Absolute independence. Government monitors in place to make sure there is no continuing collusion.
You seem to have this abiding faith in government monitors, which I do not share.
Even assuming that the government monitors act entirely in good faith and cannot be bribed or "captured" by the regulated entities, there is absolutely no assurance that they will know all relevant activity that is going on, inside the three companies and outside in the industry and economy, or that they will have the ability to understand it.
That is why Bork seized on three as the number of competing "Baby Bills" for his proposal. Three is the minimum number of companies he believes is necessary to prevent collusion.
It should also be pointed out that not all forms of collusion, aka "cooperation" and "contracts in restraint of trade," are harmful. Only certain forms of collusion (e.g., on price and on division of markets) are per se illegal. The rest are evaluated under a rule of reason that weighs the pro- and anti-competitive effects of the activity in question. The same rules should apply to these new companies as apply to the rest.
Also, if everybody owns something, nobody owns it. . . . So I would predict shared or open code would fail. . . . It's what MSFT owns that matters at this point, not the alleged skills and vision of it's development teams and managers.
Financially, I suspect that if they all own the same code, that code could not be taken into account in net asset value of the corps at nearly the same value as today, because they could not sell or license exclusive rights to anyone else, including in takeovers.
Even with my limited knowledge of software, I do understand that owning code is not the same thing as knowing how that code is built and how it operates. So, the value of the cold code is already zip; what matters is the knowledge of the people who use it to build software. This is especially true for a 40 million line program like Win 2K.
Although I have no understanding about how knowledge of the code base is shared at Microsoft, I think this basic fact presents some interesting implications for the various breakup proposals.
First, the practical difference between an undifferentiated breakup and a breakup along functional lines is not as great as it appears on the surface. In an undifferentiated breakup, each of the spinoffs will have gaps in its knowledge of the code base to which they all have access, just as I am sure they would in a breakup along functional lines. The difference is that, in an undifferentiated breakup, numerous, small gaps will appear everywhere, since different people on different teams will end up with different companies. In a breakup along functional lines, applications people's knowledge of applications code will be fairly complete, but they will know nothing about the OS code, for example. So, that OS code, which they have the right to use, will be worth relatively little to them, at least in the beginning.
The key difference between an undifferentiated breakup and a functional breakup as was previously proposed will be in the restrictions on the spin-offs to compete in each others' areas of specialization. In an undifferenciated breakup, they will all have the right to do so from the beginning, and, over time, they will. In the past proposals for a functional breakup, somebody would have to keep the OS company, for example, out of the business of the applications company.
Second, some "collusion" as you call it should be encouraged so people can share knowledge and jointly develop products for individual sale. For example, there is no reason to force the OS Company's people to withhold vital information about OS code from the Internet Company's people if they want to develop some product jointly or if they want to exchange information so each can develop products for the other's markets or that interoperate with the other's offerings.
Again, one should rely principally on the market to decide what collusion is desirable and what is not. The only collusion that should be prohibited is the same type that is illegal generally, that which raises prices or restricts output.
Third, breaking Microsoft apart probably will not affect the degree of cooperation that goes on as much as some people think. Much of the cooperation that presently goes on within the single company will still go on between separate companies, especially in the beginning. The difference will be that, over time, if the companies are separate, there will be no centralized control to keep each unit from maximizing its own profit at the expense of the others or of third parties in ways that a centralized plan handed down by Microsoft management may not allow today.
Read Bork's discussion of the similarity of the problems faced by antitrust mergers and acquisitions law and the problems faced by the law on contracts in restraint of trade to understand better what I am saying here. The basic problems are, at root, the same: what forms of cooperation, whether within firms or between firms, do you allow and which do you not? What, after all, is a merger but a specific kind of restraint of trade engaged in to achieve some efficiency, and what is a divestiture but the elimination of a restraint of trade which is less efficient than competition?
It's true that they are barely getting win2000 done now, and you can't break up the OS teams. So don't. A problem that will be mirrored in other areas in the 'baby bills' proposal.
What I have learned from thinking about your ideas is that you don't have to break up the OS team, or any other team. You just separate them into separate organizations and let the market, rather than Bill Gates' central direction, decide the degree of cooperation and competition between them.
Better yet, just have the government set some minimum criteria that have to be met to satisfy antitrust law, and let Microsoft decide the specifics of how any breakup is carried out, subject, of course, to the criteria being met. |
