1 Introduction

The COVID-19 pandemic forces policy makers around the world to make ethically loaded decisions about the lives and well-being of individuals. One such decision is often presented as a choice between “saving lives and saving the economy” (Bazelon, 2020). This catchy phrase refers to the following important question: How and when should a country move from lockdown to reopening?

In his article “Beware the Trolley Zealots,” Gil Eyal argues that this reopening question is often framed as a trolley problem, forcing us into a utilitarian “moral trap” rather than providing adequate moral guidance (Eyal, 2020). He argues that on the saving-lives track lie the lives and well-being of the old and unhealthy and on the saving-the-economy track, of the young and healthy (p. 24). Eyal targets this framing, attacking the trolley problem in particular and utilitarian reasoning in general. But he goes beyond that and refutes a common aspect of policy making and moral reasoning: Weighing probable harms and benefits of a decision to understand the trade-offs, even under uncertainty. To illustrate his point, he offers a counter-frame — a Rabbinical story — and without endorsing either of these frames, he proposes policy decisions regarding essential workers and allocation of life-saving treatments.

Eyal’s picture suffers from three fatal flaws: (1) Misrepresentation of policies and arguments, (2) misapplication and mischaracterization of the trolley problem, and (3) reaching moral conclusions without moral justifications. The reopening question is a complex problem and the policies debated by public health experts and philosophers are nuanced. Once these nuances are captured, the question does not fit into the structure of a trolley problem anymore. If it did, we could have used this thought experiment to clarify the underlying moral considerations (such as doing versus allowing harm, duties, rights, and moral claims) because, in contrast to Eyal’s presentation, the trolley problem is not a utilitarian frame. Eyal acknowledges that neither the structure of a trolley problem nor the Rabbinical story captures the complexity that we face with the reopening question. But instead of offering moral reasoning, Eyal avoids this complexity by asserting moral judgments without any justifications. At the end of Eyal’s piece, what we are left with is a refutation of risk assessment and moral reasoning — in short, necessary components of ethical policy-making.

2 Saving Lives versus Saving the Economy

As experts have often repeated, the COVID-19 pandemic is unchartered territory (WHO, 2020a): It is a novel disease without a cure, treatment, or vaccine, whose scale makes our traditional methods for combating infectious diseases ineffective. This means that when countries get hit by exponentially growing numbers of cases that quickly overwhelm their healthcare system, the only sure way to reduce COVID-19 deaths is to issue a lockdown policy. No human interaction means no transmission. Lockdown policies still allow essential workers (healthcare workers, law enforcement officers, grocery store workers, and such) to continue their work to provide for the basic needs of individuals. This is what is often called the “saving lives” option.

But a lockdown is not a sustainable policy. Putting aside its probable effects on mental health (Campbell, 2020; Giuffrida, 2020; Thomas, 2020; van Hoof, 2020), lockdown means the suspension or slowing down of all activities as we know them: food production, logistics, education, individuals’ livelihoods (shops, restaurants, hair dressers, etc.), social events, wellness, and non-pandemic related healthcare, just to name a few. Since there is no guarantee of a future treatment or vaccine, the lockdown policy must come with an exit strategy that relies on controlling the spread of the disease. The so-called “saving the economy” option — which really is more about saving our life in the fuller sense, as opposed to mere short-term survival — is about how and when to implement this exit strategy.

