The Bayh-Dole Coalition, a lobbying front for pharma and university patent administrators, claims that Bayh-Dole is a “tech transfer statute” that would be misused if federal agencies used its march-in provisions to address drug pricing. If companies cannot price-gouge, they argue, “you’re going to kill innovation.” They talk nonsense.
Congress sets out Bayh-Dole’s “policy and objectives” at 35 USC 200. This “policy” replaces executive branch patent policy for funding agreements with small companies and nonprofits. The policy at 35 USC 200 requires that the patent system be used to, among other things, promote the utilization of inventions arising in federally supported research or development. It is utilization that matters throughout the statute. The implementing regulation for federal licensing of inventions at 37 CFR 404.2 weirdly reduces the statement of policy and objective to promoting utilization only. But Congress does not leave it at promoting utilization–Bayh-Dole also expects–Congress intends–that the patent system will also be used to maximize participation by small businesses, promote collaboration between nonprofits and for-profits, and promote free competition and enterprise. In addition, Bayh-Dole expects–Congress intends–that federal agencies protect the public from nonuse and unreasonable use.
We might observe, then, that (i) technology transfer is just one element of the express policy of Bayh-Dole and must be considered in context; and (ii) utilization of inventions, with benefits available to the public on reasonable terms, is the full objective–not merely attempts to patent and license inventions, not taking licenses to inventions, not “good faith” efforts to “commercialize.” There is no free pass in Bayh-Dole because a university is trying really hard to find a commercialization partner and can’t, or has found an exclusive licensee but nothing has come of it.
Consider what Bayh-Dole’s policy and objective does. It does more than set out premises for making the law. 35 USC 200 replaces executive branch patent policy with new policy. Bayh-Dole is part of federal patent law. The statement of policy and objective places limits on the patent property rights available to owners of inventions arising from federally supported research or development. For that, see 35 USC 261. Think about the requirement to “use the patent system to promote the utilization of inventions.” This is a working requirement. There’s nothing else like it in federal patent law. Without Bayh-Dole, a holder of a patent has no obligation in federal patent law to “work” the claimed invention. The inventors have performed their public service by disclosing their invention in the patent literature, and if the patent holder chooses to sit on the patent for 20 years, then everyone else has an incentive to design around the patent. With Bayh-Dole, however, the patent system is to be used to promote utilization, not to suppress utilization, not to make utilization less attractive. The Bayh-Dole requirement restricts the patent property rights of the holder of an invention subject to Bayh-Dole.
Ask a simple question, then: how does one use the patent system to promote the utilization of an invention when the patent system is built on the public disclosure of an invention and the right to exclude all others for the term of the patent? Clearly, public disclosure might promote utilization, so one might use the patent system to publish the details of how to make and use the invention. But what about excluding all others? It cannot be. Bayh-Dole expressly reserves rights–rights that otherwise would be within the patent property rights of patent holders–to protect the public from nonuse and unreasonable use. Thus, nonuse of an invention subject to Bayh-Dole is not acceptable, not within the statement of policy and objectives. The patent system, then, is not to be used to suppress utilization of inventions.
This is a restriction on enforcement of patents. There are two major such restrictions set forth in Bayh-Dole, (1) the government license and (2) march-in. Fhe federal government license is part of the default patent rights clause, and specified at 35 USC 202(c)(4). For any invention that a contractor acquires and chooses to keep, the federal federal agency receives a nonexclusive license to practice and have practiced the invention. Not merely the claimed invention, but the invention taken broadly–anything that may be patentable, even if the contractor chooses not to include parts of the invention in a patent application. Thus–a contractor is not allowed to hold aspects of an invention as a trade secret not disclosed to the government. The government has the right to obtain ownership of any elements of an invention not properly (and fully) disclosed to the federal agency 35 USC 202(c)(1).
The government license effectively divides any patent right into a government “market” and a non-governmental “market.” Or, another way, the government license limits the patent holder’s right to enforce the patent against the federal government or anyone practicing the invention for or on behalf of the federal government. Bayh-Dole could have amended 35 USC 287 (which addresses limitations on damages and other remedies), for instance, and made the limitation express rather than casting it as a license. The practical upshot of the government license is that if a patent holder (or anyone licensed by the patent holder) wants to sell to the government or to someone that otherwise might look to the government for products or services, the seller needs to offer terms that are competitive with what the government could do itself, or with the help of others (including companies) authorized to work on its behalf. So the government license sets up a basic level of competition between the patent holder (or any assignee of the invention) and the federal government, for activities that are within the authority of the federal government.
