Appendices - Killing Happy Animals: Explorations in Utilitarian Ethics - page 153

aggregation will not help, according to Friedman, because the average
view has very implausible implications as well. A combination of the
Total View and the Average View on aggregation might be able to avoid
the counter-intuitive implications of both views, and thus, according
to Friedman, this would be the most plausible utilitarian view in differ-
ent-number choices. However, combining two different methods of
aggregation means that one accepts two different standards or factors
for evaluating outcomes. Therefore, this version of utilitarianism would
have intransitive implications. Thus, Friedman concludes that any mini-
mally plausible version of utilitarianism cannot avoid intransitivity.
Prior Existence Utilitarianism, however, can avoid intransitivity. After
all, as we have seen, Prior Existence Utilitarianism avoids the Repugnant
Conclusion. It allows us to stick to the Total View on aggregation and
thereby to avoid the counter-intuitive implications of the Average View.
It focuses solely on the total amount of welfare, and thus does not
combine several methods of aggregation. Prior Existence Utilitarianism
can avoid intransitivity at the level of what one is morally required to
do, i.e. the deontic level. This is, arguably, the level at which intran-
sitivity would be most disturbing. After all, intransitivity in the sense
that whatever one does, there is a better option, implies that there is no
rational basis for choice. This would probably lead to scepticism about
practical reasoning. Even if intransitivity remains at the axiological
level, avoiding it at the deontic level seems to be relevant. So, if transi-
tivity is indeed a reasonable requirement of a moral theory, being able
to avoid intransitivity at least on the deontic level is a big advantage of
Prior Existence Utilitarianism.
A major concern that has been brought forward against the person-
affecting restriction, and therefore also against prior existence utilitari-
anism is that it allegedly violates transitivity (Arrhenius 2009; Broome
2004, p. 146; Temkin 1996, p. 209; Bradley 2013, p. 41). For instance,
Arrhenius (2009) presents the following choice between three outcomes,
A, B and C, each consisting of two populations, which both have a posi-
tive but different welfare level:

A B C

x = 10 y = 10 z = 10

y = 5 z = 5 x = 5

Arrhenius (2009) explains:

Since the x-people don't exist in B, B is neither worse nor better than
A for them. Similarly, since the z-people don't exist in A, A is neither worse