Barry and Øverland on Singer and assistance-based duties

ABSTRACT

Barry and Øverland reject Singer’s assistance principle as being too demanding and offer a positive defence of a less demanding one. In this article, I critically scrutinize their arguments. My main claim is that they fail to show that their own principle of assistance is superior to more demanding ones such as Singer’s.

KEYWORDS:

• Appeal to costs
• duties of assistance
• moral equality
• moral options
• poverty
• Singer

Introduction

Chapter 2 in Christian Barry and Gerhard Øverland’s Responding to Global Poverty provides an insightful discussion of Singer’s (1972) classical argument for why affluent people are under a strong obligation to help the global poor.¹ Singer’s argument appeals to moral intuitions regarding a case in which a passer-by is able to save a child from drowning at insignificant cost to herself. Intuitively, the passer-by is morally obligated to save the child (and, on reflection, we see that rich people’s relationship to, e.g., the starving in East Bengal in 1971 is not relevantly different from that of the passer-by and the child). This would be explained by the truth of Singer’s Assistance Principle (henceforth: SAP):

[I]f 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 (Singer 1972, 231).²

Hence, by way of inference to the best explanation, one might conclude that SAP is true. However, as Barry and Øverland point out, SAP is not the only principle that justifies the relevant intuition. In fact, they believe that SAP attributes stronger obligations to us than we have and than are needed to explain moral intuitions about Singer’s vignette (33–34). Our obligation to rescue Singer’s drowning child is satisfactorily explained by their moderate assistance principle (henceforth: MAP):

If we can prevent something (very) bad from happening at relatively modest cost to ourselves and others, then we ought [morally] to do it (33).

Section 2 presents Barry and Øverland’s reasons for preferring their less demanding MAP to SAP. Section 3 assesses their critique of SAP, appealing to cases involving the sacrifice of body parts. Section 4 moves on to scrutinize Barry and Øverland’s argument for why intuitions about cases involving the sacrifice of external objects cohere better with MAP than with SAP. I argue that Barry and Øverland’s defence of MAP over SAP is inconclusive both when it comes to sacrifices of body parts and to economic sacrifices. Section 5 challenges Barry and Øverland’s attempt to ground MAP in a balance between moral equality of persons and autonomy. Section 6 concludes.

Barry and Øverland’s argument

One of Barry and Øverland’s arguments for why SAP is too demanding appeals to cases where one must sacrifice a limb to rescue Singer’s drowning child. Losing an arm is not losing something of ‘comparable importance’ to the child’s life, but losing an arm is a loss which is greater than relatively modest and, accordingly, unlike MAP, SAP implies that one ought to sacrifice one’s arm. In Barry and Øverland’s view, this speaks against SAP and in favour of MAP, since ‘[SAP is] implausible when we consider moral requirements to sacrifice body parts to save a child’ (35).

One might suggest that there are special concerns about bodily integrity and, thus, that even if Barry and Øverland are right about cases where rescuing badly off people requires sacrificing bodily parts, this is not relevant to global poverty, since the sacrifices in question are economic and as far as these are concerned, we must accept sacrifices greater than relatively modest ones (36–37). However, Barry and Øverland argue that generally, only modest economic sacrifices are required, so even if it may be easier to see that SAP is false when it comes to sacrifices involving body parts, ultimately, such sacrifices are no different from sacrifices involving other goods.

Barry and Øverland concede, however, that there are cases that initially may appear to support SAP rather MAP:

Bob’s Bugatti: Bob, who has most of his retirement savings invested in a Bugatti, is confronted with the choice of redirecting a railway trolley by throwing a switch in order to save a child which will result in the destruction of his Bugatti because it has accidentally been placed on the side spur of the line, or he might leave the switch as it stands so that his Bugatti remains in mint condition, which will result in the child’s death (37).

Barry and Øverland believe that Bob is required to sacrifice his Bugatti. They think, however, that there are other cases involving economic sacrifice where sacrifices greater than moderate are not morally required, and that this vindicates MAP, in part because there are, as we shall see, special features of Bob’s Bugatti which explain the requirement to sacrifice it. Having defended MAP against SAP by appeal to moral intuitions about particular cases, Barry and Øverland proceed to ground MAP in a ‘reasonable balance’ between the values of moral equality and autonomy. I now turn to an assessment of their arguments.

