I can believe that, but it doesn't by itself show the full effect of the patent system. The threat of a legal case is important in a lot of places where no actual legal case develops. For example, Stanford's CCRMA computer-music research lab has been funded over the past few decades largely by patent royalties, from licensing patents to synth-makers like Yamaha. IMO that's an example of the patent system functioning largely as intended: R&D group develops something, licenses it to a commercialization group, which develops a product and pays a small percentage as royalties to the R&D group. No court actual court case happened anywhere here, because everyone played by the rules. Yet the possibility of one was necessary, or else there'd be no reason for Yamaha to pay anything at all.
So for a full picture of the effect of the patent system, I think you need to look at total licensing royalties, including both court-ordered and contractually agreed ones.
The value of the civil legal system is entirely about the influence of the possibility of litigation on primary behavior. Obviously any actual litigation is a loss--but the possibility of recourse to litigation can allow transactions to happen in low-trust environments, and thus be a net win.
Similarly, value of NPE's is that they allow recovery of investment into R&D that doesn't lead directly to a product, thus reducing the barrier to investing in R&D.
So that would include the nearly $100 million collected by Forgent in settlements over bogus JPEG patents [1]? Or all of the other NPEs who target the small startup, the app writer, or anyone else they can think of who can't afford a long, protracted legal battle?
Yeah, this article is just part of a complex picture, but it's telling.
So for a full picture of the effect of the patent system, I think you need to look at total licensing royalties, including both court-ordered and contractually agreed ones.