Here is one of the most provocative parts of Vannevar Bush’s Science the Endless Frontier:
Science Is a Proper Concern of Government
It has been basic United States policy that Government should foster the opening of new frontiers. It opened the seas to clipper ships and furnished land for pioneers. Although these frontiers have more or less disappeared, the frontier of science remains. It is in keeping with the American tradition – one which has made the United States great – that new frontiers shall be made accessible for development by all American citizens.
The problem is not so much with the metaphor–that somehow discovery in science is like discovery in geography–but rather with the idea that comes after is that the federal government should fund “basic science,” especially in the “war against disease”:
There are areas of science in which the public interest is acute but which are likely to be cultivated inadequately if left without more support than will come from private sources. These areas – such as research on military problems, agriculture, housing, public health, certain medical research, and research involving expensive capital facilities beyond the capacity of private institutions – should be advanced by active Government support.
This, then, is the logic: government should be mindful of science; science represents a frontier that may be developed; private sources of funding are inadequate to “cultivate” all areas of this scientific frontier. The federal government should provide funding to advance basic science.
Bush went on to describe a specific means for providing government support. He recognized the problems in how the government might use the power of its funding to dictate the priorities of what should be studied, with what outcomes. He acknowledged that industry had its research and the government had its research, but these forms of research were not what he was concerned with. Bush argued that the frontier of science was not adequately cultivated by either of these existing efforts. What was needed was independent inquiry, free from the direction and focus of company or governmental needs, free from the priorities and problems of professional consensus. Bush wanted institutional funds to enable “the free play of free intellects” to expand scientific frontiers not adequately cultivated by existing funding, so that the results–and these are the fighting words–shall be made accessible for development by all American citizens.
Bush proposed a mechanism to make his idea possible–a National Research Foundation supplied with government money but independent of any government agency, run by a director and citizen board, accountable to Congress for its performance and reporting directly to the President. The government would receive a license in any patented invention made with NRF support, but other than that the money would be provided without strings. It would be “grant-in-aid” or subvention funding, like a fellowship to support independent inquiry, not procurement to buy something or someone.
But all this was ignored–instead, folks fixated on the idea that the government should have money to spend on research. Federal agencies all clamored for their share. University administrators argued that it was dangerous to concentrate federal funding for “basic” research with any single agency. Everyone should get money for research, if research was so important. Basic research could be sponsored by any federal agency; basic research could be done by any institution that assembled the resources to bid for funding. Instead of subvention allocated to individuals to support their work by an almost-independent Foundation, government officials worked out a system of procurement by which projects were proposed and funded. Thus, the government’s funding shifted from supporting independent investigators to supporting projects evaluated on the merits (to the institutional sponsor) of the project, not the person. It’s a huge difference. Rather than–is this person an independent thinker, pursuing work that has in the past proved productive? we get –is the proposed new project properly drafted and does it meet with the institutional approval?
More so: if federal funding for basic research is a form of procurement, but with the statement of work supplied by the bidding party, then the government may also negotiate for rights to inventions. Indeed, the convention of the procurement contract establishes the right of the sponsor to state what it wishes to acquire as part of the deal, take it or leave it. Once the basic form of the funding relationship was established as procurement, but of “research services proposed by a candidate contractor,” then it was a matter of how much interest the federal government should have in inventions that arose in this research. This is the origin of the “title” or “license” debate, a political dichotomy. Should the federal government negotiate to own all inventions made with its support, or should it be content with a license?
In 1947, Gordon K. Lister in “Government Patent Rights,” an article published in the Bulletin of Atomic Scientists, worked through the issues, asking, among other things, “What should the Government do about the patents and patent applications it already owns?” Lister outlines four functions of the “patent system”:
- induce disclosure of inventions
- induce the investment of venture capital
- promote enforced diversity of innovation
- reward the inventor
A patent may encourage disclosure of inventions that otherwise might be held as trade secrets or indifferently. An entrepreneur might invest in developing an invention knowing that he may “reap more than an ordinary profit.” Patents also force “diversity”–a competitor has to design around a patented invention and cannot merely copy it. Thus, patents encourage “diversity of innovation.” (Of course, a patent, often being broader than a single point of invention, discourages diversity of variations within the claimed invention, forcing any design-around to go way, way around). Finally, the patent system rewards inventors.
