Exploitation and the Desirability of Unenforced Law

Abstract

Many business transactions and employment contracts are wrongfully exploitative despite being consensual and beneficial to both parties, compared with a nontransaction baseline. This form of exploitation can present governments with a dilemma. Legally permitting exploitation may send the message that the public condones it. In some economic conditions, coercively enforced antiexploitation law may harm the people it is intended to help. Under these conditions, a way out of the dilemma is to enact laws with provisions that lack coercive enforcement. Noncoercive law would convey the state’s condemnation of wrongful exploitation without risking the harmful effects of coercively enforced law. It would also give firms and their agents a way of explaining nonexploitative pricing decisions to investors, and it may help give precise content to the moral duty to set prices and wages fairly. Governments should thus consider noncoercive law a viable component of their responses to exploitation.

One of the central normative questions for the regulation of business is how government should respond to wrongful exploitation that is consensual and arguably beneficial to both parties. A transaction or a relationship is wrongfully exploitative if one party wrongfully takes advantage of another. ¹ Some exploitative transactions and relationships are nonconsensual, harmful to the exploited party, or both. Sometimes, though, both parties to a transaction or relationship give fully informed, morally and legally valid consent, and both parties benefit (compared with a nontransaction baseline), but the transaction or relationship is nonetheless wrongfully exploitative (Wertheimer, 1996). ² Consider a wealthy household’s decision to hire a gardener for full-time, exhausting work at a wage that is insufficient to support the gardener’s basic needs. The gardener might understand the terms of the job, consent to them, and regard the job as preferable to the alternatives. The household wrongfully exploits the gardener by not paying the gardener a living wage despite being able to do so. Arguably, many companies wrongfully exploit workers by paying very low wages despite being able to pay a living wage (Goodin, 1987; Kates, 2019; Meyers, 2004; Sample, 2003). Mutually beneficial, consensual exploitation can also involve an unfairly high price for a good or service. Possible examples include some high prices for medically necessary drugs and for aid during emergencies (Wertheimer, 1996). ³

Given that there is such a thing as consensual exploitation that is wrongful and (by at least one measure) mutually beneficial, there is a difficult question of how, if at all, the legal system should respond. It is relatively uncontroversial that government is justified in banning transactions that are either nonconsensual or harmful to one of the parties. There are potential objections to governmental interference with mutually beneficial, consensual exploitation. This interference could cause unfair transactions and relationships to be replaced with fair ones—a happy outcome. It could instead harm the people it is intended to benefit. It could result in unfair but mutually beneficial transactions being stopped and not replaced. Perhaps more importantly, it could also shut down some perfectly fair transactions. Making the judgment whether a transaction is wrongfully exploitative involves a judgment about whether one party could feasibly have conducted the transaction differently. Sometimes the latter judgment requires information or insight that is private and difficult to share with regulators. In these contexts, attempts to stop mutually beneficial exploitation are likely to stop some fair transactions as well. Because of the risk that regulation of mutually beneficial exploitation will harm the people regulation is intended to benefit, Alan Wertheimer argues that mutually beneficial, consensual exploitation should usually be legally permitted. Regulation is only appropriate, on his view, when we have good reason to believe that fair transactions will replace the unfair ones (at least for the most part).

Coercively interfering with exploitation and legally permitting it are not the only possible legal responses. We could have antiexploitation laws that include explicitly unenforced provisions; that is, we could enact legal prohibitions that explicitly lack any coercive enforcement mechanism, including penalties, compelled payment of damages, directly compelled compliance, or even public findings of guilt. ⁴ Despite the lack of coercive enforcement, we could take these prohibitions seriously, regard them as genuine prohibitions (not as public advisories or “suggestions”), and demand that people and firms comply.

I argue that legislators should consider it a viable policy option to enact antiexploitation laws with explicitly unenforced provisions. Consider wage regulation. In some economic contexts, an enforced requirement that all employers pay a living wage would do more harm than good. In these contexts, governments should consider enacting a law that combines an enforced minimum wage, lower than the living wage, with a requirement that all employers pay a living wage if they can. The law could specify some conditions that clearly indicate ability to pay employees more (e.g., recent payment of a dividend). If these conditions apply to a firm, the living wage law would be coercively enforced via a threat of fines or damages. The law would explicitly state that the requirement is binding but unenforced when the identified markers of ability to pay are absent. No doubt some employers would flout the unenforced component of the wage law, and it would have limited efficacy as an instrument of distributive justice. But including the unenforced provision is unlikely to do harm, and it would serve three purposes. First, it would authoritatively clarify what counts as fulfilling the ethical duty not to exploit workers. Second, it would express the state’s condemnation of wage exploitation. Finally, it would give ethical managers a way of explaining their wage decisions to profit-oriented stakeholders who recognize law’s authority.

Whether societies should adopt this specific proposal depends on empirical questions I cannot address and whose answers may vary in different social and economic conditions, for example, the question of how an enforced requirement to pay a living wage would affect the employment rate. It also depends on difficult value judgments about how policy makers should weigh harms, risks, and benefits. (If adding an enforcement provision to a law would give N ₁ people an X dollar raise, put N ₂ people out of work, and bring N ₃ people who were not employed into the labor force, should it be enacted?) My aim is not to advocate a specific policy proposal but to defend a broader point. The idea of noncoercive law is not a mere philosopher’s thought experiment, applicable only to hypothetical societies of angels or saints. Noncoercive law is a real option that societies of flawed human beings should consider. The regulation of exploitation is a context in which laws with noncoercive provisions are likely to be desirable.

