Illusions of Bayh-Dole: Bayh and Dole argue for unreasonable drug pricing

Here’s an article by Birch Bayh and Bob Dole published in the Topeka Capital-Journal in 2002 with the headline “Bayh, Dole: Drug price controls hurt, and weren’t intended.” In the article, the co-sponsors of Bayh-Dole aim to rebut Peter Arno and Michael Davis, who propose that Bayh-Dole supports price controls on prescription drugs made with government-supported inventions. After the usual unsupportable praise for the success of the law, we hit the nub:

The purpose of the Bayh-Dole Act was to spur the interaction between publicly funded research and privately funded medical research. Our hope was that patients would receive the benefits of innovative science sooner, and that is just what the law has done.

I have the sad duty to point out that in 2002 there was absolutely no public evidence that Bayh-Dole had sped up the development of biomedical products. First, Bayh-Dole’s patent rights clauses, as we have repeatedly pointed out, have not been complied with or enforced, nor has the federal government acted on any of its rights, such as its license or its march-in rights. No one has gathered data with regard to the speed with which innovative science has passed through institutional ownership of inventions, a patenting process, marketing of assets for exclusive license (or assignment), and any subsequent race to produce product for suffering humans despite Bayh-Dole’s apparent blessing of the idea of a patent monopoly to prevent competition during and after the effort to create a mass-produced commercial product. AUTM did not and does not gather such data. AUTM does not even break out subject invention reporting from other university commercialization activity. If federal agencies gathered such data in 2002, they were not reporting publicly (and they could not reveal the details anyway, since Bayh-Dole purports to make all such information excluded from FOIA disclosure). Thus, although Bayh and Dole make a confident assertion of their law’s success, there was nothing to back it up but the confidence by which they presented their view.

We have argued here at Research Enterprise that the purpose of Bayh-Dole is to push monopolies on publicly funded research to the pharmaceutical industry. In their article, Bayh and Dole admit as much. In their Senate speeches on the introduction of S. 414, which laid the foundation for Bayh-Dole, both senators called out biomedical inventions as a focus. Before that, Norman Latker, the NIH patent lawyer who surrepetitiously drafted Bayh-Dole, used the Harbridge House report’s discussion of the pharmaceutical industry’s objection to the Public Health Service’s policies on inventions as the basis to restart the Institutional Patent Agreement program, which provided an end-run around executive branch patent policies, funneling patent monopolies to pharmaceutical companies using universities and their nonprofit foundations as the cover. Bayh-Dole was conceived when the IPA program was shut down when Congressional reviews showed it was doing sweetheart deals with pharma and wasn’t doing that well even at that. 

Bayh and Dole continue:

Our legislation was based on a broad bipartisan understanding that government-funded research was not benefiting the public before the 1980s because it was not being turned into products for public use.

It may have been an “understanding” in Congress, but the facts don’t support the creation of that understanding. The federal government’s licensing rate was reported as 4% while university licensing rates were reported as 25% to 33%. But the facts are different. The federal government’s licensing rate for biomedical inventions was 23%. The university licensing rate for government-supported inventions–almost all biomedical–under the NIH-led IPA program was 4%. Most of the 28,000 patents owned by the government in 1978 (Senator Bayh rounded up to 30,000 in his Senate floor remarks on S. 414) were defense-related inventions that the contractors involved had chosen not to own.

Government patenting served to publish the inventions, and most anyone could get a license to make, use, and sell. Since government licenses were free, few bothered with the formality. Commentators since the 1950s on have noted that there was no way to determine how broadly government-owned inventions were being used based only on licensing figures, given much use was simply undocumented. In many cases, too, the only customer for much of this stuff would be the government and thus there was no “commercial” product to develop, at any cost. What was the market for commercial nuclear weapons? For commercial space capsules?

So Bayh and Dole spun the facts into a selective fantasy rhetoric and created a “broad bipartisan understanding.” The understanding, of course, was true. It’s just what the congressmen thought they understood was not true, contrary to the evidence, false. But, hey, bluffing is part of politics. We can’t whine about it. The point is to understand who the beneficiaries of the bluffing were and are.

We wanted to promote the cooperation that can expedite the private sector’s commercial application of basic ideas pioneered by government and university scientists.

