Bayh-Dole’s march-in provisions (35 USC 203) are worthless. They were designed to be worthless, except for creating a show of public oversight and intervention that allowed Bayh-Dole to get through Congress and be signed into law. Bremer bragged about how he had to step in and make sure the march-in procedures wouldn’t operate by making them so ungainly and difficult that no federal agency would dare to use them. Which is just how things have turned out. But just how worthless is march-in?
Consider background inventions. Under Bayh-Dole (35 USC 202(f)), a funding agreement cannot require a contractor to license any invention that’s not a subject invention unless that requirement has been approved by the head of the federal agency along with a written justification. Furthermore, even if a funding agreement adds such a provision, a federal agency still cannot use that provision without determining that
the use of the invention by others is necessary for the practice of a subject invention or
for the use of a work object of the funding agreement and
that such action is necessary to achieve the practical application of the subject invention or work object.
This is all so strange. Let’s work through it.
the use of the invention by others is necessary for the practice of a subject invention
We move from “use” to “practice”–as if these terms are synonymous–but they aren’t. “Use” means use. Practice means “make, use, and sell.” So a federal agency has to determine that the use by others of a contractor’s invention is necessary for the making, using, or selling of a subject invention. Why not “the practice of an invention by others is necessary for the practice by others of a subject invention”? Oh, because look:
If the contractor licenses a subject invention and withholds a background invention so that the licensee cannot practice the licensed subject invention, the license is all but worthless except to patent trolls–and even there it pretty much has to be an assignment (such as in the form of an exclusive license granting all substantial rights). If no one can use the subject invention but for an unlicensed background invention, then all a license to the subject invention is good for is suing others. That’s not sanctioned by Bayh-Dole’s statement of policy (35 USC 200) and does not meet the requirements of practical application (35 USC 201(f)).
But it’s all much worse than this. The requirement to license a background invention has to be placed in the funding agreement–in the patent rights clause of the funding agreement–before the agreement is even accepted and therefore also, obviously, before any research has begun or any new invention made and turned into a subject invention. Presumably, to add such a provision to a funding agreement, a federal agency must not only comply with 35 USC 202(f)(1) and (2) but also with 35 USC 202(a) and (b)–which sets up changes to the default patent rights clause specified in (c).
But how would anyone know in advance what inventions might be made under a given funding agreement in order to specify what of the contractor’s inventions might be necessary for practice by others?
Worse, why would a contractor be licensing to others subject inventions and not licensing as well background rights? And if the contractor is practicing the invention itself and not licensing to anyone, then what the heck would it matter whether others need background rights to practice foreground rights? Ah, you see how worthless march-in is!
If a funding agreement does not have a clause requiring the licensing of a background invention in order for licensees to practice a foreground subject invention, and one or more background inventions are necessary to practice that subject invention, then there’s absolutely no point in march-in. The government might be able to compel a contractor to license the foreground subject invention on terms that are reasonable under the circumstances, but cannot force licensing of the background invention(s) necessary to practice the foreground, marched-in subject invention.
Only in the odd situation in which a contractor has no background invention could march-in work. And if any others hold background rights necessary for the practice of the subject invention, march-in is still worthless unless those others are willing to grant licenses, too. If they decline, then march-in is worthless–and for that matter, spending time fussing around with any subject invention from such federally supported research is also worthless. One might use patents on such inventions to create standoffs with other patent holders–and perhaps those standoffs might result in cross licensing or something. But none of that is anticipated in Bayh-Dole’s statement of policy. Nothing like “to promote standoffs in practice through the use of the patent system.”
If march-in was going to be used, actually, then these provisions about contractor background inventions would have to be standard in every funding agreement. A federal agency would have to determine special circumstances to take these provisions out of any given funding agreement. Otherwise, lacking these provisions to start, no subject invention that’s within the scope of any other contractor invention–needed to practice or an improvement or an application–can ever be march-in upon because there’s just no point to it. Others can gain licenses but cannot practice without the background right as well, and the government did not bother to obtain that right in advance.
or for the use of a work object of the funding agreement
This part, too, is interesting. A “work object” here cannot be a subject invention so it has to be something else. But what? One possibility is computer code. One could hold patents on inventions and then get federal funding to write software that used those inventions. But then there’s no subject invention at all–just code (and copyrights). Thus, there is nothing that a federal agency can march-in for.