1. How is NonCoercive Law Possible?

By a noncoercive law, I mean a law that imposes a requirement, not a mere recommendation, and that explicitly omits authorizing any form of coercive enforcement. The executive branch is not permitted to enforce the law by imposing penalties or directly compelling compliance. ⁵ Private persons cannot enforce the law coercively by suing. The law is nonetheless presented as binding. It is at least somewhat unusual for legal prohibitions concerning the conduct of private individuals or firms to be noncoercive in this robust sense. It is common for laws regulating the conduct of public officials to lack legal enforcement. In the United States, the president and Congress have legal duties, including constitutional duties, that the courts are not empowered to enforce (Sager, 1978). Jurors in Anglo-American legal systems are legally required to apply the judge’s instructions, but they cannot be punished for failing to apply these instructions in the way the judge thinks they should. ⁶ Allowing a judge to sanction a jury for reaching the wrong decision would reduce the independence of the jury’s judgment. We impose legal requirements that lack sanctions on public officials when we want those officials to exercise their own judgment. Why not impose legal requirements that lack sanctions on private citizens when we want citizens to use their own judgment in carrying out these requirements?

Legal philosophers have long recognized the possibility of binding laws lacking coercive enforcement, including laws that apply to private citizens as well as laws that apply to public officials. ⁷ How can an unenforced rule constitute a binding law? Different legal philosophers have different accounts of what law is and of what makes law binding. To illustrate two ways of explaining the possibility of noncoercive law that is binding, I will consider two recent, influential theories of law, one from the positivist tradition and one from the antipositivist tradition.

Shapiro’s (2011) account of law is representative of the positivist tradition, according to which social facts alone (to the exclusion of moral facts) determine what the law is. Shapiro argues that legal systems are social planning organizations with three specific features. First, legal systems are “compulsory” in the sense that their socially recognized claim to be able to settle questions normatively applies to people whether or not they consent to be governed (211–12). People can disobey the law, and they may sometimes be able to do so with impunity, but people cannot opt out of a legal system’s socially recognized authority. In this, the authority of a legal system differs from the authority of a private organization, such as a club or an industry association. Second, legal systems are “self-certifying,” meaning that they are free to enforce their rules without seeking permission from a superior organization (221). Florida has a legal system distinct from the federal legal system because Florida can enforce its laws without asking the federal government’s permission. The Condo Board of Del Boca Vista does not have its own legal system, because it needs help from the state of Florida to enforce its rules. Third, legal systems hold themselves out as having the aim of solving “those moral problems that cannot be solved, or cannot be solved as well, through alternative forms of social ordering” (225). There can be wicked legal systems, but an organized criminal operation that does not even pretend to have a moral aim is not a legal system.

On Shapiro’s (2011) account, laws are, roughly, rules either created by or incorporated into a legal system. ⁸ Duty-imposing laws are binding in only a limited sense, on Shapiro’s view. ⁹ They are pieces of the legal system’s overall plan for social cooperation. To be bound by law is to be bound from the legal system’s point of view (186–88). What is legally obligatory is not always morally obligatory, and people are not always given an incentive to do what the law requires. Though Shapiro thinks the ability to enforce law without asking permission from another organization is essential to a legal system, he explicitly denies that a rule must have sanctions attached to count as a law. Indeed, on Shapiro’s view, “there is nothing unimaginable about a sanctionless legal system” (169). To show this, he uses a thought experiment in which friends from a cooking club settle an uninhabited island together. If they create a legislative system to allocate the island’s resources and a judicial system to address good faith disputes about the rules, they have created a legal system, even if no enforcement mechanism is ever established.

According to Greenberg’s (2014) moral impact theory, law is binding in a more robust sense: it is morally binding. Greenberg’s account is in the antipositivist tradition, which holds that moral facts as well as nonmoral social facts determine what the law is. Greenberg defines law in terms of the effect legal institutions have on what we can and may do morally. Greenberg uses the term moral profile to refer to “our moral obligations, powers, privileges, and so on” (1308). He sums up his account of law thus: “the content of law is that part of the moral profile created by the actions of legal institutions in the legally proper way” (1323). The term legal institutions is explicitly left undefined (1323–25), but it clearly includes legislatures and courts. One way in which legal institutions can alter the moral profile is by giving content and specificity to a preexisting moral duty. Drivers everywhere have a moral duty to drive safely, and this requires that everyone drive on the same side of the road. By enacting a statute, a legislature can specify on which side people should drive. This phenomenon is pervasive. There is a universal moral duty to support public goods; tax law specifies what people must contribute. There is a universal moral duty to respect property rights; the law specifies what types of property rights exist and how they may be acquired and transferred.

One feature of Greenberg’s (2014) account is that the content of the law can diverge both from the text of statutes and from the way in which the law is enforced. Suppose that a town council enacts a speed limit law whose text prohibits driving faster than thirty-five miles per hour on local roads. Suppose that the town’s small police force would not respond to speeding unless the police see someone going extraordinarily fast—say, sixty-five miles per hour on a winding two-lane road. There might be a moral reason, such as respect for the wishes of the town’s residents, to avoid exceeding the speed limit greatly, though there is no reason to obey the posted speed limit precisely. (Most residents do not expect or want drivers to obey the posted limit precisely.) Maybe it would be objectionably disrespectful to drive fifty miles per hour in this town, even in good conditions, though it would be fine to drive forty miles per hour. If so, then on Greenberg’s account, it is legal in this town to drive forty miles per hour, but not to drive fifty miles per hour.

Whether coercive enforcement is necessary to establish a morally binding legal requirement is a complex and context-sensitive question. It is debatable when widespread noncompliance with a norm undermines the moral obligation to follow that norm (and, with it, on Greenberg’s account, its status as law). Certainly if an organization that holds itself out as a government completely fails to influence behavior, it fails to establish a legal system. But the moral obligation to comply with property law or tax law cannot depend on perfect compliance by one’s fellow citizens. People sometimes break laws of both kinds with impunity. That fact does not make stealing or tax evasion morally acceptable. So, on Greenberg’s account, if a legislature enacts a statutory provision with no enforcement mechanism, and if this act of the legislature succeeds in imposing a moral obligation that aligns with the unenforced requirement in the text, the legislature creates a law that lacks coercive enforcement. The unenforced requirement is binding law because it is morally binding.

