> I can’t imagine being two people easily, let alone a galaxy full.

Approximate being two people as living twice as long.

Why intuitions on this seem likely to be bad evidence:

1) Human moral intuitions generally don’t react a lot for people who can’t return influence (e.g. distant people, other groups, people likely to keep suffering, strangers, sufferers of omissions rather than actions [hard to apportion blame], omissions in crowds). Those who don’t exist yet are the extreme of this. So no reason our feelings should be attuned to their interests.

2) It is easier to imagine the difference between happier and sadder than between one person being happy and two from the inside. I can’t imagine being two people easily, let alone a galaxy full. So I would expect level of happiness to be accounted for in feelings much more than number living.

I said that a world of a few happy people is better than a world of many unhappy people.

Robin said that the many unhappy people would still prefer their world to the world of a few happy people, because they don’t exist in the latter world. And I gather he also thinks that we in this world should prefer the unhappy world to the happy world because there are more people (or people-who-want-to-live) in it. And finally this is presented as a reason to be enthusiastic about highly populated but possibly unhappy futures in this world.

In response I first drew the distinction between preferences that can make a difference, and preferences which are purely of abstract significance – which in being voiced express an aesthetic or ethical value, but which do not correspond to a choice of outcomes. When the people of the unhappy world are asked to express a preference between the unhappy world and the happy world, it is a purely abstract preference. They are already in the unhappy world and nothing can undo that. So it is at the very least a contingent matter as to whether they will say that their world is better than the abstract alternative.

The next issue concerns abstract preferences as expressed by us in this world. Robin asks us to judge one world as better than another basically by counting up the number of people in it who want to live (I am guessing that this is the criterion of “lives worth living”). I don’t know if he would amend this principle in additive-utilitarian fashion by making quality of life a matter of degree; in that case a few very happy people might trump a slightly larger number of miserable people. But we seem to be coming close to the view that a galaxy full of people who are miserable, but not so miserable as to commit suicide, is better than a single planet of extremely happy people. I believe this is the “repugnant conclusion” and Robin says above he does not find it repugnant, whereas I regard it as a reductio ad absurdum of naive additive utilitarianism.

The final issue concerns the consequential preferences we express in this world regarding our own possible futures. Should we regard a future for this world, our world, that is miserable but populous, as better than one which is happy but sparsely populated? Robin apparently says yes, I say no.

I hope that this addresses your second batch of questions by providing the context to what I wrote. As to whether there’s a fallacy in Robin’s argument or not – the belief there’s a fallacy comes from the sense that the proposition is absurd, and the inability to pin down the specific fallacy comes from not knowing the specific perspective from which it nonetheless appears reasonable to Robin.

“Katja – You’re trying to compare utility for an existing entity (U/1) over utility for a non-existing entity (U*/0). Divide by zero might not be the best term for it, but any attempt to compare an existent person to a non-existent person (as you seem to be asking Tony to do) runs into an undefined quantity problem.”

So the same problem I run into when I wonder whether if I had another jar I could fit more jelly beans into it – I don’t have the jar yet, so the calculation would be (extra beans)/(number of extra jars I have now) = 100/0 = oh no!
Is there a pertinent difference I’m missing, or do I need to give up making judgments about future quantities all together?

“This is one of those cases of Alice-in-Wonderland reasoning where one feels that an enormous fallacy is being committed, but the very alienness of what is being asserted makes it hard to guess what assumptions make it seem plausible to the other person, which in turn makes it a little difficult to engage with.”

If you can’t pin down a fallacy, maybe it’s the lack of one that makes somebody else find it plausible? If it’s just your feeling of alienness about the assertion, why do you think your feeling on that is likely to be more reliable than others’ feelings plus their reasoning?

In the second part of your next paragraph you say you can prefer worlds you aren’t in. You do so because of the people in the worlds who will enjoy them presumably. Why can’t I do the same for future possible worlds containing people who don’t exist yet? Why does it matter that once they do exist past possibilities of them not are irrelevant? Their judgement on the matter isn’t important to whether we should prefer such worlds.

Well, that depends. How pleasant is life behind the veil of ignorance? Is real existence better or worse?

The Fehig essay is called “A Pareto Principle for Possible People.” I haven’t read it, but it’s discussed by Benatar [author of "Better To Have Never Been"]. It’s weird as intuition pumps go (there’s a pill creates a desire that is either met or denied and the asymmetry is thus addressed without the existential predicate that leads people to squawk about non-identity).

Do you folks think it is morally preferable to be given a desire which is then satisfied relative to doing nothing or just neutral? Would you carry over the logic to creating a new life you expect to possess positive utility, or do you think there is an important distinction?