Utilitarianism

Note: This page was written some time ago, when I was still a Christian. I'm not any more, so it's likely that some material in here no longer reflects my current opinions.

(You might want to read some disclaimers.)

Utilitarianism, for all the unfortunate connotations of the word (which conjures up images of factories, high-rise buildings and all things ugly-but-functional), is an ethical system of great elegance and beauty. It is also a system of great importance: I would guess that the large majority of people in our society are more or less utilitarians, and that they are such without having given the matter a moment's thought.

It arouses strong feelings. Most proponents of utilitarianism would probably say that it's not only right, but obviously right; that those who are not utilitarians are living in the Dark Ages. Many of its opponents consider it a thoroughly evil thing, tending to lead to the erosion of vital moral principles. I think both sides exaggerate.

The issue is a particularly interesting one for the Christian: utilitarianism has much in common, in practice, with Christian ethics, but little or nothing in common with Christian philosophy.

This essay will attempt to analyse the idea of utilitarianism (which is more complicated, and richer, than it may seem at first glance), and to explain what I think is wrong with it and what I think is right with it. At the end, there are some rather disorganised comments about what a really adequate system of ethics would have to look like.

1. What is utilitarianism?

This sounds like an easy question. "The greatest good of the greatest number". Simple. Or is it? In any real situation, there are many people involved; they will all be affected in different ways; there is no reason why the "greatest number" should receive the "greatest good".

What is usually meant in practice by that slogan is something like the following procedure for choosing between two or more actions.

  1. Look at the state of the world after each action. Look in particular at the level of happiness of each person in the various situations.
  2. Add up, somehow, those levels of happiness in each case.
  3. Compare the results. The one which leads to the maximum total happiness is the (morally) right one.

The thing to notice about this is that it actually involves a lot of quite separate principles. I think it is fair to say that they are all part of the idea of utilitarianism; a system of ethics is utilitarian in so far as it accepts some or more of these principles. Someone who accepts some of them but not others may reasonably be called a utilitarian, even if the procedure above would be seen by them as a coarse caricature.

There is at least one important issue which we haven't addressed so far: we have to consider the entire future of the universe in order to make our decision. I shall consider the practical difficulties of this later; there is also a theoretical issue: we are presumably required to compute the total amount of happiness in a person's entire lifetime. So we need some sort of integral calculus for happiness.

I shall refer to different versions of utilitarianism as stronger or weaker according as they accept more or fewer of the principles above. "Strong utilitarianism", simpliciter, will mean the complete system above in all its glory. Most utilitarians are not in this sense strong utilitarians.

2. Why utilitarianism is plausible.

Utilitarianism has the awkward property of seeming entirely obvious to its proponents, and clearly wrong to its opponents. This can make for discussions with much more heat than light. If it already seems obvious to you that utilitarianism is right, by all means skip this section.

There are no ethical first principles which are agreed on by everyone. On the other hand, there is a striking level of agreement about what is actually right and wrong, in concrete cases. Of course, there are disagreements; anthropologists have turned up some pretty surprising ones. But there is something pretty close to a consensus that (in most cases) murder, lying, rape and theft are bad, and that (in most cases) generosity, healing, truthfulness and loyalty are good.

One obvious thing that these things have in common is that most of the things near-universally agreed to be good are things which make people happy, and most of the things near-universally agreed to be bad are things which make people miserable. And in most exceptional cases, there is a clear recognition that they are exceptional cases: excuses are made.

Furthermore, the actions usually reckoned to be the worst are often the ones which cause the most suffering. Rape, for instance, which causes lasting psychological trauma as well as involving physical injury, is generally reckoned to be morally much worse than theft.

So, utilitarianism seems to do a pretty good job of giving the right answers. There is also a theoretical justification for (at least) something rather like utilitarianism. It seems clear to me that, all else being equal, something which makes me happy is better than something which doesn't. After all, that's one way in which I make decisions (though, to be sure, I wouldn't in such cases call them moral decisions). Since it seems plausible that all people are ethically equal, this means that anything which makes anyone happy is (all else being equal) better than something which does not. This seems to lead naturally to something very like utilitarianism.

3. Strong utilitarianism won't do.

I am not writing exclusively for Christians, so I'm going to avoid, where possible, arguments that just say "Christianity says X; utilitarianism implies not-X; so utilitarianism is wrong".

However, what I've called strong utilitarianism has a terrible problem: it is grossly inconsistent with ethical intution (or, at least, with my ethical intuition) in certain cases.

For instance, suppose that (never mind how; all that matters is that it should be conceivable) I could, by subjecting my aged grandmother to the most appalling tortures (which I shall leave to your imagination, should you happen to have that sort of imagination), relieve a sufficiently large number of people from one minute's toothache. No matter how small the amount of suffering from which each person is thus delivered, and no matter how great the amount I cause to my grandmother, if the number of people is large enough then the total amount of suffering in the world will be decreased by this transaction. Therefore I ought to torture my grandmother. This, it seems to me, is unacceptable.