So, on both positivist and antipositivist accounts of the nature of law, binding laws, including laws applicable to private citizens, can lack coercive enforcement. Noncoercive laws concerning the conduct of public officials are common; noncoercive laws concerning the conduct of private citizens are rare. I do not see a sound reason for this asymmetry. For one reason or another, authors of laws and constitutions have often judged that government officials can be trusted to carry out certain legal obligations without a coercive enforcement mechanism. Perhaps there is no alternative to trusting them, even if we know this trust will sometimes be violated. But if it is sometimes appropriate to trust officials, why would it not sometimes be appropriate to trust private individuals? In a range of cases, noncoercive law regulating the conduct of private citizens is not only possible but desirable.

2. When is noncoercive law desirable?

Having legal prohibitions that lack coercive enforcement makes sense when legal prohibitions would serve a function that can be achieved, to some degree, without coercion, and when coercive enforcement of these prohibitions would be unjustified. If the reasons against coercive enforcement can be expected to persist, it may make sense for a legal prohibition explicitly to disclaim the availability of a coercive remedy. Law has three functions that can sometimes be performed without coercion or the threat of it. Public recommendations that lack the status of law cannot perform these functions as well (or at all).

The first of these functions is to give specificity to moral duties whose content is unclear in the absence of authoritative law or custom. The moral duty to respect others’ property is an example. In the absence of laws or customs determining who owns what things (e.g., when does a bicycle on the side of the road belong to someone, and when is it abandoned?), it is unclear what constitutes theft. For a norm presented by a government to determine authoritatively the content of a moral duty, such as the duty to respect property rights, the norm must be socially regarded as binding. Laws, including explicitly unenforced laws, can authoritatively clarify moral duties. Public advisories and recommendations do not hold themselves out as binding and are not socially regarded as binding. So explicitly unenforced law can perform a function that public advisories and recommendations cannot.

Explicitly unenforced law can also serve an expressive function. ¹⁰ By enacting a legal prohibition, a society publicly acknowledges the prohibition as a standard of the community. The value of doing this may include effects on behavior. People may be less inclined to do something that is against publicly recognized community standards, even if there is no sanction for violation. The articulation of community standards may also have value apart from the effects that articulating these standards has on the frequency of the prohibited behavior. Shiffrin (2021) argues that in a democracy, that is, in a society in which citizens are coauthors of the laws that govern them, using law to articulate public standards of conduct is a way for citizens to express their commitment to treating each other with respect. A public expression of a commitment to treating each other with respect helps people to respect themselves. In Rawlsian language, publicly expressed standards of conduct are among the social bases of self-respect. Public recommendations do not express commitment to a standard of conduct in the way that laws (enforced or unenforced) do.

Law has a third function: it gives people a way of justifying their actions to others who recognize an ethical duty to obey the law. Many recognize such a duty. ¹¹ Even on Milton Friedman’s (1970) view of business ethics, according to which executives of publicly traded, for-profit corporations have a moral duty to maximize financial returns to shareholders, pursuit of value for shareholders is to be constrained by “law and ethical custom.” ¹² Managers who decline to do something they regard as profitable but unethical may have difficulty explaining their actions to shareholders (or to other stakeholders) if their ethical views are not universally shared. Managers who decline to do something illegal can justify their actions by pointing to the legal requirement—even if that requirement has weak sanctions or no sanctions attached. Public recommendations and advisories cannot perform this function of law, because they neither hold themselves out as binding nor are recognized as binding.

One might think that when law performs any of these functions, coercive enforcement will promote law’s ends. Punishing people who do not go along with the rules can facilitate establishing social rules that give specificity to moral duties. A community arguably condemns conduct more strongly if its articulation of a norm has an attached penalty. Managers can better explain a refusal to do something that looks likely to increase profits by pointing to a law that has sharp teeth. Nonetheless, a society might want to leave a legal prohibition unenforced for various reasons.

One reason will be salient here: regulated parties may know better than regulators or police what conduct would be consistent with a law or regulation and what conduct would violate it. Depending on the effects of under- and overenforcement, asking officials to second-guess the decisions of regulated parties may be undesirable. As an example, suppose that the problem of drug-resistant bacteria grows, and scientists judge that we should select three antibiotics from a certain class—it does not matter which three—and reserve these for cases in which a human’s life or health is gravely threatened. ¹³ Law or some other form of public authority is needed to select which three drugs to limit in this way. A community norm about which drugs shall be limited will not emerge spontaneously. The decision to limit these drugs’ use cannot be presented as a mere public advisory. Offering a public advisory that people limit their use of three antibiotics would suggest that patients have a self-interested reason not to use these drugs and that doctors should limit their use of the drugs out of concern for their patients. But the proper reason to limit one’s use of these drugs is concern for the health of the community —a concern that may be at odds with some patients’ interests. Only a rule that is presented as mandatory would adequately convey the way in which public authorities want doctors to change their behavior. What is needed is a law.

Should this law be coercively enforced? It would be appropriate to penalize people who use these drugs in agriculture, and it would be appropriate to penalize doctors who prescribe these drugs when they are clearly inappropriate (e.g., for an infection the doctor knows to be viral). But there will be cases in which doctors could reasonably disagree about whether a bacterial infection presents a grave threat to a human patient. If doctors face no penalties for prescribing the three drugs when they were not needed, some doctors will overprescribe them (partly because of pressure from patients to “do something”). If regulators are authorized to impose penalties on doctors who overprescribe, there is a risk that regulators will overimpose fines (partly because of pressure from legislators and the public to “do something”). Doctors may underprescribe in response. Whether overprescription or underprescription is a more serious risk is a complex question. The answer turns both on empirical facts about what effects enforcement and nonenforcement can be expected to have and on difficult moral questions about how different harms and benefits should be weighed. If underprescription is the more serious risk, it would be best to leave the prohibition explicitly unenforced in cases that approach the borderline.