No doubt their expression of wants is accurate. But the words use cover for the reality. For “cooperation that can expedite the private sector’s commercial application” read “pass patents on publicly funded inventions directly relating to public health to favored pharmaceutical companies, which then may charge monopoly prices that exploit suffering.

Also missing is the massive effort by universities to strip inventors of any rights in their inventions–including the right (as Jonas Salk had) to release inventions to the public domain, or to choose to assign their inventions to the federal government, which would then release them for all to use, including the inventors.

And we get the basic historical untruth:

Before enactment of Bayh/Dole, virtually none of these discoveries were ever commercialized.

Before Bayh-Dole, there was the IPA program specifically built by the NIH to circumvent executive branch patent policy. If the IPA program–over 70 nonprofits participating–was an abject failure, and Bayh-Dole re-implements the IPA program as patent law–then it surely wasn’t Bayh-Dole that has made any difference. And the fact is, the IPA program was about as effective as the general average for patented inventions generally. Not better or worse. Put it this way, universities and nonprofits have obtained about 51,000 utility patents citing government support since 1981. The commercialization rate is awful.

Finally, it’s understandable that Senators Bayh and Dole may have wished to “expedite” commercial application of basic ideas “pioneered” by government supported scientists. But nothing in Bayh-Dole sets up to encourage or evaluate the speed with which commercial products appear. In the original Act, there was a preference for small businesses that limited the term of exclusive licenses granted to large (i.e., non-small) businesses, but that provision was removed almost immediately (in 1984), and Bayh-Dole’s reporting requirements don’t make any attempt to measure the speed with which privately owned inventions make it to commercialization. There’s no data available. AUTM certainly wouldn’t dare to collect it let alone report it.

In the area of biomedical research, government-funded research is clearly significant, but that alone has never brought new medicines to patients. Without the pharmaceutical and biotechnology industries’ resources and advanced applications, drug development as we know it would not exist.

This statement is only true with the addition of “as we know it.” The assertion is therefore absolutely true–without what we have we would not have what we have. That’s the gist. But the rest of the statement is simply not true. Before Bayh-Dole, we had many government-supported successes in “new medicines”–we had anti-virals and penicillin, and treatments for polio, leukemia, and other cancers. Indeed, these successes spurred government involvement in medical research, and it was that involvement that stepped on the toes of the patent medicine industry, which was also racing to engage university scientists and establish what everyone hoped would be lucrative monopolies. Freely available drugs, such as insulin (developed at the University of Toronto and licensed non-exclusively) were simply lost opportunities to speculate on the value of pressing public health needs.

It is clear that drug development as we know it would be unlikely without government complicity in creating private monopolies on medicines. It is precisely the great bitterness we have with regard to present drug development–its expense, its delays, its monopoly pricing, its penchant to turn acute conditions chronic rather than cure or prevent, its lack of innovation and development, and its relatively dismal record of failing to provide significant relief (many cancer drugs only delay the inevitable, with awful side effects, with the business proposition to extract a few additional months of payment). No wonder the effect of Bayh-Dole is put in such a happy vocabulary of “cooperation” and “advanced applications.” No question that the pharmaceutical industry has “advanced” stuff. Lord knows what we would have now as public resources if we had not through trillions of dollars into their patent monopolies.

It is estimated that for every $1 spent in government research on a project, at least $10 of industry development will be needed to bring a product to market.

That’s because industry development is so very expensive, and because the comparison between research on a given product and industry claims of the cost to develop a commercial pharmaceutical product have pretty much nothing to do with each other. The government expenditure may be over three years, while the industry expenditure is over a decade, and while the government expenditure involves people paid modest salaries–especially post docs and grad students–the industry expenditure pays top dollar for scientists and advertising executives. A recent study showed that a major drug company paid out more in advertising and administration than it spent on research and development. How much spending did Jonas Salk have to do with the polio vaccine to get it into use? Virtually none, of course. But now the cost of commercial advertising must be factored into the “drug development as we know it.”

Moreover, the rare government funded inventions that actually become products are typically 5-7 years away from being commercial products when private industry gets involved.

