You might think that if it’s not wrong not to donate to charity, then it’s not wrong to give to whatever particular charity you choose (as long as no harm is done). I’m going to argue against this view. Very often, it is wrong to give to an ineffective charity, even when it’s not wrong not to give at all.
It’d be wrong of you to refuse to greatly help others (save lives), if the cost to you of doing so were very small (muddied shoes). But how much are you 
morally required to sacrifice, for the sake of helping others? At present, there’s significant disagreement among moral philosophers about this, but most would agree that we – those likely to be reading this post – are morally required to give substantially more to help others (especially those living in extreme poverty) than we currently do. As Larry Temkin noted in an engaging talk, in just one year Americans spent $306 billion on philanthropic causes; but the majority of this money went to religious institutions, alma maters, etc., and at best only $13.3 billion went to aiding people in extreme poverty. Compare this with: $74 billion spent on toys and sports equipment, $323 billion on tobacco, $457 billion on alcohol, $300-600 billion on soft drinks, chips, and candy, $841 billion on recreation, and $2.7 trillion on eating out at restaurants. (These figures come from the 2010 US Statistical Abstract.)
According to many, once you’ve given some sufficiently large amount, it is supererogatory to give still more of yours for the sake of benefiting others. That is, it would be permissible and perhaps also morally better to give more, but not wrong to refuse to do this – giving still more would be going beyond the call of duty. And although they’re in the minority, there are some who think that every act of self-sacrifice is supererogatory, no matter how small the sacrifice and no matter how greatly others would benefit.
When it’s supererogatory, or morally optional, whether to give your $200 to charity or instead keep it for yourself, does it follow that it’s morally optional where you give your $200, if you do give it to charity? Many seem tempted to accept this inference, but I don’t think we should. (See note 1.)
Consider a thought-experiment. There is 1 person drowning in West Lake and 100 people drowning just to the east in East Lake. If you were at the same latitude as these lakes, it’d be virtually costless to save the 1 and virtually costless to save the 100, but impossible to save all 101 (you can launch a rescue boat into one lake or the other). Unfortunately you’re 50 miles south of the lakes, and it’d cost you $200 (including inconvenience) to travel up to them. Assume it’s supererogatory to incur the $200 cost to travel north. Nonetheless, it seems that if you did travel north and were at the same latitude as both lakes, you’d be morally required to save the 100 people in East Lake over the 1 in West Lake. (Once you’re up there, it feels like you’re in a simple “save the greater number” case.) In this example, it’d be morally optional whether to travel up to the lakes, but not morally optional where to offer assistance once there. (See note 2.)
One might object that this isn’t how giving to charity typically works. It’s not as if you first pay $200 to some meta-charity, and subsequently decide whether it goes to charity X or to charity Y. Instead, typically the reality is that whether to donate and where to donate are bundled together in a single choice: you can donate $0, donate $200 to charity X, or donate $200 to charity Y.
In response, I could revise the thought-experiment: West Lake and East Lake are 100 miles apart, and you’re right in between them. You could pay $200 to go west (where you could virtually costlessly save 1) or you could instead pay $200 to go east (where you could virtually costlessly save 100). Again, you can’t save all 101. Again, we’re supposing it’s morally optional whether to incur the cost of $200 at all. But it doesn’t follow that if you are going to incur this cost, it is morally optional whether you head east or west. On the contrary, it seems that if you are going to incur this cost and end up at one lake or the other, you are morally required to go east where you’d do much more good. And this does appear analogous to situations we’re often in with regard to charitable donation. While you might not be morally required to give your $200 to charity, if you are going to donate it, and you have the choice between a charity X that would use it to save 1 life and a charity Y that would use it to save 100 lives, you should donate to charity Y (see note 3). (Where this “should” entails that it would be wrong of you not to donate to charity Y.)
