Pandemic Ethics and Status Quo Risk

Abstract

Conservative assumptions in medical ethics risk immense harms during a pandemic. Public health institutions and public discourse alike have repeatedly privileged inaction over aggressive medical interventions to address the pandemic, perversely increasing population-wide risks while claiming to be guided by ‘caution’. This puzzling disconnect between rhetoric and reality is suggestive of an underlying philosophical confusion. In this paper, I argue that we have been misled by status quo bias—exaggerating the moral significance of the risks inherent in medical interventions, while systematically neglecting the (objectively greater) risks inherent in the status quo prospect of an out-of-control pandemic. By coming to appreciate the possibility and significance of status quo risk, we will be better prepared to respond appropriately when the next pandemic strikes.

Introduction

Conservative assumptions in medical ethics risk immense harms during a pandemic, even if they serve us well enough in ordinary times. Public health institutions and public discourse alike have repeatedly privileged inaction over aggressive medical interventions to address the pandemic, perversely increasing population-wide risks while claiming to be guided by ‘caution’. This puzzling disconnect between rhetoric and reality is suggestive of an underlying philosophical confusion. In this paper, I argue that we have been misled by a kind of status quo bias¹—overweighting the significance of the risks inherent in medical interventions, while systematically neglecting the (objectively greater) risks inherent in the status quo prospect of an out-of-control pandemic.²

The first section of this paper, on ‘The Asymmetry of Risk’, sets out a possible justification for conservative medical regulations in ordinary contexts, and explains why it reverses in a pandemic. The next section, ‘Caution and Recklessness’, applies this lesson to explain why various policy responses to the pandemic that were presented as being undertaken ‘from an abundance of caution’ were in fact objectively reckless with people’s lives—and, conversely, why alternative responses that intuitively strike many as seeming ‘reckless’ are better understood as objectively prudent. Expanding upon this theme, the sections on ‘Human Challenge Trials’ and ‘Early Targeted Immunity’ defend controversial methods for speeding immunity research and immunity acquisition (where most needed) early in a pandemic. The final sections on ‘Vaccine Hesitancy’ and ‘Pandemic Policy Expertise’ address objections to the overall line of argument pursued in this paper.

The Asymmetry of Risk

Autonomy-restricting medical regulations (that bar access to untested or experimental treatments and candidate vaccines) are sometimes justified on grounds of harm-minimization. This justification requires that the regulations protect the public from risks of harmful side-effects—such as those that emerged in the Thalidomide scandal—sufficient to outweigh the potential benefits of the barred medical interventions. I take no stand here on whether this attempted justification is in fact typically successful³; my claim is just that if autonomy-restricting medical regulations are to be ordinarily justified, this would seem the best available explanation. Regulators may argue that, in ordinary circumstances, the status quo is relatively safe and so untested medical innovations are riskier than the status quo. Until they are proven safe and effective, it may be reasonable to assume that the potential risks of untested products outweigh their potential benefits, and so block public access to such products until they pass stringent testing requirements.

Even this argument does not support privileging the status quo as a matter of principle. Rather, the suggestion is just that if we are already in a tolerably good state to begin with, highly uncertain gambles may be disadvantageous in expectation, with the potential harms outweighing the potential for reducing status quo risk. Of course, that antecedent assumption—that there is little status quo risk to alleviate—may be questionable even in ordinary circumstances.⁴ But the crucial point for our purposes is just that the argument is restricted in scope to specific contexts in which the status quo is relatively safe. In different contingent circumstances, such as those of an out-of-control pandemic, any sensible underlying normative principle will recommend very different practical responses.⁵

A pandemic reverses the usual asymmetry of risk. Now it is the status quo that is immensely dangerous, and a typical sort of medical intervention (such as an experimental drug or vaccine) is comparatively less so. As a result, we should expect to find many cases in which the potential benefits of innovation outweigh the potential risks. Doing nothing new, and allowing the pandemic to continue unabated, should be recognized as a far riskier prospect—for many individuals, and especially for society at large—than trying experimental or otherwise uncertain solutions for which the risks are orders of magnitude lower than the risk otherwise posed by the virus.