Dreamy fantasy. Let’s distinguish practical application from commercial product, a distinction actually made by Bayh-Dole. A great many government funded university inventions are processes, compositions of matter, and research tools. Almost any of these can be used immediately–and are, in research settings. There is absolutely no necessity that these inventions become products before they can be used. And there is little sense to expect research tools to become “products” rather than to do any sort of other thing–assist in the analysis of results, accurately measure, improve the performance of, aid in the development of, and the like. It is dreamy fantasy to set forth the idea that each invention made in a research setting must become a commercial product or the public fails to benefit. This is even true for biomedical inventions. While some are complex and dangerous, others could be made and used by individuals with only a little guidance. We see this happening in the area of food supplements, for instance, where patents do not figure so prominently and the pharmaceutical industry has less involvement. Patents held for licensing exclusively as monopolies destroys immediate research practice and technologist use of new research inventions.

In areas such as information technology, an invention that’s not used in three years will likely never be used. To get “private industry” involved by means of patent monopolies virtually assures an invention will be used only to raise money from impressionable speculators. Patents held for licensing slow things down. University licensing of patents on government inventions is among the worst, slowest, least effective pathways to make an invention available for practical application or commercial development. That’s the finding of the Harbridge House report, from 1968. No one has bothered to do such a comprehensive, insightful, even honest report since. The open source software movement does not necessarily develop product as quickly as a commercial software company may do, but it may develop better product, more reliable product, more broadly supported product than a bankruptcy-prone, follow-the-profits company may produce. Only where there are monster companies propped by government patent monopolies do we see the prospect for the persistence of a commercial version of an invention.

This is because almost all universities and government labs are conducting early stage research — not developing products. For new drugs, medical devices and therapies, the development times and costs incurred by private industry are often significantly higher.

“Early stage” is itself vocabulary of the apex predator corporations. The research that’s conducted may be “early” (as in “we haven’t a clue what we should observe”) or “advanced” (we have spent years characterizing this phenomenon and now are refining our account). The research has nothing to do with “developing products.” People might use the results of such research–even patentable results–for all sorts of purposes that don’t involve making mass-market products or hoping to sell a speculative interest in a patent monopoly to even more gullible or less risk-averse speculators.

While it is no doubt true that private industry’s “development times and costs incurred” are “significantly higher” than the expenditures on research, it’s simply an inapt rhetorical comparison. If the private industry involved was not the portion of industry fixated on private monopolies, there might be competition to develop inventions more quickly, and there certainly would be cooperation the development of cumulative technology platforms for common commercial use as libraries, as standards–you know, how the digital computer and the internet were developed. The private industry that works with selective patenting, that cross-licenses to speed access and limit the expense of designing around useful technology–that private industry’s costs appear to be significantly lower. Why not have a national policy that encourages lower costs for development rather than handing patent monopolies to the significantly higher cost players in industry?

We now get the non-sequitor conclusion:

It is for these reasons that Bayh/Dole did not intend that government set prices on resulting products. Such action, while well-intended, would break the link that we have formed over the last 20 years that is one of the economic wonders of the world. This is not idle speculation.

This is nuts. Bayh-Dole is legislation. It has no intention but for what it states. It may be that Senators Bayh and Dole did not intend for the government to set prices. But Bayh-Dole sets a standard of “practical application” and that standard is defined to mean use of inventions so that the benefits of that use are available to the public “on reasonable terms.” Reasonable terms, in any reasonable interpretation, includes reasonable pricing. While Bayh and Dole no doubt are truthful in asserting that they did not intend for the government to set prices, this too is utterly unrelated to their Act. The original version of Bayh-Dole accounted for price not only with “reasonable terms” but also by limiting the duration of exclusive license agreements granted by nonprofits to large companies–that is, limiting the time that universities could hand patent monopolies to the big pharmaceutical companies. The exclusive term was five years from the date of commercial sale or eight years from the start of the license, whichever was sooner. That limitation would no doubt expedite development. But this limitation was removed in 1984 to clear the way for universities to deal in patent monopolies with big pharma without any particular protection for the public. In fact, given that Bayh-Dole anticipates that universities will earn royalties on their exclusive licenses, the universities become complicit in the monopoly pricing that results from their exclusive licenses. Why, they will benefit from the money, and that’s virtuous, so the monopoly pricing in turn is virtuous, because it benefits university research.

In 1989, NIH instituted a requirement that any product resulting from public-private Cooperative Research and Development Agreements be priced “reasonably.” However, President Clinton’s NIH director, Dr. Harold Varmus, eliminated the requirement in 1995 after a year-long NIH investigation found that it had “driven industry away from potentially beneficial scientific collaborations with (NIH) scientists without providing an offsetting benefit to the public.”