One might reasonably ask what it is that ultimately makes it permissible not to donate the $200 at all – for it could turn out that whatever this is would in turn make it permissible to donate to charity X as well as permissible to donate to charity Y. I believe a defensible answer is that your own well-being (and perhaps also the well-being of those near and dear to you) generates special permission for you to act in a way that is not optimal from a fully impartial perspective. Perhaps you’d be permitted to save yourself (or your loved one) over 100 strangers. But suppose you’ve decided to donate your $200 to “charity,” where this refers to a long disjunction of particular charities A, B, C, …X, Y, Z, none of which would, if donated to, benefit you (or anyone near and dear to you). Indeed, this is a fairly realistic situation for donors to find themselves in – as so many charities help distant needy strangers. Thus we can see that it doesn’t follow from the claim that it’s morally optional whether to give that it’s morally optional where to give.
Next one might object that what makes it permissible for you not to donate the $200 at all is simply a matter of your desires. The thought continues: If you desire to “donate” the money to yourself (or to a friend, a family member, a restaurant server, etc.), it’s permissible to do so, and so similarly if you desire to donate the money to charity X rather than the much more cost-effective charity Y, it’s likewise permissible to do so. But unless the frustration of your desire to give to charity X constitutes a genuine cost to you, the second half of the previous sentence is very implausible. If you could costlessly save the 1 in West Lake or costlessly save the 100 in East Lake, it wouldn’t be permissible to save the 1 in West Lake merely in virtue of your desire to do so. If on the other hand the frustration of your desire to give to charity X did constitute a genuine cost to you (as certain desire-satisfaction theories of well-being imply), then perhaps it’d be permissible to give to charity X. But it’s worth noticing that apart from your having this desire there’s nothing costlier to you about you giving to charity X rather than charity Y, and charity Y is much better from an impartial point of view. Insofar as changing your desire is itself costless to you (or perhaps it’d even be beneficial to you to bring your desires closer into alignment with reason!) it’s plausible that you should desire giving to Y instead of X. (This isn’t to say you should give to Y, assuming your desires haven’t yet changed.)
Finally, as before, there remains the simple point that, while you may have a desire to donate your $200 to yourself over giving it to charity, your desires themselves may be neutral when it comes to selecting between particular charities A through Z. This happens often enough. And so in these cases permission not to donate wouldn’t imply that it’s permissible to donate to any charity, given that one does choose to donate to charity.
Many people give to charity (and not just to “charity” but also to friends, family, etc.) not on the basis of mere desires, or whims, but on the basis of deep emotional connections and commitments. Perhaps the basis of your permission whether, and where, to give isn’t simply a matter of your desires, but rather of what’s close to your heart. Perhaps deep emotional connections can generate such permissions. But first, even if so, this wouldn’t support a general inference from “permissibility whether to give” to “permissibility where to give,” since there will remain cases in which your emotional connections are neutral with respect to the particular charities you can donate to. And second, surely there are limits on the sort of permission that emotional connections could plausibly generate – they can’t plausibly permit you to act in ways that are much worse from an impartial standpoint. Return to the West Lake versus East Lake example. Suppose you’ve worked as a lifeguard for West Lake for many years, and have a serious emotional connection and commitment to helping people there. You can save 1 person drowning there, or save 100 people drowning in East Lake (all strangers to you). It seems implausible that the fact that helping people in West Lake is closer to your heart could permit you to save the 1 over the 100. Should it matter if we replaced “helping people in West Lake” with “helping people in the United States” or “helping people who have suffered from some illness you’re especially emotionally connected to (such as cancer)?” I do not believe so. But just how strong must the emotional connection or commitment be to generate significant permissions whether and where to give, and just what are the limits of the permissions generated? These are difficult questions I won’t be able to settle here.
