I. Levinson and the Size of our Preferences

Do people care about the law? Do they prefer law, in the slightly technical (or economist's) sense of treating law qua law as a preference? Many of the knottiest issues of legal compliance hinge on these questions because we cannot know how much (if any) coercion is necessary to secure some degree of compliance with law without knowing how much law's subjects treat compliance with law as a coercion-independent preference. This question is not only at the heart of various accounts of how law matters (if and when it matters at all), but is also at the center of the commentaries offered in this symposium. Attempting to join my responses to these commentaries into one integrated essay would not only tax my abilities of analysis and prose, but would run too great a risk of reducing the diversity and richness of the individual contributions of each of the commentators. Consequently I treat them individually, even while acknowledging that each of them can be understood to engage some aspect of the question of whether there is in fact a preference for law.

I begin with Daryl Levinson, in part because I find his framing of the question of legal compliance and legal effectiveness highly congenial and, indeed, it is Levinson's framing that has influenced my own framing of this collection of responses. Levinson describes the initial problem of legal compliance as being driven by the fact that “even the most morally motivated, other-regarding types may have little interest in legal rectitude” (2017, 28). Exactly. And although I agree with the conclusion that there is little popular interest in legal rectitude, it is important to recognize that this claim is, at bottom, an empirical and contingent one. Perhaps the Finns who stand obediently at “Don't Walk” signs when there is not a car or police officer in sight have more of an interest in legal rectitude than most Italians or East Coast Americans, for example (Schauer Reference Schauer2015, 3, 73), but in general both our intuitions and a fair amount of empirical data, some of it described in The Force of Law (Schauer Reference Schauer2015, 57–74), support the view that few people appear to have what Levinson (Reference Levinson2017, 28) also nicely describes as “intrinsic preferences for legal compliance.”

Although Levinson phrases the question in terms of “legal compliance” and “legal rectitude,” we might even broaden the claim. In doing so, we ask whether there is an intrinsic preference for law at all, as opposed to the particular substantive first-order goals that law may facilitate on particular occasions. Relatedly, is there an intrinsic preference for process or procedural or even institutional values in general, including, but not limited to, law? In a recent article, Joshua Fischman (Reference Fischman2014, S289) asks, sarcastically, why there are no organizations like the “Purposivist Council” or “Americans for Agency Deference,” suggesting that there are few political incentives for methodological positions apart from the outcomes that particular methodologies may produce at particular times. And if Fischman is correct, which he almost certainly is, it is because there are few popular preferences that would produce political incentives for methodology as such. And insofar as legal values can be put into the same basket of second-order considerations, Levinson's initial claim is that such second-order preferences, the Finns apart, are largely nonexistent.

Levinson, however, follows up this initial claim with another of even greater subtlety and interest. Recognizing that even coercive law rests ultimately on noncoercive coordination,Footnote ¹ Levinson is interested in the size of the “rule package” about which people bargain and coordinate. Is coordination about specific rules, whether legal or otherwise, or is the “unit of analysis” for the game-theoretic calculation a package of rules? And if the latter, might a coordination enterprise produce, even if not an intrinsic preference for legal compliance, at least a strategic willingness to comply with all the rules within a bargained-for package? If a specific rule of international law can attract noncoerced compliance for standard game-theoretic reasons, then could not the rule package we designate as “international law” do very much the same thing?

By placing on the agendas of jurisprudence, legal development, and legal institutional design the question of rule packages, and thus the question of the size of our institutional design decisions, Levinson has greatly advanced the inquiry, even though he, like myself, has no easy answers to the questions he raises. Still, it might be valuable to mention several implications of his concern with the size of the packages with which people bargain, cooperate, and coordinate to produce the foundations of a legal system.

