December 01, 2011
Kill One To Save Five Lives?

Imagine you had a choice to divert a run-away train box car onto a different track and that doing so would kill one person while saving five lives. Would you pull the lever? About 90% of participants in a study chose to kill one to save five.

EAST LANSING, Mich. ó Imagine a runaway boxcar heading toward five people who canít escape its path. Now imagine you had the power to reroute the boxcar onto different tracks with only one person along that route.

Would you do it?

Thatís the moral dilemma posed by a team of Michigan State University researchers in a first-of-its-kind study published in the research journal Emotion. Research participants were put in a three dimensional setting and given the power to kill one person (in this case, a realistic digital character) to save five.

The results? About 90 percent of the participants pulled a switch to reroute the boxcar, suggesting people are willing to violate a moral rule if it means minimizing harm.

I'd hate to face this choice in real life. I think if I didn't know any of them I'd kill one to save five. Imagine you would face no legal repercussions from either choice. What would you do? Suppose you did not know any of the people involved.

Those who didn't pull the switch were more emotionally aroused.

Of the 147 participants, 133 (or 90.5 percent) pulled the switch to divert the boxcar, resulting in the death of the one hiker. Fourteen participants allowed the boxcar to kill the five hikers (11 participants did not pull the switch, while three pulled the switch but then returned it to its original position).

The findings are consistent with past research that was not virtual-based, Navarrete said.

The study also found that participants who did not pull the switch were more emotionally aroused. The reasons for this are unknown, although it may be because people freeze up during highly anxious moments Ė akin to a solider failing to fire his weapon in battle, Navarrete said.

I'd like to see a larger study done that controls for sex, age, ethnicity, citizenship of different nations, level of education, type of education, personality type, and varying degrees of autism. What factors have impact on what choices people make?

Also, I'd love to see this controlled for who is on the two train tracks. Would someone let five unknown die to save their wife or mother? Their brother? Their best friend? A powerful or rich figure? A beautiful woman? A small child?

Share |      Randall Parker, 2011 December 01 10:08 PM  Brain Ethics Law

PacRim Jim said at December 2, 2011 12:09 AM:

In wartime, most people would sacrifice entire countries to save one inhabitant of their own country.

Sergey Kurdakov said at December 2, 2011 2:15 AM:

the problem is usually discussed by axiomatic libertarians when they try to 'calculate' moral outcomes ( just for the note - there are many different ethical approaches, such as virtue ethics, which does not require calculations and lack ( as new research shows some problems) which can not be resolved by 'calculation' ethics .

Let us face this - there is no such problem for most people. maybe one once in a century have to make such a decision. And then - there are many other factors which influence decision. Then, it of cause, would be fine to know who are people in reality ( see Stanford prison experiment - it turns that anyone can kill or make suffer other people without any problems ).

For the is more important question is why so many people do think that their calculation ethics has anything with real world and should be considered to have any implications. Even being explained several times about ethics problems and different approaches, they would return to their calculations. And then the problem, could these people be considered as humans? If the only thing they could - to calculate non existant problem? Then, after some consideration the answer might be following - if the person to 'sacrifice' is one who only can think of such approach, regardless of other possibilities ( use of wood, for example or stand which is near, instead of 'human' ), then those people should be used to save other people with no questions - because, obviously, if one calculates - then one can calculate, that one life is less worth, than 5 ones - this is just a mere outcome of 'calculations' approach itself and then, who agreed to calculate in a first place - should assume the consequences of it's own will.

Bernardz said at December 2, 2011 2:29 AM:

Another question on a similar line, is would you kill one person for their organs to save five people that need those organs.

solaris said at December 2, 2011 8:32 AM:

I have my doubts about the usefulness of this research. In practice the only way to find out what you'd do in such a life-and-death situation is to actually BE in that situation.

Brett Bellmore said at December 2, 2011 3:20 PM:

I'd reason that finding myself in such a situation in real life was so wildly unlikely that I was probably suffering a psychotic episode, and would go quietly wait in the corner for the nice men in white jackets to medicate me out of it, lest I harm somebody.

LAG said at December 2, 2011 6:40 PM:

These aren't interesting choices. One for five? Sure, almost as easy as deciding to kill one for six million (Hitler for the Jews). What's interesting is one for one. Would you kill this one to save that one? Or two for one? Would you kill a carload to save the King, the President, or the Pope? The first two might feel you should; not so sure about the Pope, but I doubt it. How about an infant for an adult? What if the infant is only minutes old? Of minutes from being born? How many infants? What about a young woman for an important old man? Ridiculous? I guess it depends on how much you value Ted Kennedy's legacy. Would you balance your political party with dead from the other?

Try to come up with more interesting questions, please.

