Peter Singer argues that we’re obligated to donate our entire expendable income to aid organizations. One premiss of his argument is "If it is in your power to prevent something bad from happening, without sacrificing anything nearly as important, it is wrong not to do so." Singer defends this by noting that commonsense morality requires us to save a child we find drowning in a shallow pond. I argue that Singer’s Drowning Child thought experiment doesn’t justify this premiss. I offer my own Drowning Children thought experiment, which should reveal that commonsense morality entails that premiss two is actually false.
Peter Singer’s Famine Affluence and Morality is undoubtedly one of the most influential and widely read pieces of contemporary philosophy. Yet, the majority of philosophers (including ethicists) reject Singer’s conclusion that we are morally required to donate to aid agencies whenever we can do so without sacrificing anything nearly as important as the good that our donations could bring about. Many ignore Singer’s argument simply because they believe morality would just be too demanding if it required people in affluent nations to donate significant sums of money to charity. Of course, merely rejecting Singer’s conclusion because it seems absurd does not constitute a refutation of Singer’s argument. More importantly, this standard demandingness objection is a particularly inappropriate dialectical move because Singer provides a valid argument for his (demanding) conclusion and, crucially, the argument only consists of ethical premisses that Singer takes his typical readers to already accept. Singer formulates his argument as follows.
Suffering and death from lack of food, shelter and medical care are bad.
If it is in your power to prevent something bad from happening, without sacrificing anything nearly as important, it is wrong not to do so.
By donating to aid agencies, you can prevent suffering and death from lack of food, shelter and medical care, without sacrificing anything nearly as important.
If it is not true that typical readers’ existing ethical commitments entail that they accept premisses one and two, then they should be able to say which premiss(es) they reject and why. Those who believe that Singer’s conclusion is too demanding will need to reject premiss two. This requires addressing Singer’s infamous Drowning Child thought experiment, which elicits a common response that Singer believes demonstrates that his readers are already committed to the truth of premiss two. As such, Singer purports to demonstrate that the ethical commitments his typical readers already accept are demanding enough to require them to donate a substantial portion of their expendable income to aid organizations. A dialectically appropriate demandingness objection would have to demonstrate why this is not the case. I aim to provide such an objection, in part, by demonstrating that the inference from the near universal intuition that we are obligated to rescue the child in Drowning Child to the truth of premiss two is unwarranted.
To be fair, although many philosophers do not attempt to directly engage with the argument Singer provides, many do. Those that have commonly reject premiss three (e.g. Fagelson 2009; Gomberg 2002; Schmidtz 2000: 684–9) and moral libertarians must reject premiss two (e.g. Narveson 2003; Pogge 2008). The existing arguments given in favour of rejecting premiss two appeal to the highly controversial claim that we only have a negative duty to not inflict harm on others and no positive duty to help others, even when we can do so at little or no cost to ourselves. Such views about duty strike me as highly dubious. More importantly, even if they turn out to be true, Singer might still be able to successfully argue for his conclusion since there are many ways people in affluent nations do causally contribute to the suffering of those in extreme poverty (Singer 1999; Pogge 2001; Pogge 2008).
I aim to do something different in this paper. I will grant Singer the truth of premisses one and three and that we can have positive duties to help others, even when we did not, in any way, causally contribute to the suffering we are obligated to end. In spite of this, I argue that Singer has not provided sufficient justification for the truth of premiss two. I do this by proposing a thought experiment analogous to Singer’s Drowning Child case, which plays the dual role of blocking the inference from the common intuition in Drowning Child to the truth of premiss two and gives us positive reason to reject premiss two.
The paper proceeds as follows. In the next section, I review Singer’s Drowning Child thought experiment, which he appeals to in an attempt to establish that his typical readers already accept the truth of the second premiss. In the third section, I offer my own Drowning Children thought experiment, which is a better analogy of the situation between people in affluent nations and those in extreme poverty than Singer’s Drowning Child. People’s intuitions in the case I provide should reveal that they think, upon reflection, premiss two is actually false. In the fourth section, I consider and rebut two objections to my argument before concluding with a heuristic for determining how much people in affluent nations are obligated to donate to aid agencies.
