Are markets coercive?

Abstract

Are markets coercive? Contemporary debate is dominated by two answers. The first, longstanding among defenders of free markets, holds that voluntary exchange is non-coercive by definition: coercion enters the picture only when rights are violated. The second, revived from Robert Hale's 1923 essay and embraced today by progressive legal scholars and post-liberal conservatives alike, holds that markets are pervasively coercive because property rights backed by state power constitute a system of mutual coercion. Both answers fail, but the Halean answer fails in the more interesting way. The libertarian answer moralizes coercion, stripping it of evaluative force; the Halean answer de-moralizes coercion, stripping it of discriminating power. If coercion is everywhere, it cannot pick out the arrangements worth criticizing. Drawing on a convergence across the philosophical literature on coercion, the paper develops a third option: identify coercion non-moralistically—as a factual matter of options restricted by another's institutionally backed power—but evaluate its significance in graduated terms, by attention to the significance of the options curtailed, the availability of alternatives, and the directness of compulsion involved. The result is a framework that can distinguish cases of genuinely worrisome market coercion from cases where the concept does no moral work.

Introduction

Are markets coercive? ¹ For much of the 20th century, the dominant answer among defenders of free markets was a confident no. Market transactions are voluntary exchanges between willing parties; coercion enters the picture only when someone's rights are violated. Government regulation is coercive; markets are not.

Challenges to this picture date back at least to 1923, when the Columbia law professor Robert Hale published a short but influential essay in the Political Science Quarterly entitled “Coercion and Distribution in a Supposedly Non-Coercive State.” His thesis was direct: the economic system defended by proponents of laissez-faire was not, as they supposed, a system of voluntary exchange free from coercion. On the contrary, Hale argued, “the systems advocated by professed upholders of laissez-faire are in reality permeated with coercive restrictions of individual freedom” (Hale, 1923: 470).

Hale's argument has enjoyed a remarkable revival in recent years, and what is striking about this revival is the breadth of the political coalition it has assembled. On the progressive left, the Law and Political Economy (LPE) movement has placed Hale at the center of its intellectual genealogy. Britton-Purdy et al. (2020: 1819, 1796) describe economic life as “a system of mutual coercion, with the degree of each person's coercive power arising directly from legal entitlements,” and insist that “legally constituted and distributed coercion is the sine qua non of market relations”. Meanwhile, on the post-liberal right, Ahmari (2023: 23, 14) draws on the very same 1923 essay to argue that market relations amount to “a naked instance of coercion” and that “there is a case for calling it tyranny”. Vermeule (2022) invokes the same structural insight, citing Hale in support of the contention that “power is always conserved”—that government regulation cannot be criticized as an imposition on an otherwise free market, because market relations are themselves structured by state-backed power.

These contemporary deployments of Hale tend to involve two subtly different argumentative moves. The first uses Hale to criticize markets: markets are coercive, coercion is bad, therefore markets—or at least certain market arrangements—are objectionable. ² The second uses Hale to neutralize a criticism of regulation: markets are coercive too, so coercion cannot be a reason to prefer markets over government intervention—it is a wash. Ahmari leans toward the first move, affirmatively characterizing private economic power as tyrannical. The LPE movement leans toward the second, using Hale analytically to dissolve the public/private, coercive/voluntary dichotomy and clear the ground for normative arguments framed in terms of equality, democracy, and power. Vermeule's “power is always conserved” is a concise statement of the second move. Both moves have real argumentative force. But they sit uneasily together. If coercion is everywhere, it cannot serve as a special criticism of any particular arrangement. If it is a special criticism, it cannot be everywhere. Contemporary Haleans tend to want both—coercion as a live criticism of markets and as a neutralized criterion for evaluating regulation—and this combination is unstable unless one has a principled way of distinguishing significant from insignificant coercion.

This paper argues that such a principled distinction is available, but that it requires abandoning Hale's framework. The argument proceeds as follows. “The libertarian answer” section briefly considers the libertarian answer to the question of market coercion—the view, common among defenders of free markets, that voluntary exchange is by definition non-coercive. The problem with this argument, as many others have noted, is that it relies on a moralized definition of coercion that builds the conclusion into the premises. The “Hale's challenge” section reconstructs Hale's challenge to this position. “The bodily integrity test” section presents what I call the “bodily integrity test": the observation that Hale's logic, applied to rights over one's own body, entails the implausible conclusion that ordinary exercises of bodily autonomy constitute coercion. The “Baseline nihilism” section argues that this problem is a symptom of a deeper structural flaw—what I call “baseline nihilism”—that renders Hale's concept of coercion evaluatively inert. The “Identifying and evaluating coercion” section develops the paper's positive contribution: a non-moralized account of coercion on which instances of coercion are identified without reference to moral rights but evaluated according to their graduated moral significance. The “Evaluating markets” section applies this framework to contemporary debates about markets. The “Conclusion” section concludes the article.

