Let’s examine the 1984 amendment to Bayh-Dole’s secrecy provision for invention use reports in more detail. It is worth pointing out that Section 202 of Bayh-Dole has been routinely amended and it would be worth at some point to show the general nature of the amendments–most of which appear to remove public protections and agency power to exercise oversight over the work of patent brokers gaining access to federally supported inventions.
The provision regarding non-disclosure of invention use information shows up in a list of clauses that must be included in any funding agreement for research:
(c) Each funding agreement with a small business firm or nonprofit organization shall contain appropriate provisions to effectuate the following: . . .
The actual wording is not specified–just that there must be “provisions to effectuate” what follows. (Please, avoid phrases such as “provisions to effectuate” with your families. These are not nice words.) It is important to see this fundamental move by Bayh-Dole. The law is directed at federal agencies, but instead of setting things out directly, Bayh-Dole specifies what a patent rights clause must do–but it leaves that doing to be drafted outside the law. If you want, it is a back door built into the facade of the law. The law can make whatever gestures it wants, but the actual impact of the law will come with how the standard patent rights clause is constructed, and that construction takes place without Congressional oversight or approval.
Bayh-Dole then includes at its core the freedom for the drafters of the implementing regulations (it turns out, funny, the same folks drafting Bayh-Dole itself) to mess around with the requirements after Congress has approved the basic idea. Given these folks could have included the language they really wanted right here in the statute (they were the same folks, after all), and didn’t, should tell us something. I wonder if this clever bit in Bayh-Dole also has been exported to all those other countries adopting a “Bayh-Dole” law. One might compare the Kennedy patent policy, where requirements for agency contracting for invention rights are set out with some definition, to be walked back when there’s a matter of public interest that takes precedent.
At 35 USC 202(c)(5), Bayh-Dole declares that information on utilization of subject inventions is excluded from federal public disclosure law. Here we go, with the 1984 amended law:
. . . effectuate the following:
(5) The right of the Federal agency to require periodic reporting on the utilization or efforts at obtaining utilization that are being made by the contractor or his licensees or assignees: Provided, That any such information as well as any information on utilization or efforts at obtaining utilization obtained as part of a proceeding under section 203 of this chapter shall be treated by the Federal agency as commercial and financial information obtained from a person and privileged and confidential and not subject to disclosure under section 552 of title 5.
The part in red is FOIA language, and so depends on what FOIA has to say. Bayh-Dole declares here that the standard patent rights clause must include for federal agencies the right to require reporting on use or efforts for use of subject inventions. How federal agencies come to have such a right is not indicated. Of course, in a freedom-to-contract world, whatever someone insists upon, and gets in a contract, becomes an “enforceable right.” But that is a right to the benefit of another’s promise. Not a natural right. Not even a statutory right. Here, Bayh-Dole dictates what can be in a federal contract with regard to patent rights. But it goes further and also dictates what can be in a federal contract with regard to how patent rights are used to promote the use of subject inventions. And how reports of such use (or nonuse) are to be treated.
Think about the situation. Bayh-Dole requires the patent system to be used to promote the use of subject inventions–established use with public benefits on reasonable terms. If there is no such use, or is unreasonable use, then Bayh-Dole claims to permit federal agencies to march in to protect the public. Reporting, then, is the fundamental means by which a federal agency monitors what is going on, and whether the policy and objectives of Bayh-Dole are being met. If those reports were for the most part public, then there might be pressure on federal agencies to intervene more readily when there was no apparent effort to develop a given subject invention–no licensing, no purposeful marketing, no investment of private funds, no development, nothing heading to market, no private use outside the lab. But if the reports are entirely secret, then the whole operation of the law is, essentially, secret, but for the parts that contractors and their supports wish to promote. So they repeat that Bayh-Dole has been a success, as a sort of mantra, knowing that the law itself does not permit the public from having access to the report data to ascertain if their claims hold up. Bayh-Dole in this version creates the environment for exploitation of the law without a public means to check any excesses or unintended consequences.