The CRADA program is a strange place to look for reasonable pricing. CRADAs are hardly the core of federal patent licensing and have next to nothing to do with federal research funding university scientists. CRADAs amount to “work for others” that make government facilities available for private company use. It was a strange requirement, and doesn’t address the primary concern with regard to reasonable pricing of prescription drugs. In the early 1960s, the pharmaceutical industry boycotted federally supported inventions in medicinal chemistry because they wanted monopoly rights, even to screen compounds for bioactivity. That the same companies might steer away also from an NIH CRADA program that asks for reasonable pricing is to be expected. The point, however, is not that it is a good thing for public money to subsidize companies pricing their products unreasonably, but that the government should stop providing welfare to such companies:

Since the elimination of this reasonable pricing requirement, this type of public-private research cooperation has rebounded.

Translation: federal welfare for companies charging monopoly prices has encouraged them to resume “public-private cooperation.” Cooperation with monopolists appears to be more important that other public benefits. Take a quick poll of those around you. How high on the list of public benefits available on reasonable terms from Bayh-Dole is such “cooperation” with monopoly-loving companies? Yeah, I thought so.

Thus, while Bayh and Dole no doubt are correct that their assertion is not “idle speculation,” again, the broader point is simply nuts. Bayh-Dole managed price through limits on exclusive licenses, but that was dropped, and on a requirement for reasonable terms for public access, but that requires government “march-in”–and those march-in procedures were designed not to operate, as Howard Bremer made clear in a public talk.

More importantly, Arno and Davis misunderstand the actual structure of the Bayh-Dole Act. Nowhere in the text of the law are there any references to “reasonable price” that should be dictated by the government. This omission was intentional, given that the primary purpose of the act was to entice the private sector to seek out public/private research collaboration rather than focusing on its own proprietary research.

It turns out that after Stanford v Roche, it is Senator Bayh who clearly did not understand the structure of the Bayh-Dole Act. This is no doubt deep irony, even sad. The “actual structure” of Bayh-Dole fails to make an express connection between “use the patent system to promote the utilization of inventions arising from federal research or development” and the definition of “practical application” with its requirement for “use” and “reasonable terms.” That omission was, too, no doubt intentional, to give the appearance that reasonable terms were required to maintain a patent monopoly, but that reasonable terms somehow did not include price, but only availability, when price would be, in any otherwise reasonable world, the primary consideration. While the government has no need to set prices, it surely has a role in the structure of Bayh-Dole to step in if prices are not reasonable. Perhaps the rejoinder would be that “reasonable” is ambiguous–and that, too, no doubt may have be intended. If so, then that’s more evidence of the horrid mess that Bayh-Dole is, and why drug development as we know it includes government-endorsed monopoly pricing of prescription drugs–prices that the government ends up paying through medicare, money that leaves the economy because we pay higher private insurance premiums, money that ends up with pharmaceutical companies and their investors. A great scheme if you can find some senators to run cover for you.

Additionally, the article’s authors mistakenly characterize the rights retained by the government under Bayh-Dole technology transfer agreements. The ability of the government to revoke a license granted under the act is not contingent on the pricing of a resulting product, nor is it tied to the profitability of a company that has commercialized a product that results in part from government-funded research. The law instructs the government to revoke such licenses only when the private industry collaborator has not successfully commercialized the invention as a product.

Sadly, this part is simply not true, other than Arno and Davis may have made some mistakes in their account. The account of Bayh-Dole given by Bayh and Dole is also inaccurate. Non-use is one of four conditions in 35 USC 203 under which the government may march-in and compel licensing. But non-use is tied to practical application, and the standard of practical application is met only when the benefit of the use of an invention is available to the public on reasonable terms. This is not the terms of the licensing agreements between universities and their monopoly-loving favorites, but rather the terms on which the favorites make the benefits available to the public. That means, among all else, reasonable price. But Bayh-Dole allows causes for government intervention other than for non-use–when an invention has not been made reasonably “available” for public health, or to comply with government regulations. Availability, too, may depend on price.

