Thursday, February 17, 2011

Fighting Over Green Patents: How To Appease China & India Without Hurting U.S. Business

Tomorrow at the Yale Climate & Energy Congress Symposium, I will be presenting on a Comment I published in the Yale Law Journal last May: Addressing the Green Patent Global Deadlock Through Bayh-Dole Reform. (I also wrote a nontechnical version of this argument for SlateLicense To Green: Can We Have Clean Energy and Patents, Too?) Rather than summarizing the whole argument here, I will just point out the three pieces that I think are novel contributions:
  1. One way to address global concerns about green patents is by changing the way federally funded green technologies are patented and licensed. A number of articles had recognized that conflicts over IP are contributing to the deadlock in climate change negotiations, but none made the distinction between the patent incentives needed for public-sector and private-sector innovation. I examine the justifications for Bayh-Dole patents as applied to green technologies and conclude that in light of available evidence, patents will impede dissemination of most green technologies.
  2. Market segmentation should be used for green technologies. The strategy of allowing strong patent protection in rich countries (to recoup development costs) while allowing broad access in poor countries has been made by scholars, advocates, and universities in the medical context (see, e.g., this policy statement from AUTM and many universities), but I'm not aware of anyone who had extended this argument to green engineering technologies. And market segmentation is even more compelling for green technologies because patent protection is less important for them than it is for pharmaceuticals.
  3. Funding agencies should use their ex ante control over who receives federal grants to influence licensing policies. Several scholars, particularly Professor Arti Rai, have looked at the impact that funding agencies can have on Bayh-Dole reform, but their focus has been on the ex post influence of these agencies on technologies that have already been developed. I argue that agencies could influence university licensing more effectively through their ability to determine who receives federal grants in the first place. For example, the National Science Foundation's "broader impacts" criterion could be used to encompass access-promoting licensing policies.
I welcome feedback, either in the comments or by email. And for readers in New Haven, feel free to stop by the symposium!

1 comment:

  1. It was interesting talking about this to a diverse group of people today. Summarizing the reactions:

    Scientists: Wow, that's interesting, and I've never thought about any of this - I think you're right that scientists should think more about how their research is commercialized.

    Venture capitalists: This makes no sense - exclusive patent licenses are ALWAYS necessary for commercialization of any worthwhile ideal. Why would someone accept a nonexclusive license?

    Research director from a small energy company: This is exactly right - we have encountered many problems from university tech transfer offices that want exorbitant payments for every innovation. And we just want to use these good ideas - we don't need the patents.

    Tech transfer officer: Incentives for tech transfer officers vary widely. Wealthier schools can afford to think about promoting access like this, but many schools are looking to their tech transfer offices to make money.

    I think everyone I talked with was willing to agree that universities (and other recipients of federal research grants) should only patent when the patent is necessary for commercialization, and the main disagreement was over where you draw that line. The problem is that there is very little empirical evidence on this. My starting premise is that patents create deadweight loss and inefficiencies, so we do not want to grant them unless there is evidence that they are needed to incentivize innovation or commercialization. But people with different starting assumptions come to different conclusions.