2. Singer’s case
In his (1972), Singer defends the claim that
(231). He argues for this by analogy, writing ‘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’ (Singer 1972: 231). He expands this defence into a thought experiment in The Life You Can Save, which I will refer to as Drowning Child:
‘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’
Drowning Child: On your way to work, you pass a small pond. On hot days, children sometimes play in the pond, which is only about knee-deep. The weather’s cool today, though, and the hour is early, so you are surprised to see a child splashing about in the pond. As you get closer, you see that it is a very young child, just a toddler, who is flailing about, unable to stay upright or walk out of the pond. You look for the parents or babysitter, but there is no one else around. The child is unable to keep his head above the water for more than a few seconds at a time. If you don’t wade in and pull him out, he seems likely to drown. Wading in is easy and safe, but you will ruin the new shoes you bought only a few days ago, and get your suit wet and muddy (Singer 2009: 3).
It is supposed to be obvious that you are obligated to wade in the pond and save the child, even if doing so ruins your new shoes and suit. The best explanation of why you are obligated to save the child, Singer contends, is precisely because premiss two is true. The only additional defence provided of premiss two is a rebuttal of two objections. First, Singer claims that our proximity, in itself, to those in need is of no moral relevance (Singer 1972: 231–2). Although some deny this assumption (e.g. Kamm 1999, 2007), I will happily grant it for the sake of argument. Second, the fact that millions of others in affluent nations are also in a position to prevent children from dying in extreme poverty, but do not, does not diminish the extent to which we are obligated to provide aid. Singer rhetorically asks
‘Should I consider that I am less obliged to pull the drowning child out of the pond if on looking around I see other people, no further away than I am, who have also noticed the child but are doing nothing?’
(Singer 1972: 233). Although some also deny this assumption (e.g. McKinsey 1981), I again grant Singer as much.1
‘One has only to ask this question to see the absurdity of the view that numbers lessen obligation’
Although premiss two is admittedly prima facie plausible, the problem with it is that it is deceptively demanding of us as moral agents, far too demanding to be intuitively compelling. If true, we are obligated to donate our money to aid agencies whenever we can do so without sacrificing anything nearly as important as a child’s life, the consequence of which Singer makes explicitly clear.
Yet if we were to take [premiss two] seriously, our lives would be changed dramatically. For while the cost of saving one child’s life by a donation to an aid organization may not be great, after you have donated that sum, there remain more children in need of saving, each one of whom can be saved at a relatively small additional cost. Suppose you have just sent $200 to an agency that can, for that amount, save the life of a child in a developing country who would otherwise have died. You’ve done something really good, and all it has cost you is the price of some new clothes you didn’t really need anyway. But don’t celebrate your good deed by opening a bottle of champagne, or even going to a movie. The cost of that bottle or movie, added to what you could save by cutting down on a few other extravagances, would save the life of another child … So you must keep cutting back on unnecessary spending, and donating what you save, until you have reduced yourself to the point where if you give any more, you will be sacrificing something nearly as important as a child’s life—like giving so much that you can no longer afford to give your children an adequate education (Singer 2009: 18).
Perhaps premiss two is true, but a proposition with such strong counter-intuitive implications requires a strong defence, one that gives us reason to think that certain ordinary moral intuitions are radically misguided. Singer believes he has provided such a defence with Drowning Child. Aren’t we morally obligated to sacrifice our new clothes to save the child because we are obligated to prevent something bad from happening whenever we can do so without sacrificing anything nearly as important? The short answer is ‘No.’ Here’s why. Although Singer’s description of Drowning Child is ahistorical, the implicit assumption is that Drowning Child is an anomalous event. People almost never find themselves in the situation Singer describes, so when they consider their obligations in Drowning Child, they implicitly assume that they have not frequently sacrificed their new clothes to save children in the past and will not need to do so frequently in the future.
