Is It Because He's Fat?

A Piece of Bigotry

Three years ago, just before Should You Kill the Fat Man? went live, we mentioned its existence on a professional philosophy mailing list in the hope of getting some useful feedback, etc. In among the responses was this little missive:

I have to say that I find this experiment offensive, and always have. There is absolutely no reason in the logic of the experiment that the person sacrificed should be a "fat man" or "a (very) fat man." There is a clear subtext that killing a fat man is somehow different from killing a man.

I am glad to see that some versions of the experiment have recently removed this piece of bigotry — I just recently saw one in which simply "a person" was suggested to be sacrificed. But the perpetuation of this gruesome story in its obviously bigoted anti-fat form is regrettable. The fact that Prof Foot may have invented it in that form is hardly an excuse.

There are a number of things to say about this. First, it can't simply be "a person", because that immediately raises the possibility of self-sacrifice (i.e., why not throw yourself off the bridge rather than push somebody else off?). Second, Judith Jarvis Thomson, not Philippa Foot, originated the fat man variation of the trolley problem. Third, there is no clear subtext that killing a fat man is somehow different from killing a man (not least, it's relevant here that the fat man variation is intended to provoke the intuition that throwing him off the bridge is not morally permissible).

The Evidence

Having said all that, the fatness of the fat man is a possible confounding variable: in other words, if somebody is "anti-fat", then this might be a factor in how they respond to the fat man scenario(s). So what does the evidence show?

Although we're not able to say with any degree of certainty - because Should You Kill the Fat Man? was not designed to shed light on this issue - there are a couple of pieces of data that suggest that if the fatness of the fat man is a confounding variable, it doesn't have a very big effect.

The first is the fact that people's responses to the fat man scenario tend to be consistent with the way they answered the earlier questions about their metaethical views. For example, the first chart to the right should show that utilitarians are much more likely to throw the fat man off the bridge than non-utilitarians, presumably because they think that this is what is required by their utilitarian beliefs (and not because they hate fat people).

The second, and perhaps more significant, piece of data concerns those people who say it would be wrong to turn the train in the standard trolley problem setup. If even a significant minority of this group state the fat man should be thrown off the bridge, it suggests that the fatness of the fat man is a relevant factor in people's thinking about this issue.

In fact, as we can see from the chart to the right, only a very small percentage of this group opt to sacrifice the fat man. Although the percentage isn't zero, given that we would expect at least some inconsistency in people's responses (consider, for example, that 20% of the same group opt not to throw the fat man off the bridge in the circumstance where he's responsible for damaging the train's brakes!), it is reasonable to conclude that for this subset of people at least, the fatness of the fat man does not loom large in their decision making process.