Ethical Theory Spring 2019

The Repugnant Conclusion


Parfit’s essay raises an important question about how to determine what counts as the best consequences of an action or policy. Specifically, is the best world the one with the most total happiness in it or is it the one with the highest average happiness?

The orthodox utilitarian position is to choose the greatest overall good. Parfit argues that this leads to what he calls the Repugnant Conclusion. It appears that maximizing average utility would be the solution. But Parfit denies that this will work. What he calls the “mere addition paradox” is meant to show that it is better to lower average utility for the sake of increasing the total amount.

That kind of reasoning, of course, leads right back to the Repugnant Conclusion. In the end, Parfit suggests a way of avoiding the Repugnant Conclusion that involves departing from utilitarianism.

The Mere Addition Paradox

I think the reasoning that leads to the Repugnant Conclusion is clear enough. If you assume that adding more people to the world improves the world, even if it lowers average happiness, you seem to be committed to move from A all the way down to Z.

The reasoning behind the mere addition paradox, by contrast, is not so easy to see. Here the idea is that you add people to the world whose happiness is lower than the average happiness of everyone else. Would anyone say that the world is worse if it moves from A to A+? After all, the only difference is that there are more people in A+ than there were in A and, while their happiness is lower than the average happiness in A, they are all satisfied with life.

If you say “I agree that A+ is at least as good as A,” then Parfit is off to the races. Divided B is better than A+ so Divided B can’t be worse than A. Divided B is the same as plain B. So plain B isn’t worse than A.

So what. Well, we agreed we could stop the slide to the Repugnant Conclusion only if we agreed that plain B is worse than A (and C is worse than B, D is worse than C … Z is worse than Y, so Z is definitely worse than A). And now we have been led by solid-seeming chain of reasoning to the conclusion that plain B is not worse than A. Oops.

Other Problems with Average Utilitarianism

Emma saw the problems with maximizing average happiness right from the start. One way of raising average happiness is to get rid of people with below average happiness. This could be people with disabilities that make life difficult or people with hyper-abilities who are perpetually dissatisfied, such as Mill’s Socrates.

Zane noted that another problem might be raised by so-called utility monsters. These are people who derive extraordinary enjoyment out of things that lower happiness for others. For instance, they might derive intense satisfaction out of really expensive things or they might be bitterly disappointed without really expensive things. Either way, maximizing overall happiness might require taking a lot of resources from others so that the monsters are happy. Parfit dismissed this as unrealistic. In addition, it isn’t obvious to me that it would be a problem unique to maximizing average happiness. Satisfying the monsters raises both average and total happiness. It probably lowers the happiness of everyone else, though.

Perfectionism as an Answer

Parfit suggests that one way to stop the slide to the Repugnant Conclusion is to adopt a different understanding of what is good and bad. Utilitarianism typically holds that good and bad are measured in terms of happiness and unhappiness (or pleasure and pain). The advantage of that is that it isn’t judgmental. Whatever you like counts. However that generates the problem: everyone will like something and so each possible new life counts for a lot.

Parfit proposes instead that we think of some aspects of human culture as especially valuable: art, say (Parfit 2004, 18–20). If the world is better for having more art in it, then we could have a natural stopping point: the population should not grow so large that there would be no resources for art.

The idea is that we should think that the world is getting better if humanity is being made better or “perfected” (hence “perfectionism”) and that having art is part of becoming better.

One obvious problem with this is that it means leaving behind one of the attractive features of utilitarianism, namely, its non-judgmental, non-elitist theory of what makes things good or bad. But nothing comes for free and maybe it’s worth paying that price.

Parfit’s Examples and Reality

We got hung up on some of Parfit’s examples. Ariel objected to what she saw as the assumption that increasing population always decreases the standard of living. It’s generally the case the the opposite is true. Cassy (I think) didn’t like his use of geographical labels to illustrate “A+” and “Divided B” in his explanation of the Mere Addition Paradox. Where does he get off saying that people in the Americas are (or were) less happy than people in Europe?

Both points bring up a problem with this kind of moral philosophy. On the one hand, authors like Parfit would want to insist that their examples are purely hypothetical. He doesn’t mean that rising population is always bad for living standards, he’s just asking you to imagine what a utilitarian would do if it were. And he’s not saying that people in America are, or were, less happy than people in Europe. He’s just asking you to imagine two populations that don’t interact with one another and he only used the names of continents to illustrate how it is possible that there could be populations isolated from one another like that.

That said, it’s quite reasonable to get caught up on details like this for two reasons. First, you don’t want to go around repeating misleading stories about how the world works. Even if the stories are only hypothetical, you want to be clear on what is real and what isn’t. Second, it’s fair to ask why we should care about a moral philosophy that only applies to completely fictional settings.

Both points are well taken. In response, I think I should say a little bit about how this essay might be applied to the real world.

We know that real states have made decisions about these questions. The government of China, for instance, decided that it would rather improve the average welfare of its citizens rather than the total welfare when it implemented the so-called “one child” policy. I’m not saying they were right or that they were wrong; I’m just saying that they faced exactly the kind of decision that Parfit describes.

From what Prof. Brown said in the PPE seminar this fall, I gather that those working in development NGOs frequently have a similar outlook. They prefer economic growth that will enhance the quality of life over public health programs that would save lots of infants and thereby create a larger population sharing a smaller economic pie. They are implicitly favoring average happiness over total happiness.

Finally, of course, there is the looming threat of climate change. If we add more people to the world’s population, they are going to produce more greenhouse gasses. At some point, we think that this will threaten a massive collapse in the world’s standard of living. There will still be a lot of people, but many of them will be living in environmentally harsh conditions. Even though average happiness will be lower in such a world than it would have been with a smaller population and less climate change, it could still be the case that total happiness will be higher.

Key Points

These are the main things you should know from today’s class.

  1. The difference between maximizing total happiness and maximizing average happiness.
  2. The reasoning that leads to the Repugnant Conclusion.
  3. Why the Mere Addition Paradox also leads to the Repugnant Conclusion.


Parfit, Derek. 2004. “Overpopulation and the Quality of Life.” In The Repugnant Conclusion: Essays on Population Ethics, edited by Jesper Ryberg and Torbjorn Tännsjö, 7–22. Dordrecht: Kluwer Academic Publishers.


There was a handout for this class: 10.ParfitRepugnant.handout.pdf