Mashup of the coronavirus and a trolley
(credit: NIAID CC BY and withvengeance86 CC BY)

On viruses and trolleys

We can catch a virus on a trolley, but there is another connection — or why we sometimes must think the unthinkable

The coronavirus, in the meantime rechristened COVID-19 by the World Health Organization, continues to hold the world’s attention. We receive updates on the progression of the disease in every main news bulletin — the number of infections, the number of deaths — as well as regular stories about people affected by the condition, or more often, by the measures to try and stem the spread.

We can easily understand the impact for real people confined to the quarantine quarters in hospitals worldwide or stuck on a cruise ship, and we can sympathize with anyone who has lost a loved one to the virus. But it is much harder to grasp what COVID-19 really means for public health, in our towns in cities, in our country, and indeed globally.

What is the significance of the numbers we hear? As these words are being written, the number of deaths has exceeded 2,000, and the number of infected people is estimated at well over 75,000. Is that a lot? Should we be worried?

The virus is very contagious, spreading easily through airborne respiratory droplets sneezed or coughed out by an infected person (who does not need to be ill). Its , which expresses how many secondary infections typically result from a single case is still quite uncertain, with estimates between 1.4 to 6.6 — from ‘no great cause for alarm’, to ‘prepare for a severe pandemic’.

Similar uncertainty seems to exist regarding the likelihood of dying of the virus. China’s Ministry of Health estimates the case fatality rate (CFR) to be between 2 and 3%, but other estimates range between 0.1% and 15%. Even a CFR of 2–3%, comparable to SARS and measles is far from trivial — much higher than that of seasonal flu, at around 0.1% ( source for these figures).

Asian person with face mask
Asian person with face mask
Will facemasks continue to provide protection? (image: Daniel Arauz CC BY)

For the time being, the virus still seems to be largely confined to China, and indeed to one province. Of the 75,000 confirmed cases in China, about 60,000 are in the Hubei province, with fewer than 1000 cases outside the country. But the uncertainty about the contagiousness and the severity is a challenge for public health: if COVID-19 becomes a full-blown pandemic and the number of cases worldwide suddenly starts rising, the consequences may be rather serious, both for the healthy and the sick.

With a few percent of employees off sick at the same time, healthy colleagues in businesses and public services can usually take up most of the slack. But if sickness and severe confinement measures (like the ones seen in Wuhan, where the outbreak first happened: lockdowns with strict travel restrictions and public transport suspended) stop many more people from working, economic life will collapse. If 20% of employees don’t turn up to work, economic activity will drop by much more than 20%, and the supply of essential goods may be severely disrupted.

Healthcare systems have limited capacity, and the spectacular feat of building a dedicated, 1000-bed hospital in just ten days is not something that is scalable. Even the most advanced healthcare services will clog up with the number of patients of a significant pandemic. And infrastructure is just one constraint: if medical professionals cannot work because they too are affected by the virus, the problem will be even worse.

Would it be possible to intervene in some way, before the virus has spread beyond the capacity of society to cope? This is what economist Robin Hanson wonders in a recent blogpost. He makes the comparison with the practice of deliberately setting fire to vegetation in so-called controlled burns. Such small fires prevent wildfires from growing so large they get out of control. Two smaller fires one after the other are easier to put out than one large one.

So Hanson (no stranger to controversy) asks whether deliberately infecting certain people, like critical workers in healthcare and elsewhere, in a planned way might ensure they have gained immunity before the wave of infections hits. If medical staff is thus protected, the health system will be able to operate at close to full capacity when it is needed, and if the peak can thus be spread out over a longer period, the worst of the disruption may be able to be avoided.

Controlled wildfire with fireman
Controlled wildfire with fireman
Lighting a small fire to prevent a large one (image: BLMIdaho CC BY)

Such an intervention would not be without downsides, though. Administering a live virus to thousands of people will almost certainly lead to some deaths. We are faced with a moral question that is closely related to the so-called trolley problem (see this post for more). In this thought experiment, originally posed by the British philosopher Philippa Foot in the 1960s, a runaway trolley is speeding down a track towards a section with workers carrying out repairs, unaware of their imminent fate. However, the subjects have the option of operating a switch which would divert the trolley to a different section, where just one person is working, and thus to sacrifice the life of one individual to save the lives of several others.

Infecting part of the population is much like pulling the lever: it will cause the death of a small number of people (and some harm to many more), but it averts the deaths of many more people as a result of a pandemic on a scale that totally exceeds the capacity of a country to handle. But deliberately causing the deaths of citizens is not something that we expect from our governments, and Hanson himself calls the idea “disturbing”.