First, what kind of inquiry is he positing? One possibility is that Levinson has raised questions that belong to the realm of cognitive science as much as anything else. There is nothing wrong with this, of course, but the point is only that questions about the size of the categories we can fathom and the size of the categories over which we can bargain are as much cognitive (Minda and Smith Reference Minda and David Smith2001; Sakamoto and Love Reference Sakamoto, Love, Alterman and Kirsch2013) as they are game theoretic. But whether game theoretic or cognitive, the question lingers about the relationship between abstract or general commitments and immediate necessities. Levinson gives the nice example of chess as an enterprise in which the commitments are either wholesale or nonexistent. When two people commit to playing chess, their commitment is to all the rules of chess—this is what I label as “wholesale”—and as soon as a player defects from any of the chess rules that together comprise the “rule package,” the two are no longer playing chess at all.Footnote ²

But is law plausibly of the same variety? In theory we can imagine a similar kind of multiparty commitment to law, where the category of law, like the category of chess, encompasses all the directives of the legal system. And if and when there exists such a commitment, will the defector who has an all-things-considered good reason for defecting on one occasion from the commitment to the full package of law be sanctioned just because he has defected from one component of the package? Will the participant in law who breaks one or a few laws be treated as a nonparticipant in the way that a chess player who repeatedly moved his rook diagonally would be treated?

In a different world, we can certainly imagine such a consequence. But in our world it appears, as a contingent empirical matter, that people are not willing to consider the rule package we designate as “law” in such a comprehensive way. Some of this might be a question of cognitive capacity or cognitive preference. In many domains, for example, the urge to particularize is strong (Schauer Reference Schauer2003), and the same common (but hardly universal) resistance to generalization may produce a resistance to the form of generalization that exists when we take “the law” as a largely undifferentiated unit.

Moreover, one of the reasons why the rules of chess are perceived as a unit is that following only some of the rules, or following the rules only some of the time, is not only simply not playing chess at all, but is also such that the other participants in the enterprise treat this rule-violating behavior as such. It is difficult to imagine saying of someone that she is a good chess player despite the fact that she occasionally moves her rooks diagonally.

In a different universe, law could be like this. We could treat following the law as an all or nothing proposition, such that the lawbreakers, even if they only broke some of the laws some of the time, would be as unwelcome in a community of laws as the occasional diagonal rook mover would be in the community of chess players. But this is not our world, nor has it ever been. Minor lawbreaking is widely accepted, not only for individuals but also for officials (Schauer Reference Schauer2010,Reference Schauer2012a; 2015, 85–92) and at the point at which we empower a community to treat some lawbreaking as de minimus we may have crossed the line away from thinking of the law as a largely undifferentiated unit. Nor is it clear that this is necessarily a bad thing. When Thoreau observed that it is “not desirable to cultivate a respect for the law, so much as for the right” (Thoreau Reference Thoreau1980, 228), he highlighted the way in which the law is necessarily imperfect. Sometimes, this imperfection is manifested in bad laws. At other times, as we have known at least as far back as Aristotle (Schauer Reference Schauer2003, 27–54), the imperfection arises as a consequence of the generality of law, and thus of the inability of general laws to achieve an optimal outcome in every instance, especially in the context of instances whose existence was not or could not have been anticipated at the time the law was adopted. And although different systems have different mechanisms for dealing with the imperfections of law,Footnote ³ it would be hard to imagine a system willing to treat its entire system of laws as an undifferentiated unit.

This is not to deny that some parts of some legal systems might be treated in this way. Self-contained domains of transactional law, such as securities law, might fit this mold. Nor is it to deny that some legal systems are more prone to treat “the law” as the relevant unit than others. That is the point, after all, in observing that some societies are more willing than others to treat any violation of law as a content-independent wrong. But whatever the cause, it seems unlikely—empirically and contingently even if not necessarily—that any legal system can survive on the basis of all the relevant participants engaging in a mutual uncoerced agreement to treat all its laws as binding. And as long as that is the case, then we can understand why all the legal systems we know, even if they rest on coercion-free agreement at their onset and at their foundations, have thought it necessary to rely on coercion to secure their continuance and their effectiveness.

