Reassessing the Exploitation Charge in Sweatshop Labor

2.1.Wertheimer’s theory of exploitation

Alan Wertheimer maintains that a transaction between two parties is unfair if the way they share the surplus conflicts with the shares that the parties would have had there been a hypothetical competitive market for the goods and services they transact (Wertheimer 1996: 230–236). Wertheimer’s account of fairness is transaction-specific and does not consider how well-off the agents are in comparison to each other apart from the terms of the transaction itself.

Many authors have criticized this theory for using competitive markets as fairness criteria. Ruth Sample, for instance, finds Wertheimer’s criterion conservative and questions his claim that exploitation consists of violating market practices: “Wertheimer does not explain why the market price should be regarded as the fair price and why nonmarket prices are exploitative” (2003: 24). Moreover, even if we accepted the fair market price as a criterion to detect exploitation, we would find cases where talking about such a price does not make sense.

For example, take a porcelain collector who lacks only one piece in his extensive collection of rare pieces. This piece stands in a store whose owner raises its price after hearing about the interest of this collector. The collector wants the piece so badly that he buys it for the new exorbitant price.3 He seems overcharged because the store owner raised the price after learning about the collector’s interest in this piece. However, to talk about exploitation, in this case, would require us to imagine the piece’s price in a hypothetical competitive market.

In such a market, the piece would have the price “that an informed and unpressured seller would receive from an informed and unpressured buyer” (Wertheimer 1996: 230). There are multiple buyers and sellers in such a hypothetical competitive market to ensure that “neither party takes special unfair advantage of the particular defects in the other party’s decision-making capacity or special vulnerabilities in the other party’s situation” (Wertheimer 1996: 232, emphasis in the original). However, if there were multiple buyers and sellers of this porcelain piece, then there might be less of an incentive to make a collection of it. After all, people who are into collections tend to collect rare items, for which, by definition of a rare item, there cannot be multiple buyers and sellers. Hence, imagining a hypothetical competitive market for such items

is difficult.

To be charitable to Wertheimer’s definition, we can ignore the clause on “multiple buyers and sellers” and focus on the part where he talks of taking special unfair advantage of special vulnerabilities in the other party’s situation (cf. 1996: 232). We can blame the store owner for taking a special unfair advantage, thus exploiting the collector when they use the information that reveals this collector’s vulnerability.

However, unless the store owner’s act also involves unfairness, it would bear the problems of an account of exploitation based on the idea of vulnerability. I explain such an account as (E2) and discuss its disadvantages below. Otherwise, if it involves unfairness, it has to account for how to come up with a hypothetical price for the porcelain piece in question.

It is, then, unclear whether the porcelain collector is exploited, as per Wertheimer. It is even dubious whether he is overcharged because we need a standard price to talk about overcharging, which is nonexistent in this case. He might well “feel” overcharged, yet this does not offer much help for the normative analysis I am pursuing here.

The flaw of Wertheimer’s account of exploitation in explaining the moral issues in exchanging rare items might disclose a theoretical weakness on its side. Nevertheless, we still have to examine the explanatory power of this theory to analyze the exploitation charge in sweatshops.

Determining a hypothetical market price for sweatshop labor is difficult. One difficulty stems from how hypothetical we imagine this market to be. In other words, we must find out how much of a deviation from an actual market situation in a given region of sweatshops we want to imagine.

If we were to imagine a hypothetical market on a global scale, say for just the garment industry, keeping other factors constant, we would imagine a global labor market of workers and capitalists from all countries worldwide. Under such circumstances, other things equal, one would expect the more densely populated regions to have lower market wages than more sparsely populated regions since the labor supply is higher in these densely populated regions. This situation might result in a neighborhood in Manhattan having a lower hypothetical market wage for the garment industry than Brahmanbaria, a city in Bangladesh similar in size to Manhattan but with a population of at least one-tenth. The counterintuitive implication would be that while workers in Brahmanbaria are exploited with the current real wages they receive, the workers in Manhattan exploit their employers.

