We are looking at the National Patent Planning Commission argument that the government should be permitted to grant exclusive patent licenses on inventions that it acquires. The basic position is that it is a good thing that the government should make its inventions available to all. But now we reach a walk back of this initial position:
While it is the conviction of the Commission that the Government should continue to adhere to its traditional policy of not engaging in activities which can adequately be conducted by private enterprise, it nevertheless believes that in suitable cases the Government should have the authority and the power to depart from the general policy hitherto followed and, recommended above and take steps to insure the proper commercial development of an invention covered by any of its patents whenever this course is necessary and in the public interest.
So, an exception to policy–“in suitable cases” to “insure proper” “commercial development” “whenever necessary” and “in the public interest.” There are plenty of qualifiers here–suitable, proper, necessary, public interest. The assertion is similar to an argument for sole source purchase contracts–no bidding, no one else need apply. Here, the assertion is that sometimes it is in the public interest that a commercial product be developed from an invention, and that for this to happen it is necessary that the manufacturer have a monopoly or no one will justify the cost relative to the possible profit.
Clearly such a product could not possibly be one that the government specifically needs, or the government would pay both for the development and then pay again (with a reasonable profit) for whomever manufactured the product for the government. So such products would have to be ones that the government doesn’t particularly need, but the government determines is in the public interest to have. Again, perhaps pharmaceuticals and vaccines might be such products, but anything else would appear to be an odd niche thing–something so necessary and so unprofitable and with such a small market that only with a temporary monopoly would any company take on the expense to develop. What other products are so public interesting, so necessary and so unprofitable and so expensive to develop that the only way to do it is with a patent monopoly created by the government and conveyed to a single government-endorsed company for “exploitation”?
There it is. Feel free to write in your list of such products. I can’t imagine the list is extensive. The products on the list will be special special cases–special cases where a patent matters for any development, and special special cases where the government determines that the public should have a product and the only way–the necessary way that the public will get that product at all ever is that the government hands its patent to a single company for whatever years are left on the government’s patent monopoly.
Now let’s take a look at “proper commercial development.” Archie Palmer in 1948 published a compendium of university patent and research policies. Palmer’s purpose was to encourage universities to adopt formal patent policies as part of their institution building. Palmer also clearly favored the idea that the patent system should be used to promote the use of research inventions. He was agnostic about who used the patent system, and thus did not indicate whether universities should own inventions, or external patent agents should own inventions, or inventors should own inventions–what he wanted was formal patent policies and increased use of the patent system to promote the beneficial public use of the outcomes of research.
There was, at the time, a great interest–if not hope–in the idea that more research, now stimulated by the federal government, would lead to many new benefits for the public. Vannevar Bush’s Science the Endless Frontier was as much a restatement of this interest and hope than its first expression. But the administrative interest in hope took a slightly different twist–“if we are to continue to get research money from the government, we had better be able to show outcomes beneficial to the public.” And this twisted concern easily became “if we are to continue to get research money from the government, we had better be able to show the potential for outcomes beneficial to the public.” Patents on research inventions nicely suited this twisted administrative concern. A patent policy, then, became necessary to deal with whatever ownership position a university administration chose to take. However that patent policy was drafted, the implication was that the use of the patent system was an integral part of a university’s expanded role in hosting research, especially research supported by public money.
According to Palmer, despite academic scientists desire to publish their findings:
Discoveries or inventions that are merely published, and are thus made available to everybody equally, are seldom adopted, despite their possibilities of commercial application.
Palmer quotes favorably Elihu Thomson, from his 1920 MIT graduate speech:
Publish an invention freely, and it will almost surely die from lack of interest in its development. It will not be developed and the world will not be benefited. Patent it, and if valuable it will be taken up and developed into a business.
This is the same argument made by the National Patent Planning Commission. Left out of this argument is the historical fact that less than 5% of those inventions patented ever reach the stage of commercial use anyway. University officials at the time estimated 1 in 1,000 of their inventions ever became a source of significant royalties–meaning, broadly available to the public as a commercial product. Thus, despite Palmer’s and Thomson’s confident assertions, what evidence supports the claim? The reality is that it does not matter whether an academic invention is published or patented (and thus published in exchange for a monopoly on it) or held as a trade secret, making an invention has little to do with what gets adopted for use, and what gets developed into commercial products. The selection criteria for companies to create new commercial products does not have to do with patenting–or, at least, patenting has little to do with success.
We might consider the proposition that patenting, especially academic or government patenting, may induce more people to spend money trying to develop a commercial product. Another way, academic or government patenting might induce people to spend money trying to develop a commercial product from an academic or government invention rather than spend their money some other way. But efforts to develop commercial products based on the bare fact that something that might have been freely available to all has been made into a monopoly is an empty proxy for actually creating commercial products that are available to all on reasonable terms. The idea that this can happen is attractive. The fact that it almost never happens is a sobering qualification on thinking such thoughts. The further fact that adapting an invention for use does not necessarily require either a patent monopoly or a commercial product version leaves the commercialization argument as a pleasant feeling thought, without underlying substance.
Palmer goes on to quote writers who make a case for academic patenting based on (i) the right of an inventor to his invention (even if there’s no money to be made) and (ii) that an academic inventor owes the public control over the quality of implementations of his invention: “the inventor should so far as possible prevent the sale of inferior or harmful imitations.” That is, the purpose of a patent on an academic invention is quality control, not exclusion of all others in favor of a monopoly. But that argument ignores entirely the argument for the cost of development justified by the hope for monopoly-protected profits.
