Consider law itself, laws and legal systems as three distinct foci of concern. Coercion may and has been used as an explanatory notion with respect to any or all of these in any combination, and so combined with (or perhaps without) logical consistency. Hans Kelsen, the preeminent legal theorist of the twentieth century, held that norms are legal norms only if they are coercive (Kelsen 1945/1961: 19). By this he meant that a scientific account of the content of a legal system would necessarily consist of a set of statements of the following hypothetical form. The antecedent of each such hypothetical will specify a condition, and the consequent a coercive sanction that is authorized or required in case the antecedent condition is met. Such a statement, or rechtsatz, could take this form: “If anyone murders another, then that person ought to be executed.” Execution is the coercive sanction. The norm of law described by jurisprudence—the science of law—need not be of hypothetical form, and need make no reference to a coercive sanction. If the norm is expressed in a statute, it might simply state: “The penalty for murder is death.” The legal norm is not to be confused with the scientific statement of the content of the law. The latter makes essential reference to coercion, or so Kelsen believed, while the former need not. Legal theory, whose task is not to inventory the legal norms of any given system of law, but to explain how those norms are both valid and a unitary system, has no further use of the concept of coercion. So, to summarize (while oversimplifying) Kelsen’s view, coercion is an essential element in the scientific statement of the content of a legal system, and in that sense a legal system is necessarily coercive. A legal norm itself need not employ the idea of coercion. But no norm is a legal norm unless it is both valid—that is, authorized by some superior norm—and a condition of a coercive sanction.

Kelsen’s account of law employs a conception of coercion that is roughly equivalent to the idea of compulsion. This usage sets him apart from recent Anglophone theorists, who distinguish between coercion and compulsion (but cf. Anderson 2008). Coercion, according to the currently dominant view, is properly understood as involving a threat, or an expression of some undesired consequence to be brought about if, but only if, the addressee fails to do as commanded. Coercion differs from altering another’s behavior by issuing a warning, in that the coercer at least pretends to have control over the occurrence of the undesired outcome (Nozick 1969). John Austin’s (nineteenth-century) account of law was founded on the concept of coercion as a threat—law was properly defined as a set of orders from a sovereign, given under threat of sanction, to subjects who habitually obey (Austin 1954/1832). Both Kelsen and Austin are key figures in the history of legal positivism. Both believed that coercion had an essential role in what we call law. But they had different ideas of what coercion is, and where it figures in.

Which conception of coercion better explains what law is: Austin’s notion of coercion as threat, or Kelsen’s notion of coercion as compulsion? This essay will not take a position on that, but it is worth noting that more recent legal theorists have been more of Austin’s mind than Kelsen’s on this point. Lon Fuller, a natural law theorist, insisted that making law involves more than merely authorizing force. Law, in Fuller’s view, is essentially a public direction to all citizens subject to law, and is therefore not in essence duly authorized compulsion (Fuller 1964). H. L. A. Hart, a critic of both Austin and Kelsen, conceived of law as in essence a matter of rules, and of rules, in turn, as human practices, organized in a distinctive way, having both an internal and an external aspect (Hart 1961). A rule can be said to exist only if someone views it as a standard of behavior, such that lapses from that standard are subject to criticism. But not everyone subject to a rule need view it that way. Distinctively legal rules, however, are backed by authorizations to apply “serious social pressure” to assure conformity. By social pressure, Hart did not mean to confine law to the application of force, at least not proximately. Law takes hold in a society when a certain set of rules of conduct are recognized and insisted upon as binding, in the sense that nonconformity may be met with serious social pressure to conform. The greater the degree to which “physical sanctions are prominent and usual,” the more inclined one will be to regard the rules as forming a “primitive or rudimentary” legal system, even though these sanctions be “neither closely defined nor administered by officials.” But a legal system and laws, in a robust sense, depend upon the presence of “secondary” meta-rules identifying the rules of conduct that are enforceable by serious pressure, providing for the introduction and retirement of rules so sanctioned, and—most importantly—rules governing the resolution of disputes involving the “primary” rules of conduct. It is this (distinctively organized) serious social pressure antecedent to the unwanted conduct that marks the existence of a legal rule, and that rule’s membership in a legal system.

For Hart, it is a mistake to think that coercion, in either Austin’s threat sense or in Kelsen’s compulsion sense, is essential to any particular rule or norm or command counting as a law. A rule may for example empower private parties to make binding agreements, or may disable a legislative body from regulations in some field, without failing to be law. But Hart agreed that coercion, at least in the sense of serious social pressure (forms of which may be culturally various, and need not include what we would describe as physical compulsion) was an essential part of any legal system and, to that degree, an essential part of an adequate theory of law.