Holding everything else stable and focusing on population density to imagine a hypothetical market might not do justice to what Wertheimer had in mind. However, the point of the example is to mark the difficulty of imagining a hypothetical market for sweatshop labor globally.

Sample maintains that Wertheimer’s account would fail to distinguish the wage-labor exchange between the MNE decision-makers and the sweatshop workers as exploitative. According to her, Wertheimer’s account would fail in these cases because “[t]here is a competitive market for labor in Pacific Rim countries, but there are more workers than there are capitalists. Thus the competitive market price for labor is relatively low” (Sample 2003: 24).

Sample’s claim might require more explanation here. The fact that there are more workers than capitalists should not suffice to say that the relevant labor relation is exploitative unless one believes that a labor-capital relationship is necessarily exploitative since the number of workers is always more than the number of capitalists in any capitalist part of the world. We could interpret Sample’s idea to mean that the capitalist/worker ratio in the Rim countries is below a certain ratio, above which it would not make sense to talk of an exploitative labor relationship according to Wertheimer’s hypothetical market price criterion for exploitation.

Cohen’s idea of “collective unfreedom” can help us here understand the significance of this capitalist/worker ratio.4 Cohen explains collective unfreedom as follows: “a group suffers collective unfreedom with respect to a type of action A if and only if performance of A by all members of the group is impossible” (Cohen 1983: 16). He maintains that although each worker in a capitalist portion of the world is free to stop being a worker and become an entrepreneur using their savings or other loans, if possible, such freedom cannot be used by all workers. Thus, workers suffer from collective unfreedom to stop being workers.

Cohen adds, “[c]ollective unfreedom comes in varying amounts, and it is greater the smaller the ratio of the maximum that could perform A to the total number in the group” (1983: 16). One would expect this ratio to be quite small in a region where sweatshops are widely used in production. So, we would not expect many workers to be free to stop being workers and start their own sweatshops. Hence, workers would suffer from a higher level of collective unfreedom in such a region.

Now, we can take Sample’s idea of the capitalist/worker ratio to follow Cohen’s idea of collective unfreedom. So, as Cohen advocates, the lower the capitalist/worker ratio in any given region, the more exploitative conditions are in that place. However, unless we want to accept that all possible paid labor is exploitative, we need to mark a threshold for the capitalist/worker ratio at which labor relations become exploitative. Therefore, we still need to say more than what Cohen theorizes about workers’ collective unfreedom.

Sample might be wrong in her assumption that there is a problem with Wertheimer’s theory just because it fails to mark sweatshop labor as exploitative. Still, her criticism gives us a new idea for interpreting the criterion of a hypothetical market. We can now take the hypothetical market to imply a given region rather than the whole world and the hypothetical market wage as the wage resulting from an “ideal capitalist/worker ratio” in this region. This idea would be a better interpretation of Wertheimer’s theory.

This result is still counterintuitive, according to Jeremy Snyder. He argues against a hypothetical market price definition of exploitation: “An interaction may be fair by the standards of a hypothetical fair market (or another standard of micro fairness), but leave workers without sufficient income to meet their basic human needs” (Snyder 2010: 199). He concludes that the fact that workers cannot meet their basic human needs in the presence of a hypothetical market price for labor is sufficient to reject Wertheimer’s theory of exploitation.

Snyder is correct that a hypothetical market wage might leave workers with insufficient provisions. However, there would be nothing inconsistent in Wertheimer’s theory to say that this wage is non-exploitative. Wertheimer could still concede that other moral wrongs might be involved (like a history of colonization or a corrupt government) or some unfortunate events (like famine) that caused the circumstances under which a hypothetical market wage is so low. Thus, he would insist that no exploitation is involved as long as the wage in question is at a level commanded by a hypothetical labor market in this region.

Let me go back to the interpretation of the hypothetical market wage that depends on the idea of an “ideal capitalist/worker ratio.” The theory has to hold that a hypothetical market wage would be obtained if such a ratio were maintained in a given region. This ratio ensures that an unpressured seller takes a price from an unpressured buyer. The difficulty of determining such a ratio is evident. If we point at a particular country or region of the world as an ideal example of such a ratio because there is no exploitation there, this will beg the question. We can define exploitation by looking at a given non-exploitative relationship only if we accept this particular non-exploitative case as a foundation for our theory. This argument is different from what Wertheimer defends.

