Some Thoughts About Temporal Trolley Problem

The emotional response to the coronavirus has been rife with bad arguments that have wider resonance. In particular, there is a strong dose of "one is one too many"-ism, which links, I believe, to a curious temporal variant of the trolley problem.

The classic trolley problem, for those who don't know it, is a moral thought experiment designed to tease out your moral values. Imagine that you are in the control box of a trolley [an urban train] service. An out-of-control trolley is hurtling towards a junction. You can pull a switch to change the trolley's route to Track B, or leave it alone on Track A. On Track A are lying some number of folks who will die if you do nothing. But on Track B are lying some other number of folks who will die if you pull the switch. What should you do?

What is effective about this framing device is that it forces a decision by making clear that these are both unpalatable options, and that not acting is itself an available option. There is no option to save everyone. You must choose what to sacrifice, and on what grounds.

A standard trolley problem involves two folks on Track A, and one folk on Track B. The greatest good principle argues that you should pull the switch to save two lives over one. A traditional moral intuition about the difference between active and passive behaviour argues that you should let the trolley follow its course uninterrupted, that you are not responsible for the lives you passively allow the trolley to take, but that you are responsible for the life you specifically step in to force the trolley to take. This aspect is made stronger in a variant in which you can stop the trolley by pushing a fat man out of the control box into the trolley's path, an even more direct killing of one folk, to save more lives.

The choice with the coronavirus is much simpler on its face. We must choose between Option 1, doing no more than usual and allowing the virus to run its course, which would only kill a few hundred thousand who were mostly already very sick and/or very ill; or Option 2, where we shut down the entire global economy, which would certainly greatly harm billions for the rest of their long lives, and very likely lead to the deaths of millions in the ensuing economic catastrophe. A traditional, staid look at the trolley problem logic makes this choice seem trivial: we should leave the trolley on its track, allowing the few to suffer and die, rather than divert the trolley, causing the many to suffer and die.

Yet most are choosing to actively cause the many to suffer, rather than passively allow the few to suffer. This is radically the opposite of conventional moral reasoning. Yet the plan's proponents passionately claim that they are in the moral right, going so far as to attempt to ban all mention of the opposing point of view, and fine or even jail those who do not conform to the qurantines. Why is such a straightforwardly immoral choice suddenly the conventional wisdom?

First, there seems to be an element of refutation of the traditional moral philosophy assumption mentioned above, that action is considered more morally culpable than inaction. Instead, there seems to be a moral bias towards action. Is this a change in attitudes, or has moral philosophy always been mistaken in believing that passivity is a moral intuition? That is beyond knowing at this time, but whatever the answer, for now at least it seems as though many of us, in this real-world setting (i.e. not in a laboratory or philosophy lecture), would rather do something to actively save a life, even if it ends up causing more harm overall.

It's also true that there are strong elements of excessive emotionalism, ignorance of the facts and their implications, and outright lying about motives. Humans feel emotionally inclined to obey authority, especially when they have been primed to feel fear, and they feel emotionally inclined to make kneejerk decisions. Many have no idea of the real facts of the situation, believing that the virus has a mortality rate above 3%, or that millions have already died. [Fewer than 9,000 have died at time of writing.] Many claim to want the greatest good when in fact they are motivated only by personal considerations. Emotionalism, ignorance, and mendacity are present in every human domain, and this is no exception.

But there also seems to be an important temporal discounting element. The true trolley problem variant is not quite as described above: instead, citizens are asked to choose between Option 1, passively allowing the few to suffer now, and Option 2, actively causing the many to die later. This interacts with the other measures: the action (shutting down the economy) is immediate, while the costs are in the future; the emotional impulse is stronger for short-term measures; economic ignorance makes the long-term costs easier to ignore; deception provides a veneer of justification for all of these other motives.

This conforms to what we know of temporal discounting. Many studies show that folks would rather get a small reward now than a large reward later. What is interesting is to see that this preference for immediate reward also applies to real-world morality as well as to laboratory incentives. Interesting, but of course also harmful for the folks who will suffer from the ill-thought-out recession that now looms over us.


sum θoətiz abaʊt ðə tempərəl trolii probləm

ðiiʸ imoʊʃənəl rispons tə ðə kəroʊnəvairəs av bii raif wið bad aəgyəməntiz ðat hav waidə rezənəns. in pətikyələ, ðeə biiʸ a stroŋ doʊs ov "um biiʸ um tʊʊ meniiy"-izəm, wic link, mii biliiv, tʊʊʷ a kyʊəriiəs tempərəl veəriiənt ov ðə trolii probləm.

