Brains, Persons, and Society *** ABSTRACTS    Cervelli, Persone e Società ***ABSTRACTS

David McCarthy
Dept. of Philosophy, University of Edinburgh

Utilitarianism, prioritarianism and fairness

The priority view, roughly characterized by Parfit as the claim that we should give greater

priority to the worse off, is perhaps the most popular view about distribution in moral

philosophy. But Broome (1991) has argued for the surprising conclusion that the priority view is

more or less meaningless.

Elsewhere I have begun a response to Broome (McCarthy, forthcoming a, b). I have suggested

that a version of the priority view, which I call ex ante prioritarianism, avoids the problems

Broome raises. But it may run into another problem.

The main way in which ex ante prioritarianism departs from utilitarianism is in its view,

following an example due to Diamond (1967), that lotteries are sometimes a good thing. More

precisely, the claim is that when two contestants to a good (e.g. a heart for transplant) are

perfectly symmetrically placed, it is better to hold a fair lottery than to give the good to one of

them directly.

However, for reasons unconnected with his discussion of the priority view, Broome claims that

utilitarianism can accept this claim. Therefore, Broome’s account of utilitarianism threatens to

coapt what I have claimed is the principled prioritarian departure from utilitarianism.

But this claim is not easy for Broome to defend. His account of utilitarianism rests on a version

of a theorem originally due to Harsanyi (1955). One of the premises of this theorem is that the

betterness relation satisfies the central axiom of expected utility, known by such names as the

sure thing principle, or the strong independence axiom. And prima facie, it looks as if this

premise is inconsistent with the claim that lotteries are sometimes a good thing.

Broome tries to avoid this inconsistency via an instance of a well known strategy for avoiding

alleged counterexamples to the sure thing principle, namely a further individuation of outcomes.

To do this, Broome is forced to claim that being treated unfairly is what he calls an individual

harm. So to prevent its distinctive claim being coapted, a defender of ex ante prioritarianism has

to deny that unfairness is an individual harm. This issue is the focus of the paper, and I will argue

that unfairness is not an individual harm. Hence ex ante prioritarianism is a viable position.

I will give several arguments for this claim, but I only have the space to discuss one here. It is

well known that the strategy for avoiding counterexamples to the sure thing principle which

Broome uses raises a worry about whether the sure thing principle is being trivialized. For the

sure thing principle to be nontrivial, the domain of possible lotteries needs to be rich in a certain

technical sense.

In Broome’s work, the technical sense is specified by what he calls the rectangular field

assumption, and he points out that the rectangular field assumption is threatened by the claim

that unfairness is an individual harm. However, this discussion needs to be superseded because it

has recently been shown that the rectangular assumption can be abandoned without trivializing

the effect of the sure thing principle. What is needed, in the context of Harsanyi’s theorem, is

what is known as a convexity assumption which is strictly weaker than the rectangular field

assumption.

However, I will argue that this is of no help to Broome. The claim that unfairness is an individual

harm still violates the convexity assumption, unless impossible lotteries are allowed into the

domain of the betterness relation.

That is possible, though hugely controversial. But it leads to a further problem. Broome takes it

as incumbent to avoid what is known as a levelling down objection. In other words, he accepts

that an adequate account of fairness cannot result in a situation where it is a good thing to make

everyone worse off in terms which ignore the alleged harm of unfairness merely to improve

overall fairness. But once the needed impossible lotteries are admitted into the domain of the

betterness relation, it turns out to be impossible to avoid the levelling down objection.

I will claim that this result greatly reduces the motivation for the claim that unfairness is an

individual harm. When combined with other arguments I will present, I conclude that we should

reject the claim that unfairness is an individual harm. The priority view is a meaningful position.

The argument just discussed contains some technical ideas which are much better known to

economists than philosophers, but they will be presented in a way which does not presuppose a

technical background.

References

Broome, John (1991). Weighing Goods. Oxford, Blackwell.

Diamond, Peter (1967). "Cardinal welfare, individualistic ethics, and interpersonal comparisons

of utility: comment." Journal of Political Economy 75: 765-766.

Harsanyi, John (1955). "Cardinal welfare, individualistic ethics, and interpersonal comparisons of

utility." Journal of Political Economy 63: 309-21.

McCarthy, David (forthcoming, a). "Measuring life's goodness." Philosophical Books.

McCarthy, David (forthcoming, b). "Utilitarianism and prioritarianism I." Economics and

Philosophy.