In a stimulating and controversial paper published in Think,Footnote ¹ Ole Martin Moen has argued that you should not give to beggars. His argument is simple and familiar. It is that the beggar one encounters in the rich world is, in the scheme of things, doing very well for herself. The London beggar is hungry, ragged, addicted, and schizophrenic; but she is like unto a king in comparison to the starving Ethiopian. If she receives only a few pounds a day and falls asleep in a doorway, she is still much better off than the millions of people in the world now dying for lack of food or clean water. It follows that a pound put in the hand of that beggar is a pound wasted: it should have gone to the person whose need is most urgent. Moen counsels you to ignore the beggar as you pass her on the street, and to give all your spare pounds instead to charities that assist the world's most needy. In general, in your action, you should aim to do the most good you can. I wish to say here a word in favour of the beggar, and to show what I think is wrong with this currently fashionable line of reasoning in applied ethics.

Let us put some initial objections aside in order to focus the argument. We assume that you, the rich passer-by, have a right to your money. We assume that it is possible to know that by donating your pound to charity you would send it to the most needy and would assist them. We assume that the long-term effect of that institutional assistance would not be harmful. We assume that you are honest, and really will so donate your pound when you get home. We assume that the beggar really is in need, and that by stopping to give to her, you would assist her. We assume that you justly possess enough wealth to meet the standard for basic social participation in your society and that your ‘excess wealth’ is everything that exceeds that requirement. Let us suppose that you have a pound a day excess. Now, who should get that pound?

Moen says, ‘we should not give money to beggars’. He means that if you do give, then you fail morally. It is a bold assertion. I will argue that it is false. A humbler interpretation is available, which I will consider for the sake of discussion: that you do not fail morally if you do not give. But I will argue that even this is false. Here are the two interpretations:

1. (1) It is false that you ought to give money to beggars.

2. (2) You ought not give money to beggars.

On standard theories of ethics, (2) is stronger than (1): it entails but is not entailed by (1). (1) and (2) are derived respectively from two general principles about moral obligation and optimality of action. (An action is optimal when there is no alternative whose consequences are better.) These are the general principles.

1. (3) You ought A only if A is optimal. [From (3) is derived (1)]

2. (4) If A is not optimal, then you ought not A. [From (4) is derived (2)]

Both (3) and (4) are falsifiable. Consider first (4). You are traveling through the icy mountains of Kyrgyzstan. Touched by the grace and hospitality of its people, you leave feeling an affinity with that land, and sympathy for the hardships of its life. Returning home, you donate every month your excess wealth to a charity that assists the education of Kyrgystan's young women. But of course, your money could have done more good elsewhere, and you have withheld it from those who need it most. You could have done better. But did you do something wrong? Did you do something that you ought not have done? I think you did not. Your obligation is a duty of beneficence: it is to assist the needy. It is wrong not to assist the needy; but it is not necessarily wrong to do less good than you could have done.

Consider now (3). Here is a counter-example told to me by John O'Neill. It is a better example than mine. Suppose that you are an effective altruist on the way to a critical and unrepeatable meeting in which you expect to secure donations for foreign aid from businesses. You expect by those donations to save many lives. Let us say that taking into account the total probability of your success, the expected savings are 100 lives. But on the way you pass a lake in which a child who has gone swimming is drowning. If you stop to save her you will miss your train, and will miss the one chance to make your highly-promising pitch. May you stop to save the child?

I think the example suggests that it would be psychopathic of you not to stop. You certainly ought to stop. That is, you ought to do something sub-optimal. Now it does not follow that the optimal act is not something that ought to be done; possibly it does not even follow that the optimal act is not something that you ought to do: for O'Neill's example might describe a deontic conflict or dilemma. But it does show that (3) is false: it is false that you ought do something only if it is optimal. My first example might likewise describe a deontic conflict: what is optimal ought to be done, but you may do something that is not optimal, even though you cannot do both. I can remain neutral about both examples, so I will: they might describe conflicts, but they might not.

