Exploiting disadvantage as causing harm

ABSTRACT In Responding to Global Poverty, Christian Barry and Gerhard Ã˜verland argue that, while exploitation is morally problematic, responsibilities not to exploit are characteristically less stringent than responsibilities not to harm. They even suggest that exploiters’ responsibilities to assist the exploited may be weaker than the responsibilities of culpable bystanders who are able to help the poor but fail to do so We think Barry and Øverland underestimate the prospects of the exploitation argument. In our paper, we suggest that exploitation can plausibly be understood as a kind of harm. If exploitation harms, then it requires special justification and can generate stringent responsibilities not to exploit that have a different ground than those generated by morally culpable failures to assist. This suggests an important way to rehabilitate arguments for poverty relief on the basis of a duty not to harm, and that there is more interesting territory to explore than Barry and Øverland’s arguments suggest.


Introduction
In Responding to Global Poverty, Christian Barry and Gerhard Øverland reject Thomas Pogge's argument that affluent countries harm the poor (Barry and Øverland. 2016). 1 Pogge claims that rich countries cause grave harm by imposing the global economic order on poor countries. This global order foreseeably fails to satisfy millions of poor people's human rights to the satisfaction of their basic needs, and there is a feasible alternative arrangement of global institutions that would dramatically reduce the number of rights violations. Since affluent countries ought to recognize all this, their imposition of the global economic order causes harm. As a result of causing harm, affluent countries have more stringent duties to remedy global unfairness than they would if they merely failed to assist the global poor. Barry and Øverland join others in claiming that Pogge's argument employs an implausibly stretched conception of causing harm. Affluent countries allow harm to befall the poor, and even enable such harm, but they do not cause harm to the poor (Patten 2005;Satz 2005). However, Barry and Øverland also suggest a way Pogge's argument can be interestingly recast. Affluent countries rely on the disadvantages of poor countries, using their superior power over those countries to secure trade agreements that unfairly favour the affluent, despite providing some benefits to poorer countries. As a result, while affluent countries do not harm the poor, they do exploit them. And perhaps this fact suffices to give affluent countries the stringent responsibilities that Pogge tries to establish. Barry and Øverland then argue that, while exploitation is morally problematic, responsibilities not to exploit are characteristically less stringent than responsibilities not to harm. 2 They even suggest that exploiters' responsibilities to assist the exploited may be weaker than the responsibilities of culpable bystanders who are able to help the poor but fail to do so (Barry and Øverland. 2016, 143-144). This means that replacing Pogge's argument with the argument about exploitation will not secure the result that Pogge initially promisednamely, responsibilities for the globally affluent that are more stringent than the responsibilities of bystanders who fail to assist the poor.
We think CB&GO underestimate the prospects of the exploitation argument. In what follows, we suggest that the exploitation which the argument accuses rich countries of can plausibly be understood as a kind of harm. If exploitation harms, then it requires special justification and can generate stringent responsibilities not to exploit that have a different ground than those generated by morally culpable failures to assist. This suggests an important possible route to rehabilitating Pogge-style arguments for poverty relief on the basis of a duty not to harm. We do not intend to establish that the exploitation argument is successful here; we are just arguing that there is more interesting territory to explore than Barry and Øverland suggest.
We begin by describing the cases of exploitation that interest Barry and Øverland ( §2), then discuss a theory of harm that regards these cases as instances of harming ( §3). Next, we argue that exploitation is generally impermissible ( §4). We end with the discussion of the stringency of the duty not to exploit ( §5) and of cases where third parties should allow exploitation despite its moral impermissibility ( §6).

Features of exploitation
Our discussion follows Barry and Øverland in focusing on cases that have certain features (Barry and Øverland. 2016, 136-139). These features are not meant to provide a general characterization of exploitation (we allow, for instance, that these features may not be necessary conditions or jointly sufficient conditions for exploitation). But, following Barry and Øverland, our interest here is in actions that have all of the following features, and our discussion of exploitation should be taken to apply to such actions.
