The Authority of Law


Scott Hershovitz

A 30-ton statue looms over the steps that lead to the United States Supreme Court. A seated figure holds a tablet, and just behind that, a sword. The statue is called Authority of Law, but it is poorly named. It depicts a view of law’s authority most closely associated with the nineteenth-century legal philosopher John Austin (1998: 14—18), who conceived of laws as orders backed by threats. Today, nearly no one holds Austin’s view. To be sure, some laws are orders backed by threats, but that is not what is distinctive about law, and it is not what law’s authority consists in.

Contrast a gunman who commands “your money or your life” with a police officer who shouts “stop or I’ll shoot.” Both issue orders backed by threats. But by virtue of her station the police officer’s order carries an implication the gunman’s does not. The police officer claims the right to demand that her subject stop, not just the power to make him do it. If the police officer has that right—she might not—her order obligates her subject to do as she says. Not so the gunman: his order does not even purport to obligate. He claims might, not right (Wolff 1998: 4).

Austin’s problem was that he understood law on the model of the gunman, rather than on the model of the police officer. And that is the problem with the statue at the Supreme Court too. It depicts law’s power. But it does not capture the element that distinguishes law from mere thuggery. It does not capture the fact that legal officials claim the right to issue orders, not just the power to enforce them. Of course, legal officials may not have authority to tell others what to do. They may just be thugs, issuing orders backed by threats. But they claim to be more. This chapter considers the nature of that claim and what we might learn about law from it.

The Paradox of the Just Law

Most philosophers hold that authority (of the practical sort) consists in a right to rule, such that subjects are obligated to obey. However, they disagree about whether the law ever has authority in this sense. Nearly no one thinks that law always has the authority it claims. For instance, almost no one thinks that people in unjust states have an obligation to obey the law, at least not a general one. But a longstanding puzzle known as the paradox of the just law aims to establish that there is no such obligation even in reasonably just states, on the ground that a just law could not have a claim on our obedience.

We can use laws that prohibit murder to illustrate the paradox. Murder is wrong quite apart from what the law has to say about it. Moreover, we expect people to refrain from murder because murder is wrong, not because the law prohibits it, and certainly not because the law punishes it. Indeed, it would be perverse to avoid murder on legal rather than moral grounds. This raises the possibility that we do not have an obligation to obey laws that prohibit murder. Perhaps stranger, the reason that we do not have that obligation is that we should not obey laws that prohibit murder. Joseph Raz explains the line of reasoning that leads to that conclusion as follows:

The more just and valuable the law is … the more reason one has to conform to it, and the less to obey it. Since it is just, those considerations which establish its justice should be one’s reasons for conforming with it, i.e., for acting as it requires. But in acting for these reasons one would not be obeying the law, one would not be conforming because that is what the law requires.

(1994: 343–44)

If this argument establishes that we should not obey laws that are just, it also establishes that we do not have a general obligation to obey the law even in a reasonably just state. However, the first conclusion is the one of interest here, both because it seems paradoxical and because it poses a serious challenge to law’s authority.

Raz (1994: 343) argues that the paradox is “merely apparent.” He is willing to accept the conclusion that we do not have an obligation to obey laws that prohibit murder, but he says the paradox is overstated because it does not apply to all just laws. He observes that a legal prohibition may be just even though there is no independent moral prohibition on the conduct proscribed. Consider a city ordinance that restricts parking at certain hours so that streets may be cleaned. If it is important to clean the streets, and it is necessary to coordinate people’s behavior to allow cleaning crews access, such a law may be just even though there would be no reason to avoid parking at the prohibited times in absence of the scheme. Here, we cannot say that one ought to avoid parking during restricted hours for the reason that parking at those times is wrong independent of the law. By hypothesis it is not. Yet, the paradox is still present here, at least in mild form, so long as the law is in fact just. Imagine a person who pronounces that, although she understands that she must cooperate if the parking restrictions are to succeed in coordinating behavior, she will avoid parking during restricted hours only because the law requires her to do so, not because doing so will aid the scheme. We would not think this person perverse, but we might think her obtuse. She implies that clean streets have no value, or that she has no role in securing them. But she is wrong, as those are the very considerations we are supposing justify the law. Here again it seems that when faced with a just law, one should act on the reasons that justify it, not simply obey it.

Now it should be obvious that the paradox only gets off the ground if acting on the reasons that justify a law is not a way of obeying it. So we must get clear about just what obedience to the law involves. Don Regan explains the standard view as follows:

I cannot be said to obey a law in the fullest (and philosophically interesting) sense unless I do what I have been told to do because I have been told to do it. Consider an example. If a governmental authority commands me to avoid the use of cocaine, and if I subsequently avoid the use of cocaine, but for reasons which make no reference to the authority’s command (I might even be a philosophical anarchist and regard the command, in itself, as a reason in favor of using cocaine), then I do not obey the authority in any interesting sense. My behavior [conforms to the law], but I do not “obey.”

