study on copyright term

I came across a posting on Ars Technica about a study by a Cambridge University PhD candidate in economics on the optimal terms for copyright.

According to Rufus Pollack's paper [PDF], the optimal term is 14 years, calculated based on the premise that the optimal term for copyright drops as the costs of producing creative work go down: It continually grows less expensive to produce, reproduce, and distribute works using new digital tools, so the therefore the costs for creative work has dropped and will continue to do so. His work presents a set of equations focused on the length of copyright term based on what empirical data there is to determine the "break even" point for the value of a work in copyright. His proposed optimal term is designed to encourage a balance between the incentive to create new work and the social good that comes from works entering the public domain. It's a very interesting set and assumptions and some even more interesting math. I expect to see this attacked, though, not only for its conclusion, but because there is a lack of empirical data for some aspects of his calculation: the data just isn't available to use. He makes plausible estimates and defensible justifications, but I still expect that will get his work attacked. A shame.