For example, scientists may know enough about the general properties of an experimental vaccine to expect that it would have a low chance of serious side-effects, while judging that it offers a decent chance of protection against the rampaging virus. People, especially those at high risk of exposure and/or high risk of serious illness from the virus, could then reasonably judge that trying the experimental vaccine offers them better health prospects than waiting and remaining vulnerable to the virus with certainty. In such a situation, regulatory obstacles unjustifiably violate both individual autonomy and harm-minimization.

The case for medical liberalization in pandemic circumstances is greatly strengthened when we take into account the societal value of information. By tracking the outcomes for those who try experimental vaccines, for example, we could sooner confirm their safety and efficacy, thereby encouraging broader adoption and slowing the pandemic’s spread much sooner than has actually been the case.⁶ The total societal cost of an ongoing pandemic is so immense that we will typically have overwhelmingly strong reasons to support any policy that could feasibly bring it to an end sooner (so long as this does not entail a symmetrical risk of extending the pandemic or causing comparably significant harms). So it bears emphasizing that whenever the risk to individuals of pharmacological side-effects is lower than the status quo risk they are otherwise exposed to in the ongoing pandemic, the societal risk from such side-effects will typically be entirely negligible. After all, unlike the virus, blood clots (and other possible side-effects) are not contagious.⁷

This all creates a strong presumption in favour of expanding access to untested or experimental treatments and candidate vaccines in the context of a pandemic. In order to justify continuing such prohibitions in a pandemic, government agencies ought to present a cost–benefit analysis showing that the consensus opinion of medical experts implies that the harm from potential side-effects straightforwardly outweighs (in expectation) the benefit of potential protection against the pandemic. When the status quo is so inherently risky, it is no longer justifiable to privilege the status quo by default. A positive case must be made on its behalf. In the absence of a positive case for judging the status quo to be definitively safer, reasons of autonomy and the value of information-gathering together provide undefeated and unopposed reasons for preferring a more liberal policy.

Caution and Recklessness

The argument of my previous section suggests the need for radical revisions to ordinary judgments about what constitutes caution and recklessness in pandemic policy. People typically privilege inaction by default, judging pauses to an intervention to be ‘cautious’, while plowing ahead with an uncertain intervention is automatically deemed to be the more ‘reckless’ option. But of course it would not be particularly prudent to sit still on the tracks while a train hurtled towards you, even if the exit ladder was overdue for a safety check.⁸ It all depends on how the risks from intervention compare to the status quo risks. So if I am right that the pandemic reverses the usual asymmetric balance of risk, this should prompt us to also reverse our usual judgments about prudence and recklessness.

Consider the Joint CDC and FDA Statement on Johnson & Johnson COVID-19 Vaccine, responding to early indications that one in a million vaccine recipients developed dangerous blood clots: ‘we are recommending a pause in the use of this vaccine out of an abundance of caution.’⁹ This echoes similar moves from many European countries a few weeks earlier, suspending their use of the AstraZeneca vaccine on similar grounds (albeit against the advice of the European Medicines Agency and the World Health Organization). Especially for the European nations, with their limited vaccine supply and slow roll-out in early 2021, this decision likely increased the total death count. If so, it seems misleading to describe such recklessness towards the virus as constituting ‘caution’, except in the pathological sense that someone irrationally terrified of pinching their finger in the buckle could count as ‘cautious’ of wearing a seatbelt.

Pathologies aside, it should be clear that what normatively matters is assessing overall risk levels, and not just protecting against one (comparatively minor) risk no matter the cost. With this correct understanding in mind, then, we can see that governments and their agencies are not generally entitled to describe vaccine suspensions as reflecting ‘an abundance of caution’, unless they can show that the policy actually reduces overall risk. If it instead increases overall risk, it would seem more objectively accurate to describe such suspensions as ‘reckless’—as they would then reveal a reckless disregard of the objectively greater threat posed by the unchecked spread of the virus.

