INTRODUCTION
“A legal system exists,” Joseph Raz claims, “if and only if it is in force.”Footnote 1 By this he means to suggest that the efficacy of law—that is, its capacity to control the population to which it applies—is a condition of its identity as such. If, for example, revolutionaries wishing to replace a monarchic order with a republican one fail to bring to heel those over whom they are asserting authority then their efforts will only count as an attempt at a new legal order.Footnote 2 So too, if the law in a given jurisdiction loses its grip, either over the whole or a substantial part of the population to which it applies, then the system in question will, to the extent of its deterioration, have ceased to exist.Footnote 3 These observations bring home the important truth that while law is in part a matter of having certain kinds of institutions—courts, and perhaps legislatures and enforcement agencies—it is also a matter of the relationship between those institutions and the population to which their rules and rulings are addressed. Law, as Hart reminds us, is most fundamentally “a means of social control” and it is this fact that sets it apart as an object of persistent moral concern.Footnote 4
That efficacy is a condition of the existence of a legal system has long been accepted. Austin, following Bentham, built the idea into his conception of sovereignty. Law, he argued, was to be understood as the command of the sovereign where the sovereign was defined as that body to which “the bulk of [a] given society are in a habit of obedience or submission to… .”Footnote 5 Without submission, Austin claimed, there could be no sovereign and without the sovereign there could be no law.
Kelsen, with whom modern discussion of the topic began, held that “the efficacy of the entire legal order is a necessary condition for the validity of every single norm of [that] order.”Footnote 6 While the rules of a legal system were valid “because they are created in the constitutional way,” they maintained their validity “only on the condition that the total order [was] efficacious.”Footnote 7 And while disagreeing with him about the details, theorists such as Hart, Raz, and Finnis have followed Kelsen in drawing this implication.Footnote 8
Given the widespread acceptance that efficacy is a condition on the existence of law, it is perhaps surprising to find that few theorists have paid the concept much attention.Footnote 9 Raz, for example, devotes only part of one chapter of his The Concept of a Legal System to the question of what it means for the law to be in force, confining himself largely to criticism of Kelsen's account.Footnote 10 Kelsen and Hart, whose discussions of the issue were the most penetrating, nonetheless left many questions unanswered.Footnote 11 And while it is true that a theory need not be complete in order to be illuminating, it is hard not to see the question of law's efficacy either sitting behind or relating to a number of central issues in the philosophy of law. Skepticism about international law, for example, often hinges on the belief that it is unable to effectively constrain state action, at least so far as the more powerful are concerned.Footnote 12 So too issues concerning rival claims to legal authority, made famous by the pluralists of the 1970s and 80s and by the rise of supra-state institutions such as the European Union, are at their heart questions about what it means to say of systems of overlapping jurisdiction that they are each in force.Footnote 13 Perhaps most importantly, however, the question of what it means for a legal system to subject a given population to its control has important implications for the nature and justifiability of law's authority. It helps to explain why law, in its structural dimension, manifests as an instrument of social power.
In what follows I will sketch an account of the efficacy of law that is broadly Kelsenian in spirit. In the first section I make the case for efficacy as an existence condition of law.Footnote 14 The second section considers the relationship between the efficacy of a legal system and the efficacy of its constituent laws. In the third I explain how the notion of obedience plays into that question, as well as saying how an account of law dependent on its capacity for control can accommodate power-conferring laws within its compass. The fourth section argues, with Kelsen and contrary to Hart, that the efficacy of law should be determined not only by its capacity to secure obedience but also through its capacity for enforcement, by way of punishment or remedies issued in response to legal wrongs. Here I also say something about how the discussion relates to the issue of the ontological status of international law.
