[updated to expand on the limits on government-side licensing once the (e) exploit is in operation–5/19/17]
There is one more circumvention in Bayh-Dole that can be used to by-pass much of the onerous apparatus of the law. This circumvention involves an exploit of the almost totally ignored contractor license and march-in provisions in the standard patent rights clause at 37 CFR 401.14(a)(e). The march-in in (e) is not the one authorized in 35 USC 203, in which the government can march-in on assignments and exclusive licenses to require contractor licensing to meet what an agency has determined to be a need for public availability or failure to comply with the U.S. manufacturing requirement. Instead, the (e) march-in has to do with a provision under which a contractor “retains” a domestic license with the right to sublicense even when the federal government obtains title to a subject invention (and goes on to patent the invention). That is, (e) requires the federal government to grant a conditional, revocable non-exclusive license to all subject inventions, with a conditional right to sublicense. For the federal government to attempt, later, to grant an exclusive license in any given subject invention, it must first overcome the march-in barriers set up in (e).
This (e) provision has no basis in Bayh-Dole. Like the (f)(2) written agreement requirement, it is something made up in the moment perhaps because Bayh-Dole didn’t manage to set out all the things that ought to go into a patent rights clause (even if never enforced). Or perhaps (e) is there in the standard patent rights clause because it wouldn’t have made it through a Congressional review if it had been placed in the text of the statute, as one more required provision in 35 USC 202(c). One wonders why Bayh-Dole isn’t reduced to just *one* requirement for the patent rights clause and then leave *all* other provisions to be drafted outside Congressional oversight. Apparently they had to make a show of enough apparatus gesturing at public oversight to get the law through, and put the walk-backs and add-ons into the implementing regulations, where there would be more special interests to support them and less bothersome meddling by elected representatives. But that’s another issue.
Anyway, let’s have a look at 37 CFR 401.14(a)(e). Here’s the basic gesture:
(1) The contractor will retain a nonexclusive royalty-free license throughout the world in each subject invention to which the Government obtains title, except if the contractor fails to disclose the invention within the times specified in (c), above.
As long as a contractor reports a subject invention, the contractor has a royalty-free license to it. Essentially, the standard patent rights clause requires the contractor, in conveying title to the federal government, to hold back the right to practice the invention (though the clause does not actually specify the scope of rights “licensed”). The effect is roughly the same as compelling the government to grant a license to the contractor.
The wording here is strange, because the contractor cannot actually “retain” a “license”–the contractor can “retain a right” (and thus not assign the contractor’s entire right in the invention to the federal government) or the contractor can require the federal government to grant back a license, and thus “receive” a license to practice the invention, or the federal government can be required by this clause to “grant” a license to the invention. Perhaps what is meant is that the federal government doesn’t have to bother “issuing” an actual written license to the contractor–the contractor can write up the license for itself, or can just point to the (e) clause if anyone challenged the contractor’s right to practice a subject invention to which the government has acquired title.
In usual Bayh-Dole fashion, the clause is a garbled mess. It would appear that this clause was drafted by amateurs–clever in a way, but thoroughly clueless. However, it does communicate the basic idea that the government is committed, by default, not to sue the contractor for infringement of a patent on an invention that a contractor has assigned to the government (so long as the contractor properly reported the invention in the first place).
The next part, though, is the basis for the circumvention:
The contractor’s license extends to its domestic subsidiary and affiliates, if any, within the corporate structure of which the contractor is a party and includes the right to grant sublicenses of the same scope to the extent the contractor was legally obligated to do so at the time the contract was awarded.
The highlighted text is the operative part. The contractor’s license includes the right to sublicense–so long as the sublicensing was set up *before the funding agreement*. The prospect, then, for a circumvention of usual practice arises. One can pre-sell licenses to future subject inventions. The license can take the form of an option–just enough so that should the option be exercised, the contractor would be “legally obligated” to grant the license. While the contractor’s license is non-exclusive, there is no limitation on the contractor’s sublicensing. Such sublicenses cannot, of course, be exclusive with regard to the patent, since the federal government also has the right to grant licenses, but a sublicense can be exclusive with regard to the contractor’s own rights–or a sole license.