Badness of Given Option: one party, B, is in a situation where some or all of B's (or B's dependents') basic needs will go unmet without another party's assistance. Claim to Assistance: The badness of B's given option generates a valid claim to assistance from A. Asymmetry of Vulnerability: A is not similarly in need of B's (or anyone else's) help. Unfairness: A takes advantage of the badness of B's given option to secure an agreement that favors A over B in a way that violates minimal standards of fairness (however they're to be understood). Mutual Advantage: Nonetheless, the exploitative option is mutually advantageous. Both A and B are made better-off overall when B accepts the exploitative offer than they would have been had A not made the offer, leaving B with the given option. Voluntariness: B is free to decline the exploitative option in favor of the given option, without interference or coercion from A. Availability of a Better Offer: A could offer B fair terms, while still benefiting from the agreement with B.

Exploitation as harm
Barry and Øverland's denial that exploitation causes harm focuses on its mutually advantageous character. Exploitation confers a net advantage on the exploited, so it benefits her rather than harming her (Barry and Øverland. 2016, 139-140). By contrast, we think exploitation might plausibly be regarded as a species of harm, despite its being overall beneficial for the exploited. As a result, exploitation can be appealed to in support of the claim that rich countries harm the poor. Classifying exploitation as harm can explain why the obligation not to exploit seems to be a morally serious obligation since the moral presumption against harm typically gives agents strong reasons, not to harm. 3 In this section, we will make the case for exploitation as harm. We will begin by discussing some ways that exploiters make the exploited badly-off, before discussing why this might suffice to make exploitation a kind of harm, even though it is advantageous for the exploited when compared to the option of not interacting with the exploiter.
Exploiters cause the exploited party to face a bad kind of choice. Exploited parties must decide between (1) avoiding exploitation, but allowing their basic needs (or those of their dependents) to go unsatisfied; and (2) meeting those basic needs by accepting conditions that violate minimal standards of fairness, thereby undermining their status as equals. Accepting the exploitative option demeans a person because it amounts to her volunteering for a position that treats her as a social inferior (foregoing a minimally fair wage or decent working conditions, for instance). Yet because the exploited person is in desperate need, she has reason to accept such objectionable conditions. Many (though not all) cases of ongoing exploitation also involve the imposition of a further bad state: the creation of ongoing and asymmetric forms of dependence. When exploiters limit the ability of the exploited to fairly pursue her interests, exploiters commonly foster a scenario in which the exploited must continue to seek out exploitation in order to avoid the badness of her given option. By creating this type of ongoing dependence on future exploitation, exploiters induce a form of vulnerability in the exploited that is extremely undesirable and conducive to the exploited's future domination by exploiters.
So, despite the fact that exploitation improves the position of the exploited, it also characteristically involves putting the exploited in some agency-impairing situations, situations that involve demeaning choices or ongoing asymmetric vulnerability. Whether the overall beneficial character of exploitation suffices to render it nonharmful depends on how harm is conceived.
Barry and Øverland's discussion seems to rely on a counterfactual theory of harm, according to which A's harming B requires that A makes B worse-off than she would have been in some counterfactual scenario. 4 Different counterfactual views specify different scenarios as baselines for comparison. A simple counterfactual view compares how A's actions leave B with where B would have been had A not acted (Parfit 1984, 69-70). A moralized counterfactual view compares how A's actions leave B with how B would have fared had A acted morally (Butt 2009, 102-115;Lawford-Smith 2014).
Both types of counterfactual views struggle to classify exploitation as harmful. On the simple view, exploitation benefits the exploited, rather than harming her. This straightforwardly follows from the fact that exploitation leaves the exploited better-off, overall, than she would have been without interacting with the exploiter. Moralized views seem to do better at vindicating the claim that exploitation harms, but they also face a related problem: if the exploiter could permissibly refuse to interact with the exploited, then exploitation will make the exploited party better-off than she would have been in at least one scenario where the exploiter acts permissibly.
To classify exploitation as harm, moralized views need to explain why non-interaction is not a permissible option or why, despite being permissible, non-interaction is not the relevant counterfactual scenario to compare B's outcomes against. One feature of exploitation as discussed by Barry and Øverland is that the exploited has a claim of assistance against the exploiter. This might explain why non-interaction is impermissible in the cases of interest. But, presumably, many other individuals have similarly strong claims to assistance. And the exploiter's duty to take on the cost to help those in need could be discharged by helping some of those others rather than the exploited. So, it seems that exploiters might be able to harmlessly exploit some, provided they provide assistance to others. Without a solution to this problem, moralized views will follow the simple view in classifying exploitation as a benefit, rather than harm, at least in some cases. 5 Non-comparative theories of harm can more easily accommodate the claim that exploitation harms. These theories require only that A's actions cause B to be in a certain type of bad state. Non-comparative theories allow that an act can leave agents better-off, overall, while also harming them (For non-comparative theories of harm see Harman 2009;Shiffrin 1999). For instance, cosmetic surgery might harm someone by physically injuring him and causing him pain, while also benefiting him, by improving his satisfaction with his appearance. Even if he is made better-off overall by the surgery, he has been harmed inasmuch as the surgeon's cuts cause him to endure pain and cause damage to his body (assuming, plausibly, that pain and bodily damage are among the states that a non-comparative theory identifies as harmful). This explains why there is a moral presumption against non-consensually performing the surgery, and why people who non-consensually impose these kinds of bad states on others have strong reasons to rectify the situation they have imposed.
