The Efficacy Problem

Abstract

Legal theorists agree widely on two necessary and jointly sufficient conditions for the existence of a legal system: a legal system exists if (i) legal officials adopt a critically reflective attitude toward the legal system’s foundational rule, and (ii) the substantive laws of the system are “by and large” efficacious. The latter “efficacy condition” plausibly applies to all posited law, paradigmatically including modern centralized legal systems and less paradigmatic instances like international law. And yet, philosophers have also frequently pointed out the difficulty in determining precisely what this efficacy amounts to. In this article, I argue that the persisting difficulty of explaining the efficacy of law results from three tempting but inadequate assumptions about posited law and that our basic assumptions need to be revised accordingly.

I. Introduction

Legal theorists widely agree on two necessary and jointly sufficient conditions for the existence of a legal system: a legal system exists if, and only if, (i) legal officials adopt a critically reflective attitude toward the legal system’s foundational rule and (ii) the substantive laws of the system are “by and large” efficacious. The latter “efficacy condition” plausibly applies to all posited law, paradigmatically including modern centralized legal systems and less paradigmatic instances like international law. The efficacy condition also has a long and venerable pedigree, uniting both positivist and natural law theorists back to medieval scholastic views and eighteenth-century utilitarianism. And yet, philosophers have also frequently pointed out the difficulty in determining precisely what this efficacy amounts to. On the one hand, critically reflective acceptance of the law by the general population would be too demanding to be realistic; on the other hand, mere coincidental conformity of behavior would be too little to constitute any real influence at all. Thus, Bentham, identifying efficacy with obedience, already complained that it would be difficult to distinguish between the presence and absence of a “habit of obedience.”¹ More recently, Gerald Postema has suggested that these considerations of the efficacy of law even “yield an antinomy, or at least force on us a stark dilemma.”² In this article, I argue that the persisting difficulty of explaining the efficacy of law results from three tempting but inadequate assumptions about posited law and that our basic assumptions need to be revised accordingly.

Since the existing literature is not entirely uniform in its usage of the terms “conformity,” “compliance,” and “obedience,” I will keep to the following use throughout this article. By “conformity,” I mean doing what the law requires for any reason whatsoever, possibly ignorant of the law and possibly accidentally. By “compliance,” I mean doing what the law requires with knowledge of the law (albeit for any motivating reason, including independent reasons as well as threats of sanction). By “obedience,” I mean doing what the law requires because the law requires it. “Conformity” is thus the broadest term and “obedience” the most restricted; obedience entails compliance and conformity, whereas conformity entails neither compliance nor obedience.

In the opening sections of this article, I outline the difficulty of explaining the efficacy of law. Three related assumptions about law are that (i) posited law is a direct means of social control, (ii) law must be efficacious to exist, and (iii) law is efficacious by influencing the behavior of its addressees. While all three of these assumptions are individually appealing, they are also cumulatively incompatible with the actually existing forms of alienation of ordinary citizens from the law: the epistemic and motivational gaps. As socio-legal studies have frequently noted, law subjects have only little and selective knowledge of the substantive legal rules governing their daily lives. Moreover, legal theorists have frequently noted that even when people do know the law, their behavior is not typically motivated by the law.

To get the full force of the problem in view, the following sections survey the most instructive attempts to overcome this difficulty. Section V takes a close look at the telling example of Hans Kelsen’s evolving views from the 1920s–1960s, and Section VI discusses three recent attempts to overcome the difficulty of explaining law’s efficacy. Each of these recent attempts proposes a different kind of influence required for law to be efficacious such that posited law can be said to be effective despite the epistemic and motivational gaps. Unfortunately, none of these attempts successfully overcome the difficulty of explaining law’s efficacy. Consequently, I suggest that we need to give up the traditional version of our three basic assumptions about law in favor of the search for a new account.