Perhaps the shape of this exploit is beginning to dawn on you. I’ll shift the horizon a bit more. Bayh-Dole divides up a patent right into a private portion and a federal portion. In the usual disposition, a subject invention is owned by the contractor and the government receives a non-exclusive license to “practice and have practiced by and on the behalf of the United States.” This government license is broad–practice means in executive patent policy “to make, use, and sell” and have practiced means “to have made, have used, have sold.” The rights are essentially the same between the contractor and the federal government (except, oddly, that the government is not granted expressly the right to import–so the government apparently can’t authorize a generic drug company manufacturing in Canada to produce Xtandi for sale in the US–what a bungle Bayh-Dole is!).
The difference in the two sets of rights is that the federal government’s scope is restricted to the government procurement market–the federal government and state governments (“the United States”), if not also municipal governments (which were expressly included in the executive branch patent policy, the ruins from which Bayh-Dole is built). The contractor, however, is not restricted to selling only to the “United States”–obviously, and can sell to anyone. Thus, in the government-side market, the contractor competes with the government and anyone the government authorizes. But the government cannot compete with the contractor outside the government market.
But in 37 CFR 401.14(a)(e)(1), title and license are switched. Now the federal government has title, but the contractor has a license that appears to be as broad as to “practice and have practiced” without limitation. That is, under (e)(1), the contractor has the same rights to practice that the contractor would have anyway! It’s just that the contractor is not responsible for the effort and cost to file a patent application and does not hold title to the patent when it issues–but has its squirrelish cheeks nut-full of rights, if only the contractor thinks about using these rights before hand. And, of course, the government by holding title is also free to compete with the contractor in the private, non-federal market.
The license is transferable only with the approval of the Federal agency except when transferred to the sucessor of that party [sic] of the contractor’s business to which the invention pertains
Here the clause appears to contemplate small business contractors. Perhaps the whole clause was intended for small business contractors only. But it is not drafted to exclude nonprofits. This assignment, then, allows a contractor to make, say, an invention management organization the successor in interest to the contractor’s “business” dealing with the subject invention. In the case of a university, this is no different from assigning a subject invention to a research foundation–except here there is absolutely no requirement to pass along as well the public covenant regarding nonprofit dealings with subject inventions, such as restrictions on how licensing income is spent (though the limits in 35 USC 200 still apply–it is, after all, an invention arising from federally supported research or development).
What’s more, there is no restriction that the assignment must be to an invention management organization. The assignment could be to a company, a big company, a company that might have an interest in commercializing the invention and taking over development research. The difference between this (e)(1) approach and the approach by which a contractor “retains title” is that (i) the contractor doesn’t have the cost and bother of filing patent applications and (ii) the deal has to be done upfront, before the funding agreement is in place. There’s a third difference, too, involving the custom march-in procedures for this approach, which we will get to.
Thus, rather than waiting until a grant has been awarded, work has been done, and an invention made and reported, under the (e)(1) approach, a university picks the commercialization partner upfront and enters into an option to license any future inventions to the partner. Since the university is likely already circumventing Bayh-Dole by requiring a present assignment of future rights from all potential inventors (and not requiring the (f)(2) written agreement that might pre-empt that future assignment), why not use exactly the same method with a commercialization partner by committing to what we might call a “present assignment of the university’s future interest in that part of the university’s business to which any future invention arising in the proposed research might pertain”? Sure. A forward-looking arrangement with a commercialization partner built with the same forward-looking arrangements that the university requires of its own inventors.