Would such a controlled exposure be morally acceptable? The question throws the fundamental tension between deontological and consequentialist perspectives from its comfortable habitat of academic philosophy right into current reality. Infecting people and knowingly causing harm and death is the kind of thing we associate with depraved Nazi camp doctors, and violates the most basic moral norms. That would make it utterly repugnant. On the other hand, if not intervening would cause many more deaths than would result from planned infection, would that not be the more morally objectionable choice?

In the dual process model for moral cognition, first proposed by Harvard psychologist, neuroscientist and philosopher Joshua Greene, deontological judgements correspond with fast, emotional and intuitive reactions, and consequentialist or utilitarian judgements with slow, conscious and deliberate reasoning. (It can be seen as a specific case of the System 1/System 2 model that psychologist and Nobel laureate Daniel Kahneman described in Thinking, Fast and Slow.)

Tempting as it is to associate the latter approach with more developed intelligence, this is not so, a paper by Dries Bostyn, a philosopher at Ghent University and colleagues argues. It is the motivation to deliberate, rather than the ability that correlates with a preference for consequentialist positions. Might it nevertheless be something that is more prevalent in Western, industrialized societies than in small tribal communities? That is what a paper by Piotr Sorokowski, a psychologist at the University of Wroclaw, and colleagues might suggest. The researchers found that members of the Yali tribe in Papua were much less willing to sacrifice one life to save five (they used a ‘falling tree’ dilemma, rather than the trolley one, recognizing that trolleys are uncommon in the Yalimo valley). But that is not necessarily indicative of deontological judgement. In this community, the consequences of causing a person’s death are harsh: not only is “a person accused of contributing to someone’s death […]killed”, their extended family and indeed their entire village are also at risk. Customs dictate that the relatives of the deceased person must “compensate for the wrongdoing by killing the same or a greater number of persons”. This provides a distinctly consequentialist explanation for their preference.

Graffiti of the word ‘Regret’ on a wall
Graffiti of the word ‘Regret’ on a wall
The writing is on the wall — regret, whatever you choose (image: duncan c CC BY)

How would a decision-maker feel after either making a deontological or a consequentialist choice? This is the subject of a paper by Jacob Goldstein-Greenwood, a psychologist at the University of Virginia, and colleagues. They explored two distinct types of regret: affective regret — the negative feelings about a decision — and cognitive regret — thoughts about alternative outcomes of a decision. The research suggests that people with a utilitarian preference (i.e. prepared to cause harm to save lives) experience stronger emotional regret, and less cognitive regret than those with deontological ethics (who are unwilling to make the sacrifice). Utilitarians will feel distress, but be less inclined to (want to) change their decision, while deontologists are emotionally OK with their choice, but are more likely to wonder whether they should have changed their minds.

None of this, unfortunately, points towards a simple, obvious way forward. There are arguments for and against both approaches, and no matter what the authorities do, there will be objections. We may find this strange, since governments make other choices of a similar kind. An obvious example is war: this generally means both soldiers and civilians are deliberately sacrificed for the benefit of some greater good. Laws and enforcement regarding traffic safety and air pollution are an example of the opposite: not actively intervening. Lower emission norms and lower speed limits would lead to fewer deaths from respiratory problems and fewer road fatalities and serious injuries, yet they remain unchanged.

It is remarkable that Robin Hanson has received not just reasoned criticism for his idea, but also insults and even, apparently, threats of physical violence — simply for what is effectively thinking the unthinkable. It is clear that many commenters did not actually read the blogpost, and many of the objections contain no argument, simply a dismissal. This does support the idea that deontological judgements are typically not deliberate and reasoned, but rather inspired by intuitions and emotion.

But in all profound decisions, and definitely if the stakes are as high as they are in respect of COVID-19, we should reject the notion of absolute no-nos. This is precisely when it is essential to think out of the box, without limitations, and dispassionately consider the pros and cons of all the options — including the unthinkable ones.

Originally published at on February 21, 2020.

Thank you for reading this article — I hope you enjoyed it. Please do share it far and wide — there are handy Twitter and Facebook buttons nearby, and you can click here to share it via LinkedIn, or simply copy and paste this link. See all my other articles featuring observations of human behaviour (I publish one every Friday) here. Thanks!

Accidental behavioural economist in search of wisdom. Uses insights from (behavioural) economics in organization development. On Twitter as @koenfucius

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store