Kal Raustiala and Chris Sprigman, Freakonomics.com:
As Mark Twain once wrote, there are three kinds of lies: lies, damned lies, and statistics. However true that may be in general, statistics can be particularly tricky when they are used to assess the effects of IP piracy. Unlike stealing a car, copying a song doesn’t necessarily inflict a tangible loss on another. Estimating that loss requires counterfactual assumptions about what the world would have been like if the piracy had never happened — and, no surprise, those most affected tend to assume the worst.
This is a good article that debunks some of the numbers floating around regarding the economic extent of piracy. That is not to say digital piracy’s impact is negligible, rather the effect has been exaggerated by those affected.
I am certainly not pro-piracy, but the notion that pirated content is a direct loss to the content provider is slightly fallacious. That is to say, people who pirate content would generally not purchase in the absence of piracy as a means of procurement. Meaning, if someone steals a $2.99 app or a $15.99 movie, it is questionable to say those are equivalent harms to the copyright owners. There are exceptions, however. For instance, when mobile applications require the use of servers to store and transfer data, pirated versions cause a substantive harm to the owner. The pirated copy is using resources that aren’t reimbursed through the sale of the app.