II. Three Standard Assumptions About Law

Because these three assumptions about law have appeared almost self-evident to most legal theorists so far, these assumptions do not receive significant (or any) explicit defense in the literature. Rather, they are usually either presupposed, mentioned in passing, or themselves used to help explain other aspects of law. I will call these three assumptions characterizing the standard picture of efficacy the means principle, the efficacy condition, and the direction-of-fit requirement.³

In metaphysical terms, the means principle is an assumption about law’s characteristic act. Whatever law’s ultimate functions are, and whatever ultimate end lawgivers may want to achieve through positive law, law achieves them by guiding the conduct of those it governs.⁴ Intuitively, the means principle just expresses what it means for a norm to be practical (i.e., about conduct) and, consequently, also what makes positing a practical norm intelligible in the first place. Positing a norm is an intelligible act because one wants the norm to control behavior. On the face of it, it would seem unintelligible to posit a norm to regulate things that cannot be regulated or to posit a norm that does not aim to guide anyone’s behavior. In addition, it is usually assumed that the general rules of law (like commands) exercise control over law’s subjects in a way that has to do with the reason-giving nature of law’s norms itself. By this, I mean no more than that legal norms themselves are supposed to be reason-giving by triggering preexisting (independent) reasons for action. Indeed, this assumption about law is so tempting that it has also provided the basis for prominent views about law’s authority and the relation between law and morality.⁵ Whatever ultimate ends law is supposed to achieve, legal norms are assumed to fulfill their function by exercising social control, i.e., by guiding the behavior of law subjects.

If a system of posited norms is a direct means of social control, it can be said to exist only if it is effective or “in force.” In other words, if it is a characteristic act of posited law to govern behavior, then a legal system can be said to be actual only if it succeeds minimally in its characteristic act. This efficacy condition is usually said to be a necessary but not sufficient requirement for the existence of a legal system. Importantly, too, the efficacy condition is usually taken to be a requirement applicable to the existence of an entire legal system rather than the existence of individual legal norms. An individual legal norm can govern conduct merely because it has been validly created (according to the specific constitutional rules of the relevant normative system) even though it cannot yet be said to be effective, as, for instance, a newly created law which has never even had the chance of being put into effect. Equally, an individual legal norm, like the norm to not cross the street at a red light, may plausibly be said to exist merely in virtue of being validly created (e.g., according to the general traffic laws in England and Wales), albeit the individual rule is largely ineffective in the greater London area.⁶ However, an entire legal system cannot plausibly be said to exist unless it is “by and large,” “generally,” or “to a sufficient extent” effective.⁷ (Since the most often quoted term of art in this context is “by and large,” I will continue to speak about the efficacy condition as requiring that law be by and large efficacious throughout the rest of this article.) Unfortunately, this requirement of “by and large” efficacy is not spelled out in more detail within the existing literature. However, for the purpose of the present discussion, I suggest that a tentative way of spelling out this requirement further would be: for a legal system to exist, at least a bare majority of the areas of law pertaining to both the public and private everyday life of its participants (such as constitutional, criminal, tort, family, education, contract, employment, and medical law) must be at least more effective than ineffective among the law’s purported subjects.⁸ Finally, the efficacy condition also entails that legal rights and duties (as well as all other first- and second-order jural relations) can only exist in an effective legal system. Not only are legal rights and obligations only enforceable in an efficacious system, but they also only exist as legal rights and obligations if the respective normative system is, by and large, effective.

To be effective, it is not enough for a legal system that the behavior of the law subjects matches the prescription of the norms. To be effective, the law must be used by the population “in the right way.” Although there is little agreement as to what exactly this direction-of-fit requirement amounts to, there is, to my knowledge, no disagreement about its importance. One obvious reason for this direction-of-fit requirement is that it would be highly counterintuitive to call a legal norm effective if the norm were modeled after the behavior. A norm specifying that whatever happens according to the physical laws of nature is legally required to occur or that only that conduct is lawful in which the law subjects actually engage could not intelligibly be said to be effective.⁹ Moreover, efficacy is usually thought of as a property of the law. Consequently, efficacy must in some way attest to the influence, of some kind, of the law, on the law subjects. For instance, it is plausibly a property of Illinois state law that it is effective in the state of Illinois. If another normative system, say an entirely unknown and obscure religious code of a small community of three people in Springfield, IL, contained norms that accidentally prescribed much of what the people of Illinois actually do and proscribed much of what they do not do, that obscure religious code would still not be effective—it would be Illinois state law that is effective.¹⁰