Before proceeding, I want to be clear about what I am and am not claiming. I am not arguing that property rights never raise moral concerns, or that workers in difficult circumstances have no legitimate grievances. I have argued elsewhere, at length, that they do (Powell and Zwolinski, 2012; Zwolinski, 2007, 2015). What I am arguing is that the contemporary debate between those who deny that markets are coercive and those who affirm it in Hale's sense is a debate between two inadequate positions, and that a better account of coercion—one that is non-moralized in definition but discriminating in evaluation—can do the work that both sides need but neither can deliver.

The libertarian answer

One influential answer to the question “Are markets coercive?” is a flat no. On this view, market transactions are paradigmatically voluntary. Coercion enters the picture only when someone's rights are violated—when force or fraud is employed, or when someone is compelled to act against her will by threats of impermissible harm. Since ordinary market transactions involve neither force nor fraud, they are non-coercive by definition.

This position is widespread among defenders of free markets. ³ Rothbard (1973: ch. 2) restricted the concept of coercion to invasions of person or property—the “initiation of physical force” against others. Nozick's (1974: 150–153) entitlement theory treats distributions arising from voluntary transfers as just, where voluntariness is understood in terms of the absence of rights violations. And the basic intuition—that a willing buyer and a willing seller engaged in unforced exchange are not coercing each other—has deep roots in ordinary moral thought.

The problems with this position are by now well understood, and since they are extensively discussed elsewhere, I will be brief. ⁴ The core difficulty is that the moralized definition of coercion builds the conclusion into the premises. Whether a market exchange counts as “coercive” depends, on this view, on whether the parties acted within their rights—and specifically on whether the background distribution of property rights is legitimate. But this means that the concept of coercion cannot serve as an independent input to the evaluation of property arrangements. It presupposes the very moral judgments it is supposed to help ground. If we want to ask whether a particular property regime is just, we cannot appeal to the fact that exchanges under it are “voluntary” and “non-coercive,” because those characterizations already assume the justice of the regime. ⁵

This is a serious problem, but it is not my primary target. The libertarian view has been ably criticized by others. My concern is with what has been offered as its replacement.

Hale's challenge

Hale's (1923) essay is, in part, a response to the libertarian picture—or rather, to the version of it that prevailed in his own time, articulated in the “individualist” economist Thomas Nixon Carver's Principles of National Economy. ⁶ Carver, like many defenders of free markets, assumed that an economy organized around private property and voluntary exchange was characterized by freedom and the absence of coercion. Government intervention, by contrast, introduced coercion into what would otherwise be a realm of peaceful, consensual interaction. Hale set out to challenge this asymmetry.

The challenge begins with the observation that private property is, at bottom, a right to exclude. The owner of a factory can prohibit others from entering or using the factory without permission. More importantly, the owner can make access conditional—offering the use of the factory's resources on terms that the owner specifies. As Hale (1923: 472) puts it, “in protecting property the government is doing something quite apart from merely keeping the peace. It is exerting coercion wherever that is necessary to protect each owner, not merely from violence, but also from peaceful infringement of his sole right to enjoy the thing owned”.

Hale emphasizes that this exclusionary power is not a fact of nature. It is a product of law, and ultimately of state enforcement. The factory owner's ability to deny others access to the factory depends on the willingness of the legal system to back up the owner's claim with force. If a homeless person attempts to sleep in an empty factory, it is the police who will remove them. Property rights, in short, are enforced by the coercive apparatus of the state.

From these observations, Hale (1923: 474) draws a striking conclusion. The distinction between “free exchange” and “government coercion” is illusory. A system of private property is itself a system of government-backed coercion. And crucially, Hale does not see this coercion as flowing in only one direction. The worker, too, has a form of coercive power over the factory owner—the power to withhold labor. “What else is ‘coercion'?” Hale asks, rhetorically.

Hale (1923: 471) is explicit that he intends “coercion” as a descriptive rather than normative concept. “To call an act coercive is not by any means to condemn it,” he writes. “It is because the word ‘coercion’ frequently seems to carry with it the stigma of impropriety, that the coercive character of many innocent acts is so frequently denied”. And again: “it seems better, in using the word ‘coercion’, to use it in a sense which involves no moral judgment” (Hale, 1923: 476).

This descriptive orientation leads to what I will call Hale's symmetry thesis: the thesis that any distribution of property rights involves coercion, and that shifting rights from one set of persons to another merely redistributes coercive power without increasing or decreasing it. ⁷ Hale is quite explicit:

To take this control by law from the owner of the plant and to vest it in public officials … would neither add to nor subtract from the constraint which is exercised with the aid of the government. It would merely transfer the constraining power to a different set of persons. (Hale, 1923: 478)

And he draws the conclusion without flinching: “Neither [arrangement] can be said to be any ‘freer’ than the other in the sense that it involves less coercion” (Hale, 1923: 478).

