Rethinking Singer: Toward a valid argument for helping the global poor

In 1971 Peter Singer wrote Famine, Affluence, and Morality. It is arguably one of the most influential and widely read pieces of contemporary philosophy. There, he famously argues that we are required to donate to aid agencies. Or, rather, more accurately, he sketches the first two versions of his now famous and influential common-sense morality argument about donating aid to those living in extreme poverty. Singer's argument has influenced many people the world over, and later versions of it helped kickstart what is known as the effective altruism movement (Singer, 2015). In the more than fifty years that have since passed, versions of Singer's argument have appeared everywhere. As Singer (2019, p. xvii) himself notes:

I have been thinking and writing for more than 40 years about how we should respond to hunger and poverty. I have presented this book's argument to thousands of students in my university classes and in my online course on effective giving, and to countless others in newspapers, magazines, a TED talk, podcasts, and television programs.

Quite possibly the most famous, influential, and most widely cited instance of Singer's argument appears in the first, 2009, edition of The Life You Can Save (Singer, 2009, pp. 15–16). It reads:

While there has been no shortage of commentators over the past fifty years who deny the soundness of Singer's argument, all of them appear to be focused on rejecting one of these three premises. Most opponents reject premise two, but there are also many critics who deny premise three. ⁽¹⁾ Interestingly, and perhaps somewhat surprisingly, no one appears to have directly rejected Singer's argument on the grounds that it is invalid. In fact, to the contrary, it is fairly common to see claims affirming its validity. ⁽²⁾ What's more, oftentimes these claims are made by people who criticise Singer's argument! ⁽³⁾ Here, for example, is what Travis Timmerman (2015, p.204) wrote in a recent article that then went on to reject premise two of Argument 1:

...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.

In this article, I intend to show that despite these commonplace claims, Singer's supposedly common-sense morality arguments to the conclusion that we are doing something morally wrong, if we do not donate to aid agencies, are invalid. That is, he has not offered us good reason to think that we would be doing something morally wrong, if we do not donate to aid agencies that prevent the death and suffering of the less fortunate. ⁽⁴⁾ Now, to be fair, Singer does, in various places, offer common-sense morality arguments related to helping the world's less fortunate that are valid, but, as I will also show, these arguments do not have as their conclusion the claim that we are doing something morally wrong if we do not donate to aid agencies. So then, ultimately, one of the important claims defended in this article is this: in over 50 years, Singer has not explicitly offered the world a valid common-sense morality argument to the conclusion that we have a moral obligation to donate to aid agencies, or one to the conclusion that we are doing something morally wrong if we fail to donate to aid agencies that prevent the death and suffering of the less fortunate. And if we take into account those arguments that are valid but that support different but related conclusions, then the claim defended in this article can be widened to this: Singer's common-sense morality arguments to the conclusions that we have a moral obligation to donate to aid agencies and that we are doing something morally wrong if we fail to donate to aid agencies are either invalid or inadequate.

Now, of course, all of this might be uncharitable. While Singer, and indeed many of his critics, may not have been as cautious as we might have wanted them to be when he, and they, explicitly offer one of these arguments, it might be the case that we can piece together a charitable and valid version of it from the ideas contained within the totality of his work on global poverty and our obligations to those in need. So, then, in an effort to be charitable, I will offer what appears to be a valid instance of Singer's argument to the conclusions that we have a moral obligation to donate to aid agencies and that we are doing something morally wrong if we fail to donate to aid agencies that prevent the death and suffering of the less fortunate. The argument that I offer is, I think, fair, valid, and respects the totality of Singer's work on helping the world's poor. This is the second aim of this paper.

As I will highlight though, the move to protect Singer's argument from the charge of being invalid comes with some costs, and, as I will motivate through a critique of this new argument, these costs are unacceptable. Ultimately, this discussion leads me to the following conclusion: while we have an obligation to help those less fortunate than ourselves, we do not necessarily have an obligation to donate to aid agencies to help them, nor would we do something wrong if we failed to do so – provided that we fulfilled our obligation to help. The defence of this position is the third aim of this paper. ⁽⁵⁾

This article will proceed as follows. I will begin, in the next section, by highlighting that the most famous and influential instance of Singer's argument (which appeared above) is invalid. I will then run through every distinct instance of the argument Singer has offered since he first sketched it back in 1971 and show that each one is either invalid or has, as its conclusion, a much broader claim than one directed at donating to aid agencies. Then, in the third section of this article, I will offer a version of Singer's argument that appears to be valid and that is in keeping with the totality of his work on global poverty. Those familiar with Singer, and his body of work, will easily identify that this argument is utilitarian in spirit. I will also critique this argument in the third section. Then, out of that critique, I will motivate and conclude, in the fourth and final section of this article, that while we have an obligation to help those less fortunate than ourselves, we do not necessarily have an obligation to donate money, nor would we do something wrong if we failed to donate money—provided that we fulfilled our obligation to help.

