SYNOPSIS: Justice's half-assed plan to break up Microsoft goes too far by not going far enough

So now it's official. The Justice Department has requested a breakup of Microsoft into two companies -- one piece that gets the rights to Windows and a defenestrated rump that gets the rights to everything else, mainly Microsoft Office.

In my last column I offered a metaphor: a baron who controls two castles along the Rhine, and extracts tolls from passing traffic. The Justice Department proposal is like letting the baron keep one castle, while giving the other to his nephew. That metaphor suggests that such a breakup would make things worse, with two robber barons instead of one levying tolls on those who pass by. Why, then, does Justice think otherwise?

As best I understand it, the argument is that neither castle is impregnable by itself. Both Office and Windows are near-monopolies sustained in part by "network externalities" -- economese for the incentive people have to use the same products that other people are using. But -- or at least so the argument goes -- the dominance of Windows is partly sustained by the dominance of Office, and vice versa. People who might use WordPerfect are deterred by the fact that Microsoft Word works better with Windows; people who might switch to Linux are deterred by the fact that there is no Linux version of Word.

If this is true, then putting the castles under separate ownership might loosen the control of both castle lords over passing traffic. Over time Windows would become only one of several competing operating systems, and Office would become only one of several competing office suites.

Now one question is whether separating the two pieces of the company would really eliminate the monopoly power of each piece. This is plausible for the downstream applications like Office; after all, rivals to Office survive even now, and would probably do much better once Office lost its home-court advantage. But remember that the objective of antitrust policy is to help consumers, not disgruntled rivals.

The idea that the proposed breakup would undermine the dominance of Windows seems much less compelling. And if Windows remains dominant -- if the upstream baron still controls a crucial choke point -- then the breakup would still be a bad thing for consumers; a baron who controls only one castle has less incentive to hold down tolls than one who controls the whole stretch of river, even if there are several downstream barons.

But suppose that slicing Microsoft at the waist does lead to a true competition among operating systems, that once Powerpoint runs on Linux, the dominance of Windows will fade away. There is still a puzzling question: if your objective is to introduce competition in operating systems, why not introduce competition in operating systems? That is, why not follow the advice of a number of Microsoft critics -- like Robert Litan of the Brookings Institution, who argues that any solution that leaves Windows intact is no solution at all -- and break Microsoft into three or four parts, an applications company and several rival sellers of Windows?

After all, the major objections to such a "vertical" breakup of Windows apply with even greater force to Justice's actual proposal. Some have worried that creating rival Windows companies would lead to the emergence of rival, incompatible versions of Windows; well, what could be less compatible than a completely different operating system? Others have worried that breaking up Windows would amount to expropriation of intellectual property, and would put a chill on future investment; well, if the Justice plan works and eliminates Microsoft's market power, how is that different?

The only explanation I can give of Justice's half-measures is political: fear that a straightforward proposal to slice up the operating-system market would seem too radical. So the plan is to achieve the same result by stealth and indirection. But half-measures in economic policy have a way of being worse than a clear decision either way. This plan could produce the worst of both worlds: it could place a chill on investment even while raising prices and creating system incompatibilities.

So here's a plea for clarity: If the court determines that Microsoft's sins warrant dismantling of its Windows monopoly, then so be it. But let's not try to break Windows while pretending we aren't.

Originally published in The New York Times, 4.29.00