I. LAW AND LAWLESSNESS IN YAP, MICRONESIA
That the validity of particular legal rules does not depend on their efficacy is easily recognized. The fact that many flout the speed limit, for example, does not transform the motorway into a rule-free zone. So too, that a statute has been ignored or overlooked in previous judicial decisions does not prevent it being the courts’ continuing duty to honor its provisions. This latter point is qualified by the fact that certain legal systems contain a custom—known as desuetudo or the rule of obsolescence—requiring that courts treat particular rules of law as invalid by virtue of their long-term nonapplication. But to say that a rule of law may become invalid by virtue of its inefficacy is quite different from saying that it was valid in the first place because efficacious.Footnote 15
From the question concerning the validity of individual rules of law we should distinguish that concerning the existence of the legal system as a whole, the legal order as Kelsen called it. Consider the way in which claims of law, that it is unlawful to consume recreational drugs, for example, depend on or can be said to presuppose the existence of an efficacious legal system. Just as it makes little sense to claim of someone that they are “in check” when not engaged in a game of chess, so too to assert that the law requires that individuals avoid illicit substances does not make sense absent the background assumption of a functioning legal order.Footnote 16 To make such a claim where the set of norms of which the rule forms a part either has never taken root or has fallen into disuse would involve a form of presupposition failure. In such a situation, Hart tells us, “the normal context or background for making [a] statement in terms of the rules of the system is missing.”Footnote 17
Considerations of this kind are standardly expressed in terms of the requirement of an “institutional context,” Anscombe famously having argued that assertions of social obligation require “institutions as background” in order to be “the kind[s] of statement[s]” that they are.Footnote 18 And while this claim is good as far as it goes, what I want to stress here is that for a given society to have law it is not enough that the relevant institutions exist; they must interact in the right way with the population over whom they claim control.Footnote 19 Consider, by way of analogy, a refereed sporting game. For such an event to take place it is not enough that some individual deems themselves the game's official and judges play according to the rules. It is important, too, that their authority is actually accepted by those engaged in the match, that those playing the game defer to their decisions. In the absence of such a relationship of authority—one in which the determinations of the referee are taken as binding—we lose the sense that the game has an official. So too if disregard for the rules of a legal system were to become both general and protracted, for example as a result of widespread and lasting civil disobedience, the notion that the society in question actually had law would become harder to make sense of. Law, like the game, requires not just institutions and officials but institutions and officials whose activities properly affect the lives of those over whom they claim jurisdiction.
It is important to understand the limits of this argument. When it is suggested that statements of law only make sense in the context of a functioning legal system, the claim is not that such assertions only have meaning in such a situation and appear otherwise as nonsensical.Footnote 20 The argument is not about the sense but about the pragmatics of legal statements; claims of law made in absence of a functioning legal system are intelligible but they occur apart from the background that makes them truth-evaluable. The argument is, in this way, both about the pragmatics of language—what is presupposed when we make claims of legal right and obligation—and about the grounding of these very rights and obligations.Footnote 21
But just as the significance of the argument should not be overemphasized so too its importance should not be underappreciated. The idea is not merely that legal rights and obligations are only enforceable in the context of an efficacious legal system, although this is of course true, but that they can only be said to exist in such a context. Even in reasonably well-functioning legal systems important rights sometimes go discarded and duties unenforced. However, if the law were to continue progressively to lose control over its subject set, we would reach a point at which it was no longer sensible to claim that those notionally under its jurisdiction had the rights or obligations pronounced on by its institutions. In the first type of situation there is a deficit in the law. In the latter there is no law.