It is typical to say that an argument is valid if and only if it is not possible for all of its premises to be true and its conclusion false. An argument is typically said to be sound if and only if it is valid and all of its premises are true or acceptable. To say that an argument is a common-sense morality argument is just to say that it is a moral argument that is ground in everyday, widely shared, common-sense beliefs. They might not be beliefs that we are conscious of, but they are typically thought to be ones that, at least when made explicit, most people would readily assent to. Understanding and being persuaded by a common-sense morality argument is not supposed to be the kind of thing that requires those who engage with it to understand any moral theory. They are supposed to be appealing to everyone and theory neutral. ⁽⁶⁾ Here, again, is the most famous and influential instance of Singer's aid argument.

It is taken to be a valid, common-sense morality argument. ⁽⁷⁾ This argument, however, is invalid. There are at least two ways in which this argument is invalid. One of them is trivial, the other less so. The trivial way is this: there is nothing contained in the argument which says that suffering and death from a lack of food, shelter, and medical care actually exists. So, in a world in which it does not, one would hardly do something wrong if they did not donate to an aid agency. This, evidently, is a trivial objection. The actual world has a large and intolerable amount of death and suffering in it that results from people having a lack of basics such as food, shelter, and medical care, and many of us, including Singer, know this to be true. So, a more charitable and accurate version of this argument containing this unstated premise (U1) is:

Unfortunately, though, this amended argument is not valid either. Why is that? Well, because donating to aid agencies is but one thing that is within our power (or, at least, within the power of many of us) that we could do to prevent something bad from happening, without sacrificing anything nearly as important. What are some examples of this? Here is a partial list of things we could do, either in isolation or combination: we could donate our time and efforts rather than our money to an aid agency. We could literally go and directly help those who need assistance in ways that respect their needs and wishes (thereby bypassing any aid agency). We could start our own charity devoted to these issues. We could work for or start our own meta-charity devoted to these issues. We could likely donate some of our organs, or blood, or marrow, for example, to help prevent some of these bad things from happening. We could commit ourselves to advocacy work, and aggressively work on, and push for, say, economic or foreign relations policies that will prevent deaths and reduce suffering among the less fortunate. We could aggressively and effectively work on addressing climate change, since doing so will prevent a great deal of death and suffering that results from a lack of food, shelter, and medical care. Several further examples could be given. ⁽⁸⁾

This instance of Singer's argument is invalid. You might say: So what? He has been offering versions of this argument for a very long time. While this instance is invalid, that is not to say that he has not offered a valid instance of it elsewhere, even though this instance happens to be the most famous, influential, and widely cited. Fair enough. Let us run through then the collection of distinct arguments he has offered on this topic, over the past fifty plus years. And let us begin with the version he offers in the updated, ten-year anniversary, second, 2019 edition of The Life You Can Save. There, the argument reads as follows:

The argument is basically the same as the last one, but it now includes a qualification about effectiveness. The third premise now refers only to donations to effective aid agencies. This qualification presumably was added to deal with critics who denied the truth of premise three in Argument 2 on the grounds that whether you can prevent suffering and death by donating to an aid agency depends on whether it is effective. As some commentators rightly note, and as Singer would likely agree, there are certainly some aid agencies that you can donate to that would not prevent bad things from happening (Singer, 2019, pp. 59–60). But now, even with that qualification, and the additional, unstated premise that such suffering and death actually exist in the world, it is still the case that there are other things that are within our power that we can do that will prevent bad things from happening. So, this argument is invalid, for the same reason as before.