Everyday said at December 2, 2011 7:50 PM:

LAG: How about a mother versus her minutes old daughter?

Food said at December 2, 2011 11:23 PM:

Most people are utilitarians, as is evidenced by economics. Putting aside private crisis choice that hits personal psychological issues, planning for self-interest and emotional involvement with one of the potential victims, people seem to be mostly utilitarians. I think this has been obvious since the French Revolution, at least. It is also operative in why our monetary policy operates the way it does, for instance.

Brett Bellmore said at December 3, 2011 5:05 AM:

No, most people are not utilitarians. Most people try to apply balancing tests, but utilitarianism is more than applying balancing tests. There's the universal aspect, (You have to take into account the welfare of people on the other side of the globe.) the calculation aspect, (What's the ratio between orgasms and a good meal? You need to know!) interpersonal balancing across different benefits... (How many people have to laugh at it, before kicking somebody in the shin is mandatory?)

Even utilitarians aren't utilitarians. Seen many who live on ramen so they can spend the excess money feeding poor children in Africa? Utilitarianism demands things like that.

Utilitarianism is just a metaphor run amok. It's not a moral theory people can actually apply.

Detroit said at December 3, 2011 5:13 AM:

...very weak "research" at Michigan State University -- a few non-randomly selected 'participants' playing a simple video game... with many huge variables ignored in that soft 'research' scenario.

Perhaps amusing, but certainly not objective nor even remotely conclusive about general East Lansing, Michigan, American or human behavior norms.

Tj Green said at December 3, 2011 7:33 AM:

Most of us act on instinct so saving the five would be most peoples reaction. Those that chose not to save the five I would blame on their religious views. Instinct is a better survival strategy for a species than religion. The problem is if you do not see yourself as part of a species.

spindizzy said at December 3, 2011 10:19 AM:

Again, I have to rise to the bait of Brett's anti-utilitarian sophistry!

> There's the universal aspect

If you want to prioritise the needs of local people over those further away then you can propose your own utility function that includes an appropriate weighting.

> What's the ratio between orgasms and a good meal? You need to know!

I don't need to know that anymore than I need a diploma in physics to catch a ball. I can just approximate and adjust, like I do with any hard problem.

> How many people have to laugh at it, before kicking somebody in the shin is mandatory?

From the point of view of a happiness utilitarian, I suppose it would have to be enough to outweigh not only the immediate pain of the individual but also the insidious social effects of such acts being condoned.

Incidentally, YouTube built a large audience with precisely this kind of comedy.

> Seen many who live on ramen so they can spend the excess money feeding poor children in Africa? Utilitarianism demands things like that.

Well, not everyone's utility function will emphasise the welfare of African children. Of those which do, few practitioners would consider living on ramen to be the most efficient route to their goal.

More importantly, your point seems to be that utilitarianism demands action from its adherents. I consider that a strength and not a weakness.

> Even utilitarians aren't utilitarians.

I'm sure that the pope sometimes takes the last biscuit. That doesn't mean he isn't Catholic.

Randall Parker said at December 3, 2011 10:08 PM:

Sergey Kurdakov,

Taking a look at how people make ethical decisions in rare situations with extreme outcomes provides insight into what people decide in other situations. By making the stakes high and simple specific factors can be studied by themselves. Then other experiments can be done with family ties and other considerations to measure their weightings.


People make utilitarian moral choices all the time when, say, they argue for taxing everyone to support research that benefits the majority. I see lots of situations where people apply utilitarian reasoning to support public goods.

I don't buy that utilitarianism demands that people send money to Africa. Utilitarianism ends up being for some goal. Someone in theory could have a utility function best satisfied by raising the average happiness of everyone in the world. Or to minimize deaths globally. Though in reality people have really complex utility functions.

Jessica said at December 4, 2011 7:18 AM:

It seems to me five adult strangers versus one adult stranger is more a matter of math than an ethical dilemma. If you made the one person on the track into a small child or a loved one, it would be much more of a dilemma.

LAG said at December 4, 2011 12:50 PM:

Everyday, what's your criteria? In some settings, the death of the mother means the death of the infant, so killing her to save the infant is fruitless and misguided. Also, if you're half a married couple, say, in a subsistence economy, then killing the mother may condemn the partner to poverty and result in the death of the infant as well. In those circumstances, the infant may be sacrificed with the knowledge that more children can be produced later.

Is the mother ill, to old to bear more children, or is this simply a trade without measures attached? Personally, in a utilitarian world I'm not sure you ever trade an adult parent for a child too young to contribute to the family.

Post a comment
Name (not anon or anonymous):
Email Address:
Remember info?

Go Read More Posts On FuturePundit
Site Traffic Info
The contents of this site are copyright ©