One might worry that in the absence of coercive enforcement, a legal prohibition would fail to express condemnation of prohibited conduct. Normally, the severity of the penalty attached to a prohibition signals a legislature’s view of the gravity of an offense. The extent of the resources devoted to detection and enforcement likewise signals the executive branch’s view of a law’s importance. If a prohibition explicitly lacks enforcement provisions, would that not signal that the government takes the prohibition even less seriously than a prohibition with a light penalty that is rarely applied? This would indeed be what the lack of a penalty expresses if the legislature had no reason other than apathy to omit the penalty. Matters are different if a legislature’s reason for omitting enforcement has to do with the competence of enforcers or with the ill effects enforcement would have. The lack of a penalty for failure to follow jury instructions in no way signals that the judge does not take jury instructions seriously. There is no penalty because we want jurors to exercise their own judgment in applying jury instructions. Likewise, we may omit a penalty for excessive uses of antibiotics because we want doctors to exercise independent judgment in applying the principle we set out. In such contexts, lack of a penalty signals our recognition that enforcement would have perverse effects.

3. Regulating Wrongful Exploitation

A similar normative structure can arise in regulating wrongful exploitation. Sometimes the rule we would ideally like market participants to follow to avoid exploitation is a rule that cannot be applied properly by courts or enforcement agencies; the rule can only be properly applied by the regulated parties. When this is the case, it is not obvious either that we should have coercive enforcement of the best rule or that we should retreat from enacting the rule we want market participants to follow in favor of a less demanding rule that we can justifiably enforce. The possibility of having an unenforced law or a law with unenforced components should be a live option. This section explains how noncoercive law can be an appropriate response to the wrong of exploiting workers through low wages. A living wage law with noncoercive components (in addition to some coercively enforced components) can advance three functions of law: giving determinate content to people’s moral duties, expressing public condemnation of wrongful behavior, and giving managers a way to explain their decisions to shareholders and other stakeholders. I then briefly explain why a parallel analysis applies to other forms of wrongful exploitation (not just to wage exploitation).

Kates (2019), Sample (2003), and Snyder (2008) have offered accounts of wage exploitation that entail the following claim: if an employer can hire a worker at a living wage while still benefiting from the employment relationship, then it would be wrongfully exploitative to hire that worker at substantially less than a living wage. The accounts differ in the precise demands they would make on employers, in part because the three authors have different conceptions of need and vulnerability. ¹⁴ The accounts also differ about what grounds the duty of nonexploitation. Sample (2003) and Snyder (2008) ground the duty of nonexploitation on the Kantian duty not to treat people as mere means. Kates (2019) understands the duty as a duty to divide the economic gains a transaction generates in a fair way. Despite the differences among these accounts, there is an important overlap. I take their arguments to have at least made plausible the claim that employers who can pay a living wage, while still benefiting from hiring, have a strong pro tanto ethical duty either to pay a living wage or to refrain from hiring. ¹⁵ For brevity, I will call this the conditional duty to pay a living wage, understanding that there are important questions about its grounds and interpretation and that some of these accounts support further moral demands.

It is important to be clear about what the conditional duty to pay a living wage does and does not imply. It does not imply that people have a duty to offer jobs as a form of charity, that is, to aid needy people by offering them jobs that will not be a net benefit to the employer. One need not create jobs that pay fifteen dollars an hour for work that one values only at eight dollars. The requirement also does not imply that one must refrain from offering jobs at a sub-living wage if the only economically viable alternative is to refrain from offering these jobs at all. If an employer would get only eight dollars per hour of benefit from hiring people for a certain task, it is not exploitative (on these accounts) to pay people eight dollars per hour to do this task. On Kates’s (2019), Sample’s (2003), and Snyder’s (2008) accounts, mutually beneficial, consensual wage exploitation can occur only when there is a range of possible wages at which the employment relationship benefits both the employer and the employee, compared with a baseline in which the employer does not hire anyone for the job. If an employer would be willing to hire people for a given task if the market wage were twenty dollars, but the market wage is in fact eight dollars, the employment relationship benefits the employer at any wage between eight and twenty dollars. ¹⁶ If an employer gains a substantial welfare surplus by hiring at less than a living wage, and the employer could pay employees more, the employer wrongfully exploits the workers. The employer should instead either refrain from hiring or pay more. Because the employer would still benefit from hiring at a higher wage, the typical way to carry out this ethical duty would be to raise wages.

If Kates’s (2019), Sample’s (2003), or Snyder’s (2008) defense of the conditional duty to pay a living wage is correct, employers have this conditional ethical duty whether or not there is a corresponding legal duty. That said, on all three accounts, what counts as a living wage is vague. Businesspeople who want to compete effectively on the market while complying with their ethical duties may want this vague duty to be made precise. By authoritatively determining the living wage in a jurisdiction, legislation can clarify the underlying ethical duty. ¹⁷ Compared with a law imposing an unconditional wage floor, a living wage law that identifies the living wage and requires employers to pay this wage if they can is advantageous. When a labor market is perfectly competitive, imposing an unconditional wage floor that is coercively enforced and above the market wage will result in some people losing their jobs and others getting a raise. It may produce further harms and benefits. ¹⁸ Whether it is good to impose a wage floor that benefits some workers and harms others is a difficult question. In some economic environments, an unconditional requirement to pay all workers a living wage would be, on balance, harmful to workers. A law requiring employers to fulfill the conditional duty to pay a living wage would not be harmful in this way.

There would be an obstacle to enforcing a law that perfectly tracks the conditional duty to pay a living wage. It is often infeasible for outside observers, such as regulators, to assess whether an employer is exploiting workers. The question whether an employer gains a welfare surplus from hiring depends on the employer’s reservation wage, the highest wage the employer would pay for labor if the market price changed. The employer knows this number; the public does not. The broader question whether an employer can pay workers an above-market wage is complex and fact-dependent. Some factors that affect ability to pay a living wage are whether the firm is profitable; whether the firm needs to build cash reserves to weather economic changes; whether the firm needs to reinvest profits in a new product to replace an outdated product; whether the firm has power in the markets for its products and, if so, whether it could make more money by changing prices; whether the firm has power in markets for factors of production, including markets for skilled labor, and, if so, whether it could make more money by bargaining for lower prices or lowering wages for skilled labor; and how well the managers making hiring decisions are themselves compensated. ¹⁹ Whether a firm can raise wages may depend on facts that only a manager intimately familiar with the firm’s finances, business strategy, and economic environment would be in a position to know. ²⁰ If enforcing an unconditional wage floor at a living wage level would put too many people out of work, attempting to enforce the conditional duty to pay a living wage would likely have the same effect.

