In one of my early articles, I presented a supposed proof of utilitarianism. This proof, I think, does not work, but the considerations should cause us to be more favorably disposed to utilitarianism. Here’s what I said about the argument, before defending each premise.
Suppose you accept
1) Hedonism, which says that the happiness that one experiences during their life determines how well their life goes for them.
2) Anti egalitarianism, which says that the distribution of happiness is irrelevant.
3) Pareto Optimality, which says that we should take actions that are better for some, but worse for none, regardless of consent.
These axioms are sufficient to derive utilitarianism. The action that maximizes happiness could be made Pareto Optimal by redistributing the gains. Anti egalitarianism says that redistributing the gains has no morally significant effect on the situation. If Pareto improvements should be taken, and the utilitarian action is morally indistinct from Pareto improvements, utilitarian actions should be taken.
This can be illustrated with the trolley problem as an example. In the trolley problem it would be possible to make flipping the switch Pareto optimal by redistributing the gains. If all of the people on the track gave half of the happiness they’d experience over the course of their life to the person on the other side of the track, flipping the switch would be pareto optimal. In that case, everyone would be better off. The person on the other side of the track would have 2.5 times the good experience that they would otherwise have had, and the other people would all have .5 a life's worth of good experience more than they would have otherwise had. Thus, if all the axioms are defensible, we must be utilitarians.
There were a few mistakes. One initial one was that I mischaracterized the view — I should have called it non-egalitarianism, not anti-egalitarianism; after all, it holds that distribution of utility is neutral, not that it is bad. This, however, is not where the critical flaw resides.
The problem is that utilitarianism doesn’t naturally follow from these three. Instead, a more limited claim follows from the three — that claim is that the utilitarian action makes the world just as good as an action you should take. After all, a world where you flip the switch in the trolley problem and utility is distributed to make things a Pareto improvement is just as good as a world where you flip the switch ordinarily — and the world where you flip the switch and bringing about a pareto improvement is one in which you do the thing that you should do. However, the deontologist can accept this and deny utilitarianism — even if the world is just as good, the action may still be wrong.
Here, the deontologist could say that, even if actions B and C both make a better world than A, you still shouldn’t take B. B might be wrong, even if it doesn’t make the world worse. The fact that some action makes the world as good as a right action doesn’t mean it itself is a right action — they are, after all, not consequentialists.
But nonetheless, I think that this should lead us in the direction of utilitarianism. Pushing the guy off the bridge to save five seems less wrong when you realize that if we just had a way of distributing utility, then it would be right to do, and would make everyone better off. If you combine this with the insight that redistributing utility wouldn’t make anything worse, then it seems clear that you should flip the switch.
Thus, for the utilitarian proof to be successful, it requires the following additional principle: if some action makes things just as good as a right action, then it is a right action. However, this could be denied, though it sounds plausible. Additionally, even if we reject this principle, pondering the fact that a non-morally relevant change would make killing one to save two a Pareto improvement should move us in the direction of utilitarianism.