Against Effective Altruism

Why I don’t help starving Africans

Every day thousands of people in Africa die of starvation. Effective altruists such as Peter Singer and Peter Unger have argued we are morally required to give a large share of our income to charity to save them, because, the loss to us is minimal and the good done to them is gigantic. According to Singer not donating to charity is equivalent to walking past a drowning child in a shallow pond because you don’t want to ruin your expensive suit: An implausible verdict.

In this article, I mount three arguments against effective altruism. One, a utilitarian or moderate deontologist reading of effective altruism would implausibly require us to have children, or, humanely breed shrimp on an industrial scale; two, the strong pull of the utilitarian grounding of effective altruism is in its opposition to suffering, but this strong pull is fictitious, so, the strong pull is lost; three, we can be morally required to plausibly save the drowning child while not being required to save the starving African.

Midway through this article I make a limited defence of the exclusive existence of agent relative value as against agent neutral value; this limited defence is the uniting moral idea behind the three arguments (though all three arguments can stand independent of it). The essence of this piece is thus: Man exists to pursue his own happiness in life not the limited advancement or maximisation of the aggregation of pleasure and pain across sentient beings.

1. Bringing Children into Existence & Humanely Farming Shrimp

The utilitarian and moderate deontologist believe you should save the starving African because of the great good which will be lost to the world if you do not, and, the minimal cost of doing so to yourself. Today it costs about £3,500 to save a life. I maintain if you can be obligated to donate £3,500 to ensure someone lives for another 40 years, then, analogously, you can be obligated to create a child at the cost of £3,500 to ensure someone lives for at least 40 years who otherwise wouldn’t. This conclusion is very implausible, hence, the verdict you can be required to donate £3,500 to save the starving African is too. What are the possible problems in this analogy?

Many people will point to two facts, one, having a child is a greater burden than giving up £3,500, and, there is an important distinction between someone losing a life and the universe not gaining another life. To address the first objection, we need only imagine that the £3,500 is enough to finance the whole life of the child until it reaches adulthood: I take it we still have the intuition we are not required to have the child. Feminist and conservative concerns can be dealt with via imagining artificial wombs etc. Should this be thought to be far-fetched, only remember Judith Jarvis Thomson’s people seeds in her esteemed article, ‘A Defence of Abortion’.

The second concern is there remains a distinction between not saving an existing life and not bringing a new life into existence. I am not too sure about this. If pleasure is itself agent neutrally good, i.e., if we all have reason to bring it about due to its nature, then, it doesn’t seem to matter whether it embodies itself in an existing sensory receptacle or a new sensory receptacle. If this is true then we are simply obligated to bring into existence new people who have net good lives. Were we to imagine the drowning child in the shallow pond example is knocked out and loses all his memories, personality and social connections, putting him back in the state of a new born baby, I reckon most people would still say it should be saved. Accepting this , however, gives us strong reason to bring a child into existence as what is lost in this drowning child example is pretty much what is lost by not bringing a new life into existence.

Moreover: If we could humanely breed a great number of shrimp all on cocaine for £3,500 which produced greater pleasure than saving the starving African, then, although it would be to the advantage of nobody at all, hedonist utilitarians such as Jeremy Bentham would have to advocate for it. Objectivist list moderate deontologists may believe they can avoid this counterexample because shrimp cannot access the higher goods of life; the higher goods lost to humans via the cost of humane shrimp farming are not outweighed by their great pleasure. I doubt this due to the fact objectivist list moderate deontologists usually allow pleasure in some amount to outweigh higher goods, therefore, enough shrimp pleasure could tip the balance against any higher goods humans might enjoy without the humane shrimp farming.

Rejecting Agent Neutral Value & Defending the Pursuit of Your Own Happiness

Howard Roark in Ayn' Rand’s ‘The Fountainhead’

Both of these counterintuitive verdicts point to the implausibility of pleasure being an agent neutral value, i.e., reason for everyone to bring it about. This is a big claim. I want to sketch here Eric Mack’s rationale for believing all values are instead agent relative values, i.e., reason for action exists solely in the person to which the pith of the value will be embodied. To make this distinction clear. Agent neutral hedonism would say me getting pleasure from cake is at least some reason for you per se to give me cake. Agent relative hedonism would say me getting pleasure from cake is no reason for you per se to give me cake. Should Mack’s contention be right then it will undermine the effective altruism of moderate deontologists such as W. D. Ross, Micheal Huemer and Amos Wollen too.

One, utilitarianism as a moral philosophy argues there are only agent neutral values which exclusively provide reason for us to act. Now any moral theory which contains any personal prerogative, i.e., admission a person can act against the agent neutral good to serve his own desires, and, any plausible moral theory will, admits agent relative value to exist and to trump agent neutral value in many cases. Lots of people enjoying stamping on your head more than you feel pain is not reason for you to allow your head to be stamped on – you have a personal prerogative to deny what is supposedly the agent neutral best. The admission agent relative value can trump agent neutral value though makes agent neutral value strange, counterintuitive to a certain extent, because, the pursuit of your own good is very clearly good for you, i.e., reason for your action, but, contrastingly, the good of another person is not obviously reason for your action, e.g., their pleasure from stamping on your head is not any reason at all for you to lie down and be stamped on; thus, we can cast doubt on it being of any value for you at all.