Advocates of non-comparative views must explain which bad states are relevant to harm. Plausible accounts will typically regard the outcomes of exploitation discussed above to be among the set of bad-making traits. For instance, Seana Shiffrin's theory, which regards agency-impairing states as the class of harmful states, will vindicate the claim that causing the bads of exploitation is harming (Shiffrin 1999, 123-124). Elizabeth Harman's view, which includes significant forms of 'mental discomfort' (and presumably other forms of psychological dysfunction or stress) among the bads will also register exploitation as harm, inasmuch as the bads described above are accompanied by bad experiences for the exploited (as they typically are) (Harman 2009, 139). If non-comparative theories are correct, then it seems very likely that exploitation will count as harm.
Various pluralist theories of harm are also possible. For instance, a theory might allow for both comparative and non-comparative harms. Or it might define all harm counterfactually, but allow that different contexts make different counterfactual baselines relevant (Tadros 2014). Since the first type of pluralist view incorporates the noncomparative view, it too will classify exploitation as a kind of harm. The second view will allow us to classify exploitation as harm provided that it can answer the challenge that the moralized view facesnamely, of explaining why fair terms of cooperation, rather than non-interaction, is the relevant baseline for assessing whether exploiters harm. But it may have more resources to answer that challenge than the more straightforward moralized views discussed above do.
For the remainder of the paper, we will assume the correct theory of harm is noncomparative. However, we think that much of what we say will apply to other accounts of harm, provided they classify exploitation as harming, in light of its imposition of demeaning choices or asymmetric vulnerability.
Even if exploitation does harm in these ways, it may seem that the kind of harm we discuss could only arise at the interpersonal level. By contrast, Barry and Øverland are principally interested in the discussion of interactions between countries and their representatives negotiating terms of global trade. But we believe that the harms in question can arise in the case of international trade and other forms of international negotiation and cooperation. One reason is that it may make sense to think of countries or peoples being put into the same kinds of harmful conditions that individual persons are. If we take seriously (as some political philosophers do) the idea that states or the peoples in whose name they govern, can be collective moral agents with interests in freedom and respect, then the parallel between interpersonal exploitation and international exploitation is fairly straightforward (Rawls 2001, pt. 1 §2).
More importantly, the terms agreed upon by national representatives will have effects on the options available to their citizens. This can foreseeably lead to those citizens facing constrained choices in their own lives, and to them being maintained in conditions of asymmetric and significant dependence. When affluent countries make choices that predictably lead to those conditions falling on individual citizens of poor countries, affluent countries cause the same exploitative harm to those individuals. This is true even when the individuals benefit overall from the options that are made available to them by their country's economic cooperation with affluent states.
So far, we have considered the claim that exploitation harms because it imposes demeaning choices and asymmetric vulnerability. We have argued that this claim can be easily supported by non-comparative theories of harm, while noting that the claim might also be sustained by other theories, including counterfactual ones. 6 Applied to the case of global trade, this yields the claim that when rich countries trade with poor countries on exploitative terms, the imposition harms them or their citizens. It seems, then, that the exploitation-based rehabilitation of Pogge's argument may be worthy of greater consideration than Barry and Øverland suggest. We now consider some implications of that argument.

Wrongfulness of exploitation
If exploitation causes harm, then exploitation requires justification to establish its permissibility; this reflects the moral presumption against harm. One possible explanation of how exploitation could meet this justificatory burden is that it protects the harmed party from the greater harm of having her basic needs go unmet. Exploitation is permissible, on this account, because the harm of basic needs deprivation is worse than the harm associated with accepting the exploitative option. We think this explanation fails.