II. Kar and the Role of Sanctions

Like Levinson, and like me (Schauer Reference Schauer2015, 75–85), Robin Kar (Reference Kar2017) agrees that some sort of sanction-independent cooperative behavior lies at the foundation of any legal system. The three of us agree as well that the tools of modern and not-so-modern game theory can help us understand how and why such cooperative behavior comes into existence (Ullman-Margalit Reference Ullman‐Margalit1977; Axelrod Reference Axelrod1981,Reference Axelrod1984; Ostrom Reference Ostrom2000) even when there are no sanctions threatened against the initial noncooperators. However, Kar diverges at this point, contending that the ordinary nonfoundational (or, more accurately, postfoundational) operation of a legal system is less driven by sanctions than I (and, seemingly, Levinson) believe, and that even beyond the foundational level, there is considerable compliance with law that is not motivated by sanctions. For Kar, such legal compliance is best explained by the existence of what he calls “evolutionary stability conditions,” conditions that may be supplemented or even initially driven by sanctions, but that are both widespread and sanction independent.

Kar's claim is, at its base, a psychological one. He maintains that there exists in most societies a “shared sense of legal obligation” that is a “psychological attitude” (Kar 41). In making this claim, Kar seemingly agrees with me that a sense of legal obligation is not the same as a sense of moral obligation, and he believes that the evolutionary stability conditions explain this psychological attitude of specifically legal obligation. But more curiously, he maintains that his claim about the existence of a shared sense of legal obligation is “empirically well grounded.” It would be nice if it were, but it is not clear from Kar's contribution just what it is that supplies this empirical grounding. The evolutionary explanation might well explain how the fact of the existence of such a sense of legal obligation came into being, but an explanation of why such an attitude might be desirable or even expected is not the same as saying that it in fact exists unless it could be established from evolutionary facts that no other alternative was possible, a claim that Kar seemingly does not make.

Thus, Kar's evidence for the existence of such an attitude appears to reside largely in his claim that law's existing use of sanctions would be “overkill” if the concern was only with outliers, but here he may be placing too much weight on just one dimension of law's coercive aspect. Given that consistency with (but not necessarily compliance with, or obedience to) the law is often motivated by various altruistic and other prosocial attitudes (Schauer Reference Schauer2015,48–52, 57–61), one function of legal sanctions is to motivate actual compliance by threatening the outliers. Such sanction-induced motivation can hardly be expected to identify the potential violators in advance, and thus what seems to Kar as overkill is simply a rational enforcement strategy. The fact that the sanction-backed laws against child molestation are necessary (we hope) for only a small segment of the population does not mean that it is overkill to threaten all of us with those sanctions.

Moreover, law also serves at times to motivate the behavior of those whose sense of prosocial obligation, especially when it conflicts with self-interest, is insufficient to produce compliance, and not just for outliers. What we know from the data on tax compliance and much else, some of which is summarized in The Force of Law (Schauer Reference Schauer2015, 61–74), is that there are many topics for which legal compliance, absent sanctions, is in fact quite low. If Kar were correct in concluding that a psychological sense of obligation were widespread, such instances of broad noncompliance would be more surprising that Kar takes them to be.

I have no quarrel with the idea that it would be very nice if people in fact had the sense of obligation that Kar assumes. However, there is a fair amount of evidence that in many societies, including much of the United Sates, such a sense of legal obligation qua legal obligation is rare. It would certainly be preferable for it to be otherwise, but the fact that there is a plausible evolutionary explanation for the existence of such a sense of obligation is not itself an empirical claim. And to the extent that such a sense of obligation to the law qua law seems rare, then the existing empirical state of affairs awaits an explanation more closely fitted to what it is attempting to explain.

