Another day another absurdly broad software patent in the mobile phone industry as we learn about Apple's patent for displaying electronic lists on a mobile device. I note that a lot of my items on this subject come from Apple not because Apple is a uniquely abusive patenter, but because their iOS devices were so genuinely innovative. Having solved the really difficult engineering problem of creating an affordable attractive looking multi-touch pocket-sized telephone with a highly responsive user interface first, Apple also got to be the first to tackle all the totally obvious issues and then start furiously filing for patents. But the policy problem is much larger than any one company or even than merely the mobile devices space.
Somewhat coincidentally—and separate from the policy questions of patent law—I've seen three different studies over the past 10 days or so that were attempting to reach some kind of conclusion about the state of innovation or the rate of technical progress that chose to use aggregate patents granted as the numerator in some equation. This seems like an egregious example of the unreality problem that often afflicts quantitative studies of social science topics. Treating "mathematically rigorous model with crap input data" as some kind of close second-best to "mathematically rigorous model with accurate input data" makes no sense, and it's clearly the case that the volume of patents is determined by the laws and norms governing the patent system rather than the quantity of inventions.