The start is here:
Three sorts of university invention
We can then distinguish three sorts of invention arising in federally funded research at universities: inventive tools, inventive tools that can be sold as products, and articles that can be used only as products. An inventive tool is anything useful for a purpose. In research, the proximate purpose is the research activity itself. Thus, resistive touch materials were developed initially to more easily mark x-y coordinates. As an inventive tool, such new technology has an immediate use–conduct better “research with” the tool. Closely related, others may use the tool to verify the claims made about the research (and thus also examine the tool for its reliability)–“evaluation of.” And others may study the tool to learn how it operates and how it might be improved, extended, or adapted–“research on.”
In terms of the Harbridge House accounting, where an inventive tool is useful for the government’s purposes, but not for much else, then allowing the contractor to retain rights to the tool for private market development beyond research uses makes sense. Where the inventive tool has been developed by government contracts with an eye for that private market all along, then allowing any one contractor to monopolize that market makes much less sense. Even if private market development was not a purpose of the federal research, if the inventive tool can be used broadly in research–then, there too we might expect that the tool should be made available non-exclusively. And this, in fact, is just what the NIH advocates for what it calls “biomedical research resources.”
Follow-through on government research funding
We are left, then with what we should do if all the government does is fund a tiny bit of initial work in an area that has both commercial and public implications, and the government does not bother to pursue either. This is not a situation of “basic” research as Vannevar Bush described it, but rather of research that has a public policy purpose (such as alleviation of suffering) but also runs parallel to companies that make their livelihood in this same “market” and rely on patents to preserve their financial positions. We might say this is a highly specialized set of conditions. Harbridge House identified two agencies that do this sort of thing–HEW and Interior:
Unattended speculative research
When research is supported in “more speculative” areas and then left unattended–no followup–then patents become an issue. Imagine randomly plowing acreage in the middle of nowhere, on the great plains, say–but some of it is good land. Break up the sod, but don’t bother to plant. Erosion becomes an issue. And while you are at it, shoot as many buffalo as you can. Why on earth would the government (or anyone) tear up a place (even with the excuse that it’s research) and have no will to follow up on what’s discovered? This is the mind-bogglingly strange policy decision at the NIH with regard to “basic” research in areas such as medicinal chemistry. Why fund academics at all, when they are already working with industry? Why not fund the drug companies directly to augment their access to new compounds wherever they might find them, even in university labs? Viewed this way, while the PHS position on patents makes sense, the underlying research policy makes no sense at all. Where the PHS did pursue development–in areas such as cancer and malaria–it appears to have done well. Where it has pulled out (even with Taxol), things have ended up in disputes and strangeness.
It was this special combination of the PHS’s determination to fund “basic” research in medicinal chemistry and not generally carry through to support the results that has led to the arbitrary, rather disastrous government patent policy set out in Bayh-Dole. To get there, the university patent brokers and drug companies made a distinctive situation appear to be the general case. They kept the idea of urgency, but shifted it to private development because the PHS may have been fighting a war, but only in the abstract and never taking things to the front lines. The patent brokers and drug companies also kept the idea of “high risk” but transferred that to inventions rather than to what Harbridge House noted was the case of leaving inventions in the ownership of universities:
And here we get to the nature of the circumvention. If universities hold patent rights, it is in the nature of the data to say that their licensing efforts are “high risk.” Another way to put it is: “more likely to fail” than if the rights are held by a company that already has commercial experience in the area of the invention. Thus, why should universities hold any patent rights in potentially therapeutic compounds. What value at all did UCLA add to the Xtandi patents? Nada, other than filing the patent applications. Wouldn’t it be much better to direct the assignment of such inventions to commercial firms outright, regardless of whether they have contracted for the research that did the inventing? Why not treat therapeutic compounds as we do the radio spectrum and license it out to bidders who then may hold a licensed monopoly? Take title to the invention, file patent applications, re-issue the invention to a capable company, require the company to pay the inventors a share of the action (how about 0.25% of net sales if ever there are any or a lump sum cash amount) and be done with it?
The university patent broker as a crappy compromise
We get, then, to the university patent broker role. Here’s Harbridge House on the idea:
If the PHS (and its successors in name) won’t change its research ways or its patent policy, then despite universities’ lack of licensing success, they can resolve the problem by allowing the PHS to continue to fund academic research with no follow-on support or interest (what a crappy sponsor–loaded with money but disinterested in seeing anything through to completion, worse than the military, incapable with regards to the frontiers of science, plowing up the plains and shooting the buffalo, but with a policy that makes a virtue of leaving it all to blow away and rot.
