The two main forms of utilitarianism are preference utilitarianism and hedonistic utilitarianism (but there are other forms). According to hedonism pleasure should be maximized, while according to preferentialism (as I will call it here) it is the satisfaction of preferences (or “utility”) that should be maximized, i.e., people should get what they wish for to the highest degree. The “crudest” form of preferentialism simply takes people’s actual wishes as given and asks how they can be satisfied, while other versions qualifies the theory by, for instance, demanding that the preferences must be “rational” or the like to count. An “economistic” version simply assumes that people want as much income and/or resources as possible.
A problem for preferentialism – especially in its cruder versions – is adaptive preferences. People who claim to have certain wishes or desires may have adapted them to the fact that there is not much in life they can really get. Or they may have adapted them to certain social expectations of them. For instance, a poor and uneducated woman might claim that she really doesn’t want much for herself and that she doesn’t deserve as much as her husband, because life without patriarchal structures are almost unthinkable. Should we, then, say that the way she “chooses” to live reflects her true preferences?
We can easily see the difference between hedonism and preferentialism. People may, and often do, want things that will not make them happier, and sometimes they even want things that will make them unhappier (of course, a hedonist could also want such things, but then the reason would be to increase happiness for someone else). Of course, as a hedonist I cannot see the good in simply giving people what they want if there is no connection to actual pleasure and pain. Now it is often the case that people – especially people who have few choices in life – want things that will actually make them happier, but we all know from experience that there are many exceptions to this rule.
Serene Khader has discussed adaptive preferences in the book Adaptive Preferences and Women’s Empowerment. Like myself, she rejects preferentialism but do not endorse hedonism. Instead she has another consequentialist theory that wants to maximize “basic flourishing”. Certainly, this theory has some affinities with hedonism, since hedonists (especially in a political context) are also interesting in providing people with the things that Khader includes in basic flourishing; basically the things that poor people (and especially poor women) lack to achieve basic well-being (the book is mainly about development in poor countries).
Why then, does Khader reject hedonism? Although the criticism against preferentialism is thorough, the criticism against hedonism is handled in a few lines. She admits that a hedonist can view a preference for, e.g., staying malnourished as an adaptive preference, since it is more pleasurable to be nourished than malnourished, but she claims that “many preferences we intuitively classify as adaptive may not produce psychological suffering” (p. 50). Khader’s examples of these intuitive cases are not, however, very enlightening. The only actual example against hedonism (as opposed to the many examples against preferentialism) is a poor and oppressed (through discrimination) worker who “gains immense subjective pleasure from the small mercies in her life” (ibid.).
Now I would say that if we can imagine such an unlikely person, who is immensely happy despite being poor and oppressed – and if we have good reasons to think that this person would be unhappier if her social and economic conditions improved – then we do not have any obligation spend resources to help that person. But, again, these people are probably so rare that we do not really have to take account of them in political discussions. If people are living in wretched conditions but still claim to be as happy as they could be, we can simply assume that they are victims of adaptive preferences, unless we get clear evidence for individual cases that this is not so.