III. Nadler and the Development of Group Norms

Janice Nadler (Nadler Reference Nadler2017) is entirely correct that people's behavior is often influenced by group norms, and by the desire of individuals to conform to those group norms. And I agree with her that on occasion law has a role to play in the development or reinforcement of those group norms. To give an example I described in The Force of Law (2015, 145–47), there is some indication that the norm requiring people to pick up after their dogs, a norm that in theory and sometimes in practice can exist and influence behavior even absent from and apart from the law, was reinforced and strengthened because of the law. This state of affairs may well have emerged as a consequence of multiple causal influences, but one of them is that the very existence of a law requiring such behavior appeared to have empowered people to scold those who violated the norm. Thus there may well have been more behavior-influencing scolding because of the law than there existed without the law.Footnote ⁴

That law can operate in this way is clear, and it is quite possible that the principal disagreement between Nadler and me is about whether we actually disagree. As the foregoing example indicates, neither here nor in The Force of Law do I deny the phenomenon she usefully stresses, and thus it is possible that our only disagreement is about the extent of this and related phenomena, a question that Nadler acknowledges is an empirical one, and as to which I agree, and a question, as Nadler properly stresses, that remains in need of a great deal more serious empirical investigation.

That said, however, there does remain a domain of interesting disagreement, and that is about whether there exists a norm of law. Nadler is right to urge us to focus more than we have in the past on the way in which law might have an effect on group processes, but much (albeit not all) of the way in which law can operate in this group-mediated way is parasitic on the existence of a norm of law itself. Of course, there are many norms in any society. Even absent the law, there are group norms that influence individual behavior. There is no need here to recapitulate the vast literature on the emergence and enforcement of social norms (see Schauer Reference Schauer2015, 223–28), but it should come as no surprise that there exist some social norms that are legally unconnected in the sense of not emerging from the law, not being enforced by the law, and sometimes not influencing the law. Such law-independent social norms would include the norm against spitting on the sidewalk, the norm mandating the sending of thank-you notes, the norm prohibiting eating until all have been served, and the former norm against women wearing white shoes before Memorial Day, as well as countless other norms of etiquette, manners, and much else.

But then the question arises about the relationship between the law and such social norms. Although it is true, as noted above, that the fact of law might on occasion empower people to criticize others for violating law-independent social norms, as in the dog poop example, and although law may sometimes serve the informational function that both Nadler and McAdams (Reference McAdams2015) stress, a central question—and a question that goes to the heart of law's power and law's interaction with social norms—is whether there is a norm of legality as such. There is indeed a norm against spitting on the sidewalk, just as there is a norm prescribing that we relinquish our subway seats to the elderly and the infirm, and just as there is a norm directing us to be kind to our aging parents. But is there a similar norm telling us to follow the law?

At least some of The Force of Law is focused on attempting to frame the question just stated in as clear and as empirically useful manner as possible.Footnote ⁵ So if we wanted to ask the question whether violating the law is bad in the same way that being mean to your elderly mother is bad, we would want to take formal sanctions out of the equation. If we are interested in the existence of a norm, the answer to the inquiry becomes noisy (in the economist's sense) if there are formal legal sanctions attached to violation. Moreover, we would also want to take the content of the law out of the equation. It is true that it is illegal to steal money from the Salvation Army collection bucket, but it would be wrong to do so, intrinsically and in the view of society, even if it were not illegal. So again it is useful, for purposes of clarity and experimental design, to attempt to separate the social or moral wrongness of an act from the illegality of that act. There are, of course, activities that are illegal but not wrong,Footnote ⁶ and the question posed here, and in The Force of Law, is whether a social norm exists such that the kinds of social attitudes that are associated with violating a social norm are also present for violating the law just because it is the law.

As some of the examples in The Force of Law suggest, and as some of the small amount of carefully designed experimental results suggest as well (Schauer Reference Schauer2015, 200–01), it is far from clear that there exists such a social norm. Both officials and citizens who violate the law for good ends are rarely criticized, and even in circumstances less catastrophic than the events that produced the Nuremberg trials, following the law for bad ends is often criticized. Thus, when we pry the fact of law apart from the content of the law, it appears as if law does quite a bit less work than is commonly supposed.