If it is within the authority of the federal government to fund research on diseases with the objective to address public health needs, then it would appear also to be within the authority of the federal government to provide the public with products and services based on those results. Again, Congress intended for the federal government to have a broad nonexclusive license to anything within the scope of “subject invention” under Bayh-Dole, whether or not a contractor sought a patent on it, or some of it. Push it. Congress intended for the federal government to use its license. There’s no point in Congress making provision for the federal government to get a license but not intending for federal agencies to use that license. Congress comes back around to amending Bayh-Dole in 1984 and adds an additional requirement at 35 USC 210(c) that all funding agreements, even with large companies, must include the government license and march-in. It’s clear that Congress considers the government license to be important. It would be truly strange to argue that Congress did not intend the government to use its license, especially when the cost for the government to obtain the benefits of using a subject invention, and to make those benefits available to the public served by the government, would be lower than the price charged by the contractor (or assignee, or licensees) for comparable benefits.
Congress intended for there to be competition, then, between the government and contractors holding patent rights on inventions made in government-supported work, in the governmental “market.” A “reasonable” price to be paid by the government, or by the public served by the government, or by insurance companies covering the cost for that public, is the cost on which the government could obtain the goods and services itself for that same “market.”
It makes no sense to claim that Congress did not intend for the government license to be ever used. Just the opposite. It makes sense that Congress did intend for the government to use its license anywhere that it would be wasting money not to. In a sense, a failure to use the government license could be seen as potential collusion with patent holders to make government work more expensive and so price-gouge the public or deny the public access to benefits arising from research and development that has taken place in projects receiving public support and pitched to the federal government (and to the public) as in the public interest. Such collusion is implicit in the argument that if the government were to act on its government license, then companies would not have any incentive to produce product. Flip the argument: Congress intended that companies providing products and services into the government “market” should price those products and services competitively with what the government could do for itself. If not, then the government has a mandate to act. That’s the policy enunciated by Congress. Intended by Congress.
The second restriction on patent property rights is march-in, also a part of the default patent rights clause, and set out at 35 USC 203. March-in addresses a patent holder’s practices in the non-governmental “market.” If a patent holder does not use a subject invention, then the federal agency has the right to march-in and require the patent holder to grant one or more licenses. Further, if a patent. holder fails timely to achieve “practical application” of a subject invention, the federal agency has the right to march-in. Bayh-Dole follows the Kennedy and Nixon patent policies to give a definition for “practical application.” Reduced to its essentials, 35 USC 201(f) defines practical application as the use of a subject invention such that the benefits of that use are available to the public on reasonable terms. “Practical application” and “utilize” are pretty much equivalent. The meat of the definition, then, is not utilization, but that (i) the benefits of use are (ii) available to the public (iii) on reasonable terms. To meet the standard of the definition, all three elements must be met. There must be benefits. These benefits must be available to the public. And must be available on reasonable terms. Any reading of the definition must give effect to “reasonable terms.” Congress intended to restrict the patent property right of the patent holder, in addition to making a working requirement, to require reasonable terms.
One standard, we have seen, for what’s reasonable is whether the federal government could have comparable work done on its behalf at the same cost or less. Reasonable terms would be terms, including price, that the government found reasonable–“that’s about what it would cost us to do the same thing–or less than it would cost us.” But march-in applies also outside the government “market.” There, the standard for “reasonable terms” is, as far as the public is concerned, that the benefits of utilization are affordable, or that the terms on offer are comparable to those that would be offered if there were competition–multiple sources offering competing versions of the products and services based on use of the subject invention–an invention arising in work pitched to the federal government as in the public interest and receiving federal support.
How do we get to this second standard? Look at what march-in does. March-in requires licensing. Licensing in turn opens up multiple companies practicing the invention. March-in compulsory licensing then promotes “free competition” in response to terms that otherwise take advantage of the public–unreasonable terms, an opportunity supported by the patent right–the right to exclude all others but for the federal government. It is worth pointing out that compulsory licensing on reasonable terms under march-in anticipates that the patent holder will receive royalties based on the activities under those march-in licenses. It is not like the patent holder is stripped of the patent; rather, the patent holder is directed to meet the policy and objective of Bayh-Dole, including to promote free competition.
