Introduction
It happens in the actual world, in philosophical examples, and in zombie apocalypse films: a group faces a dire problem, a problem too big for any single person to handle by herself. Fortunately, someone has the will and vision to lead. She seizes control and begins giving orders.
The group members have a moral obligation to obey—their leader is telling them to do the right things, after all. But what must the group members do to establish her as a de jure authority? Nothing—according to a recent wave of non-voluntarists, at least. These non-voluntarists advance views with two components, one empirical and one normative. On the empirical side, they argue that the state, or the law, is best (or uniquely) capable of accomplishing something important. On the normative side, they defend a substantive principle on which being so situated is sufficient for de jure authority.
However, widely accepted tenets undermine all such views. As even the leading non-voluntarists agree, authority is a controversial normative status. We should not invoke it to explain our duties when there is an equally plausible, authority-free alternative.
1. Setting the Stage
What is de jure authority? There are two main analyses, but they are importantly similar: to be an authority is to have a normative power, or a normative status, that all non-authorities lack. According to the first analysis, authority is the normative power to issue directives that ground an obligation for subjects to obey. Footnote 1 More formally, if A has authority over S, and A tells S to ϕ, the first analysis implies two things: S has a pro tanto duty to ϕ; and A’s instructions are the normative ground of S’s pro tanto duty.
When philosophical anarchists deny authority—of the state, of the law, or of an individual person—they often have this conception of authority in mind. Footnote 2 Consider Robert Paul Wolff’s well-known example: a ship is sinking, a captain with the will and vision to lead organizes a rescue operation, and the captain orders the passengers to man the lifeboats.
[Perhaps] … I had better do what [the captain] says, since the confusion caused by disobeying … would be generally harmful. But … I would make the same decision, for exactly the same reasons, if one of the passengers had started to issue “orders” and had, in the confusion, come to be obeyed. Footnote 3
Wolff concedes that he might have a moral obligation to obey the captain, but insists that the explanation of that fact doesn’t invoke any special normative powers on the captain’s behalf. The situation calls for swift, coordinated action, and Wolff must obey just because the captain is effectively organizing things. According to Wolff, of course, citizens are in a similar position vis-à-vis the state: citizens will sometimes have a moral obligation to obey, but not because the state has a special normative power to ground such obligations. Something else, such as an urgent need for coordinated action, must do that normative work instead.
The second analysis of authority emerged, in part, as a response to philosophical anarchism. Roughly, it holds that de jure authority is the right to enforce or administrate the law, on some formulations, without interference. Footnote 4 This analysis agrees that the captain’s directive (‘Man the lifeboats!’) does not itself ground an obligation for Wolff to obey. Rather, the captain’s de jure authority entitles him to force Wolff’s compliance, should Wolff begin to disobey. Still, this analysis holds that there is something normatively special about authorities. Authorities have, and non-authorities lack, a special normative status that entitles them to enforce laws, or to administrate without interference.
I will remain agnostic between these two analyses, insisting only that de jure authority is a special normative power or status. However, even this minimal insistence has two significant consequences. First, the primary disagreement between statists and philosophical anarchists concerns moral ontology. The main question is not whether it is important for citizens to regard their state as an authority; the main question is whether governments that meet certain conditions (laid out by theories of consent, fair play, and the like) actually have authority. Second, we should not posit authority unnecessarily. On both analyses, authority is a controversial addition to our moral ontology; we should not multiply such things if there is no need. These widely accepted points are the foundation for a test.
2. An Occam-Inspired Test for De Jure Authority
Given that authority is a normative status, we should look to normative space for evidence that it exists. Given that we should not posit authority unnecessarily, we should affirm authority’s existence only if we find an obligation or permission such that (a) authority plausibly explains it, and (b) there is no equally plausible authority-free explanation.
Imagine, for example, that you are part of a crowd facing a dire emergency, and that all must act in concert to survive. You shout orders that, if followed, would institute one of several adequate solutions to this coordination problem—but the panicky din drowns out your voice. Luckily, George ‘Foghorn’ Wilson happens to be in the crowd. His booming bass voice means that he, and only he, can make himself heard. He shouts out a different, but also adequate, set of orders. Everyone can hear, and given what’s at stake, everyone has an obligation to obey. Footnote 5
There are two ways to explain this obligation. First, Foghorn might be an authority—his special normative powers are the reason the crowd has an obligation to obey his orders rather than yours. Second, there is nothing normatively special about Foghorn, but the crowd should obey because Foghorn’s orders will get the job done, and because he alone is sufficiently loud. The test implies that, of these two explanations, we should opt for the second. The volume-based explanation is just as plausible as the authority-based explanation, and it doesn’t rely upon controversial normative posits.
Before pressing on, consider a case in which authority does not simply ‘drop out’: each citizen of Consentia expressly consents to obey the government’s laws. Footnote 6 The government makes reckless driving illegal, and, of course, the citizens have an obligation to avoid reckless driving.
It might be tempting to think that authority is again otiose—reckless driving is primarily wrong because it endangers lives and limbs. But, in this case, the authority-free explanation misses something. Think about it from the standpoint of disobedience: reckless driving is wrong because it risks lives and limbs, but Consentians who drive recklessly also fail to give the law proper normative uptake. A full catalogue of their wrong will include the fact that they made the law authoritative, and then promptly ignored the law proscribing reckless driving. So, whereas authority is a needless posit in the Foghorn case, there is no equally plausible authority-free explanation of Consentians’ moral obligations; the authority-free explanation misses something.
None of this is novel or groundbreaking—no one thinks it is a good idea to posit controversial normative powers when there is an equally plausible, authority-free way to explain our obligations. But this point of wide agreement is fatal to a recent, high-profile wave of non-voluntarist theorizing.
