This paper introduces a new internationally comparable data set that permits an empirical investigation of the effects of patent law on innovation. The data have been constructed from the catalogues of two 19th century world fairs: the Crystal Palace Exhibition in London, 1851, and the Centennial Exhibition in Philadelphia, 1876. They include innovations that were not patented, as well as those that were, and innovations from countries both with and without patent laws. I find no evidence that patent laws increased levels of innovative activity but strong evidence that patent systems influenced the distribution of innovative activity across industries.
Unfortunately, Petra Moser doesn't even appear to have a web page, much less a freely available electronic copy of the working paper; I suppose this is typical of economists and other non-computer-science academics, as I complain in DeLong's comments.
Incidentally (warning, blog post veering off onto radical tangent), how in blazes do people in these other fields get any work done? Luckily, I can get Moser's paper through my university's unlimited subscription, but it's still a pain to set up the proxy when I'm browsing from home; and what about the unwashed masses who don't have access to the resources of a large, well-funded university? Any single paper's reasonably cheap, but when you're doing a literature search you might want to read a few dozen papers and skim many more, from several different journals. The gated-access publishing model makes it basically impossible for an interested layperson to keep up with cutting-edge research in these fields, except through watered-down (and frequently distorted) accounts in the popular scientific press. One might object that cutting-edge research is too deeply technical for a lay reader to understand anyway, but I disagree --- in computer science, at least, a reasonably intelligent and well-educated programmer with an undergrad degree can get the basic gist of all but the hairiest papers. Some grotty technical details might remain out of reach, but there's still some value in reading the paper straight from the horse's mouth rather than secondhand --- the citations alone can make a huge difference in understanding the authors' contribution. I imagine that the same would be true in other disciplines.
And this does not even take into account the incredible value of a site like CiteSeer, which is only possible because most papers are freely available on the web, where they can be easily indexed and cross-referenced regardless of source publication.
On the other hand, a few academics in other fields, like philosopher Nick Bostrom and law professor Yochai Benkler, do seem to "get it": they have home pages, with full papers. Maybe there's hope after all, in the coming generation of academics?
* To quote Lowell Bergman, played by Al Pacino, in The Insider; those with sharp memories may recall that he's speaking of Wall Street Journal.