Wage Exploitation as Disequilibrium Price

Abstract

There are two opposing views concerning intuitive cases of wage exploitation. The first denies that they are cases of exploitation at all. It is based on the nonworseness claim: there is nothing wrong with a discretionary mutually beneficial employment relationship. The second is the reasonable view: some employment relationships can be exploitative even if employers have no duty towards their employees. This article argues that the reasonable view does not completely defeat defences of wage exploitation, because these do not rely solely on the nonworseness claim. They also rely on the idea, popularised by Alan Wertheimer, that exploitation is a form of disequilibrium price occurring in defective markets. The article then proceeds to criticise Wertheimer’s account through neoclassical, new institutional, and Austrian economics. It concludes that considerations for economic efficiency are irrelevant to assessing intuitions regarding exploitation.

Suppose that a corporation from a rich country outsources part of its production to sweatshops in the Global South. Working conditions are bleak, hours are long, and the pay is low compared to the wages the corporation pays in its developed home country. Workers accept to work for this corporation because it is still better than nothing. Suppose there are intuitive reasons to consider this mutually beneficial employment relationship as unfair: it seems it is a case of wage exploitation.

There are two ways to assess this interaction. ¹ One could commit to the nonworseness claim (NWC). ² The NWC states that if it is morally permissible for an agent not to enter into an economic interaction with another, ³ then any mutually beneficial and consensual interaction between them cannot count as a case of wrongful exploitation. To believe otherwise would be to prefer that neither party receive any benefits at all whenever the distribution of these benefits cannot meet a given benchmark of fairness. In a context where neither party has any specific obligation to the other, an account of fairness leading to this conclusion would be perverse. ⁴

Hence, to the extent that the corporation has no specific obligation towards its employees, it follows from the NWC that any wage it pays cannot be worse than if there were no interaction at all between them, even if the wage intuitively strikes us as unfairly low. The NWC has inspired a body of literature that shall be characterised as defending wage exploitation; that is, it rejects the intuition that a case like the preceding one is a case of wrongful exploitation. Its main advantage is that it allows this literature to remain theory neutral (Faraci 2019, 170), that is, not committed to any fleshed-out account of what wage exploitation is (see also Zwolinski 2007, 690–91).

Alternatively, one could commit to what David Faraci (2019) has called the reasonable view (RV): the economic interaction described earlier can be considered exploitative, but its mutually beneficial character warrants non-interference from third parties. The intuition that the corporation is exploiting its workers by defaulting on its obligation to pay a fair wage is compatible with the idea that third parties must refrain from interfering in any way that could make the corporation rescind its employment offer.

The present article addresses an issue that the literature on wage exploitation has, in my view, neglected and that has become apparent in the opposition between the NWC and the RV. I think both views miss an important aspect of wage exploitation: most of it occurs in competitive markets. I shall argue that this gives us reasons outside of the NWC to reject the intuition at the basis of the RV. This point has been overlooked by both sides of the debate and therefore needs to be addressed, because, as I shall demonstrate, it has tangible consequences for how certain cases of wage exploitation should be defended or criticised.

The idea that exploitation cannot occur in competitive markets has been popularised by Alan Wertheimer—note that Wertheimer (1996) himself rejected the NWC. I will first show how his account lends support to defences of wage exploitation and then demonstrate how it fails to do so. More generally, the relationship between fairness and market competition has been scantly explored, and so the present article takes on this task as well.

In the first part of this article, I will argue that Wertheimer’s account, which I will call exploitation as disequilibrium price, links fairness with market competition by identifying it with Pareto efficiency. This is because when a distribution is Pareto efficient, no further and fairer distribution, achievable through an increment in market competition, is available. Pareto inefficiency is therefore what gives exploitation its moral weight, that is, “the intensity of its wrongness” (to use Wertheimer’s words; see Wertheimer 1996, 28). The rest of the article shows that the account fails. It will do so by reconstructing it through different economic theories and their related understandings of efficiency—the neoclassical, the new institutional, and the Austrian.

The second part deals with the neoclassical and new institutional reconstructions. I show that the neoclassical theory of price as a static equilibrium between supply and demand seems to match, at first sight, what Wertheimer had in mind. However, such an equilibrium fails on the account’s implied conception of fairness. Even if the price of labour in a labour market is at equilibrium, there is nevertheless a fairer price that is available, and the reason the fairer price is never reached is indeed efficiency based. It is related to the high transaction costs that prevail in labour markets, as analysed by new institutional economics.

The third part offers a way out through a second reconstruction based on an alternative theory of price, namely, the dynamic equilibrium conceptualised by Austrian economics. Such a reformulation comes with strings attached, however. First, the account becomes unable to criticise a prime and obvious example of exploitation, that is, monopsonist pricing. Second, such a counter-intuitive result may be defended, but only through the adoption of a clear libertarian outlook. This, I shall argue, makes the account unable to reject the intuition at the basis of the RV. Inefficiency cannot, therefore, be what gives exploitation its moral weight.

For Wertheimer, moral weight is to be distinguished from moral force. While moral weight is the feature of exploitation that makes it wrong, moral force is what justifies interference from third parties seeking to prevent or correct it—the “moral upshots or reasons for action that exploitation might or might not involve for parties to the transaction or for society” (Wertheimer 1996, 28). I show that inefficiency is just as useless to provide such reasons.

EXPLOITATION AND EFFICIENCY

For Wertheimer, ⁵ exploitation can be harmful or mutually beneficial. Harmful exploitation is when A takes unfair advantage of B in a way that B is made worse off than she would have been without interaction. Mutually beneficial exploitation is when A takes unfair advantage of B in a way that still leaves B better off than she otherwise would have been without interaction (Wertheimer 1996, 207). I focus on the latter case in this article, for it is precisely what happens in wage exploitation.

Alan Wertheimer’s Account

Because wage exploitation is mutually beneficial, any unfairness must stem from how the interaction’s productive surplus has been divided. When A takes unfair advantage of B, A’s gains must be somehow excessive, and B’s must be somehow too low. Therefore it conceptually requires a counterfactual fair price where A’s gains are not excessive and B’s are not too low. To understand what this fair price is for Wertheimer, consider three examples he gives of exploitative interactions—Snowstorm Rescuer, Friends Buying a House, and Traditional Marriage.

Consider the Snowstorm Rescuer case (Wertheimer 1996, 208): Joe is a motorist stranded in a snowstorm. Imagine there is a rescuer, call her Jane, who offers to rescue Joe. Suppose that rescuing Joe comes at a cost, and thus we cannot expect her to do it for free. Still, she is, as Wertheimer says, “fair-minded but not overly altruistic” and thus offers her service at a fair price, say, 50 denarii. In that case, Wertheimer contends, even if Jane gains from Joe’s predicament, and even if he has no other choice than to interact with her, there seems to be no exploitation. ⁶ She takes advantage of him, but not unfair advantage.

Why is that? Imagine a variation of the case—Greedy Rescuer—in which Jane is not fair-minded. She is greedy and would love to charge Joe as much as she could to tow him out, say, 100 denarii. This is indeed the most that Joe would be willing to pay—at 101 denarii, he would prefer staying stranded than being rescued. However, just as Jane is about to make her greedy offer, another rescuer, Ann, comes up and also offers her services. Assuming the two cannot form a cartel, the price would immediately drop to 50 denarii, the lowest price they are both ready to accept (under that amount, they would not perform the rescue). Fifty denarii is here the so-called equilibrium price in the market made by Joe, Jane, and Ann.

Wertheimer’s chief insight is that the equilibrium price is the fair price. When that price obtains, then neither Jane nor Ann could have possibly exploited or taken unfair advantage of Joe, because everyone has been a price taker in the market they formed (Wertheimer 1996, 217–18). If, for example, Jane still tried to make her greedy offer, then Joe would immediately turn to the next available substitute—Ann (and vice versa). Market competition, through the provision of substitutes, is thus the mechanism that prevents exploitation understood as unfair advantage taking.

Furthermore, the equilibrium price will by necessity satisfy any possible conception of fairness. Consider indeed a further example: Friends Buying a House (Wertheimer 1996, 230). It is an impractical example, as the housing market is a complicated outlier in economic theory and this article is concerned with wage exploitation, so let me reformulate it by applying it to the labour market. Suppose that a worker sells an hour of labour to an employer. She offers her the price that is quoted by a specialist as the current rate for an hour of that type of labour in the relevant market, in which many employees sell their labour time and many employers are willing to buy it.

Here again, that price would be a fair or not exploitative division of the productive surplus of the cooperation between workers and employers in that relevant market. For, if employers could afford to pay a fairer wage (under any conception of fairness), then workers would obtain it from other employers providing substitute offers of employment through an increment in competition on their side. Because it cannot happen when workers get the equilibrium price of their labour, this price is the fairest wage employers can afford to pay. The equilibrium price of labour, whatever it is, can therefore only be fair. A worker paid that price cannot be exploited, for the same reasons that Joe could have been exploited by neither Jane nor Ann.

Thus, Wertheimer (1996, 236) concludes, “a hypothetical market price is more plausible” as a “candidate for principles of fair division.” ⁷ In effect, that fair price is a Pareto-efficient distribution, for Pareto efficiency is indeed achieved whenever no further redistribution can be obtained through an increment in competition, or unless someone is harmed. ⁸

Conversely, Paretian inefficiency is what gives to mutually beneficial exploitation its moral weight (what makes it unfair). When a distribution is Pareto inefficient, someone can be made better off by a further distribution of the interaction’s surplus achievable through an increment in competition. Read in this way, Wertheimer’s account has reconciled two seemingly opposed imperatives, one that is normative (fairness) and one that is economical (efficiency). His chief merit is his demonstration of the conceptual impossibility of a Paretian exploiter. Call this account Exploitation as Disequilibrium Price.

Not all cases of exploitation are like this. Consider the Traditional Marriage case (Wertheimer 1996, 289–93). Suppose Jane marries Joe on terms that strike us as exploitative—imagine, for example, that Joe relies to an unfair extent on Jane’s free care work and domestic labour. This seems unfair, because Joe seems to have certain moral obligations towards Jane stemming from their relationship that he did not have before. The way he fulfils these moral obligations should be compared not to a counterfactual where there is no marriage (a counterfactual baseline of non-interaction) but to a counterfactual of what counts as an egalitarian marriage (a counterfactual baseline of fairness).

Consider now how Wertheimer contrasts such an exploitative marriage with Greedy Rescuer. Wertheimer believes that the exploitative marriage is worse because the difference between the fair and the exploitative baseline is greater in the former than in the latter. That greater distance comes from the closeness between Jane and Joe in the marriage. The only way he can make sense of that intuition is by taking the fair baseline as the point of comparison, rather than the baseline of non-interaction. This is why, as noted in the introduction, Wertheimer rejected the NWC. One could go further: what makes Jane and Joe’s marriage exploitative, and what makes by the same token Joe’s behaviour morally worse, is Jane’s love for him. In marriages, what ensures fairness—an increment in competition—is not operating. Perhaps describing Jane’s plight in these terms could be even considered a category error. ⁹ In labour markets, however, fairness may be obtained through increments in competition, for a substitute is normally available, and thus, as Wertheimer (1996, 217) puts it, wage exploitation “generally cannot occur in a perfectly competitive (labour) market.”

Wage Exploitation and the NWC

Now that the disequilibrium price account has been clarified, come back to the opposition between the RV and the NWC. I will argue now that the literature that defends wage exploitation needs the NWC less than it appears at first. Indeed, most cases of wage exploitation occur in competitive markets. This fact renders the disequilibrium price account sufficient to assess them as non-exploitative and makes the NWC irrelevant. Like Faraci, I will focus on Matt Zwolinski’s defence of the type of employment with which this article started and which seems to be intuitively exploitative—sweatshops. ¹⁰ Zwolinski (2007, 707) does seem at first to rely on the NWC to reject that intuition:

Suppose it turns out to be true that MNEs [multinational enterprises] are earning an unusually high rate of profit from their use of sweatshop labour—high enough that they could afford to pay significantly higher wages without putting the firm at risk. Such firms could be said to be taking advantage of workers’ vulnerability to benefit disproportionately from their labour agreements. Are they guilty of an objectionable form of exploitation? Are they acting in a blameworthy manner? There is something rather odd about saying that they are.

Zwolinski then notes that sweatshop wages cannot be worse than their absence because non-interaction is permissible. That, in turn, must be the case, for otherwise, companies that do not open sweatshops would be more blameworthy than those that do. He reiterates this point with Benjamin Powell (Powell and Zwolinski 2012, 468–69) when replying to Jeremy Snyder’s (2008, 2009, 2010) objections to the NWC based on the duty of beneficence corporations owe to their employees.

Now, Zwolinski addresses the intuition that sweatshops are exploitative only because it could be a possible objection to his main argument. That main argument emphasises that the decision of workers to work in sweatshops is voluntary and should not be interfered with. Even if it is made in severe constraints, it nevertheless signals “information about an agent’s preferences” (Zwolinski 2007, 693). Workers accept sweatshop labour among all the employment possibilities they have because that one is the best and because they value their wages more than whatever they will get in their next employment opportunity.

This fact invalidates justifications for third-party interferences with sweatshops, such as boycotts and regulations, that aim to increase wages or improve working conditions (Zwolinski 2007, 691–95). These measures would indeed increase the cost of sweatshop labour and decrease the demand for it. The result is that sweatshops would hire less, thus harming some workers’ preferences by condemning them to unemployment. Even if the economic reasoning behind Zwolinski’s argument has been contested (Coakley and Kates 2013), most of the available evidence seems to confirm it (Sollars and Englander 2018, 16–19).

But this argument also points out that sweatshop workers get paid the equilibrium price of their labour. First, an equilibrium price is nothing else but the efficient satisfaction of all the preferences expressed in a given market. Second, and as stated by Zwolinski in his later work with Benjamin Powell (see especially Powell and Zwolinski 2012; see more generally Powell 2006, 2014), the fact that an increase in labour costs leads to an increase in unemployment necessarily entails that the market for sweatshop labour has reached its equilibrium price. If the only available redistribution requires harm of one of the parties involved (the potentially unemployed workers), then sweatshop wages are Pareto efficient. And if Wertheimer is right that fairness follows from efficiency, then sweatshops should be considered non-exploitative. As Zwolinski (2012, 165–66) himself admits,

on an account of exploitation that is focused on transactional fairness alone, the case against sweatshops appears thin. Alan Wertheimer’s account, arguably the most influential fairness-based account of exploitation, is founded on the idea that for at least a broad range of cases, the fairness of a transaction between A and B can be analyzed in terms of the agreement that would be reached between them in a hypothetical competitive market.… But there is little reason to suppose that sweatshops and the multinational companies are operating in a non-competitive environment.

Consider, then, the following two argumentative steps defending sweatshops:

Step 1: Workers’ voluntary choices lead to price equilibrium, and hence third parties should not interfere.

Step 2: Sweatshops have no obligation to their employees and thus cannot be blamed for paying the market wage (following the NWC).

Conclusion: No exploitation

Step 1 assesses third-party actions, because it serves to reject third-party interference, whereas step 2 assesses the employer’s actions, because it negates the intuition that sweatshops are exploitative. Following the disequilibrium price account, however, step 1, despite its focus on third-party interferences, is already enough to assess employers’ actions and already justifies the conclusion. If third-party interferences lead to lower employment levels, then sweatshops already pay their workers the equilibrium price of their labour. And that means that they cannot be exploitative, for the same reason that there was no exploitation when Jane asked John for the market price (50 denarii) to rescue him in Snowstorm Rescuer.

Now, if there are additional reasons to hold to the intuition that sweatshops are exploitative (e.g., hypothetical duties or obligations employers owe to their employees generated by employment relationships, perhaps akin to those involved in Traditional Marriage), then a defender of wage exploitation could appeal to the NWC to reject those as well. But for these reasons to support that intuition, it must also be true that sweatshops can pay higher wages without raising unemployment, as is acknowledged by their critics (see, among others, Meyers 2004, 329–30; Snyder 2008, 396, 400–401; Kates 2019, 44). ¹¹ This is, however, already proven false at step 1, following the disequilibrium price account—if sweatshops could pay fairer wages, then they already would, because they are operating in competitive markets. This is actually why third-party interference would lead to a rise in unemployment in the first place. Step 2 is therefore not needed.

Zwolinski (2007) himself seems to have been aware of this, as illustrated by the first sentence in the first quote given earlier (“Suppose it turns out…”), which he qualifies further with the following caveat:

This argument [that sweatshops have duties to pay higher wages and are exploiters to the extent that they default on that duty] seems to presuppose that there is considerable slack in the market for international labor. I have not seen any evidence to indicate that this is actually the case, and it strikes me as a rather implausible claim when made about the market for such labor generally.… Nevertheless, I am interested in seeing what follows morally, if the claim turns out to be true (note 56).

Sweatshops, like most intuitive cases of wage exploitation, occur in competitive markets. This fact makes the NWC—which has occupied a significant share of the literature on the topic—redundant. ¹²

The Disequilibrium Price Account and the Reasonable View

Come back to David Faraci and the RV. Consider now his Owner/Employee example, which he gives as a prime example of wage exploitation:

Owner offers full-time employment to Employee, for which the latter will receive compensation insufficient to meet his basic needs (food, shelter, etc.). Owner could offer Employee better compensation without lowering the net benefit of their interaction or eliminating Owner’s profits (Faraci 2019, 269).

Faraci’s intuitive reason to consider this a case of wage exploitation is that Owner defaults on her obligation to pay what corresponds to a pre-established distributive standard—in this case, a living wage, but it does not matter much—whereas she has the margin to pay for it.

But if Owner can pay such a living wage without harming other employees—through a rise in unemployment—then it means that something is not right. It means the market in which Owner and Employee are interacting has some defect. Otherwise, Employee would get that living wage from another owner providing substitute employment, say, Owner 2, through an increment in competition (here between Owner and Owner 2).

Faraci later points to the unfair division of their interaction surplus to justify Owner’s obligation to increase wages. Owner takes a cut of what is rightfully Employee’s:

The intuition behind the claim that Owner acts wrongly is not that Owner has to benefit Employee more because Employee has benefitted him some. Rather, it is that Owner is keeping an unfair portion of the benefits they create together; a greater portion of those profits is rightfully Employee’s (Faraci 2019, 177).

But Owner should not be able to do that, because here again, Employee would get what is rightfully hers from Owner 2.

The problem with Faraci’s RV—that Owner does indeed exploit Employee even if we have compelling reasons not to prevent their interaction—should now become apparent, regardless of whether it is indeed reasonable. It is at best a partial objection to the literature that he seeks to criticise, because sweatshops (like most cases of intuitive wage exploitation) operate in competitive markets and are nothing like Owner/Employee.

Regarding that particular case, the disequilibrium price account has two implications. First, the reasonableness of the RV follows from it as a matter of coherence—Owner is indeed an exploiter (and, let us grant this for now, should not be interfered with). Second, holding Owner/Employee as exploitative while considering sweatshop wages as fair would not require any intuitive understanding of a pre-established distributive standard (like a living wage), because here again, a commitment to the disequilibrium price account is sufficient.

As Faraci (2019, 181) himself puts it,

of course, sweatshop defenders sometimes suggest that market-clearing wages are fair wages, so they might simply reject this intuition. But this is far from necessary unless their view is that market-clearing wages are ipso facto fair.

As I have argued, this is, or at any rate should be, their view. Faraci acknowledges as much:

If accurate and representative, this claim might allow Powell and Zwolinski to accept the wrongful unfairness of what Owner does but draw an important moral distinction between that case and actual sweatshops. If sweatshops are defensible on these grounds, if they do not meet relevant conditions of unfairness, there is simply no need for blanket theoretical moves like the appeal to the nonworseness claim—especially when such moves threaten intuitions shared by defenders.… We do better to focus our attention on development of relevant theories of fairness and other values, and questions about the extent to which actual employment relations (fail to) meet those theories’ conditions. This is not merely theoretically superior; acknowledging an intuitive wrong-making feature but showing that it does not obtain is surely more rhetorically forceful than simply rejecting your opponents’ intuitions about what makes things wrong, especially when you yourself have similar intuitions about similar cases (181–83).

Among the possible “theories of fairness,” he then mentions Wertheimer’s account based on a “counterfactual market standard,” not noticing that this is the “relevant theory of fairness” most likely at the basis of the literature he criticises.

Finally, and crucially for what comes next, holding to the disequilibrium price account is precisely why Powell and Zwolinski (2012, 466) can reject intuitions like Faraci’s while refraining from offering an explicit definition of what exploitation or unfairness is. The advantage of the disequilibrium price account is that such a substantive and controversial theory of “fairness and other values” is not needed, as it provides, as Wertheimer (1996, 216) emphasises, an account of “transaction-specific fairness.”

For, to repeat, take any conception of fairness we may believe applies to the distribution of market interactions’ surpluses and that may be readily acceptable to the present parties. That conception will be necessarily satisfied at equilibrium price. Because again, if a “fairer” price could obtain in such a market, then it would, through the supply of a substitute and an increment in market competition. If it does not, then it cannot, unless someone is harmed. The theory neutrality of defences of sweatshops (and, by extension, of wage exploitation) therefore already follows from step 1 and the disequilibrium price account and not solely through the NWC (in step 2), contrary to what Faraci (2019, 173) seems to suggest.

Of course, such harm may be justified—for example, we may imagine an argument showing that exchanging unemployment for living wages is a worthy trade-off, thus justifying interference with sweatshops by artificially raising wages. But such a justification requires an account of distributive justice and therefore departs from the debate on what a fair wage and exploitation are. ¹³ This is why, just as Wertheimer (1996, 216–17), Zwolinski, and Powell (Zwolinski 2012; Powell and Zwolinski 2012, 467) maintain, background and structural injustices do not matter in assessing whether a wage is fair.

Provisional Conclusion

Let me summarise what I have established so far. The disequilibrium price account considers that inefficiency gives exploitation its moral weight, that which makes it unfair. This is because in markets leading to efficient distributions, no further distribution is possible, and the one that obtains is thus fair by necessity. The account therefore establishes the conceptual impossibility of a Paretian exploiter. To criticise the literature that defends wage exploitation, one needs to re-establish that possibility. Criticising only the NWC will not do on its own; we must criticise the view—as Faraci (2019, 181) himself puts it—that “market-clearing wages are fair because market forces (tend to) ensure that market-clearing wages are fair.” We also must do it on theory-neutral grounds, without appealing to an explicit conception of fairness.

In the next two parts, I show that the reconciliation between efficiency and fairness either fails or is maintained by sacrificing theory neutrality. To do that, I shall reformulate Wertheimer’s account with several economic theories of prices.

WERTHEIMER AT CAMBRIDGE: EXPLOITATION AND STATIC EQUILIBRIUM

This part reconstructs Wertheimer’s disequilibrium price account as being based on a neoclassical understanding of what an equilibrium price is, that is, a static equilibrium where supply equals demand. In labour markets as they exist in capitalism, this equilibrium obtains when labour is remunerated at the same rate as its marginal revenue product (MRP). ¹⁴

I show that this reconstruction does not work, as there is always a hypothetical “fairer” price that could be offered through an increment in competition when labour is remunerated at that rate and that therefore the conditions for non-exploitation or fairness are not satisfied. The reason that the “fairer” price does not obtain is that it would come at a prohibitive cost for all the parties involved and would lead to an inefficient distribution. A neoclassical reading thus leaves Wertheimer in front of the contradiction between fairness and efficiency it is meant to solve, because for every equilibrium that occurs in labour markets, there is a fairer one on the account’s very own conception of fairness.

The Fair Price of Labour as Its Marginal Revenue Product

The Friends Buying a House example, with the specialist who knows the equilibrium price in a given market, implies that Wertheimer considers prices to be objectively knowable (Wertheimer 1996, 230). Neoclassical economists see economics precisely in this manner, as having predictive capacity and capable of positivist objectivity (Friedman 1953). Each actual price can be compared to a knowable, calculable counterfactual equilibrium.

Some neoclassical economists have developed an empirical account of exploitation that is based on such a comparison. Arthur Pigou (1920, chapter 14) called “exploitation” a situation where the remuneration of labour falls below its MRP. Joan Robinson (1969, chapter 25) has expanded this theory by identifying it as the difference between the marginal product (the increase in production of one additional unit of labour power) and the marginal physical product of labour (the increase in production due to the hiring of one additional worker), which, in plain words, indicates that the increase in the pool of employed workers has decreased their respective wages while increasing the profits of the company (see also Heller 2007). Both explained the occurrence of exploitation through a defect in competition leading to inefficiency—for example, monopsonist labour markets.

Wertheimer does not mention Pigou as an influence, but the conceptual similarity with my earlier interpretation of his work is striking. A neoclassical reading of the disequilibrium price account would consider that the fair price of labour is simply its MRP because no “fairer” price can—apparently—be offered. This is because no employer would hire a worker if she had to remunerate her above that rate, and no employee would accept a lower rate because she could go to the next firm to get it. No fairer price is available so that one can only be fair.

The neoclassical reading can be found in several arguments stating that workers cannot be exploited if they get their MRP, even if it is inferior to a living wage or any other intuitive distributive standard. This is indeed one of the arguments marshalled by Powell and Zwolinski (2012, 467) in their defence of sweatshops:

So multinational enterprises are not earning unusually high profits off the backs of sweatshop workers. Nor is it obvious that sweatshop workers are receiving less than they ought to earn in wages. Such a claim might be credible if MNEs were, as some critics have charged, using their monopsonistic power to pay workers less than the market rate for their labor. But as we have argued earlier in this article, there is no reason to think that workers’ wages are not determined, by and large, by their productivity—just as the wages of non-sweatshop workers are.

Zwolinski (2012, 166) adds,

The compensation sweatshop workers receive for their labor is generally no higher than the amount they contribute to the firm (their marginal revenue product), and no lower than the value they place on their next best alternative employment. Their compensation is low because the values of these upper and lower constraints are low, and not because of any failure of competitiveness in the labor market.

This idea can be found in other places. For example, when denying employers’ duties to pay living wages, Jason Brennan (2017, 20) has argued,

Suppose Unproductive Bob—like many workers around the world—is unskilled and lives in an area with low capital accumulation. As a result, he is highly unproductive. The most Bob can produce per hour is $1 of value for others. In a competitive market economy with no minimum wage, Bob would thus expect to receive about $1 an hour in total compensation. But the problem here is not that people take advantage of Bob. The problem is that employing Bob doesn’t offer anyone much advantage.

Thus, Brennan concludes, there is nothing exploitative in paying Unproductive Bob $1 for his work. My objection to the neoclassical reading will focus on Wertheimer’s original account. If successful, it will by extension invalidate these types of arguments.

A Tale of Two Factories

I will argue now that if Wertheimer is correctly understood as relying on a neoclassical understanding of equilibrium price, then it fails to reconcile fairness and efficiency. If the MRP is indeed the price of labour that obtains in an efficient labour market, it is not non-exploitative or fair in the theory-neutral sense that no fairer price could be obtained. Note before I start that I am not contesting or criticising the neoclassical price theory. I only show that something like the account developed by Pigou and Robinson cannot be what Wertheimer had in mind.

Consider first that in capitalism as it exists, market-type interactions and competition are not the dominant modes of cooperation between labour and capital, between owners and workers, between sweatshops and their employees, nor even between any two workers cooperating within the same workplace. Granted, workers compete to sell their labour power, but once they are hired, they leave the labour market. Market competition, which as we saw is the process that prevents exploitation and leads to fair distributions, is absent from most productive interactions as they occur in capitalism.

To understand why, we need to leave exploitation aside for a moment and examine the reasons why capitalistic production, understood as the coordination of various productive agents, is organised, not through markets, but in hierarchical firms. Why, if markets achieve Pareto-efficient distributions, do firms exist at all? ¹⁵ This question has been at the centre of new institutional economics. The answer they give is that firms are created to save on transaction costs. Transaction costs are what market agents face only to participate in markets and are added to the costs they face in every interaction. When transaction costs are relevantly high, they maintain the productive output under its possible limits, as summarised by Oliver Williamson (1985, 1–2):

In mechanical systems, we look for frictions: Do the gears mesh, are the parts lubricated, is there needless slippage or other loss of energy? The economic counterpart of friction is transaction cost: Do the parties to the exchange operate harmoniously, or are there frequent misunderstandings and conflicts that lead to delays, breakdowns, and other malfunctions?

Firms replace markets as a coordination mechanism between various productive agents when they improve the efficiency of the production process (Coase 1937; Williamson 1985, chapter 1).

By way of illustration, let us imagine two different factories that produce widgets, Corporate Factory and Competition Factory. For simplicity, let us assume that these two factories have only two stakeholders: owners—top hat–wearing and cigar-smoking capitalists who own the machines on which workers work—and workers, blue-collared, beer-drinking proletarians who fasten screws and various appliances on the widgets. ¹⁶ The existence of managers is ignored for the sake of the argument, for it would not change much—managers are simply workers who organise production—and their inclusion would only complicate an already complex matter. Likewise, let us assume that labour and capital are homogeneous in this fictional economy (all capital and labour that exist are used to make widgets) and that capital’s profit is justified.

In each factory, production is organised around two different modes of interaction. In the Corporate Factory, production is organised through a top-down hierarchy. The necessary labour input is provided by workers who are hired from an external labour market. In that way, workers have their relationships with owners contractually institutionalised for an indeterminate amount of time. Because the hiring is done in an external market, and just as predicted by neoclassical price theory, workers are remunerated according to their MRP. The MRP corresponds, not to individual output, but to the output of the last worker whom the company found profitable to hire—if the company paid everyone her hypothetical individual productivity, the owners would not make any profit. In that sense, most wages are not proportional to individual productivity or effort; they are just a number dictated by the bottom line of the company, the quality of available capital, the structure of the market, and the types of costs each agent faces (Heath 2018; Sen 1992, 119). Likewise, workers are paid by the owners, who retain the widgets produced and then sell them for profit.

Why do owners and workers interact in this way? Imagine an alternative to Corporate Factory called Competition Factory. There workers are not hired. Rather, they are renting a working spot on the assembly line for only one day (or any unit of time relevant for production) from owners. They are competing for that spot with several other workers at the beginning of every time unit; and likewise, the owners are competing to attract workers on the spots they have available. Just like their comrades in Corporate Factory, the workers’ job is to fasten screws on the unfinished widgets. However, they do not obtain these from capitalists like in Corporate Factory. The screws are bought from a screw shop—competing with other screw shops while the workers also compete between themselves to obtain the available screws. The widgets are bought by each worker from the worker who is before her in the production process, and she sells them to the worker who is after her in that same process, after having fastened the screw she just bought, keeping the difference as an income—an income that she will use to buy new screws and widgets and pay for the spot on the assembly line she will rent the next day. Of course, she competes with other workers who are at the same level in the production process to obtain the unfinished widgets and sell them to the next worker.

Now contrast the two factories. As you can easily imagine, Competition Factory is not only an insufferable nightmare for everyone involved but also a comically inefficient and unproductive business. Think only of the time and resources wasted by all the mentioned agents with the competition process alone—they must gather all the relevant information; front the entry cost through the purchase of the screws and unfinished widgets; waste time negotiating and making counteroffers, probably through a system of automated auctions, and so on. In other words, Competition Factory, by maintaining market-type interactions, faces very high transaction costs. For efficiency’s sake, to minimise waste and make everyone as better off as she can be through the maximisation of output, it is better to suspend market competition within the firm. The institutional layout of capitalism as a competition of various firms, rather than as a competition of various individuals, is simply more efficient and more productive. Corporate Factory will produce more widgets than Competition Factory, it will create more wealth, its owners will garnish higher profits, and the workers it employs will be paid higher wages.

Come back to exploitation. Following the disequilibrium price account, which factory is not, or the least, exploitative? Competition Factory. Every interaction that occurs in the productive process—the rental of the spot on the assembly line and the purchase of the screws and widgets from other workers and shops—occurs at a fair price. Indeed, no further redistribution can be achieved through an increment in competition. Just like Joe, Jane, and Ann in the Greedy Rescuer case presented before, everyone is a price taker; no one can take unfair advantage of anyone else. These fair prices indeed result from an increment in competition compared to Corporate Factory. The division of the surplus of every single interaction will be in that relevant sense “fairer” than in the latter because, for every interaction occurring between any two productive agents, there will be no possible counterfactual where the productive surplus will be divided in a “fairer” way. Conversely, every interaction occurring in Corporate Factory has such a fairer counterfactual: its counterpart in Competition Factory. Fairness as defined by the disequilibrium price account requires the inefficient and unproductive chaos of Competition Factory.

What the comparison between the two factories shows is that there is a point where fairness starts trading off with efficiency. Imagine that the management of Corporate Factory reads Wertheimer and, convinced by the disequilibrium price account of exploitation, decides to pay its workers the equivalent of their net benefits in Competition Factory (suppose they can calculate it through a computer algorithm). They would most likely be left with a significant surplus that, if distributed, could make everyone better off without harming anyone—what Corporate Factory saves on transaction costs by decrementing rather than incrementing competition between its productive agents. Efficiency is thus gained by reducing what would have led to a fair distribution, which is market competition. If the disequilibrium price account is correct, then it follows that someone in Corporate Factory—anyone who gains from the decrement in competition—is an exploiter who benefits from an unfair distribution of the productive surplus. The unfairness of Corporate Factory is therefore necessary for efficiency, and the inefficiency of Competition Factory is necessary for fairness.

One objection could attempt to show that wages being paid in Corporate Factory are still relevantly fair or not exploitative because it has been the result of some competition—the one that occurs in labour market before workers are hired. Call it the Fair Enough Objection. This objection would, however, require another starting point for its understanding of unfairness than simply the possibility of a further modification of the division of any interaction’s surplus through an increment in competition and thus another price theory. I shall consider it in the next part.

Efficiency and Fairness: Opposed Again

Before moving on, however, I will briefly summarise the findings of this part. The disequilibrium price account has been falsified, because Corporate Factory is efficient and unfair (exploitative) and Competition Factory is inefficient and fair (non-exploitative). Someone in Corporate Factory—it may be the owners, but that is irrelevant—is a Paretian exploiter, that is, an entity whose existence was supposed to be impossible. Note, finally, that this conclusion has been reached while maintaining theory neutrality, for I have not used or implied any “general principle of fair distribution” (Powell and Zwolinski 2012, 466). Perhaps the opposition between the two factories is just an extreme example of the NWC—Corporate Factory is unfair, but we need to let this exploitation be for everyone’s sake. But even if that is true, the example of the two factories has detrimental consequences for the disequilibrium price account, for fairness and efficiency find themselves opposed again.

WERTHEIMER IN VIENNA: EXPLOITATION AND DYNAMIC EQUILIBRIUM

Perhaps a more plausible reading of the disequilibrium price account would emphasise the process of competition as a condition for fairness, rather than the efficiency of the resulting surplus distribution. In other words, it would make what has been earlier called the “fair enough” objection. Even if a further distribution is still achievable through an increment in competition, what is relevant is that there has been some competition in price formation—enough of it to prevent exploitation. To do this, I need to reconstruct it with a theory of price that is different from the neoclassical one and that particularly emphasises market competition in its formation, namely, the Austrian school of economics. As I show, such a reading maintains the reconciliation of efficiency with fairness. However, it does so at the cost of theory neutrality.

An Austrian Reading of Wertheimer

Austrian economists consider markets to be coordination mechanisms where prices convey information about the relative scarcity or availability of a given good. Any price is only but a moment in a dynamic, ever-going process that equilibrates supply and demand according to what economic agents learn on the go. These agents form these prices by acting on the partial and localised knowledge they acquire through uncoordinated economic competition (Hayek 1945; von Mises 1998, 333). Objective knowledge on the state of any given market is impossible, and that is why, according to the Austrians, we need markets. This means there is no way we can know, for every interaction that occurs, what exactly a fair (or any) counterfactual price would be. Fairness is thus more a feature of an efficient price formation process that is sufficiently competitive. Instead of saying “an efficient price is necessarily fair,” we would say “a price reached through a competitive price formation is fair enough; there cannot be anything unfair about it.”

Even if it goes against the grain of the already mentioned specialist example discussed in the first part and that hinted at a neoclassical reading of Wertheimer (1996, 230), the Austrian reading has the advantage of heuristic simplicity. Because equilibrium is by assumption never achieved and unknowable, what determines whether a price is exploitative is the process that led to it, not a static counterfactual.

An obvious example is the aforementioned case of the external labour market where firms compete to hire workers, like in Corporate Factory. Even if a fairer wage could be obtained in a different mode of organisation—Competition Factory—the competitive process that has occurred before the hiring means that the price of labour in that example cannot be reasonably considered exploitative. Jane and Joe in Corporate Factory both could turn to a competitor if the wage offered is not to their taste; so as long as there are competitors around and sufficiently low exit costs, the price they end up accepting is pro tanto fair.

As before, I shall focus on Wertheimer’s original version of the disequilibrium price account. However, the objections presented here will by extension apply to some of the aforementioned arguments defending cases of wage exploitation—and especially Zwolinski and Powell’s work on sweatshops. They indeed seem to understand the markets in which they operate in this way, not only emphasising how competitive they are (see the previous quotes) but also noticing, in a very Austrian fashion (Hayek 2002), that their primary purpose is to palliate to asymmetries of information (Powell and Zwolinski 2012, 451–53).

The Problem of Natural Monopsony

Has the Austrian reading reconciled efficiency and fairness? I show now that it can only do it by giving up on theory neutrality. The Austrian reading cannot say much on the fairness of the division of an interaction’s surplus without committing to a substantive conception of a just division.

To understand why, consider the problem of monopsony in labour markets. Monopsony is a situation when there is only one buyer of a given good; in the cases that interest us, it is when there is only one employer. It is the reverse of a monopoly, where there is only one seller. Imagine the following Mining Town example, which is an archetypal case of unfair monopsony. in a poor, developing country, there is an isolated village that lays near a deposit of a precious metal that is particularly useful for producing widgets—call this metal widgetium. A corporation arrives to mine the deposit and employs the population of the village. The corporation reaps huge benefits while paying them low wages, barely sufficient to ensure the workers’ subsistence, whereas, given its monopsonistic position, it could pay higher wages. It is a prime example of what David Faraci has assumed is intuitively exploitative, as it has the same structure as in Owner/Employee. The owners of the company mining the widgetium can indeed increase the wages of its employees without compromising its profitability (Faraci 2019, 269).

Following the disequilibrium price account, and as explained when discussing Faraci’s version of it, such an example should be assessed as a prime example of exploitation. Does this assessment come through the Austrian reading? The answer is that it does not. This could strike, at first sight, as paradoxical. Are not monopsonies characterised precisely by an absence of competition? ¹⁷ But here is the subtlety: the unfairness that gives exploitation its moral weight as identified by the Austrian reading is not the absence of competition but a defect in it.

An absence of and a defect in competition are two different explanations of why the mining corporation is a monopsonist. The first is purely economic—coming to the village is very expensive; it is a market with very high fixed entry costs. There is no way anyone would be willing to front that cost without sufficient initial capital to do so. From an Austrian perspective, the the wage the company will offer for the inhabitants’ labour is the price that in its eyes made coming worthwhile in the first place. The company has simply read the price mechanism before others, seen the low price of the available labour, and seized the opportunity, thus behaving as economic agents are supposed to behave. From that perspective, there is nothing unfair in that example. ¹⁸

Conversely, a defect occurs when market competition was prevented from running its course by the application of non-economic forces. A typical example—but there are many others—is when a company uses the power of the state to restrict the number of buyers or sellers to its advantage or other types of legal tools that rig competitive rivalry between market agents. Imagine for instance that the entry costs in the Mining Town example are high not because of the isolation of the village but because some regulations are favourable to the first corporation or, perhaps, that corporation has friends in the government that prevented competitors from settling in the area or has used armed guards for that same purpose. A monopsony counts as a defect in competition only when it was born out of a unilateral abuse of power. ¹⁹

To better grasp the difference, contrast the neoclassical with the Austrian reading. The latter looked forward. It provided us with a counterfactual equilibrium price, and if it did not obtain, then we knew that exploitation had occurred. With the Austrian reading, we have no such counterfactual. We have to look backwards and see whether the price the mining corporation pays followed a defective competitive process. Absence, however, is not a defect.

The reason the Austrian reading is not theory neutral is what makes the difference between absence and defect. A defect in market competition, necessary for unfairness or exploitation, follows an uneven enforcement of property rights. The account therefore requires the assumption of a background structure of independently defined rights, with exploitation gaining its moral weight from their unequal enforcement.

The non-exploitative counterfactual here is the price obtained in a market in which property rights are equally enforced. But this counterfactual relies on a “general principle of fair distribution” (Powell and Zwolinski 2012, 466): the libertarian idea that whatever follows from a voluntary transfer of claim rights is just and that any interference with such a transfer is unjust (which is what a defect in competition is). Fairness is here reconciled with efficiency, but at the cost of theory neutrality.

Provisional Conclusion

Let me summarise once again. I have shown that the literature that justifies wage exploitation relies on the disequilibrium price account as formulated by Wertheimer. Criticising that literature requires criticising the account’s reconciliation between efficiency and fairness.

I have done so by undermining two possible economic reconstructions of the account. The first one, the neoclassical reading, has failed to reconcile fairness and efficiency. The second one, the Austrian reading, has maintained that reconciliation has to abandon theory neutrality, for it does rely on a substantive view of what a fair price is. The disequilibrium price account cannot therefore reject the intuitions some have about examples of wage exploitation simply by appealing to its reconciliation of fairness and efficiency.

How does this result bear on the literature supported by the account and which defends intuitive cases of wage exploitation as being non-exploitative? There are three ways out. The first is to rely on the neoclassical price theory, but such reliance would make the defence invalid by nullifying the reconciliation between efficiency and fairness that stands at its core. The second is to rely on the Austrian price theory, but that means forfeiting any claims to theory neutrality. The discussion would collapse into one where the intuitions concerning “fairness and other values” (like Faraci’s regarding living wages in Owner/Employee) become relevant again. A third solution would be to rely exclusively on the NWC (employers have no duties towards their employees and that is the end of the story). But that would mean running into the objection made by Faraci with his RV.

EFFICIENT UNFAIRNESS AND INEFFICIENT FAIRNESS: DOES INEFFICIENCY HAVE MORAL FORCE?

There is, however, a remaining problem. As has already been shown, Faraci’s Owner/Employee illustrates the reasonableness of the RV in defective markets. Its reasonableness in competitive markets, however, still needs to be demonstrated. Consider therefore a variation, called Employer/Employee: Employer buys Employee’s labour in a competitive market. However, because of Employee’s low productivity, the wage she gets does not correspond to what could be defined as a living wage (or any intuitive baseline you deem relevant). Suppose now that there are compelling intuitive reasons to deem that interaction to be unfairly exploitative. Can it be reasonable to think Employer is a wrongful exploiter while still thinking third parties ought not to interfere?