The argument is elegant, and it is easy to see why it has been so appealing to critics of free-market liberalism. It appears to turn the libertarian's own favored concept against the libertarian's own favored institutions. But as I will argue in the next two sections, the argument's elegance conceals a structural defect.

The bodily integrity test

Hale's argument has a clear logical structure. It begins with an observation about the nature of property—that property is a right to exclude backed by state power—and draws a conclusion about its coercive character. The structure can be stated as a simple argument:

(1) Property is a right to exclude others from the use of a thing. (2) The right to exclude is backed by the coercive power of the state. (3) The exercise of this right restricts others’ liberty—their range of options is narrowed by the owner's power to exclude. (4) Therefore, property is coercive.

The argument is perfectly general. Nothing in it limits its scope to any particular kind of exclusionary right. It applies wherever a legally enforced right to exclude others exists. And this is precisely where the trouble begins.

Consider rights of bodily integrity—rights against assault, against being forced to labor, against sexual violation. These are among the most widely accepted and least controversial rights in all of moral and political philosophy. One need not be a libertarian, or subscribe to any particular theory of self-ownership, to affirm them. ⁸

And yet these rights have exactly the same logical structure that Hale identifies as coercive in the case of property. My right to bodily integrity is a right to exclude others from the use of my body. It is backed by the coercive power of the state—by laws against assault, rape, forced labor, and involuntary organ harvesting. And its exercise restricts others’ liberty: because I have a right to refuse, others cannot use my body as they might wish.

If Hale's argument is sound, then bodily integrity rights are coercive in exactly the same sense as property rights—and at both levels at which Hale locates coercion in the case of property. ⁹ The state coerces would-be assaulters, kidney thieves, and rapists through the threat of criminal sanction. And the rightholder, by withholding access to her body or her labor, coerces the would-be user: a person who refuses to work unless paid a wage is coercing the would-be employer, a person who refuses to have sex unless certain conditions are met is coercing the would-be sexual partner, and a person who declines to donate a kidney to someone who will die without it is coercing the person in need.

These conclusions are difficult to accept. The right to refuse sex is not coercion in any morally meaningful sense. The right to refuse labor is not coercion in any morally meaningful sense. The right to keep one's organs inside one's body is not coercion in any morally meaningful sense. ¹⁰ Nor is the state's enforcement of these rights: no one takes the criminal prohibition of rape, forced labor, or involuntary organ harvesting as a reason for moral concern. If Hale's framework nonetheless classifies all these cases as coercion on a par with the most egregious workings of private power, it is not our intuitions about bodily autonomy that have gone wrong. It is the framework—which has lumped together what careful moral thought must keep apart, and which therefore cannot do the critical work its contemporary admirers need it to do.

A Halean may grant all of this. Yes (she will say), the concept is descriptive; the awkwardness of the verdict reflects the reader's assumption that coercion must be morally bad; the descriptive identification remains illuminating. But to grant this is to grant exactly what I've been arguing: coercion in Hale's sense is morally inert, and so cannot ground a critique of any particular institutional arrangement. I take up this point more fully in the next section. Here I want to address one more textually grounded objection.

One might attempt to resist the extension to bodily integrity by pointing to Hale's own text. When discussing the worker's power to withhold labor, Hale (1923: 474, emphasis Hale's) describes this as “their actual power (neither created nor destroyed by the law) to withhold their services”. This passage suggests that Hale recognizes a distinction between the worker's natural power over her own body—a power that exists independently of the legal system—and the legally created powers of property. A Halean defender might seize on this: the argument targets legally created exclusionary rights, not natural bodily powers.

But this response fails for three reasons. First, Hale himself calls the withholding of labor “coercion” in the very same passage—“What else is ‘coercion’?” he asks—so the extension follows even on Hale's own terms. Second, bodily integrity rights in modern legal systems are legally enforced, and if anything more aggressively than property rights. Assault, rape, and kidnapping carry harsher penalties than trespass or theft. The state's coercive apparatus stands behind bodily integrity no less than behind property. Third, and most fundamentally, the distinction between “natural” and “legally created” power is itself a normative distinction. To say that some powers are natural and therefore exempt from the coercion analysis is to introduce a moral baseline—a prior judgment about which exclusions are legitimate—which is precisely what Hale's descriptive framework is supposed to render unnecessary. ¹¹

This problem did not escape the attention of Hale's most sympathetic interpreter. Barbara Fried, in her comprehensive reconstruction of Hale's thought, acknowledges the bodily integrity difficulty—but only in the endnotes, where she writes that bodily integrity “carries with it a strong, universally shared, intuition as to certain basic legal rights that ought to go along with it” (Fried, 1998: endnote p. 254). That even Hale's most careful defender relegates this problem to the notes rather than confronting it in the main text is evidence that it is real and intractable within the Halean framework.

