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.

NEXT: Covid-19, the Evil Genius, and How to Think about Societal Risks

Editor's Note: We invite comments and request that they be civil and on-topic. We do not moderate or assume any responsibility for comments, which are owned by the readers who post them. Comments do not represent the views of or Reason Foundation. We reserve the right to delete any comment for any reason at any time. Report abuses.

  1. I think patents should be an extent.

    Let’s say, for example, China developed and patented a medicine which cured COVID-19. They then refused to sell the medicine to the US patient base…at any price.

    The US government would and should be able to manufacture the medicine, despite the patent.

  2. “But will Gilead be allowed to do so?”

    No. Either the government will seize the patent and litigate it later or public opinion would force them to back off.

    Its a panic, no law survives a big enough mob.

  3. A modest proposal: Patent the virus, and also recognize it as a legal person and give it rights of publicity and depiction under California law. Then combine the residuals for each use of that annoying microscope image, false image claims against the worst of the apocalyptic press, and claims for infringement against any deep pockets that facilitated the reproduction in different hosts, and then use that fund to pay the patent price of the lifesaving medication that keeps many thousands of people from dying for lack of air.

    (Best to get the madcap satire in before the surge hits.)

    Mr. D.

  4. I believe a similar problem happened during WW I, 100 years ago, over the Wright Bros patent contest with Curtis, and it was solved by the government buying the patent.

    Why not just do that? Buy the damned things! If need be, by eminent domain. Isn’t that a good statist and constitutionalist solution?

  5. Were I the judge in this controversy, I would use yet another analogy (h/t David Friedman): Negotiating with the company over the price of its drug now, when large numbers of lives may depend on getting it quickly, is like the owner of a ship sinking in a storm negotiating on the spot with a tugboat owner who can rescue his ship, but wants to profit opportunistically by doing so.

    The sensible solution seems to me the same as in that scenario: Gilead should be forced to provide the drug to all affected by the emergency, but then the question of how much the government should pay them should go before a court, which will decide on a price that is fair to both sides.

  6. The speculative patent problem is related to one of the problems with software patents. There are lots of software patents that are obvious once a certain level of computer technology is invented. But since there’s a certain time at which the technology is invented, there’s going to be a first person to file the patent too.

  7. Sometimes patents cover inventions that will not be useful until some later time, when some complementary technology arrives.

    An invention that isn’t useful isn’t patentable.

  8. Patents are ultimately sometimes necessary evils. Restrictions contrary to the natural order of the free exchange and use of information. They are not good things in and of themselves but exist to benefit society and enterprise. They should be set aside the second they start to harm instead.

  9. The argument here is poorly made and it is based on an extreme belief if the rationality of markets. Although many objections are correctly identified, the rebuttal is consistently superficial.

    (1) Unforeseeability:

    The argument here is just the efficient markets hypothesis on steroids. Although applications of a patent based on future events (like a pandemic) are unforeseeable, somehow, as if through magic, the invisible hand takes this into account. This is foreseeable unforeseeability.

    Has the professor done an EMPIRICAL inquiry into how entities actually make investment decisions? Looked at the actual NPV calculations that pharmaceutical companies and other patent holders ACTUALLY use in making investment decisions? And been skeptical of those calculations (since pharmaceutical companies and other patent holders have a reason to engage in a certain amount of hype to attract investment, like anyone else). No?

    Instead, the so-called professor proceeds upon faith. Somehow, in some way, the current pandemic was “priced in” to investment decisions in a plausible manner, writing, without evidence that “the possibility of such uses increases incentives to invent.”

    BS. The burden is on YOU to prove that the unknown unknowns are properly taken into account. And the burden is on you to show that these “incentives to invest” (In what? Inventing, or gathering patents… these are separate but sometimes related activities) are more optimal. What about “over-incentivizing” such behavior? Surely, it is not optimal to allocate ALL of GDP to invention. So, presumably it is possible to “over-incentivize.” And, presumably, a bubble is possible in this market like any other.

    The professor deals with none of these OBVIOUS issues. Which makes him particularly unsuited to opine about patents, which are themselves supposed to be non-obvious. If this professor cannot understand obvious objections, how is he even qualified to opine on intellectual property law at all??? His analysis is entirely superficial.

    (2) Speculative patents

    The section on speculative patents is just the same issue as unforeseeability. What meaningful distinction has the professor made between “unforeseeable” rewards to patents and “speculative” patents. Bizarrely, after arguing in the previous section that “unforeseeable” rewards of patents ought to be honored, the professor contradicts himself saying that “speculative patents” (where the speculation is based on the unforeseeable development of a complementary technology, rather than, say an unforeseeable pandemic.) But, if the professor were CONSISTENT, there is no difference between the unforeseeable development of a complementary technology and the unforeseeable/speculative development of a pandemic. Both involves wild guesses about the future. If wild guesses about a pandemic should be rewarded, why not wild guesses about a complementary technology? Does not the “the possibility of such uses increases incentives to invent” argument apply here if it applies in the unforeseeable case?

