Singleton, the DMCA, and Interoperability

The last major topic of Solveig Singleton’s defense of the DMCA that I want to address is the interoperability issue. Here’s her take on it:

The main obstacle to more interoperable DRM or reverse engineering is not the DMCA, it is a business problem. DRM has an advantage in security and speed to market when only one company need be involved in its development. A more open process is slow and the result is not usually cutting-edge. There are endless negotiations, a host of issues with compatibility with legacy equipment, and serious trust issues. CSS protection for DVD’s was jeopardized partly because a licensee, Xing, neglected to encrypt a key (though the system had other weaknesses as well). The more players, the more risk.

Let us examine the example of Apple and the iPod/iTunes model more closely. Apple has the idea of selling music at a low price and making money on the hardware, just as radio broadcasters once sponsored programming in the hope of selling radios. To offer a library of music, Apple needs to convince music rights-holders that the files aren’t going to show up free everywhere. This they are most likely to be able to do if they control the security technology. Furthermore, would Apple bother if it anticipates other companies coming in and cutting into the profits they want on the hardware? Unlikely; this was the reason that radio broadcasters were moved to advertising to fund radio, but this won’t work in a digital world if advertising can easily be stripped out. Finally, what gains do we get if, say, Real Networks hacks the iPod, Apple puts out a fix, Real Networks hacks it again, and Apple fixes it again? There’s nothing there for consumers or entrepreneurs but quicksand. We have not Schumpeter, describing the process by which the candle is replaced by the light bulb, but Hobbes’ war of all against all. Meanwhile, there is plenty of competition in the tunes market without breaking Apple’s code. There are many different kinds and levels of competition.

Now, what’s most interesting about this passage is that it concedes one of the central contentions of my paper: that the DMCA doesn’t merely assist in the enforcement of copyrights, but actually creates a new kind of quasi-property right in technology platforms. In my paper, I spend a fair amount of time talking about the IBM PC example as a model for how interoperability and reverse engineering worked before the enactment of the DMCA. IBM would have loved to prevent competitors from building PC clones, but copyright law didn’t give them any way to do that.

IBM isn’t an isolated example. There are tons of other examples throughout the computer industry wherein one company pioneered a new format or platform, and other companies (or open source projects) reverse engineered the platform to build compatible, competing products. Here are a few more examples: AMD reverse engineered Intel’s x86 instruction set in order to build compatible processors for desktop PCs. The Samba open source project reverse-engineered Microsoft’s file and print server protocols in order to allow Linux to exchange files and print jobs with Windows computers. The netatalk project reverse-engineered Apple’s proprietary Appletalk protocols to allow open source operating systems like Linux to work on Mac networks. And Microsoft copied many of the proprietary features of Netscape Navigator when it was building Internet Explorer.

In short, one of the models for the development of a new standard is for one company to create a standard that becomes an industry leader, and then for others to copy that standard for their own products. This doesn’t in any way infringe on the ability of the creator of the standard to innovate: he doesn’t have to consult with anybody, nor is he prevented from changing the format as he pleases. However, it avoids unnecessary technological lock-in. Consumers whose data is locked up in a particular format aren’t compelled to purchase products from that particular vendor for the rest of their lives in order to access their data.

If I’m interpreting Singleton’s argument accurately, she claims that it’s a good thing that we’re abandoning that model–that we’d have been better off if Phoenix, AMD, Samba, et al had been legally prohibited from cloning the then-dominant products and enabling intra-platform competition–that the fact that the DMCA prevents potential iPod competitors from allowing consumers to access their iTunes songs isn’t a bug but a feature.

Singleton asks, “would Apple bother if it anticipates other companies coming in and cutting into the profits they want on the hardware?” I think the answer is obviously “yes.” “Coming in and cutting into the profits” is what competitors do. It’s what AMD did to Intel, what Phoenix and Compaq did to IBM, what Internet Explorer did to Netscape. And it’s what Sony, Creative, Dell, and others should be able to do to Apple’s iPod. The fact that Apple created a platform doesn’t give them a right to extract monopoly rents from that platform in perpetuity.

It strikes me as simply incredible to contend that Apple would not have created the iPod (or iTunes) without a government guarantee that other companies would not be permitted to build compatible products. The creator of a successful technology platform reaps enormous profits even if competitors quickly move in and create clones. IBM was very successful in the PC business despite the entrance of PC clones. Intel continues to be the dominant creator of x86 processors despite more than a decade of competition from AMD and other companies. Incumbents have tremendous advantages in name recognition, economies of scale, and control over the underlying format. Creating the iTunes/iPod platform would be tremendously lucrative with or without the legal monopoly over that platform granted by the DMCA.

Indeed, the history of the computer industry consists of what Singleton characterizes as “Hobbes’ war of all against all.” There’s a reason Intel’s Andrew Grove wrote a book titled Only the Paranoid Survive. Computer companies have never had the luxury of creating a successful technology platform and then resting on their laurels while they collect monopoly rents.

But the DMCA changed that. Now, if your technology platform encrypts copyrighted content (which more and more platforms do) the DMCA gives you a quasi-property right in it. You get to decide who’s allowed to interoperate with the platform and who isn’t. In my opinion, that’s a bad thing. Had something like the DMCA been on the books in the 1980s and 1990s, the computer industry would have been less innovative, as companies like IBM and Intel could have used it to exclude competitors like AMD and Compaq from building interoperable products.

Now, it’s true you would have still had inter-platform competition. Apple would still have competed with IBM and Motorola would still have competed with Intel. But the problem with this type of competition is that it’s extremely inconvenient for users. Switching from a Mac to a PC, or from Word to Word Perfect, or from Lotus 123 to Excel, is a costly and time-consuming process. So wherever possible, our legal system should be set up to allow intra-platform, as well as inter-platform competition. I don’t think we should be celebrating the fact that in order to switch to a non-Apple MP3 player, I would have to re-purchase all my iTunes Music Store songs from another vendor.

Frankly, I don’t think that Ms. Singleton’s paper really addresses this critique very well. The DMCA represented a radical shift in the legal balance between platform incumbents and challengers. This shift has little to do with enforcing copyright, as such. Dell MP3 players and Linux DVD software are not piracy tools, but prohibiting them has dramatically reduced the possibilities for intra-platform competition of the sort that was the norm pre-DMCA. I would be very interested in more details from DMCA supporters about why we should regard this shift as a feature, rather than a bug.

SHARE: