The failed Bayh-Dole bargain, 1

Dr. Rebecca McFadyen and Dr. Tara Nealey–both working for an IP law firm–have published an article “Will old Bayh-Dole be taught new tricks?” Me, Dr. Barnett–not working for an IP law firm–thinks they have got it wrong. Consider:

Before 1980 the US government did not have a uniform IP policy with respect to all federal granting institutions that funded research or technology transfer to the private sector.

This is not particularly meaning-full prose. What is a “federal granting institution”? Are we talking federal agencies and the Department of Defense or something else? And what is this about funding “technology transfer”? Bayh-Dole certainly doesn’t anticipate funding “technology transfer”–though universities have figured a way to do this, sort of, sometimes, from licensing income, even if they have to violate Bayh-Dole’s patent rights clauses here and there to do it. No matter, federal agencies uniformly decline to enforce the patent rights clauses in any of their substantive terms, so there is that.

But the article is out there, and ought to be read, regardless of the garbled bits. Let’s put a stick in this mess and swirl it around a bit, and see what we’ve got here.

What was there before 1981, when Bayh-Dole went into effect? We had the Kennedy patent policy, modified by the Nixon amendments, backed up by the 1947 Attorney General report authored by David Lloyd Kreeger, along with the Truman Executive Order 10096. Collectively, these documents present a rational patent policy, as the Harbridge House report on government patent policy found in 1968. What was that policy? Federal contractors could own inventions arising in federally funded work if they had capability and a real non-governmental market, with four exceptions: (1) the agencies intended to develop the technology for general use; (2) health research; (3) government was the primary user or developer (i.e., no non-governmental markets); (4) the contractor supervised the work of others involved in the federal research.

Thus, federal contractors–companies of any size–that had real markets also had the right to retain ownership of inventions made in federal work, except in a few situations. First, if federal agencies are using contractors to develop a technology for general release, then allowing each contractor to hold patent rights in their bits of invention to shut out all others runs against the whole purpose of spreading the work around to multiple contractors. Instead, the government takes ownership of all inventions involved and releases rights so that all the contractors involved have access to the developed technology–whether platform, standard, or product. It makes no sense at all to allow contractors to take out patents and hold those against all other contractors and the rest of the industry if the goal is to release a technology for all to use–water desalination technology, say.

Federal ownership of inventions meant that federal agencies could obtain patents, or not, but that the inventions were to be dedicated to the public–either via the public domain or by non-exclusive, royalty-free licensing. The basis of licensing was to determine that a company was technically qualified and to manage the quality of resulting products (such as to meet a specified standard) to protect the public, as it were. The federal disposition of inventions the government acquired was entirely uniform under the Kennedy/Nixon patent policies.

So was the premise that companies with real non-governmental markets could own inventions arising in federal work. The exceptions don’t make the policy “non-uniform” any more than the exceptions in Bayh-Dole do. Either we still don’t have a uniform federal patent policy (and we don’t), or we had one pre-1981. Work through those options for a bit, and come back when you’ve got it figured.

The key bit in all of this is the health research position. When the government funded health research, the results were to be made available to all, with the government holding the rights to any otherwise patentable inventions. Even here, companies and nonprofits could request a determination of greater rights by submitting a plan that showed how their holding exclusionary patent rights (but for the government’s practice of the inventions) would better serve the public interest that would open access facilitated by government ownership of those rights. Follow?

Pre-Bayh-Dole: companies could own inventions other than in federal public health research, or when the federal agencies were developing the technology for public release, or where there was no non-governmental market, or when they were hired to supervise the research of others. In any of these cases, the contractors could petition to own rights anyway, but had to show that their ownership would better serve the public than open access. Norman Latker, patent counsel at the NIH, worked for years to reverse this policy with regard to health research. Rather than contractors having to explain how their holding patent rights would better serve the public, Latker wanted them to have rights without making that case–replacing that case with a set of public protections that would kick in if the contractors didn’t serve the public better than with open access.

This is just what Latker did in 1968, with the Institutional Patent Agreement program, which was a clever end-run past the Kennedy patent policy (and the Nixon amendment in 1971). Under the IPA, a nonprofit could “contract” with the government for a set of standard patent rights terms that would apply to all subsequent NIH contracting. Those terms required the nonprofit to take ownership of every invention arising in NIH-supported research that the contractor chose to patent, and set up requirements for exclusive licensing of such inventions by the contractor. Basically, the contractor had to try non-exclusive licensing first and show that it didn’t work, or had to study the situation and show that non-exclusive licensing wouldn’t work. In practice, this all boiled down to “we didn’t bother to try.”

For non-exclusive licensing, the only substantive difference with federal practice was the nonprofit’s money interest in the license. This money interest then was split  between the contractor and the inventors, setting up a claim that the money would motivate nonprofit inventors to be more active in contractor-led licensing. The money interest also ramped up institutional conflicts of interest between making technology available and making money by flipping the patent rights to a new private owner (such as through an exclusive license for all substantial rights in an invention), and between regulating research for proper conduct (including published claims and making data available for inspection) and making money by riding the hype around patented inventions.

So you see–the battle zone since at least 1963 has been federally supported health research, the centerpiece of Vannevar Bush’s Science the Endless Frontier. Bush argued that the only patent policy needed for his new National Research Foundation was that the government received a royalty-free license to any invention. And there’s a good argument that such a policy was workable, provided that the government’s license was to practice and have practiced–what the Kennedy policy stipulated, what Latker repeated in the IPA provisions, and what Bayh-Dole (also drafted by Latker) repeats in 35 USC 202(c)(4). To practice means to make, use, and sell an invention. This provision then follows the fundamental idea that Bush puts forward, that the federal government has an interest in the progress of science by means of research. It follows that the government has a legitimate “market,” a governmental market, for the production and distribution of new health products. Any company may participate in such a market, but to do so, the company has to compete with the federal government. In this way, competition in health research is baked into the pre-Bayh-Dole government patent policy.

Even here, the pre-Bayh-Dole patent policy is remarkably uniform. Companies own where they have an actual market, and can request ownership otherwise, having made the case that their ownership would better serve the public, expect in health research, where contractors have to make that case or acquiesce in federal open access. What was not uniform was whether a given federal agency would find a contractor’s petition for patent rights to be persuasive. It’s difficult to see how one could make such s review “uniform” other than by imposing procedures to displace the exercise of judgment. This bit of “uniform” is where Latker took the fight over contractor patent rights, first in the IPA program and then in Bayh-Dole, and then in the regulations implementing Bayh-Dole (which Latker also had a hand in drafting), and then in Reagan’s 1983 Memorandum (which Latker also claimed credit for drafting) and 1987 Executive Order 12591 that overturned the Kennedy/Nixon patent policy.


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