    Of course it does! That our so-called professor does not recognize the contradiction in his own position and attempt to reconcile it is a basic intellectual failure.

    (3) Bad patents.

    Our so-called professor writes: “But non-practicing entities at least sometimes engage in useful R&D that the patent system should reward”

    Don’t tell me. Prove it! Give examples. Do something. But don’t just act like you can just allow this part of your argument to be assumed to be true. Trolls (which is the proper name for so-called non-practicing entities) can INCREASE THE COST of innovation by diverting engineering resources to a courtroom and to lawyers and expert witnesses rather to the development of technology. Basically, the ASSUMPTION here is that patent trolls are, on the whole, good, such that their marginal “contributions to innovation” ought to be rewarded and thus help sustain them.

    Also, regarding Thanos, isn’t this company a complete counterexample to the professors earlier “faith” in markets to act efficiently, implicit in his argument for (1) ? After all, investors diverted $700 million resources from a productive uses to Elizabeth Holmes based on nothing more than hype, when they theoretically had every incentive to be more intelligent and invest those resources in patents that might be useful in a pandemic. That the professor fails to recognize how the failure of the market to evaluate the Theranos business model properly (which was based on nothing more than hype) calls into question the claim that they should be rewarded for that which is “unforeseeable.” It turns out that these market actors aren’t actually good at predicting the future. Thus, any arguments that they ought to be rewarded for the “foreseeably unforeseeable” ought to be discounted as well.

    Of course, I don’t expect our professor to necessarily agree with me. But the complete failure to defend assumptions about the usefulness of incentivizing patent trolls or relate Theranos to the point about unforeseeability and investor investment behavior is just so bad as to be intellectually indefensible.

    (4) Urgency.

    The professor addressed this in the most ridiculous way possible, focusing on licensing. But what is OBVIOUSLY just as significant as licensing is being “well-positioned” to take advantage of a license, if it were granted. Patents create barriers not merely in using the licensed technology, but in developing the capacity to use licensed technology in the first place. This is CLEARLY an extremely significant issue when it comes to evaluating the utility of incentivizing patents on speculative technologies that would otherwise not be patented at all. After all, a company that is not well-positioned to take advantage of a license will not be able to move with the urgency we would like in a pandemic. How could our so-called professor miss this elementary issue???

    Maximizing production:

    The professors discussion of healthcare providers paying for hotel rooms is entirely UNRELATED to maximizing production. No one has EVER argued that healthcare providers would produce more if only they didn’t have to pay for their hotel rooms. In fact, this is simply has no significant EMPIRICAL impact on the marginal provision of healthcare. (The real constraint is getting more doctors and nurses trained, not housing the ones we have.) In OBVIOUS contrast, we can see much more directly how having to reward patent holders leads to an EMPIRICAL observable decrease in the quantity and quality of medical care provided. That our so-called professor came up with an example that is completely irrelevant to the real world is just sad.


    Ok, the label here is just inexcusable. Patent law itself redistributes (from society to patent holders). So, to say that not enforcing patent law is redistribution is a category error. It is, at best, re-redistribution, not redistribution. That our so-called professor fails to deal with the fact that the use of ideas is naturally non-rivalrous and that law is coercively being used to create a scarcity that would not otherwise exist in order to reward patent holders is inexcusable.

    There is a basic economic term for this scarcity, and that is deadweight cost. Poor and middle class people value patented goods at more than the marginal cost of production, but are deprived access to that good anyway. This deprivation benefits NO ONE, including the patent holder but is a consequence of the patent holders inability to engage in perfect price discrimination.

    Characterizing deadweight cost as “mere redistribution” implies that the patent holders rights are somehow “natural” and that when goods are provided to people who value that good at more than the marginal cost of production that this is “unnatural.” If we feel bad for the poor and the middle class who are thus unable to access the healthcare they need, the government should “subsidize” their access. What this FAILS to recognize is that the patent right itself is the thing that is “unnatural” and the patent right itself is actually a subsidy. One speculatively designed to increase innovation (with much waste and in a ham-handed manner).

    The only redeeming quality of this section is the mention of using eminent domain to extinguish patent rights, since this is a valid partial solution to the deadweight loss problem (which goes unnamed). But on the unredeeming side, the professor fails to mention the possibility of using a reward system instead of the patent system. Also, if eminent domain is used, what price should be paid? The price during the pandemic or the price pre-pandemic? Based on the professors ex ante argument, the price that should be paid is the pre-pandemic price, since we should not compensate investors for their ex-post expectations, but only their rational ex ante expectations. The failure to address the price to be paid is a major flaw in this blog post, because based on the reasoning in section (1) the better price would be based on ex-post expectations (because there is an assumed speculative ex-ante reliance) but based on (7) the principle should be ex-ante expectations. A major flaw (and there are too many to count) in this analysis is thinking that investors should “have it both ways” and be rewarded based on ex post expectations, but only be limited by ex ante expectations/knowledge).