There is absolutely nothing in Bayh-Dole that instructs the government to revoke licenses. That’s the amazing thing. Bayh-Dole works by requiring federal agencies to use a standard patent rights clause in federal contracts. But Bayh-Dole has absolutely no mechanism that requires federal agencies to enforce that patent rights clause. In fact, at almost every substantive point, Bayh-Dole provides a means for federal agencies to ignore the provisions that would protect the public–they can take no action, can waive requirements, can decline to ask for reports. Federal agencies have no obligation to march-in. Bayh-Dole does not instruct them to march-in. It merely requires them to include march-in in their default contracting clause. Then they can ignore it–and they do. In the version of Bayh-Dole that Senator Schmidt introduced the year after Bayh-Dole, there was a provision for third parties to initiate march-in. University front groups lobbied to remove that provision and ultimately to defeat the bill entirely. Bayh-Dole is a law that requires contracts that don’t require enforcement. Nice cover. Crappy outcomes, except for pharmaceutical companies and patent brokers.

We feel strongly that Bayh-Dole or other technology transfer laws that have developed since 1980 should not be viewed as laws that are intended to control the profitability of the private sector, or that keep score on the government’s contribution to knowledge that is eventually commercialized.

Who can argue with an expression of strong feeling? According to Bayh and Dole, their law is about releasing any controls on the profitability of the pharmaceutical firms. That indeed appears to be the case. That’s what monopoly pricing can do, that’s the effect of passing publicly directed inventions as patent monopolies to select companies. That profitability, though, comes at the expense of the suffering public. That’s a shame.

It is therefore understandable that our good senators would appeal to us not to even bother to “keep score” on whether the government’s contribution matters at all. What do you make of that? It appears that Bayh-Dole is so successful that no one should examine the claim of its success. Not only should the government trade in monopolies at the public expense, but the public shouldn’t bother to ask whether that trade is worthwhile. Why not just give money to the pharmaceutical companies and their investors and save all the bureaucratic show of awarding research dollars to university scientists and then stripping them of their inventions through the now woefully ineffectual mechanism of “technology transfer.”

The laws that we passed are about encouraging a partnership that spurs advances to help the American people. We are proud to say it’s working. Indeed, the government of Japan has just adopted its own Bayh/Dole Act hoping to achieve similar results from its publicly funded R&D institutions.

True, that’s what the laws are about. It’s just that Bayh-Dole is a dismal failure at that. They can say they are proud about it–that’s true, too–but if they just said “it’s working” it would be false. It’s one of the clever tricks of rhetoric–“I want to say that…” + “anything” = technically true, regardless of the truth of “anything.” While Japan did pass its own Bayh-Dole Act, and no doubt hoped for “similar results,” it is not at all clear that Japan recognizes the disaster that Bayh-Dole has been, nor even that they have done more than institutionalize research inventions behind the same patent paywalls that 99% of university inventions are now trapped behind.

Let’s start by repealing the illusions of Bayh-Dole, the faux Bayh-Dole, the brain-dead stupid Bayh-Dole, the Bayh-Dole that serves as a pipeline of private monopolies on publicly funded inventions for some few hundred of pharmaceutical inventions at the expense of over a hundred thousand research inventions tied up in institutional bureaucracies. Then let’s enforce Bayh-Dole according to the rule of law rather than the hand-me-down wisdom of folks that had a hand in drafting the law but have proven they don’t know the law or if they do they aren’t willing to be even close to accurate or honest. Enforcing Bayh-Dole includes implementing all its requirements. Enforcing Bayh-Dole means requiring federal agencies to audit and enforce all substantive terms of the patent rights clause. Enforcing Bayh-Dole means marching-in on nearly all subject inventions now held by universities, because nearly all of them fail the standard of practical application. Enforcing Bayh-Dole, too, means requiring the government to use, not waste, its non-exclusive license to practice and have practiced–make, use, and sell, and have made, have used, and have sold–all subject inventions.

Once we have repealed the fake Bayh-Dole and enforced the real Bayh-Dole, then we can settle in to protecting the public–make use reports public, at least use reports from nonprofits. Restore limits on exclusive licenses to big companies. Clarify the (f)(2) written agreement requirement to make express that nonprofits have no right to any invention as a condition of federal funding. Then we would make some progress in developing an approach to invention management that served the public and not just the profit-taking ability of a handful of pharmaceutical firms.

This entry was posted in Bayh-Dole. Bookmark the permalink.