Brian Tamanaha, a legal theorist who approaches the subject from the perspective of sociology, thinks that this claim about the conditions under which law can be said to exist is too strong. Using as an illustrative example the case of Yap, Micronesia—which was for a number of years under U.S. trusteeship—he suggests that we should drop the “requirement of general obedience” from our understanding of law.Footnote 22 “Yap,” he claims, “had a legal system, with a legislature, a handful of judges and attorneys … and a complete legal code based in its entirety on laws transplanted from the United States.” Large parts of the code, however, were completely unknown to the island's population:
few lay people had any knowledge of the content of the laws or of the operation of the legal system, a large proportion of social problems were dealt with through traditional means without participation of the state legal system, and indeed on most of the islands there was no legal presence at all.Footnote 23
“Although the overwhelming majority of the populace lived in general disregard of the vast bulk of the rules of the legal system,” Tamanaha suggests, “it is wrong to say that their state legal system did not exist.”Footnote 24 “The existence of state law in Yap,” he claims, was “a social fact … despite its failure to meet the efficacy requirement.”Footnote 25
What should we say in response to Tamanaha's argument? The first thing to note is that a legal system may well fall far short of the jurisdiction claimed for itself and yet exist because it has control over part of its de jure subject set.Footnote 26 This much was true, for example, of the Soviet legal system in its final years and, indeed, this may have been the best explanation of the situation in Yap under U.S. trusteeship. For, as Tamanaha himself notes, although the law was ignored by many of those who lived on the island, especially in the countryside, it was nonetheless effective in regulating “the affairs of running a government” as well as much of life in the capital.Footnote 27 So even accepting that efficacy is a condition of law's existence we can say that the state legal system in Yap existed, albeit with a much more limited jurisdiction than that which it claimed for itself.
But perhaps Tamanaha's argument is of a different form. His suggestion may not be so much that the law in Yap applied to a narrower range of the general population than that over which it claimed authority, but that the system's reach never extended beyond the acts of its agents and institutions. Indeed, this may be what Tamanaha means when he says that in Yap “social practices gave rise to the existence of a state legal system … based upon the self-identification of state legal actors.”Footnote 28 Should we be willing to accept the possibility of a legal system whose power extends only to the very institutions—courts, legislatures and other officials—whose actions give rise to its requirements in the first place?
In a way the question is ill conceived because we can only make sense of questions concerning the law's de facto authority if we know what areas of social life the order in question aims to regulate: to understand whether a system of norms effectively controls behavior we have to know what kinds of behavior it sets out to control. All of this presupposes the existence of institutions whose job it is to authoritatively make, determine, and apply the law. Without, in particular, courts in place identifying what is to count as the law for a given jurisdiction, we do not have a body of rules about which we can ask: “Are they effective?” So the question of efficacy assumes but does not extend to the initial composition of legal institutions.
One way of responding to this claim is to note the possibility of self-regarding normative systems—for example, those that comprise unincorporated associations or societies—whose only function is to control the behavior of their members. Could we not try to imagine a legal system similarly confined in its scope, to the regulation and control only of the behavior of its agents? It is, I think, hard to see how we could admit this much without losing the important distinction between the rules of a club—inward facing, voluntary—and law—a system of institutional control over a population.Footnote 29 This distinction helps us to explain, for example, why law and not the club will always be an appropriate candidate for assessment in terms of its political legitimacy. It also helps to explain why law and not the club involves a form of social regulation that is necessarily as opposed to contingently morally hazardous.Footnote 30
For these reasons, then, we should reject the argument that attempts to make plausible the idea of a legal system whose control encompasses only the regulation of the “affairs of … government.”Footnote 31 All legal orders, at least to some degree, extend their authority over a population outside of the institutional or official realm and they do so as a matter of necessity. But what, more precisely, does it mean to say of a system of law that it is in force in this way? What are the conditions under which we can claim that the law has effective control over its subjects?
There are three aspects to this question. The first involves specifying the relationship between the effectiveness of individual legal rules and that of the system as a whole. While a legal system being in force is not the same thing as each and every rule of the system being efficacious, the analysis of the effectiveness of individual legal norms must play into the analysis of the existence of the institution as a whole. The second issue concerns what it means to say of law's obligation-imposing requirements that they are, generally speaking, effective. This includes saying something about how power-conferring rules fit into an analysis of law premised on its capacity for control. Finally, it needs to be considered whether, and if so in what way, enforcement—dependent, as it is, on a breakdown of obedience to the law's primary obligation-imposing requirements—plays into the question of the system's efficacy. I treat each of these issues in turn.