If it were not for all the hype about what Bayh-Dole does–much of it grossly distorting–we might believe that Bayh-Dole assumes that nonprofit holders of patents on subject inventions will license non-exclusively as a default. That’s the standard, expressly, for federal agencies on their side of Bayh-Dole. 35 USC 209 goes to great lengths to provide a rationale for when a federal agency is permitted to consider exclusive licensing. For federally owned inventions, Congress intended that federal agencies use non-exclusive licensing as the default. Why would it be much different for, at least, nonprofits? Nothing in Bayh-Dole states that the only justification for the law is that nonprofits will offer exclusive licenses as the default and so attract patent monopolists rather than distribute the rights as licenses throughout industry and so promote “free competition.” Not there. Just made up. How does one maximize the participation of small companies? Non-exclusive licensing. How does one promote collaboration between nonprofits and “commercial concerns”? Non-exclusive licensing–creating standards, interoperability, shared research and development. How does one use the patent system to promote free competition? Non-exclusive licensing.
If the argument is that there is no need for Bayh-Dole if nonprofits are expected, as a default, to offer non-exclusive licenses on reasonable terms, then the argument runs toward the idea that Bayh-Dole is failed policy, which it might well be. Maybe federal open access, in general, better serves the public than does nonprofit open access. The problem–as university patent administrators put it–was that it was difficult to know if a federal agency would, for any given invention made at the nonprofit in work supported by federal funding, allow the nonprofit to keep principal rights in the invention, and even if a federal agency might allow the nonprofit to keep the invention, it was uncertain how long the federal agency would take to finalize its decision. That was the “uncertainty of title” argument.
That argument, however, fails to get at whether university patent administrators were able to make a persuasive case for nonprofit control of patents. Answer–for the most part, they weren’t. They did not have good evidence that their patent and license programs were all that effective, and they did not have good arguments that patent exclusivity was all that necessary to the transfer of technology or the uptake of that technology by industry. They did have a point–that DHEW was consistently denying university requests to retain title to inventions made in work receiving federal support–but they didn’t have a response that showed that they had made a persuasive case for nonprofit title in terms of what would better serve the public interest.
Given the claim made by some studies (sketchy methodologies, but . . . ) that nearly all drugs developed in the recent past have benefitted from the results of federally supported research, most of it published openly, it is not clear at all that patent exclusivity in such results matters much at all. Even the Stevens et al. study that traced commercial products to patents held by non-commercial entities did not bother to consider which of those products depended on an exclusive license from those entities. That is, an exclusive license may have been involved, but it would take more care to establish that the licensee would have refused to use the invention if it had been offered a non-exclusive license, especially one that was royalty-free (as were government non-exclusive licenses before Bayh-Dole).
A good way to tell that an exclusive license, at least for a limited time, is necessary is to offer access to an invention on non-exclusive terms–perhaps even royalty-free. If no one adopts the invention in, say, three years from the date of patent issue, then consider whether exclusivity would induce someone to invest in the work needed (if any) to make the benefits of using the invention available to the public on reasonable terms. University patent administrators generally don’t do this. They do the opposite. They hold out for an exclusive licensee, or refuse to license at all. They almost never open up an invention to non-exclusive licensing after seeking an exclusive deal. The gist of their argument has been “if we offer a non-exclusive license first, then we lose the opportunity to offer an exclusive license.” Of course, that’s not true. What’s true is that if anyone takes the non-exclusive license, then it is not possible to offer an exclusive license. But if someone has taken the non-exclusive deal, then it’s clear that exclusivity wasn’t necessary for transfer, and we are back to whether federal open access would better serve the public than Bayh-Dole, and whether a default of non-exclusive licensing would better serve the public than a default attempt to find an exclusive licensee, no matter whether a nonprofit contractor or federal agency acquired any given invention.
All this licensing talk applies, as far as Bayh-Dole is concerned, only to nonprofits. Small companies have no particular need to grant licenses. The point of including small companies is that it preserves their interest in making their own products, with some way of fending off other companies. For that, perhaps some patent exclusivity matters, but even there, if a big company wants access, and all that stands in the way is an infringement case that will likely get settled anyway, once the patent has survived a challenge for invalidity, for a reasonable royalty on a non-exclusive license, then we are back to non-exclusive licensing on reasonable terms. Patent exclusivity might sound nice to some investors, but for the most part it is administrative drama–often expensive administrative drama.