Since the use of patent rights to promote the publicly beneficial use of subject inventions is the heart of Bayh-Dole, how patent rights get used is the key metric of the law. Bayh-Dole, then, is about more than letting contractors acquire title to federally supported inventions: it is also about the promotion of use of those inventions. And further, it is about the public benefits on reasonable terms arising from the use of those inventions. And finally, it requires federal agencies to contract for inventions such that they have the right to require reports of the use of these inventions. Bayh-Dole encompasses all four elements–ownership, promotion, publicly beneficial use on reasonable terms, and reporting of use. What’s odd, then, is the amendment that shifts “may” to “shall” with regard to non-disclosure, as that significantly changes the scope of the law.
In the original Bayh-Dole statute, federal agencies were to contract to obtain reports of invention use but were authorized to have the discretion to withhold information from public disclosure if it met the conditions of 5 USC 552–FOIA. That is, in the original Bayh-Dole, the effort was to make it clear that agencies did not have to make all the information in such reports public. Put another way, the original wording aimed to defeat the idea that all report information must be publicly disclosed. Some did not have to be disclosed, if it met a given standard under FOIA. Amended Bayh-Dole, by contrast, appears to make mandatory the withholding of all information in reports by replacing “may” with “shall.” That is, amended Bayh-Dole aims to defeat the idea that some report information should be made public as a matter of reporting the effects of the law and contractor diligence with promoting the use of subject inventions.
Who advocated for this change? Who stood to gain? Did anyone point out the significance of the change? The new intent? If the purpose was simply to make sure that agencies did not disclose sensitive commercial information (budgets, release dates, applications), the original language already covered that issue, so no change was necessary. The effect of the change, apparently, was to prevent federal agencies from having any discretion to determine what constituted information that was subject to FOIA.
Here’s what Sen. Thurman’s report on what would become the 1984 amendments (in S. 2171) had to say:
Subsection (6) provides assurance that agencies can protect information provided to the Government by contractors on their invention utilization efforts.
Here’s what subsection (6) provides, by way of amendment, in PL 98-620:
(6) by striking out “may” in section 202(c)(5) and inserting in lieu thereof “as well as any information on utilization or efforts at obtaining utilization obtained as part of a proceeding under section 203 of this chapter shall”
The “may” to “shall” change slips in with the expansion of confidentiality to include information obtained in march-in procedures. The idea then might be that march-in procedures might require divulging confidential information that otherwise would not be required to be reported under 35 USC 202(c)(5), and this additional information is better treated by “shall” than “may.” Imagine march-in as a secret trial with all the evidence sealed away.
Perhaps a march-in procedure is seen, then, as an adversarial relationship, more like an accusation against the contractor’s efforts than a positive decision by a federal agency to step in and address a public need where there’s no obvious indication that the contractor has accomplished anything by holding a patent monopoly. If march-in is more like “step-in,” then the process involves enlisting the contractor (and assignees and exclusive licensees) to assist the federal government, or if they are unwilling, the government may go it alone. In this “step-in” view, the contractor never had “ordinary” ownership of a patent right–just a first shot at doing something better and more timely than the government could do, and if that first shot misses, or doesn’t get taken, or takes years of fussiness to even get attempted, then if the government wants something done, it ought to be able to step in and do it. After all, the government also has a broad license to “practice and have practiced”–the government does not need to “step in” at all. If it has the authority from Congress, it can do pretty much whatever it wants with an invention directly, or by contracting with others to do the work for it. “Step in” in this view is a gesture to allow a contractor to remain the primary point of contact, and perhaps to earn some money from reasonable, non-discriminatory non-exclusive licenses as directed by the federal government. “Step-in” rights = nice federal government, helping poor contractor who hasn’t done anything.
One might see, then, the addition of “shall” as a defensive measure by folks more interested in making sure that a contractor (and the contractor’s beneficiaries) keep full control over inventions without federal intervention for any public purpose. Not just no “march-in” but no “step-in” neither. Yes, there are public purposes enumerated for march-in, but a close reading combined with the implementing regulations make it clear that the threshold to establish a basis for government action is nearly impossible to trigger in the time frame in which a useful action would need to be taken.
Think of it this way. If march-in can be delayed so long that any federal intervention is unlikely to address a pressing public need, then the rationale for the federal intervention ends up appearing merely to penalize and disrupt the contractor (and ilk), not to address any public need. Only if the march-in is immediate is public need foremost. Once the march-in procedures become so difficult to implement that a contractor can delay action for an extended period does it become apparent that march-in has become, primarily, an adversarial proceeding. That appears to be the effect of changing “may” to “shall”–something that was made to be agency discretionary becomes, apparently, an absolute requirement.
