Public access to federally-funded research results?


The topic of implications of federal funding for research was brought up again recently in this Federal Register notice. The Office of Science and Technology Policy wants to receive public comment on a range of issues related to access to academic publications that were funded by Federal grants. The notice mentions the NIH model

One potential model, implemented by the National Institutes of Health (NIH) pursuant to Division G, Title II, Section 218 of Pub. L. 110-161 ( requires that all investigators funded by the NIH submit an electronic version of their final, peer-reviewed manuscript upon acceptance for publication no later than 12 months after the official date of publication.

and seeks comments on a range of issues regarding how to structure this broader policy, how to make articles available, how to ensure compliance, etc. This notice seems broader than the NSF-specific discussion I wrote about earlier because it appears to apply to all Federal agencies that fund open research.

It seems to me that there are at least two access-related issues here: whether the agencies that funded the research (and perhaps, reciprocally, all other federal agencies) get open access to the published work, and separately, whether the public at large has similar levels of access. This policy stands to have the largest impact on for-profit journal publishers who currently charge libraries and individuals quite a bit of money to access their materials. If the US government breaks publishers’ monopoly on content (perhaps after a small time interval), that may seriously undermine the viability of for-profit academic publishing.

I see this as a positive development: arguably the value that a journal adds rests mostly in the quality of peer reviews it can solicit, and thus for-profit publishers derive profit from reselling content to a community that that community generated. Given that the advertising, distribution, and copy-editing services can no longer be considered legitimate expenses, there is no longer any good reason to incur the high fees charged by these journals. This policy may give open-access and overlay journals a boost, while causing for-profit publishers to reconsider their business models.


  1. The public at large is the agency that funds this research. The federal agencies merely have the important task of deciding how those funds are distributed. For this reason, and many others, open access for the public as a whole makes sense.

  2. By that argument, anything produced at a public university should also be included in the same category. Furthermore, if public money is used by private universities (e.g., scholarships), then their work should also be in the public domain. This is beginning to sound a lot like GPL for research.

  3. A better analogy would be the patent system. The patent system was set up to benefit both inventors and the public. It benefits inventors by providing a mechanism to protect their intellectual property and benefits the public by requiring that the text of all patents be made public. All patents are freely available to be read by the public at large. Of course, access to the text of the patent does not mean that the public at large can use the invention for free.

    I see many upsides, and few downsides, to requiring that the results of publicly funded research be made public. This requirement does not mean that the public then has the right to use it for free any more than in the case of patents.

  4. When I have an idea in the shower, it’s simply not clear who paid for it. Suppose, like us, you get a mix of revenue from commercial customers and SBIR grants from NIH.

    We use the SBIR money to develop software that we can then sell to customers. That’s the whole point of the SBIR program — to fund tech transition from the lab to the marketplace. We explicitly get to own all the resulting intellectual property. In our particular case, we release royalty-free software with source and publications. And while that’s attractive to the funders, it’s not an SBIR requirement (other than that our publications funded by NIH follow the NIH rules for open sourcing).

    Tech transition is also the point of DARPA (I think), but it’s not how their business model works. When you get DARPA funding, they require “government use” rights for software developed on their nickel. Then it’s not clear exactly what the government owns. A typical strategy is to charge research and prototype development to DARPA, while charging any commercial development to venture capitalists or paying customers.

    Of course, everything’s negotiable. It really has to be if a company comes in with a base of existing software and then apply it to a DARPA problem. The government doesn’t immediately own all of IBM’s speech recognition work, for instance, just because they fund IBM as part of the GALE project. So it’s not quite GPL for research.

  5. I wonder if it is possible to define a cleaner separation of federally-funded research (the results of which should arguably be available to everyone) and privately-funded development based on those research results.

    Thus in the SBIR case, the government funds company X to develop some idea. The company makes available publicly the software/ideas/etc. that were developed using this money; while they retain considerable competitive advantage in time and in the understanding of the problem, they do not retain a monopoly. After all, the goal of SBIR is not to fund a specific company, but to stimulate activity in a market. Funding specific companies is just a means to that end.

Comments are closed.