It can be permissible to impose harm in order to prevent greater harm. For instance, if Truvi can only rescue Lucian from drowning by breaking his arm as she pulls him from the river, she may do so. 7 It is tempting to think of exploitative offers as imposing this kind of harm-in-order-to-rescue. Exploiters impose harm that gives the exploited the opportunity to meet basic needs that would otherwise go unmet. But there is an important difference between the two cases. While in both the rescue and the exploitation cases imposing a lesser harm results in the prevention of more serious harm, only in the rescue case is the lesser harm necessary to prevent the greater harm. Truvi could not save Lucian without breaking his arm. It is only when the lesser harm is necessary to prevent the greater harm that the imposition of the lesser harm can be permissible. The offer imposed by exploiters is not necessary to rescue the exploited from the harms associated with unmet basic needs. Exploiters could offer a fairer option to the exploited that would still leave both parties advantaged by the interaction. As a result, exploitation is objectionable, despite the fact that it protects the exploited from basic needs deprivation.
Exploiters may claim that their status-undermining offers are in fact necessary to save the exploited from destitution. Suppose Layla will only hire Adam if it allows her to generate a very large return on her investment. And suppose that such a return is achievable only when Adam is hired at a wage that only persons in desperate need would accept, which Adam is in. Suppose further that if Layla does not hire Adam, no one else will offer Adam any employment or assistance. Given Layla's unwillingness to settle for less than a very large return, the exploitative offer is, in some sense, necessary for the prevention of the harm of basic needs deprivation. But this does not make Layla's exploitation permissible, because it is Layla who makes it the case that the exploitation is necessary. She could still benefit while settling for lower profit. 8 It is important to note here that exploiters' assistance-based responsibilities are not doing the normative work in our argument for the impermissibility of exploitation. 9 To reiterate, on our account, the wrongfulness of exploitation is in the imposition of unnecessary harm. Any such imposition is wrongful, even when the rescuer does not have assistance-based responsibility to rescue (say because they have already fulfilled their assistance-based responsibilities by helping other victims). Take the exploitative hiring case above. Suppose that Layla had already discharged her assistance-based responsibilities (she is a patron of a charity against Malaria). Despite this, it would be impermissible for her to make Adam an exploitative offer. Doing so would impose a demeaning and seriously unfair offer on Adam, which is a way of harming him that is not necessary for helping him.
In short, our view implies that Layla can permissibly hire Adam on non-exploitative terms or not hire him. Hiring Adam on blatantly unfair terms is impermissible. This may seem counter-intuitive since not hiring Adam is worse for him than hiring him on exploitative terms. Despite this, we think ours is a plausible view to maintain. For one, permissibility can come apart from what is a better or a worse outcome. The fact that Layla would rather not hire Adam than settle for lower profit is unfortunate but does not render exploitation permissible. Secondly, the fact that Adam would prefer the exploitative option over the morally permissible option of not being hired, while troubling, does not invalidate our argument. An action may impermissibly harm someone even if the harmed person would not prefer the action not to have taken place. 10 Thirdly, as we argue in the last section, what is impermissible for particular agents may not be impermissible for third parties, who may need to enable or permit exploitation by others in order to alleviate poverty.
A different concern with our argument is that it ignores the consent of the exploited. They choose exploitation over the badness of their given option. And consent often renders harm permissible, especially when that harm confers benefit or protection from another harm (as in the case of cosmetic surgery discussed above). One response to this is to deny that the exploited have consented in a morally meaningful waythe conditions of their choice are insufficiently free to give rise to authoritative consent (Olsaretti 1998). But even if one accepts that the exploited have properly consented to the exploitative option they are offered, they have not consented to the unfair choice set that exploiters impose. The harm of exploitation is in the imposition of this choice. Given the choice between the exploitative choice set and a set of options that included fair terms of cooperation, the exploited would clearly opt for the latter. But despite the availability of better options, exploiters see to it that the exploited face a set of options limited to their given option and an exploitative option.

Exploiters' responsibilities
Supposing that exploitation is a harm, we have argued it is impermissible since it is not necessary to alleviate the greater harm of unmet basic needs, suffered by the exploited. What implications does this have for exploiters' responsibilities? Recall that Barry and Øverland think an exploitation-based argument would fail to vindicate Pogge's claim that affluent countries have stringent obligations not to impose unfair global cooperation schemes. On Barry and Øverland's account, the duty not to exploit may even be weaker than the responsibilities of culpable bystanders to assist. As a result, they maintain that the responsibilities of the affluent are best captured as duties to assist.
