I just want to make clear that I also believe that sharing copyrighted music is a bad thing, just so you don't think that I am some pro-piracy nut.
However, let's concider the source on these statistics. The RIAA and the MPAA. These are lobbying groups for an industry that feels threatened by filesharing. They are a special interest group. Accepting their numbers is akin to accepting tobacco industry statistics on second hand smoke from the 1990s. The studies they fund are equally suspect.
I will take the second paragraph of this testimony as an example:
Piracy is the greatest obstacle the film industry currently faces, costing our industry approximately $3.5 billion annually due to hard goods piracy of DVDs and VCDs alone. Deloitte and Touche estimates that approximately 400,000 films are illegally downloaded every day. CacheLogic, an Internet monitoring group, has estimated that over 60 percent of all Internet traffic in the U.S. is attributable to peer-to-peer usage. In Asia, over 80 percent of all traffic on the web is from P2P. Furthermore, well over 90 percent of all the content on P2P networks consists of unauthorized copyrighted files.
There are several flaws here in these funny numbers.
1. Dollar value: This number is acheived by multiplying number of downloads times the price of the DVD that the item is on. An assumption here is that the everybody who downloads a dvd would purchase it if they could not download it. There is no evidence to support this claim, anywhere. He does not even take the idea that people might rent into account, nor does he take into account the fact that many people may find downloading an easier method for media shift (the practice of putting content that you purchased the right to listen to on another media, such as CD to iPod), something that is expressly allowed by fair use clauses laws in current US copyright law.
2. Percent of traffic being p2p is not related to the amount of traffic that is pirated goods. Bittorrent is the primary method of distribution for Fedora Core 6, which is 3.5 gigabytes at 2800 downloads a day. If you assume that a song is about 4mb, that equals two million songs a day. That is bigger than the traffic for the entire aMule network and we are only talking about one single linux distro. I imagine when Debian etch is released later this month, their bt traffic will dwarf fedora.
3. 90%. That is not even internally consistant in the article. It is a scare tactic. When he sites his evidence later in the article, he admits that the supreme court accepted that "certain" p2p protocols had a very high amount of copyrighted content, I think it would be safe to assume they meant grokster, which is why grokster isn't around anymore. However, the 90% number is not from the supreme court. It is from one of the MPAA studies (assuming he did not make it up since he never cited an actual study). He used the common rhetorical tactic of association by proximity to associate his unsubstantiated number with a body that enjoys recognized authority. In order to distract the reader/listener, he conducts a classic bait and switch and quickly switches to the fact some studies have found pornography on said networks, including "child pornography". Guess what, websites distribute pornography and child pornography, too, yet that does not make it a valid argument to shut down the delivery of http over port 80. The subject is not child porn, it is music piracy.
Do you have any neutral sources, maybe government funded studies or academic studies?