GAO reports Bayh-Dole is a do WTF you want law

Bayh-Dole is a do WTF you want law. Don’t take my word for it. You don’t even have to go read the law to see what I have been talking about. Here, check out this report from the GAO, for 1998.

The repeated theme of the report is that “we talked to someone” and “they think Bayh-Dole is doing wonderfully” along with “but they don’t have any way to check that their thinking is accurate.” And of course, the university thinking about Bayh-Dole is mostly inaccurate, except for one key point–universities can do pretty much whatever they want to, as long as they insist they are complying with Bayh-Dole. Here’s a major heading:

Check out that left column summary: “The administration of the Bayh-Dole Act is decentralized and relies heavily on voluntary compliance by the universities.” Of course, the Bayh-Dole Act applies to patents on inventions made with federal support and to federal agencies, and dictates how federal agencies must contract for rights to inventions made with federal support. But what happens if the universities haven’t bothered with compliance? Nothing, apparently. Universities haven’t bothered with the (f)(2) written agreement requirement, routinely assign inventions under the cover of exclusive patent licenses, fail to (knowingly) require their assignees to accept the nonprofit patent rights clause, and spend the income from licensing and from other exploitation of subject inventions however they wish, disregarding the standard patent rights clause.

The results: university inventors are disenfranchised under the color of the law when they should be protected and given a primary role under their own Bayh-Dole-provided inventor patent rights clause; universities propagate patent monopolies when they should be breaking up patent rights to promote free competition and public benefits on reasonable terms; and universities use the income they receive any way they wish rather than deducting only the costs to administrate subject inventions and the rest for scientific research or education. As a result, universities have accumulated over 50,000 US utility patents with federal funding notices–and licensed hardly any, and of the hardly any, even fewer have resulted in a commercial product based on the licensed invention.

But the GAO had to ask. Department of Commerce?

But Commerce does not maintain any overall Bayh-Dole database.

That was 1998. We are 22 years later, and we are no better off. What about the other seven agencies interviewed by the GAO:

Yeah, it does read “had made analyzed.” None of the agencies had made analyzed the impact, but were “pleased” with how Bayh-Dole was working. Then they spouted aspiring thoughts about what Bayh-Dole should be doing. Whoop. No data here! But the law is kick-ass wonderful.






The agencies “relied on the universities to ensure that all federally funded inventions were meeting the requirements of the law and regulations.” Inventions are to meet the requirements! Hoot! That’s a pretty indirect way of saying that no one is supervising how the universities (not the inventions) are meeting the requirements of Bayh-Dole. The agencies are asleep at the wheel. Do WTF you want. Just don’t make us wake up.

Now look at what the GAO reported from their visit to ten top universities:


That’s non-compliance. The definition of “subject invention” is a matter of federal patent law. It’s not whatever a university decides to call a subject invention. There’s no “presumption” possible. The university has to demonstrate that the invention meets the requirements for “made in the performance of work under a funding agreement.” But does the GAO raise its eyebrows? No.

Or this:

“Most” of the universities have “developed information to inform researchers.” That means some haven’t. But the standard patent rights clause requires universities to have such programs (37 CFR 401.14(a)(f)(2)):

The contractor shall instruct such employees through employee agreements or other suitable educational programs on the importance of reporting inventions in sufficient time to permit the filing of patent applications prior to U.S. or foreign statutory bars.

Not a blink from the GAO. Now for the single most important provision of Bayh-Dole, 35 USC 204, the requirement to require exclusive licensees to use or sell in the US to agree to use product manufactured substantially in the US. 35 USC 204 asserts that it takes precedence over anything else in the Act, and Bayh-Dole takes precedence over every other statute but for Stevenson-Wydler. What does the GAO find?


First, the universities are, apparently, granting exclusive licenses to make, use, and sell. That might be an exclusive patent license, but Bayh-Dole prohibits nonprofits from assigning subject inventions without also requiring the nonprofit patent rights clause–and exclusively licensing all substantial rights–make, use, and sell–assigns the invention, whatever fussiness a university might make about keeping title to patents. And of course, universities have “no practical method” for enforcing US manufacture. Do WTF you want.

And the small business preference required of non-profits by 37 CFR 401.14(a)(k)(4)?

Here’s the start of the provision:

It will make efforts that are reasonable under the circumstances to attract licensees of subject invention that are small business firms

The universities go, “aw, shucks, companies rarely compete for the right to license.” But the requirement is that universities make an effort to attract licensees (plural) of “subject invention” (singular, typical Bayh-Dole garble). It’s not that universities license exclusive to small businesses more than to others, it’s that for each subject invention, the university will make efforts to attract small businesses–and perhaps then license to them non-exclusively. We might observe, then, that a non-exclusive licensing program directed at small businesses is the core of the requirement. But universities don’t bother, because they say they license to more small businesses than large. The GAO doesn’t even bother to tell us whether the universities were talking expressly about subject inventions or what they call “inventions” in general. Under AUTM rules, any license of more than $1,000 is a “commercial” license. All a university has to do is “license” unpatented software or research materials to scores of companies each to run up its count of small business licensees–and that count won’t have anything to do with Bayh-Dole subject inventions.

So, universities aren’t in compliance, the GAO lacks the awareness of the law to notice, and the universities go their merry way.

What then about Bayh-Dole’s effects, given that the universities and agencies ignore the law?

Ho-ho. They ignore the effects, too!

There’s just the AUTM licensing survey, which as the GAO notes, does not break out the federally funded inventions.

Keep in mind that at the time Bayh-Dole was passed, the universities (really, their patenting agents) claimed a commercialization rate of 25%, but for federally funded inventions under the NIH IPA program, their rate through to product was 5%. One would think that to assess Bayh-Dole, folks would want to see that universities were performing at least as well (badly) as they were under the IPA program, which Bayh-Dole was created to replace. But no. No one bothers to evaluate the effects of Bayh-Dole, but still they all insist that Bayh-Dole is “working well.” Unicorns and rainbows time.

Universities confirm that the federal agencies just don’t care about enforcement:

You see the pattern. Party like it’s 1998. Do WTF you want.


This entry was posted in Bayh-Dole and tagged , , , . Bookmark the permalink.