All this sounds good in the personal and singular sense. Things change, however, if companies are substituted for individuals. Companies might choose to disclose inventions rather than hold them as trade secrets, but often the decision whether to invest in a new product does not depend on a monopoly position derived from any particular invention. One invention might involve a patent, but ten others that are also necessary to the product’s success might not require patenting at all. One can see how companies, in securing swaths of patent-controlled technology, might make other companies design way around or remain excluded from the market. But as each company cuts swaths through an emerging area of technology that requires contributions from many companies, we get gridlock–no one can acquire all the rights necessary to make any product (and this is what happened, for instance, in the early airplane industry in the U.S.) So diversity of innovation ends up collapsing into standards, or to cross-licensing, or to patent pools, so that anyone with any rights has access to sufficient rights to make product. The enforced diversity of innovation, then, is a limited argument–it’s more an effect of the patent system than a “primary function.”
As for the fourth primary function–rewarding inventors–that nearly goes away once companies are involved hiring people to do the research and requiring them to sign away their invention rights as a condition of employment.
And all this becomes strange when institutions–universities or governments–get involved. The publication issue is moot. Universities expect to publish without any inducement, and if a government classifies new technology, no offer to itself of a monopoly makes any difference. As for investment capital, a university is all but forced to look outside itself for that capital, where a company may simply make a decision where to spend its resources. Universities, then, might look to speculative investors as well as to established companies for money. But if a university relies on a patent, and insists that the patent should carry the value of the activity to develop an invention, then the university looks for situations in which investors insist on the presence of a monopoly for each and every invention in order to justify product development–unlike many company settings in which a patent monopoly for each invention is simply unnecessary. One might see that a university patent licensing program then could be parasitic (if not predatory) on industry, patenting up swaths of technology that companies in a given industry would collectively leave unpatented–used, but not the source of competitive advantage; left to the commons as an enlightened self-interest, to prevent gridlock or meaningless patent litigation. Finally, universities tend to reward inventors–but only with a share, typically a minor share, of whatever the university makes from its patent licensing program. This inventor share, however, tends to reinforce an institutional desire to make a patent license carry the entire value of the university’s effort to “transfer technology”–increasing the problems of patenting everything and forcing industry to accept exclusive licenses as a precondition of access.
Lister, however, focuses on these considerations when the federal government issues patents to itself. The disclosure and inventor reward elements can “be largely disregarded.” Government does not need to offer itself a monopoly in order to publish inventions it acquires. Nor does it need to offer inventors a share of income for what they invent–they are hired by the government, or supported by the government and have accepted the terms of their support, much as they might at any company. They choose to do their work, not in the hope of inordinate profits, but for a fair wage and satisfaction with research discoveries (or so we might psychologize).
Now Lister gets to (2) and (3)–investment capital and innovation diversity. These, he argues
can be effectuated only by preserving the patent monopoly, that is, by granting to an individual or corporation an exclusive right to use the invention covered by the patent.
In essence, the government must operate a second shadow patent office. Using the first patent office, the government issues patents to itself (by requiring inventors to assign rights to their inventions, permitting the government to have patents issued to the government, not to the inventors). The government then must operate a second patent office, ad hoc, to re-issue the patents as “preserved monopolies” to private parties to exploit. Even here, however, the assumption is that the private party that receives such a patent re-issue will work the underlying invention (at least some bit of it) and not merely trade in the patent monopoly itself–acquire the monopoly for less than the market values it at, and so sell (or sublicense or assign) the patent to someone who values it more than what one owes the government for it. Such activity would be a trade in patents, not an inducement to invest. The diversity of innovation would not be that a competitor is prevented from copying a product that is on the market, but rather is prevented from copying a product that is not on the market–the re-issue patent owners simply trade the patent monopoly based on differing premonitions of the future value of the patent.