ðə klasik trolii probləm, foə ðoʊz hʊʊ not noʊʷ it, biiʸ a morəl θoət eksperimənt dizain tə tiiz aʊt ðiis morəl valyʊʊwiz. imajən ðat ðii biiʸ in ðə kəntroʊl boks ov a troliiʸ [an uəbən trein] suəvis. an aʊt-ov-kəntroʊl trolii bii huətəliŋ təwʊəd a junkʃən. ðii kan pʊl a swic tə ceinj ðə troliiy-iis rʊʊt tə trak B, oə liiv it aloʊn on trak A. on trak A bii laiyiŋ sum numbəʳ ov foʊkiz hʊʊ wil daiʸ if ðii dʊʊ nuθiŋ. but on trak B bii laiyiŋ sum uðə numbəʳ ov foʊkiz hʊʊ wil daiʸ if ðii pʊl ðə swic. kwes ðii ʃʊd dʊʊ wot?

wot biiʸ efektiv abaʊt ðis freimiŋ divais bii ðat it fʊəs a disiʒən bai meikiŋ kliə ðat ðiiz bii boʊθ unpalətəbəl opʃəniz, and ðat not aktiŋ biiʸ itself an aveiləbəl opʃən. ðeə bii noʊʷ opʃən tə seiv evriiʸ-um. ðii must cʊʊz wot tə sakrifais, and on wot graʊndiz.

A standard trolley problem involves two people on Track A, and one person on Track B. The greatest good principle argues that you should pull the switch to save two lives over one. A traditional moral intuition about the difference between active and passive behaviour argues that you should let the trolley follow its course uninterrupted, that you are not responsible for the lives you passively allow the trolley to take, but that you are responsible for the life you specifically step in to force the trolley to take. This aspect is made stronger in a variant in which you can stop the trolley by pushing a fat man out of the control box into the trolley's path, an even more direct killing of one person, to save more lives.

The choice with the coronavirus is much simpler on its face. We must choose between Option 1, doing no more than usual and allowing the virus to run its course, which would only kill a few hundred thousand who were mostly already very sick and/or very ill; or Option 2, where we shut down the entire global economy, which would certainly greatly harm billions for the rest of their long lives, and very likely lead to the deaths of millions in the ensuing economic catastrophe. A traditional, staid look at the trolley problem logic makes this choice seem trivial: we should leave the trolley on its track, allowing the few to suffer and die, rather than divert the trolley, causing the many to suffer and die.

Yet most are choosing to actively cause the many to suffer, rather than passively allow the few to suffer. This is radically the opposite of conventional moral reasoning. Yet the plan's proponents passionately claim that they are in the moral right, going so far as to attempt to ban all mention of the opposing point of view, and fine or even jail those who do not conform to the qurantines. Why is such a straightforwardly immoral choice suddenly the conventional wisdom?

First, there seems to be an element of refutation of the traditional moral philosophy assumption mentioned above, that action is considered more morally culpable than inaction. Instead, there seems to be a moral bias towards action. Is this a change in attitudes, or has moral philosophy always been mistaken in believing that passivity is a moral intuition? That is beyond knowing at this time, but whatever the answer, for now at least it seems as though many of us, in this real-world setting (i.e. not in a laboratory or philosophy lecture), would rather do something to actively save a life, even if it ends up causing more harm overall.

It's also true that there are strong elements of excessive emotionalism, ignorance of the facts and their implications, and outright lying about motives. Humans feel emotionally inclined to obey authority, especially when they have been primed to feel fear, and they feel emotionally inclined to make kneejerk decisions. Many have no idea of the real facts of the situation, believing that the virus has a mortality rate above 3%, or that millions have already died. [Fewer than 9,000 have died at time of writing.] Many claim to want the greatest good when in fact they are motivated only by personal considerations. Emotionalism, ignorance, and mendacity are present in every human domain, and this is no exception.

But there also seems to be an important temporal discounting element. The true trolley problem variant is not quite as described above: instead, citizens are asked to choose between Option 1, passively allowing the few to suffer now, and Option 2, actively causing the many to die later. This interacts with the other measures: the action (shutting down the economy) is immediate, while the costs are in the future; the emotional impulse is stronger for short-term measures; economic ignorance makes the long-term costs easier to ignore; deception provides a veneer of justification for all of these other motives.

This conforms to what we know of temporal discounting. Many studies show that people would rather get a small reward now than a large reward later. What is interesting is to see that this preference for immediate reward also applies to real-world morality as well as to laboratory incentives. Interesting, but of course also harmful for the people who will suffer from the ill-thought-out recession that now looms over us.