Body parts

Barry and Øverland are right to assess the plausibility of SAP in light of cases involving the sacrifice of limbs. However, the picture is more complicated than they suggest. Consider:

Artificial Arm: In order to save the child, Ann has to jump into the pond, where a crocodile will bite off her right arm. However, once the rescue mission is completed, super-surgeons will graft an artificial, but equally good, arm to her right shoulder. However, she will be charged a fee for the operation. The pain from having her arm bitten off together with the fee adds up to costs that are more than moderate.

Here I am inclined to think that Ann is under an obligation to save the child (though it is excusable if she fails to do her duty). Admittedly, it is not clear that she is under such an obligation, so perhaps there is something special about one’s body. There is also another possibility, however: That our intuitions pertaining to the sacrificing of limbs reflect difficulty, not costs.

Being costly and being difficult often go together. In Singer’s example, rescuing the child is neither. However, though I cannot offer a full analysis of the concepts of cost and difficulty, I believe that as they are commonly used, the two can be separated (Cohen 2011, 16). Suppose I have taken some drug which makes me relax. In that case, it might be easy, though costly, for me to write a check donating my savings to Oxfam. In writing the check I would neither have to overcome physical obstacles, nor – perhaps thanks to the drug – to prevail over any urges not to make a huge economic sacrifice. Conversely, if I enjoy the challenge of making large companies donate money to Oxfam, because I know they are very unlikely to do so, it might be difficult for me to help poor people in this way, though not costly. The distinction between cost and difficulty bears on our assessment of Artificial Arm. Most would find it psychologically very difficult to let their arm be bitten off. They would do so even if they knew that it would not be costly for them, since they would have a replacement arm and be compensated for their pain, etc. In a variant of Artificial Arm, where Ann will be compensated in this way and will only be charged a small fee, I suspect many will still find it doubtful that she is under an obligation to rescue the child. Perhaps what is driving our intuitions here is not appeal to cost, but appeal to psychological difficulty:

The Moderate Appeal to Difficulty Assistance Principle: If we can prevent something (very) bad from happening and it is only moderately psychologically difficult for us to do so, then we ought [morally] to do it (unless the costs for us are comparable to the badness of the event we prevent).

Since it is generally painful to lose one’s limbs and generally not painful to lose money, and since it is generally psychologically difficult for people to do something which they know will be painful, generally, in cases involving sacrifice of one’s limbs to save others, doing so will be more than moderately psychologically difficult. To test this suggestion, consider a case where sacrificing a limb is costly but not psychologically difficult:

Magic Wand: You wake up recalling standing in front of a crocodile-infested pond deliberating whether to save the drowning child. A sorcerer informs you that he has a magic wand which can make things happen in the past. If you express the wish that you had jumped into the pond, saved the child, but also had your arm painfully bitten off by the crocodile, replaced by an arm no worse than your own and that you now have to pay a bill from the hospital, the sorcerer will make it the case that this is what happened.

What should you say to the sorcerer? Because the case is not one in which any strong psychological resistances need to be overcome – the pain will be in the past and you will have forgotten it at this point – it is not one in which it is psychologically difficult to save the child. However, due to the hospital bill, it is more than moderately costly. Yet, it seems you should sacrifice your limb in this case. Indeed, if I am about to say to the sorcerer that I do not wish I had jumped into the pond, others might permissibly enforce my duty to rescue the child, even if this means that I lost a limb (cf. 36). Accordingly, SAP appears plausible in light of the present example. I conclude that even when it comes to the sacrificing of limbs, the correct principle of assistance might be more demanding in terms of cost than MAP.³

Some might resist this inference on the grounds that my vignettes involve loss of replaceable bodily parts only, and that what Barry and Øverland might take to be distinctive about body parts such as arms etc. is that generally, and contrary to what I have specified to be the case in Artifical Arm, they are irreplaceable. However, while limbs might generally be in some sense irreplaceable (though I doubt that they are in a strict sense), economic goods can be so too. Hence, in effect, the present reply involves a shift of attention away from the distinction between bodily parts and economic goods to the distinction between irreplaceable and replaceable losses, which is not the distinction underpinning Barry and Øverland’s argumentative strategy.