Apparently (again) all information in any report is to be subject to withholding under FOIA. But my use of “apparently” is deliberate, because it is not all that clear that Bayh-Dole, in its amended form, follows FOIA requirements. Again, if the words don’t much matter, we were done a long time ago. Do WTF you want, so long as you don’t bother a federal agency or a powerful person who might stir a federal agency to the point of bother.
Bayh-Dole’s 1984 amendment changes a provision that makes clear that federal agencies may withhold the parts of utilization reports that are confidential (or privileged) to one that mandates that reports in their entirety are to be treated as confidential and privileged information and not subject to public disclosure under FOIA. That is a big change. The question is–does the change comply with FOIA requirements for how federal agencies may withhold information? That is, does Bayh-Dole comply with FOIA? Let’s look, then, at what is necessary for a federal law to exclude information from the Federal Freedom of Information Act.
The 1984 amendment preserves the sentence structure in the original version of Bayh-Dole appropriate to agency discretion but changes the modal verb–may to shall–and thus changes the text from a clarification that an agency does not have to release all use report information under FOIA to a requirement that it must not release any information.
The clause previous to 202(c)(5) uses the same “Provided” construction, and it is helpful to consider it to gain an idea of what is going on in 202(c)(5), and why the change from “may” to “shall” matters. In 202(c)(4), Bayh-Dole requires the standard patent rights clause to include a provision that requires the federal agency to have license to any subject invention to which a contractor “elects rights.” The clause continues:
Provided, That the funding agreement may provide for such additional rights, including the right to assign or have assigned foreign patent rights in the subject invention, as are determined by the agency as necessary for meeting the obligations of the United States under any treaty, international agreement, arrangement of cooperation, memorandum of understanding, or similar arrangement, including military agreement relating to weapons development and production.
The “Provided” clause authorizes an agency to include in a funding agreement any “additional rights” to meet federal international obligations. The “may” here is interesting. Why not “shall”? Federal agencies are left with the discretion whether to worry about compliance with international obligations where a license would not be sufficient. Bayh-Dole does not mandate that federal agencies include requirements for additional rights beyond those of a non-exclusive license to practice and have practiced. But the “Provided” qualification leaves it open for federal agencies to do so on a case-by-case basis. Not exactly a uniform provision, but there it is. The standard patent rights clause, 37 CFR 401.14(a), does not include this provisio. A federal agency would have to insert it as part of tailoring the patent rights clause for a given funding agreement.
Now consider the next provision, our 202(c)(5). Originally, it had the same structure. The funding agreement must provide the federal agency with the right to require “periodic reporting on the utilization or efforts at obtaining utilization” made by the contractor, assignees, or licensees. That requirement is similarly qualified by a provisio, that the agency may treat the information as exempt from FOIA. The clause here does not state that this provisio is to be included in the funding agreement–it’s just a free standing statement that a federal agency is not obligated to disclose information received in response to the federal agency’s right to require reporting. The agency may treat information received as “commercial and financial information obtained from a person and privileged and confidential and not subject to disclosure under section 552 of title 5.” To be able to treat information in this way, of course, the information would have to meet the requirements pertaining to such information under FOIA. That’s what the “may” signifies.
Just as with the previous clause, 202(c)(4), our provisio in (5) states a contingency, something that provides an agency with flexibility that might not be otherwise present in a funding agreement but might be necessary for the agency to meet obligations under other authorities–in this case FOIA rather than a treaty or other international agreement. Makes sense.
Now consider the 1984 change to 202(c)(5):
shall be treated by the Federal agency as commercial and financial information obtained from a person and privileged and confidential and not subject to disclosure under section 552 of title 5.
The “shall” turns the provisio into a requirement rather than a statement of agency discretion necessary to comply with the law. Now we have something altogether different. Any information received comes within this requirement and must be treated by the agency as if it is “commercial and financial” information that is both “privileged and confidential.” Thus, the provisio goes on, the information meets a disclosure exemption under FOIA.
