One other point about the purported “lack of evidence” that software patents harm innovation. I’m probably more sensitive to this type of argument than most because I’m also working on a paper on eminent domain abuse, and you sometimes see precisely the same style of argument in that context. People will argue “sure, eminent domain sometimes screws over individual landowners, but there’s no evidence that it harms the economy as a whole.”
There are two problems with this sort of argument. First, as I said before, it’s not obvious what “empirical evidence” would look like. Eminent domain abuse occurs in almost every state in the union, and it would be extremely hard to set up a good controlled experiment.
But the more fundamental point is that individual examples of injustice are themselves evidence that something is wrong. When city governments steal peoples’ land to make room for a shopping center, that is, in and of itself, evidence that eminent domain abuse is harmful. If we can pile up enough examples of such abuse, that’s evidence that the system is causing harm even if the harm doesn’t show up in national GDP statistics.
Similarly, I don’t think anyone would seriously claim that what happened to RIM, or what’s happening to Vonage, is just. So those are two data points in support of the thesis that the software patent system is screwed up. Here are 26 more. When you’re talking about issues like this that aren’t susceptible to clear-cut quantitative measurements, the plural of anecdote really is data—especially when the anecdotes are so lopsidedly in one direction.