Moreover, it may be slightly too easy to assume that social norms have the effect they do without regard to sanctions. Criticism, gossip, social exclusion, and the like—shaming and blaming—are not punishments meted out by the state, but they nevertheless provide strong incentives for people otherwise disinclined to adapt their behavior to social norms. And although there may be other differences between the operation of social norms and the operation of law in a strict and narrow sense, the existence of social norms enforced by the full arsenal of social pressures cannot be considered a counterexample to the basic importance of sanctions in the enforcement and effective operation of society's social rules.

IV. Herzog's Existence Theorem

I confess. I am a lumper. Don Herzog (Reference Herzog2017) engagingly calls forth the venerable distinction between splitters and lumpers, the former concerned with seeing the differences among the superficially similar, and the latter focused on identifying the similarities among the superficially different. Herzog proudly rides under the splitter banner, part of a group whose rallying cry is often, “It's more complicated than that.”Footnote ⁷ We lumpers recognize that life is more complicated than we often assert, and that general propositions about anything can always be subdivided, refined, complexified, precisified, and qualified. However, we lumpers also believe that in doing so we may lose the virtues of seeing pervasive themes and connections, and may lose the intellectual advantages that analytic isolation—extracting one aspect from a messier whole—can at times bring. Indeed, the committed lumper is enamored of the scientific method, and thus of controlling for a large number of variables in order to see the effect of a single variable, even if that single variable may rarely or never exist totally alone in the actual world.

The lumper's manifesto out of the way, it is time to turn more carefully to Herzog's engaging and important comment. As befits his splitter credentials, Herzog recognizes, correctly, that people may have various attitudes to the law, and thus treat law's directives in different ways depending on the context and the situation of the addressee. He says, for example, that some people may treat the law as authoritative for reasons of its democratic provenance, and consequently will be inclined to obey the law qua law without regard to the fear of the sanctions that are the focus of The Force of Law.

That such people and possibilities exist is undeniable, and thus the only real difference between Herzog and me is about the relative proportions of sanction-induced and non-sanction-induced obedience to law in particular domains. In support of his view that there are some domains in which genuine obedience to lawFootnote ⁸ exists without the support of sanctions, Herzog provides a valuable and personal example of the fact that, under circumstances in which sanctions appeared highly unlikely, the University of Michigan Law School nevertheless obeyed a directive—Michigan's Proposition 2 amending the state constitution to prohibit even constitutionally permissible race-based affirmative actionFootnote ⁹ —that the law school believed to be morally and otherwise deficient.

The example is a nice one, if only because Herzog, unlike too many others, recognizes the value of trying to isolate obedience by examining an example in which the subject's all-things-other-than-the-law-considered judgment diverged from the content of the legal directive. But for the Michigan referendum, the Michigan Law SchoolFootnote ¹⁰ would happily have done what Grutter v. Bollinger (2003) held that it was permitted (but, of course, not required) to do. And thus Herzog has located a very good example of a law that required its subjects to set aside their law-independent preferences. And he has also done so in a setting in which, if he is to be believed,Footnote ¹¹ the possibility of formal legal sanctions was essentially nonexistent.

Of course, the Michigan law faculty, a group of which I was once proudly a member, is hardly representative. One might expect, for example, that legal academics would be disproportionately respectful of and influenced by the legitimacy of the law. Assuming that there are for some people the kinds of democratic legitimacy arguments for obedience to law qua law that Herzog explains, it should come as little surprise that legal academics at an elite law school would be especially attracted to and influenced by arguments for obedience emanating from the fundamental democratic legitimacy of the law.

Even if we put aside the characteristics of the elite law school, it is still the case that most of the Michigan Law School's decision makers on the question of obedience (Herzog apartFootnote ¹² ) were lawyers. As Jack Goldsmith (Reference Goldsmith2009) has documented in the context of the behavior of government lawyers with respect to national security issues, lawyers tend to think that law is important. And for lawyers, the law is their comparative advantage. If you have a hammer, every problem looks like a nail, as the old adage goes, and thus it should come as little surprise that a group of lawyers might be inclined to treat legal reasons as being more important than other reasons, and might be inclined to treat legal reasons as more important than nonlawyers treat legal reasons.