What the examples show is that ‘ought’ is not controlled only by optimality, but by something else as well. In my view it is simply the misery of the beggar – the inhumanity of her condition – that compels you to assist her. You are guilty if you do not. Charity requires of you that you assist the indigent. That some are in a still worse state than others does not necessarily relieve you of your duty to assist the less worse-off, even if you cannot assist them all.

The word ‘charity’ might feel moth-eaten: it recalls plastic cans of pennies, discarded novels, and bad television. But its root is the beautiful Latin word caritas, which means ‘love’. St. Paul's dictum, Caritas numquam excidit was inscribed by the neglected critic of utilitarianism, Friedrich Nietzsche, on his father's gravestone: ‘Die Liebe hört nimmer auf’ – ‘Love is never expended’. Charity is not merely – perhaps not even primarily – about doing the most good you can: it is about relations of solidarity and fellowship amongst people, and the acts and virtues that are constitutive of those relations. It is even possible for a world that has more suffering, but also more fellowship, to be a better world.

In a 1943 essay on utopias and the difficulty of depicting a happy society, George Orwell wrote:

Fellowship is a necessary good for ethical life, and certain actions are constitutive of fellowship. This fact is partly what lies behind the counter-examples I have given. To pass by a person who is suffering and in need, and to ignore her, is to degrade and humiliate her. To degrade and humiliate a person is wrong. It is to dishonour human fellowship. This might show that proximity is not always ethically irrelevant: distance does sometimes matter. It is possible that failing to donate excess wealth to those who are most poor is always wrong; but even if it is, so to fail is not necessarily to degrade and humiliate those people. There are in addition reasons of virtue: to pass by a person in need who calls upon your humanity is to harden your heart: it is to become an insensible and calculating person. And there are teleological reasons too: it is to normalize in the public arena indifference to suffering and unfairness.

One might object here that I have not yet shown that optimality does not govern ‘ought’: I have not shown that (3) and (4) are false: I have merely shown that there are several distinct goods, and that what minimizes suffering is sometimes on balance sub-optimal. We should think of virtues of compassion and relations of fellowship as consequences which like well-being partly determine the goodness of an action. ‘Doing the most good’ should be expanded therefore to include relations and virtues as well as welfare. Though (1) and (2) are false, either (3) or (4) is not. Call this optimalism.

I do not need to deny this – my purpose was only to show that Moen is wrong about begging – but I should like to, and will say a brief word about it here. It is an enormous subject, and I cannot say much. One reason for thinking that it is not true is that it collapses all moral concepts into a single concept: that of failing to do sufficient good. I have appealed in my argument to concepts like violation, dishonour, and transgression. They are to my mind an irreducible part of the richness and complexity of moral thought. if you think that an action can be wrong because it violates something important, then you cannot believe in optimalism.

Another reason for thinking it false is that there are certain goods that must be honoured, and not merely promoted. That is because to promote them, to think of them merely as outcomes which one tries by one's action to increase, is to vitiate them. In his long prose-poem, ‘The Dream of John Ball’, William Morris wrote:

and this dictum became a slogan of socialism, and the motto of the Clarion Cycling Club. Now Morris did not mean that we should optimize fellowship. To think of fellowship as something one might optimize is to kill it. Is there a reason of this? Perhaps it has to do with the particular value of instances of the good. Instances of wealth are substitutable; that is why we can think of optimizing wealth. It is easy to imagine scenarios in which we make some people poorer, but others richer, and the world thereby better. But instances of fellowship are not like that. Fellowship is like friendship. Friendship is good; and in general, the more friendship the better. But if you think that each instance of friendship might be sacrificed to increase total friendship, then you shall have no friendship at all. I'm afraid that I cannot say any more than this.

To return to the main argument; my conclusion is this. You ought assist the needy. But it does not follow that you ought assist only the most needy: that proposition is false. To pass by, to avert your eyes, to withhold your assistance: it is wrong, even though good may come of it.