Lister finds that the federal government may run into problems operating such an ad hoc shadow patent office. Can the federal government legally grant such monopolies, but not to the inventors based on inventive merit? Obviously, Bayh-Dole claims the federal government can do so–yet one might ask where is the Constitutional authority for granting such monopolies? Lister concludes:
As a practical matter it seems likely that the Government has only two alternatives that it can adopt with respect to the patents it owns, namely dedication to the public or adoption of a procedure whereby it charges a relatively modest fee to anyone who wishes to use the invention covered by the patent.
In fact, these are the two alternatives laid out in the Kennedy patent policy in 1963:
Government-owned patents shall be made available and the technological advances covered thereby brought into being in the shortest time possible through dedication or licensing and shall be listed in official government publications or otherwise.
That “or otherwise” is one heck of a catch-all construction. In other words, Lister’s argument (and the policy position of the Kennedy patent policy) is that the federal government should not trade in patent monopolies–if the federal government acquires a patent, it should be for a public purpose other than trading on the value of the patent. For instance:
- To create an import barrier to protect a domestic market
- To regulate safety or quality or true advertising with regard to a new product
- To set up cross-licensing if others create patent gridlock in a new area of technology
- To create a higher bar to patentability of future inventions
But Lister does not note any of these possible functions and instead notes the argument that if (2) and (3) cannot be supported, then “these functions of the patent system are largely vitiated by a policy which provides for assignment of patents to the Government and thus the issuance of such patents becomes simply so much waste motion.” Lister observes that “it is rather difficult to arrive at a wholly satisfactory alternative” and concludes that the problem would be best answered if the government published inventions through a special statutory procedure rather than obtained patents–much like publishing an invention through the patent system but obtaining no grant of exclusive rights for doing so.
This argument, one might see, is at the root of Bayh-Dole as well–the patent system is to be used to promote the use of inventions, but leaving it entirely open how patents might be exploited. One could use patents to create the situations involving investment capital and innovation diversity; one could use patents to trade on the value of patents, without working any invention; or one could use patents to break up the patent monopoly while doing other things to promote the use of the underlying inventions.
In the Kennedy patent policy and in the NIH IPA program based on it, the expectation was that the monopoly in a government-supported invention, if retained by a contractor, would be broken up before the term of the patent expired–within three years under the Kennedy patent policy; or the sooner of three years from first commercial sale or eight years from the date of an exclusive license under the IPA program that worked to circumvent the Kennedy patent policy while creating the appearance of complying with it. Under the original Bayh-Dole, it was the sooner of five years from first commercial sale or eight years from the date of exclusive license, but only for nonprofit licenses to unsmall businesses, and by 1984 the limitations had been removed entirely (and so small companies got no more than a handwaving gesture of benefit from nonprofit licensing).
The policy under Bayh-Dole is that the government should not decide how to exploit patent rights in inventions made with federal support until institutions have made a decision not to exploit a given invention. And within Bayh-Dole, the federal government is given an express authorization to undertake a trade in patent monopolies through exclusive licenses that in essence assign the inventions to private parties and thus re-issue, as it were, the patents to those parties, subject to a review of the merits of a plan regarding exploitation of the invention, such plan subject to later revision.
The problem Vannevar Bush faced was what to do if the federal government expanded its research funding beyond its procurement and programmatic needs–to fund independent research for the sake of the benefits of independent research. For Bush, writing in 1945, independent research meant individuals exploring and owning what they might invent, subject to the expectations of academics anywhere, to publish or act on their inventions, granting the government a license if they chose to patent their inventions. Virtually no university handled its own patent work in 1945–inventors were sent to Research Corporation or to a university-affiliated research foundation–and many of those foundations sent the inventions on to Research Corporation. Thus, in Bush’s formulation, if the federal government funded more “free play of free intellects,” the only thing the government ought to bargain for is not to be sued from practicing what might be invented.