External objects

I now turn to Barry and Øverland’s discussion of SAP in light of cases involving the sacrifice of external things, i.e., vintage cars and life savings. As noted Barry and Øverland agree that Bob is required to sacrifice his Bugatti to rescue the child, even though that amounts to giving up ‘most of his retirement savings’ (37). Barry and Øverland, however, think that we cannot infer a general duty to make more than moderate sacrifices to rescue others. To support this claim they appeal to:

Bob’s Internet Banking: Bob is sitting in his remote house doing some internet banking. Unbeknownst to his only neighbours (the Smiths), he can see and hear them through the open door on the veranda. He notices that they are discussing the state of their terminally sick child, Jimmy… They urgently need a new and expensive treatment to cure Jimmy. They therefore need to get a helicopter to bring their son to a private hospital far away. They live in a society that has no universal health coverage, they cannot afford the helicopter service, let alone the operation, themselves, nor are they able to finance it or acquire the funds from relatives and friends. Bob understands that he can transfer the money for the helicopter service and the operation with a click of his mouse … Clicking over the money would save Jimmy, but most of Bob’s savings for retirement would be gone. Bob decides not to click the mouse (38).

Barry and Øverland submit that it is permissible for Bob not to donate the money – a sacrifice that would be greater than moderate. Hence, there is no general duty to rescue the lives of others when doing so requires greater than moderate costs.

Suppose we share Barry and Øverland’s intuitions about Bob’s Bugatti and Bob’s Internet Banking. How can we explain the difference between them?⁴ Here is a suggestion: In Bob’s Bugatti (and in Singer’s Pond), the bad situation is simply a freak accident, and there is no reasonable way in which one can prevent freak accidents from occuring. In the case of Bob’s Internet Banking, the bad situation is not a freak accident in the sense that parents have a duty to look after their children, and doing so involves ensuring that, within reasonable limits among other things set by the medical costs involved and the parents’ economic opportunities, one is able to pay – perhaps indirectly through a medical insurance scheme – for one’s child’s operation should such an operation be needed. There is a division of moral responsibility, and, assuming that we are within the reasonable limits determining parental obligations, Bob might think it is unfair that he should pick up the slack from the child’s parents.⁵ To test this explanans, consider the following variation on Bob’s Internet Banking:

Bob’s Internet Banking 2: Bob is sitting in his remote house doing some Internet banking. Unbeknownst to a terminally sick child, Jimmy, who is lying on the ground just outside Bob’s property, Bob can see and hear Jimmy through the open door. Jimmy, who has no parents and who had always looked after himself up until he was bitten by a poisonous spider, repeats to himself over and over with a voice which is barely audible that he urgently needs a new treatment to survive, and that he can get this treatment in time only through being transported via a helicopter. Bob and Jimmy live in a society that has no universal health coverage, Jimmy cannot afford the helicopter service, let alone the operation, nor is he able to finance it or acquire the funds from relatives and friends. Bob understands that he can transfer the money for the helicopter service and the operation with a click of his mouse. Clicking over the money would save Jimmy, but most of Bob’s savings for retirement would be gone. Bob decides not to click the mouse.

This case strikes me as really no different from Bob’s Bugatti. This is as one would suspect, if my explanans were correct.

Barry and Øverland, however, offer a different, tentative proposal as to what makes it the case that Bob is under a requirement to sacrifice his Bugatti. In their view, two different factors might be able to explain the moral difference between Bob’s Bugatti and Bob’s Internet Banking. First, Bob ‘acted very imprudently, exposing himself to the undue risk of serious loss’ by investing nearly all of his savings in the Bugatti, since cars ‘are easily damaged, expensive cars are very attractive targets for theft, and are of little use value relative to other material assets such as dwellings’ (41). Second, ‘the decision to drive around in such expensive cars imposes costs on others. Drivers of expensive cars increase third-party insurance for everybody, including those who spend little on cars, since forecast cost of reparation increases’ (41).

This explanation is unconvincing. Consider a variation of Bob’s Bugatti where this purchase neither exposes himself to undue risk of serious loss, e.g., he insures himself against theft etc., nor imposes net costs on others, since he compensates everyone for any costs it imposes on them. In this case, Bob’s duty to sacrifice his Bugatti to rescue the child is no less stringent.⁶ Similarly, if we consider a variation of Bob’s Internet Banking, where internet banking involves exposing oneself to risk of serious loss, e.g., because of hacking, and increases the costs to others, e.g., because doing so makes some of their activities more expensive in the same way that Bugatti purchases drive up insurance premiums, this does not affect our intuition about Bob’s Internet Banking.