Let us journey back then, and consider other, distinct versions of Singer's common-sense morality argument. We will proceed in chronological order, beginning with the arguments (or, rather argument outlines) that appear in Famine, Affluence, and Morality. While Singer does not offer a neat standard form version of his argument in that piece, it is pretty clear that he offers two versions of it in the paper: one that relies on a strong version of the second premise, and a weaker one, with a weaker second premise. Here is the first, stronger one:

There appear to be two unstated premises (U1 and U2) in this version of the argument, which Singer clearly seems to endorse, despite not noting them as premises. These are ones that were noted before. First, that there actually are people who are suffering and dying from a lack of food, shelter and medical care, and second, that by donating to aid agencies, you can prevent suffering and death from lack of food, shelter, and medical care, without thereby sacrificing anything of comparable moral importance. Also, in this version of the argument, the move to the (unstated) conclusion (UC2) that if we do not donate to aid agencies we are doing something wrong, presumably must pass through an intermediate and unstated conclusion (UC1) that we ought, morally, to donate to aid agencies—given the way premise two is phrased.

This version of the argument, however, is also invalid. It does not follow from premises 1–U2 that we ought, morally, to donate to aid agencies. As in earlier cases, the conclusion does not follow because there are many other actions within our power that can prevent death and suffering. Moving onto the second, weaker version, which also appears in Famine, Affluence, and Morality, we see that the same criticism applies. Here is the weaker version of the argument:

The weaker version of premise 2, and the corresponding weaker version of U2, do nothing to eliminate the possibility that there are many other things that are within our power that can prevent death and suffering.

The next time Singer offers a distinct version of the argument is in 1979. It appears in the first edition of his book, Practical Ethics. In fact, the same version of the argument appears in all subsequent editions. ⁽⁹⁾ Here is that argument, as it appears in Singer (2011c, p. 200): ⁽¹⁰⁾

Interestingly, once we add in the claim that some people are actually experiencing extreme poverty, the argument is valid. But notice that even with the additional unstated premise, the conclusion makes no reference to donating to aid agencies. So then, even if this argument is valid, it is not an argument to the conclusion that we have a moral obligation to donate to aid agencies, or that we would be doing something morally wrong if we did not.

Here, now, appears to be the last, distinct instance of Singer's common-sense morality argument that appears in the literature to date. This one appears on p. 147 of a 2011 book chapter, entitled Changing Values for a Just and Sustainable World. It reads:

This instance of the argument is very clearly invalid. Even if the premises of this argument are true, it does not follow that we ought to prevent the deaths of children affected by global poverty. For one thing, there is nothing in the premises linking the deaths of children to a lack of food, shelter and medical care, and even if there was, there would still not be a link between the lack of food, shelter and medical care causing their deaths and this being the result of their poverty. But notice that even if we update this argument to prevent these criticisms, at best, if valid, this argument merely establishes that we ought to prevent the deaths of children affected by global poverty. It does not claim that we have a moral obligation to donate to aid agencies or that we would be doing something morally wrong if we did not.

That concludes this historical journey through the collection of common-sense morality arguments Peter Singer has explicitly offered in support of the ideas that we (as affluent individuals) have a moral obligation to donate to aid agencies and that we are doing something wrong if we fail to donate to them. Hopefully then, it is clear to conclude that Singer has not offered a valid argument for these conclusions in the past 54 years, despite many (perhaps carelessly) claiming otherwise.

Now, perhaps it was a little unfair to have simply singled out Singer's standard form arguments, and to have criticised those rather than to have also engaged with the body of work in which they are embedded. Perhaps the appropriate thing to do then is to construct and consider a version of the argument that respects his entire body of work on global poverty, and that guards against the kind of criticisms that have been made in this section. Another reason to favour this approach is that, at least in many cases, the standard form arguments considered and critiqued in the section seem to have been intended for a general audience, and were offered somewhat casually. This suggestion, of course, does nothing to absolve those who, in formal, academic journals and settings, have affirmed the validity of these precise instances, but perhaps it is nonetheless the appropriate way for us to move forward. So, it is to this we now turn.

Here is an attempt at offering a valid argument to the conclusion that if you do not donate to effective aid agencies you are doing something morally wrong. This argument also attempts to respects Singer's body of work on global poverty, in a way that tries to respect the historical ordering of his ideas and situate his position within the effective altruism movement. ⁽¹¹⁾

This argument seems to be valid. It also has, as its conclusions, that we ought, morally speaking, donate to effective aid agencies, and that if we do not donate to them, we are doing something morally wrong. These, I take it, are the conclusions that Singer wants, and that he thinks we ought to accept.