Palmer quotes Elmer L. Sevringhaus, a University of Wisconsin physician who had become by 1945 the research director at Hoffman-La Roche, a pharmaceutical firm:
Rapid development of discoveries which are of academic interest may be secured when patent rights assure a commercial producer of protection in the field.
Here, the argument is that a patent monopoly will speed development of inventions into commercial products, without reference to whether more such products will be produced, and without reference to what happens to those inventions that end up behind a patent paywall but are never licensed, or if licensed never take the form of a commercial product. In fact, the good doctor’s assertion is not much of an argument at all. There’s no evidence behind it to demonstrate its truth or general applicability. Severinghaus argues that the distaste for medical patents among academic researchers fails the movement of inventions from laboratory to the “non-technical public”:
Unless some practical control for a period of perhaps a few years can be assured, it is difficult to secure adequate backing for the adaptation to quantity production and the prompt distribution of the new article. The patent laws provide just this necessary guarantee; the result is not only profit to the inventor and the manufacturer but also a social gain.
Again, it sounds nice. But notice the qualifications–“perhaps a few years,” “it is difficult,” “adequate backing,” “adaptation to quantity production,” “prompt distribution.” These are arguments for a public covenant to run with academic patents that limits the duration of the monopoly, demands a mass produced product, and demands distribution to a non-technical public. These, too, are nice ideas–but these are exactly the requirements that make the licensing of an academic or government patent unattractive. These idealistic justifications for patenting were necessary for the passage of the Bayh-Dole Act, and were the first things abandoned in the regulatory implementation and the round of revisions passed three years after Bayh-Dole went into effect. We might observe that the patent monopoly is required by the licensee not so much to provide competitive “protection” but rather to justify the huge liability of the idealistic diligence requirements imposed by the institutional patent owner.
Behind these arguments is a claim about what pharmaceutical companies need in order to take in an academic invention in favor of their own in-house opportunities to develop products. These arguments are then generalized to be made to appear to apply to inventions generally, where nearly the only application these arguments have is in the context of the pharmaceutical industry practice. The arguments do not even apply to pharmaceutical inventions generally–only to how a pharmaceutical company views the choice of profit-generating activities from the available opportunities. There, undoubtedly, an invention available to all is less appealing than an invention in which a company already owns patent rights. But even here, there’s the problem of development.
Palmer works through various arguments against academic patenting, but ends with a committee resolution that patenting academic inventions is worthy because the patent laws are endorsed by the federal government for public benefit, that an academic inventor who patents is justified in money making because of the public good that comes about, and anyway, making money from patents is justified because academic researchers need more money to support their efforts. It is this last argument that, by 1960, proved to be the most compelling reason for universities to create patent licensing programs.
Palmer ends with a summary:
The first argument for university patenting is to preclude easy patenting of improvements and blocking the practice of the original invention in improved form. The second argument is for other “cases” in which a monopoly or at least a limited monopoly is necessary because others own necessary inventions or because of the cost involved to develop the invention. Neither of these arguments holds up. If a company owns inventions on which a new academic invention depends for commercial exploitation, then why does the new academic invention have to be licensed exclusively? Why not non-exclusive licensing, leaving free all uses that do not require commercialization? Why not non-exclusive licensing for commercial use, leaving others free to develop alternatives to the existing inventions on which commercial development might depend?
That leaves us with the argument for the high cost that may be involved in some cases (“often”?) to develop an invention as a commercial product. Notice how the idea of the patent monopoly is buried in a cascade of qualifications and options:
it is manifest that the public interests may be best served by giving a license… even exclusive if necessary, either royalty-free or on a royalty or cash basis.
Rephrasing: it is obvious that among the ways to serve the public interest, the best way may, some times, involve granting a patent license royalty-free and non-exclusive, or for a royalty. On the face of it, it is manifest that at some point one of these options may well be the best way. But the list of options itself makes the statement meaningless. The effect of the statement is to normalize the idea that monopoly patent licenses on a royalty-paying basis should be–and therefore are–perfectly acceptable, and may be, obviously, in some cases, necessary (in some way, raising the odd question of what makes such a thing necessary), and thus (if necessary) the best way to serve the public interests. What does Palmer mean by exclusive patent licensing being necessary? How does trading in a patent monopoly serve public interests? The idea appears to be that companies, motivated by profits, might be motivated as well to serve some public idea–develop an academic invention as a product–so long as there is a monopoly on the invention.
What happens, then, when the patent expires? If the argument justifying the monopoly holds up, then we might expect that the exclusive licensee will drop the product in favor of other, proprietary products. How is the public served if the product is available only so long as a company can extract a monopoly price for it and prevent all efforts to copy the product, improve it, or incorporate it into other products? I can see how the grudging company might be served, or how an institutional patent owner seeking money for research might be served, or even how an inventor desiring money in recognition of having made a patentable invention might be served. But how do these interests translate into public interests? One possible rationalization is that the public is served if a research institution obtains more money to expand its research efforts. Research as a self-sustaining industry that can exist and even prosper, if only with a very few results involving the commercialization of patentable inventions. Apparently one such invention every two or three decades is sufficient justification for impounding all the rest “just in case” one proves valuable.
Part 3, the last part, is here.