None of this is to challenge Herzog's example. It is to suggest, however, that the legal obedience practices of the Michigan Law School faculty and senior administration might not be especially representative of the legal obedience practices of the population at large, and thus might not be representative of the necessity of sanctions to influence obedience in many or most settings. For the splitter, of course, this is of little moment, for she gladly acknowledges that different incentives and goals and preferences may be at work in different segments of society. But for the lumper, the lessons of the Michigan Law School example about law itself, or even about most of law, may have only a more limited value.

V. Ellickson and the Role of Nonstate Law

Robert Ellickson's (Reference Ellickson2016) commentary here is both complimentary and complementary. As to the former, nothing need be said except “thank you,” but as to the latter, the question is how we should think about the complementary roles of the state and nonstate actors and forces in preserving order and enforcing social norms. Ellickson suggests that McAdams and I have overstated the role of the state, and accordingly understated the importance of nonstate organizations and nonstate aggregations of social pressure.

In one sense, Ellickson is plainly correct. The state indeed does loom large in my analysis, and in the approach of McAdams as well, and I acknowledge that it would have been better to take more account of the kinds of phenomena that Ellickson describes. But in another sense, Ellickson's worry is a product of what may be too narrow a conception of law in the first place. As I suggest in The Force of Law (2015, 161–63), a connection with the political state may indeed be one of the characteristics of law as it is commonly understood, however narrow and at times unproductive that common understanding may be. However, if law is a family resemblance concept (Hart [1961] Reference Hart, Bulloch, Raz and Green2012, 13–17; Schauer Reference Schauer2012b, Reference Schauer2015, 37–41), such that there is no property that is essential to law, then a connection with the political state may be among the indicia of law but still not essential to it.

What makes this methodological and definitional issue important is the existence of many nonstate organizations that share many of the characteristics of the state-connected legal system. Most corporations and associations, not only legal ones like the National Football League and the Daughters of the American Revolution but also illegal ones like the Mafia, have the union of primary and secondary rules that Hart ([1961] Reference Hart, Bulloch, Raz and Green2012, 79–99) believed was at the center of a legal system. They have rules of recognition telling the membership and the officials what the rules are (and are not); they have rules of adjudication to deal with disputes; and they have rules of change that enable such organizations to persist over time. Moreover, as Ellickson properly points out, such organizations have the ability to impose sanctions and to do many of the other things that we commonly associate with the state-connected legal system.

Ellickson wants us to appreciate that the force exerted by such organizations ought to be considered as part of the full array of behavior-influencing forces, and with this I have no disagreement whatsoever. Indeed, our only disagreement may be over whether to call such organizations “law.” Ellickson has long resisted this move (Ellickson Reference Ellickson1994), preferring to understand such organized arrays of rules and enforcement as nonlaw institutions of great social importance. But it is not clear why he resists just calling such institutions “law.” It is true that these institutions are typically are not connected with the political state, but once we recognize the degree of organizational and coercive force exerted not only by various informal but still structured institutions but also by more formal ones such as Apple, Royal Dutch Shell, Toyota, Harvard University, the United Nations, and Amnesty International, for example, it may be best simply to think of these institutions and organizations, in part, simply as law. In the looseness or even nonexistence of their formal connection with the political state these institutions and organizations may lack some of the attributes of the prototype of law, but one might say much the same thing about ineffective state legal systems (e.g., Somalia), or about state legal systems with very few secondary rules apart from the primary rule of absolute individual power (North Korea, perhaps).