The Reasonable View and the Distinction between First- and Third-Party Obligations

Faraci’s RV is based on the distinction between first and third party’s obligations regarding exploitation (Faraci 2019). Remember from the introduction that the RV is that we may consider that employers have a first-party obligation to decrease their profits to pay fair wages and also that third parties have an obligation not to interfere if that puts the employment relationship at risk. The distinction makes sense in Faraci’s original Owner/Employee or in Mining Town. Even if, in these cases, there seems no economic reason why third-party interventions would have detrimental effects—as these would only correct defective markets—it is worthwhile to highlight that first and third parties have separate obligations (Owner/the mining corporation has an obligation to increase wages independently of any regulation and is thus an exploiter if it fails to do so voluntarily).

In the case of Employer/Employee, it seems unnecessary to make such a distinction, because first and third parties will face the same ethical conundrum. As already explained, in competitive markets, if wages could be increased by decreasing profits, then Employee would get that trade-off from Employer 2. Past the equilibrium price that so obtains, any further increase will have the same consequence regardless of whether it follows first- or third-party correction: it will only make some workers better off—the ones who are hired because their productivity is at least equal to the new higher rate. But it will harm workers who would want to work at that rate despite having a productivity level falling below it and who now find themselves unemployed. Any first-party correction through voluntary provisions of higher wages will therefore have the same effect as third-party interference (see Zwolinski 2007, 696–700; Powell and Zwolinski 2012, 457–60).

The obligations of either party in cases like Employer/Employee are, in my view, better analysed through what Wertheimer has called the problem of moral force. Remember from the introduction that the moral weight of exploitation is what makes it wrong. Conversely, its moral force is the set of reasons that justify third-party interference with it on the grounds of its wrongness (Wertheimer 1996, chapter 9). The RV would be here that there exists exploitation that lacks such moral force. At first, Employer/Employee seems to be such a case, as “there may be good reasons—good moral reasons—that A has a right to take unfair advantage of B, wrong though it may be” (Wertheimer 1996, 286). These good moral reasons are the same as Zwolinski’s criticism of third-party interference with sweatshops: a concern for B (Employee) herself. If the wage she receives for her labour is the equilibrium price of a competitive labour market, then any correction might lead to her losing her employment. Inefficient unfairness thus has moral force; efficient unfairness does not. Moral force in its original formulation appears, then, a concept whose sole purpose is to prevent the harmful correction of efficient but intuitively unfair market distributions.

How to Assess Efficient Markets?

I will now show that just as for moral weight, inefficiency is not what gives exploitation its moral force. Efficient unfairness may very well have moral force, and first- or third-party corrections do not require inefficiency to be justified. To make my point, I will criticise Wertheimer’s contradictory treatment of two examples of exploitation, with one lacking moral force (low wages) and the other having it (blackmail).

What does Wertheimer think of cases of efficient unfairness such as Employer/Employee? Unsurprisingly, he argues for the position deduced earlier (Wertheimer 1996, 301–5). In defective labour markets, that is, in cases like Owner/Employee, he does advocate minimum wage laws and unionisation as remedies to low wages, in a move that recalls Pigouvian defences of collective bargaining (Takami 2009; Richard 2020, 290). Such third-party interferences are indeed justified for what he calls “strategic reasons,” when regulations correct existing inefficiencies and are thus beneficial to workers. ²⁰ But as far as cases like Employer/Employee are concerned, he is clear that

it must be stressed that the strategic argument for the prohibition of mutually advantageous exploitation applies only when the prohibition is, in fact, beneficial to B ex ante, as when A would otherwise be able to exploit a monopoly position vis-à-vis B. The strategic argument has no purchase on contexts in which the terms on which A and B transact are set by a fully competitive market (Wertheimer 1996, 303).

For in that case, “we have every reason to think that A would refuse to transact with B, to B’s detriment” (Wertheimer 1996, 303).

Wertheimer seems here to leave no space for the RV. If exploitation has moral weight, then it will have occurred in a defective market, and it will have by extension moral force. Consider, however, how he defends the criminalisation of blackmail. At first, he believes it to be a case of exploitation that has moral force and thus that it should be forbidden (Wertheimer 1996, 301). ²¹ Here Wertheimer makes the same argument from “strategic reasons.” Such legislation makes victims of blackmail as a group collectively better off because, first, they allow them to take legal action against blackmailers and, second, they disincentive the collection of compromising material on other people that could then be used against them.

But Wertheimer seems not to follow his own account to its end. Consider what would happen if we were to legalise blackmail. The Pareto improvements due to such legalisation should be obvious. It would indeed entail a market for security and insurance on private information, on which potential individual victims could purchase greater control of their secrets at an equilibrium price. But such an arrangement would strike us as unfair, because blackmailing, in general, is unfair. This leads to the correct conclusion—blackmailing is unfair and is a type of exploitation that has moral force no matter what, even if its prohibition leads to an inefficient result. The issue of the efficiency of a market for private information and how fair a price on it would be is here completely irrelevant.

Blackmail is therefore a case when efficient unfairness may have moral force. Purchasing fairness at the cost of market efficiency (or vice versa) might be a worthy trade-off. Here again, moral neutrality fails—for it is arbitrary that this trade-off is acceptable in the case of blackmail and unacceptable in the case of intuitive wage exploitation.

Provisional Conclusion

We ought to conclude that inefficiency does not ground the moral force of exploitation either, and in general, purchasing fairness at the cost of market efficiency (or vice versa) might be a worthy trade-off. This vindicates the reasonableness of Faraci’s RV in cases of wage exploitation occurring in competitive markets. We may reasonably hold cases of efficient unfairness—like legalised blackmail—to be indeed unfair, while also considering that some cases should be interfered with (e.g., blackmail should be illegal), while some other cases are better left alone—sweatshops and low wages (perhaps). Regardless of whether these views are sound, there is, in Faraci’s (2019, 170) words, “theoretical space” for all of them. That is because none of the reasons we have to accept or reject them are solely related to efficiency but also stem from pre-existing judgements on each particular case.

CONCLUSION

I have argued that philosophical defences of wage exploitation are based on Wertheimer’s account of exploitation as price disequilibrium. The account sees the unfairness of certain exchanges as stemming from Pareto inefficiency. I have then shown that such defences fail on their own conditions by breaking the reconciliation between efficiency and fairness on which they are based or by having to abandon theory neutrality to maintain it. Having shown that inefficiency cannot provide any moral weight, I have ended the article by showing that it cannot, by extension, provide moral force. Efficiency should now appear for all its normative irrelevance in debates on fair wages and exploitation.