For another striking example, consider the following quote from The Guardian: ‘Germany … was the first country to refuse to allow people over the age of 65 to have the AstraZeneca vaccine because of the absence of evidence of how well it worked in older people, indicating a more cautious approach than most.’¹⁰ It is objectively inaccurate to characterize as ‘cautious’ a policy that reduced vaccine access for the most vulnerable age group in the midst of a pandemic.

The UK’s Joint Committee on Vaccination and Immunisation recommended against vaccinating children under 16, despite granting that ‘the benefits from vaccination are marginally greater than the potential known harms’, noting that ‘the UK public places a higher relative value on safety compared to benefits’.¹¹ But this contrast between ‘safety’ and ‘benefits’ makes little sense when the benefits from vaccines are specifically safety benefits, i.e. protection against the risks from Covid, not enhancements above the baseline of good health. The choice was not between safety vs. benefits, but between safety vs. safety. So the assessment of which risk was greater should be made fairly, without rhetorically leaning on the scales in this way.

Finally, consider the debate from early 2021 over whether to prioritize the timely delivery of second doses (as per US policy) or to follow the British in delaying second doses in order to allow more of the population to sooner receive a first dose of the vaccine.¹² It seemed common to conceptualize the prioritization of second doses as the ‘safe’ option, even in the absence of cost–benefit analysis or other evidence to suggest that it reduced overall risk from the virus.¹³ The alternative policy of ‘First Doses First’ was admittedly higher variance, in that it conceivably could have yielded results that were either better or worse than the (more predictable) status quo.¹⁴ But lower variance does not by itself render an option ‘safe’, as it may instead be invariably bad.¹⁵

Of course, some readers may judge these particular cases differently. That is fine—I do not here mean to settle any empirical disputes. My purpose in this section is instead to cast doubt on conventional assumptions about what determines whether a policy is ‘cautious’ or ‘reckless’. My suggestion is that this cannot be settled independently of the first-order disputes over the merits of the competing policies. Rather than blindly favouring the status quo as intrinsically ‘safe’, we should judge the safer option to be whichever one best reduces overall risk. Specifically, I have argued that (1) it would be a serious moral error to focus on a single source or type of risk (neglecting greater overall risks) in making such evaluations; and (2) we should be open to the possibility that a higher-variance option may best reduce overall risk. Both of these principles seem to be routinely overlooked in public discourse concerning pandemic risk. We need to chart better paths, and the faint risk that any such attempt might lead us astray is simply not a good reason to settle for a disastrous status quo.

Human Challenge Trials

The failure to approve Human Challenge Trials for promising vaccines at the very start of the pandemic offers an especially clear example of harmful biomedical conservatism. The Moderna vaccine was designed in mid-January, reported 94.5 per cent efficacy in mid-November, and was finally authorized for emergency use by the FDA in mid-December of 2020.¹⁶

Vaccine field trials take a comparatively long time, as they require waiting for a sufficient number of participants to suffer natural infection to confirm the vaccine’s efficacy relative to placebo. Human Challenge Trials instead directly expose participants to a ‘challenge strain’ of the virus, potentially confirming vaccine efficacy much sooner. Bioethicists have vocally championed Covid challenge trials on this basis,¹⁷ and the advocacy organization 1Day Sooner <1daysooner.org> enlisted over 30,000 global volunteers, while their open letter was signed by moral theorists as diverse as Peter Singer and Christine Korsgaard.¹⁸

The obvious case against Covid challenge trials was that they involve deliberately infecting people with a potentially deadly virus. That sounds bad, on its face. What if a participant died? That would, of course, be tragic. Yet, at the peak of the pandemic, thousands of Americans were dying of Covid every day. Speeding vaccine approval in the slightest (let alone by several months) could have saved many thousands of lives in the US alone,¹⁹ not to mention a sooner end to all the social and economic harms that accompanied the pandemic.²⁰ Prioritizing the paternalistic prevention of moderate risks to a few willing volunteers over everything else at stake here arguably constitutes a kind of moral insanity.