These three assumptions about law just outlined appear to be almost universally endorsed—and reasonably so. At least on the face of it, giving up (1) the means principle might seem unappealing for two reasons. First, to hold that law was no direct means of social control would fly in the face of a long tradition of political thought that treated law as both an important subject of moral interest and a useful political instrument. Moreover, giving up the means principle would contradict the self-understanding of most legal officials regarding their own practice, as well as the intuitive view of most law subjects. Giving up (2) the efficacy condition would seem unattractive, too. Giving up the efficacy condition would make the existence conditions of a legal system implausibly easy to satisfy. Some legal systems clearly do not exist and thus do not currently impose valid legal norms on anyone—such as legal systems of a long-gone past and the legal systems of an imagined future, whether or not some conservatives might wish back the former and some revolutionaries wish for the latter. Moreover, it is difficult to see how one could hang on to the means principle without also hanging on to a version of the efficacy condition. Giving up (3) the direction-of-fit requirement would be unappealing for similar reasons as giving up on the efficacy condition, as this would again make the existence conditions of a legal system implausibly easy to satisfy.

III. The Gaps Between Posited Law and its Subjects

The standard picture of efficacy faces two related challenges: the epistemic and motivational gaps. Conjointly, these gaps between the law and its subjects in modern legal systems make it difficult to explain how an existing legal system can satisfy (1) the means principle, (2) the efficacy condition, and (3) the direction-of-fit requirement without denying the reality of either gap. Although these gaps have been frequently alluded to in the jurisprudential and socio-legal literature (at times even as commonplace), their extent is equally frequently downplayed in theoretical discussions of the efficacy of law. Moreover, outside the narrow jurisprudential and socio-legal context, the existence of the epistemic and motivational gaps between law and its subjects appears largely unacknowledged.

As is frequently recognized in passing by legal scholars and philosophers, law subjects usually have little and highly selective knowledge of the actual law of their jurisdiction. This fact has been widely acknowledged in the literature for at least a century¹¹ and has been observed in studies of employment, criminal, family, education and medical law in the United States (US), the United Kingdom, and Australia between the 1970s and 2010s.¹²

Two recurring findings in this context are worth emphasizing: first, empirical studies of legal knowledge repeatedly find that test subjects’ correct responses to questions assessing their legal knowledge are little (and sometimes no) better than mere chance; second, empirical studies of legal knowledge have also repeatedly found that test subjects tend to mistake their own moral judgment for the law.¹³ The fact that test subjects tend to mistake their own moral judgment for the law further complicates comparisons between jurisdictions, for it is difficult to estimate whether test subjects in some jurisdictions actually have better knowledge of the law or whether the law of the respective jurisdiction follows the law subjects’ common moral judgments more closely. When people do consult lawyers and find out about the applicable law, it is usually ex post (and not ex ante).¹⁴

As the existing empirical literature on law subjects’ legal knowledge from the 1970s onward shows, law subjects in paradigmatic municipal legal systems like the United Kingdom, Australia, and the United States clearly do not “by and large” know the law governing their private and professional relations within their respective jurisdictions. Denying the epistemic gap would require rejecting either the methodology or interpretations of data collected by virtually all researchers investigating legal knowledge over the last six decades. For obvious reasons, I do not seriously contemplate this possibility.

Apart from its empirical reality, the possibility of such an epistemic gap is part of the very idea of a system of posited norms. On the dominant view of law, and across traditional positivist and antipositivist divisions, law is a system of both primary and secondary rules. The most important of these secondary rules is the rule of recognition, which specifies what counts as law in a legal system. This rule of recognition is not itself validly created as law but itself specifies the criteria of validity. On the dominant view, this rule of recognition exists by being practiced by the most important actors of the legal system, such as judges, and perhaps in some cases by legislative and executive officials. However, except for a small legal elite, law subjects have only a rough idea of the specific content of the rule of recognition of their community. The very fact that law, as a union of primary and secondary rules of recognition, makes it possible to posit some rules as validly created also means that laws can necessarily be validly created without being acknowledged, understood, or even known by the wider population of law subjects.¹⁵ In fact, even jurists are rarely aware of the actual content of the substantive legal norms in every area of law, save their areas of professional specialization. And reasonably so. No one could possibly remember all individual laws governing all areas of life (from criminal, tort, and contract to family, housing, employment, and education law), and no one could possibly consume all new legal rules upon their promulgation.¹⁶