Even if we could determine an ideal capitalist/worker ratio by looking at a particular part of the world to help us determine a non-exploitative hypothetical market wage, we would not be able to use it in other cases. Imagine that in a specific region where living conditions are harsh, power plant companies struggle to employ engineers because engineers prefer to live in parts of the world where living conditions are better than in this region. If a capable engineer asks for an exorbitant wage from a power plant company in such a region, we could consistently hold that this is an exploitative offer because the wage this engineer demands would be higher than the average wage engineers receive in the part of the world where the capitalist/worker ratio is ideal.5 However, we would then lose Wertheimer’s condition of a hypothetical market: an unpressured seller takes a price from an unpressured buyer. This result would constitute another reason we cannot take an existing location to indicate an ideal capitalist/worker ratio.

Another method would be giving another criterion, such as workers’ capacity to meet their basic needs, to indicate this ratio. However, now, there would not be any need for the hypothetical market criterion because this new criterion, i.e., meeting the basic needs, would do the conceptual work. Thus, the hypothetical market criterion for determining fairness does not help clarify Wertheimer’s exploitation theory.

Indeed, Wertheimer provides his readers with an alternative way to interpret fairness in a transaction, although he refuses to use it to develop this position further. According to this interpretation, a transaction between two parties is unfair if the way the surplus is shared conflicts with the shares that the parties would end up having, according to the rational bargaining theory (Wertheimer 1996: 218–221).

Sollars and Englander develop an account of fairness based on this idea. They define the reservation prices of the seller and buyer of a product as R _S and R _B, respectively. R _S is the price below which the seller would not wish to transact, and R _B is the price above which the buyer would decline the transaction. For a transaction to occur, R _S has to be less than or equal to R _B; R _S ≤ R _B. As a result of any transaction, a surplus will be defined by their difference, R _B – R _S. According to their theory, a fair transaction would divide the surplus equally; (R _B – R _S)/2. Any deviation from this amount would make the transaction exploitative.6

Their exploitation theory helps them distinguish between a company that makes enormous profits using sweatshops in their production line and another case where a relatively small domestic company uses them. In both cases, workers might find it challenging to make ends meet. Nevertheless, while we may rightfully blame the big company for exploiting its workers, we might not say the same for the small company.

The difference stems from the surplus ratio (R _B – R _S). The big company is thought to have a higher R _B, so it is expected to yield a higher wage for its workers. If this company does not share half of the surplus with its workers, it is charged with exploitation. The small company could also be charged with exploitation unless it shares half of the surplus. However, even if the small company renounces all the surplus favoring its workers, this might make a minimal change in workers’ wage levels. So, although the workers of the big company might make more money than those of the small company, the former are exploited while the latter are not, according to this theory (cf. Sollars and Englander 2018).

This account seems more stable than the hypothetical market criterion for fairness. Nevertheless, the problem lies in the initial presuppositions of the theory. Rather, indeed, the problem lies in what the initial presuppositions neglect. Rational bargaining theory neglects the background circumstances that affect the reservation prices of the transacting parties. As Wertheimer contends, “[...] rational bargaining always reflects the prebargaining position or endowments of the parties [...]” (Wertheimer 1996: 220). This theory ignores why sweatshop workers have a reservation price that is insufficient to fulfill their basic human needs.

A defender of Wertheimer’s theory of exploitation could respond that those background circumstances enable a transaction between large MNEs and poor sweatshop workers in the first place, and thus including these circumstances in a theory of exploitation to mark them morally wrong would be ignoring the basic requirements for foreign direct investment. However, this rejoinder misses the point of the need for a theory of exploitation that purports to explain a mutually advantageous albeit morally wrong relationship.

If a mutually advantageous relationship involves unjust background circumstances in which the parties find themselves, a theory of exploitation is expected, at least not to ignore them. I do not mean to defend that any transaction made under unjust background conditions is necessarily exploitative. A theory of exploitation must explain why and when they are irrelevant if that is what the theory claims. Unfortunately, Wertheimer’s theory of exploitation falls short of this research goal.