Now consider–the commercialization partner receives (and pays for, say) an option to acquire the university’s “future business” (that is, licensing of patent rights) in the area of any subject invention that may arise, once the university obtains federal funding for a particular grant. That relationship alone might induce the commercialization partner to contribute to the research proposal–shape the work so that there’s an even greater likelihood of useful information and inventions. This approach may be way more effective at both producing worthwhile subject inventions and getting them quickly developed.
One worry in this approach might be the dreaded Tax Reform Act of 1986, with its awful Revenue Procedure 2007-47, which evidences the both dreaded and awful fingerprints of cleverly clueless university patent administrators advising the IRS about how to bungle its own bungling of nonprofit research assets. The primary problem under 47-2007 is determining if research should result in a finding of “private use” of facilities funded with tax-free bonds. The thing that will trigger that use, for corporate funded agreements, is that the value of the licensed “resulting technology” is determined in advance. That’s easy to navigate (if one is forced to use facilities financed with tax-free bonds and will use more than allowable assignable space for private use)–just negotiate the value of the license after an invention is made and reported. Put a range for that value in the option, and have the price paid for the option be less than the low end of that value, to be credited to the final negotiated price. That ought to work. Of course, you will need to work with your most transactionally savvy attorney to navigate these waters. But there’s plenty of depth to work with, and the rocks are clearly marked.
There’s a second worry, and that has to do with the section (e) march-in. Let’s look at the march-in text next.
(2) The contractor’s domestic license may be revoked or modified by the funding Federal agency to the extent necessary to achieve expeditious practical application of the subject invention pursuant to an application for an exclusive license submitted in accordance with applicable provisions at 37 CFR part 404 and agency licensing regulations (if any).
A federal agency has the right to march-in. This might look pretty grim, but since this is under the umbrella of Bayh-Dole, you must have faith that the drafting will make it difficult for the federal agency to do what appears to be in the government’s interest. But first, note that if your friendly commercialization partner is the one that seeks the exclusive license, then that partner can tie up the rights however it wants. That’s something one might address in the option agreement, too.
A second accommodation of the government’s march-in right here is to achieve practical application expeditiously. That would mean adding some diligence clauses with language suitable to show that no one in the industry could reasonably do anything more expeditiously than what the commercialization partner has committed to do. It is important that the option agreement and license use the phrase “achieve expeditious practical application.” That ought to be enough. But now the walk-back. Bayh-Dole apparatus is nothing without a walk-back on every gesture toward the public interest.
This license will not be revoked in that field of use or the geographical areas in which the contractor has achieved practical application and continues to make the benefits of the invention reasonably accessible to the public. The license in any foreign country may be revoked or modified at the discretion of the funding Federal agency to the extent the contractor, its licensees, or the domestic subsidiaries or affiliates have failed to achieve practical application in that foreign country.
The license becomes irrevocable where there’s practical application. That’s a good start on the walk-back, but we have another whole section to go. A long walk, so to speak, a long march-in:
(3) Before revocation or modification of the license, the funding Federal agency will furnish the contractor a written notice of its intention to revoke or modify the license, and the contractor will be allowed thirty days (or such other time as may be authorized by the funding Federal agency for good cause shown by the contractor) after the notice to show cause why the license should not be revoked or modified. The contractor has the right to appeal, in accordance with applicable regulations in 37 CFR part 404 and agency regulations (if any) concerning the licensing of Government-owned inventions, any decision concerning the revocation or modification of the license.
In essence, the license “retained” in (e)(1) gets mapped to a government license issued under the federal licensing part of Bayh-Dole, which gets implemented in 37 CFR 404. So the standard patent rights clause for the non-profit/small business side of Bayh-Dole includes what amounts to a compulsory non-exclusive license with a conditional right to assign and sublicense granted upfront by the federal agency as a future government license. It’s just that the compulsory federal license in the standard patent rights clause waives nearly all the requirements for federal licensing–the contractor does not have to submit a plan (under 35 USC 209 (f)) to achieve practical application, or submit every sublicense to the federal agency (under 37 CFR 404.5(b)(4)) and the like. The license in (e)(1) is a bare, naked conditional license subject to the possibility of limited federal agency march-in. Follow? Yes, of course you follow.