One might suppose that whether or not to designate nonstate organizations of systematically structured rules and enforcement as law or not is merely a question of terminology, but it is not. At some level the question is about labels, but it is also about the tasks we entrust to some kinds of organizations but not others, about the people who perform certain tasks and not others, and about the subjects we think appropriate for some organizations but not others. It is hardly a coincidence that lawyers are asked to serve as the proceduralists in a wide variety of nonstate organizations, or that those with legal training are frequently thought most suitable to the drafting of nonstate organizational rules.

But although all law is not force, neither is all of force law. Ellickson is correct to point out that much force is administered outside of anything we might remotely think of as law, and his Pulp Fiction example makes the point very well. But although caring for the injured is often the job of family members, friends, police officers, scout leaders, passersby, and the clerk at the 7-11 who dispenses nonprescription painkillers, it is still useful to recognize the special province of medical care and the medical establishment (including not only physicians, but also nurses, EMTs, pharmacists, etc.). So, too, might it be useful to recognize the special province of law, including nonstate law, even though the application of force and the enforcement of social norms is often in the hands of other social actors and mechanisms. Ellickson is on sound footing in pointing out that state law has no monopoly on force and violence, even legitimate force and violence, but just as it is important to see that the boundaries between law and other social institutions is permeable, so, too, is it important to recognize that permeable boundaries are still boundaries, and still serve the separate, even if imperfectly, social institutions with different functions and different characteristics.

Hadfield on Law's Functions and Definition

Like Ellickson, Gillian Hadfield (Reference Hadfield2016) also worries about the comparative importance of the state in the imposition of penalties and other sanctions. But although Ellickson is focused substantially on nonstate enforcement of nonstate rules, Hadfield's primary attention is on nonstate enforcement of rules that nevertheless emanate from the political state. As she puts it, “there is no shortage of examples where the state's role has been limited to providing a clear statement of what the rules are, leaving enforcement to private means” (2017, 20).

I agree that private enforcement is often important and, of course, a full theory of social coercion would need to take far more of it into account than does a book on law. Drawing this distinction, of course, suggests that there is a difference between the two, a distinction I happily embrace. Hadfield herself appears to acknowledge the distinction in her focus on the private enforcement of state law as opposed to the private enforcement of morality, etiquette, manners, fashion, and many other topics as to which there are social norms with private enforcement.

Even if we focus on the private enforcement of law more narrowly understood, however, we still need to be attentive to the distinction between enforcing the law because it is the law and enforcing a host of norms that happen to be embodied in law, and that are enforced by the formal state legal apparatus, but that exist, are often followed, and are often enforced outside of the formal state legal apparatus. Consider, for example, the norm against making misleading representations in the purchase and sale of securities. Violation of that norm is also, in most cases, a violation of the law, but it is also the case that many violators are punished outside of the formal state law by virtue of exclusion from certain markets, and by the unwillingness of other market participants to trade with them. Given this widespread phenomenon, we must now ask whether this latter form of punishment is punishment for violating the law or punishment for violating a law-independent norm that happens also to be embodied in the law. The distinction is important because it goes to the question, one discussed above in my response to Daryl Levinson, especially, and at great length in The Force of Law. Once we recognize that the norms of formal state law are often perceived by their subjects to be mistaken, either in general or as applied on a particular occasion, the question then arises whether the “mere” fact of law, separated from its content, can be effective without sanctions, and, if so, how, when, how often, and for whom. Law qua law can, of course, be enforced by nonstate actors, as is obvious to anyone who has been scolded by his or her fellow citizens for crossing the street against a “Don't Walk” sign in some parts of the world, even when no police or dangers are apparent.Footnote ¹³ However, much of the data at our disposal suggests that such behavior is rare, and that nonstate enforcement of state-created norms is most often the enforcement of the content of the norm and not the enforcement of law as such. As long as this is the case, then it may be that it is a mistake to fail to distinguish content-independent of the law, even state law, from the content-dependent enforcement of norms that have also been made illegal. If we fail to attend to this distinction, we may find that we are overestimating the prevalence of private enforcement of law, which is as much of a mistake as underestimating it.