Folks want to repeal Bayh-Dole–and that would be good–but Bayh-Dole is like a shield and folks still have to get at what motivates corrupt practices under Bayh-Dole. Bayh-Dole does not require corrupt practices–it just creates the conditions that make it harder for anyone to act to stop them.
Here’s a typical tweet:
I’d say to President Trump, “Repeal the Bayh-Dole Act.” That act gave government workers the right to patent their discoveries, so to claim intellectual property for discoveries that the taxpayer paid for.
The president cannot repeal laws. That’s for Congress. But the president can issue executive orders that alter regulations that implement Bayh-Dole. For instance, an EO could require federal agencies to enforce the patent rights clause, or use an alternative patent rights clause for all research and development directed at public health, or relieve federal agencies of all Bayh-Dole compliance (since they so obviously don’t do it) and shift that to a federal agency that does not award grants and contracts for research and development. That’s stuff a president can do. Even if the president leads an effort to repeal Bayh-Dole (which would seem to be a low priority at the moment), one is left with a huge hole–what will fill it with regard to inventions made in federally supported work? That hole will take years to deal with. And there’s the damage caused by Bayh-Dole practice, especially in the universities and nonprofits. That may take decades to unwind and rebuild.
Bayh-Dole does not give government workers any right patent their discoveries. US patent law gives inventors the opportunity to patent their inventions. For federal government employees, this right goes back to U.S. v Dubilier (1933). That case involved two government scientists who made and patented an invention relating to radio receivers, with approval of their supervisors to use government resources to do it. Here’s a bit of the Supreme Court ruling:
Where the contract of employment does not contemplate invention, but an invention is made by the employee during the hour of his employment and with the aid of the employer’s materials and appliances, the right of patent belongs to the employee, and the employer’s interest in the invention is limited to a nonexclusive right to practice a “shop right.”
These principles are settled as respects private employment, and they apply also as between the United States and its employees.
The Supreme Court makes the situation clear:
The proposition that anyone who is employed by the United States for scientific research should be forbidden to obtain a patent for what he invents is at variance with the policy heretofore evidenced by Congress.
If public policy demands such a prohibition, Congress, and not the courts, must declare it.
In a way, Bayh-Dole is one bit of the response by Congress to Dubilier. Repealing Bayh-Dole, then, does not end up prohibiting inventor control of inventions–rather it removes one bit of inventor protection and lots of bits of executive branch effort to make Bayh-Dole appear to be a law that voids the rights of inventors working under government contractors–and that despite a later Supreme Court ruling, Stanford v Roche (2011) that insists that Bayh-Dole can do no such thing. Asking for repeal of Bayh-Dole to get at government worker ownership of inventions made where the worker is not employed to invent is barking up the wrong tree. The barking is good. Picking a tree is good. Having a plan for what happens if you get a squirrel also would be good. Having all the goods together, then pick a better tree.
The federal government response to Dubilier is Executive Order 10096 (1950), which asserts the federal government’s right to take title to inventions made by government employees–or at least require a license. Contractor and contractor employee rights in inventions made under federal contract were both a matter of statute (e.g., Atomic Energy Act, NASA Space Act) and of federal contracting, rationalized and made uniform in the Kennedy patent policy (1963). Kennedy’s policy was modified by Nixon (1971) and codified (1975) as the Federal Procurement Regulation, but the main points remained the same until Reagan made a substantial revision in 1987. That, combined with replacing the Federal Procurement Regulation with the Federal Acquisition Regulation in 1984 made a total mess of federal invention contracting policy.
Under the Kennedy and Nixon patent policies, companies with real markets and products–that is, they operated in non-governmental markets–could choose to own inventions made under federal contract,
(i) except for contracts directed at public health, where they had to make a case that their dealing in patent monopolies better served the public interest than did open access. In many cases, government contractors who could own inventions chose not to–for whatever reason (no market, or patents were not helpful, say)–and the government then might patent these inventions and release them for all to use. In effect, the government used the patent system to publish inventions (there’s that “promote the progress of the useful arts”).
There were a few other areas where the Kennedy protocol for companies applied:
(ii) where a federal agency was developing a product for release to all it made sense that the federal agency could take ownership of any inventions involved from all the contractors involved, and then when the product had reached the point of practical application, all contractors–and everyone else–would have access to the product and the patent rights. In effect, the federal agency created a de facto technology standard by aggregating patent rights that otherwise might be held by each of the contractors involved, making it awkward if not impossible for anyone to use the product except for government purposes. By aggregating rights, the federal agency prevented any one contractor from disrupting access by any other contractors to the use of any invention made in the distributed work. This approach worked well especially for the Department of Agriculture and the Department of the Interior. The thing to contemplate is why the Public Health Service and the NIH have not used this approach in the development of health treatments.
