Wednesday, March 23, 2011

Do Patents Disclose Useful Information?

I recently uploaded a revised draft of my current working paper, Do Patents Disclose Useful Information?, on SSRN. I am grateful for all the feedback I have received on this paper so far, and I welcome more—please feel free to email me. Here is the abstract:
Courts often state that patents are justified by disclosure theory, the idea that patents are awarded as quid pro quo for the public disclosure of inventions. Economists have long argued that disclosure theory should be accorded no weight in the design of the patent system because patented inventions would have been disclosed anyway. Even the few legal scholars who dispute these economic arguments agree that, in practice, patents are currently not useful as technical sources for other innovators. This Article challenges all these arguments, arguing not that disclosure theory is a valid justification for the patent system, but rather that the benefits of full technical disclosures are stronger than is generally believed, and that these benefits probably outweigh any incremental loss in innovation incentives. Using results from a new survey of nanotechnology researchers and case studies of individual patents, this Article shows that patents do contain useful technical information that is not available elsewhere. This technical content could be improved, however, in at least three ways: stronger enforcement of disclosure requirements (including through an obligation to respond to good faith reproducibility questions from skilled researchers), elimination of legal barriers to using patents as technical sources, and improved access to patents through peer production. These changes will help defuse the tension between patents and the open culture of science.


  1. As an examiner I can assure you that on occasion patents do disclose useful information to those skilled in the art. I read them, I read the NPL. At least as much as is found in most NPL is found in a patent on the same thing and you can tell a generally useful disclosure from the junk. Not however, I'm in the useful arts, as opposed to your everyday business method or software art. In those arts I'm quite sure that the usefulness of the information presented is much lower and the amount of useful information disclosed is lower.

    However, that isn't to say that many, if not most, applications are not disclosing anything all that useful to those of skill. Generally we get reams of patents for a small tweak on something already made. The majority of these are pretty much useless as nobody else is going to use the tweak anyway, ever.

  2. So, in addition to the suggestions they made in the article, you really need someone to go through and determine which of the patents issuing are generally worth reading to one of skill. That is, if you want to make patents really useful in this patent glutted day and age, you need someone to sift through the noise and deliver the signal. I'm not saying do not allow the other apps because nobody cares about them, but there is definitely a need for people to sift through the patents issued and determine which should be showcased for having disclosed something useful to those of skill. Also, I had to post this many times because your comment system s u c k s.

  3. Hi 6 - Thank you for taking the time to read my article, and I'm sorry for the trouble you had with Blogger's comment system (blame Google, I guess?). You're right that there are probably big differences in the value of the technical information in patents from different fields. My paper focuses on nanotechnology patents, and I should probably be more clear that the results may not apply to business method or software patents. And you are also right that figuring out which patents are useful can take time, though my hope is that this will be easier if the patent literature becomes more integrated with the traditional scientific literature.

  4. "though my hope is that this will be easier if the patent literature becomes more integrated with the traditional scientific literature."

    And how, prey tell, will that happen until someone either writes an entire book showcasing patented techs and people are interested in the book, or the PTO starts making an art-specific publication full of what a few hired experts consider to be the cream of the allowed crop each week?

  5. I've considered writing such a book on my own initiative, but it is a massive undertaking and I have no experience in writing such books so my first one probably won't turn out like a golden swan.

    Also note that even if I do, unless we get the willful infringement nonsense off the books that any impact such integration efforts will have a much diminished effect or no effect at all.

  6. I think it would be fantastic if patent examiners or the PTO started showcasing outstanding patents, but I think the patent literature is also becoming more accessible to scientists as patents become easier to search while searching for scientific papers. For example, I think it is great that Google Scholar now includes patents in searches by default. I discuss other mechanisms (including the willful infringement issue) in Part IV of my paper.