Learning from Latker’s 1984 “Federal Initiatives for Innovation” Talk

In 1984 Norman Latker, who as NIH patent counsel drafted the Bayh-Dole Act on the sly, gave a talk (“Federal Initiatives For Innovation“) to the American Intellectual Property Association. At the time, Latker worked for the Department of Commerce, and Sen. Dole got Bayh-Dole amended that year to shift control over its regulations to Commerce so Latker could continue to work the law into something he wanted it to be.

In his talk, Latker makes a case for contractor ownership of inventions made in federally supported work. But the case simply doesn’t hold up. And in failing at his logic, Latker reveals deeper problems with federally supported research, the administrative concept of “technology transfer,” and American industry that any hoo-haw about Bayh-Dole obscures.

Latker argues that the U.S. has been slipping in its dominance in the development of new technology compared to the rest of the world. There’s a growing “loss of U.S. markets” for technology products, and that’s bad. So the federal government is launching programs to turn things around. After quickly running through tax law changes to encourage private investors to get involved with new technology (why?–more of that later), rolling back antitrust law to permit industry research and development consortia, and more research parks, Latker turns to “commercialization of federally funded technology.” All these programs won’t work, if inventions are not commercialized:

Fifty percent of all the R&D, 70 percent of the basic research, and one-sixth of all U. S. scientists in federal laboratories are supported by federal funding. Right now there is more evidence that the results of that research are being used by the Japanese than by the United States. This seems to be confirmed by the Japanese complaint that S. 2171, the Dole Bill, which gives title to federally funded inventions to contractor is an attempt to restrict their access to our technology (which basically it is). (Laughter)

The “Dole Bill” that Latker refers to (S. 2171) is a bill introduced in 1983 by Sens. Dole, Laxalt, and DeConcini that would have extended the Bayh-Dole Act to all contractors by adding a new Chapter 19 to Title 35:


The proposed chapter specially calls out the extension of only one of the purposes named in Bayh-Dole’s statement of “policy and objective”:

it is the further policy and objective of the Congress to ensure that all inventions made with Federal support are used in a manner to promote free competition and enterprise.

It is odd–what gets emphasized is not the U.S. manufacturing requirement (which is restricted to exclusive licenses to use or sell in the U.S.), and not maximum participation of small businesses in federally supported research and development. Just this free competition bit. And, according to Latker, free competition was not his purpose in extending Bayh-Dole to all contractors–locking out the Japanese from free access to U.S. based research was. So, it’s odd, isn’t it?

In S. 2171, “Practical application” is expanded to include “available to the public on reasonable terms or through reasonable licensing arrangements.” By proposing this expansion, Latker makes evident that the Bayh-Dole definition of “reasonable terms” does not include reasonable licensing arrangements. That’s at least interesting. The Kennedy patent policy does lay out a consideration for both–“reasonably available” to the public if developed to the point of practical application or licensed on terms reasonable under the circumstances if practical application has not been achieved.  Bayh-Dole just has “reasonable terms” in the definition of practical application and does not consider terms in licensing other than at march-in. In Kennedy, licensing on reasonable terms–non-exclusive licensing, essentially–was a way to preclude a need for march-in. In Bayh-Dole, licensing on reasonable terms is set out as the remedy available to the government once it does march-in.

S. 2171 also replaces “to practice and have practiced” in the government license with “make, use, and sell the subject invention throughout the world by or on behalf of the United States.” That replacement makes clear that Latker, at least, thought “to practice and have practiced” might be the equivalent to “make, use, and sell” or “all substantial rights in an invention”–just as the Nixon patent policy had done in amending the Kennedy patent policy in just this same way.

And S. 2171 makes express that federal agencies would have the right to “waive all or any part of the rights of the United States” provided to the federal government under S. 2171. That is, federal agencies would have been expressly authorized to waive, without process, any public protections–government license and march-in especially–not only for non-small, non-nonprofit organizations but also for those small and nonprofit organizations, but have to go through a convoluted process to pull back from the default what should be rights of the United States. That’s also an interesting administrative move: Bayh-Dole offers up an apparatus of public protections–U.S. manufacturing, limitations on patent property rights, a license–an undivided right, essentially–in any subject invention, and march-in. This public protection apparatus is placed to balance the federal government’s limitation on having an ownership interest in such inventions. It would be appear to be a bargain that the public might rely upon. But no–S. 2171 would make all those public protections a matter of agency whim. An agency could decline to assert any part of the public bargain that would have allowed any contractor to have exclusionary rights in inventions made in work identified as being in the public interest and so receiving federal funding.

