The scholarly consensus is not unanimous, however. Michel notes that Oberholzer and Strumpf (2004) found no correlation between file-sharing and P2P use in 2002, and a more recent paper by a Harvard student found that file-sharing benefited more obscure artists.

The dominant impression gained from reading these studies is that finding accurate correlations between file-sharing and loss of revenue for the music industry is tremendously difficult. Michel points out, for instance, that his data would be better if it focused only on broadband users—but such data was not available for the time period he wanted. And then, of course, having a broadband connection is not the same as using it extensively, or even for using it at all for P2P. And of course it would be nice to know exactly how many copyrighted songs each particular household downloaded using such software, but this number too is almost impossible to get.

Despite the sophisticated regression analyses and economic modeling done by these authors, the initial data points are not ideal starting places for this sort of analysis. And academic papers are unlikely to sway those who have either made up their minds on the issue or simply want to justify behavior they think think should be legal.