I’ve been discussing where you should donate your money, assuming that it’s permissible for you not to donate to charity at all (either because you’ve already given so much or because the duty of beneficence is very undemanding). I’ve argued that at least as long as you’re donating to charities that are equally good from the standpoint of your own interests (and the interests of those near and dear to you, and your desires, and what’s close to your heart), you should donate to the ones that do the most good from an impartial standpoint. The moral option not to give to charity at all doesn’t entail that it’s morally optional where to give. But for all that, probably most of us would have to give considerably more before the moral option not to give kicks in. I suspect the truth is that the duty of beneficence is pretty demanding by comparison with commonsense standards, and that few of us really do give enough. I doubt I do.
I am grateful to Caley Anderson, Tom Ash, Gulzaar Barn, Roger Crisp, Michelle Hutchinson, Brian McElwee, Andreas Mogensen, Caleb Ontiveros, and Peter Singer for helpful discussions.
Notes:
1. Some qualifications: First, I am not here discussing legal options or rights; if there weren’t and ought not to be laws compelling people to donate money to Oxfam, this wouldn’t settle the question of whether it’s morally optional to donate there. Second, though the following discussion is couched in terms of saving lives, it’s highly debatable whether life-saving charities are best from an impartial point of view. Arguably, charities that primarily improve quality of life (such as deworming charities) are better. It’s also worth considering charities that focus on animal well-being. My examples could be revised accordingly without damage to the main argument. Third, I recognize that there is a lot of uncertainty about what charities are best or “do the most good” from an impartial point of view. More radically, it may be indeterminate which are best, or perhaps none is best but several very good charities are on a par; I believe the main points of this post are consistent with these possibilities. Importantly, it remains the case that some charities are much better than many others.
2. I’m not the first to offer examples illustrating this point. In The Limits of Morality (p.16), Shelly Kagan imagines that you find a building on fire, and are told that someone may still be trapped inside. Plausibly, it is permissible for you not to enter the building to try to rescue them. Suppose, however, that you do enter the building and find a child and a caged bird; you can rescue one or the other, not both. Plausibly, it would be wrong to save the bird rather than the child, though it would not have been wrong for you to have refused to save the child by refusing to enter the building out of concern for your own safety. (In Kagan’s example the cost to you is much greater than the cost to you in the lakes example, as well as greater than the cost to you in more everyday cases of donating to charity.)
3. Typically it costs considerably more money than this to save lives. Arguably the most cost-effective life-saving charity (the Against Malaria Foundation) saves a life roughly every $3,000 received.
Hi Theron,
This is a very appealing and, on its face, intuitive argument. But it has a bizarre (if familiar) implication: When faced with the option to 1) save the many at cost, 2) safe one at cost, or 3) do nothing (and incurring cost is supererogatory), 2 is impermissible while 3 is permissible. That is to say, it is more morally acceptable to stand by and do nothing than to save one person at cost. Not only is this perverse, it is in tension to the converse principle that motivates making 2 impermissible in the first place, where, e.g., one is *required* to save someone (in normal cases) when there is no cost to oneself (surely, in costly cases, it should at least be *allowed* to save one rather than none!).
This is not a problem in Kagan’s sort of diachronic case, where one’s obligations shift from when one is initially incurring costs (saving none is OK) to when one is choosing whom to save at no further cost (one must save the many); at t1, one chooses between 3 and ~3, while at t2 one chooses between 1 and 2, so it is nowhere implied that 3 is preferred over 2. But the second version of the case (more closely analogous to charitable giving) is synchronic where all 3 options are considered simultaneously, and so leads to this strange result.
You might press the conditional: one is required to save the many *if* one decides to give, which dissolves the tension by bifurcating 1 and 2 from 3. But as that is not a diachronic ‘if’, I don’t think the bifurcation can do much work here. Indeed, I could redescribe the case as follows: one is required to do nothing, *if* one decides not to save the many.
Alternatively, you might say that giving to charity really does (at least sometimes) have the structure of a diachronic case. At t1, one decides to give *something*; at t2, one decides to whom to give. These might be very close together, but no matter – at t2, option 3 (do nothing) has been removed, leading to no strange implications.
