The Informationist

In a competitive operating systems market, vendors would integrate new functionality at their own risk. Technically smooth integrations that made consumers happy would win market share away from competitors—who would then have to integrate the new functionality as well or wither to insignificance. Technically premature integrations would annoy consumers, degrade OS performance and system stability, and drive market share away from the integrator. The market mechanism would either reinforce or negate the underlying engineering decision.

The antitrust argument has long been that the structure of the operating systems market skews incentives. Consumers cannot punish Microsoft for engineering missteps stemming from premature integration (or, as I have argued, for replacing engineers with marketers as the key decision-makers in product development). Microsoft, on the other hand, retains the standard incentives to make life difficult for its competitors. In this environment, it would be irrational for Microsoft to do anything other than integrate prematurely—thereby curtailing innovation and competition. There is no evidence that Microsoft is an irrational corporation.

But the key to Microsoft’s rationality lies in its ability to leverage the IP rights that we gave it. Largely due to a lack of analytic thought, we have arrived at a situation where a typical commercial software product bundles: (i) patented algorithms and (ii) trademarked logos and icons into (iii) copyrighted source code subsequently both (iv) maintained as a trade secret and (v) compiled into copyrighted object code, then circulated subject to (vi) a shrink-wrapped license. That combination of legal protections makes software among the most heavily regulated products on the market—and almost entirely with stealth regulations whose regulatory nature many people miss (or deny). It is precisely the courts’ willingness to enforce these regs, though, that enables both Microsoft’s basic business model and the activities about which its antitrust critics complain.

The proper legal remedies (at least under U.S. law) for a company that leverages its patents or copyrights to harm markets that those IP rights don’t cover lie in IP law, not in antitrust law: they are known as “patent misuse” or “copyright misuse.” Under a misuse remedy, the courts simply announce that they will not enforce IP rights that their holder has misused unless and until the effects of that misuse have been purged.

As long as Microsoft’s critics continue to push after-the-fact antitrust remedies to solve what is really a problem in IP policy, many of their remedies will have minimal effect—and make them look foolish. Behavioral remedies (e.g., decoupling) can matter at the margin; structural remedies will be hard to implement. Neither addresses the fundamental problem.

If we really want to see a vibrant free market in software that motivates broad-based innovation, we’re going to have to face the underlying problem: our approach to IP rights in software is strangling us.

I develop this argument in greater detail in: