Summary and Keywords
The concept of exploitation is often invoked in situations where relatively impoverished people are treated unfairly in economic and social contexts. While the claim that exploitation involves taking unfair advantage is broadly accepted, there is little consensus about what fairness requires and whether unfairness is seriously wrong in the context of exchanges. One family of accounts claims that exploitation involves the maldistribution of resources, either because exploitative transactions result in distributions that violate substantive norms of fairness, or because procedural flaws in the way exploitative transactions come about entail that their outcomes are unfair.
A second, domination-based approach to exploitation claims that the moral flaw embodied by exploitative relations is the exploiter’s disrespectful use of his power over the exploitee. While exploiters’ domination of others may lead to maldistributions, defenders of the domination-based approach argue that distributive unfairness is neither necessary nor sufficient for exploitative relations.
These approaches both face two kinds of challenges. The first concerns the scope. Neither appears to provide necessary and sufficient conditions that are adequate to capture all and only cases commonly described as exploitation. The second concerns the normative status. Exploitation is typically assumed to be morally impermissible, yet neither approach seems to satisfactorily explain how exploitations that nevertheless generate significant welfare gains for both parties can be wrong.
The charge of exploitation is commonly made in the context of wage labor. In the 1990s and early 2000s, labor rights advocates argued that sweatshop labor in the garment industry was wrong, primarily because the workers were exploited (Arnold & Hartman, 2003; Zwolinski, 2007). Increasing globalization in the 20th century and the rise of populist political movements in Western Europe have seen a reemergence of interest in the concept. More recently, the term has been applied not only to the situation of workers in low-income countries, but also to the working- and middle-class persons in high-income countries. The association of problematic labor conditions with exploitation enjoys a long history that can be traced to Marx’s condemnation (1867) of the relationship between capitalists and proletariats as exploitative. However, concerns about exploitation also arise outside the context of labor. It is often said that colonial powers exploited their colonies (Ypi, 2013), that large corporations exploit their power to influence government regulations, and that women are exploited in sex work (Sample, 2003).
Whether the moral issues that arise in these cases can be captured by the concept of exploitation depends on what exactly exploitation is. A substantial proportion of the literature on exploitation has focused on providing a conceptual analysis of the concept in the form of necessary and sufficient conditions. Many such analyses begin with the claim that exploitation involves taking unfair advantage, and they proceed to provide an account of distributive unfairness and advantage taking. More recently, some accounts of exploitation as a form of domination (unrelated to distributive fairness) have emerged (Vrousalis, 2013).
Both distributive and domination-based accounts are moralized accounts of what exploitation is. The analyses they provide depend on the normative concepts of fair distribution and domination. In this sense, then, to say that A exploits B is to say that A is involved in behavior that is pro tanto wrong. Recent debates have focused not only on whether exploitation involves pro tanto wrongs, but also whether it is all things considered morally impermissible. In the context of exploitative transactions that are nevertheless beneficial to both parties compared to the status quo, it is not at all clear that exploitation is morally impermissible or that it should be legally prohibited (Wertheimer, 1996; Powell & Zwolinski, 2011). The two main parts of this article focus on how existing accounts have explained these two aspects of exploitation: first, what it is, and second, why it is wrong.
What Exploitation Is
At the most fundamental level, exploitation is a relational concept1 that involves morally flawed interactions between two agents, the exploiter (A) and the exploitee (B). These agents may be individuals or group agents. Exploitation is most often invoked in the context of market transactions involving goods or services, but it also can occur systemically and in nonmarket-based settings. For example, it is possible for friends and lovers to exploit the attachments of those who care about them.
According to distributive accounts, the moral flaw in exploitation is the maldistribution of benefits in the relationship. A exploits B when A takes unfair advantage of B. Distributive accounts of exploitation come in two main forms. Substantive approaches claim that the distribution of benefits in exploitative relations is unfair when it violates a particular distributive criterion, for example, a division in which benefits are equally distributed between the parties or a division that ensures all parties’ basic needs are met. According to procedural approaches, fairness does not require the satisfaction of a particular distribution. Instead, any distribution of benefits in exploitative relations can be unfair if it comes about in a flawed way. Under procedural approaches, it is not the distribution itself that makes the relationship unfair, but rather the process by which that distribution is reached. There are several distributive criteria used in substantive accounts, and there are many ways in which the process leading to a particular distribution can be morally flawed.
For those who endorse domination-based approaches, the moral flaw embodied by exploitative relations involves the exploiter’s disrespectful use of their power over the exploitee. While exploiters’ domination may lead to maldistributions, defenders of the domination-based approach argue that distributive unfairness is neither necessary nor sufficient for exploitative relations. The remainder of this section outlines and assesses the most prominent analyses of exploitation under each of these three approaches.
Though procedural and substantive distributive accounts differ in many respects, they share a common structure. First, they claim that exploitation does not involve mere maldistribution. Rather, in order for A to exploit B, A must gain at B’s expense. It is uncontroversial that condition gain at another’s expense is not sufficient for exploitation. Gifts, reparations, and rescue of those in duress all involve gain for one party at the expense of another, yet none are exploitations.
But why think gain is necessary for exploitation? Alan Wertheimer (1996, p. 209) has forcefully argued that “A does not actually exploit B unless A gains from the interaction, even if A seeks to gain from the transaction.” He points out although parties can attempt to exploit, or “[act] exploitativly towards,” others when they intend to gain, if they do not realize this gain, they do not exploit (Wertheimer, 1996, p. 209). This does not entail that such actions are morally permitted. Just as attempted theft is not theft, but nevertheless wrong, so too is attempted exploitation wrong (but not exploitation proper). It is, therefore, relatively uncontroversial that A’s gain is necessary for exploitation. Far greater controversy surrounds accounts of what constitutes fair distribution.
Theories of exploitation that are based on substantive accounts of fair transaction claim that transactions are exploitative only if the distribution involved in the transaction diverges from a fair distribution, where a fair distribution is defined with reference to a distribution principle. Distribution principles define how a particular good ought to be allocated between the parties. They involve two components: first, a definition of the distribuendum (i.e., an account of the good that is being distributed); and second, a definition of the distributive criterion (i.e., an account of how that good ought to be allocated). Thus, a transaction between A and B is unfair if A gets more of a particular distribuendum, and B less, than either would have received if the distribuendum had been distributed in accord with a particular criterion. In what follows, I will examine four prominent substantive theories of fair transaction: the classical Marxist account, equal utility gain, basic needs accounts, and fair market value approaches.
Table 1. Theory, Criteria, and Distribuenda
Embodied labor time
Fair market value
Table 1 summarizes these accounts’ distributive criteria and distribuenda.