Two issues. A) it’s not clear to me that’s how most decide to give to charity. Some do, to be sure, but just as many will keep not giving as a live option when considering donations. Moreover, B) even when one does decide that way, the mere decision to give doesn’t have the finality of shutting off options that, say, having already purchased the rescue boat or rushed into the burning building has. In those cases, it’s too late – you’ve already incurred the cost, and declining to incur costs is no longer a possibility. But even after one has decided to give to charity, it is still a genuine option to change one’s mind and do nothing. And so, one still has the perverse implication that it is more acceptable (better?) to do nothing rather than save one at cost.
A final option might be to insist that 1 is required with respect to 2, but not with respect to 3; whereas 2 is required with respect to 3. Obligations are not absolute on this picture, but only relative to some (single?) alternative option. Obligations would then have an odd structure and wouldn’t be transitive (one ought to A over B and B over C doesn’t imply one ought to A over C, on this picture, even when A, B and C are all possible), so one is trading one strange implication for another, but granted the new implications are only theoretically strange (and have some acceptance).
Thanks for the interesting post.
You say: “It seems implausible that the fact that helping people in West Lake is closer to your heart could permit you to save the 1 over the 100.”
I think, for your argument to work, you really have to be talking about instrumental rationality rather than moral obligation. You have to assume that the intention to save the 1 is some general intention to do what’s impartially good. Then this all makes sense.
But if you’re doing the act of saving the one rather than saving the 100 for any other reason, it’s not clear how this is different for any other permissible act you could do with the $200, such as buy a new pair of shoes, or invest it, etc. You seems to be endorsing the principle, “You either have to be completely selfish or maximally altruistic.” But this is a strange principle, and I don’t think it’s what the defenders of a limit to impartial obligations would want to defend.
Suppose, instead of spending $200 dollars to save the 100, I spend it on a bus ticket to visit my ailing grandmother. Suppose I do this for purely altruistic reasons, because it will make her feel better, even though I’m not very interested in doing it and won’t enjoy it. In this example, agains assuming that it’s permissible for me to spend $200 on my shoes instead, it’s really hard to imagine that it’s impermissible to visit my grandmother.
You might think this example supports my case because it smuggles in intuitions about interpersonal relationships, but then denies them. However, I think it shows that altruistic acts can have all sorts of characteristics, and if maximization isn’t always required simpliciter, it shouldn’t be required once one chooses to act altruistically.
This seems to be describing a case unlike the normal case of the ineffective-charity-giver. The condition “if you are going to incur this cost” presumes that a person would give £200 to a charity, whether it be charity A (about which they care enormously) or whether it be charity B (about which they care much less). But this surely does not describe the typical person who views charity as super-erogatory, but wants for some reason to give to an ineffective charity. Rather, they have a particular interest in some ineffective charity (or class of charities) A, but no interest whatsoever in giving £200 to charity B. Still, maybe you’re not interested in actual givers and just hypothetical people who *do* say “I will give £200 to charity be that charity an ineffective I most want to give to or some other effective ones I don’t care so much about… now which one will I give to…?”
Re. the modified East-West Lake case, the statement “it seems that if you are going to incur this cost and end up at one lake or the other, you are morally required to go east where you’d do much more good. And this does appear analogous to situations we’re often in with regard to charitable donation” just seems to be begging the question. I don’t see what motivates or is to make plausible this presumption. Presumably the person who thinks that giving to charity/helping people out of lakes is simply supererogatory will just not think that they are so required, and I don’t see why they should.
You seem to think that pulling people out of lakes/giving to charity is not really supererogatory: rather it sounds like an obligation, which is defeasible if there’s too much of a cost to your well-being. Once the cost of going to the lakes is cancelled out, you seem to think that it’s just obligatory to pull people out of lakes.
Dear Owen, Cody, and David,
Thank you so much for your very thoughtful responses to my post! I’ve now written up replies to each of you. (There’s some overlap in these replies; I’ve written up these responses the way I would if I were replying to each of you individually via email.)