If exploitation harms, then the duty not to exploit has the characteristic stringency of the duty not to harm. This means that those in a position to exploit ought to take on significant cost not to harm-by-exploiting and that the duty not to exploit can be enforceable by third parties. 11 To see the stringency of the duty not to exploit compare it with the duties of culpable bystanders. Consider two affluent countries, A and B, which are similarly situated and neither of which have fulfilled their duties of assistance to a poor country, C. A and B both have a choice between three options in bilateral trade deals with C: leaving C to suffer through its given option (No Trade); exploiting C, resulting in thousands less deaths than under the given option (Unfair Trade); or giving C fair and favorable terms of cooperation that result in many more lives saved than under either of the previous alternatives (Good Trade). 12 Suppose that A chooses Unfair Trade while B chooses No Trade. A's action results in more lives saved than B's. It may seem straightforwardly implausible to think that A's responsibilities towards C are more demanding than B's; as a result, it may be tempting to conclude that exploiters have less demanding responsibilities than culpable bystanders. This may be Barry and Øverland's view, since they sometimes suggest the benefits which exploiters confer on the exploited might count against the exploiters' duties of assistance (Barry and Øverland. 2016, 143-144). If that were true, culpable bystanders would have more stringent duties to address victims of global poverty than exploiters do, since exploiters have assisted the exploited through the beneficial transaction.
Despite the apparent appeal of this line of thought, we believe it is mistaken, provided that exploitation harms. It seems morally perverse to both condemn an act as impermissible because it harms someone while counting the benefits the act confers on the same person towards the harmer's duties to assist others (including that person).
Exploitation can sometimes lessen the amount that exploiters owe to particular exploited parties. By making the exploited better-off, exploitation can reduce her need for assistance. But while this can lessen how much help particular individuals are owed by an exploiter (or by any other third party), it does not reduce the amount of overall assistance that exploiters are obliged to provide. Once exploited parties have been assisted, exploiters are obliged to help other people in need, and they are not entitled to offer less help to those people because their exploitation improved a lot of the exploited. So, while benefits to the exploited might shift where people are obliged to assist, they should not count against the help that exploiters are obliged to offer to parties in need.

Third parties' responsibilities
One troubling possible implication of our view is that it may suggest that third parties with the ability and responsibility to influence whether others engage in exploitation (henceforth 'regulators') should prevent exploitation, even when doing so is contrary to the interest of the global poor. If regulators can force the morally required option, good trade, then our view permits them to do so, provided the enforcement satisfies familiar constraints of necessity and proportionality. But our view seems to run into trouble when they are not able to bring about the morally required option. For instance, it might suggest that the regulator should force a globally affluent country to choose No Trade, where no harm is done, over Unfair Trade, where the affluent harm by exploiting. This would implausibly suggest that regulators should deprive the exploited of opportunities to improve their livelihood, in the name of the exploited's own dignity.
Our view does not have this implication. To see why, suppose that a regulator cannot cause the affluent to accept Good Trade, where no harm is done and the most benefit is conferred upon the disadvantaged parties. All the regulator can influence is whether the poor end up in the exploitative option, Unfair Trade, or the worse given option, No Trade. By causing the poor to end up in Unfair Trade, the regulator contributes to harm. But this harm is necessary to avert greater harmthe one which will result from the poor parties' being forced to endure the given option (worse poverty). And, as we discussed above, it is typically morally permissible to harm, when doing so is required to avert greater harm. So, while the affluent may not exploit, third parties may permitor even encourageexploitation, when doing so is required to prevent greater harms to the seriously disadvantaged.

Conclusion
We think that common sense suggests exploitation is wrongful, and that exploitation harms the exploited, even if it can leave her better-off overall. Drawing on a suggestion from Barry and Øverland, we have tried to sketch a way to make sense of that idea. We have also begun thinking through its implications for the responsibilities of affluent countries that exploit, and third parties that can affect whether exploitation comes about. Obviously, a great deal more would need to be said to develop a full theory of exploitation as harm, and to determine whether it applies to the case of the imposition of the terms of global trade. The issues discussed in Barry and Øverland's book provide a fruitful framework for beginning to explore such a theory.