Naturally the path of appeal of any march-in follows the government side rather than the university/small business side of licensing regulations. It’s just that here the march-in is directed at restricting a non-exclusive license rather than forcing a non-using or under-performing or non-compliant contractor to grant non-exclusive licenses (or even an exclusive license). The appeals procedure is set forth in 37 CFR 404.11–an appeal to an agency head or via alternative dispute resolution. The federal licensing march-in, however, appears to have to follow from a breach of federal licensing requirements, most of which are waived for an (e)(1) non-exclusive license. It’s all rather strange–the condition on which an (e)(1) non-exclusive license might be “revoked or modified” is that the federal government wants to grant an exclusive license and that such a license will “achieve expeditious practical application” of the invention:
to the extent necessary to achieve expeditious practical application of the subject invention pursuant to an application for an exclusive license
It’s just that the requirements for granting an exclusive license under 35 USC 209 include
(4) granting the license will not tend to substantially lessen competition or create or maintain a violation of the Federal antitrust laws;
Thus, it’s difficult to see how a federal agency could even consider revoking or modifying an (e)(1) license unless the contractor and its licensee(s) were doing nothing with the invention–that is, there was nonuse. Otherwise, the agency would have to show how terminating an active non-exclusive license with one more more sublicensees using and/or developing an invention would somehow *not* lessen competition. Among other arguments, there would appear to be plenty of room for appeals to thwart any government march-in, so long as the invention is being used (hence, there already is practical application) or is being developed (and hence, there is competition with any applicant for an exclusive license).
The (e)(1) license, though it is not so stable as holding title to a subject invention, is less expensive by far and readily defensible from march-in attack, so long as there is practical application or expeditious efforts toward practical application. Also rather oddly, it does not appear that the (e)(1) license is exposed to standard 35 USC 203 march-in, as the only condition set out for revoking or modifying is government lust to grant an exclusive license where there’s not been expeditious efforts toward practical application.
There is one more element to the section (e) circumvention. Rather than choosing just one commercialization partner, a university could create a consortium of future licensees, and make the scope of the program any listed federal grant. Each company in the consortium would have an option to acquire a sublicense in any subject invention made within the grants in the program. Whenever a principal investigator wished to have his or her proposed research added to the program, the grant proposal would be added to the consortium option agreement. Now a university has a pool of companies that pay a consortium fee–an option fee–to acquire rights to any subject invention within the program.
There are all sorts of interesting licensing dynamics possible then within such a program. If only one company exercises an option on a given subject invention, then the university and company can negotiate the value of a sole license–a sublicense done just once (all the other terms can be set in the option agreement, but for the garblebuss requirements of Rev Proc 2007-47). In this fashion, one can create a marketplace for licensees before there is any invention to “market.” If companies want, they can request that the university acquire title and elect to retain title to any given subject invention, and deal with the university under the normal Bayh-Dole provisions in the standard patent rights clause and be exposed to the standard march-in requirements under 35 USC 203 (if the license is exclusive). But the companies could also wait and see if the federal government will take title, and if so, then if the federal government obtains a patent, the companies exercising their option get a non-exclusive license that’s even better than any license they’d get under the normal nonprofit/small business contractor Bayh-Dole arrangement or under the heavy federal licensing requirements on the federal side of Bayh-Dole.
Within any opportunity to acquire a license to a subject invention, then, companies in the option consortium could have various alternatives. A given invention could be offered FRAND–non-exclusively, under fair, reasonable, and non-discriminatory terms, so everyone gets the same deal (and therefore does know the deal that anyone else also is getting–no need for secrecy). Or an invention could be offered under a bid-for-value license, in which option holders bid for first access (say, a six month lead) or for what the FRAND value should be (a high FRAND value might limit how many companies exercise their option, and so give standing to the companies that really want to work the invention). Or an invention could be offered on a sole basis–and companies bid (if there are enough) on that basis. Of course, if the sole license is offered as an (e)(1) sublicense, anyone losing the bid might go applying to the federal agency for an exclusive license, unless that’s something that’s also anticipated in the (e)(1) option.