There is another option: we could enact a law against exploitation in hiring that tracks the ethics (as identified by Kates [2019], Sample [2003], or Snyder [2008]) and that includes both enforced and unenforced provisions. Such a law would specify a living wage that employers should pay if they can. ²¹ The law would include some clarification about what explanations of the inability to pay the living wage are acceptable. The alleged obligation to maximize value for shareholders will not qualify. ²² The law may specify a range of cases in which employers clearly can pay a living wage and in which paying a lower wage will therefore result in a penalty. Perhaps, for instance, recent payment of a dividend is clear evidence of the ability to pay a living wage. In a range of cases, an outside observer will not be able to tell whether a firm has a legally acceptable justification for paying less than the living wage. Only a manager or another insider to a firm can judge whether the firm is living up to its legal obligations. Under these circumstances, it may make sense for the wage law to include an explicitly unenforced requirement to pay a living wage if one can, even if none of the clear indicia of ability to pay are present. ²³

The unenforced component of this living wage law would fulfill three functions: 1) it would authoritatively clarify what counts as a living wage; 2) it would express society’s public condemnation of exploitative wages; and finally, 3) it would give managers a way to justify nonexploitative wage policy to investors who do not recognize an ethical side constraint against exploitation but who do recognize a duty to obey the law.

This defense of the potential value of antiexploitation laws with unenforced provisions rests on two features of the conditional duty to pay a living wage (on Kates’s [2019], Sample’s [2003], and Snyder’s [2008] accounts). First, the ethical duty to refrain from exploitation exists independent of law but is vague. The law-independent duty’s vagueness gives law a role in making the duty more determinate. Second, outside observers of an economic relationship, including government officials, are often unable to assess whether the relationship is wrongfully exploitative. This obstacle to reliable enforcement is the obstacle to imposing a fully enforced legal duty that tracks the ethical duty. These two features are by no means unique to the topic of wage exploitation. They can also arise in exploitation of other types.

Consider the ethics of pricing medically necessary drugs in places that lack a publicly supported health insurance system that makes medicine available to everyone who needs it. Arguably, pharmaceutical companies have an ethical duty to price medically necessary drugs so that people who need them can afford them, to the extent such affordable pricing is feasible (Hughes, 2020). This ethical duty is vague. There is room for reasonable disagreement about what constitutes affordable pricing. It is sometimes, though not always, difficult or impossible for regulators or the public to assess whether pharmaceutical companies are fulfilling the duty to make prices affordable if they can. Sometimes it is obvious that a large price increase was not a business necessity and could have been avoided. Given the complexities of the pharmaceutical industry, including the costs and financial risks of research and development, sometimes detailed knowledge of a company and its place in the industry is needed to assess whether the company could have priced its drugs more affordably.

What form of price regulation would be best for patients in the long run is a difficult question both morally and empirically. Overly demanding regulation may reduce investment in the industry and the rate of innovation. One possible approach would be to combine coercively enforced regulations (tailored to promote affordability without unduly deterring investment) with an unenforced requirement to adopt affordable pricing schemes to the extent this is possible. The unenforced requirement would serve three moral functions: first, it would clarify what counts as affordable pricing; second, it would express society’s commitment to norms of nonexploitation; third, it would give ethically upright managers a way of explaining their pricing decisions to investors, and it would signal to investors that they should not expect managers in this industry to pursue unconstrained profit maximization. Such an arrangement would be a clear improvement over nonregulation of pharmaceutical prices. Whether it would be the best policy depends on many empirical and normative questions, including the feasibility of moving to a publicly supported health insurance system that covers all or almost all the costs of medically necessary drugs. Under such a regime, pharmaceutical pricing could still be exploitative, but taxpayers, not individual patients, would be the exploited.

4. Objections and Replies

This section addresses several objections to the proposal that there should be explicitly unenforced laws against some forms of exploitation. To fix ideas, I will present these objections as objections to a specific policy proposal: to wit, that there should be an explicitly unenforced law identifying a living wage and requiring employers to pay workers at least this wage if they can. This unenforced living wage would be accompanied by a lower, enforced minimum wage, set at a level that would not cause excessive unemployment. It would also be accompanied by enforced requirements to pay a living wage when the employer meets specified criteria that clearly indicate an ability to pay (e.g., recent payment of a dividend or recent share buyback). The requirement to pay a living wage if one can would be unenforced when the listed indicators of ability to pay are absent. This proposal would aim to address a form of wage exploitation that Kates’s (2019), Sample’s (2003), and Snyder’s (2008) accounts of exploitation identify as wrong. Most objections to this proposal would also apply to attempts to address consensual exploitation of other sorts and to other accounts of the wrong of consensual exploitation.

The aim here is not to advocate a policy proposal; the aim is only to explain why including noncoercive provisions in antiexploitation laws should be considered a viable option in some economic conditions. The proposal to include unenforced provisions in a living wage law is appropriate only if a uniformly enforced law at least as demanding as the underlying ethical requirement would cause undue harm and should therefore be ruled out. ²⁴ Under these conditions, the live options for legislators are three: 1) not enacting an antiexploitation law at all, 2) enacting a coercively enforced antiexploitation law that is less demanding than the ethical duty of nonexploitation, or 3) enacting an antiexploitation law that tracks the ethical duty and includes unenforced provisions. I claim that the last of these three options is preferable.

