This essay is an analysis of abortion in utilitarian terms. Compared to some writings on abortion, it is very short. And it is short for good reason: utilitarianism really has very little to say on this issue. Intuitionists will predictably take this as "proof" of the "inadequacy" of utilitarianism. The utilitarian, however, after noticing the various muddles produced by the intuitionist - the arguments over whether the fetus is a "person", whether one person has the "right" to the use of another's body and/or whether someone has the "right" to determine what occurs in their own body (and in the case of both, the interminable debates as to what is to be done about the dilemma), and whether having sex in the first instance amounts to an "invitation" and the effects of this - might take this issue to be a good example of the inadequacy of intuitionism.
[It may also be noticed that utilitarianism avoids altogether a problem which has plagued many attempts to justify abortion from a more conventional moral framework. The problem is this: "if it is sometimes permissible to kill a fetus, where is the dividing-line between this and killing a normal baby (or adult)?" The problem emerges because abortion is held to be sometimes "permissible", but killing a normal baby (or adult) not, and it is quite hard to point to a hard and fast (morally-relevant) distinction: (e.g.) at what stage does consciousness develop? Utility avoids the problem because it does not share the assumptions - it does not say that it is never right to kill a normal baby (or adult), in fact the considerations in each case would be quite similar (with the exception of alarm in the case of adults).]
Utilitarianism, of the hedonistic variety, is (we may recall) concerned only with pleasure and with pain. Therefore we shall be concerned with the amounts of pleasure and pain in situations where abortion is permitted as contrasted with the amounts of pleasure and pain where abortion is forbidden. It might be suggested that the main consideration would be the interests of the fetus: not only can its future life be expectedly happy (or at least having a balance of happiness over suffering) it might also be the case that the abortion itself is painful, particularly if it occurs later in the pregnancy. However this focus on the fetus is unwarranted: any suffering involved in the abortion itself can be avoided by simply aborting the pregnancy sooner (before the fetus has even developed the capability of suffering), or with painless techniques. The direct suffering of the fetus can therefore be no argument against abortion generally, only the bad practice of it. A more significant consideration exists if we posit that the future life of the fetus involves a probable balance of happiness over suffering for the fetus. This would seem to be a definite point against abortion, though not, we shall see, a dominant one.
The second party we might consider are the parents and other family, and guardians if the alternative to abortion is adoption. According to some studies, having a baby appears to decrease the happiness in a relationship - even in those cases where the pregnancy is desired [see Eysenck on happiness]. But again, this need not be considered too much, it is not a dominant consideration.
As is the case with many issues in a utilitarian system, the rightness or wrongness of the act in question turns mainly not on the effects of the act on the agent, nor on the being(s) directly affected by the act, but on the less direct effects on the community at large. The issue of abortion, stripped of the language of "rights" and emotional sway over "murdering babies", actually becomes one of the desirability of increasing or decreasing the population.
Given that there must be some population size which is felicifically-optimal, it is clear that Utility will proscribe new births above this amount. Some may not realize, though this will hardly apply in the foreseeable future, that below this population size utility will prescribe reproduction. Utility would give positive duties in this case - you ought to have a child, and it would be wrong of you not to - in a manner that might surprise maintainers of other ethical systems. So the utilitarian who suggests that the future happiness of the child, combined with the estimated value of the effects on others, is such that Utility opposes abortion, must admit that this would imply that Utility prescribes an increase in population - and that this would (prima facie) apply to anyone capable of producing a child. So Utility is generally against abortion only when it is generally for raising the population, in which case the obligation on the woman with the unwanted pregnancy not to have an abortion is not so very much greater that than that upon the average woman to get herself pregnant. In terms of utility, the actual act of abortion is not a particularly significant one.
A brief mention must be made of why it is that the comparative effects on the community at large are dominant in this issue, and why the other considerations are not. It must be remembered that the raising of a child in a modern developed country has a very large cost in financial terms, which is highly significant. The amount required to raise one child in a developed country could raise, at a guess, at least a dozen in a poorer part of the world. So if increasing the human population is the aim, this can be achieved more effectively elsewhere. However in these days of increasing environmental pressures, institutionalized animal abuse, and terrible inequality, increasing the human population is not generally what we should be aiming for. [If everybody became a utilitarian, would the human race become extinct? No, because, if everybody were utilitarian, these problems would not exist to the same degree.]
In utilitarian terms, a general prescription either for or against reproduction is very hard to justify: each case would have its own relevant and specific features. For example, it could be argued that all the many man-made problems in the world do not suggest that good people (i.e. the kind of people who would be influenced by ethical considerations) should not reproduce - but rather that they should reproduce, to (in the longer term) increase the number of people trying to fix all these problems. "Where would ethical theory be today, if James Mill had not decided to reproduce?", some might ask. However, this requires the assumption that reproduction is the cheapest method of recruiting moral agents, even granting (as it must) that it has a high cost in time and effort - and this would require empirical support (which would, if found, be the death-knell of most projects to improve society via reasoned argument.)
If a utilitarian found herself unexpectedly pregnant, then it seems that she should get an abortion: if Utility was in favour of her reproduction, she should've been trying to get pregnant, so the pregnancy should not be a surprise. If Utility is against her reproduction, she should have an abortion - the mere fact of her personal desires changing, as they might, should not be allowed to prejudice the decision... assuming that the desire is not too strong to be defeated without serious psychological harm, and that abortions are easily available. (Followers of Hare might wonder if abortion is contrary to a number of prima facie principles which she may hold. If so, since we are critically examining the issue, she should change her prima facie principles.)
In the world as it is, the utilitarian shall not be particularly concerned with the issue of abortion per se, for it is clearly outweighed by so many other issues. In better circumstances this may change, but by such time the issue will likely be irrelevant anyway: pregnancy will likely be subject to much stricter controls (but probably not for Russell's reasons). To suggest, as many non-utilitarians do, that the rightness or wrongness of abortion can be determined generally and simply by consideration of the effects on those directly involved - the parents, the doctor, and the fetus - is dangerously shortsighted and simpleminded.