(iii) where regulations required public use, or the invention was directed toward where there would be such regulations. Again, it makes sense that if a federal agency dictates what the public must do, then it is difficult to leave the supply of what’s necessary for compliance with a single company holding a patent on what was developed for a public purpose–not only is there a problem with a sole source but also with monopoly pricing that takes advantage of the requirement that the public must use the invention. If we walk this back, anticipating such a result, then we might see that the federal agency choice of contractors to do the work to invent target technology ends up granting those contractors the benefit of the force of law to establish their market. All this runs against the role of a public agency issuing regulations with the force of law. Of course, from a favored company perspective, such an arrangement would be pretty sweet. And if one wanted to take risks, then buying out contractors for their inventions with a bet that those inventions might some day be in the way of federal regulations could also be an attractive speculative play.
(iv) and where the contractor was engaged to supervise the work of others. In such a case, the contractor gains the benefit of access to other contractors’ work–and talented people might invent in such circumstances, as they observe what others are doing. But they supervise on behalf of the government–so why should they end up with exclusive control over patent rights, and not the federal government? All makes sense, no?
For nonprofits–like universities and their research foundation fronts–and for research-only companies that had no real products and markets but lived on federal research contracts, they always had to make a case for holding patents as better for the public than open. When the fed govt took title to inventions, it generally made them open to all. That’s Kennedy patent policy. Fed agencies would not sue citizens for using inventions made in publicly supported work. Or take a financial interest in those inventions. Or play favorites.
The thing that rankled the nonprofits–especially the patenting foundations that had sprung up, emulating Research Corporation and the Wisconsin Alumni Research Foundation–was the hassle of making a case for their patent exploitation of each research invention. Meanwhile, in the early 1960s, pharmaceutical companies boycotted federally sponsored inventions in health care because of their fear of patent “contamination.” The federal government could claim privately made inventions held by a pharmaceutical company if the company took a license to a federally supported invention and allowed the government to have an interest (ownership, license) to anything developed that enabled that invention–whether or not federally funded. Even if an invention was developed entirely with private money, the company could not be sure that the federal government would not claim an interest in it.
The NIH worked around the problem by reviving its Institutional Patent Agreement program in 1968. Under the IPA, the NIH contracted with nonprofits to require them to take ownership of any invention made under NIH contract they chose to patent. The nonprofits then could license exclusively to pharmaceutical companies without federal government claims to related inventions and follow-on inventions.
There’s your first patent pipeline to pharma. The IPA program ran until 1978. NSF joined in 1974. The IPAs were shut down when a review showed they were ineffective and contrary to public policy. Nearly all the nonprofit licenses were exclusive. But only 4% of inventions were used. Given that most of the inventions involved were biomedical in nature, the licensing rate for the IPA was worse than what universities claimed as their licensing rate for non-federal inventions (25% to 30%) and the rate of federal agency licensing for their biomedical inventions (23%). The IPA was ineffective, by comparison.
When the IPAs were shut down, Norman Latker, the NIH patent counsel who created the IPA program, started drafting Bayh-Dole, which was signed into law in 1980 and became effective in 1981. Bayh-Dole restored the practices of the IPA program but also authorized federal agencies to grant exclusive licenses directly, increasing the patent pipeline to pharma. In an odd twist, nonprofits then *had* to patent if they wanted open access. I know–it makes no sense. But they had to pay $10K+ (the typical cost of preparing and prosecuting a patent application) to give everyone access because if they didn’t, the fed agency could play favorites and license exclusively, cutting everyone else out. Not that many nonprofits chose to offer non-exclusive licenses–but it was clear that if they did not patent, then federal agencies could shut off public access by means of exclusive licensing or sitting on inventions until–if ever–an exclusive licensee showed up.
By making it difficult for contractors to release inventions and still follow the regs, Bayh-Dole tipped practice to dealing in patent monopolies. That put federally supported inventions behind patent paywalls until a favored company came along to sweep up rights. if ever.
But Bayh-Dole is not the law that lets federal employees receive royalties. The law that does that is Stevenson-Wydler (1980). See 15 USC 3710c. Like Bayh-Dole, Stevenson-Wydler anticipates that folks dealing in patents made in federally supported work will license those patents. That’s what a royalty is–any consideration for a patent license. It’s not income arising from using an invention–that would be what Bayh-Dole calls “income earned by the contractor with respect to subject inventions” (35 USC 202(c)(7)(C)). Thus, if a contractor is a company, and the company makes, uses, and sells product based on an invention, it does not owe the inventor anything under Bayh-Dole. But for some reason, if the contractor is a nonprofit, then the contractor must share royalties with the inventor (Bayh-Dole–no specification of how much is to be shared) or if a federal agency holds the patent, then inventors must get a share of licensing income. There’s nothing in either law that makes the institutional holder of a patent license any invention, nor aim to maximize royalties.