Nonexcludable Innovations and COVID-19
Some of the most widely discussed COVID-19 interventions include vaccines, drugs, and medical devices—typical interventions for many diseases, whether the cause of a pandemic or not. These interventions share a further similarity—they’re all generally excludable. That is, the owner of a particular invention can generally exclude others from practicing it without permission. In a classic account of IP, it is this excludability that encourages their development in the first instance. But as Professors Amy Kapczynski and Talha Syed have explained, some knowledge goods are more excludable than others. In this post, we describe why many successful COVID-19 interventions—washing hands, wearing face masks, even the proning of patients in a hospital setting—are generally nonexcludable and thus likely to be underincentivized by IP-based market rewards. Policymakers tasked with encouraging COVID-19 innovation should attempt to correct for this asymmetry in excludability.
Intellectual property law relies on a relatively simple structure: give people incentives for producing knowledge goods by giving them the right to exclude others from making or using those goods for a period of time. This allows creators to charge higher prices than they could in a fully competitive marketplace. For some knowledge goods, however, excluding others is difficult or impossible even under expansive IP laws; they are “nonexcludable.” If innovators invest in knowledge-good production only up to the point that their own marginal benefit exceeds the marginal cost—ignoring the benefits to others who can’t be excluded—then they will invest less in nonexcludable goods than is socially optimal.
Some knowledge goods are nonexcludable (or only partially excludable) because they simply aren’t the subject of IP, i.e., they’re not protectable subject matter. Laws of nature, natural phenomena, abstract ideas, and facts about the world, for example, cannot be patented or copyrighted or protected through regulatory exclusivity (though they can often be kept secret for a time). But as Kapczynski and Syed have explained, even where IP is available, it may be difficult to enforce. Monitoring use of the knowledge good may be technologically infeasible, normatively unacceptable, or otherwise unviable with existing institutions. Nonexcludability thus systematically distorts the incentives for innovators to create new knowledge goods.
What kinds of nonexcludable interventions can help address the pandemic?
A number of interventions that have emerged in the context of COVID-19 have limited excludability. Consider mask wearing. At present, the best evidence suggests that wearing masks in public helps reduce the spread of COVID-19—whether it is an exhaustive review of the existing literature, empirical studies done in the context of seasonal coronaviruses, mathematical models of COVID-19, or other scientific articles. After initially recommending against the use of masks in public by most Americans, the CDC has now recommended the use of cloth masks “in public settings where other social distancing measures are difficult to maintain,” such as grocery stores.
But this knowledge about the benefits of wearing of masks in public—particularly DIY cloth masks—is highly nonexcludable. The holder of a patent on a method of wearing masks in public to prevent the spread of COVID-19 (should it even issue) would not reasonably be able to enforce that patent. Systematically enforcing patent violations would not be practical, given the vast numbers of Americans wearing masks in public (73%, in a recent AP poll). It might be easier to sue entities (like grocery stores) who require shoppers to wear masks, on grounds of inducement of infringement, but this is difficult to imagine, given the social pressure on such a patentholder not to enforce their rights. And the same argument applies to knowledge about the benefits of other public health measures—it is hard to imagine enforcement of a patent related to hand washing or social distancing.
As another example, consider proning, in which COVID-19 patients are simply placed on their stomachs (in the prone position) rather than on their backs. According to a series of studies (including multiple on COVID-19 specifically—see here, here, and here for a selection, but also here in the pre-COVID-19 context), proning may increase patients’ oxygen saturation and may help patients avoid being placed on a ventilator. It may even lower their risk of death. l
Proning is somewhat more excludable than mask-wearing because it occurs primarily within the hospital setting. As a result, a patentholder on a method of using proning to improve oxygen saturation and outcomes in the COVID-19 context would be able to identify institutions within which these activities are occurring, even if that number would be vast. Many medical centers or specialty societies have proning protocols, which would permit them to be identified as well. But a patentholder would still need to identify specific cases of infringement within these contexts (made somewhat more difficult by health information privacy laws), and the optics of suing hospital systems for providing the best available care during a pandemic are not helpful.