But things went south quickly. Instead, the title policy dominated–government should own inventions. And where the “license” policy countered, the argument was made by institutions, not by inventors or on their behalf. That is, the government should be satisfied with a license from the company or institution that acquired the rights to an invention made with federal support. As Lister noted in 1947, once the government became involved as a potential owner of inventions as research deliverables, the patent system function of rewarding inventors was no longer relevant. Bush’s idea of an independent National Research Foundation also was no longer relevant, to be replaced with its doppleganger, the National Science Foundation.
One can see how problems might develop. The federal government justifies funding in non-procurement, non-programmatic areas of research on the premise that private funding is inadequate to cultivate that area. Look then at medicinal chemistry. Pharmaceutical companies provided bits of funding–significant for the faculty researchers involved, but not particularly robust funding–and offered to screen discovered compounds for biological action at no charge (except for an option to an exclusive license in any compound that proved interesting). Was such funding inadequate to cultivate the area? Should the government extend its interest to support much more research in medicinal chemistry, with a premise that the results should be “made accessible for development by all American citizens” and not just as patent monopolies for drug companies, subsidized by public funding?
Bayh-Dole answers that subsidizing the pharmaceutical industry through government research funding is an appropriate function of government. That’s the law. No wonder folks like Bremer and Latker were ecstatic that they got things through Congress to make the answer not merely PHS policy, not merely government-wide executive branch policy, but baked into statute. Baking the statute meant cutting off the debate with regard to title vs license. Asserting for three decades without evidence that Bayh-Dole has been a success has served to prevent the discussion from regrowing.
But the problem remains, and it starts not with patent rights but with the move of the federal government into private research activities on the premise that these activities are inadequately cultivated by the general public and its various resources–companies, foundations, investors, bank loans, accident, and the like. Bush tried to craft that move to keep investigators free of government claims or the institutional exploitation of those government claims. In that, Bush failed almost entirely, but his work was used to justify the expansion, and with the expansion the procurement contracting, and with that, the title vs license debate, and within the title debate whether there should be some special public covenant that restricted what a patent owner might do with a patent on an invention made with federal support. Early policies acknowledging that private risk capital might be required imposed a public covenant. Bayh-Dole does way with nearly all express statements of a public covenant, while retaining the language that stipulates such a public covenant in its statement of policy (35 USC 200) and related definitions (such as 35 USC 201(f)) and basis for action to remedy covenant failure in the private market (35 USC 203).
The effort of Bayh-Dole has been to implement not merely a “license” policy, but to cut nonprofits into the mix as brokers for conveying patent monopolies unbroken to private companies, starting with the pharmaceutical industry. That is, the government should allow universities–many instruments of state government–to trade in patent monopolies on the argument that these universities can do a better job than the federal government in trading on monopolies. That is, states should run shadow, ad hoc patent offices to convey monopoly rights to companies because the federal government has not been effective at doing so–primarily because the federal government had a policy of not doing so. The irony is that once Bayh-Dole authorizes federal agencies to run their own ad hoc patent offices (35 USC 209), there is no particular need for universities to have the benefit of the “license” policy. The federal government could take title and re-issue patents nationally, avoiding the hit-and-miss incapability of most university patent licensing offices, the fragmentation of ownership, and the utter lack of uniform institutional standards for determining who to grant monopolies to, and how to enforce those grants.
Bayh-Dole, as a creature of political rhetoric, still retains within it the elements of inventor ownership (soon to be eradicated by NIST without any statutory authority for doing so), of public covenant (though mostly disabled by a tangle of procedures), of government prerogatives (which federal agencies refuse to enforce or act upon), of uses for patents that break up the patent monopoly (and which most university administrators refuse to choose until forced to do so). But Bayh-Dole is designed to destroy these very things, even while it must carry the scars as it were of their presence in past discussions of the government’s role in funding private research but without a governmental purpose in procurement or program operations.
