Scott Alexander’s review of What We Owe The Future seems to imply (as I read it) that the standard person-affecting-view solves the Repugnant Conclusion. This is true for the standard way the Repugnant Conclusion view is told, but a simple modification makes this no longer the case. I’ve heard this point made a lot, though, so I figure I’d make a short post addressing it.
The standard Repugnant Conclusion, told to discredit total utilitarianism, is as follows. You have some population, each starting out at a reasonable net happiness:
Then you add a bunch of people with much lower but still positive happiness, which under total utilitarianism is a net benefit:
Then you impose equality while adding a little bit of happiness to each person to make it a net benefit under utilitarianism:
And here you end up with a society preferable under utilitarianism relative to the first but maybe intuitively worse. This can also be repeated infinitely until you have people just above the “net zero happiness” threshold.
The person-affecting view normally told to avoid this problem is a view that is neutral with respect to creating or removing future happy people (or even, as Scott writes, slightly negative about creating), but is in favor of making existing people happier. So the first trade would not be taken.
But a simple modification shows that this doesn’t actually solve anything. We just have to make the first trade add a small amount of happiness. Or, if you accept Scott’s “slightly negative” approach, an amount of happiness slightly more than (# of new lives * negative margin). So this first world (axes modified from above):
Becomes this second world:
Which becomes this third world:
If you accept the “slightly negative” argument, say margin=0.1, this prevents you from extending this to preferring a world with many people at 0.001 happiness and instead limits you to 0.101. I disagree with this threshold argument (which has been made before more explicitly), but regardless it doesn’t actually get around the meat of the Repugnant Conclusion.
Scott has some more ethical commentary that suggests he might not care if individually ethical decisions lead to cumulatively terrible results. Regardless, I just wanted to address this one point. The standard person-affecting view does not escape the Repugnant Conclusion.
The standard person-affecting view doesn’t solve the Repugnant Conclusion.
Scott Alexander’s review of What We Owe The Future seems to imply (as I read it) that the standard person-affecting-view solves the Repugnant Conclusion. This is true for the standard way the Repugnant Conclusion view is told, but a simple modification makes this no longer the case. I’ve heard this point made a lot, though, so I figure I’d make a short post addressing it.
The standard Repugnant Conclusion, told to discredit total utilitarianism, is as follows. You have some population, each starting out at a reasonable net happiness:
Then you add a bunch of people with much lower but still positive happiness, which under total utilitarianism is a net benefit:
Then you impose equality while adding a little bit of happiness to each person to make it a net benefit under utilitarianism:
And here you end up with a society preferable under utilitarianism relative to the first but maybe intuitively worse. This can also be repeated infinitely until you have people just above the “net zero happiness” threshold.
The person-affecting view normally told to avoid this problem is a view that is neutral with respect to creating or removing future happy people (or even, as Scott writes, slightly negative about creating), but is in favor of making existing people happier. So the first trade would not be taken.
But a simple modification shows that this doesn’t actually solve anything. We just have to make the first trade add a small amount of happiness. Or, if you accept Scott’s “slightly negative” approach, an amount of happiness slightly more than (# of new lives * negative margin). So this first world (axes modified from above):
Becomes this second world:
Which becomes this third world:
If you accept the “slightly negative” argument, say margin=0.1, this prevents you from extending this to preferring a world with many people at 0.001 happiness and instead limits you to 0.101. I disagree with this threshold argument (which has been made before more explicitly), but regardless it doesn’t actually get around the meat of the Repugnant Conclusion.
Scott has some more ethical commentary that suggests he might not care if individually ethical decisions lead to cumulatively terrible results. Regardless, I just wanted to address this one point. The standard person-affecting view does not escape the Repugnant Conclusion.