As a final example, consider the development of evidence regarding the use of already-approved drugs like hydroxychloroquine (approved for use in conditions including malaria, lupus, and rheumatoid arthritis) for COVID-19. Famously referred to by Professor Rebecca Eisenberg as “The Problem of New Uses,” pharmaceutical companies themselves have little incentive to study their existing drugs for new uses. Patents on the new method of use can be obtained but are hard to enforce. Once a drug becomes subject to generic competition (as has been true of hydroxychloroquine), prescriptions of the old drug for its new indication can be filled with a generic, and trying to enforce patents against doctors or patients using the drug for the new use would be a practical and reputational nightmare. The incentives to develop negative information about existing drugs—such as the harmful side effects of hydroxychloroquine in COVID-19 patients—are even smaller.
To be clear, the limited excludability of knowledge about mask wearing, proning, or new uses of old drugs does not mean that no one will invest in these innovations. After all, as detailed above, preliminary evidence has been acquired about the effect of all of these interventions for COVID-19, and additional studies are ongoing. Rather, the concern is that firms relying on market incentives will invest less in these kinds of innovations than is socially optimal. As Kapczynski and Syed explain, there is no reason to believe that excludability is correlated with social value, which means that IP-based market rewards will underincentivize many valuable innovations.
Should policymakers make nonexcludable knowledge goods more excludable?
One putative solution to these misaligned incentives is to make more COVID-19 interventions excludable. This would, under a classic incentive theory, increase investment in developing such interventions—even basic public health interventions—by essentially offering a bounty that doesn’t currently exist. For example, the problem of insufficient incentives for researching new uses for old drugs is at least partially addressed by allowing developers to both claim patents on the new use and garner regulatory exclusivities for the new use from FDA. Both types of incentives are grounded in creating excludabilities that did not previously exist.
As successful—or not—as such policies have been, this kind of IP-based incentive in the COVID-19 context should be matched with a non-IP allocation policy (e.g., a healthcare system with universal low-cost coverage) to promote widespread adoption and use, without respect to price. As noted above, excludability is meant to encourage innovation by promising supracompetitive profits. But for obvious reasons, the policy aim of developing COVID-19-related interventions is not necessarily to make them maximally profitable but to ensure they reach the greatest number of patients.
In yet other cases, increasing the excludability of a given intervention may do little work because, as discussed earlier, enforcing these provisions may be impracticable due to the institutional siting or nature of the intervention. Barring the use of a medical device in a hospital setting may be simple enough; attempting to lay an exclusive claim to face masks in a world replete with old T-shirts, sewing machines, and people with a lot of time on their hands, is a fools’ errand. The most valuable COVID-19 interventions—ones that broadly traffic on the public health outside the medical care delivery setting—are those least likely to be affected by increasing their excludability.
What non-IP tools can policymakers use to increase incentives for nonexcludable innovations?
Increasing excludability through IP is fortunately only one option in the innovation policy toolkit. Two weeks ago we detailed the benefits of different non-IP incentives in the pharmaceutical context; these alternative innovation policies play an even more important role for inventions that cannot be effectively protected through IP. Most significantly, direct government spending—whether through R&D grants, public-private partnerships, procurement from private firms, or other institutional arrangements—can help correct for failures in market-set incentives. As Kapczynski and Syed note, their argument strengthens the case for public R&D funding and makes clear that the “scope of public funding should not be restricted to basic research, but rather should extend into other domains that also involve valuable but highly nonexcludable information goods.”
In bills enacted on March 4, March 27, and April 24, Congress allocated nearly $9 billion for COVID-19-related R&D, and this public funding goes well beyond basic research. But even though the gap in IP rewards for nonexcludable innovations is now well established in the innovation policy literature, we are unaware of any systematic effort to use public funding to balance this asymmetry. Kapczynski and Syed offer specific policy suggestions that are as applicable to addressing COVID-19 as to other areas of R&D, including dedicated funding for nonexcludable research, incorporating nonexcludability assessments into the NIH peer review process, increasing public health funding, and using FDA regulation to require production of nonexcludable information about drugs and medical devices.
To be sure, government-set rewards are not a panacea—policymakers have failed to correct for IP’s distortions beyond the asymmetry in excludability, and these failures of political markets can be as harmful as market failures. But this crisis has sparked an unprecedented interest in biomedical innovation policy from the public, the media, and policymakers throughout federal, state, and local government. Innovation policy scholars should use this opportunity to emphasize a simple but important point: public R&D funding should focus on socially valuable innovations that receive little incentive from IP.
Labels: COVID19