4. Weaker versions of utilitarianism, and their problems.

Of course, there are ways round this problem. For instance, we could model happiness and misery with a modified number system, containing values incommensurable in the sense that no integer multiple of one was as big as the other (for mathematicians: in other words, we could work with a non-archimedean valuation). Or we could replace the idea of adding up utilities with some other operation: take the single biggest happiness or misery, and just look at that, or something. (Actually, either this second option actually reduces to the first, or else it doesn't work. Proof left as an exercise to the reader.)

So, we can get around that particular problem. Alas, there are others, though I wouldn't claim any of them as an actual refutation of (weak) utilitarianism. I shall take the utilitarian principles I enumerated above, and describe some objections to them.

Is this really plausible? It doesn't seem so to me. If I kill someone, isn't there something intrinsically bad about that, even if (as might be the case) the killing turns out to be right in terms of maximising utility? I think most people would agree that a killing of this sort would be at best a necessary evil.

Suppose I tell a defamatory lie about you to an acquaintance of mine, who has never had and never will have any sort of interaction with you, and swear him to secrecy. This makes no difference whatever to your future happiness. Does that make it OK? It seems clear to me that it doesn't (and if you disagree, there is really nothing I can say to convince you).

Isn't there, in fact, something fundamentally good about truth and bad about falsehood? Some such idea seems to underlie the near-universal agreement that lying is in itself bad.

And what about, say, someone who takes great pleasure in annoying other people? Suppose I get enormous satisfaction from causing you minor but genuine unpleasantness. Does that mean that it's right for me to do so?

I wouldn't like to claim that this is obviously wrong. But is it really obviously right?

What about criminals? If I am in the process of raping your wife, do you really have to consider my well-being as carefully as your wife's in deciding how to go about stopping me? (Perhaps the answer is "yes". It certainly doesn't seem like an easy question to answer.)

These two principles, taken together, give rise to the problem in the previous section. But there is another problem. Is it really obvious that different sorts of happiness are commensurable? How do you compare the pleasure person A gets from an hour of wild sex with his wife, the contentment person B has from the knowledge that his money in the bank is earning him piles of interest for his retirement, the wonder person C feels on contemplating the starry sky, the thrill person D has when listening to her favourite piece of music, person E's enjoyment of an evening listening to a stand-up comic, and so on? And how do you weigh those up against person P's toothache, person Q's unhappy marriage, person R's fear of cancer, person S's resentment of unfair treatment long ago, person T's frustration at having spent three weeks chasing a bug in his computer program? I don't know, that's for sure. I don't even know how to do similar comparisons when all the people involved are myself: in difficult cases it feels a lot more like tossing a coin than like choosing the best of a neatly ordered set of options.

5. The practical problem.

Let's pretend, for the sake of argument, that all those problems are resolved, and that I'm fully persuaded that utilitarianism (of your favourite variety) is correct. I now have a decision to make; for instance, I have to decide whether to cycle home in the dark without lights (thus endangering a few people slightly, maybe) or to be late home (thus upsetting my wife and perhaps not managing to get anything for dinner). This is a trivial example; it should be easy to work it out. ... Not a bit of it. I have to work out the entire future of the whole universe (possibly radically different in the two cases: remember the butterfly effect), work out exactly how happy each person is in each case and for how long, and add it all up. Good grief.

In practice, what the utilitarian recommends is entirely different. I should make guesses as to the likely effects of the actions I'm considering, estimate the resulting levels of happiness, and do the best I can at adding them up in my head. Anything more is impossible, and in any case I can't be blamed for things I can't predict.

That last remark, if actually made by a utilitarian, would amount to an abandonment of one of the key principles of utilitarianism which I haven't mentioned so far: you don't do things so as to have done the Right Thing; you do them because that has results which are good. (This is hard to explain; I apologise if I don't seem to be making any sense.) In other words, when making an ethical decision you aren't out to maximise your own righteousness: to the utilitarian, that is a horribly selfish way of thinking. You're acting for the common good.

But we can't have it both ways. Either I take into account all the effects of my actions (impossible), or I abandon the attempt to maximise overall utility -- for the future consequences of my actions are in most cases much greater than those in the foreseeable future. Perhaps it's possible to get round this by claiming that there is some irreducible random element in exactly what happens, and that beyond some point in the future the consequences of my actions will be swamped by the results of amplified random noise; if this is so, I need only consider a finite portion of the future, and things look less bleak.

6. What's right with utilitarianism.

If you are a Christian, or some other sort of non-utilitarian, you have probably been reading the foregoing sections with either boredom or glee, depending on whether you've already thought of all those arguments against utilitarianism yourself. I'd now like to suggest that there is much to be said in favour of utilitarianism, despite its problems.

The first point is one I've made already: utilitarianism, in so far as we can actually apply it (which means, in practice, only looking at a small chunk of the future and only looking at a small region of the universe), actually does a pretty good job of giving answers to ethical questions. And, subject to those approximations, it's quite easy. Most of us are capable of guessing "what will happen if...", and of imagining others' responses to the ensuing situations; and in many cases it's possible to compare the resulting utilities without too much trouble.

Secondly, utilitarianism provides a valuable corrective against the sort of excessively rule-based ethics which come naturally to the Christian, and perhaps to anyone who lives in a society with a very well-defined set of laws. The two approaches to ethics are complementary, and I think we need both.