II. THE EFFECTIVENESS OF THE LAW AND THE EFFECTIVENESS OF ITS LAWS
A partial failure of the law to effect control no more calls into question the existence of the system than the failure on the part of one person to move their bishop diagonally means they are no longer playing chess. But, equally, just as a near total disregard for the rulebook does raise the question of whether we can sensibly say of someone that they are playing the game, so there exists a limit in terms of the noncompliance that the law can tolerate while maintaining its identity. Kelsen put the point this way: a legal system exists just in case “its norms are by and large effective.”Footnote 32
What should we make of this definition? As noted earlier the criterion is not one that needs to be satisfied in relation to the law's claimed or de jure jurisdiction. A legal system may well exist even if it effectively regulates only part of the population to which its norms are addressed; the failure of the state or other political organization that rules through law may be fractional and is at any rate not a binary matter.Footnote 33 The idea is that effective control determines the limits of the legal system and a legal system must control some more than de minimis population in order to exist.
The second point to note about the definition is that it marks only a vague boundary between situations in which the law is in force relative to a target population and those in which it is not. But this should come as no surprise. Most distinctions worth marking in the philosophy of social and political institutions—for example, between legitimate and illegitimate forms of government—are similarly vague. Moreover, we should not expect more from this aspect of the theory of law than we do from others. Just as the boundary between legal and other kinds of institution—for example, religious or revolutionary forms of association—is imprecise, so too is the distinction between a society that is effectively ruled by law and one that is not.Footnote 34
Finally, the definition specifies a particular relationship between the efficacy of individual rules of law and of the legal system as a whole. Because the effectiveness of the system is construed in terms of the by and large effectiveness of its norms—a term of art for Kelsen, about which I will say something below—the analysis of the former will depend on the analysis of the latter. But which kinds of legal rule bear on the law's capacity to control conduct and under what conditions can we assert that they effectively do?
III. LAW'S DOMINION
Kelsen embraced a reductive theory of legal norms according to which all such rules were ultimately to be understood as instructions to judges to apply sanctions in concrete cases (“if such and such conduct has come to pass then the judge must apply the specified sanction”).Footnote 35 The model most closely tracks our understanding of the criminal law, intimately concerned as it is with punishment. Even here, however, there are problems. Such laws are first and foremost instructions to the public not to commit the relevant wrong and only in their second act—the one it is hoped will not come to pass—do they function as premises in judicial decisions designed to sanction the wrongdoer.Footnote 36 Further issues arise when attempting to assimilate power-conferring laws—such as those involved in the formation of contracts, for example—into the model without losing sight of the point or purpose of such institutions: to allow individuals to create binding obligations, as opposed to forming part of a set of conditional instructions to judges to award damages in the case of breach.Footnote 37
These are problems in Kelsen's understanding of the individuation of legal norms, but what he said about the effectiveness of legal rules was more illuminating:
By effectiveness of a legal norm, which attaches a sanction to an certain behaviour and thus qualifies the behaviour conditioning the sanction as illegal, two facts may be understood: (1) that this norm is applied by the legal organs (particularly the law courts), which means, that the sanction in a concrete case is ordered and executed; [or] (2) that this norm is obeyed by the individuals subjected to the legal order, which means, that they behave in a way which avoids the sanction.Footnote 38
Shorn of the trappings of Kelsen's reductive theory of legal norms we arrive at the following: first, that the effectiveness of legal rules is determined either by submission to their requirements on the part of the law's subjects—“that this norm is obeyed by the individuals subjected to the legal order”—or by their enforcement in the courts—“that this norm is applied by the legal organs.”Footnote 39 Second, and as an implication of the first condition, that the question of the effectiveness of legal rules is confined to those that impose obligations, for laws that create powers—for example, those that allow us to make contracts or deeds—cannot be obeyed although they can of course be complied with. As the question of enforcement raises a range of relatively independent issues I will discuss it in the next section. Here I want to focus on the notion of obedience as well as issues around power-conferring laws.