Baseline nihilism

The bodily integrity problem is the sharpest illustration of a deeper structural flaw in Hale's account. To see this flaw clearly, it will be helpful to draw on the distinction, developed most fully by Wertheimer (1987), between moral baselines and nonmoral baselines for evaluating coercion.

The basic idea is as follows. To determine whether a proposal is coercive, we need to know whether it constitutes a threat or an offer. A threat makes the recipient worse off if she does not comply; an offer makes her better off if she does. But “worse off” and “better off” are comparative—they require a baseline, a reference point. A moral baseline compares the recipient's situation to what it would be if people acted as they are morally required to act. A nonmoral baseline compares it to some other standard that does not depend on moral entitlements.

Hale's account, reconstructed in these terms, employs a nonmoral baseline. ¹² Property is coercive because the owner's exercise of exclusionary rights restricts non-owners’ options relative to what they would be under alternative institutional arrangements. It does not matter whether the owner has a moral right to exclude. What matters is only that the exclusion narrows options.

The problem is that this generalizes catastrophically. Every institutional arrangement restricts some people's options relative to some other possible arrangement. Under private property, non-owners are excluded from resources. Under collective ownership, individuals are excluded from the freedom to use and dispose of resources as they see fit. Under any system at all, some people will have fewer options than they would have under some alternative. As Frye (2022) has shown for collective property, shifting from private to communal ownership does not eliminate restriction of freedom—it merely reconfigures it.

The consequence is the symmetry thesis in its starkest form: coercion, on Hale's account, is entirely zero-sum. It can be redistributed among persons, but never increased or decreased in total. This means that the overall level of coercion in society is a constant—it is the same under capitalism and under communism, the same in Sweden and in North Korea. ¹³ I do not mean, of course, that Hale considers these arrangements morally equivalent. His point is narrower: coercion specifically cannot distinguish among them. No institutional arrangement is more or less coercive than any other. The concept of coercion, so understood, is completely useless as a tool for evaluating or comparing social systems. ¹⁴

This is not an implication that I am foisting on Hale against his will. Hale himself embraced it. And for Hale, a legal realist deeply suspicious of normative concepts, this was not a troubling conclusion but an illuminating one. Fried (1998: 57), in her comprehensive reconstruction, stresses that Hale's argument “said nothing about the desirability of the coercive force exerted by private property rights”—a point “worth stressing, since Hale was so frequently misunderstood on this point by his own contemporaries and later scholars”. Fried (1998: 70) is candid that Hale's analysis was “probably largely tactical: an effort to dismantle constitutional obstacles, erected in the name of freedom, to a more egalitarian distribution of wealth”. As a tactical move in early 20th-century constitutional law, the argument had real force. Hale's point—that the common-law baseline is itself a legal construct—is well-taken.

But contemporary theorists who invoke Hale want more than a ground-clearing exercise. They want coercion to do normative work. Ahmari (2023: 23) calls market relations “tyranny” and finds them “manifestly unjust”—while simultaneously acknowledging that “coercion is inevitable in human affairs” (Ahmari, 2023: 178). The LPE manifesto faults law-and-economics for “giv[ing] up the urgency of both criticizing coercion and inequality and asking how they might be justified, if at all” (Britton-Purdy et al., 2020: 1796). These are normative claims. They depend on coercion being a morally loaded concept—one that can ground objections to existing arrangements.

Here we can see the instability I identified in the introduction. The first move—coercion as a live criticism of markets—requires that coercion be normatively significant. The second move—coercion as a wash between markets and regulation, the symmetry thesis politically deployed—requires that coercion be ubiquitous. If coercion is ubiquitous, it cannot serve as a special criticism of capitalism. If it is a special criticism, it cannot be ubiquitous. The two moves pull in opposite directions, and Hale's framework offers no way to reconcile them.

It is telling, I think, that the most philosophically sophisticated thinkers working on these questions have quietly moved away from the language of coercion. Elizabeth Anderson's Private Government (2017) frames the problem of workplace authority in terms of domination—arbitrary, unaccountable power—rather than coercion. Pettit's (1997) republican framework focuses on non-domination. Gourevitch (2015) frames the labor question in terms of republican freedom. ¹⁵ These are better concepts for the work these scholars want to do, precisely because they pick out morally salient features—arbitrary power, subjection to another's will—that mere option-restriction does not. Whether or not these authors intend it as such, the shift from coercion to domination is consistent with the thesis that coercion in Hale's undiscriminating sense cannot do the normative work they need it to do. I will argue below, however, that we need not abandon the concept of coercion. The graduated framework developed in the next section preserves the insight driving these authors’ turn to domination—indeed, it makes domination a factor in the moral evaluation of coercion. The move is from Hale's flat coercion to graduated coercion, not from coercion to domination.