Thirdly, considering "the greatest good of the greatest number" can be an effective way of defeating prejudices and selfishness. This ethical symmetry is, after all, quite close to such principles as "Do to others as you would have them do to you" and "Love your neighbour as yourself".

These last two points really go together. A great part of the ethical teaching of Jesus consisted of pointing out that the Scribes and Pharisees, in their devotion to the rules, were losing sight of the importance of the welfare of others -- that is, of utilitarian considerations. Of course the Christian ethical system (in so far as there is one) goes way beyond utilitarianism; but one cannot get beyond it without first getting as far as it.

Fourthly, utilitarian arguments are, so to speak, portable. If you need to discuss ethical questions with someone else who doesn't share your system of ethics, you can often get some way towards agreement by considering the utilitarian question first. Then you can discuss the corrections that need to be made.

7. Some concrete suggestions.

I haven't put forward a coherent system of ethics myself. I don't have one. Still, it seems worthwhile to make a few suggestions and observations.

I am, as has presumably been apparent for some time, a Christian. This more or less commits me to a number of ethical propositions; but it has little to say about the more abstract questions of what's really going on "underneath". It is tempting to infer from the absence of an ethical system in the Bible that there is none (beyond the requirement to do the perhaps-opaque will of God); but there is no justification for this, even if one takes a very high view of the authority of Scripture. If you are going to rescue someone from drowning, you just throw them a rope and tell them to grab it; you don't try to give them swimming lessons.

Firstly, about rules. There are a lot of rules in the Bible, and many Christian communities have come up with others (either deduced from the ones in the Bible, or not). I distrust rules, even when they come from the very most reliable sources. It seems improbable to me that any finite collection of rules can really give a perfectly accurate account of what one should and should not do.

This is realised in practice (though often not in theory) by just about everybody. This is why there are all those other rules which aren't found in the Bible: "we can see that this must be right; it's right for the same reason as <thing found in Bible> is right."

Anyway, what should an adequate theory of ethics look like? In the first place, I think it has to consider both actions and states-of-the-universe. In other words, an action can be good or bad apart from its consequences, and a state of affairs can be good or bad apart from its causes and the actions it tends to produce.

Next, I think any theory of ethics has to acknowledge that happiness and suffering are in themselves good and bad (respectively). This is why utilitarianism does as well as it does. But clearly (well, it's clear to me) happiness and suffering, pain and pleasure, aren't the whole story. For instance, adding a component for accurate knowledge provides, at one fell swoop, the wrongness of lying, the value of education and the explanation of why it might be better to be an unhappy Socrates than a contented pig. Possibly some sorts of aesthetic quality could come in here; but I am inclined to say that their moral value is purely a matter of their effects on people's happiness and suchlike.

What about the rightness and wrongness of actions? Again, I think there are probably several different factors here. One might be obedience, or something of the sort: obedience to any proper authority, and especially to God. Generosity, or kindness, or something of the sort, could be another: these are really qualities that apply to actions as well as to people. It's good to be a generous sort of person; it's also good to do generous things, whatever sort of person one is generally.

Of course (putting my Christian hat on) there is a lot more to life than ethics; and an important part of the Christian message is that one cannot really live a good life without the Spirit. In other words -- well, sort of in other words -- ethics really is much too complicated to be captured by any finite set of principles, and we need some sort of direct connection with God, who can see the whole thing without needing to parcel it up into laws and principles. This is reminiscent of the theorem in mathematics which says that there is no algorithm which will reliably tell you whether an arbitrary statement of mathematics is true or false. It would appear that there is also no algorithm which will reliably tell you whether a given course of action is right or wrong.

Some disclaimers

  1. I am not a professional philosopher. The opinions expressed above are not intended for use in the operation of safety-critical facilities; any injury you may sustain by heeding them is your own responsibility.
  2. The essay was written several years ago (somewhere around 1990, I think) and I wouldn't necessarily put everything the same way now as I did then.
  3. There is a thing called "rule utilitarianism"; every now and then I get e-mail telling me I've neglected it. I have neglected it because I think it deserves neglect; my apologies to any rule utilitarians out there. (Rule utilitarianism, lightly caricatured, says that the right way to live is as follows. First, adopt actual utilitarianism, which rule utilitarians call "act utilitarianism", and use its principles to decide on a set of rules you'll live by. Then forget all about utilitarianism, and just follow the rules. Leaving aside the prima facie absurdity of this scheme, it seems to me that if you fill in the details then you end up either with nonsense or with act utilitarianism.)
  4. I received the following delightful message from someone who may possibly be called either Joel Cockhill or Patrick Smith:

    I really do not think that it is fitting to write essays on subjects that you do not properly understand and then post them on the net simply causing confusion. Your essay on Utilitarianism was most misleading containing many misunderstandings of the thesis. If students were to read this it would confuse and misinform. Please either research your pieces properly or stick to subjects on which you are qualified to write.

    My polite request for criticisms specific enough to heed or challenge went unanswered. So, be it noted: Someone On The Net thinks that my understanding is all wrong.