Classical Marxist Exploitation
The classical Marxist account of exploitation is one of the earliest and most well known (Marx, 1867). Karl Marx was interested primarily in the relationship between those who owned productive assets and employed others (capitalists) and those who worked for them (the proletariat). In its simplest form, the Marxist account claims that A exploits B when an exchange between the two involves unequal amounts of labor time (Elster, 1986). For Marx, the distribuendum (that which is being distributed in the exchange) is labor time and the distributive criterion (how it is to be distributed) is equality. Of course, very few transactions involve only exchanges of labor time. Employment contracts involve exchanges of labor time for money, and most day-to-day transactions involve the exchange of money for goods. Marx’s labor theory of value provides a way to translate these exchanges into the currency of labor time. Roughly speaking, the labor theory of value claims that the value of any good or service is determined by the amount of labor time embodied in its production. Suppose A buys an apple from B for $1. To evaluate whether the transaction is fair, we need to know how much labor could be purchased with $1, and whether this amount of labor is equal to the amount of labor required to produce the apple. If the amount of labor required to produce the apple is equal to the amount of labor that can be purchased for $1, then the exchange is fair, and consequently not exploitative. If the amount of labor time required to produce the apple is greater than the amount of labor that can be purchased for $1, then the labor time embodied in the goods exchanged is unequal and, because A gives B less embodied labor than he receives from her, A exploits B.
The classical Marxist account is attractive for a number of reasons, perhaps most prominently because it incorporates concerns for both equality and contribution, in the form of the human labor used to produce goods. Equality is closely linked to the idea of fair division. Suppose that A and B want to divide a cake fairly. If neither has a greater need for cake, makes a greater contribution to cake baking, or a stronger claim to the cake than the other, it seems difficult to explain how a fair division of the cake can be anything but an equal division. In this idealized setting, equality and fairness are near-synonyms. However, we might reject the claim that a fair division is an equal division if we know of exceptions to these considerations. For example, if A produced the cake with little or no help from B, then it seems the time and effort that A has invested in the cake’s production entitle him to a greater share of the final product. By claiming that a fair division is one that involves an equal amount of labor time, Marx’s account incorporates both the intuition that fairness involves equality and the intuition that fairness is sensitive to individuals’ proportional contributions.
Unfortunately, this simplified version of Marx’s account faces a significant challenge. The labor theory of value upon which it is based has been widely discredited. As Robert Nozick (1974, p. 258) famously pointed out, “[F]ound natural objects (valued above the labor necessary to get them); rare goods . . . that cannot be reproduced in unlimited quantities; differences in value between identical objects at different places; differences skilled labor makes; changes caused by fluctuation in supply and demand; [and] aged objects whose producing requires much time to pass” all represent counterexamples to the claim that the value of a good is determined by the amount of labor embodied in its production.
From a normative perspective, the goods Nozick lists appear to have different values than would be assigned by the labor theory. That is, it seems that we ought to value these goods in a different way than we would under the labor theory. In these cases, differences caused by time, supply, location, and other elements frustrate Marx’s account. The labor theory of value also faces a descriptive problem because in most cases, it diverges from the observed value of goods on the market. Marxists have made many attempts to solve the “transformation problem,” whereby labor values can be converted into market values, but none have been successful.2 These issues have led most economists to reject the labor theory of value, and with it, the classical Marxist account of exploitation.
However, despite these problems, Marx’s influence can be seen in general families of theory that are broadly Marxist in spirit. The first, and perhaps most notable, are the approaches developed by analytical Marxists such as Jon Elster (1982), G. A. Cohen (1979), and John Roemer (1982a), (whose account is outlined later in this article). The second approach includes recent attempts at reviving a cost-of-production account of fair transaction, such as the account defended by Mark Reiff (2013). Finally, a large number of alternative definitions of unequal labor exchange that do not imply a commitment to the labor theory of value, but in which labor remains the central normative variable, have been defended by economists.
Roberto Veneziani and Naoki Yoshihara (2017) have produced an axiomatic taxonomy of unequal exchange accounts in which they identify the following core concern shared by all unequal exchange theories: “[T]he exploitation status of every agent is unambiguously determined by comparing the labour that the agent ‘gives’ to the economy [spends in productive activities], and the labour she ‘receives’ [in a commodity bundle]” (p. 3). Interestingly, as the authors point out, this core concern “allows for the possibility that all agents in the economy be either exploited or exploiters” (p. 24). While there are good normative reasons to limit the domain of exploitation to relational interactions, this requires the addition of further conditions. The core ideas behind unequal exchange accounts, and indeed some unequal exchange accounts, such as Morishima (1974), do not entail relational interactions.3
A sensible response to the problems encountered by both the classical Marxist account and equal exchange accounts is to remove the normative focus on labor while retaining the egalitarian distributive criterion. One possibility that has received frequent mention (e.g., Wertheimer, 1996; Ostmann, 2017), but little support in the existing literature, is the claim that a fair transaction is one in which each party gains an equal amount of utility4 from the transaction.
According to this approach, we first identify the prices at which each transactor is indifferent to transacting. These are prices at which the transactors are no better off transacting (under preference satisfaction accounts of well-being) than they would be if they didn’t. In economics, these prices are called the transactors’ reservation prices. Consider again a case where A buys an apple from B. Suppose that the most A is willing to pay is $2 per apple, and the least B is willing to accept is $1 per apple. These prices reflect the transactors’ respective reservation prices, and at any price between $1 and $2, both parties are better off, in the sense that they would prefer transacting at any of those prices to not transacting. Any exchange at a price greater than $1 and less than $2 is a strictly Pareto-improving transaction: a transaction in which all parties are better off compared to not transacting and no party is made worse off by transacting.
As I will discuss later in this article, there are good reasons to think that any plausible theory of fair transaction must posit a distribution that lies between the parties’ reservation prices. To use economic terminology, it must be a price that is on the contract curve, the set of points representing all possible mutually beneficial (strictly Pareto-improving) transactions. The equal welfare account makes a stronger claim. It claims that a fair transaction is the transaction in the set of mutually beneficial transactions that is equally as beneficial for A as it is for B. Note that this is not necessarily a transaction that gives each party equal monetary gain ($1.50, in this example), since the function mapping each transactor’s financial gain into utility may not be linear. There is good reason to think that the marginal utility of money is decreasing. The utility one gains when moving from no salary to a living wage, for example, is far greater than the utility one gains from moving from one living wage to twice a living wage.
Nevertheless, for the sake of simplicity, let us assume that in our apple case, the relationship between utility and money is linear for A and B: that is, 1 unit of money increases the utility by 1 unit. In this case, the equal welfare approach would entail that a fair price for the apple is $1.50. No party is made better off than the other, and the egalitarian association with fairness is preserved without any appeal to labor. In this case, the equal utility approach delivers a pretty compelling verdict.
Unfortunately, there are many contexts in which the verdict delivered by the equal utility approach does not fit well with intuitions about fair exchanges. Suppose that instead of buying an apple, A is on the way to an important job interview when his car breaks down. Fortunately, it stops right in front of the only mechanic’s shop in the area, run by B. It turns out that the problem is a broken fan belt. Suppose that B can buy the part for $10 and would be willing to replace it for another $10. In this case, B’s reservation price is $20. However, for A, arriving at the interview on time is very important. Suppose A would be willing to pay up to $5,000 for the repair. Under the simplifying linear utility assumption, the fair price would be . Many similar examples can be produced; for instance, life-saving operations and the purchase of goods with sentimental value produce the same results.
The general problem for the equal utility approach is that it focuses solely on demand for goods while ignoring aspects of the exchange related to supply. If we know that B normally changes fan belts for $30—earning a $10 profit—then how is it that she is able to charge A more? One reason she could charge A more is that B is a monopolist; she is A’s only option. This observation suggests a move to market-based accounts of fair transaction that incorporate sensitivities to both supply and demand.
Fair Market Values
In one sense, markets are ideal environments for thinking about fair prices. The fair price for a good should, it seems, reflect both how prevalent or rare the good is—that is, the supply of the good—and the value that others place on having the good for themselves—that is, the demand for the good. So, while the demand for some goods, like clean air or water, is very high, these goods are also easily obtained (in high-income countries), and this ready supply means that they do not command very high prices. Alternatively, the supply of some goods, like my childhood finger-painting projects, is very restricted. There are only a few of my “early-period finger works” in existence. However, because the demand for these goods is very low (to nonexistent), they do not command a high price on the market. Some goods are both limited in supply and highly demanded. Courtside seats to an NBA game and centrally located housing in cities like London or New York are examples of rare and highly demanded goods. Thus, it is not surprising that obtaining these goods requires a great deal of money. Prima facie, at least, it seems that for these goods, the market price simply reflects how much one should have to give up to obtain goods that everyone wants.
Yet, claiming that the market price is the fair price encounters a serious complication. On the one hand, it cannot be the case that fair transactions are synonymous with actual market transactions. All transactions are actual market transactions, but some of these are unfair. Associating fairness with actual transactions would leave us unable to criticize any transaction as unfair. The most salient alternative is to claim that fair transactions are based on perfectly competitive market prices. In economics, perfectly competitive markets are characterized, inter alia, by the presence of an infinite number of buyers and sellers, no asymmetries of information, no barriers to entry, and the absence of transaction costs. These factors present a problem for the use of perfectly competitive markets in a theory of fair transaction.
Many real-world features that we ordinarily take to be decisive in determining fair prices are assumed away. In perfect competition, all agents are price takers and there is no room for bargaining over price. Yet in many contexts, we want to know what price would be fair given that parties can bargain. Furthermore, it is not at all clear that all effects of informational asymmetries, limited markets, or transaction costs are indeed unfair. It appears that neither actual nor perfectly competitive markets can be used to provide a baseline against which transactions can be judged as fair or unfair.
Alan Wertheimer, perhaps the most prominent proponent of a market-based approach, has argued that a transaction is fair when it takes place at the ‘fair market value’, where fair market value is defined as “the price that an informed and unpressured seller would receive from an informed and unpressured buyer if the [good] were sold on the market” (Wertheimer, 1996, p. 230). He argues that when goods exchange at these prices, no party “takes special unfair advantage of particular defects in the other party’s decision-making capacity or special vulnerabilities in the other party’s situation” (p. 232). In other words, the fair price is the price the goods would receive in a second best scenario, where fair prices are those found in a moderate idealization of actual markets that includes constraints on pressure and informational asymmetries. Wertheimer’s claim is that while the market’s ability to price goods according to their availability and desirability goes some way toward generating a fair price, this is insufficient to ensure a transaction is fair because pressure and asymmetric information can create unfairness in market prices.
However, though pressure and asymmetric information may serve as solid heuristics for identifying unfair or exploitative situations, these constraints (combined with the market mechanism) are not sufficient to ensure fair transactions. First, consider asymmetric information. Some informational asymmetries are fair. In bargaining situations, if A knows B’s reservation price, he may be able to push for a better deal for himself. Traders are able to spot deals or acquire higher-quality goods because they have invested time acquiring information about the products they sell.
But fairness does not require the elimination of these informational asymmetries. Information, as economists are quick to point out, is itself a commodity that has value, and arguably, those who invest in gathering information are entitled to a return on their efforts. Indeed, the good that universities are selling (for rather high prices) is information. This does not mean that informational asymmetries are never unfair—it is because they quite often are unfair that they serve as a helpful heuristic. But it does mean that informational asymmetries alone do not entail transactions are unfair.
Similar considerations apply to pressure. Not all transactions in which pressure plays a role are unfair. A may be willing to pay B more for an apple if he forgot to eat breakfast. His hunger will likely be a factor that determines his reservation price. But fairness does not require that B sell A an apple for the price that A would have been willing to pay had he eaten breakfast. It seems that when A is responsible for the pressure he faces in a transaction, this pressure does not lead to unfairness. As with asymmetric information, pressure is a useful heuristic that is correctly associated with unfair transactions. But not all transactions involving pressure are unfair.
There are two factors that shape our intuitions about the role of information and pressure: their degree and their cause. While mere hunger may not suffice to make a transaction between A and B unfair, if B takes advantage of A’s starvation to charge him a higher price, this does seem unfair. The degree of pressure (or asymmetric information) makes a difference. And, if A does better when purchasing something from B because A has invested time learning about the product, the effect of this asymmetry is not unfair. But if A gets a better price because he lies to B about the product, the ensuing transaction is unfair (and fraudulent). The cause of asymmetric information (or pressure) makes a difference.
To recap: a market-based approach improves on the equal utility account of fairness by incorporating a concern for both the demand for goods and how they are supplied. However, as Wertheimer argues, for markets to generate fair prices, they must be fair markets. For him, this means they must exclude the influence of pressure or asymmetric information. I have suggested that fairness does not require the complete omission of these factors, but rather, whether a market is fair depends on both the cause and degree of asymmetric information and pressure. Another way to put it is that whether a market is fair depends on whether the distributions of information and pressure among the parties are just. This brings us to the final substantive account of fairness: the basic needs approach.
Proponents of basic needs approaches claim that transactions are unfair when they “neglect what is necessary for [the transactors’] well-being or flourishing” (Sample, 2003, p. 57). Various basic needs accounts employ different understandings of basic needs (Snyder, 2008). Basic needs may be characterized in terms of capabilities (Sen, 2000; Nussbaum, 2000), living wages, or as a certain minimum level of utility (Crisp, 2003), but for all basic needs accounts, the distribuenda is some form of well-being. Additionally, all basic needs approaches to fairness use a sufficiency-based (or threshold) distributive criterion. They claim that transactions are fair when a transaction supplies both transactors with a level of well-being sufficient to meet their basic needs. Because one transaction is unlikely to meet the transactors’ basic needs on its own, basic needs approaches must include a proportionality principle that relates individual transactions to a person’s yearly income.5
Although basic needs accounts are, prima facie, plausible, they quickly encounter problems. First, it is pretty easy to show that they are neither necessary nor sufficient for fair transaction. Surely, unfair transactions remain possible in worlds where all parties’ basic needs are met—or even in existing transactions between two wealthy persons. The mere fact that transactors’ basic needs are met is insufficient to ensure the transaction is fair. Similarly, transactions between two individuals who are both too poor to meet each other’s basic needs the transaction may be fair. Since not all transactions between impoverished persons are unfair, meeting basic needs must not be a necessary condition of fair transaction.
There is a further problem for basic needs accounts. The distribuendum is not constrained by the transactors’ reservation prices. Suppose that to meet B’s yearly basic needs, A must pay $2.50 for an apple. If the most A is willing to pay is $2, then (assuming he is morally motivated) in order to avoid transacting unfairly with B, he will simply refrain from transacting. Yet, if B’s reservation price is anything less than $2, there are possible transactions that both would prefer to not transacting.
Of course, basic needs proponents may claim that fairness matters more than allowing these Pareto-improving transactions. But if it is morally acceptable for A to refrain from buying an apple, then it is hard to see how his making B better off by engaging in one of these Pareto-improving transactions (in terms of her own preferences) can be wrong. We will revisit this issue in the discussion of the moral status of exploitation in the section, “Why Exploitation Is Wrong.” For now, though, it appears that the basic needs account of fair transaction faces significant problems.
If basic needs accounts can be dismissed so quickly as plausible theories of fair transaction, why do they retain their intuitive appeal? We might think that unmet basic needs are closely associated with whatever else fairness requires, and so, like information or pressure, considerations of need serve as helpful heuristics. However, this suggestion does not seem to really do justice to the depth of connection between basic needs and fairness. Instead, I would suggest that basic needs are associated with fairness because they place limits on the degree of pressure or informational asymmetries that may be permitted for transactions to be fair. That is, it is not the case that fairness requires the meeting of basic needs, but rather that fairness requires advantaged persons to refrain from using the unjust disadvantage of others (in the form of unmet basic needs) to gain at these others’ expense. So, while appeals to basic needs are not very successful in providing a substantive account of fair transaction, they may play an important role in procedural accounts, which claim that transactions are unfair when their outcomes reflect procedural problems, such as previously unmet basic needs or prior injustices.
Many philosophers have pointed out that exploitation involves multiple kinds of advantage (Ferguson, 2016a; Steiner, 2010; Wood, 1995). Under the “taking unfair advantage” approach to exploitation, if A exploits B, A gets more and B less in a transaction than either would have received in a counterfactual fair transaction. Setting aside any particular account of fairness, let us suppose that the fair price for an apple in our apple case is $1.50, and A pays B an unfair price of $1.40. Then A is unfairly advantaged (and B disadvantaged) by $0.10. Let us call this kind of advantage gain (and the related form of B’s disadvantage loss).
How is it that A is able to gain at B’s expense? If both are trying to get the best deal they can when bargaining, then it must be that B’s ability to bargain is weaker than it would have been in a scenario where she could secure the higher fair price. B is, therefore, relatively more vulnerable to A than she would have been in the fair scenario. A’s relative lack of vulnerability is a second sense in which A is advantaged. He gains more than he would have because of this weakness in B’s ability to bargain. Let us call this kind of advantage bargaining advantage.
Finally, of course, A might have a bargaining advantage over B without using it to gain at her expense. A might see that B’s vulnerability means A could get an apple for $1.40, but instead, he chooses to constrain this advantage and pay $1.50, the fair price. But when A does not choose to constrain his bargaining advantage, he takes advantage, or uses this bargaining advantage, to gain from B. So, there are three ways that the word advantage may be used in the context of exploitation: as gain, as bargaining advantage, and as the use of these features.
What does this have to do with procedural fairness? Well, as we have seen, according to substantive accounts of fairness, transactions are fair when the final distribution of gains between the transactors coincides with a particular distributive criterion. But procedural accounts claim that any outcome can be fair, provided that the conditions of bargaining satisfy certain conditions. Procedural accounts distinguish between the structure of unfair transaction and substantive normative commitments. John Roemer’s (1982a, 1982b) account of exploitation is perhaps the first to emphasize this distinction, but it is also a prominent feature of Hillel Steiner’s (1984, 1987) accounts of exploitation as well. In what follows, I discuss these accounts.
Roemer’s Property Relations Account
For Roemer, a coalition [B]6 in a larger society N is exploited if and only if
R1. There is an alternative, which we may conceive of as hypothetically feasible, in which [B] would be better off than in its present situation.
R2. Under this alternative, the complement to [B], the coalition N [B] = [A], would be worse off than at present (Roemer, 1982a, pp. 194–195).
For Roemer, “if a coalition [B] can do better for its members under the alternative of withdrawing [from the economy], and if the complementary coalition [A] does worse after [the withdrawal], then [B] is exploited under that particular specification of the rules of the game” (Roemer, 1982a). He offers three pertinent examples of withdrawal conditions. Feudal exploitation occurs when a coalition would be better off withdrawing from society with its own endowments. Capitalist exploitation occurs when a coalition would be better off withdrawing with its “per capita share of society’s alienable, or transferable, nonhuman property” (Roemer, 1982a, p. 202). Finally, socialist exploitation occurs when a coalition takes “its per capita share of all endowments, alienable and inalienable [that is, “internal” goods, like skills and talents]” (Roemer, 1982a, p. 212). These three withdrawal conditions coincide with common theories of distributive justice, but many others are possible. For example, one could create a withdrawal condition based on basic needs.
Roemer’s separation of the structural and normative components of exploitation also makes it possible to criticize each aspect individually. Since he does not commit himself to any particular normative withdrawal conditions, most critics of Roemer’s account have focused on the adequacy of his structural conditions. Here, I will focus on two issues: dependence and dominance.
First, consider dependence. Julius Sensat (1984) outlines the following problem for Roemer’s account. Suppose there exist two equally populous and fully independent islands, A and B, each of which is unknown to the other. Suppose further that the collective wealth of A is $7 million and the collective wealth of B is $3 million, and that this wealth is distributed equally within each island. Taken together as N, the total value of the islands’ wealth is $10 million. Now, if B “withdrew” with its per capita share of the wealth in N, B would be better off and A would be worse off, satisfying Roemer’s conditions. The prewithdrawal situation may be objectionable from an egalitarian perspective, but it can hardly be described as a case of exploitation. Therefore, R1 and R2 are insufficient for exploitation.
The lesson of the island case is that exploitation and unfairness require more than mere maldistribution. They require a certain kind of relationship between the parties.7 Roemer acknowledges this problem, noting that “exploitation can be said to exist only if one coalition is gaining at the expense of another” (Roemer, 1982b, p. 300). He has suggested that the following condition can address the dependence problem:
R3. If B withdrew with their own endowments, A would be worse off.
This condition ensures that the gain A extracts depends in some way on the presence of B. But is dependence alone enough to ensure that the relation between A and B is exploitative? Consider a second case, also proposed by Sensat (1984), of a two-person society comprising Able and Unable (who cannot work). If Able works 1 day to turn 1 unit of seed corn into another unit of seed plus 2 units of daily subsistence for herself and Unable, then we can say that Unable is exploiting Able. If Unable withdrew with his per capita share, he would be worse off and Able would be better off. Furthermore, Unable clearly depends on Able. Thus, by (PR1–3), Unable exploits Able. But this fails to distinguish between relevant differences in social power. Unable is an exploiter regardless of whether Able is his slave, Able is his employee, or Able and Unable are friends and Able acts out of a concern for Unable.
The lesson of this second case is that something more than maldistribution plus dependence is required for exploitation. Roemer is aware of this lacuna and originally included the following requirement:
R4. [A] is in a relation of dominance to [B]. (Roemer, 1982a, p. 195)
However, he also concedes that since his goal was to produce a purely distributive notion of exploitation, the inclusion of an undefined notion of dominance is, in a sense, ad hoc.
Dominance poses a serious problem for distributive accounts.8 It means that whether a transaction is exploitative or unfair cannot simply be determined by the way the gains are distributed. Some other aspects of the relation must be cited if exploitation is to be distinguished from other forms of interaction. This problem has led some philosophers to abandon the fairness-based account altogether. I consider some domination-based accounts later in this discussion. First, however, I will consider one more procedural account: Hillel Steiner’s rights-based approach.
Steiner’s Rights-Based Account
S1. A gains more, and B less, than either would have gained
S2. because of a prior injustice.
Note that the account of fairness implied by these conditions is procedural because it allows that any distributive outcome can be fair, provided that the outcome is not influenced by prior injustice.
Steiner’s account bears a great deal of similarity to Roemer’s. It is structural, in the sense that precisely what a rights violation is, is left unspecified. Although Steiner has developed his own left-libertarian account of distributive justice that can be used to specify what counts as a prior injustice (in Steiner’s case, property rights violations), he also acknowledges that “what can count as such a prior injustice will, of course, depend on some wider theory of distributive justice. Each individual’s entitlements . . . will vary according to which particular account of distributive justice (Rawlsian, Dworkinian, left-libertarian, Nozickian, Rothbardian, etc.) one endorses” (Ferguson & Steiner, 2017).
In this sense, Steiner’s “prior injustice” plays the same role as Roemer’s “withdrawal condition.” Like R3 in Roemer’s account, Steiner’s S1 implies that A’s gain must come at B’s expense since what one party gains in a transaction necessarily comes at the expense of the other. However, unlike Roemer’s account, which focuses on groups and their aggregate endowments, Steiner’s account places the locus of exploitation at the level of individual transactions.
Critics of Steiner’s account have focused primarily on the adequacy of his structural conditions. Here, I will focus on three critiques related to the necessity and sufficiency of the conditions (the primary focus of these critiques is S2).
First, as both Steven Walt (1984) and Benjamin Ferguson (2013) point out, the account suffers from a problem of backtracking counterfactuals that threatens to undermine the determinacy of its fairness baseline. Consider a modified version of the original apple case. Suppose A buys an apple from B because the apple that A had brought with him for lunch was stolen by C. One plausible interpretation of Steiner’s conditions is that the fair price of an apple in this case is the price that A would have paid for the apple had C not stolen A’s apple. Yet, if C had not stolen his first apple, A would not transact with B at all! The problem is that some transactions occur only because of prior injustices, and this fact means that there is no clear counterpart transaction to serve as a baseline for the fair transaction in the closest possible worlds where these prior injustices do not occur. The conditions S1 and S2 appear insufficient to distinguish ordinary and exploitative transactions.
Second, as both Nicholas Vrousalis (2013) and Matt Zwolinski (2012) point out, it is unclear that any kind of prior rights violation is necessary to conclude that the transaction is unfair or exploitative. Consider Vrousalis’s example:
Rescuer. A finds B in a pit. A can get B out at little cost or difficulty. A offers to get B out, but only if B agrees to pay a million euros or to sign a sweatshop contract with A. B signs the contract. (p. 148)
He concludes, “this exchange is paradigmatic of exploitation: if this is not an instance of exploitation, then nothing is” (Vrousalis, 2013, p. 149). Vrousalis claims that we can reach this conclusion without needing to know why B is in the pit—B may be in the pit because of a prior injustice, but it is also possible that B finds herself there “through just steps, from a just starting position” (Vrousalis, 2013, p. 149). If this conclusion is correct, then contra S2, prior injustice is not necessary for exploitation.
Finally, Walt (1984) notes that Steiner’s conditions allow many transactions to be ‘infected’ by a single prior injustice, leading to many downstream cases of exploitation. As Walt puts it (appealing to Steiner’s own property rights–based account of justice):
[M]ost if not all titles have a far from “clean” causal ancestry. . . hence if a title to objects resulting from a rights violation is invalid and exploitation involves a rights violation, then it follows that most if not all unequal bilateral exchanges are cases of exploitation. But this is implausible. The presence of exploitation is unquestionable. Its omnipresence is questionable. (p. 242)
Once again, the conditions appear insufficient to distinguish exploitative transactions from ordinary transactions. Collectively, these three problems seriously undermine the plausibility of Steiner’s account.
It is time to take stock of the ground covered so far. In the introduction, I noted that the standard approaches to exploitation have begun from the claim that exploitation involves taking unfair advantage. I next discussed four substantive approaches to fairness: the classic Marxist account, equal utility gain, market-based approaches, and finally the basic needs account. Then I considered two procedural accounts developed by Roemer and Steiner. Neither the substantive nor the procedural accounts avoided serious problems. These problems have prompted two responses to fairness-based accounts.
Nicholas Vrousalis has argued that the distributive approach should be abandoned and replaced with a domination-based approach. According to this view, “exploitation is not tantamount to unfair advantage-taking,” and there can even be distributively fair exploitings (Vrousalis, 2013, p. 149).
The second response is to bite the bullet and reject the claim that many paradigmatic cases of exploitation are, in the end, really morally problematic. As Benjamin Powell and Matt Zwolinski write, “there are good reasons for thinking that the standard cases of sweatshop labor. . . are not wrongfully exploitative,” because of “the immense difficulty in specifying a general principle of fair distribution, something no critic of sweatshops has yet managed to do” (Powell & Zwolinski, 2011, p. 468). Whether Powell and Zwolinski’s pessimism is warranted depends, in part, on whether the domination-based approach can be successful.
In Vrousalis’s account (Vrousalis, 2013, p. 132),
A exploits B if and only if A and B are embedded in a systematic relationship in which (a) A instrumentalizes (b) B’s vulnerability (c) to extract a net benefit from B.
Vrousalis’s conception of instrumentalization is similar to—but narrower than—the notion of use defined in the section “Procedural Fairness.” For Vrousalis, instrumentalization is a particular kind of use, which is “inappropriately ‘playing on’ . . . some set of . . . sufficiently salient attributes (in virtue of which the agent is who he is or does what he does)” of another agent. B is vulnerable to A if A controls some good necessary for flourishing that B lacks and that can only be obtained from A (Vrousalis, 2013, p. 134). Finally, A extracts a net benefit from B when A enjoys an overall increase in his well-being. Thus, A exploits B when A inappropriately uses important aspects of B’s dependence on A’s control of a good to increase his own well-being.
Note that, at least prima facie, a transaction can satisfy these conditions without being unfair. If B is the only (or even the most conveniently located) seller of apples, then A will be vulnerable to B with respect to the provision of apples. And if A buys an apple from B at a price above B’s reservation price, B instrumentalizes this vulnerability to extract a net benefit from A. Indeed, not only does a significant number of transactions appear to satisfy these conditions,9 but such cases “do not seem unfair, and because they do not seem unfair, [they] do not seem wrongful” (Zwolinski & Wertheimer, 2016).
However, Vrousalis offers an alternative explanation for exploitation’s wrongfulness. He claims that exploitation is a form of domination, but not necessarily unfairness. For Vrousalis, A dominates B “if A and B are embedded in a systematic relationship in which A takes advantage of his power over B, . . . in a way that is disrespectful to B” (Vrousalis, 2013, p. 138). In this sense, then, exploitation is neither a matter of maldistribution of gain nor is it a matter of mere power differentials.10 It also requires that A utilize this power in a disrespectful manner. So, for Vrousalis, if A economically exploits B—that is, he “instrumentalizes B’s economic vulnerability to enrich himself”—then A dominates B; that is, “A affects B in a way that is disrespectful to B” (Vrousalis, 2013, p. 141).
Let’s revisit the apple case yet again. How can B’s sale of an apple to A in a mutually beneficial transaction be disrespectful to A? Surely it is not the case that all possible transactions of this type are disrespectful. So it cannot be the mere use of economic vulnerability for gain that is disrespectful or dominating. If this is Vrousalis’s claim, then the shift from an appeal to unfairness to domination does not help his case because, just as such transactions are not all unfair, they are also not all disrespectful. Therefore, it must be a narrower set of such transactions that is disrespectful.
Vrousalis offers a definition of disrespect that is of some help here: A is disrespectful to B when A’s actions toward her fail to express attitudes that are “dialogically endorsable”; that is, when A’s motives are such that he has “reason to feel shame or guilt for putting or allowing a particular form of justification for some putative act on the table of discourse” (Vrousalis, 2013, p. 140). So the subset of mutually beneficial transactions that are exploitative must be those for which the motives behind the transactors’ exchanges are unjustifiable to the other.
This reading of Vrousalis’s claims about domination certainly supplies a wrong-making property for transactions that is distinct from unfairness. Agents may have reason to reject the attitudes behind others’ motives for transacting without disagreeing with the way the transaction apportions gains. For example, B may be willing to sell a family heirloom to A for $500, but not if A intends to destroy the heirloom or use the good to mock B’s family or cultural heritage. However, while such cases embody wrongs distinct from unfairness, they are not closely linked to cases ordinarily described as exploitative. A more apt description may be fraud, which occurs when A deceives B about the properties of an exchange in order to obtain B’s ostensive consent to transact.
The problem for Vrousalis is that a broad understanding of exploitation as the instrumentalization of vulnerability for gain is incapable of motivating the claim that such acts are wrong, while a narrower reading that focuses on forms of disrespect distinct from unfairness appears to pick out acts that are not ordinarily described as exploitative.
The Prospects for Analysis
The previous sections of this article have outlined three broad approaches to exploitation. According to both substantive and procedural fairness–based accounts, exploitation involves taking unfair advantage. According to domination-based accounts, exploitation does not necessarily involve maldistribution of resources, but rather the disrespectful use of another’s vulnerability for gain. However, all the particular approaches that were surveyed encountered problems of one form or another.
Some of the issues that the various accounts encountered were what might be described as technical problems. For example, the backtracking counterfactuals problem frustrates the designation of a fairness baseline in Steiner’s account, and difficulties with labor as both a theory of value and a normative focus do the same for the Marxist and equal exchange approaches.
Other problems can be classed as normative problems. For example, while the distribuenda and distributive criterion of the equal utility approach are clearly defined, it is difficult to motivate the idea that there is something unfair about unequal utility gain in situations like the broken fan belt case. A different kind of normative problem occurs with domination-based accounts. While these accounts can identify a plausible wrong-making property of transactions, it is not clear that this property is properly associated with exploitation rather than with other forms of impermissible behavior. One avenue, which I have left unexplored in this article, involves the possibility of an account that characterizes exploitation as the subset of maldistributive transactions that also involves attitudinal deficiencies like domination. Such a combined account may overcome some of the challenges to existing distributive and domination-based accounts that I have outlined here.
These normative problems raise questions related to what Alan Wertheimer (1996) has called the moral weight and moral force of exploitation. Is exploitation morally impermissible? If so, why? Is it a serious and weighty wrong that should involve legal censure, like theft, or is it a wrong that we should legally permit, like lying? The final section will outline these challenges in more detail and consider some of the answers that have been given to these questions.
Why Exploitation Is Wrong
One reason that conceptual and normative analyses of exploitation are difficult to produce is that many paradigmatic cases of exploitation involve ostensively consensual transactions that leave the transactors better off than they were before the transaction (and this is the case for all accounts of well-being that enjoy widespread support). It is difficult to claim acts that make all parties better off should be morally or legally prohibited. As Ferguson (2016b) has shown, the claim that the fairness of a transaction is a necessary condition of its permissibility conflicts with welfarism, the principle that the goodness of a state of affairs depends on the utility of the individuals in that state (Sen, 1979), when two other, relatively weak conditions are added. These conditions are Betterness of Permissibility, which claims that morally permissible acts are morally better than morally impermissible acts, and Optionality of Ordinary Transactions, which claims that transacting with another person is not obligatory simply in virtue of the transaction being mutually beneficial.
Consider again the apple transaction, in which A’s (purchasing) reservation price is $2 and B’s (selling) reservation price is $1. Let us also assume, for the sake of argument, that the fair price is $1.50. Although B would prefer getting $1.50 for the apple to getting the unfair price of $1.20, she would still prefer getting $1.20 to not transacting. This is also true for A. Consider the three possible actions that A and B might take and the preference orderings over these options for each (shown in Table 2).
