I was not officially invited to join the discussion, but I wanted to sneak in a thought or two in reaction to some of the themes already raised.
First, it may well be that the incentive perspective has been the dominant approach in American IP law, but it would be incorrect to characterize it as the prescription of economic analysis. To restate what is probably obvious to all, a utilitarian (or “economic”) social welfare function, as applied to IP, measures only one dimension: the total net value from creation, use of, and access to the information goods. Incentives-to-create matter only indirectly, as a feasibility constraint, to guarantee that information goods that are costly to produce and to distribute will be supplied. But what the social welfare function is set to maximize— the ultimate goal under the utilitarian framework—is not to promote more inventions but rather to promote more value from use.
Second, it may well be that the tools of economic analysis do not currently have the necessary resolution to prescribe the precise limits of IP rights, such limits that would strike the perfect balance between incentives and access, and would lead to maximization of social welfare. But the fact that the principle identified by economic theory—the tradeoff between access and incentives—is difficult to implement with precision does not mean that other principles for limiting IP rights should trump it. Choosing not to look at incentives, or to “override” the incentive/feasibility calculus, would not make it go away. Ironically, the conclusion I draw from the complaint that economic analysis is too crude is that we need to focus on it more, to further refine it, rather than to look elsewhere.
Third, having read some of Madhavi Sunder’s illuminating writings on this issue, my impression is that some of her examples for excessive IP rights are cases in which the added incentive to creators is either non-existent or too small to justify the restrictions on the types of benefits derived by users or remixers of those creations. I tend to agree with Madhavi’s sentiment, and yet I think it can be perfectly stated within the utilitarian framework. In fact, it suggests – despite my second comment above – that the economic analysis does have surprisingly fine resolution to accommodate subtle observations about incremental effects on welfare and incentives. It can deliver clear cut recommendations in cases in which the incentive effects are very likely to be minimal.
Fourth, Sunder provides interesting examples to demonstrate the cultural context of information goods. I view these insights as nice complements to the economic framework. For example, culture manifested in “participatory communities” (YouTube, fan fiction, and the like) can help us identify network effects in consumption of information, which significantly increase the use value of the information and, other things equal, ought to shift the incentive/access tradeoff in favor of more access, and in favor of protection to the providers of platforms for such communal activity. Or, culture can help us identify an additional source of utility derived not from consumption of the information good but rather from production of it. Adding this source of value to the utilitarian social welfare function would, again, likely shift the tradeoff towards less protection and more access. Economics as a discipline cannot determine what goods or activities provide value to individuals. It welcomes any insight from other disciplines regarding sources of value, including insights from cultural perspectives.
Finally, I am not sure why the Disney example is relevant to the cultural issue. To be sure, Jungle Book and other stories were plucked out of the public domain, but the cultural value enjoyed by kids—the songs, the animated representation of the characters, and the nexus of ancillary artifacts distributed through costly marketing investments—was created by Disney. Why would it be wrong to accord Disney protections (correctly calibrated) for these costly creations? Whether the original script comes for the public domain as in Jungle Book, or from collaboration with other writers as in Toy Story, does not seem to matter.