It is important not to confuse the explanatory enterprise of showing law to be more than and distinct from mere coercion from the normative enterprise of redeeming the apparent coerciveness, or coercive element or dimension, of law. That law is coercive as a straightforward matter of descriptive fact is a commonplace view (Lamond 2000: 39). The coerciveness of law as a descriptive matter is often invoked as a way of introducing or of framing the question whether the state—regarded as a legal system effective in some territory—can be morally justified, and the (separate) question of whether law imposes a moral obligation of obedience. As John Rawls put it, “political power is always coercive power [and] this raises the question of the legitimacy of the general structure of authority” (Rawls 1993: 136–37). In particular, the assumed coerciveness of law, as a descriptive fact, is often taken to frame the questions of justification and legitimacy in a way that casts a burden of proof upon the defender or apologist for the state. Hart, for example, wrote, “We are committed … to the general critical principle that the use of legal coercion by any society calls for justification as something prima facie objectionable” (Hart 1963: 20–21). The rhetorical significance of casting the burden of proof upon the state is that the issues of justification and legitimacy are to be decided against the state, in case the argument in favor is inconclusive or contestable. The defender of the legitimacy of legal authority thus bears a “risk of non-persuasion” (Gaskins 1992) that the skeptic does not. Moreover, even if the apologist for the state succeeds in carrying this burden, many will maintain that the “objectionable feature that attaches to each instance of coercion persists even after a demonstration that the particular deprivation of freedom is justified [for it] is simply outweighed in such cases” (Husak 1983: 355, emphasis in original; see also G. Dworkin 1968).

The thought that the defense of the state has to bear a burden of proof has also influenced the kind of justification that political philosophers have believed to be required. In particular, many prominent twentieth-century philosophers have believed that the inherent coerciveness of law entails that any adequate defense of the state must command unanimous assent, at least at some level. The clearest statement of this unanimity requirement is put forward by Thomas Nagel: “In view of the coercive character of the state, the requirement [of unanimity as at least to some “higher-order principle”] becomes a condition of political legitimacy” (Nagel 1991: 150–51). But a similar stringency can be located in the work of Rawls, and Jeremy Waldron finds common to a wide range of liberal, libertarian and socialist theories the “requirement that all aspects of the social world should either be made acceptable or be capable of being made acceptable to every last individual” (Waldron 1993: 36–37). It is perhaps worth noting also a curious way of running this requirement in reverse: legal theorist Ronald Dworkin identifies the law of certain legal systems (i.e., those of most English-speaking countries) with whatever would justify—at least to an ideally perspicacious mind—their past applications of coercion. Whatever at a given time “figures in or follows from” the most attractive account of the legal history and practices of a given polity is identical with its law at that time (R. Dworkin 1986: 225). From this idealized judicial perspective, the law of modern Anglophone nations (if it exists at all) is necessarily coercive, but also necessarily justifiably and legitimately so.

Analyzing the Concept of Coercion

Assessing the normative sense that the concept of coercion has been assigned must wait upon further analysis of the concept itself. This analysis can best proceed by concentrating on a canonical form of the coercion claim, and the contexts in which such a claim might be asserted or rejected. A coercion claim is any statement approximately equivalent to one of the form:

By ping, person A coerces or attempts to coerce person B into ψing.

The ping by A will typically involve a declaration or threat of conditional form, which can be termed the coercive proposal. “Proposal” is meant to be a morally neutral term. Some writers have thought it important to distinguish threats from offers, and to analyze coercion claims in terms of threats; but the more promising approaches have avoided being drawn into disputes about where and how the line between threats and offers is to be drawn. Coercion claims—including statements that someone has been coerced or subjected to coercion—can occur in one or more of three distinct kinds of context. Coercion claims in these types of contexts at least potentially raise questions about justification. The three can be called, respectively, justification-supplying, justificationdefeating and justification-demanding contexts.

A coercion claim occurs in a justification-supplying context if the claim is offered in order to satisfy a demand for justification of another action that need not itself be coercive, but could be. “I was coerced!” is a representative coercion claim made in a justification-supplying context: it functions as an excuse more often than as a straightforward justification. An affirmative defense of duress raised to avoid criminal liability creates this kind of context. The proper dimensions of the duress defense in criminal law are a matter of controversy; and it might be argued that a coercion claim can never, in and of itself, supply a justification for wrongdoing, as opposed to a mere excuse (a distinction whose import is, itself, a matter of dispute).