Barry and Øverland mention one additional factor which might influence our intuitions – namely, the distinction between immediate and final responsibility (41). One can have the former without having the latter if one is morally required to carry the immediate costs of rescuing someone, while others are then morally required to shoulder some of those costs, such that one is not responsible for covering more than one’s fair and smaller share of the final costs. Barry and Øverland then suggest that we consider the following case, in which the potential rescuer, unlike Bob, has not imposed any ‘undue risk on himself’:

Bob and the Avalanche: There is an unexpected avalanche that is certain to bury a child unless Bob acts. Bob is able to redirect the avalanche, but the only way he can do so will lead it to the new house in which he has invested most of his savings (42).

Barry and Øverland do not think Bob is morally required to rescue the child if he will not – or, alternatively, cannot be reasonably sure that he will (cf., 43) – be compensated by others later. Barry and Øverland also conjecture that Bob and the Avalanche explains why we do not think that one is morally required to sacrifice a limb to rescue the child. For what makes it the case that Bob is not morally required to redirect the avalanche – assuming he is not – is that he will have to cover all the costs of rescuing the child himself. This follows from the fact that what is special about cases where one must sacrifice ‘body parts’ to rescue others is that one’s losses ‘cannot be easily or fully compensated’ (43).⁷ This goes against my suggestion above in the Artificial Arm case.

I want to make two observations in response to Barry and Øverland’s discussion of Bob and the Avalanche. First, consider the fact that if one were certain that one’s economic losses would be fully compensated, one would definitely be morally required to rescue the child. I have submitted that we are not equally certain about a case where one’s limbs would be torn off, but where full post-deed compensation would follow. Hence, the present case does not shed full light on sacrifices of body parts.

My second point is that while I have argued against Barry and Øverland’s account of why Bob’s Bugatti and Bob and the Avalanche are morally different, I concede that the two cases seem morally different. My explanation of this – without necessarily endorsing the explanandum – appeals to Scanlon’s distinction between subjective and objective rankings of interests (Scanlon 1975). It might be true that we are obligated to serve a religious person’s interest in avoiding starvation, but not obligated to serve this person’s interest in building a place of worship – objectively speaking, the former interest is higher-ranking – even if, subjectively speaking, the latter interest is the one that ranks the highest, i.e., the believer prefers to starve if, alternatively, she cannot build a place of worship. A similar difference might lie behind our different assessments of Bob’s Bugatti and Bob and the Avalanche. While Bob’s interest in keeping his collector’s item in mint condition is subjectively stronger than his interest in having a home, objectively speaking one’s interest in having a home ranks higher than an idiosyncratic interest in vintage cars.⁸ Suppose Bob’s Bugatti is a caravan and that if he redirects the trolley, he will become homeless and destitute. In that case, Bob’s Bugatti seems equivalent to Bob and the Avalanche, as one would expect if the present explanation is correct. I conclude that, in cases not involving an unfair distribution of the costs of rescue, the correct principle of assistance might be more demanding in terms of subjective costs than MAP.

Moral equality and autonomy

So far the discussion has appealed to various moral intuitions about vignettes. Barry and Øverland also propose a rationale for MAP:

[MAP] ‘seems… to strike a reasonable balance between two important values, moral equality and autonomy. Since all people have equal moral worth, it is reasonable to expect that all individuals would be required to make certain sacrifices to protect others from very bad things happening. However, because our interest in autonomy is also important, we are nevertheless entitled to give some priority to our own concerns. Therefore, what individuals are morally required to sacrifice to help others in need is significant, but also limited’ (44).

This brief account of the underlying rationale is incomplete. First, the notion of equal moral worth as such implies nothing about our duties to assist others. Suppose that Adam believes that autonomy is so important that, though no person might harm others, there is no duty to help others. Suppose that Beatrice believes that we have extensive duties to help others. However, for racist reasons she believes that while we have a duty to make moderate sacrifices to prevent something very bad from happening to members of out-group races, we have a duty to make greater sacrifices to prevent something very bad from happening to members of in-group races. While Adam’s view is flawed, because he rejects even minimally stringent duties to help others, unlike Beatrice’s view, it is compatible with the moral equality of persons. Hence, there is nothing in the idea of equal moral worth as such that speaks to the stringency of the duty to assist and, accordingly, there can be equal moral worth whatever degree of autonomy in Barry and Øverland’s sense that we enjoy. Hence, there is no such thing as a ‘reasonable balance’ between the two values, since that would involve an unavoidable trade-off between the two values.⁹