What, then, does it mean to say of duty-imposing rules that they are effective?Footnote 40 More than that the behavior of individuals coincides with their requirements. As Donald Regan puts the point: “obedience involves something more than … behavioral compliance with [a] directive.”Footnote 41 Mere congruence between the activities of individuals and the stipulations of the law—as per, in partial modification of the original story, the tide happening to turn at the appropriate moment in King Canute's performance—is not enough to show that the latter effectively regulates the former. For this to be the case the law must play a role in determining the behavior of those subject to its rules.Footnote 42
Equally, however, obedience should not be seen to imply that the relevant behavior comes about simply because the law mandates it. Most people, for example, avoid killing or doing serious injury to others quite apart from the fact that these acts are criminalized.Footnote 43 Nonetheless, it would be wrong to say of such individuals—on the assumption that they are in general law-abiding—that they do not obey the law.Footnote 44
How, then, to square the circle? Obedience to law should be thought of as setting a background condition on permissible action. While the law may be effective in certain types of cases because individuals treat its existence as supplying their primary reason for action—for example, when they act only to avoid the punishments associated with the breach of its rules—many act consistently with the law's requirements at least in part for reasons that stand apart from the fact that they are legally mandated. And for these people the law's effectiveness should be gauged counterfactually, based on whether they would comply with its requirements even if the reasons that otherwise compel their action were not present. In this way obedience involves taking the law to count decisively in favor of the action it prescribes, although this need not in fact involve the law making a practical difference to behavior.Footnote 45
Where does this analysis leave those aspects of the law that confer powers as opposed to imposing duties? It might be thought that defining efficacy in terms of obedience leaves large swaths of the law—for example, the institution of contract or the law-making powers of legislatures—unaccounted for in determining whether or not a legal system is in force.Footnote 46 Such a result would indeed be counterintuitive, but does it follow?
Take private powers first. While the rules that determine how to form legally binding agreements cannot themselves be analyzed in terms of their ability to control behavior—the law does not require that individuals contract and so it does not count against the efficacy of the system if they do not—a consequence of the use of such rules is the creation of new duties. A contract to pay £20 for the delivery of a box of fruit, for example, creates obligations on the part of one party to pay the money and on the part of other to deliver the fruit. The efficacy of the institution can then be measured, in the first instance, by the extent to which people keep to their contracts. It is, in other words, the duties created as a consequence of agreement by which we first measure the effectiveness of contract law.Footnote 47 Even specifying the efficacy of law in terms of control, then, we can still sensibly talk of the force of institutions that involve private powers.Footnote 48
What of public powers? The efficacy of these laws can be measured, again, not by the extent to which the relevant power is exercised—how many times, for example, should a legislative body use its law-making power in a particular session for that procedure to count, in the relevant sense, as efficacious?—but by the extent to which the laws that are created as a result of its exercise are themselves effective within the system.Footnote 49 A power to create law is not rendered ineffective because it is not used. It is, however, if the rules created through its use fall on deaf ears.
Raz disagrees with this point, holding that the failure to exercise power may itself be directly relevant to the efficacy of the system. He asks us to consider a situation in which “members of some racial or ethnic group are considered as second-class citizens and do not enjoy political rights”:
Suppose that the government appoint some of their members as a semi-autonomous legislative assembly, and that these persons, in protest against the regime, refuse to make use of their powers. Is not their action as detrimental to the existence of the legal system as any act of violating duties in civil disobedience?Footnote 50
I think the right response here is to say that it might be but not because the legislative body has failed to exercise its law-making powers. The failure of the assembly to carry through on its agenda no more brings the system into disrepute than the famous logjams in the United States Congress question the efficacy of American law. The crisis is in the first place one of politics and just so long as the rules that each institution has produced remain efficacious, so the law retains its authority. Protest of the form that Raz describes might, however, lead to a situation in which ordinary forms of deference to the system are disrupted. The actions of the minority members are presumably intended to delegitimize the legislature and if effective could interfere with obedience to its laws on the part of the more general population. Disobedience of this kind would uncontroversially go to the efficacy of the system. It is, in other words, a possible consequence of the act of protest on the part of the minority members of the assembly that speaks to whether the law retains its force, but not the bare refusal to legislate in and of itself. By distinguishing between the fact and consequence of inaction in this context, then, we can allow Raz the intuitiveness of his example while at the same time maintaining that the efficacy of a legal system is to be gauged in terms of its capacity for control.