This last response would only satisfy some, and I wish not to push it further since I do not necessarily need it to make my main argument. So, regarding Wertheimer’s theory, I can say that until a better account of hypothetical market criterion is given, his theory works best with the approach suggested by Sollars and Englander, and their method concedes that we can rightfully blame some sweatshops, viz., some of those that make big profits, for wrongfully exploiting their workers. Moreover, many other sweatshops, especially ones that do not make huge profits, will be marked as non-exploitative in Wertheimer’s approach.

2.2. Improving the fairness account using the vulnerability criterion

One way to improve (E1) is to add the “vulnerability” element to this definition to better distinguish cases like the porcelain collector from ones like sweatshops. This amendment would enable us to consider the background conditions and filter out cases in which the alleged exploitee is not necessarily vulnerable in a significant way to the transaction in question. The new definition of exploitation would look like this:

Robert Goodin and Thomas Christiano both have theories of exploitation that one can call vulnerability accounts. Goodin claims that “flagrant violation of duty to protect the vulnerable constitutes the essence of interpersonal exploitation” (Goodin 1987: 188). Moreover, Christiano defines exploitation as a violation of a duty to the vulnerable (Christiano 2015: 263). Both authors emphasize the vulnerability of the exploitee to the exploiter.

The general criticism against vulnerability accounts is that they would create false positives, viz., mark many of our ordinary transactions as exploitative. For instance, Richard Arneson argues that there is nothing objectionable in using someone’s vulnerability for advantage as long as there is no unfairness involved in the interaction (Arneson 2016: 10). After all, our specific vulnerabilities and necessities enable many of the transactions in which we get involved. For example, in Arneson’s “Cancer Treatment” example, a cancer patient consults the only qualified surgeon in his town (Arneson 2016: 10). The surgeon offers to operate on the patient and save his life for a better-than-fair price. In this example, although the patient is vulnerable to the surgeon, the interaction is not exploitative. Of course, one might argue that this is what the Goodin-Christiano line is arguing for, i.e., the surgeon is not violating her duty of making a fair offer to the vulnerable patient. Again, however, we would fall back to the fairness account, which would explain why this transaction is not exploitative.

It seems plausible to combine the unfairness element with the vulnerability criterion to benefit from both criteria’s strengths.

It might seem evident that if A can exploit B, then something about B’s situation already creates vulnerability on her account. Hence, adding vulnerability to the definition of exploitation might seem redundant. However, talking about an exploitative relationship for some types of vulnerabilities would not make sense. For example, assume I sell chess lessons online and get paid in terms of donation, viz., my customers pay me whatever they deem suitable as a fair price.8 Now and then, some customers would pay a meager price or not pay at all while benefiting from the lessons. The situation fits (E3) well; I seem to be exploited by these customers. In this case, although I have made myself vulnerable to unfair transactions, I am not vulnerable in any special sense or any other sense regarding the background conditions of the parties involved. So, adding vulnerability as a separate criterion does not add much to the exploitation charge unless the definition of vulnerability is related to some background conditions.

Satz seems to agree that “ underlying extreme vulnerabilities of the transacting parties” (Satz 2010: 97, emphasis in the original) rather than any random vulnerability can lead to exploitation. She adds that “widely varying resources or widely different capacities to understand the terms of their transaction”” (Satz 2010: 97) cause such extreme vulnerabilities.9 Therefore, the exploitation account must build the connection between the exploitee’s background conditions and vulnerability to avoid false positives.

2.3. Arneson’s theory of exploitation

Arneson gives an account of exploitation that would fit (E1) and include the background circumstances into the definition of unfairness. He formulates a prioritarian criterion of the best outcome regarding the distribution of social surplus in an interaction. He holds that the best outcome is obtained by measuring the weighted well-being of the parties involved in the transaction in question. Moreover, the weighted well-being score “is fixed in this way: obtaining a benefit (or avoiding a loss) for a person has more value, (1) the greater the well-being gain it achieves for the person, (2) the worse-off in lifetime well-being the person would otherwise be absent this benefit, and (3) the more deserving the person is in life-time terms” (Arneson 2016: 17). So, a transaction between parties A and B would be less fair the farther it is from the best outcome criterion that Arneson provides.