Giving to aid organizations is, in this respect, unlike Drowning Child. Every individual in an affluent nation, so long as they have some expendable income, will always be in a position to save the lives of people living in extreme poverty by donating said income. It may be quite clear that one has a moral obligation to sacrifice $200 worth of new clothing a single time to prevent a child from drowning. It is much less clear that one is morally obligated to spend one’s entire life making repeated $200 sacrifices to constantly prevent children from drowning. So, we may be obligated to save the child in Drowning Child, but still be disposed to believe that premiss two is false. I will expand on this asymmetry in the next section by providing an altered version of Singer’s thought experiment that more closely resembles the position those in affluent nations are in with respect to providing aid to those in extreme poverty. I suspect that most people’s intuitions in such a case will show that they reject premiss two of Singer’s argument.
3. The most relevantly analogous case
People almost universally have the intuition that we are morally obligated to rescue the child in Drowning Child, but are not morally obligated to donate all their expendable income to aid agencies. Singer attempts to explain away this intuition as a mere psychological difference, a difference that results from our evolutionary history and socialization and not a moral difference (Singer 1972: 232–3; Singer 2009: 45–62). To a certain extent, I think Singer is successful. I grant that, ceteris paribus, there is no moral difference between one’s obligation to save the drowning child and one’s obligation to donate to aid organizations.2 However, there is a moral difference between the sacrifice required to save the child in Drowning Child (as it is imagined) and the sacrifice Singer believes people in affluent nations are required to make in order to donate the supposed obligatory amount to aid organizations.
This moral difference is easily overlooked because Singer’s Drowning Child thought experiment is, in a crucial way, under-described. Once the necessary details are filled in, its inability to support premiss two will be made clear. My following Drowning Children case is not under-described and gives us reason to believe that there are times at which it is morally permissible to not prevent something bad from happening, even when one can do so at a comparably insignificant personal cost.
Drowning Children: Unlucky Lisa gets a call from her 24-hr bank telling her that hackers have accessed her account and are taking $200 out of it every 5 min until Lisa shows up in person to put a hold on her account. Due to some legal loophole, the bank is not required to reimburse Lisa for any of the money she may lose nor will they. In fact, if her account is overdrawn, the bank will seize as much of her assets as is needed to pay the debt created by the hackers.
Fortunately, for Lisa, the bank is just across the street from her work and she can get there in fewer than 5 min. She was even about to walk to the bank as part of her daily routine. On her way, Lisa notices a vast space of land covered with hundreds of newly formed shallow ponds, each of which contains a small child who will drown unless someone pulls them to safety.3 Lisa knows that for each child she rescues, an extra child will live who would have otherwise died. Now, it would take Lisa approximately 5 min to pull each child to safety and, in what can only be the most horrifically surreal day of her life, Lisa has to decide how many children to rescue before entering the bank. Once she enters the bank, all the children who have not yet been rescued will drown.
Things only get worse for poor Lisa. For the remainder of her life, the hackers repeat their actions on a daily basis and, every day, the ponds adjacent to Lisa’s bank are filled with drowning children.
The truth of premiss two would entail that Lisa is obligated to rescue children until almost all of her money and assets are gone. It might permit her to close her account before she is unable to rent a studio apartment and eat a healthy diet. However, it would require her to give up her house, her car, her books, her art and anything else not nearly as important as a child’s life. That might not seem so counterintuitive if Lisa has to make this monumental sacrifice a single time. But, and here’s the rub, premiss two would also prohibit Lisa from ever rebuilding her life. For every day Lisa earns money, she is forced to choose between saving children and letting the hackers steal from her. Lisa would only be permitted to go to the bank each day in time to maintain the things nearly as important as a child’s life, which I take to be the basic necessities Lisa needs to lead a healthy life.4
I propose that it’s a viable option that morality permits Lisa to, at least on 1 day over the course of her entire life, stop the hackers in time to enjoy some good that is not nearly as important as a child’s life. Maybe Lisa wants to experience theatre one last time before she spends the remainder of her days pulling children from shallow ponds and stopping hackers. Given the totality of the sacrifice Lisa is making, morality intuitively permits Lisa to indulge in theatre at least one time in, let’s say, the remaining 80 years of her life. In fact, commonsense morality should permit Lisa to indulge in these comparably morally insignificant goods a non-trivial number of times, though a single instance is all that is required to demonstrate that premiss two is false and, consequently, Singer’s argument is unsound.