In addition to limited and selective knowledge of the law, law subjects are also rarely motivated to behave in conformity with the law by the law itself or its potential sanctions. Again, this motivational gap has been virtually commonplace both in legal theory and in the sociological study of law for a long time.¹⁷ At least in part, the motivational gap is a consequence of the epistemic gap. But the motivational gap also frequently applies when law subjects do know the law. This fact has been repeatedly acknowledged in the jurisprudential literature.¹⁸ In any moderately just legal system—indeed, in any legal system that is not wildly unjust—most people behave in conformity with legal rules because they are simply not inclined to do what the law incidentally proscribes or because they have independent moral reasons for their behavior. While the law and its potential sanctions might be required to make most people pay their taxes, vast swathes of, e.g., criminal, tort, contract, family, and medical law do not require their sanctions to achieve “by and large” conformity. Most people abstain from killing others, from taking others’ possessions, keep their promises, pay their debts, and materially support their children not because it is legally required but because it simply does not cross their mind to kill someone, because they think it immoral to damage others’ property, and because they are fulfilling their role-dependent social obligations, e.g., to their children or their customers.

Moreover, the motivational gap is not merely an accidental empirical fact, reflecting the currently predominant, motivating reasons of people living in specific jurisdictions. Like the epistemic gap, the motivational gap is also part of the very idea of a system of posited norms. On a plausible view of human nature as only somewhat vicious and not entirely evil, the motivational gap will likely be bigger the more reasonable the individual norms of the posited legal system, and the more reasonable a legal system, the more desirable it would be for the motivational gap to grow. Put programmatically: in the ideal society with an ideal legal system, much-posited law should not be motivating at all.

Denying the reality of the motivational gap would thus be difficult. For one, denying the motivational gap would imply denying the epistemic gap, too. And on any plausible view of human nature, most people conform to vast amounts of law in moderately just legal systems not out of either acceptance of the law or fear of punishment but out of independent motivating reasons.

IV. The Difficulty of Explaining Law’s Efficacy

The difficulty of explaining law’s efficacy lies in reconciling the three orthodox assumptions about law with the epistemic and motivational gaps. In order to exist, posited law must not merely match the accidentally conforming behavior of its purported subjects; to exist, posted law is supposed to effectively control the behavior of its subjects.

If law is a system of rules, one might expect that the efficacy of law is simply an application of the general notion of following a rule. To follow a rule is not merely to show a regularity of behavior—in H.L.A. Hart’s famous example, going to the movies each weekend can be a regular behavior but need not be a rule. By contrast, taking off one’s hat in church, for instance, is to follow a specific rule, namely the rule that one takes off one’s hat in church. Thus, one plausible way of identifying rule-following seems to be an “internal point of view,” i.e., a critically reflective attitude toward the rule. Such an internal point of view means accepting the rule as a reason for acting, criticizing others for failing to act in accordance with the rule and accepting corresponding criticism of one’s failure to comply with the rule as legitimate. However, it clearly is not the case that the general population adopts such an internal point of view vis-à-vis the substantive rules of a legal system. As the epistemic and motivational gaps show, the general population in modern legal systems is far too alienated from the substantive legal rules to adopt a critically reflective attitude toward individual legal norms.

Alternatively, one might have thought about substantive legal rules as demands issued by the sovereign at the threat of sanction, such that compliance with legal rules requires nothing like critically reflective acceptance of the rule but merely the wish to avoid sanctions.¹⁹ However, as the epistemic and motivational gaps make clear, the general population in modern legal systems is also too alienated from the substantive legal rules to treat them like commands. When people act in accordance with tort, criminal, or family law, they usually neither know the substantive rules of law nor do they act from a motivation of avoiding sanctions. Usually, law subjects act out of law-independent motives. (As mentioned above, empirical legal research has shown that even when people have explicit knowledge of tort law, these legal rules do not have an observable deterring effect.²⁰)

Could one then just relax the standard picture? In other words, could one reconcile the standard picture of efficacy with the epistemic and motivational gaps by simply dropping the widely presumed requirement that law be, by and large, efficacious? Unfortunately, one cannot. So far in our discussion, the difficulty in explaining the efficacy of law has been the difficulty of squaring a set of standard assumptions about the law with empirical facts about the alienation of law subjects from the law. But if we were to relax the efficacy requirement enough to make this squaring possible, we would end up with a more strictly theoretical problem: we would cease to have a way of drawing a distinction—not just in practice, but in theory—between an existing and an inexistent legal system. After all, a tempting way of distinguishing between existing legal systems (like US federal law), past legal systems (like Roman law), and borderline cases (like legal systems of fragile states during civil war) is by pointing to their respective efficacy—or lack thereof. And on a sufficiently relaxed version of the standard picture, the very possibility of drawing such a distinction is lost. In other words, the above-outlined alienation of law subjects from the law becomes theoretically (rather than merely empirically) interesting precisely because it does not create an efficacy crisis for existing legal systems: the appropriate philosophical response to the above-outlined alienation of law subjects from the law is not to conclude that there is no such thing as existing law, but to reconsider what it means for a system of posited norms to be “efficacious.”

And yet, the efficacy of law cannot be mere coincidental conformity of behavior either. Merely coinciding behavior would not constitute any influence at all. What kind of influence, then, must positive law have on its addressees in order to exist? In the following section, I discuss the most instructive historical attempt at explaining the efficacy of law.

V. Hans Kelsen’s “Normative Efficacy”

The development of Hans Kelsen’s view on law’s efficacy between the 1920s and 1960s demonstrates the difficulty of explaining the efficacy of posited law particularly well. Moreover, Kelsen’s changing views arguably provide the most serious theoretical treatment of the efficacy of law so far. Consequently, it is worth briefly but accurately outlining the development of his view here.

In the original (German) edition of the Allgemeine Staatslehre, Kelsen initially proposed an account of efficacy as motivation through coercion. In the context of discussing his view about the identity of state and law, Kelsen there claimed that the efficacy of law is best understood as “actual, psychological coercion” effected in the law subject by their mental representation of the norm that specifies a potential sanction.²¹ At this point, Kelsen still emphasized that the efficacy of a legal system is really the causal efficacy of the mental representations of norms (Normvorstellungen) of the law subjects. In the first edition of the Reine Rechtslehre of 1934, Kelsen repeated this claim about efficacy as motivation through coercion.²²

About a decade later, in his General Theory of Law and State (partially a translation of the earlier Allgemeine Staatslehre from 1925), Kelsen rejected his earlier idea of efficacy as motivation through coercion. While the General Theory of 1949 still holds on to the means principle²³ and the notion that law is a coercive order,²⁴ Kelsen there denied that a coercive order could be thought of as effective psychological compulsion. By contrast to his earlier discussion, Kelsen’s General Theory of 1949 recognizes that lawful behavior is by no means only, or even usually, motivated by fear of sanction.²⁵ Instead, Kelsen now claimed that a legal system is called efficacious simply if the law subjects’ conduct conforms (to some unspecifiable degree) with the legal norms and that no information at all is thereby implied about their actual motivation, and “in particular, about the ‘psychic compulsion’ emanating from the legal order.”²⁶ Kelsen still insisted that law is nevertheless adequately called “coercive,” for it is distinguished from other social techniques by regulating the application of violence.²⁷ Kelsen’s reason for this significant change of mind was his reading of Eugen Ehrlich’s sociological studies of law, which emphasized the motivational gap.²⁸

Thus, Kelsen started in 1925 by defending a view that endorsed the means principle, the efficacy condition, and the direction-of-fit requirement but ignored the motivational and epistemic gaps. In response to Eugen Ehrlich’s sociological studies of law, Kelsen abandoned this earlier view; by the middle of the twentieth century, his General Theory accepted the means principle, the efficacy condition, and the motivational gap but struggled to make sense of the direction-of-fit requirement.

