The Disunity of Utilitarian Psychology: Runaway Trolleys vs. Distant Strangers

Guy Kahane, Jim A.C. Everett, Brian D. Earp, Lucius Caviola, Nadira Faber, Molly Crockett, and Julian Savulescu in Practical Ethics:

Last week, we invited people to find out “How Utilitarian Are You?” by filling out our newly published Oxford Utilitarianism Scale. The scale was widely shared – even by Peter Singer (who scored predictably highly). The Oxford Utilitarianism Scale does a pretty good job of measuring how well people’s views match up with “classical” utilitarians (think Bentham and Singer), which is the form of utilitarianism we used to anchor the scale. But that’s not all it does. It also teases apart two different dimensions of utilitarian thinking, tracking two ways in which utilitarianism departs from common-sense morality. Our new research recently published in Psychological Review links these two factors to distinct components of human psychology.

The first peculiar aspect of utilitarianism is that it places no constraints whatsoever on the maximization of aggregate well-being. If torturing an innocent person would lead to more good overall, then utilitarianism, in contrast to commonsense morality, requires that the person be tortured. This is what we call instrumental harm: the idea that we are permitted (and even required) to instrumentally use, severely harm, or even kill innocent people to promote the greater good.

The second way that utilitarianism diverges from common-sense morality is by requiring us to impartially maximize the well-being of all sentient beings on the planet in such a way that “[e]ach is to count for one and none for more than one” (Bentham, 1789/1983), not privileging compatriots, family members, or ourselves over strangers – or even enemies. This can be called the positive dimension of utilitarianism, or impartial beneficence.

What are the psychological roots of utilitarianism? Why does utilitarianism attract some people but strongly repel so many others? Psychologists have tried to answer these questions by using the now famous (or infamous, depending on your perspective) ‘trolley problems’.

More here.