The head of the patent office had some advice for critics of software patents yesterday: "Give it a rest already." Basically, he made a case that software patents are a positive force because they spur innovation. Tim Lee is unimpressed:
This argument ducks the central question in the software patent debate: do patents, in fact, provide a net incentive for innovation in the software industry? Many entrepreneurs say that just the opposite is true: that the disincentive to innovation created by the threat of patent litigation dwarfs any positive incentive effects created by the ability for a firm to get patents of its own.
Empirical evidence backs this up. For example, in a 2008 book, the researchers James Bessen and Michael Meurer found that for nonchemical patents, the costs of patent litigation began to exceed the benefits of holding patents in the 1990s. Software and business patents were particularly prone to litigation.
More recent research has estimated that litigation by patent trolls costs the economy at least $29 billion per year, and that figure may be as high as $83 billion.
I don't know if these numbers are correct, but I'd add another argument to the mix: We already know what would probably happen if software patents didn't exist. That's because, for the most part, they didn't exist until the early 70s, and thanks to fights between the courts and the patent office, they didn't become common until the late 80s. And yet, the era from the 50s through the 80s was about as dynamic and innovative as you could possibly imagine. Lack of patents simply doesn't seem to have had the slightest effect on the growth of the software industry.
The world is different today, of course. But I see little evidence that software patents are any more necessary now than they were during the adolescence of the computer industry. Rather than spurs to genuine innovation, they've evolved into little more than virtual armaments that big companies use to fight virtual wars with each other. And virtual wars are no better for economic growth than real ones. Honestly, it's long past time for software patents to be put out of their misery and for software companies to focus their attention on inventing new stuff, not wasting countless man-hours of time building defensive patent portfolios with no real-world value aside from providing protection against other companies who are building their own defensive patent portfolios for the same reason. This particular arms race got out of hand a long time ago.