Owen: I’m sympathetic to much of what you say in your last three paragraphs, but I am not as worried as you by the issue you outlined in the three paragraphs before that. Let’s focus on the synchronic version of the two lakes case. Assuming that doing nothing is permissible in virtue of the cost to you, I do not find it odd that it is more morally acceptable to do nothing than to save one at cost as long as we’re clear about the fact that there’s a third option to save many at the same cost as saving the one. If this third option disappeared, then of course it would be perverse to say doing nothing is more morally acceptable. Similarly, it sounds odd to say that “one is required to do nothing, if one decides not to save the many” if one views the decision not to save the many as removing the option to do so from the mix; but I don’t think it sounds odd to say this if we assume that it’s permissible to do nothing, and that you really do have all three options on the table. We might imagine a synchronic version of the Kagan burning building case. Suppose the level of risk to you is such that it’d be permissible to do nothing, and permissible to run into the building and save the bird if there were no third option. But if there were the third option of running into the building and saving the child it would seem to me impermissible to run into the building and save the bird. If you’re going to incur that risk on yourself, it seems wrong to incur it for the bird over the child (and for perfectly legitimate, non-speciesist reasons!). It seems morally objectionable that you failed to do something morally better that would have been no costlier to you than what you actually did. Similarly, it may be supererogatory whether, in the synchronic two lakes case, you rescue anyone. But it would be morally objectionable to save the 1, since again you failed to do something morally better (save the 100) that would have been no costlier to you than what you actually did. On the other hand, if you don’t save anyone at all, no one can complain that you failed to do something morally better that would have been no costlier to you than what you actually did.
Cody: I am not endorsing the strange principle “You either have to be completely selfish or maximally altruistic.” First, I should clarify that my entire post is dealing with cases in which you are permitted not to help. I tend to think that beneficence is fairly demanding, and that for such permission not to help to plausibly kick in you would first have to sacrifice quite a lot. In other words, you’re required to help before you reach this point. This kind of moral requirement is inconsistent with the strange principle you mentioned. But perhaps you were wondering whether I intended for the principle to apply only after you reach the point at which it’s permissible not to sacrifice more to help. Even then, I don’t want to endorse the principle you mentioned. I distinguished between several different possible bases of the permission not to help: cost to you, cost to those near to you, desire frustration, frustration of deep emotional commitments or what’s close to your heart. I then pointed out that these bases may well permit you not to donate to charity, but they’d not necessarily render it equally permissible to give to any charity for the reason that these bases may be (are often) neutral with respect to giving to many different charities. So “permissible not to give” doesn’t entail a blanket permission about where to give. Of course, there are some altruistic activities besides donating to charities, such as paying money to visit your grandmother. It may be that several of the bases I mentioned would deem it permissible to spend your money in that altruistic way, even though it’s not best from an impartial perspective. My point, in the bit you quoted, is that I am skeptical that some of these bases (the “desire” one, and the “closes to your heart” one) would plausibly permit you to act in ways (altruistic or not) that are *much worse* from an impartial perspective. Of course, it’s debatable whether *any* of the bases could plausibly permit you to act in a way much worse from an impartial perspective assuming you’d lose only slightly in terms of these bases (e.g., if you’d suffer only a slight cost to yourself or to a loved one, or one of your desires or projects were set back only slightly). But maybe it’s not implausible that, *once you’ve sacrificed a lot of your own well-being*, you’re permitted not to sacrifice *any* more, even if this is much worse from an impartial point of view. Nonetheless, it does strike me as implausible that, no matter how much you’ve sacrificed (in terms of your well-being, desires, projects, or whatever), it’s permissible for you to act in a way that’s *much* worse from an impartial perspective *merely* for the sake of only slightly or moderately advancing a project close to your heart (i.e., even though doing so won’t benefit you at all). Perhaps I’m mistaken about this. Even if so, my broader point about blocking the inference from “permissible whether to give” to a blanket permission about where to give would still go through (for the reason mentioned earlier). Your example about buying shoes versus buying a ticket to see your grandmother is an interesting one; it might initially seem implausible that it’d be permissible to buy yourself shoes but impermissible to visit your grandmother. However, note that there are a few possible explanations of the appearance of implausibility: one is the perceived cost to you of not helping your grandmother (because you care about her), another is the perceived cost to someone near and dear to you (your grandmother). But if we assume and take care to appreciate that there’d be no benefit to you of visiting her, and no benefit to her either (and no benefit to the relationship itself), then it doesn’t seem very implausible to me anymore that it’d be permissible to spend on yourself but not on visiting her. Of course, making these assumptions would make the case less realistic, but that’s what you’d need to do to make the case a relevant test of the claims I offered.
