By iterating the same beneficence-and-equality reasoning down a series New A, New B, New C, …, New Z… — Derek Parfit

By iterating the same beneficence-and-equality reasoning down a series New A, New B, New C, …, New Z, we are led—via the transitivity of 'better than'—to the New Repugnant Conclusion that an enormous population with lives only slightly above the Bad Level is better than a world like New A (and hence better or not worse than A+).

By Derek Parfit, from Les raisons et les personnes

Key Arguments

Parfit introduces New C and lower outcomes in the 'New Alphabet': 'Now compare New B with New C. Once again, better-off groups would be worse off. But these groups would be worse off by an amount far smaller than the amount by which as many worse-off groups would be better off.'
He asserts, 'By the same reasoning, New C is better than New B. Such reasoning carries us to New Z.'
New Z is described as 'an outcome there is some enormous population whose lives are not much above the Bad Level.'
Transitivity then forces a ranking: 'New Z must be better than New A, since every step down the New Alphabet has been judged a change for the better, and better than is transitive.'
Recalling that 'as I argued, New A is better than A +', Parfit notes that 'Taken together, these claims imply' the New Repugnant Conclusion, which he explicitly formulates.
He allows that some may hold only that New A is 'not better than' but 'not worse than' A+, in which case they are committed to a weakened version: 'They must claim that, of A + and New Z, the second would not be worse.'
He emphasizes that, although this New Repugnant Conclusion is 'in one respect less repugnant than the Repugnant Conclusion' (because lives in New Z are somewhat above the Bad Level rather than barely worth living), he still finds it 'very repugnant.'

Source Quotes

If there was a change from New A to New B, worse-off groups would gain very much more than better-off groups would lose. Now compare New B with New C. Once again, better-off groups would be worse off. But these groups would be worse off by an amount far smaller than the amount by which as many worse-off groups would be better off. By the same reasoning, New C is better than New B.

But these groups would be worse off by an amount far smaller than the amount by which as many worse-off groups would be better off. By the same reasoning, New C is better than New B. Such reasoning carries us to New Z. In this outcome there is some enormous population whose lives are not much above the Bad Level.

Such reasoning carries us to New Z. In this outcome there is some enormous population whose lives are not much above the Bad Level. New Z must be better than New A, since every step down the New Alphabet has been judged a change for the better, and better than is transitive.

In this outcome there is some enormous population whose lives are not much above the Bad Level. New Z must be better than New A, since every step down the New Alphabet has been judged a change for the better, and better than is transitive. Remember next that, as I argued, New A is better than A +.

New Z must be better than New A, since every step down the New Alphabet has been judged a change for the better, and better than is transitive. Remember next that, as I argued, New A is better than A +. Taken together, these claims imply The New Repugnant Conclusion: In the first of two possible outcomes, there would be two groups of ten billion people. One group would have a quality of life far higher than that of any actual life that has been lived.

Remember next that, as I argued, New A is better than A +. Taken together, these claims imply The New Repugnant Conclusion: In the first of two possible outcomes, there would be two groups of ten billion people. One group would have a quality of life far higher than that of any actual life that has been lived.

Of these two outcomes, the second would be better. Some people may believe that New A is not better than A +, but is merely not worse. These people must accept a weakened version of this new conclusion. They must claim that, of A + and New Z, the second would not be worse. This new conclusion is in one respect less repugnant than the Repugnant Conclusion.

They must claim that, of A + and New Z, the second would not be worse. This new conclusion is in one respect less repugnant than the Repugnant Conclusion. In Z people’s lives were barely worth living. In New Z, people’s lives are somewhat better. But this new conclusion seems to me very repugnant. Lives that are not much above the Bad Level cannot be well worth living, or close to being well worth living.

Key Concepts

Now compare New B with New C. Once again, better-off groups would be worse off. But these groups would be worse off by an amount far smaller than the amount by which as many worse-off groups would be better off.
By the same reasoning, New C is better than New B. Such reasoning carries us to New Z.
In this outcome there is some enormous population whose lives are not much above the Bad Level.
New Z must be better than New A, since every step down the New Alphabet has been judged a change for the better, and better than is transitive.
Remember next that, as I argued, New A is better than A +. Taken together, these claims imply
The New Repugnant Conclusion:
Some people may believe that New A is not better than A +, but is merely not worse. These people must accept a weakened version of this new conclusion. They must claim that, of A + and New Z, the second would not be worse.
This new conclusion is in one respect less repugnant than the Repugnant Conclusion. In Z people’s lives were barely worth living. In New Z, people’s lives are somewhat better. But this new conclusion seems to me very repugnant.

Context

Central part of Section 148, where Parfit constructs the 'New Alphabet' of outcomes and shows that accepting each local improvement, together with transitivity, yields a new, repugnant-sounding evaluative conclusion about very large, low‑quality populations.