This summer, in the FCC's Comcast Order, three Commissioners made a claim that immediately rang false:
"Although once relegated to serving, in most cases, the savviest Internet users with unsavory or even unlawful purposes, BitTorrent and other peer-to-peer technologies, such as Gnutella, have entered the mainstream."
This claim was not supported by the "supporting" third-hand data cited. All file-sharing programs have some lawful uses, and some closed or filtered programs are certainly used mostly for lawful purposes, (when they are used), but I am aware of no data supporting a claim that the "mainstream" uses of, for example, the Gnutella protocol are now savory and lawful. For example, in Grokster, evidence showed that 97% of the files actually selected for downloading by program users were or were highly likely to be, infringing.
Fortunately, Illinois State University's Digital Citizen Project has been gathering empirical data about how many file-sharing programs are actually used. When I checked back in at its website, I noted that the Project has released another recent study, Dimensions of P2P and digital piracy in a university campus, by Alexandre M. Mateus and Jon M Peha ("Dimensions").
Dimensions reports "results from the first large-scale quantitative assessment of online media transfers based on actual observation of P2P exchanges on a college campus." The authors also report that their data on infringing uses "are... lower bounds on the actual quantities that we want to assess." Dimensions notes two reasons for this conclusion.
First, the study relied mostly upon an Audible Magic appliance to identify infringing file-sharing. For obvious reasons, the database against which it checks files consists mostly of recent, high-value content like popular music, movies, and TV shows. Older or less popular protected works of these types are less likely to be included, and many types of works, including graphic arts, photographs, software and games--and adult movies--are not included at all. The authors thus used other techniques, like examining file metadata, to assess whether non-flagged content was noninfringing or "adult."
Second, the monitoring program was widely and extensively publicized and debated. "It is therefore possible that this public discussion deterred some people from engaging in illegal transfers, thereby affecting our results."
Nevertheless, infringing uses of the monitored protocols were found to be pervasive. Here are some particularly interesting findings:
- "[W]e found no evidence that large numbers of students use P2P for these legal purposes and not to transfer copyrighted material." (p.1)
- "Most users are intensive users of P2P to transfer copyrighted material." (p.21)
- "Some might suggest that there are many people who use P2P for the legal transfer of software, such as Linux, or for the transfer of adult material (which may or may not be copyrighted), but do not engage in the illegal transfer of copyrighted material. However, we found no evidence of this among college students." (p.29)
- "[A]s for the legal transfer of software, the percentage of P2P users found transferring Linux out of those that do not transfer copyrighted media is not statistically different from zero." (p. 29).
In short, analysis of actual data confirms that file-sharing is still mostly piracy and "porn," (used in its loose sense as an unflattering synonym for "adult" movies and images). The admirable efforts of those trying to develop lawful uses of these technologies should not obscure reality: the "mainstream" uses of too many file-sharing applications are still highly unattractive--the unlawful and the unsavory still dominate, spiced, sometimes, with the sort of inadvertent file-sharing that "shared" over 150,000 New Yorker's tax returns with identity thieves and the plans for President Obama's new Presidential helicopter with Iran.