IV. ENFORCEMENT
I have argued that the law's effectiveness is to be measured in the first place by adherence to its primary obligation-imposing requirements: whether people keep their contracts, for example, or whether criminal activities are avoided. Is this, however, the only register by which we are able to discern the law's efficacy? Kelsen, as we have already seen, thought not: the efficacy of a legal norm was to be measured either by obedience or in the alternative by its “application in a concrete case,” that is, by its enforcement in the courts.Footnote 51 For Kelsen the law's effectiveness does not run out if individuals fail to keep to their bargains, for example, just so long as the courts are in a position to ensure that damages are paid. So too the law retains its force even when subjects disobey the requirements of the criminal law if the associated punishments are imposed.
Hart thought otherwise. While distancing himself from many aspects of Austin's austere command theory of law—in accordance with which law was to be understood as the command of the sovereign backed by force—he accepted that Austin's criterion for the existence of the sovereign, “the … formula of a general habit of obedience to orders[,] does designate one necessary condition [for the existence of a legal system]: namely, that where the laws impose obligations or duties these should be generally obeyed or at any rate not generally disobeyed.”Footnote 52 Because the fact of obedience is on this conception necessary for the existence of law, and because questions of enforcement arise only on the breakdown of obedience to the law's primary norms of obligation, so Hart did not countenance answerability in the courts—the imposition of punishments or liabilities—as a means to law's efficacy.
Who had the better view? I think that the obvious answer here is Kelsen but that the account requires some modification. First, though, I need to explain why we should reject Hart's approach. The theory, based as it is entirely on the notion of obedience, has the implication that a legal system in which the law effectively, even ruthlessly, punishes or otherwise coerces those who fail to honor the system's primary rules of obligation is, to the extent to which it does so, inefficacious. But a legal order does not cease to be effective just because large numbers end up behind bars or subject to other kinds of penalties. If it did, the United States of America, a country whose per capita prison population remains the highest in the world, would be, to the extent that it engages in the practice of incarceration, inefficacious. No one, however, doubts the effectiveness of that legal regime and a large part of the reason for this has to do with its capacity for enforcement.
The same point can be put differently: if Hart's view were right any legal system that fell below a baseline requirement of general obedience would, by virtue of this fact, cease to exist. This would be so, moreover, regardless of whether the regime in question were able to effectively respond to the fact of noncompliance by sanctioning wrongdoers. So on the Hartian approach there is an equivalence between a system that is widely disobeyed and that cannot bring to heel those over whom it asserts authority and one that, despite widespread disobedience, is able to do so. In both cases we would have to say that the legal system has lost its force. But while we might have reason to consider a legal system that controls behavior primarily through its mechanisms of enforcement to be pathological or unhealthy it seems wrong to deny it the title of law simply on this basis. The failing of Hart's understanding is that it refuses to countenance the occurrence of the legally stipulated response to the fact of wrongdoing for what it is: a sign of the functioning of the legal system.Footnote 53
So Hart's approach ought to be rejected. The efficacy of law should be measured either by the fact of obedience, or in the alternative, through the notion of enforcement.Footnote 54 But how should the latter concept be understood? Kelsen seems to have thought of the idea in factual terms, requiring that a “sanction in a concrete case [be] ordered and executed” in order to satisfy the criterion.Footnote 55 In this way the law would count as efficacious either if it was obeyed or if those who had offended against its requirements were brought before the courts and effectively sanctioned.