The university patent licensing shop then is recruited as a “buffer”–as an intermediary that funnels PHS inventions to drug companies, restoring for federal funding the monopoly practices that the industry relies upon, and saving appearances that the public money is not merely a subsidy because the companies pay licensing fees to the universities, and for a successful drug those fees will be in general far greater than the amounts that the government paid to support the exploratory research that produced the patentable invention. As far as it goes, it makes some sense–it’s a compromise in a good political sense of something crappy all the way around but better than a chronic standoff.
Why, then, did the political compromise that was framed initially as the revived IPA program in 1978 have to be expanded to government-wide, where it was reviewed, the expansion effort stopped, and the IPA program itself shut down? I will venture some thoughts in the absence of smoking documents. First, the IPA program was criticized as failing to extract sufficient payments from drug companies. The companies were getting sweet-heart deals from the university patent brokers. Good money for the universities, but paltry. That’s the same argument that private investors made when they took over the UCLA patent licensing office with a shadow office of their own. But the IPA program was also a ruse–it had the appearance of a public covenant, that patents made with federal support in subvention research were different in character than other patents, and should be used not only to speed development of commercial products but also to limit the profits from those products in areas of public health. But the IPA in practice did no such thing. The apparatus of public interest was for show and didn’t operate, just like the public interest apparatus in Bayh-Dole is never enforced.
The effort to make the IPA program government-wide–realized with Bayh-Dole–was also an effort to protect the ruse. By focusing on patentable inventions, attention was deflected from the arena of public health; by making all inventions appear to be “early stage” requiring vast private sums, the pharmaceutical operating conditions would appear to be the same as those faced by all industries; by emphasizing the idea that all inventions were “early stage,” exclusive licenses looked reasonable as defaults. Anyone who wished to impose special conditions on the pharmaceutical industry’s use of federally supported inventions–such as to regulate pricing–would have to do it for every industry.
If we went at things another way, we would repeal Bayh-Dole except for medicinal chemistry and other disease and injury-specific inventions. (Or, put another way, we would enforce Bayh-Dole except for medicinal chemistry.) For medicinal chemistry we would develop a federal patent policy that directly addressed the problem that Harbridge House identified sixty years ago–what to do with a federal agency that tears up the plains and shoots the buffalo but doesn’t follow up to plant anything or pack the meat? One response is to rethink its allocation of research money; another is to rethink its role as a research patron–anyone knows that lousy patrons make for lousy work; yet another is to try to lipstick the pig and create a patent policy responsive to the snout we have.
If we follow the lipstick option, acknowledging that we are hacking a defective research policy architecture, then we might create an option under which drug companies can obtain exclusive positions on federally supported inventions and rather than paying out licensing fees (or even having to negotiate licenses), they are assigned (or re-issued, as it were) patents in exchange for, say, reasonable terms for public benefit. Those reasonable terms might include moderate pricing after recovery of their total allowable development costs (which they would then report), where moderate pricing would be set as some amount over the cost of manufacture, similar to what a generic drug might sell for in an environment in which there are multiple vendors of the drug. Reasonable terms might also include making the manufacturing information and any required proprietary rights available for use by generic manufacturers if the company ends its own manufacture of the drug.
Would drug companies take such terms? Right now, they like what they have got and show no signs of being open to changes in Bayh-Dole–the non-enforcement is working out just great. Thus, one returns to the other two options–either the federal government should get out of the business of supporting public interest research that it will not follow up on, or it should shift its research policy to follow up on the research it does support. Otherwise, the patent policy ought to be the one that Vannevar Bush proposed–leave the inventors alone to decide what to patent, and give the government a royalty-free license to do what it will for government purposes.
At present, the government-purpose license right that the government receives in each subject invention is itself outside of Bayh-Dole. 35 USC 207 and 209 concern only inventions that are federally owned. There’s no guidance whatsoever on inventions that the government does not own but holds a broad license to practice and have practiced. If one wanted to gain some leverage on the pharmaceutical industry’s happiness with Bayh-Dole, then establishing a federal mandate to use the government’s broad license to practice and have practiced each subject invention in the area of biomedical practice is the place to start. The federal government (and state governments, and even municipal governments) could license rights under patents the government does not own for contractors to make, distribute, and sell products within the scope of governmental authority and purpose.
The government license is the essential piece of the bargain. The antitrust violation that keeps drug prices 10x higher than one would expect is the result of government inaction on the rights it holds. That is, it is the federal government itself that creates the monopoly position, not by issuing patents on federally funded inventions and failing to enforce the public covenant apparatus, but rather by failing to act on its own right to serve the public through the government-side market for new therapeutic drugs. Exercising that right will change the discussion and perhaps we’d see the drug industry more than ready to consider alternatives to Bayh-Dole.