Adam recently posted that the British government is contemplating pretty outrageous penalties for Internet pirates. Naturally I wanted to chime in with an outrageously long comment about the nature of property and the (im)possibility of intellectual property (IP) rights. That’s just the libertarian thing to do! There’s a lot to say and it will take more than a few beers to sort this whole thing out, so I’m going to limit myself which means I’ll just raise more questions than I answer…
So let’s start with why IP doesn’t make a lot of sense. IP is information and information wants to be free. My use of a song doesn’t prevent you from using it. IP isn’t scarce in the way a car is. Besides, there’s some evidence that government enforcement of IP does more harm than good. And pirates end up spending more to buy IP than other people anyways (I usually listen to music for free on Grooveshark, but the other day I thought, “I’d really like to tell Willie Nelson that I think his music is great,” so I bought an album even though I could have listened to it for free!). And besides, musicians can make money by performing live.
But just because it isn’t a tangible thing, doesn’t mean we are forbidden from attaching rights to it, even in anarchy. Property rights are a “bundle of sticks.” I own my land, but you might have a right to the sunshine I would block if I built a skyscraper. Likewise, a society can come to some sort of quasi-unanimous agreement that the creator of a song has the right to control its use, even in the form of digital files.
Now, government enforcement of IP laws is fraught with difficulties even before we get to public choice issues. Should a patent be 20 years or 19.5? If the optimal patent length is 16 years, then the current system is a net subsidy and so creates economic inefficiency. But determining the optimal length in a world of benevolent political actors is an incredibly complex problem. How stringently should patents be enforced? How do we account for the differences in conditions that affect different patents (or copyrights, etc.)? This argument doesn’t say “don’t do IP,” it just says, “hey, this whole venture has its own set of costs we need to account for. It’s conceivable that we conclude that the optimal patent is probably between 5 and 20 years, but if we’re off by more than 4 years the costs of the error will outweigh the net benefits of the patent.
Then there’s the public choice problems. We don’t want IP law to be some Mickey Mouse operation set up to hurt consumers.
But (and that’s a big but!) we have to return to this issue of property rights. When I buy an apple, I’m concerned with the physical thing, but really I’m buying a bundle of rights. The rights are what’s being exchanged, and then later exercised. These rights are socially determined and often-but-not-always-or-even-mostly enforced by government. Yes, if I steal your car the government will probably get involved. Yes, the government provides a back-stop to rights enforcement in a lot of areas. But rights are ultimately a social-political construct that can exist in anarchy. What does this mean? First, it means that we could conceivably have intellectual property rights . Second, it means that we could have such rights in a state of anarchy.
Obviously the nature of the good will affect the viability of such a system. Enforcing IP laws is difficult enough when some third-party can come in and say “you’re a pirate and you’re going to jail.” In a common law situation where you have to make the plaintiff whole, it’s difficult to say what that means. Reputation plus property rights might keep comics from stealing others’ material, but it might just separate the comedy industry into auteurs with sophisticated audiences and Carlos Mencia with less sophisticated audiences.
We can safely label a law or institution as legitimate if it is unanimously accepted. In the case of IP, such unanimity seems unlikely. In any case, I still suspect that government involvement in IP does net harm although I’ll grant that it’s a (probably impossible to answer) empirical question.