The actual strangeness here in Bayh-Dole at 35 USC 202(f) is that the law contemplates the federal government providing funding for development of some “work object” (such as a computer program) but not at the same time making some requirement that the work object be made available for use by others. What does such funding amount to? Funds to build a private, non-inventive “work object” that no one else should have any expectation of using. No requirement for practical application, no possibility of march-in.
and that such action is necessary to achieve the practical application of the subject invention or work object.
So a federal agency has to show that a background invention is necessary for the practice of a foreground subject invention or “work object” and–this is the stultifier–requiring the license “is necessary” for practical application. What does “is necessary” mean here?
If “is necessary” means that practical application cannot take place without a requirement to license background inventions, then the provision is empty on arrival because practical application does not require licensing, ever. A contractor can use a subject invention and make benefits available to the public on reasonable terms. A federal agency, up front, before any funding agreement has been made, cannot possibly determine that licensing is necessary to achieve practical application. It just cannot–at least if “is necessary” means “that’s the only way it will ever happen.”
If “is necessary” means “necessary because that is the means by which the federal agency has decided how practical application should be achieved,” then “is necessary” means pretty much “because a federal agency requires it” and the rest is just fluffy justification.
As it is, the definition of practical application simply doesn’t apply to “work objects.” The definition is specific to inventions, and inventions are defined by Bayh-Dole to be patentable (“is or may be”) or protectible as plant varieties–and made under a funding agreement. Thus, the law here is just nonsense garble. Yes, we can say that practical application here first has its statutory definition (for the subject invention) and then immediately also some common folk meaning (for the work object). I’m just not sure how to “practically apply” a “work object.” But then, this is Bayh-Dole–so what do you expect? Clear drafting? Coherent practice? Public interest? Bah.
It’s inexcusable that Bayh-Dole does not require the standard patent rights clause to require that a contractor license background invention rights when licensing or assigning foreground subject invention rights. It doesn’t matter whether the licensing is voluntary or as a result of march-in. Otherwise, all the licensing involves is the conveyance of a freedom from being sued for using a subject invention. That conveyance does not provide rights necessary to use the subject invention in any jurisdiction in which the background rights operate. Thus, in those jurisdictions, no one can achieve practical application (without infringing, at least) and no march-in licensing will do a lick to address that problem.
It’s just as bad with assignment. If the contractor assigns a subject invention and does not also provide a license to its background inventions necessary for the practice of that assigned subject invention, then by its action, the contractor has all but dictated that the subject invention cannot be used. Oh, sure, the contractor can wait for the new assignee to realize it needs the additional rights, or the new assignee can grant licenses to everyone and let them have to come back to the contractor a-begging for permission in the background rights. But this sort of thing is just invention thuggery and there’s no point in federal funding getting involved to enable such behaviors.
Thus, 35 USC 202(f) is in the wrong place. It should have been in 202(c) as a default condition, and the modification for exceptional circumstances should have to remove background licensing from the funding agreement on a showing that somehow licensing an invention without the necessary background rights for practice will result in practical application anyway. You can see how Bayh-Dole was rigged to have a public interest apparatus that was useless. The gesture is there, but it’s only a gesture–a cardboard army of useless provisions that cover for the underlying expropriation of inventions made with public support to pipeline to pharma and to patent speculators, and otherwise to keep out of circulation for two decades so that federal research will not have an immediate impact that might adversely affect pharma or speculation.
Without a default requirement to license background rights when licensing foreground rights, march-in is that much more reduced. More worthless than the worthlessness at first impression. Anywhere a contractor builds on its existing base of inventions, march-in is meaningless–anywhere a subject invention is an improvement or application or relies on an exclusive position controlled by the contractor, march-in is meaningless.
Federal agencies still could create default patent rights clauses that stipulated background licensing whenever a contractor assigned or licensed a foreground subject invention. All the apparatus for doing so is in Bayh-Dole. It’s just that federal agencies refuse to use the apparatus, refuse to look out for the public interest–and for that matter, refuse to look out for the interests of industry. All a bare license to a subject invention without the background invention license is good for is trolling industry or keeping an invention from use. How is patent trolling or stifling an invention an acceptable practice under Bayh-Dole’s policy at 35 USC 200? Well, it isn’t. But there it is anyway.