Arneson’s theory of exploitation is superior to Wertheimer’s in including the lifetime well-being of the parties when making an exploitation claim.10 This criterion of how agents would do in lifetime well-being absent the transaction in question would imply a comparison in background circumstances. “This account will yield the result that Poor’s driving hard bargain with Rich when Poor has a bargaining advantage is more fair than Rich’s driving hard bargain with Poor when Rich has a bargaining advantage (on the assumption that greater wealth tends to lead to greater lifetime well-being)” (Arneson 2016: 18).

Nicholas Vrousalis criticizes Arneson for creating false negatives, i.e., missing to mark exploitative cases as exploitative. The case Vrousalis uses as a counterexample to Arneson’s account is a version of Ant-Grasshopper cases.11 In this version, although the Grasshopper is much worse-off compared to the Ant, she is undeserving to enjoy the benefits of their interaction because she spent all summer lazing around, and this fact makes her “completely undeserving (absolutely or comparatively, pluralistically or monistically)” (Vrousalis 2016: 535). Hence, according to Arneson’s theory of exploitation, even if Ant charges the Grasshopper exorbitantly, she is not accused of exploitation because of Arneson’s account’s “desert” element.

Vrousalis also criticizes Arneson’s account of exploitation for creating false positives, i.e., marking non-exploitative cases as exploitative. He compares cases of unfair free-riding that do not include domination to those that include domination (Vrousalis 2016: 536). Vrousalis maintains that Arneson’s account marks these cases as exploitative, although they are intuitively not exploitative. He gives the example of someone, A, who escapes from persecution and hides in B’s boat. When B rows her boat from one coast to another, A free-rides B’s rowing efforts and hence takes unfair advantage of her for A’s benefit. Vrousalis maintains that although A seems to be exploiting B according to Arneson’s unfairness account, this is intuitively wrong.

Vrousalis gives another example where some villagers take unfair advantage of other villagers while at the same time dominating them. In this thought experiment, villagers take turns to stand sentry at the village’s gates against bandits. Some villagers refuse to serve in the scheme, though, and thus free-ride the efforts of others. They do so knowing that the villagers who live close to the village’s periphery will suffer the most when bandits attack. To defend the village safely, those contributing to the sentry scheme need the free riders’ contribution. So, the free riders have power over the contributors. Vrousalis holds that in this example, the free riders can be rightly accused of exploiting the contributors’ efforts to maintain safety.

Vrousalis also argues that even if Arneson were right that these free rider cases were both exploitative, something of normative significance would be lost by not distinguishing them from each other: “Only power-grounded advantage-taking constitutes exploitation, on this view” (Vrousalis 2016: 536).

Putting aside its theoretical handicaps, using Arneson’s prioritarian criterion of fairness in (E1) must be tested to see if it renders sweatshops exploitative. This fairness account seeks to achieve the best weighted-well-being score in a transaction. For example, compare two possible wage levels that an MNE can pay to its workers in a sweatshop that it runs, w and w + x, where w is the current wage paid, and w + x is the wage that the MNE can pay without causing much of a change in other aspects of its budget.12 Paying w + x to the workers would provide higher marginal well-being than the marginal well-being that paying w to workers would give to higher-level managers and consumers who would otherwise buy the same products slightly cheaper.

This calculation is accurate, at least in the short term. However, in the long term, paying w + x at the cost of lower manager compensation packets and a slightly higher product price may pull down wages paid to sweatshop workers. If further empirical work shows that this is the case in the sweatshops in question, then the first criterion of Arneson’s account would not agree that paying w + x creates greater well-being for workers. Whether the first criterion of Arneson’s prioritarian fairness account supports paying w + x to sweatshop workers depends on empirical data relevant to the particular circumstances in question.

