Epstein's Comments: Mobblog on the New Servitudes

It is of course an honor to be invited to comment on Molly Van Houweling’s article, on the New Servitudes, and, I think, important to do so, because in my view the extension of licenses to bind third parties to business arrangements counts as a welcome and not a troublesome development for the law of property in general.

My point of departure in this discussion is with the law of servitudes as it relates to the law of land. In the beginning, when there was no system of public notice by recordation, the question of binding third persons without notice was a serious issue. Huge portions of early law was devoted to the question of how to make sure that the chain of title was kept in order, so that no person could convey the same interest in property to two persons at the same time. The limitation on the number and types of servitudes was critical to improving conveyancing. Easements in this world were preferable to restrictive covenants because they gave notice to outsiders by influencing the physical layout of the property in question. It is easy to observe a path that runs across one’s land to water.
These restrictions on servitudes, however, came at a real cost. They prevented the efficient segmentation of real assets by voluntary agreement. Introduce the recording system, and the evidence of multiple interests is now visible in reliable form. The conveyancing externality is largely contained, so the gains from trade can be allowed to multiply. This is simply an instance of the Coasean insight that devices that lower transactions costs increase the gains from trade.
It turned out, of course, that even with respect to land all covenants and easements were not enforced. But it is important to isolate the reasons why this might be the case. The first of these was said to cover covenants that did not touch and concern the land. I see no case for making this an external constraint on the process, but no danger from that taking place either. Why? There is no intrinsic gain to having the person who buys might land covenant to bind his successor in title to cut my hair or mow my lawn. In both cases, separable transactions are likely to be more efficient by far, so that the law should be understood to prohibit only those transactions that are not likely to take place at all. It is a conceptual imperfection of little cost.
The other constraints relate to matters that attack voluntary contracting more generally. Here I refer to such matters as the antitrust laws that deal with monopoly or the civil rights laws that attack racially restrictive covenants. Accept the dangers in these practices, and there is a reason, not unique to covenants, to restrict these practices. But otherwise the law of covenants is folded into the law of property and contract more generally.
I do not think that there is any reason why this model should not apply to personal property. Historically, there were reasons why covenants did not usually run with chattels. These were usually property of limited value and often of short duration. There is no great harm, one might think in blocking the creation of interests that no one wants to create. The situation is therefore a bit like prohibiting any landowner from bundling haircuts with real property.
But once we have the situation with computer programs, it is not just the transfer of property that is at stake; it is the reproduction of the property for use by second and third persons that really matter. Now the case for controlling the use is much stronger, for otherwise, the transfers by the original licensee will cut into the market for sales by the original licensor. These licenses do not strike me as creaky. Their terms are easily understood if one wants to read them. In most cases reading them is not necessary because it is not that further copying is prohibited. There are some quibbles that arise about boxes that do not give notice of contents, but these are relatively small, especially with repeat users, so I think that notice is given to the world and we have superior internalization of costs and benefits that we get with any system that imposes for any reasons limitations on how these contracts work, except for those which are derived from the antitrust or the civil rights laws, and those issues don’t happen here.
The point here is important because there are all sorts of ways that licenses could take place. Molly has worked with Creative Commons, which is all to the good. But its dominant technique is to craft licenses that allow reuse with acknowledgment and not payment. The basic system depends on the network of contracts with notice that is at stake in all these cases. The ultimate lesson is clear. Since technology negates the notice issue, we should encourage the use of any and all forms of licenses that do not offend doctrines from outside the property area. There are no reasons of standard property, contract or intellectual property law to deviate from this result.