This judgment does not rely upon controversial utilitarian premises. For this is not a case of utilitarian sacrifice, violating the rights of a few for the sake of the many. Challenge trials do not involve conscripting the unwilling, or deliberately deceiving participants about the risks. Rather, we are talking about volunteers who freely choose to accept the risks in order to immensely benefit the rest of society. To prevent them from making such a beneficial, autonomously chosen sacrifice makes about as much sense as banning altruistic kidney donations (or charitable donations more broadly). When considerations of autonomy and beneficence both point in the same direction, as they do here, it is very hard to see what further considerations could possibly outweigh their combined force.

We should be especially wary of anti-beneficent paternalism: forcing individuals, for their own good, not to promote the general good. As a general rule, we should not prevent people from performing beneficent acts that help others more than they harm the agent themselves. Reasonable people can dispute the conditions under which individuals might be forced to sacrifice their own interests to better promote the general good. But forcing them in a way that is contrary to the general good is straightforwardly unreasonable. Applied to the present case: it is unreasonable to oppose Covid challenge trials out of concern for the participants’ interests, or the sheer level of risk that they would be (voluntarily) undertaking.²¹

Some might wonder whether our immense uncertainty and ignorance regarding COVID-19, especially early in the pandemic, would have undermined the possibility of obtaining genuinely informed consent. But this rests on a confusion about the nature of informed consent.²² In addition to consenting to specific well-understood risks, participants can also consent to a procedure that they understand may expose them to unknown risks. So long as researchers do not deceive participants, for example by deliberately downplaying the level of risk involved, there need not be anything exploitative about conducting such research. Competent adults freely act in ways that will have unpredictable outcomes all the time; there is nothing inherently problematic about this.

Granted, it would certainly be unfortunate if some participants later regretted their decision, or if their consent was the result of some unwitting confusion. We should try to avoid such outcomes insofar as we easily can (e.g. by designing clearer consent forms). But the moral severity of this misfortune should be kept in perspective. It is not a grave moral evil (compared to thousands of deaths and millions of reduced-quality life-years), and the mere possibility of occasionally imperfect consent is not sufficiently weighty to register as a serious reason to oppose challenge trials.

Ethical objections²³ to Covid challenge trials thus seem to reveal a staggering lack of moral proportionality. Whatever ethical considerations might technically be raised against challenge trials, they involve no egregious moral violations, and only the most egregious moral violations could possibly outweigh the extraordinarily strong moral reasons we had to pursue a sooner end to the pandemic. (Of course, if they would not actually serve to achieve this end, then that is another matter. It is not my task in this paper to settle such empirical questions.)

Challenge trials could have had other benefits in addition to speeding initial vaccine approval. For example, given their greater efficiency, it would have been feasible to test the efficacy of a wider range of vaccine dosage regimens—including single doses, fractional dosage, ‘mix and match’ doses from different vaccines, and longer waits between doses—to help inform subsequent vaccine distribution policy.

Moreover, the basic argument for challenge trials carries over to non-vaccine research for which the expected social value of the information gained similarly outweighs the risk of harm to the volunteers. Given the scales involved, this condition will generally be met for any research that has a feasible chance of reducing the population-wide toll of the pandemic. For example, it could have been well worth pursuing early research into ‘variolation’, or inoculation via low-dose viral infection. In the historical case of smallpox, variolation reduced fatality rates from 30 per cent down to 1–2 per cent,²⁴ and medical experts have noted that coronavirus symptom severity seems similarly influenced by initial viral load.²⁵ The expected value of early-pandemic research into variolation is thus high, as it safeguards against worst-case scenarios in which we are unable to swiftly develop effective vaccines, and it enables policies of Early Targeted Immunity (see next section). Other challenge trials might have tested for cross-immunity from cold coronaviruses, or tested candidate preventative measures such antiseptic nasal sprays.²⁶ We might expect most such trials to yield negative results, but discovering any cheap, safe and effective medical interventions early on could have made a big difference—easily sufficient to outweigh the aggregate costs of all the experimental research.²⁷ So, again, there were no sufficiently weighty moral reasons to prohibit altruists from volunteering to participate in such socially valuable research.