The second criterion seeks the distribution that will provide the best well-being in life-time terms. In line with the first criterion, sweatshop workers would lose much more than high-level managers and customers in general, and in lifetime terms, if w is paid to the workers instead of w + x.13 This scenario would be valid unless the long-term consequences of higher worker wages drive profits down to drive the MNE in question down in the competition among the regional companies. Such a consequence would result in many sweatshop workers losing their jobs and thus having less well-being in lifetime terms. Hence, whether Arneson’s account’s second prioritarian criterion supports higher worker wages also depends on the pertinent empirical data.

The third criterion is more challenging to add to the calculation. This difficulty would be due to the ambiguity of comparing the “desert” element in this account of fairness to the other two criteria. The main determining factor in the Marxist and liberal divide in the approach to exploitation stems from their economic approaches to the calculation of desert.14 According to the liberal approach to desert, the supply and demand curves of labor power determine the wage laborers deserve in that particular market. So, according to the liberal approach, if the supply and demand curves depict wage w as the equilibrium point, then the workers deserve to be paid w instead of w + x. Alternatively, if there is room in the liberal theory of desert, as some authors claim, to include w + x in the deserved wage interval, then the third criterion would hold that the fair wage for sweatshop workers is w + x and not w.15

As a result, according to Arneson’s theory of exploitation, whether sweatshop labor is exploitative depends on empirical evidence. There is evidence in both directions.16 Some evidence shows that paying higher wages to sweatshop workers will lead to worker layoffs in that sweatshop. However, even in this case, Michael Kates argues that this is not necessarily harmful to less developed countries’ sweatshop workers overall. According to him, some layoffs due to wage increases might bring greater benefits for the less developed countries’ sweatshop workers (cf. Kates 2015).

There is empirical evidence favoring higher wages paid to sweatshop workers, leading to their greater well-being. The same evidence favors the idea that the overall population working in sweatshops would be worse off in a lifetime without a higher wage. These two premises are pertinent to Arneson’s first two criteria in his prioritarian approach to exploitation. Moreover, even theoreticians working within the limits of classical liberal theory concede that some increase in the market-determined wage level is possible. So even if we accept the liberal economic criterion of desert for the third element of Arneson’s theory of exploitation, the theory concedes that paying a wage level (w) below the economically feasible maximum (w+x) would plausibly assign the charge of exploitation to the MNE managers.

Therefore, there is some support in the relevant literature to claim that Arneson’s account of exploitation would find some standard cases of sweatshops to be exploitative while marking others as non-exploitative.

3.1 Sample’s theory of exploitation

Sample presents her attitude-based account of exploitation as follows: “The basic idea is that exploitation involves interacting with another being for the sake of advantage in a way that degrades or fails to respect the inherent value in that being” (2003: 57). So, according to her account, A exploits B if and only if A benefits from interaction with B while degrading or disrespecting B.

She provides three possible ways A can disrespect B: 1) A can fail to respect B by neglecting what is necessary for B’s well-being or flourishing, 2) A can fail to respect B by taking advantage of an injustice done to B, and 3) A can fail to respect B by commodifying or treating as a fungible object of market exchange, an aspect of B’s being that ought not to be commodified (Sample 2003: 57).

Sample asserts that respect for other persons does not require us to love them, and in this sense, respect is a limited relationship with others. Just as respect requires us to engage in a limited but positive manner, she argues; the duty to refrain from exploiting others requires us to constrain ourselves in specific ways when interacting with them. So it is not enough not to harm the person we interact with; ignoring their needs is also as disrespectful as harming them. Such an act of ignoring the needs of others can be exploitative if we also benefit from the interaction, Sample maintains. Moreover, this is why she argues that exploitation is worse than neglect: We fail to fulfill our duty to respect the inherent value of our interactor when we are involved in exploitation, even if it is mutually beneficial.