If we piece the logic together, it may be that Bayh-Dole is failed policy. But it is not acceptable to argue that since Bayh-Dole is that failed policy–or at least a policy that university patent administrators don’t like–then a different policy, one not in Bayh-Dole, should be made up as a private story about what the law should have done, and claim that private story is what Bayh-Dole “was intended to do” and therefore must be the “actual” law. That’s just private tyranny, taking over public law to say something it doesn’t say. The university patent administrators and their attorneys tried that in Stanford v Roche, and the Supreme Court read the law back to them and showed them point by point that they–including Sen. Bayh, who filed his own amicus brief arguing for a reading of the law that had no support in the law, were just making things up.
As for march-in, granting an exclusive license to a subject invention does not address either utilization or practical application. An exclusive license that conveys all substantial rights in an invention–make, use, sell–acts as an assignment of the invention. Changing ownership of an invention does not necessarily even “transfer” technology. An assignment just changes–once there is a patent–who has standing to exclude others from practicing the invention. Bayh-Dole is not, then, a statute that sets as an objective for a patent holder to convey rights to an invention to another party. Doing so merely changes who has obligations under Bayh-Dole. It’s not necessarily even “technology transfer” to flip ownership of an invention. And assignment is not utilization or practical application. We might say, in general, that assignment has nothing to do with taking “effective steps to achieve practical application.” Assignment just changes who is responsible for taking effective steps. If assignment alone was sufficient to avoid march-in, then contractors and assignees could play hot potato and pass round patent rights, restarting the “reasonable time” clock with each new assignment.
The point of technology transfer is not merely to flip an invention to a new owner, and certainly not with the added condition that the new owner has to pay the original owner’s estimate of the value of the exclusive rights or no one will ever get a license. Offering to grant an exclusive license to all substantial rights–make, use, sell–may be an “incentive” for some parties to take that exclusive license. But that exclusive license may not be at all necessary for other parties to adopt the invention, use it, and develop it. The offer of an exclusive license may work against adoption of the invention by these other parties.
If a university grants an exclusive license to an invention–shifting then the ownership of the invention to the licensee (=assignee)–then everyone else in the industry has to avoid the invention, exclude it from standards, design around it, not use it. Yes, there is an argument that suppressing industry access to advances made in federally supported work creates incentives for further inventing to get around that suppression, but you see where this heads, don’t you? Federally supported results, when patented by nonprofits and held for exclusive licensing, are more likely to be avoided by industry. Federal research, when patented and offered for exclusive license by default, becomes less useful and less used. Is this what Congress intended with Bayh-Dole? There’s nothing at 35 USC 200 about a policy to limit the utilization of inventions made under federal funding agreements by industry by holding out for exclusive licenses or no deal. Nothing there that suggests that maximizing the participation of small businesses is best done by granting only one of them an exclusive license, or at least trying to. Nothing there that suggests that “free competition” comes about because a nonprofit plays favorites with one company and the rest have to work to make the Bayh-Dole invention obsolete. That’s a strange sort of “free competition” for Congress to intend.
The “tech transfer” contemplated by Bayh-Dole is end-to-end. From invention to use with benefits available to the public on reasonable terms set by free competition or by reference to terms that would be offered if there was free competition even if patents are used to suppress competition for a limited time in favor of more rapid development.
Federal agencies have a statutory mandate in Bayh-Dole to use the government license whenever the government could have the product it needs made and delivered for less than what a company benefiting from a university exclusive deal prices it at. Congress intended for federal agencies to “compete” as it were with private concerns in doing their agency jobs, in their area of work to benefit the public welfare. Federal agencies have a statutory mandate in Bayh-Dole to use their march-in authority where they find nonuse or unreasonable use, including a company offering the benefits of using an invention made in federally supported work on unreasonable terms, including price and price discrimination. March-in does not dictate price. March-in introduces competition where patents have been used to suppress competition and pricing exploits the absence of competition and is unreasonable.