I would further argue that, in the event that we lack sufficient volunteers to support such socially valuable but risky research, it would be ethical to financially compensate participants at whatever level of payment is necessary to make participation worthwhile for them. Such payment raises worries about exploitation via ‘undue inducement’—that the prospect of payment risks inducing poorer individuals to participate against their best interests, undermining their capacity to rationally consent. While we certainly ought not to deliberately exploit anyone, it is worth stressing that reducing (or even eliminating) the benefits to participation seems like a very backward way to address this worry—it is unlikely to be appreciated by would-be participants, after all, and for good reason: reducing the quality of the options available to them is not a very helpful thing to do! A more helpful way to address concerns that participation might be against the interests of the participants would be to increase the rewards for participation, so that it is more clearly worthwhile (offering a superior risk premium than one finds for everyday risky jobs in mining and forestry, say).²⁸ Given the immense social value of the research, we could—and should—sufficiently reward the participants for their service as to render it a ‘win-win’ for all involved.²⁹

It is again conceivable that, despite our best efforts, some participants end up offering imperfect consent. But again, imperfect consent is merely something that we should try to improve upon inasmuch as we feasibly can. We should not deliberately exploit it, but nor must we prevent it at all costs. It would reveal a gross lack of moral proportion to treat the mere possibility of imperfect consent as a decisive consideration that takes lexical priority over all else in a pandemic. In particular, it is not a sufficiently weighty reason to block socially valuable research that could save thousands of lives and improve millions more.³⁰

Early Targeted Immunity

Something that was largely overlooked in the early stages of the pandemic was the potential value of a policy of Early Targeted Immunity for individuals in social roles that entail either an especially high risk of transmitting the virus to vulnerable populations (e.g. medical and elder-care workers) or a high social cost to shifting to remote work (e.g. K-12 teachers). Indeed, a wide range of ‘essential’ service roles, from childcare providers to grocery clerks and Uber drivers, seem likely to yield disproportionately high social value from early immunity, greatly slowing the spread of the virus. Targeted immunity could also be used to contain outbreaks via ‘ring vaccination’: ‘A [Deliberate Exposure to Induce Immunity] program that is targeted toward individuals who work or live in close proximity to recent or anticipated outbreaks could create a “ring” of immunity that prevents or reduces spread, and might be able to do so with relatively few participants’ (Streiffer et al., 2021, sec. 3.1).

Early targeted immunity could be provided through viral inoculation: deliberately exposing volunteers from the target group to a low dose of the virus, followed by strict quarantine until no longer infectious.³¹ Depending upon the results of early research into variolation, it is entirely possible that such low-dose exposure could even be in the medical interests of the volunteers, by protecting them against the greater risk of an accidental high-dose exposure. But even if not, it could easily be in the interests of society as a whole, and it would thus be worth compensating the participants in order to make it in their overall interest to provide this social value.³²

Given their low risk, young adults might be especially keen to secure early immunity and subsequently take on essential service roles (and safely socialize in their time off) rather than being locked down largely out of fear of the threat they would otherwise pose to older members of society. Risks could be further reduced by preceding viral inoculation with experimental vaccination—we now know that the major Covid vaccines are safe and effective, but even from the vantage point of our earlier uncertainty, experts may well have judged that experimental vaccination had high expected value (with significant potential benefit and minimal risk of any serious downside) for anyone who was otherwise certain to become infected with the virus.

This is obviously all speculative. But a central theme of this paper is that high uncertainty alone should not bias us against a pandemic policy proposal. The question is whether—taking the full range of possibilities into account, weighted by their probability—the expected value of the proposal should be judged positive or negative on net. This is an empirical question, best answered by the relevant disciplinary experts. As a philosopher, I restrict myself to claiming that this is a question that ought to have been asked, and that we can see this because it is prima facie plausible that the correct expert judgment could well have been that the proposal in question had immensely positive expected value (however uncertain, or high-variance). Given that this was a reasonable possibility, we can see that to not even ask the question was to irresponsibly condemn to death many thousands who very well might have been saved by a more careful survey of the policy options, guided by expert cost–benefit analysis.