Ignoring the basic needs of someone whose needs are at stake is acting disrespectfully towards that person, holds Sample.17 So if A chooses to interact with B and B’s basic needs are at stake, B is vulnerable to A. From this moment on, if A does not meet B’s basic needs, A exploits B when A also benefits from the interaction. Sample admits that if, in the only possible mutually advantageous interaction, the basic needs of B cannot be met but are taken into account, then the interaction is not exploitative (2003: 75).18

The needs of people are related to their well-being, according to Sample. Therefore, when interacting with someone vulnerable, we must consider their well-being. The understanding of well-being that Sample bases her account of exploitation on is the capabilities approach advocated by Martha Nussbaum. Hence “exploitative interactions are those in which the capabilities of our interactors are ignored in the pursuit of our own advantage” (Sample 2003: 81). Still, “nonexploitation does not require that we ensure the capabilities of our interactors in individual transactions. Rather it requires that we in some way take their needs into account” (Sample 2003: 81). Hence, Sample reminds us that not ignoring the capabilities of our interactors means taking their needs into account.19

Such is different for repeated transactions, however. Sample argues that in the case of repeated transactions, where A is taking advantage of B and B is vulnerable to A, B’s capabilities must be ensured: “If an employer fails to compensate an employee in a way that provides her with adequate income when such compensation is possible, then the relationship is exploitative” (Sample 2003: 81).

Sample does not cash out the exact circumstances under which such compensation is possible for the employer. Thus, her account needs to be more explicit about how much compensation an employer of a local sweatshop that makes quite a small profit owes to their workers. It is more evident, though, that managers of MNEs ought to fulfill the capabilities criterion of Sample’s account. If such compensation were not possible for large MNEs, it would not be possible for any organization because of the high-profit range of these enterprises.

Sample concedes that her account of exploitation bears the result that since exploitation is a violation of a duty to respect others, agents might prefer not to interact whenever the interaction would be exploitative, even if it is mutually advantageous. In her account, avoiding interaction is permissible, and she clearly emphasizes that exploitation is worse than neglect. On the other hand, Sample notes that if an agent “always” prefers to neglect when the other option is mutually advantageous exploitation, then this agent does not fulfill the imperfect duty of beneficence and is morally blameworthy (Sample 2003: 72).

According to Sample, her account of exploitation explains why we take exploitative interactions to be worse than neglect. She believes that interacting with someone else burdens us with special duties towards that person and equips our interactor with specific claims against us. She maintains that this aspect of her account is in line with the intuition that killing is worse than letting die (Sample 2003: 61).

Another advantage of her account, she claims, is that it can explain exploitative systems besides exploitative transactions. In such systems, exploitative behavior can be part of the routine and be accepted as ordinary, yet the exploitation claim should not be waived. In these systems, the exploitee might not feel exploited, even when they indeed are.

Like other theories of exploitation, Sample’s account has faced criticisms and counterarguments. Wertheimer criticizes all the three accounts Sample gives of why an interaction might be exploitative. For the first account of exploitation, Wertheimer draws on Sample’s example of the teller. Sample claims that I would exploit the teller if I interacted with her to profit from this interaction but still tolerated the wages upon which she could not decently live (Sample 2003: 69). Wertheimer disagrees. He maintains that “it is not clear why the mere fact that A enters into an arguably limited transaction with B requires A to be quite so responsive to B’s life needs” (Wertheimer 2007: 216).

For Sample’s second account, Wertheimer gives the example of a person, B, who has recently suffered a malicious attack on her home. She interacts with a carpenter, A, to get her home fixed. According to Wertheimer, there is nothing wrong if A takes advantage of the interaction with B as long as A does not take unfair advantage, viz., A does not charge B an exorbitant price for the work. In this example, although B has suffered an injustice, there is nothing wrong with A taking advantage of the interaction with B.

Against Sample’s third account, Wertheimer gives the example of a female sex worker, B, who sees herself as a professional who interacts with A, someone who seeks B’s services. In this example, B knows how A regards her but does not care about it and even finds A pathetic for his needs. Wertheimer argues that B is not exploited in her interaction with A.

Leaving the possible handicaps of her account of exploitation aside, for the time being, Sample gives a detailed explanation of the first type of disrespect in her definition of exploitation, viz., one in which A fails to take B’s well-being into account in their interaction.20 Her criticism of Wertheimer’s theory of exploitation was that the criterion of a hypothetical market misses marking sweatshop jobs as exploitative. Moreover, she argues that these jobs are exploitative because of her first account of disrespect in her definition of exploitation.