Of course, the federal government operating under (e) has the right to grant licenses to anyone without having to use the march-in procedure in (e). But that government right is itself limited by Bayh-Dole. For instance, the government, when it obtains title, may only grant licenses “in the public interest” (35 USC 207(a)(2)):
(2) grant nonexclusive, exclusive, or partially exclusive licenses under federally owned inventions, royalty-free or for royalties or other consideration, and on such terms and conditions, including the grant to the licensee of the right of enforcement pursuant to the provisions of chapter 29 as determined appropriate in the public interest;
If a contractor is already granting licenses and achieving expeditious use of a subject invention, then any determination of the “public interest” has to take this activity into account. If the government grants its own licenses (say, undercutting the contractor’s own efforts), one might argue that the government’s action is not in the “public interest.” There’s more. Bayh-Dole places limits even on the government’s ability to grant “partially exclusive” and “exclusive” licenses:
(1) granting the license is a reasonable and necessary incentive to—(A) call forth the investment capital and expenditures needed to bring the invention to practical application; or (B) otherwise promote the invention’s utilization by the public;
If the contractor is using the (e) exploit, then condition (A) fails, since exclusivity (or partial exclusivity) is clearly *not* necessary to call forth the investment capital. And (B) fails as well, because if companies are taking licenses under (e), then any government-granted exclusive or partially exclusive license is also not necessary. (We will leave the goofiness of “partial exclusive” for another time).
One more limitation (among others) on the federal government’s ability to grant an exclusive or partially exclusive license:
(4) granting the license will not tend to substantially lessen competition or create or maintain a violation of the Federal antitrust laws;
If a contractor is exploiting (e), then any federal government license that would undermine the contractor’s licensing program could be considered anti-competitive–especially if the government granted a royalty-free exclusive license that had the effect of undermining the licensing program undertaken by the contractor. Thus, even though the government has in theory an unrestricted right to grant whatever licenses it wants, Bayh-Dole restricts that right, and once a contractor uses the (e) exploit, those Bayh-Dole restrictions on the federal government become even tighter.
Of course, if a contractor grants a sole license under (e) (a non-exclusive license done once), then the licensee might gain, effectively, an exclusive commercial position by obtaining an exclusive license (or, as it were, partially exclusive license) from the federal government for the same invention. With some patent stacking twiddle, the licensee can come away with all the rights it can carry in its cheeks without paying more than it would for an exclusive license from either the contractor or the federal government. Sure, a bit more work, but there’s a good argument that the federal government might have to grant that license, under Bayh-Dole’s requirements, rather than hold it back or try to undermine the licensee’s position.
In any case, one can see how this circumvention is built into the apparatus of Bayh-Dole. It involves pretty much the same rights as the more well known approach (electing to retain title) but trades off some risks (cost of patenting, nonprofit use of licensing income, restrictions on assignment, 35 USC 203 march-in) for others (federal patenting and licensing practices–feds could blow the patent rights or license royalty free, the restricted march-in under (e)(2)). Under the (e)(1) license, a university still may grant a sublicense with the same commercial scope as under the title-first approach–it’s just that the sublicense will not involve an assignment of the invention (as many university conventional exclusive licenses are) and will require “expeditious” efforts at practical application. The government, of course, could grant its own non-exclusive licenses–but with stiffer terms (requiring a plan, and the like). The exclusive commercial positions that matter, then, would not be based on the original license, but in any other patent positions that any licensee already has or can develop and protect by patent before other licensees figure out what has to be done. If the set of patent positions in background technology and improvements taken by competing licensees becomes a tangle, then the competitors can work out standards or cross-licenses to get on with things–or stare at each other for twenty years, having soiled their own stinkin’ dens.
