not need to come from the operating system. Microsoft itself was all bothered by the threat Netscape posed by the idea of the browser exposing APIs. Well, Microsoft did not like the idea simply because it meant it might lose its monopoly power, but the idea that APIs should be offered by middleware rather than the operating system is a very good one. Just look at the value Java offers. The Java concept permits applications to run across many platforms (large and small, specialized and general, consumer and enterprise). If applications are limited to APIs coming from the operating system then consumers are restricted in their choice of operating systems rather than given the choice of OS and given the choice of the operating environment. Java is not the last of cross platform environments. It is only the first really successful one. The Netscape browser (even without Java) is another example.

The only reason Microsoft suggests that the operating system be the only source of APIs is because they own it. From a technical standpoint, their position is unfounded. The operating system should rightly be a commodity. Linux has more than one GUI. There is no reason why Windows could not support more than one GUI either. (In fact, if you accept Caldera's legal arguments in their antitrust action against Microsoft, DR-Dos will support Windows95/98 if Microsoft would let it.)

Microsoft wants to be the only company offering APIs to anyone. But, there is absolutely no reason for this. And, there is no reason why everyone else should be seriously disadvantaged if they want to offer middleware that provides APIs for applications. Microsoft itself has suggested that IE is superior because it exposes APIs to developers. That might be an advantage that IE offers. But, there is absolutely no reason why only Microsoft can offer such advantages. That possibility should be just as realistic for any competitor who might want to offer a platform capability or runtime environment. There is absolutely no reason why only the OS vendor should be able to do so. To limit the offering of APIs to the OS vendor is only to restrict superior technology.

So. If you split the Windows code 6 or 8 ways, you end up with a group of competitors writing their version of the consumer OS to the common standard. Do you think that HP or IBM might make their version a bit more reliable than Microsoft has? Both HP and IBM know how to write reliable and dependable operating systems. Microsoft Corporation has proven to be incapable of that technical task. If HP and IBM fix their code, then the other 4 or 5 companies will have to as well. Finally, you would have companies competing on the quality of the product instead of what kind of illegal marketing scheme is required to achieve dominance in the market. The 6 or 8 would have captive markets for their own sales of PCs and would compete for the business of lessor OEMs.

Where do the consumers stand?

Where do the consumers stand? How do they benefit? Assuming the 6 or 8 companies stay in the standards group and comply, all consumers benefit. Applications written to the "Windows" standard work on all versions (some Windows applications from Microsoft do not work on NT and Windows95 today). Java and browser based applications might work across all platforms without one company or another trying to sabotage the technology. Those applications that require cross platform capabilities can succeed. Applications that only have merit at the consumer level also succeed.

But, what happens if one of the 8 wants to bundle an application (I.E. pick anyone you want). If IBM wanted to bundled Lotus Notes with their version of Windows, they could do so. Consumers who do not want Lotus Notes could just buy their PCs from HP or Dell. So, will IBM bundled Notes? Not very likely. What about a browser? Well. Customers who prefer some other browser, would not buy from IBM. When consumers have a choice, forced bundling does not work well.

There is one concept that seems to escape obvious mention in the current Microsoft antitrust trial. Without monopoly power, application bundling just does not work. Consumers will just go elsewhere if the do not like the packaging. That is why Microsoft absolutely has refused to sell Windows without IE. They know lots of consumers would take that option and pick and choose their own applications. Employees of Microsoft have pointed this out themselves. So. Microsoft refuses that choice of consumers.

What happens to Microsoft employees?

What happens to the employees of Microsoft? Well. Microsoft Corporation is gone. But, their expertise will be needed by the 6 or 8 companies. Most employees would relish the opportunity to work for a company other than Microsoft. Many developers would take the opportunity to start their own companies if startups would have a fair chance at marketing their technology. Some employees might have a hard time getting jobs. If the only reason they were successful at Microsoft was because Microsoft had monopoly power, they may find real competition in the job marketplace. Any idiot can bundle a lousy product with a monopoly product and force its sale. But, the rest of the industry knows how difficult it is to be successful in the marketplace when real competition is present. With real competition you have to provide value and good marketing. With real competition, you can not just cheat.

But what about the stockholders?

But, what about the stockholders? What do they get? Simple. The existing stockholders get the non-monopoly value of the Microsoft assets as determined by competitive bidding under the control of the court. They either get cashed out or receive stock in the companies buying the 6 or 8 copies. No corporation nor stockholder has any legal right to the monopoly value of their assets.

As always, the devil is in the details. But, this article does try to lay out a possible solution.

Lewis A. Mettler, Esq.