A paper from University of Kansas economist Koleman Strumpf (whose work we've featured here for years) empirically examines the impact of file-sharing on box-office revenues.
The paper, Using Markets to Measure the Impact of File Sharing on Movie Revenues, has a compelling methodology (tracking Hollywood Stock Exchange and Bittorrent stats) and is well-argued, and demonstrates that, as best as anyone can work out from the figures, file-sharing has a negligible (and sometimes positive) impact on box-office take, especially when a good movie leaks a little ahead of the official release, generating some buzz for the film.
File sharing provides a useful laboratory for investigating the economic importance of intellectual property protection. There are two main empirical chal- lenges: overcoming the non-random timing of the arrival date of illicit copies and dealing with low statistical power due to limited sample size. This paper uses markets to address these issues in the context of movies. I show forward looking markets can be used to establish the unobserved counter-factual of how movie revenues would change on any possible file sharing release date, particularly those prior to the theatrical premier. Using movie-level tracking stocks in conjunction with the arrival date of illicit copies, I find that file sharing has only a modest impact on box office revenue.
Using Markets to Measure the Impact of File Sharing on Movie Revenues [Koleman Strumpf] (via /.)