4.1 Effects of Unenforced Laws

One reason against having explicitly unenforced laws applicable to private persons is the fear that such laws would have no effect. This objection is distinct from the ill-formed objection that unenforced laws would fail to achieve full compliance. The purposes of an unenforced antiexploitation law, or an unenforced provision of an antiexploitation law, are to give more precise content to the duty to refrain from exploitation, to express public commitment to this duty, and to give managers of profit-oriented firms a way to explain ethical decisions to superiors. These purposes can be achieved without full compliance. That said, one might think the law must have some effect on behavior to achieve its purposes. Is there reason to think that law can influence behavior when there is no coercive incentive to comply (including the incentive of informal social pressure) and when there is money at stake?

Though there are relatively few instances of explicitly unenforced laws concerning private citizens’ behavior, indirect evidence suggests that such laws can affect behavior. People can act against their apparent self-interest and in accordance with morality when money is at stake. For example, many people return lost wallets, and a recent study showed that in many countries, people are more likely to return lost wallets if those wallets contain money (Cohn, Maréchal, Tannenbaum, & Zünd, 2019). Tyler’s (2006) Chicago study of compliance with several commonly broken laws provides evidence that people who obey these laws often do so for moral reasons—including recognition of the law’s legitimacy—rather than for fear of sanctions. ²⁵ Evidence from experimental psychology suggests that when a law is democratically enacted, many people are inclined to comply with it, even when the sanctions attached are too weak to motivate compliance by themselves (Tyran & Feld, 2006). Further evidence comes from research on compliance with COVID-19 stay-at-home orders. Cell phone location data indicate that stay-at-home orders in the United States in spring 2020 affected people’s behavior significantly, though the orders were irregularly enforced and largely dependent on voluntary cooperation (Alexander & Karger, 2021).

Perhaps the most striking evidence is from research on tax compliance. Though the theoretically available penalties for tax evasion are severe, including imprisonment, they are rarely applied. The consensus among researchers studying tax compliance is that the deterrent effect of these sanctions could not fully explain the rates of tax compliance we observe (Lederman, 2003: 1457). Moreover, fear of informal social sanctions cannot explain most people’s tax compliance, as most people’s tax returns are not publicly disclosed. Because fear of legal and social sanctions cannot fully explain observed rates of tax compliance, many people must be motivated by respect for law, apart from fear of sanctions.

There is a distinctive reason to be skeptical about the effects of unenforced antiexploitation laws as applied to for-profit businesses. Could profit-maximizing firms be motivated to comply with an antiexploitation law when they have no financial incentive to do so, either from fear of an enforcement action or from fear of reputational harm? Won’t competitive pressure, whether from other firms or from rival managers within the firm, strongly tempt managers to violate unenforced laws with which compliance is costly? It is important to be clear about the sense in which firms are “profit maximizing.” As a normative matter, firms are neither morally nor legally required to maximize profits by breaking the law. Again, even Milton Friedman’s business-friendly shareholder theory holds that “law and ethical custom” constrain the pursuit of profit. The requirement is not merely to follow the law when following the law is profitable; the requirement is to follow the law. As a descriptive matter, the aims of actual firms and actual managers vary. As Ayres and Braithwaite (1992) explain, based on empirical investigation, actual firms broadly pursue three ethical approaches: some firms aim to maximize profits unconditionally, without regarding law as a constraint; some firms adhere to the law strictly and aim to maximize profits within that constraint; and some firms have ethical aims beyond compliance with the law and show willingness to refrain from maximizing profits when doing so advances these ethical aims. To assert that all firms are in the first of these three categories would be factually false. The firms that are most likely to change their behavior in response to an unenforced law are those that are law-abiding, even when obeying the law is not profit maximizing, but not otherwise ethically motivated.

Indirect evidence suggests that unenforced wage laws could, in fact, have some effect on businesses’ behavior—though admittedly, there would be far less than full compliance. This evidence comes from the United Kingdom’s experience with a campaign for a voluntary living wage. Since 2011, the Living Wage Foundation (LWF) has offered accreditation to employers that pay all employees a living wage, the rates for which LWF determines. ²⁶ As of the end of 2016, LWF estimated that approximately 120,000 workers had received higher wages because of their employers’ enrollment in the program (Heery, Hann, & Nash, 2017: 810). By one estimate, in 2014, six million workers in the UK were making less than the LWF living wage. So the voluntary living wage campaign only benefited a small fraction of low-wage workers, but a large absolute number of workers benefited. The UK experience with a voluntary minimum wage suggests a lower bound for the effect an unenforced but well-publicized living wage law could achieve. A legal requirement enacted by a democratically elected legislature and presented as a requirement would express a stronger demand than a recommendation (whether advanced by a nongovernmental organization or by the government). Unlike a recommendation, a law would authoritatively clarify what counts as satisfying the duty to pay a living wage if one can. A law thus might have a greater influence. It would also satisfy the moral demand that the law express condemnation of wrongful exploitation. A government that enacted a noncoercive living wage law in conditional form (“You must pay a living wage if you can”) might certify that businesses comply by paying the living wage, giving those firms a reputational boost. ²⁷

The percentage of workers who would receive a raise because of a nonenforced (but still mandatory) living wage law may well be in the single digits. The percentage would be low both because many employers would not comply and because many employers, being genuinely unable to pay a living wage, would not be legally required to raise wages. Even if a law would materially benefit only a small percentage of low-wage workers, it would still benefit many people, because there are very many low-wage workers. If the alternatives would be to have no antiexploitation law at all or a coercively enforced law that harms the people it is supposed to help, why not enact a noncoercive law that would help many people? If the antiexploitation law that would benefit workers the most would have a mix of enforced and unenforced provisions, why not enact that law?

4.2 Threats to the Law’s Legitimacy

One might worry that leaving law explicitly unenforced could have a different effect. If an antiexploitation law of the form described herein is drafted well, the public will rarely be able to identify specific cases in which firms are breaking the law. The law’s requirements would be explicitly unenforced when, and only when, it is difficult for third parties to assess compliance. Nevertheless, people may deduce from the lack of enforcement and the presence of the profit motive that violations of the unenforced provisions are probably common. If people perceive that a law is being widely flouted with impunity, might that undermine their faith in the legal system as a whole? Might it reduce their willingness to obey laws of other kinds?