Two, agent relative and agent neutral value appear to have no common scale on which they can be weighted so as to ensure an optimum set of choices, indeed, any reason given on behalf of agent neutral value will likely be the same reason for saying agent relative value is trumped by agent neutral value in everyday instances, but, we have rejected this conclusion, thus, we have much reason to question it when the moderate deontologist invokes it when great agent neutral value is at stake.

Underlying the very idea of agent relative value though is the very strong intuition that each of us properly pursue our own good in our own lives. Ayn Rand in her ‘Objectivist Ethics’ puts the general idea very well: ‘The achievement of his own happiness is man’s highest moral purpose’. Certainly, the vast majority of people think with this in mind when making decisions, and, the majority of the constraints on our behaviour which we all embrace are deontic in character, that is, concerning the right as opposed to the good, or, questions of value. I’m hardly throwing out the moral universe then in arguing for this. Individuals don’t exist to maximise pleasure over pain, or, the good generally. No. Individuals properly exist for their own sake - they may go out for dinner even if that means a starving African goes without dinner. Howard Roark put this with the greatest force after he dynamited Courtland homes, built for the poor, when his designs were stolen through breach of contract:

‘I came here to say that I do not recognize anyone's right to one minute of my life. Nor to any part of my energy. Nor to any achievement of mine. No matter who makes the claim, how large their number or how great their need. I wished to come here and say that I am a man who does not exist for others.’

Now we must face the obvious rebuttal to this line of individualist thinking.

2. Utilitarianism is Not Against Mass Suffering Per Se

It will be argued by utilitarians that a moral theory is very implausible should it allow someone to enjoy an ice cream as opposed to save a million people from torture. In debating this topic at Oxford’s Socratic Society, Bentham’s Bulldog raised this very point. I think there are some very strong deontic reasons as to why the individual should save the thousand from torture, yet, I admit, pending further development in the moral theory of individualism, this is one of the counterintuitive results.

Yet should this counterintuitive verdict be thought to be decisive against the individualism I have presented then parity of reasoning dictates it is decisive against utilitarianism too, because, they too favour inflicting the worst suffering imaginable provided it is outweighed by a greater amount of pleasure in the process, e.g., a thousand trillion people enjoying an ice cream is fine even if a million people are tortured due to it. Utilitarians have arguments for this, but the main traction of the objection arises from the extreme suffering itself, so, this remains against utilitarianism as it is not against it per se. In fact, nothing, I repeat, nothing, stops utilitarianism from creating a universe with the most suffering it.

Huemer has argued the extremes of both schools of ethics points to moderate deontology as the correct form of morality. Even accepting this to be the case; as per my examples of bringing children into existence or breeding shrimp, I do not believe morality can require you to give up £3,500 to save a life from starvation in Africa. At most it is going to require you to provide the means for them to commit suicide, which, I imagine, they already have. Should people still deny the counterexamples above because of the distinction between bringing people into existence beings versus not helping existing beings, I should note the £3,500 could just as easily be required to go to eliminating wild animal suffering of existing animals; again, very implausible, therefore, the effective altruist demand to save starving Africans is too.

3. The Drowning Child Need Not be Left to Drown

I have argued against any obligation to help starving Africans. Does this mean I must counterintuitively be against any obligation for us to save the drowning child in the shallow pond? No. The proper basis for saving the drowning child is reciprocity. You should save the drowning child because you would want to be saved were you in the same predicament, and, were you in that same predicament, the only offer you could make to properly motivate the person considering saving you would be to say you’ll return the favour. Of course, it is not a one-for-one relationship, instead, people who are up for saving drowning people are in a reciprocity pool, i.e., me saving a drowning person obligates others who were not saved directly by me to save me, because, the understanding of my saving is the saved will save any others in similar instances in exchange. I have outlined in details how this reciprocity pool idea works here.

Why does this reciprocity pool not extend to helping starving Africans? The basic reason is it is not to the advantage of us in Europe to be in a reciprocity pool with starving Africans, because, we would give them £3,500 to save their lives and get little or nothing in return. Sure, I could fall into a shallow pond in Africa and not be saved because I was not in a reciprocity pool with them, but this is a price I am willing to pay. (And, anyway, I could call out ‘I’ll pay you £100 to save me’ and I would be quickly saved). It must be admitted reciprocity being the basis for helping the drowning children does have counterintuitive results, e.g., helping a disabled man with no family would not be morally required, because, no person would enter into a disadvantageous reciprocity pool with him.

Conclusion

The individual properly pursues his own happiness in life not minimising the starvation which goes on in Africa. The effective altruist argument against this proposition leads to the implausible result we are obligated to have children or finance humane shrimp farms, or, spend most of our money on averting wild animal suffering. And, of course, Singer’s argument implausibly requires us to live at a subsistence level anyway. At a deeper level, effective altruism of the utilitarian kind relies on a rejection of agent relative value which is at the height of implausibility. Once agent relative value is recognised, however, the moral case for agent neutral value becomes far less steady and with it so does the case for effective altruism of either the utilitarian or moderate deontologist variety. The case against an obligation to help starving Africans stands.