Exploitation as degradation or disrespect, as in (E4), can explain the exploitation of sweatshop workers at the bottom of the global production chain. However, besides its theoretical flaws, the degradation account focuses only on the worst-off to miss the exploitation of rather better-off workers, whose basic and serious needs have been met. This theoretical choice might or might not be a drawback of (E4) depending on whether a better interpretation of disrespect that covers the adverse circumstances of better-off workers is given.

3.2. Vrousalis’ theory of exploitation

Another attitudinal account of exploitation, which might cover the plight of the better-off workers as well, is given by Vrousalis. He gives a domination account of exploitation. He believes exploitation is “a form of domination for self enrichment” (Vrousalis 2013: 1). He adds that when A dominates B, “[a] necessary condition for domination is power-induced injury to B’s status or some form of servitude on B’s part” (Vrousalis 2016: 529). Like Sample, Vrousalis denies that unfairness in the division of social surplus created by the parties’ interaction has anything to do with exploitation. So his definition can be formulated as follows:

Arneson has a handful of counterexamples against the domination account. He argues that B is vulnerable to A in each of these examples, and A’s behavior towards B can reasonably be called domination. Nevertheless, although A enriches herself, none of these examples correspond to our intuitions about exploitation. One such example is what he calls the “Utility Company” (Arneson 2016: 10). In this example, a utility company is the monopoly heat supplier in a town with a cold climate. The residents have no choice but to buy heat from this company. The company charges the residents fairly and makes a profit from the sales. According to Arneson, although the heat company might be said to dominate the residents, there is no exploitation in this case because the company charges them fairly.

The defender of (E5) could respond to Arneson, holding that the residents are not exploited in the “Utility Company” example because they are not experiencing power-induced injury to their status, nor are they going through any form of servitude under the heat supplier. However, these defenders would need to show how and when someone would experience these forms of domination apart from being charged unfairly in an exchange. This is what Vrousalis claims to be doing. He also accuses the defenders of the unfairness view, (E1) or (E3), of confounding unfair treatment with exploitation. “On the fairness view, by contrast, ‘unfair treatment’ and ‘exploitation’ are used interchangeably. What is the extra purchase of saying that A exploits B, on this view? Arneson’s answer is ‘not much.’ Indeed, he uses ‘unfair treatment’ and ‘exploitation’ as synonymous throughout his essay [...]” (Vrousalis 2016: 537).

Vrousalis argues the distinct wrong that (E5) points should be analyzed under a specific category. Thus such cases would need a different response than those covered by unfairness views like (E1) or (E3).

According to (E5), MNE managers exploit sweatshop workers because they have dominating power over the workers. According to this power relationship, MNE managers can get sweatshop workers to agree to the managers’ terms. That is why such a relationship creates some servitude for the workers.

One alleged drawback of Sample’s theory of exploitation was its lack of explanatory power for the plight of better-off workers. Vrousalis’ theory can explain their situation as exploitative due to their employers’ dominating power over them. This result does not imply that sweatshop workers and better-off workers are exploited to the same degree. Since it is plausible to claim that domination comes in degrees, it follows that exploitation as domination also comes in degrees. Hence it is consistent to expect a correlative degree of moral indignation towards different degrees of exploitation.

So, besides explaining the exploitation of sweatshop workers at the bottom of the supply chain, (E5) can explain the plight of better-off workers who are also claimed to be exploited. According to this definition, managers exploit these better-off workers as long as they have no say in their work conditions.

The domination account of exploitation is better equipped to mark many standard cases of sweatshops as exploitative. However, what seems to be an advantage of the domination account becomes a disadvantage in sweatshops. The domination account can place sweatshops on a scale marking them as more or less exploitative. Nevertheless, many sweatshop critics refer to the concept of exploitation to determine with which workplaces to interfere. If the domination account marks all labor as exploited, we still need a further criterion to determine which workplaces are “exploitative enough” to deserve a proper protest and interference. Until such a criterion is defined, the domination account is of little use to sweatshop critics.