This is where opinions vary drastically. On one extreme is the Economy-First Approach that, as Eyal notes, Texas’ Lieutenant-Governor Dan Patrick defends, arguing that we should lift the lockdown immediately and power through the pandemic letting those at high-risk to “take care of themselves” (Rodriguez, 2020). This approach puts the elderly and those with co-morbidities at extremely high risk in order to benefit the young and healthy.¹ But this is not the approach adopted by public health experts and philosophers — certainly by none of the experts that Eyal refers to.² A common position among experts is a type of Gradual Exit Approach. The Gradual Exit Approach, as explained by bioethicist Ezekiel Emanuel in an interview that Eyal refers to repeatedly, would require a proper nation-wide lockdown, after which steps are taken to minimize the risks for everyone through an efficient tracing, testing, and isolation system, safe individual behavior (such as mask wearing, social distancing, and handwashing), and safety measures (such as re-organizing restaurants and workplaces for distancing). Then the society can gradually reopen by first letting the low-risk groups return to their schools and jobs. Only when the system works reliably, such that spikes of infection can be rapidly controlled and the risk is reduced to a minimum, can the high-risk groups return to their social lives as well (Bazelon, 2020).

Rather than engaging the moral complexity of implementing an exit strategy, Eyal writes as if the question debated by the experts is whether “we should pull the shunting switch, however ‘gradually,’” and send the train down the nursing homes tracks, rather than through the workplaces" (p. 23). By completely dismissing the Gradual Exit Approach and the opportunities that it offers, he forces us into a debate with two extreme choices: The Economy-First Approach and the Indefinite Lockdown Approach.³ In reality, the moral questions that the experts are grappling with are much more complex, as they are concerned with the details of a Gradual Exit Approach: How much risk is “minimum risk,” when should we prioritize living our lives in its full sense over saving more lives, what should be the mandatory restrictions on social life and on what grounds can they be justified, how do we balance individuals’ freedom to take risks and bans that restrict this freedom to benefit the high-risk groups, and is it fair to redirect all resources to saving a few more life years?

3 The Pandemic Trolley

By artificially narrowing the problem to two extreme choices that no public health expert defends, Eyal creates his own trolley problem: On the Economy-First Exit Track, we have the old and unhealthy strapped to the tracks and on the Indefinite Lockdown Track, we have the young and healthy (p. 24). Shall we pull the “switch, however ‘gradually,’ and send the train down” to kill the old and unhealthy rather than the young and healthy (p. 23)? Who should we sacrifice?

I will show below that the real decision society has to make does not fit into a trolley problem structure at all. However, let me pretend for a moment that it does, but incorporate into the analysis the “however ‘gradually’” qualifier Eyal mentions in passing but simply ignores. The starting point (or rather current track) of our trolley is a lockdown.⁴ So imagine the trolley is heading towards a continuous, indefinite lockdown where the first victims will be the poor (who do not have enough savings to live from), those who are domestically abused (there is no more escape from home), and those who need non-pandemic related care. As the track continues, the number of victims will grow and vary due to the disruption of goods and services resulting in lack of drugs, food, mental and physical care, education, training, scientific progress, and social interactions. Keep in mind, while we hope that the trolley will stop abruptly at some point (if/when a treatment or vaccine is found), we cannot estimate how far down the line this will happen or even be certain that it will happen. This is the Indefinite Lockdown Track. Now, we can pull the lever fast and send the trolley flying into the Exit Track, where the elderly and those with co-morbidities will perish alongside some young and healthy members of the society. This would be an Economy-First Exit. Thus far, this is Eyal’s setup. However, we can also pull the lever gradually towards the Exit Track and by doing so slow down and start controlling the speed of the trolley. While the old and unhealthy are still on this track, now we can slow down the trolley to take these individuals to a safe place away from the tracks. In fact, we can always pull the lever again to turn the trolley to the Lockdown Track and back and forth. And this would be the Gradual Exit Approach that Eyal simply skirts.

Clearly, the trolley structure is not a good model for the reopening question. The Gradual Exit Approach gives us the ability to control the speed of the trolley and the option to switch between Lockdown and Exit Tracks as needed — it is not the stark, binary choice of the trolley problem. Similarly, some individuals lie on both tracks, again in stark difference to the trolley problem formulation: The elderly and the unhealthy will be most severely affected from both an indefinite lockdown and a fast exit.