David: As noted in the discussion with Owen, we can distinguish between diachronic cases and synchronic cases. My initial two lakes example (and the Kagan example in note 2) is a diachronic case – you first have the choice between making a sacrifice and doing nothing, and then later you have a choice between doing more or less good from an impartial point of view (where the options in this second choice are equally costly to you). The modified two lakes example is a synchronic case – you have a single choice point with three options: do nothing, spend $200 and save 1, spend $200 and save 100. In the diachronic case, the phrase “if you are going to incur this cost” just means, “if you are going to make the sacrifice at the first choice point.” As I noted in the post (and as Owen noted in his response), this is not really how giving to charity works. In the syncrhonic case, the phrase “if you are going to incur this cost” just means “if it is the case that you pick one of the two sacrificial options.” (It doesn’t mean or imply that, if you don’t take one of the sacrificial options, you *would* instead take the other sacrificial option.) Now moving on to your criticism of what I said about the modified two lakes case. I confess that I did not offer much in defense of the claim that, if you pick one of the two sacrificial options, you’re required to pick the one in which you save 100. But I think that, once all the details of the case are understood properly, this is an intuitively plausible thing to claim about this case. As I said in reply to Owen, I can also appeal to a similar intuition about the synchronic version of the Kagan case. If it’s between doing nothing, running into the burning building to save the bird, and running into the building to save the child, it seems to me that while it may be permissible to do nothing, it would not be permissible to run into the building to save the bird. It seems morally objectionable that you failed to do something morally better that would have been no costlier to you than what you actually did. This is consistent with saying that it’s supererogatory to run into the building. Similarly, it may be supererogatory whether, in the synchronic two lakes case, you rescue anyone. But it would be morally objectionable to save the 1, since again you failed to do something morally better (save the 100) that would have been no costlier to you than what you actually did. On the other hand, if you don’t save anyone at all, no one can complain that you failed to do something morally better that would have been no costlier to you than what you actually did.
Hi Theron, I bookmarked this a long time ago and finally got to read it. Just a comment related to your exchange with Owen Schaefer. Instead of worrying about a diachronic/synchronic distinction, it seems potentially fruitful to think about wide/narrow scope. I understood you to be arguing for a wide-scope requirement: roughly, morality requires that if you save, you save the many. The conditional understood in this way is compatible with morality being neutral about whether, if you don’t save the many, you do nothing (to take the example in Schaefer’s third paragraph). Having said that, I still find your position hard to understand. You concede that unspecified giving is supererogatory (in the proposed circumstances); I just don’t know what it means for an act-type to be supererogatory, unless every instance is supererogatory.
Thanks Teru, this is a very helpful comment. But I want to defend a view according to which morality is *not* neutral about whether, if you don’t save the many, you do nothing. That’s because in some cases it’s wrong to save the few when you could have saved the many at no greater cost to yourself – that’s objectionable in a way that simply doing nothing isn’t. On your last remark, I do think I could have been clearer in some of the statements of my position. I should not say that unspecified giving is supererogatory (in the proposed circumstances), if that implies that it is permissible to give anywhere. Instead, I should say that it is supererogatory to give to some places (rather than keep the money for oneself).
Comments are closed.