In the second edition of the Pure Theory of 1960/1967, Kelsen finally gave a proper name to this struggle, albeit without solving it. There, he pointed out that the notion of efficacy encompasses paradigmatically two things: either (i) the application of legal norms by officials or, in more ideal cases, (ii) the conformity of the law subjects’ behavior with the law.²⁹ The former includes the application of a sanction in cases of nonconformity and thus captures a way in which a valid legal norm might said to be effective despite having been violated (and thus despite the law subject having been ignorant or unmotivated by the law)—however, such sanctioned nonconformity is but a small aspect of social life, compared to the combination of unsanctioned nonconformity and accidentally conforming behavior, and thus cannot be exhaustive of law’s efficacy. Importantly, Kelsen noted that conforming behavior could be induced by motives unrelated to the law; if the actual motive for conforming behavior is, e.g., a coinciding religious norm, Kelsen insisted it would be religion that was effective, not the law.³⁰

In the Pure Theory of 1960/1967, Kelsen, therefore, proposed a distinction between two meanings of “efficacy”: a causal meaning of efficacy, which would be the efficacy of a norm whose representation in the mind of the law subject causally influences their behavior (a requirement he abandoned in the General Theory of 1949 as a response to the motivational gap) and a normative meaning of efficacy. According to Kelsen in 1960, “normative efficacy” is the efficacy of a legal system in which the law subjects’ behavior, by and large, corresponds to the norms of a legal system even though their actual (causally effective) motives may not be the law.³¹

This distinction between causal and normative efficacy is no mere logomachy: Kelsen’s indication that legal theorists tend to use efficacy not merely in a causal sense but also a normative one seems perfectly on point. But of course, coining the term “normative efficacy” is not—and could not seriously have been meant to be—an actual solution to the puzzle about the efficacy of law. At best, we can now rephrase the efficacy problem in Kelsenian terminology. Instead of asking how we might reconcile the standard picture of efficacy with the epistemic and motivational gaps, we can ask: what is “normative efficacy”? Or, in less Kelsenian jargon, we can ask what kind of “influence” (as per the direction-of-fit requirement outlined above) is required for law to be effective. Kelsen’s late view of the 1960s holds on to the means principle and the efficacy condition while acknowledging the motivational gap. The notion of “normative efficacy” does not solve the puzzle of how a system of posited norms could be effective without satisfying the direction-of-fit requirement, nor does it explain how the direction-of-fit requirement might be satisfied in the presence of epistemic and motivational gaps. However, Kelsen’s notion of “normative efficacy” is possibly the closest that legal philosophy has come to an explicit acknowledgment that the efficacy problem arises from the very nature of law as a system of posited norms.

What kind of “influence,” then, is required for law to be efficacious? In the following section, I discuss three recent attempts at conceptualizing the influence of law on its subjects that promise to overcome the difficulty faced by the standard picture. Although each view discussed will turn out to be slightly more successful than the last, I argue that none of them fully succeeds in providing a convincing explanation of the efficacy of law.

VI. Three Recent Attempts at Explaining the Efficacy of Law

Recent literature in jurisprudence contains three accounts of the possible influence of law on its subjects that attempt to reconcile the standard assumptions about the efficacy of law with the epistemic and motivational gaps. While none of the respective authors explicitly phrase their discussion in terms of the three basic assumptions and two gaps discussed above, it is helpful to restate their respective accounts in these terms to get a better appreciation of the difficulty involved. As I argue below, none of them ultimately succeeds in reconciling the standard picture with the epistemic and motivational gaps; rather, each of them is forced to give up one or more of the assumptions about law that characterize the standard picture. However, the mere fact that they need to give up on some standard assumptions is not, itself, an inherent flaw of their respective views. As I have already suggested above, it seems impossible to fully reconcile all three standard assumptions with the epistemic and motivational gap.

A first attempt to solve the puzzle about the efficacy of law is to conceive of law’s efficacy in counterfactual terms. This counterfactual model has two core aspects: one regarding general obedience to law’s norms and one regarding the enforcement of laws. Regarding general obedience to legal norms, law may be said to be effective counterfactually if it could make a difference to the law subjects’ motivations if other motivating reasons were absent. That is, law would not be efficacious by actually influencing its addressees but would still count as efficacious as long as it would, theoretically, influence its addressees in the absence of other motivating reasons.³² This first aspect of the counterfactual model might overcome the motivational gap, but not yet the epistemic gap. So, if that were all, we would still be only halfway to a solution.³³

