Doing Good
Peter Singer's thought experiment: you would save the drowning child. You make the opposite decision almost every day.
Peter Singer’s thought experiment is simple and, once heard, difficult to shake.
You are walking past a shallow pond. A small child is drowning. You could save the child at the cost of ruining your shoes and being late for work. You save the child. Obviously.
Singer’s observation: we make the opposite decision almost every day. For a few hundred dollars — the price of shoes — a donation to a high-impact charity can save a child’s life from malaria, diarrhoea, or preventable blindness. But we tend not to donate. We hold the distant, statistical life to a far lower standard than the visible, proximate one. This is not a logical position.
The arithmetic of giving
The economic case for charitable giving is compelling in a way that most ethical arguments are not, because it is concrete.
Marginal utility diminishes as wealth increases. The difference between having nothing and having $100 is enormous. The difference between having $100,000 and having $100,100 is trivial. This is not a philosophical abstraction — it is a description of how human welfare actually works.
This creates what might be called altruistic arbitrage. A dollar transferred from someone in the wealthiest quintile of the global population to someone in the poorest quintile produces, by any reasonable welfare measure, dramatically more value than that dollar provided in its original location. The maths is not close. Someone earning a median income in a developed country sits — by global standards — in the top one to five percent. Relative to the global poor, the purchasing power of that income is several orders of magnitude larger.
The effective altruism movement has quantified this with varying degrees of rigor. The numbers shift, but the order of magnitude does not. Distributing malaria bednets costs roughly $3,000–$5,000 per life saved — possibly less. A donation of that size from someone in a wealthy country is, for them, a serious but not life-altering expense. For the recipient, it is everything.
The five questions
Knowing that giving can do significant good is not sufficient. Not all charitable spending produces equivalent outcomes. William MacAskill, in Doing Good Better, proposes five questions for evaluating where effort and resources are best directed:
How many people benefit, and by how much? Some causes help many people moderately; others help fewer people dramatically. Both matter. The product is what you’re optimizing.
Is this the most effective intervention available? Within any cause area, effectiveness varies enormously. Certain health interventions are many times more cost-effective than others targeting the same disease. Counterfactual comparison within a cause area is as important as the choice of cause itself.
Is this neglected? The most important causes are not necessarily the most visible ones. Funding and attention cluster around emotionally salient problems — often not the problems where additional resources produce the most impact. A neglected cause may be far more tractable for additional resources than an already heavily-funded one.
What would have happened otherwise? This is counterfactual thinking, and it is difficult. If you volunteer for a role that any of a hundred qualified people would have taken, your counterfactual impact is near zero — the outcome would have been the same without you. Counterfactual impact is the actual measure of what your contribution achieved.
What are the chances of success, and how good would success be? This applies especially to systemic interventions — policy change, scientific research, institutional reform. Expected value calculations apply: a low probability of a very large outcome may outweigh a high probability of a small one.
Expanding the circle
Who counts as a moral patient changes over time. Historically, the moral circle has been tribal — family, then community, then nation, then race. Each expansion of that circle has, eventually, been recognized as correct. The extension of moral concern to people we cannot see, who live in countries we will never visit, is a continuation of that pattern — not a departure from it.
The fact that distance and visibility reduce our intuitive concern for distant suffering is a fact about our psychology, not a fact about the world. The child drowning in the pond is not morally different from the child dying of a preventable disease in a country we cannot name. The only difference is proximity.
None of this requires extreme sacrifice. The case for giving does not depend on giving until it hurts. It depends on taking seriously the arithmetic: that a portion of one’s income, given well, can produce significantly more welfare than it would have provided in its original location. The mechanism is simple. The only things standing in the way of acting on it are the cognitive biases that treat visible suffering as more real than invisible suffering, and the assumption that doing nothing is a neutral choice.
It is not.