The most interesting part about the (e)(1) circumvention, however, is the idea of building one’s base of company participants before there’s any invention–even before there’s any federal funding agreement. Build that company base because you want to partner with these companies around the benefits of federal research, not just when you have a patent loaded and cocked as a threat to industry that attracts speculators who thrive on monetizing such threats. An (e)(1) based practice, even if it spills over into the conventional elect-to-retain-title practice, presents a completely different and completely workable way of managing subject inventions.
With an (e)(1) circumvention, a contractor avoids paying for patent work and gets the benefit of potentially many company relationships, which may then become patent licensees willing to pay even more, if only “reasonable” amounts. (I know, that “reasonable” is a big let down for the monopoly minded, but hey rack it up to putting public interest ahead of institutional interests).
One could imagine a version of Bayh-Dole in which (e) was the only requirement. The federal government took title to subject inventions and contractors could elect a license in the invention, which they could then sublicense as they pleased. The government would restrict its use of the patent to the government-side market for procuring government goods and services, and would only extend that license if the private-side market needs were not being met. If the government chose not to patent a subject invention, then that was that–everyone had access. But if the government did patent an invention, then it paid the patenting costs and controlled the construction of claims, and issued itself the shiny new monopoly, which it then would break up on request to permit a contractor to exploit the invention itself, or sublicense rights to others.
In this (e) version of Bayh-Dole, universities would not have to go out of their way to force inventors to assign their inventions made with federal support to the university–the assignments would be directed to the federal agency that funded the work. University technology transfer offices would be out of the patenting business–their focus would be on sublicensing federal patent rights. Companies doing business with the government could obtain a license directly from the government, of course, and so would need a sublicense from the university only for the non-government market.
Much of the present apparatus of Bayh-Dole would drop away, if (e) were king. There would be no need for all the requirements on patenting–those would be government decisions. There would be no need for reporting on utilization and march-in, because the government would have the rights it needed to promote use itself, if use was so very important. There would be no need for much of anything.
The government sponsor, not host contractors, would decide whether a patent would promote public interests. If it would, the government would issue itself a patent and promptly break that patent right up when the contractor requested its (e) license. It would be the contractor, then, who would then decide whether the public interest was best served by a private monopoly (in which case companies would have to bid for that right from the government), or by competition (in which case the contractor with its (e) license would be one competitor). If industry found it easier to work with a university contractor with the right to sublicense than with a federal agency, then industry would beat a path to the door of the contractor’s licensing office.
A further consequence of this approach is that university licensing offices would have to rig for non-exclusive licensing protocols. That would be a good thing, as non-exclusive protocols are much better aligned with academic research and public missions free from institutional conflicts of interest. Non-exclusive licensing protocols also attract many more players and way fewer schemers and trolls. A non-exclusive license is not much use to a troll or a speculative monopolist.
From this analysis, then, you might begin to see how things could be different, using the apparatus that’s already present in Bayh-Dole’s implementing regulations, but not in Bayh-Dole. Given that the commercialization rate with Bayh-Dole appears to be 100x lower than the best rates recited for non-federally supported inventions managed by independent agents prior to Bayh-Dole, perhaps there is a good reason to consider alternatives to Bayh-Dole–if only as a competing strategy. Let the investigators proposing federal research choose which model they want–Bayh-Dole 202(a) contractor title or Bayh-Dole patent rights clause 401.14(a)(e) government title. All one would have to do to streamline the implementation of (e) is to remove the limitation that rights have to be legally obligated before the grant is awarded. Perhaps make it that the rights have to be legally obligated before the grant officially ends. And that change does not require an amendment to Bayh-Dole, since it is simply part of the implementing apparatus and can be done following the procedures for executive branch regulatory changes, with the opportunity for public comment.