(1990: 15)

It will help here to borrow a distinction from Raz (1999: 178–79), and stipulate that one conforms to an order if one does what it requires, whereas one complies with an order if one both conforms to it and takes the order as one’s reason for action. Regan, then, tells us that in the fullest sense one obeys an order if, and only if, one complies with it.

If obedience to law means compliance with it—doing what the law requires and taking the law as one’s reason for action—then there is little to be said in favor of obeying just laws. A morally decent person would no doubt conform her behavior to law on many occasions, but she would rarely if ever obey it. She would certainly not obey the core of the criminal law, and again that seems paradoxical, because one would think that if any laws are deserving of our obedience, it is those that prohibit grave offenses, like murder, rape and mayhem.

We seem to be stuck, but the way out of the paradox is as simple as the way in. The paradox only arises if acting for the reasons that justify the law is not a way of obeying it. The standard view identifies obedience with compliance, so that acting for the reasons that justify a law is not in fact a way of obeying it. But suppose we reject the standard view and identify obedience with conformity—with simply doing as the law requires whatever one’s reasons. Then the paradox of the just law disappears. That murder is wrong is a reason not to murder and a reason to obey the legal prohibition on murder, for obedience now involves nothing more than not murdering. Indeed, the paradox is flipped on its head—the more justified a law is, the more justified we are in obeying it, which is a much happier state of affairs.

Now we face a choice. There are two ways to understand what obedience involves, one which leads to a paradox and one which does not. That alone might seem reason for preferring the second, but it is a weak reason at best. If the paradox of the just law is genuine, it is what Quine (1976: 1) called a veridical paradox—a “conclusion that at first sounds absurd but that has an argument to sustain it.” It is not the kind of paradox that involves self-contradiction and therefore cannot be accepted. We should be open to the possibility that just laws have no claim on our obedience, and we should not define obedience down just to conclude otherwise. If we are to reject the paradox on the ground that the conception of obedience that underlies it is too demanding, we must have good reason for preferring a weaker one. Fortunately, we do.

Let us start with an observation: in the normal case one does all that the law requires if one acts as it demands, whatever one’s reasons. The law certainly does not regard your behavior as defective if you refrain from murdering for the reason that murder is wrong. No prosecutor will investigate. No charges will be brought. No judge will take you to task for failing to pay proper heed to the law. We would have indeed a puzzle if the law prohibiting murder required that we have the prohibition as our reason for not murdering. Then it would be perverse to discharge the obligation the law purports to impose. But the law does not put us to the choice of failing to conform to its prohibition on murder or acting perversely. It asks only that we not murder, and it leaves the reasons up to us.

This is typical, but there are deviant cases. Sometimes an authority requires that one act as ordered and have the demand as one’s reason for action. Military orders may have this structure. A sergeant may be keen to have a private do as she says, and for the reason that she said so. She may be justified in making such a demand if the lives of the soldiers in the patrol depend on the private following orders without taking time for reflection. In that situation, it would not be at all perverse for the private to do as the sergeant demands even though doing as the sergeant demands requires taking her demand as the reason for so doing. But situations like this are unusual. Most authoritative orders do not even implicitly specify the reasons for which one must act.

Now we can see the problem with the standard view. If we interpret an authority’s right to obedience as a right to compliance, every authoritative order speaks to reasons as well as actions. The result is that many exercises of authority carry a demand which the authority itself has not seen fit to make. But that is bizarre. If the law does not care why people refrain from murder so long as they do, why should we insist that people subject to the law’s authority are under an obligation to refrain from murdering for the reason that the law prohibits it? The concept of authority should not demand more than authorities do.

If we identify obedience with conformity, authorities are in the driver’s seat. In the rare instance that they want compliance (and are justified in wanting it), they can leverage their right to conformity into a right to compliance. If an authority issues an order that requires subjects to take the order as their reason for action, anyone who conforms with the order will necessarily comply with the part of the order that specifies the action required. Suppose, for example, that a sergeant orders a private to clean a latrine and to do so for the reason that she said so. If the private is under a duty to conform to this order, then he is under a duty to comply with the part of the order that requires him to clean the latrines, as conforming to the entire order requires complying with that part of it. In the far more common case where an authority does not care why its subjects conform so long as they do, an authority can do just what it intends—issue an order that speaks to actions, not reasons.