Let me step back from the details to consider where we now stand. We began with a question—are markets coercive?—and have encountered two answers that fail. The libertarian answer, examined above, moralizes coercion so that voluntary exchange is non-coercive by definition—the concept is drained of independent evaluative force. Hale's answer universalizes coercion so that every institutional arrangement is equally coercive—the concept is drained of discriminating power. What we need is a third option: an account that identifies coercion without moralizing it, but evaluates its significance without treating all instances as equivalent.

Identifying and evaluating coercion

The libertarian answer moralizes coercion and thereby makes it do no independent work: whether an arrangement is “coercive” simply restates whether rights have been violated. Hale's answer de-moralizes coercion and thereby makes it do no discriminating work: all arrangements are equally coercive and therefore none can be criticized on that basis.

The striking thing is that these two inadequate positions do not exhaust the field. The philosophical literature on coercion is rich and varied—it includes accounts built around conditional threats (Nozick, 1969), moralized baselines (Wertheimer, 1987), subjection to another's will (Hayek, 1960; Pettit, 1996), the absence of acceptable alternatives (Olsaretti, 1998), and the enforcement of asymmetric power relations (Anderson, 2010). These accounts differ in important ways, and I do not propose to adjudicate among them here. But they converge on two points that both Hale and the libertarian get wrong. First, whether an interaction counts as coercive can be determined without building in prior moral entitlements—it is a substantive question about the character of the interaction, not a label that merely tracks antecedent rights. ¹⁶ Second, instances of coercion vary in moral significance—some are deeply troubling, others are trivial, and a theory that cannot make this distinction is not a theory anyone should use to evaluate institutions. A mugger who demands “your money or your life” is engaged in paradigmatically serious coercion; a homeowner who forbids strangers from cutting across her lawn is, at most, engaged in coercion of a trivial sort.

A Halean might respond that this convergence is compatible with Hale's basic insight—that all these accounts agree property restricts options, and that a graduated account simply develops Hale rather than rejecting him. But this misreads what the convergence involves. The accounts converge not only on the claim that coercion admits of degrees but on the claim that what matters about coercion is something other than the bare restriction of options. Each asks a further question—about conditionality, about will, about alternatives, about domination—that Hale's framework treats as irrelevant. To adopt any of these accounts is to abandon Hale's central thesis: that coercion is a flat, undiscriminating category, constant across institutional arrangements. That thesis is precisely what every sophisticated account of coercion denies.

What I want to do in this section, then, is not to defend one particular account of coercion against its rivals. It is to draw out the implications of this convergence—to show that the best work on coercion, across diverse philosophical traditions, yields a framework that can do the evaluative work that both Hale and his critics need. That framework has two components: a distinction between identifying coercion and evaluating its significance, and a set of factors that determine where on the spectrum of moral significance any given instance falls.

Identification and significance

The key move is to distinguish between identifying an instance of coercion and evaluating its moral significance. Whether something counts as coercion is a factual, non-evaluative question. How much it matters morally is a separate, evaluative question.

This two-step structure has a well-established precedent in the philosophical literature on freedom. Carter (1999) defends a non-evaluative account of when a constraint on freedom exists (ch. 8). Kramer (2003: 4) puts the same point sharply: particular freedoms and unfreedoms “can be verified or disconfirmed without any recourse to moral-political considerations”. Kramer (2003: 374) also extends the analysis to evaluation: “the extent of anyone's overall freedom is a matter that cannot be fully ascertained in the absence of evaluative assumptions”. Not all freedoms are equally important; each, for Kramer (2003: 428–429), gets a “qualitatively-oriented multiplier” reflecting its contribution to individual development, well-being, and autonomy. That some liberties matter more than others is, of course, familiar from Rawls (1971: 201), who treats the liberty debate as one about “the relative values of the several liberties when they come into conflict” and gives lexical priority to a specific set of basic liberties (Rawls, 1971: §39).

I want to make exactly the same move with coercion. Whether an instance of coercion exists—whether someone's options are restricted by the exercise of another's power, backed by institutional enforcement—is a factual question. Property exclusion restricts options; so do bodily integrity rights. In this much, Hale was right. Carter (1999: ch. 8) himself endorses the point in the context of freedom: property laws constrain the freedom of non-owners, full stop, regardless of whether those laws are morally justified. The poor are less free than the rich in this non-moralized sense.

The concept of coercion, so understood, picks out a specific class of option-restrictions: those imposed by human agents through the exercise of institutionally backed power. Not all restrictions on what people can do qualify. A landslide that blocks a road restricts options but is not coercion; a zoning law that blocks a road is. The distinction matters because agent-imposed restrictions raise questions of justification—questions about the legitimacy of the power being exercised, the adequacy of the alternatives left open, and the significance of the options curtailed—that natural restrictions do not. The concept of coercion identifies the domain within which these evaluative questions arise; the factors I develop below determine how they should be answered.

But—and this is the crucial point—how much moral significance any instance of coercion carries is a further question, and one that admits of degrees. Being denied access to someone else's kidney is coercion in the same non-moralized sense as being denied access to a factory. But these are not morally equivalent. A theory of coercion that treats them as equivalent—as Hale's must—has gone wrong somewhere.

Deontic and axiological coercion

Garnett (2018) has helpfully distinguished between two concepts that often travel under the name “coercion.” The first is deontic—coercion as a type of moral wrong, a violation that triggers a presumption of illegitimacy and potentially vitiates consent. The second is axiological—coercion as a type of human bad, a restriction of freedom or autonomy that bears on the quality of someone's life and choice situation but is not itself a moral wrong.

This distinction illuminates the dialectic between Hale and his critics. Hale is right that property involves axiological coercion—it restricts options, narrows the range of what people can do, and in that sense makes their situation worse than it would be in a world of unrestricted access. This is a genuine observation, and defenders of property rights should not deny it. But it would be wrong to infer—as Hale's contemporary admirers do—the further conclusion that property involves deontic coercion, the kind that constitutes a moral wrong and grounds a presumption against the arrangement. Whether an instance of axiological coercion rises to the level of a deontic wrong depends on further moral considerations—the significance of the options curtailed, the availability of alternatives, and the character of the power being exercised.

This framework concedes substantial ground to Hale while preserving what the libertarian account gets right. With Hale, I affirm that property restricts freedom and that this restriction is a fact worth noting. With the libertarian, I affirm that not every restriction of freedom constitutes a moral wrong. The question is not whether markets involve coercion in the broad, axiological sense—they plainly do—but when that coercion is morally significant enough to ground a presumption against the arrangement. And that question, unlike the question Hale asks, admits of graduated answers. ¹⁷

Three factors of moral significance

What determines the moral significance of a given instance of coercion? I do not pretend to offer an exhaustive theory; the point is to identify factors that clearly matter on any adequate account and that suffice to demonstrate the inadequacy of Hale's undiscriminating approach. I propose three such factors, each of which admits of degrees.

First, the significance of the options curtailed. Not all options are equally important. Being denied the freedom to choose one's religion is more significant than being denied the freedom to trespass on a stranger's land. Being denied the opportunity to earn a living is more significant than being denied access to a luxury good. This is not a controversial claim; it reflects the widely shared judgment that some freedoms matter more than others. ¹⁸ Coercion that curtails options central to a person's basic well-being, autonomy, and development is more morally significant than coercion that curtails trivial or peripheral options.

Second, the availability of alternatives. The moral significance of being denied a particular option depends in part on what other options remain. A worker who faces a single employer offering subsistence wages in a company town is in a very different situation from a worker who can choose among dozens of employers offering competitive wages. In both cases, the worker's options are restricted—she cannot simply walk into any factory and use its resources. But the restriction is far more significant when alternatives are scarce. Olsaretti (1998: 75) captures part of this point with her distinction between primary choices (whether to play the game at all) and secondary choices (which moves to make within the game). A worker who must work or starve faces a constrained primary choice. A worker who can choose among many acceptable employers makes a voluntary secondary choice. But “the fact that the worker chooses to work for one employer rather than another does not establish that he is not forced to sell his labor in the first place”. Morally worrying coercion tends to arise where both primary and secondary choices are constrained—where not only must the person participate, but the available options within the game are themselves unacceptable.

Third, the directness of compulsion. There is a morally relevant difference between being physically forced to labor—having one's body commandeered for another's purposes—and being subject to the remote, impersonal effects of a legal regime that assigns property rights. Both restrict options. But forced labor involves what Pettit (1996: 578) calls “coercion of the body”—direct physical control over another's movements and actions—while property exclusion operates at a distance, through the legal system's willingness to enforce boundary claims.

This third factor connects to a deeper insight from the republican tradition. Forced labor is not merely a restriction of options; it is an instance of domination—subjection to the arbitrary will of another. On Pettit's (1996: 578–580) account, domination requires that someone have the capacity to interfere with another on an arbitrary basis, with impunity and at will, in choices the other is in a position to make. The forced laborer satisfies all three conditions: the master controls the worker's body arbitrarily, without effective check, in the most intimate domain of human agency. By contrast, the property owner who excludes a stranger from her land does not, in the ordinary case, dominate that stranger. The stranger is not subject to the owner's arbitrary will; the stranger typically has no relationship with the owner at all. The exclusion operates through the impersonal framework of property law, not through subjection to another person's discretionary power. ¹⁹

This distinction helps explain why the bodily integrity cases are so compelling as a reductio of Hale. Coercion that involves the body is, in the typical case, more morally significant than coercion that involves external property—not because of a controversial self-ownership thesis, but because the body is the site of personal agency, and restrictions on bodily freedom tend to involve precisely the kind of domination (subjection to arbitrary will, loss of control over one's most intimate sphere) that makes coercion morally troubling. ²⁰ Hale's framework cannot capture this difference. The graduated significance approach can.

The framework applied

The three factors—significance of options, availability of alternatives, and directness of compulsion—work together. An instance of coercion that scores high on all three is a paradigm case of morally significant coercion: it curtails important options, leaves few or no alternatives, and involves direct domination. An instance that scores low on all three is trivially coercive—technically a restriction of options, but not one that should trouble us morally.

Most instances of market coercion fall somewhere in between. And this is exactly as it should be. The question “Are markets coercive?” should not have a simple yes-or-no answer. Markets involve coercion—restriction of options backed by institutional power—in the broad, non-moralized sense. But the moral significance of that coercion varies enormously depending on the circumstances: what options are curtailed, what alternatives exist, and whether the restriction involves domination or merely the impersonal operation of a legal framework.

This is what Hale's framework cannot do and what his contemporary admirers need it to do. It is the principled distinction between significant and insignificant coercion that allows the concept to do genuine evaluative work—to ground criticisms of some market arrangements without committing us to the implausible view that all market arrangements are equally objectionable.

Evaluating markets

What does the graduated significance framework yield when applied to specific market arrangements? I want to sketch three cases that illustrate how the framework might be applied. The treatment is necessarily schematic; my aim is not to resolve the complex empirical questions these cases raise but to show that the graduated framework asks the right questions—questions that Hale's undiscriminating concept of coercion cannot even formulate.

Case 1: Work-or-starve in a company town. A worker faces a single employer offering subsistence wages and hazardous conditions. The worker must accept or face destitution. All three factors point toward high moral significance: the options curtailed are central to survival and well-being (first factor); there are no meaningful alternatives (second factor); and the worker is subject to the employer's discretionary power—the employer can fire at will, impose arbitrary conditions, and effectively govern the worker's daily life (third factor). This is the kind of case that Hale's contemporary admirers most often have in mind, and the graduated framework agrees that it is genuinely troubling. It is genuinely troubling not because all property is coercive, but because the specific conditions—monopoly power, subsistence stakes, employer domination—make this particular instance of coercion morally significant. It is worth noting that even Friedrich Hayek—among the 20th century's most prominent defenders of free markets—recognizes that conditions of this kind can be genuinely coercive. In his discussion of employment in The Constitution of Liberty, Hayek acknowledges that where a worker faces a single employer and no realistic alternatives, the employer's power may indeed constitute coercion (Hayek, 1960: 203–204).

Case 2: Competitive labor markets with a safety net. A worker can choose among many employers, each offering above-subsistence wages, in a society with unemployment insurance and a basic social safety net. The worker's primary choice is still constrained in Olsaretti's sense—one must work for a living—but the secondary choice is genuinely voluntary, and the primary constraint is substantially mitigated by the safety net. The first factor is moderate (the options curtailed are significant but not survival-level). The second factor is low (many alternatives exist). The third factor is low (no single employer dominates). On the graduated framework, this arrangement involves coercion in the non-moralized sense—property exclusion still restricts options—but the moral significance is substantially reduced. As Pettit (2006: 139–42) argues, competitive markets serve precisely this function: where workers can exit one employment relationship and enter another, no single employer holds the kind of arbitrary power over the worker that would constitute domination, and the coercion involved ceases to be morally troubling. ²¹

Case 3: Mandatory arbitration clauses. An employer requires employees to sign an arbitration agreement as a condition of employment. On Hale's framework, this is simply another instance of coercion, indistinguishable in kind from any other exercise of property-backed power. The graduated framework asks more precise questions. How significant is the option curtailed? Access to jury trial is significant, though opinions differ on how much. How available are alternatives? If most employers impose arbitration, few alternatives exist; if only some do, the constraint is lighter. Does the arrangement involve domination? It depends on the broader labor market conditions and the employee's bargaining position. Ahmari (2023: 46–50, 79–80) illustrates the concern vividly: workers who must accept arbitration clauses as a condition of employment may find the dispute resolution process prohibitively expensive and controlled by the employer, leaving them with no realistic avenue for vindicating their rights. The answers are not predetermined by the theory; they depend on the empirical circumstances. But the framework identifies the right questions—the ones that actually track the moral weight of the restriction—rather than collapsing all cases into the undifferentiated category of “coercion.” And the empirical questions are not idle: the prevalence of mandatory arbitration across an industry directly affects the availability of alternatives, and therefore the moral significance of the coercion involved.

These cases illustrate a point that deserves emphasis. The concerns motivating Hale's contemporary admirers are not without merit. Workers do sometimes face objectionable conditions. Economic arrangements do sometimes reflect and entrench unjust power relations. Ahmari's worries about employer domination and the LPE movement's concerns about inequality and democratic accountability point to real problems. But those problems are better articulated through the graduated framework—which asks about the specific conditions that make particular arrangements morally troubling—than through the blunt instrument of Halean coercion talk, which cannot distinguish a company town from a competitive labor market.

