The Volokh Conspiracy
Mostly law professors | Sometimes contrarian | Often libertarian | Always independent
Patents During a Pandemic
Should patent law function differently during a pandemic? Let's place aside for now the question of whether research efforts during the pandemic should be entitled to new patents, and focus instead on the argument that previously existing patents should be unenforceable or weaker during a pandemic to improve society's ability to mobilize against the virus.
Sendhil Mullainathan and Richard Thaler make this argument, citing a case in which producers of 3-D-printed valves needed to replace ventilator parts received a threat of a lawsuit. Meanwhile, an attempt by Labrador Diagnostics LLC to enforce two patents in its portfolio against a maker of COVID-19 testing kits received widespread criticism, and the patent holder ultimately backed down. Separately, Israel has approved importation of a generic alternative to an HIV drug thought possibly helpful against COVID-19, even though the drug remains under patent in Israel until 2024.
The question with greater stakes is what happens if a patented treatment proves to be effective against COVID-19. The maker of remdesivir received orphan drug status for the drug (and thus exclusivity and other benefits) on the ground that it met the statutory definition of a rare disease, based on the number of people currently suffering from COVID-19 in the United States. After criticism, Gilead declined the "rare disease" status. Remdesivir remains under patent, and Gilead will presumably defend its patent rights more aggressively. But will Gilead be allowed to do so? One might think that if the treatment proves successful, Gilead would be among the most valuable companies in the world. Surely, one might think, it would be able to earn a trillion dollars, a fraction of what is at stake in dollars and lives in the United States alone. But Gilead's market capitalization has risen only around $10 billion since the beginning of December. Perhaps this reflects skepticism about the drug's prospects, but not that much skepticism. It surely also reflects skepticism that Gilead will be able to price it aggressively. Analysts seem to anticipate a price target of only around $260, a bargain if the drug is really a life-saver.
The argument for allowing robust patent protection is mostly the same as the argument for patent protection generally: Patents encourage innovation. If the rule is that patents are unenforceable when they are most needed, then there will be reduced incentives to create inventions that may be especially needed in some future emergency. When an inventor is considering developing a ventilator part, a diagnostic test, or a treatment, we would like the inventor to consider the full social benefits of the invention, not just the social benefits to be accrued in ordinary times.
Here are some counterarguments, in my view none sufficiently powerful:
Unforeseeability. If the demand for an invention is entirely unforeseeable, then inventors will not have taken that demand into account in deciding whether to invent in the first place. Thus, patent revenues are a windfall, and society would be better off not enforcing the patent rights. A pandemic, one might argue, is unforeseeable. But we know that pandemics are not unforeseeable. Even though remdesivir failed against ebola, Gilead knew that it might work against some later coronavirus, and that might well have factored into its development decisions. Certainly as to drugs for medical treatments, this argument seems especially weak. What about patents on surgical masks or other medical tools? Sure, it's doubtful that the patentees explicitly took into account the possibility of a pandemic, but a rule that limits patents to explicitly foreseen uses might well lead patentees to expect less revenue. In patent law, inventions are patentable even if they are discovered serendipitously, in part because inventors expect that R&D will sometimes lead to serendipitous discoveries. The same argument applies as to serendipitous uses. Inventors may not know exactly what their inventions will be used for, but they recognize that sometimes they might receive revenue for uses of their invention that they did not precisely anticipate, and the possibility of such uses increases incentives to invent.
Speculative patents. Sometimes patents cover inventions that will not be useful until some later time, when some complementary technology arrives. John Duffy and I argued that such "speculative patents" should be patentable only if they meaningfully accelerate the arrival of technologies. For example, even if I develop a nonobvious idea for using a time-traveling Delorean in some clever way, I shouldn't be able to obtain a patent in the hope that a flux capacitor is invented, unless I meaningfully accelerate either the invention of time travel or at least my application of time travel. At least, the invention should not be counted as nonobvious on the ground that time travel seems fanciful, if my invention is simply free-riding on the future invention of time travel. Returning to pandemics, suppose that a drug company could quickly and cheaply invent many antiviral drugs that might be useful against future pandemics, depending on what proteins might turn out to be important in the pandemic. If the drug would be nonobvious after the pandemic comes (because any person having ordinary skill in the art could create it), then it should be nonobvious if the drug is developed earlier. Indeed, one might argue, contrary to existing law, that nonobviousness should be measured as of the time when an invention is first useful on the assumption that the invention hadn't been made, rather than as of the time of invention. But there is little reason to think that remdesivir fits this argument.