March-in, like the government license, is an essential part of Bayh-Dole’s policy on technology transfer. Without march-in and the government license, Bayh-Dole is half-law. Congress did not intend a half-law, as if federal funding for research and development were merely a condition-free public subsidy to offset losses by the owners of companies. The public part of the bargain that releases nonprofits from having to make the case that their control of inventions made in federally funded work is better for the public than is federal open access is that there must be competition with the federal government in applying the invention to federal government purposes, and that in areas where the federal government has no authorized purpose, the terms on offer must be those that would be competitive, if there were competition in the form of multiple companies offering products and services based on use of the invention. If those terms are not competitive, then march-in creates competition. Competition addresses nonuse and unreasonable use. That’s the essence of the technology transfer bargain in Bayh-Dole, one that many university patent administrators refuse to comply with. The result: tens of thousands of inventions made in federally supported work are not available to the public–have not been “transferred”–and many of the small percentage that have been licensed have not achieved practical application–use with benefits available to the public on reasonable terms.
If the idea of “reasonable terms” would bring an end to most university exclusive licensing, that would appear to be just what Congress intended. That would be the policy of Bayh-Dole, not a misuse of Bayh-Dole. The misuse of Bayh-Dole is the failure of federal agencies to use the government license and to march-in early and often where there is nonuse or unreasonable use, including unreasonable terms such as exorbitant pricing and price discrimination that denies public access, especially to health needs.
At its logical, statutory core Bayh-Dole requires inventions made in federally supported research or development to be worked, and expects the patent system to be used to promote the utilization of inventions, not the suppression of inventions. It takes some care to figure out how to use a system of excluding access to promote use. The obvious ways are non-exclusive licensing–breaking up the patent right in order to create competition–and limited exclusivity–for a limited time, covering less than all substantial rights such as the exclusive right to sell. If a contractor fails at utilization or fails to offer reasonable terms as if there were competition, then Bayh-Dole’s logical, statutory remedy is to introduce competition–either by the government doing things itself for less, or through compulsory licensing to introduce competition to remedy nonuse or unreasonable use.
The public protections in Bayh-Dole–government license and march-in among them–are essential to Bayh-Dole’s technology transfer policy (“to promote the utilization . . . “). Their use is what Congress intended, what the statute mandates. Bayh-Dole is not just a technology transfer statute, but for the part that is, the public protections are essential to the bargain. If protecting the public through competition is a policy failure, the Bayh-Dole is failed policy and should be repealed. If, however, protecting the public is good policy–and surely innovation will not grind to a halt if patents are not used by nonprofits to price-gouge the public for 10% of their medicines–then Bayh-Dole should be enforced (on federal agencies) and federal agencies in turn should enforce the Bayh-Dole authorized patent rights clauses. That’s what Congress intended, even if university patent administrators would prefer price-gouging and even if lobby groups think it would have been better if Congress, in passing Bayh-Dole, had intended price-gouging as the public’s cost to have all their medicines developed by companies that prefer price-gouging as their way of doing business.
Think of it this way. Let’s say the folks at the Bayh-Dole Coalition are right, that without patent-based price-gouging, there won’t be any more innovation. That would mean–we are at their circus here, so enjoy it–that eventually medicines on patent now will come off patent, and there will be competition (or antitrust investigations), and when medicines have come off patent in the past, the price of the medicines drops by 10x or more. In the case of the prostate cancer medication Xtandi, the price per pill may drop from $80 to $3, at four pills per day, we are looking at saving over $100,000 per patient per year. The public may not get as many medicines, but the cost will certainly come down by tens of billions of dollars. Most–maybe 70%–of the patents that companies are obtaining on drugs are used to extend patent monopolies with minor variations on a given theme. A company may surround a medication with scores, even hundreds, of patents to suppress competition even after the initial patent has expired.
It is difficult to find a way to justify the idea that Congress intended price-gouging in passing Bayh-Dole. Bayh-Dole provides for public protections to run alongside a working requirement. For nonprofits, this means that technology transfer must follow through from licensing each invention to each invention being worked, with benefits available to the public on reasonable terms. One way to achieve reasonable terms is by creating competition through non-exclusive licensing or limited exclusive licensing. Another way is to price new products as if there were competition, even where patents are used to suppress competition. And a third way, where products are priced unreasonably above competitive prices, is to introduce competition, either through use of the government license or through march-in–or both.