On the face of it, there is nothing at all that would indicate that information pertaining to the use of a subject invention by a contractor must necessarily involve any commercial or financial information. We are dealing principally with nonprofit contractors here, and while nonprofit standing doesn’t have anything to do with deciding if information might be “commercial,” the idea that a nonprofit might seek “commercial” dealings with its patent rights is not a requirement of Bayh-Dole. If a nonprofit negotiated or received commercial information from a licensee–a royalty rate, or a report of sales–that information would be “commercial” and may well be “confidential” (depending on other factors, such as the operation of state public disclosure law in the case of a public university).
But the fact that, say, a subject invention has not been licensed is not (in normal usage–who knows what a court might decide?) commercial information and not the kind of information that is confidential–universities routinely advertise inventions that are “available for licensing.” Reporting to the government that an invention has not been licensed is not, on the face of it, generally confidential. Similarly, if an invention has been licensed or assigned, a university would report in response to an inquiry that the invention is not available for license from the university. That fact would be open and public, not secret. And yet here we have a clause in a law that now directs federal agencies to treat all such information as confidential: whether an invention has or has not been licensed, and whether the invention has or is being used, or developed for use, and whether the benefits of the use are available to the public, and if so whether on reasonable terms. If benefits are available to the public, is not the fact of such benefits also public? If the public receives benefits, are not the terms also public? Is there anything possibly confidential in reporting to the government these apparently public facts?
The altered Bayh-Dole declares, apparently, that all such use information is to be treated as confidential. The shift from “may” to “shall” turns Bayh-Dole into a law that prevents the public from assessing whether the law actually accomplishes what it states as its objectives. That appears to be the case, at least. But let’s look carefully at the mechanism that Bayh-Dole uses. Bayh-Dole does not declare invention use reports to be not disclosable. To do that, all 202(c)(5) needs to do is:
Provided, That any such information shall not be disclosed by the Federal agency.
That would park use information squarely within FOIA’s exemption at (b)(3), just as Bayh-Dole does with information pertaining to reports of subject inventions and patent applications at 35 USC 205. But Bayh-Dole does not do that here at 202(c)(5). Instead we get roundabout language–language that wasn’t roundabout at all when the modal verb was “may” rather than “shall.” The question then is whether the ruins of the old clause–the now roundabout language–actually does what it gives the appearance of doing.
I know, this is a “legal” question and so won’t be decided without a dispute getting to a judge and I don’t have the quarter million dollars to put the question. So instead we will treat the matter as one for interpretation, using such tools as reason and an unceasing commitment to the idea that laws regarding the disposition of inventions and discoveries made in publicly supported research ought to be readable by the public that is doing the supporting. There may be public policy reasons to argue that laws on such matters should be so sophisticated and complicated and opaque that only a special priesthood of lawyers have any right to pronounce on them–if so, then let someone else trot out that argument and we can deal with it then. Meanwhile, let’s work the situation as best we can, with the resources available, which at this point don’t include attorneys.
In the quote of 202(c)(5) in its amended glory, I have highlighted the little word “as” for further consideration. The requirement is not well stated: the patent rights clause apparently is required to specify that even information that is not confidential and not privileged is, apparently, to be treated in the same way as information that is confidential or privileged. That is the effect of the “as” that’s left over from the original text of Bayh-Dole–“may be treated as” becomes “shall be treated as.” If there was no difference in the effect of the words, there would be absolutely no reason to change the words. There, then, must be a difference in effect of the words.
(One might argue that university administrators “understood” Bayh-Dole to mean “shall” and the law was amended to comply with their understanding–but then we are left with problem that the meaning that matters is what university administrators intended, not what Congress intended. This same nutsiness was in play in Stanford v. Roche, where a blot (let’s call them) of university attorneys claimed that the law must be interpreted as they wanted to read it, and the Supreme Court said, no, what matters is what the words say and Congress intended, and took the whole blot of them to the woodshed for a paddling, from which they emerged unrepentant, declaring that the case had been wrongly decided).
In the original statute, a federal agency may withhold use report information if that information is confidential (such as financial information subject to trade secret claims received from a licensee) or privileged (such as information exchanged with legal counsel, such as regarding possible infringement or regulatory actions). In the 1984 amendment, a federal agency apparently must withhold report information even if it isn’t confidential. That is, the effect of the “shall” on the reporting requirement on the patent rights clause is to turn information that is not confidential or privileged into information that is confidential and privileged, simply by being provided to the government. This is very strange. Is it the correct interpretation?
