The fundamental disagreement here is about whether something can meaningfully be good without solving any preexisting problem. At least, it must be good in a much weaker sense than something that does solve a problem.

Thank you for your comment!

The version of negative utilitarianism I have in mind is not one that ignores net-positive lives, but one that denies their existence in principle, so the $u_i$ live in $R_{-}$.

It's easier to see in a preference framework: for a fixed set of preferences, maximising preference satisfaction is exactly the same as minimising preference frustration. But as soon as we can change the set of preferences, those two approaches are radically different. Those forms of NU are about minimising the amount of frustrated preferences, which can only be positive in principle. Or, equivalently, about maximising the opposite of that, which is always negative. This satisfies aggregation and Pareto axioms.

Thanks!

I was going to answer the same as here: the form of the "very repugnant" conclusion applying to NU seems much less repugnant.

The paper is available to download for free on the Springer website, at least for me and I'm not logged in or in a university network.

I haven't read yet. I'm curious because it sounds surprising that negative utilitarianism doesn't avoid a repugnant conclusion in some form.


Edit: I was somehow logged in.  I guess you can PM me for a copy. I'll read it tomorrow.