▻ FIXING THE TERMS OF DISCUSSION
There is no canonical formulation of utilitarianism. For one thing it comes as direct or indirect utilitarianism; and the direct form has two branches, act- and rule utilitarianism. One formula will not fit them call. For another, even for any particular form there is no consensus on exactly how it should be formulated.
Before the Question can be tackled, some conceptual work needs to be done. Bear with me a bit – I’ll keep it to a minimum.
▻ ACT UTILITARIANISM
I am going to work with act utilitarianism. This is partly to keep the discussion within reasonable limits. I assume that this is the form of utilitarianism with which most people are familiar. I also assume that my discussion can be adapted to the other forms of utilitarianism.
(1) An action is obligatory if and only if its (actual or probable) consequences are better than those of every alternative;
(2) an action is right if and only if its (actual or probable) consequences are as good as those of every alternative;
(3) an action is wrong if and only if its (actual or probable) consequences are worse than those of some alternative.
The obligatoriness, rightness or wrongness of an action depends purely on its consequences. No action has any moral character independently of its consequences. This means that an action cannot be intrinsically obligatory, right or wrong.
☛ Maximisation & distribution
Utilitarianism mandates maximizing some kind of ‘good’ – but what ? What’s the metric for maximisation ? The older, Benthamite version required the maximisation of happiness or the balance of pleasure over pain.
The two main current versions take the satisfaction of preferences or the promotion of interests as the metric. Patrick Stevens is perfectly correct to say that no objectivity need be involved here. The preferences may be simply those that people happen to have but they could also be the preferences they would have if they had complete information, perfect computing power, fully organised preference schedules – or at least more information, a greater capacity to reason straight, and preference schedules without glaring inconsistencies.
Interests may be just a matter of a person’s preferences but interest-utilitarianism usually assumes that interests comprise ideals, mental, emotional and physical well-being: of some of these goods the person may be the best judge, but of others not.
Maximisation usually goes along with the requirement of a pattern of distribution. Thus Betham’s formula, ‘The Greatest Happiness [maximisation] of the greatest number [distribution]’.
Now we can tackle the question.
The Question asks why utilitarianism presupposes ‘that an objective utility can exist’. The point has already been made that there is no necessity for utilitarianism to base its metric on what is, or what it takes to be, objectively good for a person. It may do so but there is no requirement. It may run on the satisfaction of preferences that are as subjective as anyone could imagine.
But the Question raises some serious issues about the claims of utilitarianism to define a rational decision-making mechanism for public policy. I mention three : disagreements within utilitarianism, the problem of making interpersonal comparisons of utility, and the problem of incommensurability.
☛ Disagreements within utilitarianism
If the aim is (as the Question puts it) to ‘benefit most members of the population’, a public policy can presumably only run on one version of utilitarianism. What if half the society wants policy-making to be determined by preference-utilitarianism and the other half on interests-utilitarianism ? There is no guarantee, rather the opposite, that the same policies can fulfil the criteria of both sorts of utilitarianism. And what rational choice can be made between them ?
☛ Interpersonal comparisons of utility
Consider a point made by Lionel Robbins :
[S]uppose that we differed about the satisfaction derived by A from an income of $1,000, and the satisfaction derived by B from an income of twice that magnitude. Asking them would provide no solution. Supposing they differed. A might urge that he had more satisfaction than B at the margin. While B might urge that, on the contrary, he had more satisfaction than A. We do not need to be slavish behaviourists to realize that here is no scientific evidence. There is no means of testing the magnitude of A's satisfactions as compared with B's. If we tested the state of their blood-stream, that would be a test of blood, not satisfaction. Introspection does not enable A to discover what is going on in B's mind, nor B to discover what is going on in A's. There is no way of comparing the satisfactions of different people. (L. Robbins, ‘The Nature and Significance of Economic Science, London : Macmillan, 1932, 123-124.)
If Robbins is right, there is no way that, in the words of the Question, the utility of any policy ‘can be determined and known’.
If Robbins is wrong, and interpersonal utilities can be known there is no guarantee that when individual welfare functions - individual preferences or interests - are aggregated, the result will be coherent. This, crudely stated, is the so-called Arrow Problem (after Kenneth Arrow). Individual welfare functions may not be combinable to produce a consistent result.
The older, Benthamite utilitarianism had the advantage of using a single metric, that of happiness or the balance of pleasure over pain. JS Mill complicated matters by introducing a distinction between higher and lower pleasures, which can plainly clash. GE Moore recognised in the early 20th century that there is a plurality of values with rivalrous claims to be maximised. The point came to be realised that values (even taken just as subjects of preferences) can be and often are incommensurable. Knowledge is a value, beauty is a value, health is a value, friendship is a value, pleasure is a value, justice is a value, but there is no way in which these values can be legitimately compared or commensurated. We cannot formulate a social policy on the basis that health is more important than knowledge, or beauty more important than friendship. There is no common scale by which they can be measured. All we can do is to decide that in situation A one value is to have priority over others. If there is disagreement, how can it rationally be settled ? Preferences clash; there is no objective way of resolving the disagreement.