Are there problems with the way IP is priced? Professors Christopher Buccafusco (Chicago-Kent) and Christopher Sprigman (Virginia) provide a novel experimental insight on this question in Valuing Intellectual Property: An Experiment, which was published in the November 2010 issue of the Cornell Law Review. I have been thinking recently about the prices that are set for intellectual property because of Amy Kapczynski's project on theorizing the costs of these prices, which she presented at Yale Law School today (and which I'll blog about in more detail once a draft of her paper is posted). But while Kapczynski provides the first thorough critique of price itself (challenging the premise that IP should be priced whenever transaction costs are low), Valuing Intellectual Property contributes to the second-order critiques that emphasize the particularly high transaction costs in the IP context.
Buccafusco and Sprigman conducted a clever experiment to study how people set prices for creative works by creating a market for 10 poems in a $50 poetry contest. They found that the "authors" who wrote the poems and the "owners" who were told they owned one of the poems were only willing to sell their poem (and the corresponding chance of winning the contest) for over $20, while "bidders" would only pay around $10 to buy a poem. This effect was the same whether the participants could see all 10 poems or not. And even when the "contest" became a random lottery (so that each poem had a 1 in 10 chance of winning), authors and owners would only sell for over $15, while bidders would only pay around $5 (the actual expected value).
This experiment is the first demonstration of the endowment effect (where people value something they own more than an equivalent thing they don't) for non-rival, created goods. Even though the authors were reminded that they would get to keep their poems (which would be emailed to them), they still priced the poems' values as contest entries more highly than the bidders. The implication for IP is that the deadweight loss caused by copyrights and patents may be even larger than previously expected. Although the endowment effect is well established in behavioral law and economics, these inefficiencies are particularly troubling in IP, where the marginal cost of information is zero. Buccafusco and Sprigman argue that their results suggest market failures in licensing IP, which supports the use of liability rules over property rules and a more expansive fair use doctrine. More broadly, their results demonstrate another problem with using price as a signal of value for information goods.