Second, one would like to see Barry and Øverland’s notion of autonomy explained in such a way that it follows that the more stringent the duties that we have to assist, the less autonomy we have. In standard senses of ‘autonomy’ – e.g. ‘an agent governs her own action if and only if she is motivated to act as she does because this motivation coheres with (is in harmony with) some mental state that represents her point of view on the action’ (Buss 2013) – no such connection exists. However, Barry and Øverland have a special sense of autonomy in mind. Let us call this special sense ‘moral autonomy’. An agent’s moral autonomy is greater, the greater priority he is morally permitted, but not morally required, to give to his own interests.

Unfortunately, and this is my third point, it is implausible that we have a morally relevant interest in maximum moral autonomy so construed, e.g., it is not as if a morally relevant interest is being served if morality gave us a moral permission to give absolute priority to our own interests over those of others. In response, it might be suggested that we have a morally relevant interest in a certain threshold amount of moral autonomy, which is compatible MAP, but incompatible with more demanding assistance principles. But if so, the metaphor of weighing breaks down. Of course, there might be other ways to flesh out the interest in moral autonomy, but I suspect that these will be problematic as well.

Conclusion

In this article, I defended three main claims. First, when it comes to the sacrifice of limbs, the correct principle of assistance might be more than moderately demanding in terms of costs. Second, something similar is true about the sacrifice of external objects when it comes to subjective costs. Third, MAP cannot be grounded in a balance between moral equality and autonomy. The wider upshot of my argument is that while Barry and Øverland do point to some weaknesses in Singer’s defense of SAP, the correct principle of assistance might in some ways be more demanding than their own preferred and less stringent principle of assistance, MAP. Hence, the obligations of rich people to assist the global poor might be more stringent than Barry and Øverland suggest.

Acknowledgments

A previous version of this paper was presented at KU Leuven, 8 June 2016. I thank Christian Barry, Siba Harb, Bashshar Haydar, Laura Valentini, Fiona Wollard, and two anonymous reviewers for helpful comments.

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

¹ All references without an author name below are to Barry and Øverland (2015).

² Barry and Øverland helpfully identify three different, non-equivalent formulations of Singer’s assistance principle (26–27). I focus on what they refer to as ‘SAP_1’. SAP₃ is less demanding than SAP₁, but still more demanding than MAP, and as Barry and Øverland argue, it is unclear exactly how demanding SAP₂ is.

³ Section 3, however, tentatively supports the view that the correct principle of assistance is at most moderately demanding in terms of psychological difficulty.

⁴ I am not asking: how can we explain the duty to save in Bob’s Bugatti and the absence of such a duty in Bob’s Internet Banking, i.e., I am not assuming that the putative moral intuitions reflect moral truths.

⁵ Given the use to which Singer wants to put his ‘child in the pond’ case – I should save the child even if I have already saved nine other children, each of whom should have been saved by one of nine other people passing by – he cannot press this line of response, since that would stand in the way of some of the duties to the global poor that he wants to attribute to us (Singer 1971, 234).

⁶ Barry and Øverland define stringency as follows: ‘A person’s responsibility to φ is stringent, in our sense, to the extent that it (1) constrains her and (2) can demand much of her’ (15–16). A duty is more stringent the more it requires the person to sacrifice.

⁷ If loss of body parts are not ‘easily or fully’ compensable, they are not irreplaceable, since the fact that full compensation is not easy to provide is consistent with its being possible to provide. As a general matter, it is false that bodily parts are irreplaceable in the relevant sense as shown by the fact, inter alia, that, given the limited set of options available, life-saving amputations are generally considered to be a net benefit to the recipient.

⁸ Bob’s Bugatti does not involve Bob’s being reduced to poverty and homelessness – he only looses ‘most of his retirement savings’, when he loses his Bugatti. If it did, I would fail to see why one should think differently of this case and Bob and the Avalanche in the first place.

⁹ Admittedly, there might be other values that do clash with the value of autonomy, e.g., the value of well-being. However, my focus here is restricted to Barry and Øverland’s proposal.