    (7) Time inconsistency.

    This is the only somewhat useful paragraph in this otherwise largely useless blog post. But as I explained above, the idea that investors have a valid ex ante expectation of a possible pandemic is implausible. At the end of the day, our so-called professor is REALLY arguing that inventors should be rewarded for their newly developed ex post expectations (because, implausibly – under a false efficient markets hypothesis type assumption where clear market failures like Theranos should not exist) they somehow anticipated and gambled on that ex post expectation. So, they should get their cake, and eat it too.

    On the contrary, people are individually RIGHT to be outraged by patent holder price gouging in a time of national emergency. (After all, if an individual themselves will endure an enhanced risk of death or serious injury now due to lack of production problems caused by patents, whether such outrage is somehow collectively socially non-optimal, it is still an individually optimal and thus individually rational behavior to be outraged by the threat in front of them. Also, it is perfectly predictable that people would behave in this manner.)

    Meanwhile, our “all-knowing” hypothetical investors who anticipated a pandemic, as rational beings, should ALSO have anticipated such individually rational (and entirely predictable ex ante) outrage would limit their returns in that case and curtailed their investments accordingly. So, if investors are as smart as the professor implicitly theorizes, then not allowing them to gouge now is what is consistent with their rationally formed ex ante expectations and allowing them to gouge would only provide them with a windfall.

    If we wanted to maximize future investments in innovation, providing such a windfall is not an efficient. Instead, those resources should be used to reward FUTURE innovation exclusively. Rational individual outrage at price gouging in times of pandemics, after all, isn’t something that a rational investor would fail to recognize and account for.

    Overall, this blog post is nothing more than apology for patent holder price gouging in a time of national emergency. Maybe that explains its bad quality.

    1. Thank you for your comments. Tone aside, you make some good points. I agree that there are difficult empirical questions about the degree to which people will invest more in inventions that will be particularly useful in emergencies that may occur with low probability. Probably not much if they expect not to be able to benefit in the event of an emergency, but perhaps they would if they anticipated being able to earn money during the emergency. An analogous question is why businesses didn’t stockpile goods like masks before the emergency in the hope that they might be able to sell them at high prices in the event of an emergency; I think that the answer is that stockpiling will not occur because stockpilers anticipate being unable to charge market price. And so we’re left with our inadequate national stockpile.
      Regarding eminent domain and patent rewards, I agree completely (and have written extensively on the topic).

      1. I think that the answer is that stockpiling will not occur because stockpilers anticipate being unable to charge market price.

        That is an answer, but a more conventional one might be that an investor looking at a highly unpredictable duration of capital tied up in inventory and ongoing warehousing fees would opt for a less speculative place to park their money.
        An example (though featuring a public rather than private investment) is the Canadian province of Ontario, which was a hotspot for SARS and in its aftermath stockpiled 55,000,000 N95 masks ‘for the next one’. In 2017 after paying $3,000,000/year storage costs an audit found that most of the equipment had passed its expiration date and was no longer usable. This is a situation that a profit-driven business would choose to avoid.

  10. One option for an emergency is an implied licensing procedure. Whoever wants to can make it, but the patent holder gets royalties. Perhaps there could be a law specifying there procedure and royalties in advance. In the absence of such a rule, the courts could determine a fair amount after the fact, with no punitive or other “punishment” type damages. (That is, in an emergency other parties have a right to make it, with only the amount royalties at issue.)

  11. I also agree with David Walker that markets are notoriously inefficient and unreliable at pricing the risk of rare future events. As I’ve said many times on this forum, our ability to predict the future is much less reliable, and we can have much less confidence in our predictions, than our subjective beliefs (which often make us feel very confident) would tell us.

    I think this fact has a number of implications. One of them is a general conservativism. Laws, especially ones that have been part of a society for a long time, often involve predictions of future consequences and prophylactics for circumstances a particular generation doesn’t currently see. Because of this, judges should be less quick to claim that they are irrational because it disagrees with their own perceptions and their own predictions of the future.

    A second is recognizing that markets are terribly inefficient at preparing for rare events. Left to the markets, the firm that sets aside extra capacity for the future – to innovate as well as to prepare for disaster – will be beaten into oblivion by the firm that lives and is maximally efficient only for today, and will go under long before the disaster ever occurs Or the new thing is ever perfected. Indeed, it is only because markets aren’t efficient in a classical sense that disaster preparedness, or innovation, ever happens.

  12. If speculative patents had been enforceable, Edison and Swann would have run up against prior patents for the light bulb. There were experiments with lighting using an electrically heated filament in a vacuum at least a decade before these inventors worked on it. They failed, perhaps because the vacuum was not good enough or the earlier researchers did not persist to find an adequate material for the filament. Speculative patents would have ensured no one but the first researcher could have profitably worked on improving the technology for 17 or 20 years.

Please to post comments