The objection rests on empirical claims that are speculative and difficult to assess. The only straightforward way to test these claims is to enact antiexploitation laws with noncoercive provisions and see what happens. Antecedently, there is reason to be skeptical that wage laws with noncoercive provisions would have this corrosive effect. Every society has laws that are widely flouted with impunity, and this does not normally undermine law’s legitimacy. A pattern of nonenforcement of laws against serious forms of violent crime can and arguably should undermine respect for the legal system (Laufer & Hughes, 2020). It sends the message that the basic rights of the victims do not matter to the authorities. ²⁸ By contrast, if police regularly decline to pursue cases of bicycle theft, that pattern does not tend to undermine respect for the legal system. It reflects two obvious facts: first, that pursuing bicycle theft is inherently difficult, and second, that police departments have limited resources and caseloads that include unsolved violent crimes. Personal property rights matter, but the right to bodily integrity matters more. The police appropriately give priority to crimes other than bicycle theft. Doing so is unlikely to undermine the perceived legitimacy of the legal system or the willingness of generally upright people to refrain from bike theft. I suggest, as a hypothesis, that nonenforcement of law is less likely to undermine respect for the legal system when the government’s reasons for nonenforcement are consistent with taking both the law in question and the rights of victims seriously. The suggestion is speculative, but so is the objection.

A legislature’s reasons for enacting a wage law with explicitly unenforced provisions would differ from a police department’s reasons not to pursue cases of theft. Scarce law enforcement resources are not the main issue. Wage laws can be enforced via a private cause of action, rather than by agents of the executive branch. The reason to leave some provisions of a wage law unenforced is the fear that enforcement, whether by lawsuits or by public agencies, would harm workers collectively. The cause of the harm, again, is the interaction between contingent economic conditions (in which an exceptionless, enforced requirement to pay a living wage would put many people out of work) and the inability of both regulators and workers to identify reliably when a violation of the moral duty of nonexploitation has occurred. A government shows respect for workers and their interests if it enacts a law that sends the message “we think underpaying workers is seriously wrong when employers could in fact pay more, but we aren’t going to enforce that rule for certain types of business, because we think that enforcement would backfire and harm workers.” Antecedently, there is no reason to expect that enacting such a law would undermine the legal system’s overall legitimacy.

4.3 Can Enforcement Be Improved?

One might think that it is, in fact, possible to institute a coercively enforced living wage law that closely tracks the underlying ethical requirement. Perhaps the information asymmetry between businesspeople and regulators can be remedied. We could compel businesspeople to disclose to regulators the information necessary to judge whether they are engaging in exploitation, according to whatever criteria for exploitation we have set. In the context of a fair wage law, for instance, we could require employers who do not pay a living wage to disclose the facts about the business that demonstrate the impossibility of paying a living wage. Why not compel these disclosures and give antiexploitation law teeth?

This arrangement would impose costs of two sorts on firms that are not, in fact, exploiting their workers. First, the process of providing the disclosures would itself be costly. Second, firms risk being mistakenly found liable. If a large firm employs many low-wage workers, and its managers believe in good faith that the firm cannot pay these workers higher wages, it may be willing to pay the costs of demonstrating to the government that it cannot pay more. For smaller firms that are genuinely unable to pay a living wage, the costs of disclosure and the risks of liability would likely deter hiring at less than a living wage. ²⁹ These firms would respond to a requirement to demonstrate their inability to pay a living wage in the same way they would respond to an unconditional, enforced requirement to pay a living wage: they will stop hiring low-wage workers. If their business models require low-wage labor, they will shut down.

One might try to mitigate the burden of explaining decisions to regulators by allowing firms to appeal to industry custom as a defense. Firms would be required to prove their need to pay less than a living wage, as identified by the legislature, only if they depart from industry practice. There are two problems with this alternative. First, the reason for imposing a living wage law may be in part to change behavior in some industries. Second, determining whether a firm is engaged in exploitation often requires information about that firm’s distinctive practices and its niche within the industry. It may be just for some firms in an industry to pay workers less than average because these lower-paying firms face higher risks or cater to poorer customers. Demanding disclosure from small firms that depart from industry practices on wages and prices might shut these firms down—and harm the workers they would have employed. It is a difficult question, both morally and empirically, when a uniformly enforced living wage requirement would do more good for workers by raising wages than it would harm workers by reducing employment. Under market conditions in which a uniformly enforced living wage requirement would be bad for workers, a requirement that employers paying a sub–living wage defend that decision to regulators would harm workers for the same reason.

A third possibility would be to shift the burden of proving unfairness onto workers. A government could create a cause of action allowing workers to sue for underpayment. To prevail, workers would have to show that their employer chose not to pay a living wage despite being reasonably able to do so. The approach would be analogous to the tort law of negligence, which requires injured plaintiffs to show that the defendant failed to take reasonable precautions. This approach faces a dilemma. Would underpaid workers have a substantial chance of suing under this law and either winning in court or obtaining a favorable settlement? Whether the answer is yes or no, the law has an untoward consequence.

If it is likely that underpaid workers would regularly sue successfully, there would also be a substantial chance of successful suits that should not have succeeded. Again, judges and juries are fallible. For many employers who are genuinely unable to pay a living wage, choosing to hire at less than a living wage would be taking a legal risk. Employees might sue, and a court might mistakenly disagree with the employer’s judgment that paying a living wage is not possible given the business’s circumstances. Some such employers will choose not to hire at all, rather than hiring and risking a costly lawsuit. Put another way, there would be a chilling effect on nonexploitative low-wage work, that is, low-wage work with employers who genuinely cannot pay more. In some areas of law, chilling effects are unproblematic. It is not so bad if fear of a negligence lawsuit motivates some businesses to take extra precautions that are not necessary or deters entrepreneurs from creating businesses that may appear unduly dangerous. Chilling effects are problematic when it is important for people to be able to engage in conduct that comes close to the legal line. In the United States, where freedom of speech is constitutionally protected, courts often strike down laws because of potential chilling effects on speech. ³⁰ In economic conditions in which a uniformly enforced living wage law would put people out of work and shut down businesses, it is important for employers to be able to set wages that are close to the legal line. Chilling effects must thus be taken seriously.