Now, is this a good common-sense morality argument? Well, unlike many of the arguments presented in the last section, it is, at the very least, less obvious that this one can be thought of as a common-sense morality argument. In particular, it seems much less so than Argument 1—Singer's most well-known and influential instance. If a common-sense morality argument is meant to be one that simply contains premises that express everyday, widely shared, commonsense beliefs, then it is less clear that this argument fits the bill. This is because it is at least less obvious that premise 4 and possibly premise 5 express everyday, widely shared, common-sense beliefs. Also, and perhaps more importantly, if understanding and being persuaded by a common-sense morality argument is not supposed to be the kind of thing that requires those who engage with it to have an understanding of any moral theory, and if it is also supposed to be theory neutral, then it is at least less clear that this argument satisfies these requirements either. In particular, it is much harder to believe with the addition of premise 4 that this argument is theory neutral. Premise 4 certainly has a whiff of utilitarianism about it. If the argument then is not theory neutral, or if accepting premise 4 requires some understanding of utilitarianism, then in what sense is this argument still a common-sense morality argument? Now, of course, to all of this, one might say, so what?! Does it really matter whether this valid and possibly sound argument is also a common-sense morality argument? Isn't it enough that we have been given a reasonable one that justifies belief in these interesting and important conclusions? Well, possibly. But here are some things to note. If Singer's goal is not just to offer a reasonable argument to these conclusions, but, more specifically, to offer a common-sense morality argument to these conclusions, then, as these remarks should make clear, it is at least not obvious that he has provided us with that, or with the resources to construct one. More strongly though, given the venues and the mediums through which Singer has offered these arguments, coupled together with his public facing activist work, it does seem reasonable to think that he, in fact, has it as one of his goals to offer a common-sense morality argument to these conclusions. But, on the other hand, if all of this is mistaken and, in fact, it is not his goal to offer a common-sense morality argument for these conclusions, then it seems odd why he has not then just simply offered a simple and direct argument to these conclusions from his brand of hedonistic utilitarianism. ⁽¹²⁾ For example, something like the following argument:

Setting all these concerns aside, there is yet another worry about Argument 8. This one concerns its soundness. While the argument may very well be valid, it is not clear that all of its premises are true or acceptable. In particular, it is not clear that premise 4 is true or acceptable. It is not clear that we ought, morally speaking, aid in the most effective way, or that many people would find this claim acceptable. The reason for this is that this premise is in tension with the identity conferring beliefs and commitments that many people have, and that the satisfaction of this premise would, for many people, alienate them from who they take themselves to be or want to become. This concern echoes Bernard Williams' (1973, 1981) critique of utilitarianism—particularly his claim that moral theories which disregard an agent's integrity risk alienating individuals from their own projects and commitments. Demanding that we always aid in the most effective way threatens to reduce moral agents to instruments of utility, sidelining the identity-conferring projects that give life meaning. ⁽¹³⁾

Here, in more words, and hopefully more convincingly, is the worry. Some of the projects, activities, attitudes, and beliefs that we hold or participate in are expressions of our identities. They reflect what we take to be good and valuable. Holding or participating in them is crucial for the maintenance of these identities. Some of the projects, activities, attitudes, and beliefs that we attempt to adopt or engage in are part and parcel of identities we hope to make our own. They too reflect what we take to be good and valuable. And holding or participating in them is crucial for the development and acquisition of these identities.

For example, if we and others are likely to regard ourselves as an ally to the LGBTQIA+ community, then this piece of our identity can only be said to be genuine if it is the case that some of the projects, activities, attitudes, and beliefs that we hold or participate in are expressions of support for the LGBTQIA+ community, and, moreover, that all of these things are motivated out of the care we have for this community and the value we recognise in it. Similarly, if we are trying to shape and develop ourselves in such a way so that being an ally to this community becomes part of our identity, then, again, it will be important that we adopt and act in ways that express support, and that these expressions are motivated out of our care for the things we value.

Now an obvious consequence of this is that we can hardly lay claim to an identity, or be said to be attempting to forge a new one, if it is the case that we are unable to act in ways that reflect holding that identity or that are important to its acquisition. For example, one cannot claim to be an ally to the LGBTQIA+ community if one's projects, activities, attitudes, and beliefs do not, in any meaningful way, express support for that community.