One complication with this proposal is that the law in many jurisdictions contains mechanisms that place within the hands of individuals or institutions the decision as to whether to pursue wrongs in the courts. State authorities, for example, will often have the discretion not to prosecute crimes and the law does not compel individuals to vindicate their rights in civil cases. This raises the possibility of situations in which the law is neither complied with nor enforced despite the fact that the latter option remains available. Such cases fail to satisfy Kelsen's understanding of efficacy, but intuitively there is a difference between a situation in which the law remains unenforced because it cannot be and that in which it is unenforced because it has been decided that it should not be.
How, then, to accommodate this distinction? Instead of understanding the notion of enforcement in factual terms we can think of it as a capacity. On this reading the law will count as efficacious either if it is actually enforced or if it is capable of being enforced. So a discretionary decision on the part of the state not to pursue a crime, for example, would not count against the law's effectiveness just so long as it could be punished. So too the civil law would not count as inefficacious merely because people choose not to pursue their rights in court as long as they would be able to do so effectively if they so decided.
This modification of the account also helps us to deal with a further problem. Kelsen appears to have assumed that the notions of obedience and disobedience were binaries, that the subject necessarily either obeyed or disobeyed the law.Footnote 56 But the account of obedience sketched earlier, being cognitivist in character, opens up a third possibility. If, as I have suggested, obedience depends on intentional submission to the law then there will exist a category of cases in which the individual neither obeys nor disobeys the law, for example where they happen to comply with the law's requirements in ignorance of its terms.Footnote 57 How to judge efficacy in such cases? Here too I think we can point to the possibility of enforcement. The law will count as efficacious in such a situation if, were the individual to fail to conform to its requirements, the law could be effectively enforced against them. This makes the question of the law's efficacy in large part one about its potential for control.
Applying the account systematically we can say that a legal system will count as effective if most of its laws are either obeyed, enforced, or otherwise stand capable of being enforced. In the standard case a legal system will secure its efficacy through a combination of these techniques. Some laws will be effective because the population submits to their requirements and others will be honored in the breach, through answerability in the courts. The relationship between these aspects is dynamic. Legal systems may at times be able to achieve widescale obedience, a possibility that will often be secured in part through the threat of enforcement and consequent incentive to respect the law. At other times the actual mechanisms of enforcement will themselves become a bigger part of the picture; the less that the law can extract from its subjects in the way of compliance the more reliant it will become on the possibility of force to ensure its survival.Footnote 58 Indeed, the disjunctive nature of the account raises the following possibility—that a legal system might secure its existence almost entirely through enforcement with next to no compliance with its primary norms. The coming to be of such a reality is admittedly farfetched (although it is worth noting that in many actual legal systems aspects of the law are experienced in this way by a subset of subjects).Footnote 59 Some might worry, however, that the possibility of a system of pure enforcement—a system in which there has been a near total breakdown of respect for the law's primary rules of obligation—is so far removed from our ordinary understanding of law so as to count as a reductio against a theory that countenances the possibility.
