In a variant of the classic ethical dilemma known as the trolley problem, a heavy runaway trolley is heading towards five people stuck on its tracks. You have to decide whether you let it continue down the tracks and kill the people or instead push one person in front of the trolley, stopping it and saving the group but killing the person you pushed. The choice consistent with simple utilitarian ethics is of course the latter, since it results in fewer people being harmed than the alternative, even if you actively cause the death of the person you pushed.
When someone’s utilitarian tendencies are measured using this question, they seem to correlate with some pretty surprising traits. For example, Bartels & Pizarro found that the utilitarian response can be predicted by psychopathic and Machiavellian personality traits as well as feeling that life is meaningless; more recently, another group wanted to test whether blood alcohol content changes people’s decisions in this context, and indeed, tipsiness is also correlated with the utilitarian solution.
These results seem counterintuitive at first, but less so once you think about how this kind of judgments actually are made. In morally difficult and emotionally distressing situations, people make decisions based on a variety of contradicting feelings and intuitions which certainly don’t form anything like a coherent set of ethical principles. This means that a lot of different factors ranging from subtle moods to inebriation levels are going to influence the responses to extreme dilemmas, often in very complicated ways.
In the trolley situation, then, there might be two opposing drives at play: 1) the obvious emotional reluctance to push a non-consenting fellow human being to their death, an action universally considered highly immoral, and 2) the realisation that four more lives will be lost unless the trolley is stopped, which leaves the world in a worse state and can therefore also be seen as the morally worse option. Someone familiar with consequential reasoning might stop the trolley because their understanding of 2 is stronger than average, and they have a genuine utilitarian desire to minimise bad stuff; someone else might stop the trolley simply because they don’t really feel 1, even if they don’t much care about 2 either. This in turn could be because they feel that life is empty and meaningless anyway so whatever, they have diminished empathy and don’t feel bad harming people so whatever, or they happen to be drunk so whatever.
This also means that these studies have very little to do with whether utilitarianism is a generally reasonable ethical position and the result of healthy intuitions or perhaps secretly a club for callous misanthropes who think pushing people in front of trolleys is a fun pastime. A recent paper in Cognition by Kahane et al elaborates on this:
Indeed, the sacrificial dilemmas typically used in current research represent only one, rather special, context in which utilitarian considerations happen to directly conflict with non-utilitarian rules or intuitions. To be willing to sacrifice one person to save a greater number is merely to reject (or overrule) one such non-utilitarian rule. Such rejection, however, is compatible with accepting extreme non-utilitarian rules in many other contexts—rules about lying, retribution, fairness or property, to name just a few examples, not to mention non-impartial moral norms permitting us give priority to ourselves, and to our family or compatriots, over others. Indeed, to reject a specific non-utilitarian moral rule (or even many such rules) is not yet to endorse the core of utilitarianism: the positive aim of impartially maximizing the greater good of all.
The group set out to measure a bunch of other indicators of utilitarian morality and investigate their relation to the responses to the trolley dilemma. They found that the decision to push doesn’t actually predict other utilitarian preferences at all, and is in fact negatively correlated with many of them. The people who stopped the trolley were, on average, less eager than others to give money to charity, not bothered by non-utilitarian behaviour in other less personal contexts, and not particularly sympathetic to the suffering of people far away even when it was explained in terms emphasising utilitarianism. In line with previous research, they also scored higher on anti-social traits and were less willing to jump on the tracks themselves instead of sacrificing a stranger when given the option in a modified version of the dilemma.
It seems that a lot of previous research, in which questions similar to the trolley problem are frequently used, is on the wrong tracks. (Oh, ha ha.) The entire dilemma is generally presented as a counterargument to utilitarianism – I can see why this could mean that the folks who bite the bullet are the true 100% Heavy Metal Ütilitarians so this is the best way to tell them apart from all the others who shouldn’t be counted, but that was a good guess only until we started getting data that these people don’t seem like the type to care about maximising wellbeing at all (and now, finally, pretty straightforward evidence that they reject more utilitarian intuitions than the average person). Sacrificial dilemmas like this are a bad, noisy way to assess whether or not someone will make utilitarian-ish decisions in realistic situations, and what kind of traits lead to endorsing a more consistent utilitarian approach. This is not to say that these dilemmas aren’t important questions in utilitarianism, just that they don’t really seem to measure the things they’re supposed to measure.