Until this point, I focused on the structure of the trolley problem but what about its function? What is the purpose of employing the trolley problem? According to Eyal, it is “an individualist, utilitarian fable that presents the solution to be ‘obviously’ some form of calculus of known benefits and losses” (p. 24). This is far from the function of the trolley problem. The trolley problem as constructed by Phillipa Foot, detailed by Judith Jarvis Thomson, and further developed by Frances Kamm (among others) is not a simple numbers game — in fact, none of these philosophers are even consequentialists, let alone utilitarians (Foot, 1967; Kamm, 2006 & 2015; Thomson, 1976). The trolley problem is about distinguishing the morally relevant underlying factors — such as intentions, causation of harm, duties, rights, moral claim, and the lives at stake — when making a moral decision.⁵ In contrast to Eyal’s presentation, the trolley problem does not assume certainty; in fact, as Foot herself discusses, it is a tool to evaluate decisions both under uncertainty and certainty (Foot, 1967). And finally, Eyal claims that the trolley problem forces “diverse temporal rhythms into a single instant of choice” but what it really does is to force the decision-maker to take into account the foreseeable future consequences of their action.

If we could employ the trolley problem for the reopening question, it would have helped us in evaluating the underlying issues such as the morality of taking an action (i.e., forcing people to stay in) that costs individuals’ livelihoods and probably their mental and physical well-being, the morality of taking an action (i.e., lifting the lockdown) knowing that it will result in avoidable deaths,⁶ the moral claim of individuals to healthcare, to their freedom, and to be protected from a variety of risks, state’s duties to its citizens, and of course the trade-off that each action presents in terms of lives or life-years lost.

4 Making Moral Decisions

Eyal’s main point is neither about revealing the details of actual policy choices we face nor about creating the most accurate trolley analogy to help us think through difficult decisions. Rather his main argument is that the utilitarian calculation is an inappropriate tool to employ when lives are at stake. He argues that we need to resist public health experts and philosophers framing this policy decision as a utilitarian trade-off with “known benefits and losses” (p. 24).⁷

Let us first take a look at how the utilitarian reasoning approaches this policy decision and then turn to Eyal’s “counter-frame” and his moral conclusions. The utilitarian reasoning aims to take into account all relevant harms and benefits that will follow from a policy decision, give them appropriate weights in a common unit of value, estimate how probable they are, compare them, and choose the policy that results in least harm. Utilitarian policy making would engage in this demanding task to answer various complex ethical questions: At what cost to the society should we take the option that is likely to save most lives right now? (Note that if we operate with the policy of “saving most lives right now at all cost,” we would have to spend all our resources on healthcare and law enforcement rather than, for example, education or unemployment benefits. This policy would also outlaw driving, make it mandatory to exercise and eat healthy, and keep the elderly in lockdown in every flu season. This is certainly not how societies think about the trade-off and such a position would require a strong moral justification.) Is it morally right to sacrifice a young child’s life in order to save an unhealthy elderly person who lived a good life and is expected to live only a few more years or months? What about the ethics of redirecting resources from other deadly diseases such as tuberculosis, which makes about 10 million people fall ill every year and kills about 1.5 million people annually (WHO, 2020b)?

These are not novel questions. Public health experts and bioethicists must make difficult policy decisions about scarce resources all the time. Organ transplantation is an obvious example but even the decision to allocate resources between preventive vaccines for children and life-saving end-of-life treatments for the elderly involves this type of trade-off. Ethicists have developed various arguments and tools to answer the underlying moral question: What is the morally right and just public health policy to implement? Utilitarianism is one approach to answer these questions with a formula to maximize the good and minimize the harm. However, many other non-consequentialist moral reasoning (like Kamm’s use of the trolley problem and her comment on the pandemic in Kamm, 2020) and theories of justice (such as Rawlsian theory of justice and egalitarianism) also estimate the consequences and take into account the trade-offs that a policy offers when making a decision.⁸