Jeremy Waldron articulates a concern of this type. We need to be able to distinguish, he suggests, “between a legal system [and] the activity of an occupying army”:
An occupying army may have a chain of command and an organizational apparatus of its own … but it may not be appropriate to say that that apparatus exists as a legal system among the members of a social group that comprises not only the occupiers but also those whose lives they govern, … those who feel nothing but the sharp end of their bayonets.Footnote 60
A legal system should not be said to exist, Waldron suggests, “unless most of the population do abide by the … rules of the system.”Footnote 61 If they do so “only to the extent (say) that cats and dogs do,” that is “to the extent that their physical behavior is actually forced into conformity by leashes, chains, muzzles, kicks, etc.,” then they cannot be thought of as governed by law.Footnote 62
Waldron is right, I think, to distinguish between law and the situation of the occupying army; a legal system is importantly different from a system of pure force. But he is wrong to believe that it is only through the notion of generalized obedience that we are able to do so. Even in a situation where law makes contact with the population almost entirely through its mechanisms of enforcement, its so doing is importantly different from the naked application of force. Enforcement, for example, functions as a response to the breach of a rule addressed to a wrongdoer. Mere force does not. So while the army's bayonet may be brandished for many purposes, the law involves the use of its instruments of control in a way that is reactive to a predetermined wrong. Enforcement, moreover, is by its nature mediated by formal mechanisms that determine the appropriateness of this as opposed to any other reaction. The law's response is in this sense institutionalized whereas the acts of the army need not be. That is enough to distinguish even a legal system that operates principally through “leashes, chains, muzzles [and] kicks” from a system of pure force.Footnote 63
None of this is to suggest that life under law is going to be preferable to life under an occupying army, and for some it might not be. The bayonet after the breach may be no less painful than that which comes unannounced. What it is to say, however, is that although enforcement might manifest as a subspecies of force it is not its equivalent. As such we can continue to countenance the possibility of a legal system almost entirely dependent for its effect on its mechanisms of enforcement, while at the same time holding that an occupying army that makes contact with a population only through “the sharp end of [its] bayonets” does not for this reason rule them through law.Footnote 64
A further objection to the account involves the idea that it concedes too much to the reductionist aspects of Kelsen's theory of law. While we have been careful to distinguish Kelsen's conception of the individuation of laws from his theory of efficacy, it might still be suggested that by adopting aspects of the latter we engage in a similar form of reductionism to that which is involved in the former. If Kelsen's account of individuation—that all laws must ultimately be understood as conditional instructions to judges to apply sanctions—misses out on the way in which the law aims, for example, to secure compliance or facilitate exchange, is it not similarly distorting to hold that law is in some sense indifferent as between individuals keeping to their contracts and having damages orders enforced against them, or between the criminal law securing obedience and punishing wrongdoers?
The objection involves a misunderstanding of the claim being made. The argument is about the existence conditions for law as a social kind and not about the point or purpose of particular legal doctrines. We can consistently hold that the law counts as efficacious, for example, if people either stick to their bargains or the courts are in a position to ensure apposite remedies for breach while also recognizing that the aim of the law of contract is to ensure that agreements are honored. So too we can maintain that punishment goes to the effectiveness of the legal system while also understanding that the criminal law aims to ensure compliance. The problem with Kelsen's argument was that his structural claims obscure our ability to comprehend the telos of the law, in preventing wrongs or ensuring that agreements are upheld. But the question of force, with which we are here concerned, is independent of that of the law's purpose.
Finally, it is worth considering where this account of the relationship between efficacy and enforcement leaves institutions such as international law that are famously lacking in enforcement mechanisms, at least when compared with state law.Footnote 65 How does the conception of efficacy that we have been developing apply to institutions of this kind? I think that the answer is as follows: such systems, because missing, for example, courts of compulsory jurisdiction, will be far more dependent on obedience than national legal systems to secure their existence. This makes the question of whether primary rules of obligation are respected much more central to the integrity of the international order than national law because in the case of default the latter can usually turn to institutionalized mechanisms of enforcement to secure its existence whereas the former cannot.
What does this tell us about the normative status of international law?Footnote 66 Just so long as a minimum standard of obedience is met it will count as an existent system of law, albeit one whose standing as such is more fragile than that of its national counterparts. This strikes me as the right explanation of the international order, as distinctly legal but nonetheless more dependent on acquiescence, and in this way more unstable, than other systems of law.Footnote 67
V. CONCLUSION
In this paper I have sought to argue for the importance of efficacy to our understanding of the nature of law, as well as to develop and defend an account of that concept against a number of important objections. In doing so I hope to have reinforced the importance of understanding law not only as a system of rules but also, and equally as important, as one of domination or control.
