Recently, I was presented with the problem of whether I would rather kill a 30 year old man in a third-world country, or a 30 year old man in a first world country.

I said that the three factors which would influence such a decision were:

  • Forecasted years of life remaining
  • Amount of happiness which the person would experience per year, on average
  • Impact on total human happiness by that person

I then reduced it to the equation, where the first-world country person = x and the third world country person = y,

score(x) = (life expectancy in a first world country - 30)*f(mean happiness of person x per year)*g(impact on total human happiness of person x)

score(y) = (life expectancy in a third world country - 30)*f(mean happiness of person y per year)*g(impact on total human happiness of person y)

Where f() and g() are undetermined functions. And then I said

If score(x) > score(y), then kill person y

If score(y) > score(x), then kill person x

Else, random_number = random integer between 0 and 1.

If random_number == 1, then kill person x

Else, kill person y

Is this logic sound, assuming happiness is the most significant metric? Are there any other factors which should have been included? Should total happiness over time have been multiplied with or added to impact on total human happiness?

  • Pretty much every attempt to quantify philosophy fails because of subjective definitions/values of things like "happiness". Also, there's really no such thing as "Forecasted years of life remaining" because you can't accurately determine an individual's life expectancy (it depends on how group them vs other people). You can fix that by assuming you have an oracle that tells you life expectancy and happiness caused. However, if you accept that human beings are selfish, you should use a metric that maximizes YOUR happiness, not the world's happiness. Jun 26 at 14:19
  • The maths are at the same time too simple and overly detailed. Just assume each person has a predicted contribution to global happiness for staying alive this time.
    – tkruse
    Jun 27 at 0:25

2 Answers 2


Apart from the question how ethical it is for a third party to decide who gets to live and who's to die. Or how you'd measure subjective concepts such as "happiness" or for that matter life expectancy.

So I'd assume that you measure them indirectly via statistics, which creates the problem that you contribute to that statistics if you propose that method.

So say you look at the stuff people have and determine the happiness from that and the average life expectancy and come up with the conclusion that 1st world countries have more stuff and live longer and thus decide to kill the person in the 3rd world. Then you'd even further reduce the life expectancy in the 3rd world.

And people in your vicinity being killed might also negatively contribute to your happiness. So you might end up creating a self-fulfilling prophecy. Maybe not for one case but if you've already formulated that it kinda implies that you'd generalize it and do it more often.


This is just a variant of the trolley problem. None of the differences provided are legitimate for making a triage decision on whose life to save.

Any decision to prefer people from richer countries to people from poorer countries is obviously unethical so don't even start designing such formulas. If you do, it means you do not understand ethics.

Consider the impact of having to make this decision 10 million times between comparable US citizens and people from Mexico. Do you think the population of any of those countries would consider the maths as fair if it meant most kill decisions would target one countries citizens over another?

  • Yes. How else would you make a descicion? Jun 27 at 2:50

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.