Vaccine Hesitancy

Perhaps the strongest argument for medical conservatism, at least so far as vaccine policy is concerned, invokes the fear of vaccine hesitancy. Already, by early 2021, large swathes of the US public were irrationally reluctant to become vaccinated, and one might reasonably fear that any hint of ‘rushing’ the process, or failing to react strongly to the slightest hint of problematic side-effects, could have further exacerbated this problem.

I have three main responses to this concern. The first is that, as a delicate and highly uncertain empirical matter, it should be evaluated by relevant disciplinary experts, to feed into a cost–benefit analysis. Recall that the FDA waited a whole month after the release of the successful Moderna trial results to actually approve the vaccine. Did they consult with sociologists and psychologists to determine that this would so assuage the vaccine-hesitant that it would outweigh the thousands of deaths caused by the delay? When European nations over-reacted to rare blood clots by pausing the AstraZeneca vaccine, did they have better reason to expect that this would assuage or rather exacerbate vaccine hesitancy?³³ While one may always speculate that policies that are bad on the merits may have sufficiently good extrinsic (‘backlash’) effects to outweigh their intrinsic demerits, policy-makers should be guided by actual evidence, and it is far from clear that they were in these instances. Absent such evidence, it would seem more appropriate for public health policy to be decided on the basis of the medical merits rather than ungrounded sociological speculation about backlash effects.

This conclusion may be reinforced by a second concern, which is that the integrity of these institutions requires them to avoid spreading or reinforcing public health misinformation in short-sighted attempts to promote public health goals. This follows from the more general principle that public institutions ought not to engage in strategic deception of the public.³⁴ The idea that vaccine risks outweigh (either empirically or normatively) the risks of being unvaccinated during the pandemic is an instance of public health misinformation that is troublingly prevalent in our society. When public health institutions implement alarmist vaccine suspensions or other forms of vaccine obstructionism on strategic grounds, this communicates and reinforces the false message that the vaccine risks warrant such a response. Rather than trying to manipulate the public by pandering to unwarranted fears, public institutions have an obligation to communicate accurate information and promote the policies that are warranted in light of that information.

Third, and perhaps most controversially, there is a moral case to be made for prioritizing the interests of the ‘innocent’ (reasonable members of the public who desire earlier access to vaccines to be able to protect themselves and those around them) over those of the vaccine-hesitant (who have freely chosen to reject the available protection).³⁵

To see this most vividly, focus on some particular individual—call her Sophie—who died from Covid as a result of being deprived of early access to a vaccine that she strongly wished to take. Her government’s obstructionism was then causally responsible for her death: had they not blocked her access to the vaccine, she would have survived. Moreover, it is entirely foreseeable that people will die as a result of such policies, so it further seems that the government is morally responsible for her death. They have, in effect, indirectly killed her (and others), by blocking access to life-saving vaccines.

Now suppose that someone seeks to defend the obstructionist policy by arguing that it helps to reassure fearful members of society that the vaccines have been scrupulously investigated and are safe for them to (eventually) use. Further suppose, for sake of argument, that delays really would save more lives on net by winning over vast numbers of the vaccine-hesitant. We can still ask: is that worth it? Could you justify that to Sophie?

It would be one thing if we had to explain to Sophie that we could not save her without endangering a greater number of innocent people (that is, people facing comparably grave threats through no fault of their own). That would seem perfectly understandable, however unfortunate it might be. But that is not the situation here. Those who refuse vaccines are not ‘innocent’ in the relevant sense, as they are freely choosing to reject the protection that is available (or would be available if not for their unreasonable attitudes). Those who die of Covid as a result of freely rejecting vaccines are responsible for this outcome: the heightened risk they faced was self-inflicted.³⁶ And as a general moral principle, we should not harm innocent people, like Sophie, merely in order to convince benighted fools not to harm themselves (Graham, 2020).

To further illustrate the principle, suppose that extreme anti-vaxxers, constituting 10 per cent of the population, strap bombs to their chests and threaten to kill themselves en masse unless the government immediately and permanently outlaws all Covid vaccines. Should we appease them, and let the pandemic continue since it would not do as much harm as this mob was threatening to self-inflict? Surely not. Even if harms to innocent victims are smaller in magnitude than the threatened self-inflicted harms, the harms to innocent victims matter more.

