Tom Barnett’s speech on the Interoperability between Antitrust and IP Policy is the first major DOJ policy statement on the competitive effects of digital rights (or restrictions) management (”DRM“). The importance of DRM cannot be overstated. Arguably, DRM is the pivotal information technology issue of our time. Let me highlight just a few aspects of the effects of DRM, and note how ambiguous they are in terms of consumer welfare.

• From an economic point of view, DRM introduces engineered scarcity by transforming non-rivalrous information goods into rivalrous goods. (IP/DRM introduces scarcity into a world of knowledge goods that, once created, can be replicated to satisfy virtually any demand.)
• Engineered scarcity drives up the price for information goods. The profits from higher prices sustain the legal distributors and some creators of DRM’ed information goods.
• DRM has the potential to spur innovation and increase the production of some information goods by allowing producers to recoup their R&D costs.
• DRM has the potential to diminish innovation and decrease the production of derivative or complementary information goods and of interoperable hardware.
• DRM has the power to create fully integrated verticals, locking content into a specific “trusted” distribution and hardware environment.
• DRM shifts control over content from the consumer to the holder of the content distribution rights.
• DRM shifts control over general computing hardware from the consumer to the holder of the content distribution rights.
• DRM allows for unilateral, post-acquisition changes in the utility of user-acquired information goods.
• DRM is an exclusive connection technology. It connects information goods with a certain hardware environment, with certain persons (the authorized users), and with its origin (the content distribution rights holder).
• DRM pushes markets towards high entry barrier platform competition and away from low entry barrier application competition.

Barnett deserves applause for taking on these incredibly complex issues.

As an initial matter, I am delighted to see that Barnett fully endorses the significance — maybe even the primacy — of dynamic efficiencies.

It follows from the Schumpeterian view that antitrust law, with its focus on improving consumer welfare, has a keen interest in protecting innovation. Fostering innovation requires recognition of the benefits of dynamic efficiency and the dangers of focusing myopically on static efficiency.

I also agree that in some R&D intensive industries that require massive up-front investment in order to create leapfrog information goods, strong (but narrow) patent protection is a proven way to ensure that the innovator internalizes a fair share of the benefits that he or she creates (innovative drugs come to mind.) I am skeptical, however, that — without more — a generalized case for strong copyright can be made on the basis of the incentives theory alone, and I am even more skeptical that arguments for strong copyright protection necessarily support arguments for DRM if consumer welfare is the evaluative standard. As an initial matter, Barnett subscribes to a strong version of the antitrust-IP harmony theory.

[S]trong intellectual property protection is not separate from competition principles, but rather, is an integral part of antitrust policy as a whole. Intellectual property rights should not be viewed as protecting their owners from competition; rather, IP rights should be seen as encouraging firms to engage in competition, particularly competition that involves risk and long-term investment.

Now the goal of IP policy is not to create incentives through exclusive rights. It is to increase the quantity, quality, and variety of information goods available to the consumer. Consumer welfare is the goal that IP and antitrust policy share. However, the means by which antitrust and IP policy promote the policy goals of “more, better, and different” are not identical. IP law relies on exclusive rights, antitrust relies on open competition. Some exclusive rights may foster open competition, others may serve primarily to exclude competitors from the marketplace. The difference in means may ultimately be more significant than the common goal. So I am somewhat less sanguine about the IP/antitrust harmony thesis.

Barnett then walks through three objections to IP/DRM combos that are advanced by the “access faction” against the “asset faction,” using iTunes as an example and explains why he is not convinced by them.

1. “[C]onsumers are locked into buying songs only from the iTunes service and that they will have to pay too high a price for iTunes songs.” Not true, says Barnett, because users are free to upload songs from other sources, they can “re-record an iTunes song in an MP3 format,” and iTunes has brought prices for songs down.
2. Apple is following a printer/ink strategy, selling cheap devices and expensive songs. Barnett is not convinced, because, if anything, Apple is selling expensive devices and cheap songs.
3. “A third theory is that, darn it, ‘information just wants to be free.’ That quote is so much in use on the Internet that I could not pin down its original source. Wikipedia attributes it first to a participant at a computer hacker’s conference in 1984. In any event, this argument is not based on competitive effects and consumer welfare. Information may want to be free, but information creators want to be paid — they will not create without rewards.”
4. Lastly, the song/player combo harms competitors, which is neither here nor there from an antitrust standpoint, unless the continued survival of a competitor is essential to maintaining competition in the marketplace.

This is a powerful and articulate defense of the “asset faction’s” key arguments, and I am curious to see how the “access faction” will respond. I find Barnett’s arguments as to (2) and (4) most convincing. As to (1), it all depends on the facts of the individual case. A semi-open system is certainly less of an antitrust issue than one that is fully locked down. As long as free mp3s play alongside a4ps, the player is indeed a dual use good, and the ability to de-DRM one’s collection of songs via burn and rip somewhat diminishes the lock-in effect. However, DRM policies may change retroactively, but I understand the argument that antitrust enforcement should deal with those issues if and when they arise. The weakest point, in my view, is (3), which is also the most fundamental.

First, I disagree with Barnett’s categorical claim that “information creators want to be paid — they will not create without rewards.” Some do, some don’t. Where did Barnett turn to, when he researched the origin of “information wants to be free?” He turned to Wikipedia, an encyclopedia created entirely by unpaid volunteers. The explosion of the blogosphere, the existence of Linux and other open source projects, the success of Amazon, driven by user-contributed book reviews, and the world of online academic publishing (think SSRN) belies the claim that rewards are a necessary condition for the creation of knowledge goods, at least the kind of direct monetary rewards that IP/DRM are designed to generate. As I said initially, I agree that some information goods require incentives qua exclusivity grants, but we should be cautious to make more general claims on that basis.

Second, while there is no reason to focus on exclusive rights as the only — or even the primary means — for compensating content creators, it is here that I see the greatest shortcoming of the “access faction.” The “access faction” lacks a convincing creator remuneration model. Unless the proponents of less IP/DRM and greater access can make a (more) convincing case for the commercially profitable creation of information goods that does not rely on transforming non-rivalrous goods into rivalrous goods via IP/DRM, it will be hard to overcome the (properly qualified) claim that “information creators … will not create without rewards.” IP/DRM is a proven tool to compensate artists, there can be no doubt about that. But it may not compensate artists very well and it may impose extraordinary opportunity costs on society. Antitrust enforcement should ensure that there is commercial and legal room for alternative creator remuneration concepts to develop, because in practice, IP/DRM is not just a tool (nothing ever is), IP/DRM is also a particular way of doing business. And, quoting Barnett quoting Schumpeter, antitrust should never protect existing business models from leapfrog competition

which strikes not at the margins of the profits and the outputs of the existing firms but at their foundations and their very lives.

This is an extraordinarily important and fruitful debate. I am glad to see that the DOJ leadership is getting involved in it.

NOTE: Cross-posted at the Law & Society Blog.

Technorati Tags: antitrust, IP, DRM, itunes, barnett, doj, policy

Comment feed: RSS 2.0. You can leave a response, or trackback from your own site.