A different problem arises if there is no real chance that underpaid workers would sue and win. This is the more likely horn of the dilemma. Low-wage workers rarely sue under clear, mechanically applicable labor laws. We cannot expect low-wage workers to use lawsuits to enforce labor laws that are (with good reason) vaguely written. If workers will not use the cause of action the law makes available, then the enforcement provision would serve no purpose. It would arguably be an attempt at deception, aiming to give the false impression that the law is effectively enforced. Whether or not deception is intended, workers are likely to perceive that the government has set up an enforcement mechanism that they are unable to use effectively. Workers may infer that the government is not serious about protecting their interests in fair wages. It is more respectful of workers, and it shows more seriousness about the law, for the government to say openly that it is making certain provisions of wage law unenforceable out of concern that enforcement would have perverse effects on workers collectively.

4.4 Unfairness to Efficient, Law-Abiding Firms

One might object that unenforced laws are unfair to firms that are more efficient than their competitors and that try to comply with the law in good faith. Consider a law requiring firms to pay an above-market living wage if they can. More efficient firms that benefit more from each employee’s labor are more likely to be able to pay an above-market wage. If they are law-abiding, they will face a burden that their less efficient competitors are not legally required to accept. This will reduce the competitive advantage their efficiency gives them. They will also face a disadvantage in competition with firms that choose to flout the fair wage law and can do so with impunity. One might think it unfair for a labor law to impose these relative disadvantages on efficient, law-abiding firms. ³¹

This objection is misplaced. A law requiring firms to pay a living wage if they can does not create a moral duty out of whole cloth. The ethical requirement to pay a living wage if one can exists whether or not there is a corresponding legal requirement. On this point many accounts of the ethical duty of nonexploitation agree. The moral role of an unenforced living wage law or a living wage law with unenforced provisions would be to clarify authoritatively what counts as satisfying the underlying ethical duty. One could argue that the alleged underlying moral duty is unfair. To be sure, this norm sometimes requires firms to raise wages at the cost of reducing profits. But a conditional requirement to pay a living wage if this is possible will never require a firm to institute a wage increase that seriously jeopardizes its viability. A firm cannot pay an above-market wage if doing so would seriously jeopardize the firm’s viability.

So the question is whether it is unfair to demand that efficient, law-abiding firms accept competitive disadvantages that reduce their profits but do not jeopardize their viability. The answer is that it is unfair to a degree, but that this unfairness must be weighed against the problems with the alternatives. In the economic contexts for which unenforced antiexploitation laws are intended, rigidly enforcing a wage floor or some more complex rule concerning wages would stop some mutually beneficial transactions without replacing them with fairer transactions. This would harm both workers and employers. Legally condoning wrongful exploitation would be unfair to workers. Moreover, the unfairness to these workers would involve a threat to their long-term ability to sustain themselves. These workers would be paid wages that do not enable them to meet their basic needs reliably, though their employers could pay them enough to meet their needs. There is a trade-off here between two forms of unfairness. Enacting the unenforced fair wage law subjects law-abiding firms to a form of unfairness that reduces their profits but does not jeopardize anyone’s life, health, or livelihood. Declining to enact a fair wage law subjects workers to a form of unfairness that may put their health or even their lives at stake in the long term. The latter form of unfairness is more serious.

Though concerns about unfairness to firms do not speak against the institution of unenforced antiexploitation laws, fairness concerns do point to a need for caution in designing the enforced provisions. A living wage law might include an enforced requirement to pay a living wage if certain clear indicia of ability to pay are present. Recent payment of a dividend clearly indicates ability to raise wages. If a law took dividend payment as the only indicator of ability to pay a living wage (or at least to raise sub–living wages), it would unfairly disfavor dividend-paying corporations in relation to firms that return profits to investors by other mechanisms. A well-tailored antiexploitation law would ensure that the coercively enforced provisions apply equally to firms of different types (e.g., sole proprietorships, partnerships, and corporations). Some level of regulatory arbitrage is inevitable. ³² The harms of regulatory arbitrage must be weighed against the harms that may result from making the coercively enforced provisions of law simpler and more demanding. In some economic conditions, a more demanding coercive standard (e.g., an unconditional wage floor that is a living wage) would be worse for workers.

5. Conclusion

Once we recognize that noncoercive laws against mutually beneficial, consensual exploitation are viable options, it appears that there are two possibilities governments should consider. First, perhaps governments should make the legal demands of antiexploitation law more extensive, while leaving some or all of these new demands unenforced. Second, if a legislature judges correctly that existing laws against exploitation are causing undue harm by preventing too many beneficial transactions, the legislature should hesitate to repeal the laws. Instead, the legislature should consider removing or limiting the enforcement provisions of these laws, leaving the underlying prohibitions intact.

Which of these possibilities governments should pursue, if either, depends both on empirical questions about the effects revising the law would have and on normative questions about how different harms and forms of unfairness should be weighed. Governments should not simply assume that the only useful laws are laws with sharp teeth. Laws that explicitly lack coercive enforcement can give precise content to the moral duty to refrain from wrongful exploitation. If there is a space in which there is no fact of the matter (absent authoritative law or custom) whether certain pricing practices are fair, the unenforced law would give more precise content to the moral duty not to take wrongful advantage of bargaining power. Noncoercive laws can also express society’s condemnation of wrongful exploitation without causing undue harm by misapplying sanctions. Finally, noncoercive laws can give people (such as business managers) a way of explaining their actions to others who disagree about what social norms there should be but who believe that there are moral reasons to obey the law.