If we accept Vannevar Bush’s argument that funding “science” is a “proper concern” of government, then we might ask how government might determine whether any particular area of science lacks adequate non-federal “cultivation.” Are there areas of research that the government would do well to stay out of? For these areas, government money might chase out private money rather than augment it. For these areas, too, the promise of government money might invite in people who make their careers chasing government money rather than doing something independently useful–and thus also swamp out the privately supported investigators.
We might also ask what the government should do with regard to funding arrangements and inventions when it does choose to cultivate an area of science. Should such funding look like procurement (as now) or like fellowship (which even Bayh-Dole disclaims an interest in)? It may well be that the way to deal with Bayh-Dole is to leave it on the books but move federal funding for subvention research to fellowships. There would be huge immediate gains. Fellowships do not carry the same indirect costs that sponsored research does. Fellowships do not require wasteful reviews of proposals, most of which are rejected. Fellowships to not require people to cater to whatever happens to be the consensus trendy topic for research. Fellowships do not require the horrible overhead of Bayh-Dole or even an expectation of institutional ownership of inventions.
For all of those folks arguing that government funding for basic research is essential to innovation and national prosperity, what could they say if the federal government shifted its subvention funding for research from procurement to fellowship? Perhaps they would argue that doing so was unfair as it cut institutions out of their indirect costs and their patent ownership claims–as if institutional administrative costs and patent licensing business was actually what’s essential to innovation and national prosperity. That would be good for a laugh.
With a fellowship approach, the federal government could leave inventions as they happened, neither taking title nor demanding a license. What would matter, for an investigator to receive any additional fellowship support, is what the investigator had done with the support already provided, including what the investigator (and any inventors working with the investigator) had done with any inventions they had made. One might then review past decisions and actions to reveal the potential for independent science, rather than review statements of intentions and future work. If one can state future work with sufficient detail to be reviewable as prospective science for government grant review committees, then that work is already pretty dull stuff. As I’ve been repeatedly told, to win a federal grant, one has to have already pretty much done the work to be able to write the proposal with sufficient clarity to be competitive.
It may be, then, that the federal government should not have moved into medicinal chemistry as an area of research, and should have left that area to the pharmaceutical industry to support. Given that the federal government did move into medicinal chemistry, it ought to have made a commitment to follow on from exploratory work to screening to development of product ready for commercial distribution–as it had done for leukemia and malaria. Why not for other areas? Why just fund exploration, but then demand to own whatever might be discovered, only to give the discovery away to everyone? I’m not ready to believe that monopoly rights are the only, best, or most effective way of developing compounds that prevent disease, cure disease, or repair injuries; but it is a worthwhile to ask whether the government botched its entry into medicinal chemistry research and thus created a situation that it was reasonable for the pharmaceutical industry to resist–and even for NIH insiders to undermine PHS policies. The NIH funding program appears as scorched earth for private investment. The NIH flooded an area of research with funding and so made private initiatives that involve substantial expense unattractive, if not impossible within the business framework of pharmaceutical companies.
We arrive then at the idea that Bayh-Dole is a patch on a botch. The patch turns public money into a subsidy for pharmaceutical company investors, giving them a better betting game as they play for other people’s money and support a monopoly approach to drug development and pricing. The patch also requires institutions to strong-arm inventions from inventors to ensure that no one has any opportunity to find alternatives to the monopoly model.
In Bush’s scheme, a university inventor, being a free intellect allowed free play, would have decided what inventions to recognize, what inventions to publish and which to patent (and therefore publish through the patent system). Inventors would then have controlled the quantity of proprietary claims on their work. They also would have the choice of how to manage patents if they wished to obtain patents. They could manage patents to document priority and technical expertise; to control quality or safety; to preserve a commons against those who would use patents to block or disrupt it; to commercialize an invention to their own advantage (or to that of a company). They could break up a patent right and permit use by all while controlling the right to sell. Or they could trade in patents as monopolies, producing monopolies from their inventions and selling off the right to re-issue the monopoly to others. Any of this. But at least it would be decisions made by individuals free of institutional controls; free of institutional demands to patent everything and institutional demands to operate as ad hoc patent offices, dispensing monopolies to favorites for a share of the action.