The second aspect of the counterfactual model speaks to the epistemic gap. In cases of ignorant, accidental conformity with the law, the counterfactual model claims that the mere possibility of punishment of nonconforming behavior (were it to occur) still counts toward the efficacy of law. That is, as long as nonconformity with the law could be punished, widespread ignorance of and merely accidental conformity with the law is no threat to law’s efficacy.³⁴ Thus, a crucial difference between this counterfactual model of influence and Kelsen’s view of 1960/1967 outlined above is that according to the counterfactual model, it is not merely the actual application of norms but also the potential application of norms by officials—specifically, the application of sanctions for violations of duty-imposing norms—that co-constitutes efficacy.

One difficulty of this approach is that it implies giving up the direction-of-fit requirement. If coincidentally conforming behavior counted toward the efficacy of law as long as nonconformity could be punished, a legal system could be said to be effective, although virtually no one knew its laws, no one was actually motivated by its laws, and the content of its laws were purposively designed to overlap with people’s actual behavior. Now, this is not to say that the counterfactual model could not be the correct view, but it is important to note that it also gives up the traditional direction-of-fit requirement.³⁵

The counterfactual model also achieves a vacuity reminiscent of Kelsen’s final treatment of law’s efficacy. The efficacy condition entails that legal norms do not exist as valid norms unless they belong to an effective legal system; according to the counterfactual model, a legal system is efficacious if its norms are enforceable, but the enforcement of legal norms—whether actual or potential—already presupposes an effective mechanism that is realistically capable of enforcing such norms. To have an effective mechanism realistically capable of enforcing norms is just to have an effective legal system. Indeed, standard examples of insufficiently effective (and thus inexistent or at least borderline) instances of law, such as fragile states during civil war, are characterized precisely by the absence of the realistic possibility of enforcing norms. As a result, proponents of the counterfactual model buy plausibility at the expense of explanatory power. In their view, a legal system is efficacious if its norms are enforceable,³⁶ and legal norms are enforceable if the legal system is efficacious.³⁷ Like Kelsen’s notion of “normative efficacy,” this observation seems true—but it is also hardly informative. At best, it is another way of restating the general puzzle about the efficacy of law. The only significant upshot of adopting a counterfactual model of law’s efficacy seems to be a new way of restating the original question: what, exactly, is the “efficacy” of a legal system such that it becomes possible to potentially enforce compliance from a population largely unmotivated and ignorant of its laws? And this was precisely the question a theory of law’s efficacy was supposed to answer.

A second attempt to solve the puzzle about the efficacy of law would be to insist that law influences its addressees not directly but indirectly. This attempt relies on two key concepts: mediation and congruence. Law might be effective, and the direction-of-fit requirement satisfied, if law’s influence on law subjects were mediated by “informal social customs, conventions and practices.” This mediated influence on the law subjects through informal social practices might work as long as the posited law was sufficiently congruent with a society’s social background practices.³⁸ Without some sufficient degree of congruence, a society’s informal customs could not fulfill their mediating role, i.e., help make the law indirectly intelligible to the law subjects and thereby indirectly guide their conduct.³⁹ On this view, then, law subjects naturally come to understand the law if (i) the individual legal norms are tailored to mediating schemata that law subjects easily understand, and (ii) these schemata are learned by participating in informal social customs.⁴⁰

If mediation were fully effective and it was really the case that law subjects come to “grasp in sensu diviso much of the law that pertains to them”⁴¹ simply by relying on the help of informal social practices, then we could deny the epistemic gap.⁴² Whether or not one finds this suggestion plausible, it is still only halfway to solving the efficacy puzzle, for even if law subjects really grasped much of the law, the motivational gap remains. And if the motivation for conformity with legal norms were determined largely by the informal social practices themselves, law would still not be effective—the informal social practices would be. Moreover, depending on how strictly one understands the requirement of congruence, one would also have to give up the direction-of-fit requirement, for it would seem as if law could only be effective by modeling itself onto preexisting, informal social practices, i.e., the actual general behavior of the law subjects.