Table 2. Transactions and Preference Orderings
Fair transaction at $1.50
Unfair transaction at $1.20
The Optionality condition entails that A and B are not morally obliged to engage in one of the transactions; that is, it entails that it is morally permissible for them to pick option N. Since welfarism claims that the moral goodness of a state of affairs depends on persons’ utilities, and both parties prefer U to N, welfarism entails that U is morally better than N. Finally, the Betterness of Permissibility condition claims that permissible actions are better than impermissible actions. This means that since N is permissible and U is morally better than N (by the betterness condition), U also must be permissible. But this result conflicts with the claim that fairness is a necessary condition for transactions’ permissibility.
The paradox relies on four conditions, and consequently, it can be resolved by rejecting any one of the four. However, most responses to the tension it expresses have argued for the rejection of either the welfarism condition or the claim that fairness is necessary for permissibility. For example, both Ferguson and Wertheimer suggest that the welfarism condition should be rejected. However, neither provides very convincing justifications for why a concern for fairness should trump considerations of welfare in these cases.
Prioritizing fairness over welfare not only entails that exploitation is morally impermissible, it also implies that it is a pretty weighty kind of wrong, for even when a particular transaction promised great increases in well-being, the defense of the fairness condition implies that the moral value of these increases is outweighed by the wrong of unfairness. Furthermore, if fairness is so important that it should always outweigh well-being increases, then there is at least a pro tanto case to make for legally prohibiting unfair acts. Placing this kind of absolute weight on fairness seems unattractive. There are two alternatives to prioritizing fairness.
The first alternative is to argue that exploitation is wrong for reasons that are distinct from fairness considerations—for example, by arguing that exploitation is wrong because it involves domination or disrespect. Yet, as we saw in the previous discussion, acts involving forms of domination or disrespect that are distinct from unfairness do not appear to correspond with those actions commonly described as exploitations. Unless this challenge can be addressed—and good reasons for prioritizing these alternative wrongs over welfare can be provided—this alternative also seems unattractive.
The second alternative is to claim that the wrong of exploitation consists in its unfairness, but to deny that unfairness makes exploitation all things considered wrong. With this approach, unfairness would only be a pro tanto wrong-making feature of transactions. Whether a transaction is all things considered wrong would depend, inter alia, on how unfairness trades off against welfare gains. A potential disadvantage of this approach is that it appears ad hoc, for the following reason. Either the weights attached to fairness and welfare will be fixed, which seems arbitrary, or they will be allowed to vary. However, if they are allowed to vary on a case-to-case basis, the approach also seems ad hoc. In order to avoid being ad hoc, defenders of this approach must develop a justification of contextual weights for each value. Although this task seems possible, existing accounts of how these values should be traded off against one another are controversial, to say the least.
A more troubling problem concerns the implications of the weighting approach for some of the most poignant cases of exploitation. If we say that in some transactions, welfare gains can override unfairness, making a transaction all things considered permissible, then there will be cases where exploiting the rich is impermissible but exploiting the poor is permissible.
Consider the apple transaction again. Suppose that the weights we attach to fairness and welfare gain are such that when A exploits (unfairly transacts with) B by paying her $1.20 rather than $1.50 for an apple, he barely acts impermissibly. In other words, although A gains $0.20 over B’s nonagreement point, the welfare associated with B’s gain is insufficient to morally outweigh the unfairness of the transaction. This means that if the transaction had been slightly less unfair, or if B had gained slightly more welfare, the weights attached to these values would not have entailed that the transaction was impermissible.
Now, there is good reason to believe that the marginal utility (and more generally, welfare) of most goods is decreasing—that is, persons derive less welfare from each additional unit of goods they receive. This means that as B gets poorer, the welfare she derives from the $0.20 increases. At some point, when B is poor enough, the welfare she derives from the transaction will outweigh the unfairness of the transaction, making it morally permissible. More generally, given finite weights for fairness and welfare, for any impermissible exploitation of a rich person, it will always be possible (in theory) to find a poor person whom it is permissible to exploit. This seems strongly counterintuitive and provides a good reason to reject the weighting approach.
Let’s recap. I have been considering whether and why exploitation might be morally wrong. If exploitation is wrong because it is unfair, and fairness is a necessary condition for a transaction to be permissible, then (provided that we accept two other conditions) we must reject welfarism. This conclusion can be avoided by either associating the wrong of exploitation with something other than unfairness, or by claiming that the unfairness involved in exploitation is only a pro tanto wrong that can be outweighed in certain cases.
However, neither strategy is very attractive. Domination independent of maldistribution does not appear to pick out common cases of exploitation, while allowing a trade-off between unfairness and well-being generates counterintuitive results about who is wronged by exploitation. It appears that if we want to retain an association between exploitation and unfairness, and wish to say that exploitation is wrong because it is unfair, then we must reject welfarism.
This article has discussed conceptual analyses of what exploitation is and normative arguments about why it is wrong. The first portion covered three kinds of theories about what constitutes exploitation. The first two are distributive approaches. Substantive fairness approaches claim that A exploits B when A takes unfair advantage of B by violating a particular, substantive criterion of fairness. These accounts included the classical Marxist account and other equal exchange accounts, as well as equal utility gain, fair market value, and basic needs approaches. The second kind of distributive theory claims that A exploits B when A takes unfair advantage of B by violating a procedural criterion of fairness. Procedural accounts include Roemer’s property relations account and Steiner’s rights-based account. The final kind of theory claims that exploitation involves the use of domination to extract a net benefit from vulnerable persons.
The second portion of the article discussed whether and why exploitation should be considered wrong. In those sections, I showed that exploitation brings common intuitions about the moral value of increases in well-being and the moral value of fairness into conflict.
This article was presented at VU Amsterdam. I thank those present for their comments and suggestions.