Bad patents. Some of the criticism of Labrador Diagnostics's aborted enforcement effort was based on the ground that the company is a "patent troll" and that its patents were derived from work with now-discredited Theranos. But nonpracticing entities at least sometimes engage in useful R&D that the patent system should reward, and even a company that perpetuated frauds may have legitimate intellectual property. If we should be tougher on nonpracticing entities or if we should have more exacting requirements of utility, that should be true whether there is a pandemic or not. It is doubtful that the category of "inventions for pandemics" is so much more susceptible to bad patents that it makes sense to have a special rule making such patents unenforceable.
Urgency. In ordinary times, one might have leisurely discussions about licensing patents. But we might not want manufacturers to spend, say, a month or two clearing all intellectual property issues before starting manufacturing, because a pandemic demands immediate action. This seems to me a reasonable argument as to whether an injunction should be granted against production, and indeed, under the fourth factor of the eBay v. MercExchange test, one can make a strong argument that the "public interest" argues against both temporary and permanent injunctions. But that does not explain why the patentee should be unable to obtain a damages remedy. The argument might be that people will be unwilling to produce valuable goods when they don't know what their liability eventually will be, but of course they do so all the time. Moreover, if small-scale producers are worried about liability, they could insist on contracts that indemnify them for intellectual property violations. If inventions are genuinely life-saving, purchasers like hospitals should be willing to take on such risk (and, if not, we may have a problem with the incentives of hospitals rather than a problem with IP).
Maximizing production. We will, of course, have more production if IP rights are suppressed than we otherwise would. And, as I have emphasized previously, generating more production of life-saving medical products may be the single most important policy goal right now (and unfortunately is largely unaddressed in the $2 trillion spending package). But we would also might have better emergency healthcare during this pandemic if states were allowed to commandeer without compensation hotels for housing healthcare professionals and other facilities for building hospitals. Yet we generally assume that even in times of need, healthcare providers should pay for what they consume. That's true even for goods with low marginal cost; electricity may be mostly fixed costs, but no one says that hospitals should pay less for electricity during times of emergency. If the justification for patents is sound, healthcare providers should be required to pay for the intellectual property they use.
Redistribution. What likely troubles people most about the possibility of strong patent protection for remdesivir or other COVID-19 treatment is that only the relatively rich may be able to afford it. Even though we accept that the wealthy have more purchasing power than the power, we may resist disparities when it comes to purchases essential to life itself. Such arguments, however, apply to all critical healthcare. If these arguments are sound, then the government should pay for health care that it believes is essential, at least for the poor, maybe even for the poor of other countries. Perhaps the government should take patents with eminent domain, paying to the companies what they otherwise would have received from private exploitation. But requiring redistribution from inventors to the poor displaces the government's responsibility and reduces incentives to invent.
Time inconsistency. We often assume that the correct policy, in patent law and elsewhere, is the one that would be chosen ex ante. A policy is time inconsistent where the best policy ex ante will differ from the best policy ex post. A common suggestion for monetary policy is that the government should tie its policymakers' hands ex ante to prevent the policymakers from engineering inflation ex post. Similarly, in patent law, if all of the above is correct, we should create rules that make clear that patents are fully enforceable in future pandemics. At the same time, though, the time inconsistency argument implies that when we have failed to tie policymakers' hands ex ante, and a different policy is preferable ex post, we should adopt the policy that maximizes social welfare ex post. Thus, one might argue, if the stakes of a pandemic are sufficiently high and patent law is vague as to whether patents are enforceable in pandemics, we should do what is best ex post (suppress patent rights). But that argument has two problems. First, this violates a norm in patent culture that we should conceive of the system from the ex ante perspective. This violation will lead to expectations of further cheating off the ex ante optimum in the future. Second, there is no vagueness in patent law here. There is only outrage, misplaced at least as to valid patents.