The graduated framework also illuminates the trade-offs involved in institutional design. On Hale's zero-sum account, the total amount of coercion never changes; it merely shifts from one party to another. But once we recognize that coercion varies in moral significance, a more interesting possibility emerges: institutional arrangements can be evaluated by whether they employ moderate or trivial coercion to reduce or eliminate more significant coercion elsewhere. Consider bodily autonomy rights. We do not hesitate to enforce such rights even though they restrict the options of would-be aggressors, because the coercion those restrictions involve is trivial compared to the coercion they prevent. A similar logic underlies the strongest case for private property and market exchange. In conditions of Lockean abundance, the original acquisition of a parcel of land imposes negligible coercion—closing off one plot among many scarcely restricts anyone's options—while making possible production, investment, innovation, and the dramatic improvements in living standards that follow. As Schmidtz (1994) argues, this case extends even beyond conditions of abundance: private property creates incentives for stewardship and productive use that expand the total set of options available, so that the coercion involved in exclusion is more than offset by the opportunities property institutions generate. On the graduated framework, the question is not whether property coerces—it does—but whether the coercion it involves is justified by the reduction in more significant coercion it makes possible. For well-functioning market economies—those with competitive markets, adequate safety nets, and the rule of law—there is at least a strong prima facie case that it is, though the answer will depend on the specific institutional details. ²²

It is worth noting, finally, that the graduated framework has implications for the evaluation of regulation as well. If the question is whether government regulation is “coercive,” the answer is the same qualified yes: regulation restricts options, and its moral significance depends on what options are curtailed, what alternatives remain, and whether it involves domination. Workplace safety regulation that prevents employers from exposing workers to hazardous conditions reduces morally significant coercion (threats to bodily integrity with few alternatives) even as it introduces a less significant form (restriction of employer discretion in a domain where employers have many other options). Conversely, occupational licensing that bars low-income workers from entering a trade restricts options central to livelihood while offering no compensating benefit to the workers themselves—and so constitutes a more significant form of coercion that demands stronger justification (Kleiner, 2000). Zoning laws that prevent the construction of affordable housing restrict the options of the most vulnerable while protecting the relatively trivial interest of existing homeowners in neighborhood aesthetics (Kogelmann, 2026). The graduated framework does not presume that regulation always reduces or always increases the moral significance of coercion; it asks the right questions about each case. Hale was right that redistributing legal entitlements does not change the total amount of coercion in the non-moralized sense. But he was wrong that this makes coercion evaluatively inert, because redistribution can change the total moral significance of coercion—and that is what matters for institutional evaluation. This is the kind of comparative assessment that Hale's framework, with its zero-sum commitment, cannot support—but that any serious assessment of institutional alternatives requires.

Conclusion

Are markets coercive? The answer I have defended is: yes, in a non-moralized sense, but the moral significance of that coercion varies enormously.

The libertarian who answers “no”—who defines coercion in moralized terms so that voluntary exchange is non-coercive by definition—builds the conclusion into the premises and drains the concept of independent evaluative force. Hale, who answers “yes, and equally so for all institutional arrangements,” answers correctly on the descriptive level but renders the concept normatively inert. A word that describes everything condemns nothing.

The better answer lies between these poles. Coercion is a real feature of market arrangements—property rights restrict options, and that restriction is backed by institutional power. But some restrictions matter more than others. Coercion that curtails options central to well-being, that leaves people with few or no acceptable alternatives, and that involves subjection to another's arbitrary will is morally significant in ways that routine property exclusion in a competitive market with a functioning safety net is not.

This graduated framework does what both the libertarian and Halean accounts cannot: it distinguishes cases where market coercion is genuinely worrisome from cases where it is not. It concedes the kernel of truth in Hale's insight—that property involves constraint—while preserving the concept of coercion as a tool capable of genuine evaluative work. And it directs attention to the specific conditions that matter: the adequacy of alternatives, the importance of the options at stake, and the presence or absence of domination.

Paradoxically, this framework provides a sounder defense of markets than the libertarian moralized approach it replaces. The libertarian who insists that voluntary market exchange is non-coercive by definition will persuade no one who does not already accept the moralized premise—and will have no resources to explain why competitive markets are preferable to monopolistic ones, since neither involves “coercion” on the moralized view. The graduated framework, by contrast, can explain precisely why well-functioning markets deserve support: competitive markets with adequate safety nets minimize the moral significance of coercion by expanding alternatives, protecting vital interests, and dispersing power. They do not eliminate coercion—no institutional arrangement can—but they substantially reduce its moral significance. Whether they do so more effectively than all conceivable alternatives is a further question, one that the framework helps to formulate but does not by itself answer. What the framework does establish is that the case for markets need not rest on the implausible denial that markets involve coercion at all. This is the honest and, ultimately, the stronger starting point for evaluating market institutions. It is also one that Hale's framework, for all its initial appeal, cannot even formulate.