There’s no positive requirement under Bayh-Dole for federal agencies to require reporting of how owners of subject inventions use the patent rights they acquire. Federal agencies could–and do–fail to require reports, and perhaps (do you think?) don’t read them when they get them, or do much of anything if they don’t like what they read. Perhaps their reasoning now is, since no one else will see these use reports, why should the federal agency care what’s in them?
Why even require reports at all, if practical application does not matter and there’s no term limit on exclusive licensing?
In describing related rule-making surrounding President Regan’s executive order expansion of Bayh-Dole to large company contractors, Norman Latker argued that waiving various government rights was a good thing:
In addition to mandating contractor ownership, the memo also authorizes the agencies to waive any of the rights retained by the government or the obligations of the performer if the agency determines that this is in the public interest or the contract involves a substantial contribution by the contractor for the work undertaken. So, an agency could, for example, waive its license to use for mission purposes, its reporting requirements, the march-in rights, et cetera, under the circumstances spelled out in the President’s statement.
Thus, while Bayh-Dole mandates that a number of requirements be placed into the standard patent rights clause with the expectations that doing so creates a uniform program of patent management across all agencies, the law is written with the expectation (apparently), that a federal agency could waive requirements placed in a funding agreement–requirements it could not waive if made a matter of statute. Thus, apparently, a federal agency can choose not to require periodic reports on utilization and be done with it, despite the requirement that reporting is a contractual right to be included in each funding agreement. Of course, if agencies vary on such matters as government purpose rights, march-in, reporting, and the like, then the only thing that remains “uniform” about the policy is that contractors can keep title to subject inventions when contractors acquire title to subject inventions. The rest of it–enforcement of anything in the standard patent rights clause–is a matter of agency discretion.
Bayh-Dole creates the impression that there are checks and balances for contractor exploitation of inventions made by, say, faculty members, with federal support, but really there’s nothing firm at all. Agencies are set up not to care, because in most any given circumstance that matters, they don’t have any statutory obligation to enforce the terms of a funding agreement. That, at least, is the idea put forward by Latker, and Latker drafted Bayh-Dole and oversaw the drafting of its implementing regulations.
Is the reporting requirement then merely necessary to preserve the appearance of public oversight, even though there is, in effect, no public oversight? Had Bayh-Dole been proposed as a law to vest inventions made with public funding in the contractors that hosted the research, without any requirements on how those contractors then managed those inventions (and with no concern that the contractors have any capability to manage inventions), and no public oversight, do you think Bayh-Dole would have passed? Probably not. So they put in all these trappings of public accountability and public interest to get the law passed–but put them as contract requirements rather than statutory requirements, so agency by agency folks could back off the requirements as they wanted, without bothering Congress. It’s clear that the 1984 amendments that removed limits on nonprofit exclusive licensing to big companies was driven by university patent brokers and pharmaceutical companies:
This bill has several important effects on all universities that conduct Federally-funded research. It eliminates the existing limit on the duration of exclusive licenses they can grant to large businesses. This factor is particularly important with technologies such as pharmaceuticals, where long development times and major investments are usually required prior to commercialization
That, from Sen. Thurmond’s report.
The 1984 and later amendments have successively removed the underlying mechanisms that would enable public accountability–even as the law is set up, apparently, to let agencies sleep on their rights. The justification is that pharmaceutical companies require fewer restrictions, and therefore everyone must have fewer restrictions. What’s good for big pharma monopolies must be good for everyone’s monopolies, must be good for America. It’s a familiar refrain.
It has taken over thirty years, but when NIST “clarifies” the (f)(2) provision in the standard patent rights clause to conform with another blot of university attorneys asserting that (f)(2) is “really” an assignment provision, we will be pretty much there–with Bayh-Dole being a shell of what it once was and now simply an instrument to take ownership of inventions made with federal support away from inventors, investigators, and the federal government and handed to patent brokers to make what money they can from them. Actual use, or public benefits on reasonable terms, is merely collateral effect.
This is a study in ten parts. Here are the links.