The drowning child: Peter Singer's argument that changed how we think about giving

A shallow reflecting pool in a public park, sunlit and serene — the kind of place Singer's thought experiment begins Photo: Unsplash — freely available for use

In 1971, a cyclone and military campaign sent ten million Bengali refugees into crisis. Peter Singer, a young moral philosopher at Monash University, wrote a paper asking why the Western world was largely doing nothing.

The paper was called "Famine, Affluence, and Morality." It was published in 1972 in an obscure academic journal. It opened with a few dry sentences about the Bengal famine. Then it posed a question that moral philosophy has been trying to answer ever since.

The pond

Here is Singer's thought experiment, quoted almost exactly as he wrote it:

"If I am walking past a shallow pond and see a child drowning in it, I ought to wade in and pull the child out. This will mean getting my clothes muddy, but this is insignificant, while the death of the child would presumably be a very bad thing."

Nobody disputes this. Of course you pull the child out. Ruining a pair of trousers is nothing compared to a child's life. If you walked past and did nothing, we would not merely think you were unkind — we would think you were a moral monster.

Singer then asks: what is the morally relevant difference between that pond and a child dying from a preventable disease on the other side of the world?

The child overseas is just as real. The suffering is just as bad. You have the ability to prevent it — not by wading into a pond, but by donating money you'd otherwise spend on things far less important than a human life. The cost to you is real but small, just like the muddy trousers.

Singer's conclusion is stark: if there is no morally relevant difference, then we are morally obligated to give — and to keep giving — until we reach the point where giving more would cost us something of comparable moral significance.

The principle, stated plainly

Singer offers two versions of his core argument.

The weak version (which he says is uncontroversial):

If it is in our power to prevent something bad from happening, without thereby sacrificing anything of comparable moral importance, we ought, morally, to do it.

The strong version (which he thinks is the correct one):

We ought to prevent bad things from happening unless doing so requires sacrificing something of comparable moral significance.

The difference is subtle but important. The weak version says: don't sacrifice anything morally significant. The strong version says: don't stop until you're giving up something roughly as important as what you're preventing. Under the strong version, buying a new phone when that money could fund life-saving medical treatment is not morally neutral. It is, in Singer's framework, morally equivalent to walking past the drowning child.

Four premises that drive the argument

Singer's argument rests on four claims, each of which seems defensible in isolation. Together they lead somewhere uncomfortable.

P1. Suffering and death from poverty are bad. This is not contested. Preventable childhood death is a harm.

P2. Our donations can prevent some of that suffering. Global health interventions — oral rehydration therapy, malaria nets, vitamin A supplementation, deworming — save lives at low cost. The evidence is robust.

P3. Typical consumer spending is not morally significant. The things we spend money on — coffee, streaming subscriptions, a nicer pair of shoes — are not things the loss of which would constitute a serious sacrifice.

P4. Distance and familiarity do not change our moral obligations. Singer is explicit: "It makes no moral difference whether the person I can help is a neighbor's child ten yards away or a Bengali whose name I shall never know, ten thousand miles away."

If you accept all four premises, you have accepted Singer's conclusion.

The objections — and Singer's replies

The paper provoked fifty years of counterargument. Here are the strongest objections and how the debate has played out.

"It's too demanding"

The most common objection: Singer's argument requires too much of ordinary people. If we must give until giving more would harm us comparably, most of us would need to radically restructure our lives.

Singer largely accepts this. He thinks the conclusion is demanding and that most people won't reach it — but that doesn't make it wrong. Our intuitions about what's reasonable are shaped by what our peers do, not by what morality actually requires. "The fact that we give less than we should," he has written, "is not an argument for thinking we should give less."

"Distance matters"

Some philosophers argue that proximity creates special obligations — we are more responsible for people near us, and that's morally legitimate, not mere bias.

Frances Kamm and others have argued that the pond case is an anomalous emergency where you happen to be the only available rescuer. Ongoing global poverty is structurally different: there are many potential rescuers, institutional solutions exist, and individual donations are a drop in the ocean.

Travis Timmerman made a pointed version of this argument in 2015: "Sometimes There Is Nothing Wrong with Letting a Child Drown." His point wasn't that drowning children don't matter, but that when millions of people could each contribute a small amount, the moral mathematics change. No single individual is as uniquely positioned as the pond walker.

Singer's response: the bystander objection actually makes things worse, not better. If there are a hundred people who could each easily save the drowning child and none of them do, each of them is still culpable. The presence of other potential givers doesn't diminish your obligation.

"We have special obligations to family and community"

John Arthur (1982) argued that Singer ignores the moral weight of our entitlements and special relationships. We are allowed to prioritise our own projects, our families, our communities. Morality that demands we treat every stranger's life as equal to our child's life is asking us to become something other than fully human.

Kwame Anthony Appiah added a cultural dimension: people are embedded in communities with particular traditions and loyalties, and a morality that ignores this will not and cannot command widespread adherence.

Singer's position is that special obligations are real but limited. You may spend somewhat more on your children's education than on strangers' survival, but the asymmetry has limits. At some point — when the stakes are life itself — the preference for proximity becomes indefensible partiality.