The worry then is that if it is true that we ought, morally speaking, aid in the most effective way, then this, in many cases, could alienate us from ourselves, the people we regard it valuable to become, and from the commitments that flow from what we care about and value. In a sense, if we were obligated to act in accordance with what premise 4 demands, then unless our identity and values align themselves perfectly with providing aid in the most effective way, then we would simply be reduced to a vessel that assists in the redistribution of goods to those that need them the most.

Now let us talk about these considerations within the context of helping the world's less fortunate. Suppose you believe, as many do, that so much extreme poverty is the result, or at least largely the result, of historical and structural injustices, and that because of these injustices economic and political structural reform, among other things, is necessary. Suppose further that you care deeply about living in a just world and that these reforms take place. What is also true of you is that your time, and skills, and strengths, and other resources are finite. Naturally then, when it comes to discharging your responsibility to help those in dire need, when doing so is within your means, you face a choice. You may have to decide between helping to alleviate suffering and death that results from extreme poverty in a way that upholds or helps to develop identity conferring commitments, by say, championing policy which you have reason to believe will contribute to the reformation of economic or political structures in just ways, or, instead, to donate an equivalent amount of money or some other resource, you would have otherwise used in the aforementioned way, to some effective aid agency, which does more to relieve death and suffering than the alternative but that does hardly anything, if at all, to promote the just structures you regard as important. What then should you do? Well, if premise 4 is true, then you would be required to donate to the effective aid agency, at the expense of either upholding or acquiring your identity conferring commitments. What's more, if this kind of reasoning could always be applied, then the endorsement of this premise would require many of us to abandon some of our identity conferring commitments and it would strongly encourage us to never attempt to form ones that would conflict with aiding others in the most effective way.

All of this seems sad, and it is surely too much to ask of anyone who is clearly doing some good, perhaps even a great deal of good, so as to help those in need. It is not clear that such a person would be acting impermissibly. So then, from all of this, one might conclude that premise 4 is either false, or, at the very least, unacceptable.

Moreover then, on the basis of all of the considerations discussed in this section, it seems reasonable to conclude that, in light of Singer's entire body of work on global poverty, he has not provided us with a common-sense morality argument that is valid and acceptable to the conclusions that we have an obligation to donate to effective aid agencies or that we would necessarily be doing something morally wrong, if we do not donate to them.

This article began by highlighting that every distinct instance of Singer's common-sense morality argument to the conclusions that we have an obligation to donate to effective aid agencies and that we would be doing something morally wrong, if we do not donate to them, is invalid. This is despite the fact that it has been commonplace for many, including those who criticise Singer's argument, to claim that it is valid. This article then presented what appears to be a new, and valid instance of Singer's argument that respects his entire body of work on global poverty. This argument was then criticised for not being a common-sense morality argument and also for having at least one unacceptable premise. So, on the basis of all of this, it seemed reasonable to conclude that Singer has not provided us with a common-sense morality argument that is valid and acceptable to the conclusions that we have an obligation to donate to effective aid agencies or that we would be doing something morally wrong, if we do not donate to them.

Where does all of this leave us? Well, since most (or, at least many) people readily accept premise 2 of Argument 1 or something that very closely resembles it, either because they take it as a true or acceptable common-sense belief or because they endorse any reasonable and widely endorse normative moral theory, which has it, or something like it, as a straightforward consequence of its central principles, it seems reasonable to conclude that we all have a moral obligation to help those in need, where doing so comes at little or no expense to us. Of course, saying this much does not obligate us to provide one form of assistance over another. The kind of assistance we offer will often depend on factors such as the nature of the problem, our abilities and resources, the range of options that are before us, their difficulty in being executed, their likelihood of success, and, importantly, on our relevant identity conferring commitments. Sometimes it will be appropriate to help by way of donation, sometimes it will not. Importantly, as I hope this article has made clear, while we have an obligation to help those living in extreme poverty and anyone less fortunate than ourselves, we do not necessarily have an obligation to donate to aid agencies (or to donate monetarily) to help them, nor would we do something wrong if we failed to do so, provided that we are offering some meaningful and justifiable form of help. The help we provide need not be, or need not always be, the best we can do by the people requiring our assistance, but it should earnestly do some good, and it should be something that we can justify with good reason. If doing this good coincides with and supports or develops some important and valuable identity conferring commitments, then it seems like in many cases we will have done well enough.