Eyal not only opposes utilitarianism but he also rejects engaging with the trade-offs (p. 22-23). He invites us to stop “pretending that we know how to transform this uncertainty into calculable odds” during this pandemic (p. 23) and consider a counter-frame where the “society is mobilized to defend its weakest members” (p. 24). The counter-frame he presents is based on a Rabbinical story where Moses goes after a runaway lamb putting the whole herd at risk. Eyal says that “[f]raming the epidemic with this story would imply that society is mobilized to defend its weakest members, the infirm, the elderly, the immune-compromised, even when there is a price to be paid” (p. 24). Of course, this conclusion is a huge jump from the story. In the story, we have a herd that is abandoned without a choice and put at risk by the shepherd. This “society” — meaning, the herd and the shepherd — does not decide to defend its weakest members. If anything, Eyal’s retelling of this story suggests that the leader of the society makes a decision single-handedly to put the society at risk (taking much less risk himself, if any!) without consulting them and without their consent. Eyal also points out that the Rabbinical story captures complete uncertainty, presenting us an example of “moral leadership in a time of radical uncertainty [as opposed] to the hubris of presuming to know with certainty the calculus of utilities” (p. 25).⁹ In other words, Eyal suggests that just like the shepherd, we also have no way of calculating the risks for any given member of the society during this pandemic and trying to do risk assessment would be “hubris.”

Eyal does not endorse or ask us to endorse this Rabbinical frame either but nonetheless he reaches some moral conclusions about essential workers (p. 25) and allocation of ventilators (p. 26). But what would this Rabbinical story imply about these issues and how does Eyal reach moral decisions without engaging in any principled moral reasoning?

Eyal acknowledges that when we apply this story to the pandemic, putting essential workers at high risk does not sit well with the idea of defending all members of the society. He does not follow through with this line of reasoning but if he did, then this idea would entail that essential workers (many of whom, as Eyal recognizes belong to disadvantaged groups) should be allowed and even required to stay home without losing their jobs or their income. Instead, unlike Moses, Eyal has no problem “sacrificing the few for the many” (p. 24): He concludes, without any premises, that the right policy is to simply compensate essential workers for the “greater risks” that they have “no choice” but to take (p. 25). But what is his moral justification behind pushing, even incentivizing vulnerable members of the society to take such risks for themselves and their families? In contrast to his earlier claim that “the pandemic could serve as a lever for changing” “the American status quo” (p. 23), he simply accepts the desperation of essential workers as a given and suggests that giving a financial “thank you” is the right policy. In fact, if we embrace Eyal’s idea of complete uncertainty — that is, we cannot estimate the risks and attach any probability to any outcome — then can we even say that the essential workers are at “greater risk”? If we cannot assess the risks, then why should we compensate or reward them? It turns out, Eyal himself does not believe in complete uncertainty either.

There is another moral conclusion that Eyal draws and that is about the allocation of ventilators. Rejecting trade-offs, Eyal argues that the decision about whose life will be saved with a ventilator should only be based on “individual right to medical care and the dignity of the human person” (p. 26). From this, he concludes that the only relevant criterion for allocation is “who is more likely to be saved.” This is of course very puzzling and inconsistent. The idea of calculating the odds of who can benefit more from an intervention is an exercise of weighing consequences and engaging in trade-offs. In fact, if we must reason only from human dignity and a right to medical care, then allocation by first-come-first-serve or lottery would be more justifiable.

Eyal erects a straw man by turning a complex policy question into a mock trolley problem. Urging us to “do the thinking for ourselves” (p. 25), he ends up discarding all moral reasoning. For him, this is letting go of the “hubris,” but one might call it avoiding to face the hard facts, engage in risk assessment, and present a consistent moral reasoning. A morally right policy need not be utilitarian, but a policy decision cannot disregard its implications, which include the harm and the good the policy will bring about. And most importantly, we cannot abandon all moral reasoning simply because every choice we face during this pandemic involves unknowns and heart wrenching consequences.