What Singer wasn't arguing

Two points are frequently misunderstood.

He was not saying all charities are equally effective. The 1972 paper predates the modern evidence-based charity movement by decades. Singer was establishing that we have an obligation to give, not how to give most effectively. He became a founding figure of effective altruism precisely because he later combined his moral argument with rigorous analysis of where money does the most good.

He was not demanding sainthood. Singer has always acknowledged that most people will give less than his argument strictly requires — and that giving something is far better than giving nothing. The paper's purpose was to shift the moral framing, not to make donors feel crushed by a standard no one meets.

Why this argument produced effective altruism

William MacAskill encountered Singer's paper as an undergraduate at Oxford. He found he couldn't answer it. He went on to co-found the effective altruism movement — a community of researchers, donors, and practitioners dedicated to figuring out where to direct resources for the greatest possible impact.

Toby Ord, a philosopher at Oxford, read Singer and took out a spreadsheet. He calculated that as a philosophy lecturer he would earn around £1.5 million over his career. He calculated what that could do in global health terms. He pledged to give everything he earned above a modest threshold for the rest of his working life, and co-founded Giving What We Can — the pledge organisation that has now channelled over a billion dollars to effective causes.

These are not cult members. They are people who read a 14-page philosophy paper and found they couldn't construct a satisfying objection.

Why knowing this changes almost nothing

Here is the strange part. Most people who encounter Singer's argument cannot refute it. They sit with it, turn it over, try a few objections — and find they don't hold. And then they go back to their lives largely unchanged.

This is not hypocrisy, exactly. It's something more interesting, and worth understanding.

We feel for faces, not numbers. Research by psychologist Paul Slovic found that when people are shown a photograph of a single child in need, donations are generous. Add a second child and donations go down. Present statistics about millions suffering and emotional response collapses almost entirely. Singer's argument works at the level of logic, but human generosity is driven by something else — vivid, specific, emotionally immediate images. The drowning child in the pond works as a thought experiment because it has all of those qualities. Real poverty doesn't.

We can't feel the scale. Related to this: people donate roughly the same amount to save 2,000 animals from an oil spill as to save 200,000. Our moral instincts simply don't scale with numbers. This is a genuine problem for Singer's framework, which is built on adding up welfare across millions of people. If your gut can't distinguish between 100 lives and 100,000 lives, reason alone is a very thin lever.

Knowing and doing are different things. Aristotle had a word for this: akrasia — the condition of knowing the right thing and doing something else anyway. You might accept Singer's argument entirely at 11am and still spend money on something unnecessary at noon. Beliefs and behaviours are run by different parts of us. Philosophy can change your beliefs. It is much less reliable at changing the automatic, habitual, everyday patterns that govern how you actually spend your life.

We follow what people around us do. What feels like a "normal" amount to give is almost entirely set by what your peers give. If the people around you give 1% of their income, giving 2% feels generous. Giving 10% feels strange. Singer's argument does nothing to shift those reference points — and social norms turn out to be far more powerful than philosophical conviction in shaping actual behaviour.

When the argument threatens us, we find reasons to dismiss it. When people encounter a valid argument that demands uncomfortable change, they rarely change. Instead, they tend to find reasons the argument is wrong — "charity creates dependency," "systemic change matters more," "the money doesn't actually reach people." Some of these objections have genuine merit. But they tend to appear after the discomfort, not as independent conclusions. The conclusion is driving the objection, not the other way around.

The grief at the edge of the argument

There is something philosopher Bernard Williams noticed that goes deeper than all of this.

Williams argued that a morality like Singer's — one that asks you to weigh every purchase against lives you could save — doesn't just demand sacrifice. It demands that you become a different person. We each have commitments, relationships, and projects that give our lives shape and meaning. A framework that asks you to treat all of those as potentially less important than a stranger's survival isn't refining your character. It is, in some sense, dissolving it.

People who take Singer's argument seriously — who really sit with it and can't answer it — often describe something that feels like grief. Not guilt exactly, but a kind of loss: the sense that fully accepting the argument would require giving up a version of themselves they value. The career they chose for its meaning rather than its salary. The trip they're saving for. The way they've imagined their future.

And this is where point five and point six meet. Most people, when they reach that edge, don't leap. They step back. They find — consciously or not — a reason the argument doesn't quite apply to them, or doesn't apply yet, or applies but only up to a point they haven't reached. Not because they are bad people. But because the alternative is genuinely frightening: a moral obligation with no clear floor, pointing at a life that looks almost nothing like the one they planned.

Singer himself is sympathetic to this. He has never claimed the argument is easy to live by. What he resists is the conclusion that difficulty makes it wrong. The child in the pond doesn't stop drowning because wading in is inconvenient.

There is one thing Singer's argument takes for granted throughout: that we know what suffering is. That it is recognisable, comparable across people and places, and that preventing it is straightforwardly good. That assumption does a lot of work — and it turns out to be harder, stranger, and more surprising than the argument lets on. What suffering actually is, whether a shrimp experiences it, whether childhood trauma is its most hidden and devastating form, and whether reducing it even works the way we think — these are questions the pond doesn't answer.

Continue reading: What is suffering, actually? →