What to make of such a waiver? It would appear that such a waiver is implicit in Bayh-Dole. Beyond the express provisions of waiver, such as for U.S. manufacturing, it would appear that federal agencies are free not to enforce any other parts of the standard patent rights–with or without notice or process. Perhaps the only way to preserve Bayh-Dole’s pretense of “uniformity” is if all federal agencies choose not to enforce any of the law’s public protections. If that’s the case, then it explains why federal agencies focus on the meaningless-to-innovation administrative requirements of the law–proper form and timing of an invention disclosure document, say–and not that an invention has been developed more quickly and better for public benefit than might have happened with open access.

The point of the waiver is to allow a federal agency to up the “incentive” for private development when–apparently– i) public need is not sufficient to create a market opportunity; ii) a commercial opportunity with competition and standards based on open research is not sufficient; and iii) exclusionary control via patent but with public protections is not sufficient. We are left then with Bayh-Dole “working” only when there are not public protections, and no guidance whatsoever to establish “uniform” practice at federal agencies with regard to enforcement or waiver of public protections. Bayh-Dole does not even require federal agencies to ask about compliance or to audit for compliance.Bayh-Dole is said to “work” only if a contractor has exclusive control and the public and government must purchase product (if there ever is any) on whatever terms the contractor sets–outrageous, with limited features, or otherwise. Thus, the present huffing and puffing by self-proclaimed Bayh-Dole advocates that march-in was “never intended” to address unreasonable pricing. If march-in was never intended to be used for any reason, then obviously march-in was “never intended” to address any particular reason, such as unreasonable pricing. Latker drafted a law that appeared to require one thing yet put the control of that thing in federal agency hands–for Latker, the NIH. Latker wrote a law authorizing Latker to do pretty much whatever he wanted, at the cost of administrative procedures and administrative compliance spectacle that he could waive on a whim.

S. 2171 set no boundary for when a federal agency would refuse to waive and how to differentiate such refusal from all the other times that some federal agency has waived public protections. In effect, S. 2171 reveals that Bayh-Dole’s public protections are really toothless political spectacle. The only public bargain in play is an appeal that chosen companies produce products however they may, based on federal research, on whatever terms they wish to set.

A much simpler law could be drafted that makes such a thing clear. If federal policy is really to go this way, then we should have that. Repeal the public protections in Bayh-Dole. All of them. The federal government takes no interest in inventions arising in federally funded work. One might then ask why the federal government should have any ownership interest in inventions made by federal employees, on the same reasoning. The ownership apparatus of EO 10096 and the related licensing apparatus for federal agencies in Bayh-Dole would both be disappeared.

There’s no point in a federal policy that claims federal agencies should not deal in federal contractor patents but claims that federal agencies should deal in patents on inventions made by federal employees. The argument does not rest on a distinction between independent contractor and employee (patent law does not give employers any special privilege, as employers, to own employee inventions). The argument rather rests on whether the federal government is ever a suitable owner of inventions for the purpose of exploiting exclusionary rights. If it is, then what differentiates ownership obtained via research contracting from ownership obtained via the work of federal employees? The best Latker could come up with was that technology transfer was better performed by an organization “closer” to the inventor than any federal agency.

While such an argument might appear plausible, in practice there’s little to support it. Organizations close to inventors–employers even–may be bungling messes, may have expertise and plans that run counter or skew to any given invention, may be poorly located to serve the companies and industries that might benefit from access to such inventions, and may have so much politics and process and risk aversion going on that even in competent forms, they have no chance to transfer anything, close to an inventor or otherwise. Only in a fantasy world created for political purposes does practice close to an inventor uniformly result in more effective technology transfer.

Much of the changes sought by the “Dole Bill” ended up in the 1984 amendments to Bayh-Dole, but not the extension of the law to all contractors. Instead, Latker drafted for President Reagan the 1983 Memorandum that effectively revoked the Kennedy/Nixon patent policy and replaced it with almost-Bayh-Dole, Bayh-Dole but for its preemption of federal statutes and lacking itself statutory standing. For interpretation of the Reagan policy, one looks not to Congressional intent but to executive branch intent. There’s a difference there.


This entry was posted in Bayh-Dole, History, Technology Transfer, Vannever Bush and tagged , . Bookmark the permalink.