And so it goes in the less-extreme case. The vaccine-hesitant are being foolish and facing self-inflicted risks as a result. It is not right or fair to harm innocent, sensible people in order to protect fools from self-inflicted risks. It would certainly be a shame if people died as a result of their own foolish fears, but it is even more of a shame that completely innocent people have died through no fault of their own. Contra utilitarianism, we should care more about the latter than the former.

A complication: some innocent individuals may be indirectly harmed by the vaccine-hesitant getting infected and then spreading the illness to them. So we should build into our analogy that some innocent bystanders might also get hurt in the blast. If it would actually save more innocent people to appease the mob, and if it would not have bad long-term effects (incentivizing similar hostage-taking in future)—both big ‘ifs’!—then it may be that appeasement is called for in such special situations. But it would seem generally unlikely that the necessary conditions would in fact be met.

So, while we should certainly want to reduce vaccine-hesitancy all else equal, it would not typically be ethical to do so via the means of killing innocent people. And vaccine obstructionism amounts to killing innocent people. The moral bar for justifying such obstruction is thus much higher than policy-makers seem to standardly appreciate.

Pandemic Policy Expertise

I want to wrap up this paper with some brief reflections on what it means to ‘follow the science’, or more generally, to take expert opinion appropriately into account when determining pandemic policy.

It is of course vitally important for policy-makers to work with accurate empirical information, both about the current state of affairs and about the possible and likely consequences of candidate interventions. Scientific experts are the best source we have for providing policy-makers with these essential inputs. But, crucially, science per se tells us nothing about what should be done with these inputs. Especially when making difficult trade-offs, or evaluating how to respond rationally to risk and uncertainty, there is no reason to expect doctors or scientists to be especially well-placed to respond appropriately to their data.

If we blindly defer to doctors and scientists, the resulting policies will be distorted by whatever implicit normative bridging principles they happen to unreflectively hold. These are likely to be unduly conservative (since most people exhibit a wide range of conservative biases). They may oppose challenge trials and other beneficial policies as ‘too risky’, not because they have a more accurate conception of what the risks actually are, but because they lack moral understanding of when risks of that magnitude can be justified. To ward off bias and ill-considered assumptions, policy-makers need to feed their empirical data through a robust process of cost-benefit analysis. It is not enough to one-sidedly consider potential costs of action, and then lazily assume that the status quo is automatically superior.

There are philosophers and decision theorists who have spent much of their lives thinking carefully about how to make rational decisions under conditions of uncertainty, and how to go about making difficult trade-offs in extremely high-stakes contexts. For policy-makers to wantonly violate all that these experts know about how to make good decisions is—and should be more widely appreciated as—every bit as deplorable as wantonly disregarding accurate scientific inputs into the decision-making process.

Conclusions

We have seen that much pandemic policy (and ordinary moral thought about these policies) is rife with status quo bias. Possible risks from medical intervention are inflated in perceived moral significance, while objectively greater risks from the pandemic are comparatively neglected—even when these status quo risks involve thousands upon thousands of excess deaths that could have been prevented. There are presumably systemic reasons to do with institutional incentives that can help explain the causal provenance of this moral disaster.³⁷ It will be important for social scientists to clarify the causal mechanisms that reinforce status quo bias in public health policy, and for changes in institutional design to set us on a better track next time around.³⁸ But a crucial first step is to make the moral case that change is needed, as our current institutions have made disastrous and indefensible moral errors. That first step was the task of this paper. I can only hope that this conclusion becomes sufficiently widely appreciated—and acted upon—that we do not end up repeating these mistakes when the next pandemic strikes.

Conflict of Interest

None declared.

Endnotes

Acknowledgements

Thanks to Lara Buchak, Daniel Fogal, Daniel Greco, David Killoren, Govind Persad, Philip Pettit, Helen Yetter-Chappell, anonymous referees and the associate editor for this journal, audiences at NYU and Princeton University, and blog commenters at philosophyetc.net and on Facebook, for helpful comments and discussion. The author also thanks the University of Miami Fellowship in the Arts & Humanities for supporting his work on this project.

References