Arneson, R. (2013). Exploitation and outcome. Politics, Philosophy, & Economics, 12(4), 392–412.Find this resource:
Arnold, D., & Hartman, L. (2003). Moral imagination and the future of sweatshops. Business and Society Review, 108(4), 425–461.Find this resource:
Cohen, G. (1979). The labour theory of value and the concept of exploitation. Philosophy & Public Affairs, 8(4), 338–360.Find this resource:
Crisp, R. (2003). Equality, priority, and compassion. Ethics, 113, 745–763.Find this resource:
Elster, J. (1982). Roemer versus Roemer: A comment on “New directions in the Marxian theory of exploitation and class.” Politics and Society, 11(3), 363–373.Find this resource:
Elster, J. (1986). Karl Marx: A reader. Cambridge, UK: Cambridge University Press.Find this resource:
Ferguson, B. (2013). The paradox of exploitation: A New Solution (Doctoral thesis). London School of Economics.Find this resource:
Ferguson, B. (2016a). Exploitation and disadvantage. Economics & Philosophy, 32(3), 485–509.Find this resource:
Ferguson, B. (2016b). The paradox of exploitation. Erkenntnis, 81(5), 951–972.Find this resource:
Ferguson, B., & Steiner, H. (2017). Exploitation. In S. Olsaretti (Ed.), Oxford handbook of distributive justice. Oxford: Oxford University Press.Find this resource:
Griffin, J. (1986). Well-being. Oxford: Clarendon Press.Find this resource:
Marx, K. (1867). Capital: A critique of political economy (Vols. 1–3). Chicago: C. H. Kerr and Company, 1906–1909.Find this resource:
Morishima, M. (1974). Marx in the light of modern economic theory. Econometrica, 42(4), 611–632.Find this resource:
Nozick, R. (1974). Anarchy, state, and utopia. Oxford: Blackwell.Find this resource:
Nussbaum, M. (2000). Women and human development: The capabilities approach. Cambridge, UK: Cambridge University Press.Find this resource:
Ostmann, F. (2017). Fairness in international commercial transactions (Doctoral thesis). University College London.Find this resource:
Powell, B., & Zwolinski, M. (2011). The ethical and economic case against sweatshop labor: A critical assessment. Journal of Business Ethics, 107(4), 449–472.Find this resource:
Reiff, M. (2013). Exploitation and economic justice in the liberal capitalist state. Oxford: Oxford University Press.Find this resource:
Roemer, J. (1982a). A general theory of exploitation and class. Cambridge, MA: Harvard University Press.Find this resource:
Roemer, J. (1982b). Property relations vs. surplus value in Marxian exploitation. Philosophy & Public Affairs, 11(4), 281–313.Find this resource:
Roemer, J. (1985). Should Marxists be interested in exploitation? Philosophy & Public Affairs, 14(1), 30–65.Find this resource:
Sample, R. (2003). Exploitation: What it is and why it’s wrong. Lanham, MD: Rowman & Littlefield.Find this resource:
Samuelson, P. (1971). Understanding the Marxian notion of exploitation: A summary of the so-called transformation problem between Marxian values and competitive prices. Journal of Economic Literature, 9(2), 399–431.Find this resource:
Sen, A. (1979). Utilitarianism and welfarism. Journal of Philosophy, 76(9), 463–489.Find this resource:
Sen, A. (2000). Development as freedom. New York: Anchor Books.Find this resource:
Sensat, J. (1984). Exploitation. Nous, 18(1), 21–38.Find this resource:
Snyder, J. (2008). Needs exploitation. Ethical Theory and Moral Practice, 11(4), 389–405.Find this resource:
Steiner, H. (1984). A liberal theory of exploitation. Ethics, 94(2), 225–241.Find this resource:
Steiner, H. (1987). Exploitation: A liberal theory amended, defended, and extended. In A. Reeve (Ed.), Modern theories of exploitation (pp. 132–148). London: SAGE.Find this resource:
Steiner, H. (2010). Exploitation takes time. In J. Vint, J. S. Metcalfe, H. D. Kurz, N. Salvadori, & P. A. Samuelson (Eds.), Economic theory and economic thought: Essays in honour of Ian Steedman (pp. 20–29). London: Routledge.Find this resource:
Veneziani, R. (2013). Exploitation, inequality, and power. Journal of Theoretical Politics, 25(4), 526–545.Find this resource:
Veneziani, R., & Yoshihara, N. (2017). The theory of exploitation as the unequal exchange of labour (Unpublished Draft).Find this resource:
Vrousalis, N. (2013). Exploitation, vulnerability, and social domination. Philosophy & Public Affairs, 41(2), 131–157.Find this resource:
Walt, S. (1984). Comment on Steiner’s “liberal theory of exploitation.” Ethics, 94(2), 242–247.Find this resource:
Wertheimer, A. (1996). Exploitation. Princeton, NJ: Princeton University Press.Find this resource:
Wood, A. (1995). Exploitation. Social Philosophy and Policy, 12(2), 135–158.Find this resource:
Ypi, L. (2013). What’s wrong with colonialism. Philosophy & Public Affairs, 41(2), 158–191.Find this resource:
Zwolinski, M. (2007). Sweatshops, choice, and exploitation. Business Ethics Quarterly, 17(4), 689–727.Find this resource:
Zwolinski, M. (2012). Structural exploitation. Social Philosophy and Policy, 29(1), 154–179.Find this resource:
Zwolinski, M., & Wertheimer, A. (2016). Exploitation. In E. N. Zalta (Ed.), Stanford encyclopedia of philosophy (Summer 2017 ed.). Available online.Find this resource:
(1.) Modal conceptions of exploitations also exist, according to which persons have the nonrelational properties of being exploiters and exploitees. In this article, I focus primarily on relational exploitation because it is of greater normative relevance. Nevertheless, a short discussion of modal exploitation can be found in the section “Classical Marxist Exploitation.”
(2.) Paul Samuelson famously quipped that the transformation problem could be solved as follows. “Contemplate two alternative and discordant systems. Write down one. Now transform by taking an eraser and rubbing it out. Then fill in the other one. Voílà! You have completed your transformation algorithm.” (Samuelson, 1971, p. 4).
(3.) Modal nonrelational conceptions of exploitation should not be confused with the idea that exploitation is a systemic phenomenon, in the sense that it “refers to the relationship between a person and society as a whole as measured by the transfer of the person’s labor to the society, and the reverse transfer of society’s labor in the person” (Roemer, 1985, p. 31). While systemic exploitation in this broad sense is compatible with modal exploitation, it is also compatible with relational exploitation. For example, it is possible to define exploitation over a certain set t of aggregate transactions while nevertheless maintaining that B is exploited by the whole set A of individuals who engaged in the transaction in t.
(4.) Equal utility gain is a particular example of a broader family of theories that claim fair transactions are those in which the welfare gains of the parties are equal. Utility—the subjective satisfaction of individual preferences—is but one understanding of what well-being consists in. See Griffin (1986) for an overview.
(5.) For example, if $30,000 net income per year is sufficient to meet B’s basic needs, then one proportionality principle might be $30,000 divided by 52 weeks (perhaps minus holiday time), divided by 40 hours, divided by the number of apples B sells per hour on average. If B averages 5 apples per hour, then the fair price for an apple would be (plus the expenses B incurs in its production and sale).
(6.) I have used the notation of the present article in place of Roemer’s original notation.
(7.) Those who agree with this verdict will likely reject certain unequal exchange accounts discussed in the section “Classical Marxist Exploitation,” which conceive of exploitation as a monadic property. While these accounts may provide plausible conditions for a just economy, they do not appear to characterize exploitation.