I have purposefully not made a suggestion as to how many (if any) children Lisa is obligated to rescue. I did so to make my argument as neutral as possible, as I want it to be consistent with any normative ethical view ranging from moral libertarianism to a view that only permits Lisa to indulge in a comparably insignificant good a single time. For what it’s worth, I am inclined to hold that Lisa is obligated to rescue a great many children, though significantly fewer than is required by premiss two. However, no part of my argument hinges on this claim.
I now consider, and rebut, two objections to my argument.
First Objection: Drowning Children really does support the truth of premiss two because Lisa is intuitively obligated to make the sacrifice it requires.
To be sure, impartial consequentialists will not be convinced that premiss two is false and their theory-laden intuitions may conflict with the commonsense intuition in my Drowning Children case. But my aim in this paper is not to provide a dialectically effective argument against this particular kind of consequentialism. Rather, my main goal is to demonstrate that Drowning Child does not justify premiss two in the way Singer believes it does. Specifically, Singer’s argument fails to demonstrate that commonsense assumptions about morality require us to donate as much to aid agencies as his argument entails. Singer’s goal was to argue for his conclusion without assuming the truth of impartial consequentialism (Singer 1999: 302–3). Unfortunately, it fails to do just that. A significantly stronger defence of premiss two is required if Singer is going to be successful in showing people like me, a commonsense consequentialist, that morality is as demanding as his argument entails.
Second Objection: Our moral intuitions are not reliable.
One might resist my argument by denying that the intuitive judgments about Drowning Children are reliable. In fact, Singer famously rejects the reliability of intuitions about first-order normative judgments5, so I suspect that he too will think intuitive responses to Drowning Children carry little argumentative weight (Singer 1974: 516; Singer 2005; Singer and Lazari-Radek 2014: 67). Because Singer rejects the reliability of first-order normative intuitions, it might seem hypocritical of him to appeal to Drowning Child as a justification for premiss two, but it is not. Recall that Singer’s goal was to demonstrate to non-impartial consequentialists that, by their own lights, they are obligated to donate a significant portion of their income to aid organizations (Singer 1999: 315–6, 505). That is, Drowning Child was invoked to make salient to people moral beliefs they supposedly already had, including the truth of premiss two. That’s fair enough. The problem, though, is that I may use Drowning Children in the same way as Singer uses Drowning Child. I invoke Drowning Children to show that people are not committed, by their own lights, to the truth of premiss two. My argument for that claim need not assume that our first-order normative judgments are reliable, although I don’t rule out that possibility either. More generally, one cannot call into question the reliability of our judgments in Drowning Children without also calling into question the reliability of our judgments in Drowning Child. Either our judgments in both cases are reliable and, consequently, we have reason to reject premiss two or neither is reliable and, consequently, Singer has provided no justification for premiss two.
If my argument is successful, it should provide compelling reason to accept the following two distinct claims. First, Singer’s Drowning Child case actually fails to justify the truth of premiss two by his readers’ own lights. People’s intuition that they are obligated to save the child in Singer’s case (as they imagine it) is consistent with premiss two being false. Second, my Drowning Children case actually provides positive reason to reject premiss two. These two claims are independent of one another and I take each to be significant in their own right. To sum up, the intuitive pull of premiss two is more apparent than real. Few moral truths may seem more obvious than that one is obligated to sacrifice $200 to save a child’s life at least once. But it’s far from obvious that one is obligated, for his or her entire life, to constantly sacrifice everything comparably insignificant to a child’s life. The truth of the second premiss hinges on the truth of this latter claim, not the former. How much are we obligated to donate to aid organizations? I am not sure exactly, but it should be the same amount we would be obligated to sacrifice were we to find ourselves in Lisa’s position.6