However, the indirect model is arguably an improvement over the counterfactual model of efficacy. For one, it avoids the vacuity of both Kelsen’s notion of “normative efficacy” and the counterfactual model. For another, it recognizes that posited law cannot influence law subjects by directly guiding their conduct, factually or counterfactually. If there is a philosophically satisfying way to explain the influence of posited law that constitutes its “normative efficacy,” it must be an explanation of some real influence that is not directly action-guiding.

A third attempt to solve the puzzle about the efficacy of law is to reconceive the notion of efficacy in terms of interpretive behavior: instead of thinking about law’s efficacy in terms of its counterfactual or indirect governance of action, we might think of law’s efficacy as its governance of interpretation. On this view, a legal system would be effective not in virtue of compliance with its substantive norms or in terms of (actual or counterfactual) motivation but in virtue of providing a generally accepted scheme of interpretation for human interaction. On this view, it would be wrong to say that the law’s demands are effective by motivating its subjects through threats of (potential) enforcement. Instead, posited law would ultimately be effective by constituting a generally accepted scheme of interpretation for classifying some acts—most importantly, acts of violence—as lawful and others as unlawful. But whether or not lawful violence ever occurred and whether its possibility actually ever motivated anyone to act in conformity with the law’s substantive demands would, strictly speaking, be irrelevant.⁴³ On this interpretive model of efficacy, the influence required for law to be effective would be an influence on people’s by and large interpretive behavior rather than their by and large compliance with individual legal norms.

This interpretive model of efficacy would be incompatible with the traditional assumption that law is essentially a means of social control—at least, in the way this traditional assumption is usually expressed. Instead, the means principle would have to be significantly rephrased to state that law is a collective scheme of interpretation; only secondarily (or, in metaphysical terms, accidentally) could law then be a means of social control. In other words, law would not fulfill its function(s) by means of social control; rather, law would fulfill its function(s) by means of providing a collective scheme of interpretation.

So far, the interpretive model of influence seems to me to be the most promising attempt to explain the efficacy of law. However, besides giving up the standard version of the means principle, the interpretive model still faces another challenge: while it plausibly overcomes the challenge posed by the motivational gap, it does not yet answer the challenge of the epistemic gap. It is one thing to argue that law is effective by influencing the interpretive behavior of some acts (particularly acts of violence) as lawful and others as unlawful. However, it is another thing to argue that a substantive body of norms of which people have little and highly selective knowledge can actually constitute such a collective scheme of interpretation. In its (still rudimentary) form, as recently proposed by Christoph Kletzer, the interpretive model still leaves open how this interpretive notion of influence should be reconciled with the epistemic gap, such that the law’s scheme of interpretation can truly be said to be “by and large” effective among the law’s addressees.

VII. Conclusion

The standard picture of efficacy is characterized by three basic assumptions about law: that law is a direct means of social control, that a legal system must be effective in order to exist, and that law is effective by influencing its addressees. However, the standard picture of efficacy is also difficult to reconcile with the epistemic and motivational gaps between posited law and its subjects. The challenge of explaining the efficacy of law is to show how a system of rules can influence its subjects without requiring them to either know or be motivated by the substantive legal rules of the system: just what kind of influence must a legal system have in order to exist? As I have tried to highlight in this article, this is not a question about borderline cases of efficacy, such as international law, fragile states, or colonial legal transplants. It is a puzzle about the efficacy of law in paradigmatic municipal legal systems—among them the “western,” democratic and liberal states with longstanding and self-confident traditions of the rule of law.

So far, the few existing accounts of what this influence might be are unable to hold on to the standard picture of efficacy while also providing a satisfying explanation of what law’s efficacy amounts to. This lack of a satisfying account of law’s efficacy is not merely an unfortunate gap in the literature; taking the difficulty of explaining law’s efficacy seriously means that we also need to be open to revising some of our most fundamental assumptions about posited law—and, consequently, our views previously based on these assumptions.

If we revise our assumptions about posited law, then we might eventually be in a better position to see why the epistemic and motivational gaps do not threaten the efficacy of a system of posited law. What, exactly, our revisions should amount to, I leave to another occasion. Due to the obvious lack of space, my aim in this article has been preparatory only, namely, to provide a clear philosophical outline of the efficacy problem and the inadequacy of the few existing accounts. And if we have a better appreciation of the philosophical difficulty involved, we might be one step closer to an answer.