Glynn S. Lunney, Jr. (Tulane University School of Law) has posted Empirical Copyright: A Case Study of File Sharing and Music Output on SSRN. Here is the abstract:
In copyright, we are guided by a simple intuition: More revenue leads to more original works. But the relationship between revenue and creative output is not so simple. Broadening copyright in order to increase the revenue associated with any given work may ensure the expected profitability, and hence the creation, of additional works at the margins. At the same time, however, because copyright protection is uniform, broadening copyright also increases the revenue associated with works that are not at the margins, works that would have been profitable and so brought forth with less or even no copyright protection. As copyright broadens, it increases the “excess” incentives associated with these preexisting works. As these excess incentives grow, they may, at some point, lead popular authors to substitute leisure for work, and so perversely lead to fewer works from our most popular authors. Broader copyright may thus entail a trade-off between two marginal effects: More original works from new authors along one margin, but fewer original works from the most popular existing authors along a second. If the second effect outweighs the first, then more revenue may lead to fewer original works. Conversely, less revenue may lead to more original works.
While this may seem radically counterintuitive, it also happens to be true. To explore the relationship between copyright protection, revenue, and creative output, I treat the rise of file sharing and the parallel fall in music industry revenue as a natural experiment in radically reduced copyright protection. Using a hand-coded data set, covering songs in the top fifty of the Billboard Hot 100 from 1985 through 2013, and regression analysis, I show that the sharp decline in music industry revenue that paralleled the rise of file sharing was associated, ceteris paribus: (i) with fewer new artists entering the market; but (ii) also with more hit songs, on average, by those new artists who did enter. Moreover, because the second marginal effect was larger than the first, the decline in revenue since file sharing began was associated with a net increase in the number of new hit songs, ceteris paribus. Thus, for the music industry, the rise of file sharing and the parallel decline in revenue has meant the creation of more new music.