3. Non-Voluntarist Defences of Political Authority
The non-voluntarist theories we will target are a diverse lot, but they share a two-fold structure. They contain one empirical argument and one normative argument.
The empirical argument contends that the state, or the law, is best (or uniquely) capable of achieving an important end. Of course, the fact that a person or an organization is the only agent that can achieve an end does not automatically entail its authority. There is a gap between the capacity to Φ and the authority to Φ, even when Φing is important. The normative argument tries to close that gap by defending a substantive normative principle on which agents that are best (or uniquely) situated to accomplish an important task have de jure authority. Footnote 7
Which theories exhibit this sort of empirical-normative structure? Christopher Wellman argues, on the empirical side, that states are necessary to save us from the state of nature; his theory of Samaritanism is supposed to explain how that makes states authoritative. David Estlund follows Wellman in arguing that the state is uniquely situated to stave off the state of nature; his theory of normative consent is supposed to explain why, given this empirical conclusion, a citizen’s non-consent establishes her government as an authority just as surely as her consent would. Joseph Raz is vulnerable too—at least arguably. For Raz, the empirical argument is that many of us will better conform to our authority-independent reasons (Raz calls them ‘dependent reasons’) by obeying an authority than by trying to work things out ourselves; his Normal Justification Thesis (NJT) is a substantive normative principle on which such command-givers are authorities. Footnote 8 And there are others. Footnote 9
We will examine Wellman’s political Samaritanism, turn next to Estlund’s theory of normative consent, and conclude by generalizing our results so that they apply to all non-voluntarist theories with this empirical-normative structure.
3.1 Wellman’s Political Samaritanism
Wellman’s empirical argument is that the state of nature is a place of dire peril, and that only a state can rescue us. Footnote 10 His normative argument deploys Samaritanism in two ways. One deployment defends the state’s right of enforcement (the second construal of de jure authority). The other defends the state’s power to issue directives that ground an obligation for subjects to obey (the first construal of de jure authority).
Wellman’s defence of the state’s right of enforcement rests upon a plausible principle.
Samaritan Coercion: A is morally permitted to coerce B if (i) coercing B is the only way to provide highly important benefits to others, and (ii) A’s coercion of B is not unreasonably costly to B. Footnote 11
Granting that the state of nature is a place of dire peril, Samaritan Coercion implies that the state may force individual citizens to obey for the sake of others, provided that this coercion is not unreasonably costly. As Wellman puts it, “my state may justifiably coerce me only because this coercion is a necessary and not unreasonably burdensome means of securing crucial benefits for others.” Footnote 12 Why should we think that the state’s coercion is not unreasonably burdensome? Wellman suggests that avoiding the state of nature, a system of criminal justice, access to roads, and other such benefits far outweigh the cost of governmental coercion—“it is as if the state forced each of its constituents to give up a hundred dollar bill but in return gave back ten twenties.” Footnote 13
Grant that states have the right to coerce, subject to all Wellman’s conditions. What explains that fact? One explanation is that there is something normatively special about the state—the state has authority, a special, controversial normative status that all non-authorities lack. The problem is that there is an equally plausible authority-free explanation as well. Somewhat ironically, the explanation is Wellman’s own: there is nothing normatively special about the state, the principle of Samaritan Coercion explains its right to coerce. Samaritan Coercion, that is, obviates the need for de jure authority altogether.
Samaritan Coercion is, on Wellman’s view, the permission-granting cousin of a natural duty: whereas a natural duty requires everyone to do (or to refrain from doing) something, what we might call a ‘natural permission’ entitles anyone in the right position to do something—in this case, to coerce. Footnote 14 We don’t need authority to explain why you must obey when the random stranger tells you to abide by your natural duties (‘Don’t Murder!’); we don’t need authority to explain why someone has the right to act in accordance with a natural permission either. One of Wellman’s own examples illustrates the point. In order to save Amy’s life, Beth may non-consensually ‘borrow’ Cathy’s car. Because Amy’s life is at stake, and because being temporarily separated from her car is (let’s suppose) a small burden for Cathy to bear, Beth’s actions are morally permissible. Footnote 15 Just as it fully explains why Beth has the right to coerce, Samaritan Coercion fully explains why the state may coerce. Authority drops out as a needless posit.
What about the second deployment of Samaritanism? To argue that governmental directives are the normative ground of citizens’ duty to obey, Wellman pairs Samaritanism with fairness. Footnote 16 Rescuing people from the state of nature is a Samaritan task that falls upon us all; fairness dictates that we all do our part in the rescue. A citizen’s fair share of the rescue, Wellman argues, is obeying the law. Footnote 17
Others have contested the steps in this argument. Footnote 18 Here, let’s grant that citizens have a duty to do their fair share to rescue others from the state of nature, and that obeying the law constitutes a citizen’s fair share. Still, what explains our duty to obey? Does an adequate explanation invoke special normative powers on the state’s or the law’s behalf?
To answer, we must distinguish two kinds of cases: cases in which there are good authority-independent reasons to obey the law, and cases in which there are not. There are many reasons that, on certain token occasions, it might be important for an average citizen (call him Joe) to obey the law. Perhaps the law mirrors Joe’s natural duties. Or, perhaps, Joe’s disobedience would somehow diminish the state’s capacity to rescue people from the state of nature. These are excellent reasons to obey, but they are authority-independent and make authority unnecessary.
But what about cases in which Joe’s obedience would have no authority-independent point? Initially, at least, it seems like Wellman should concede that Joe is permitted to disobey: the raison d’être of the state is to prevent the state of nature, so it is odd to claim that Joe must obey laws that don’t serve that (or any other authority-independent) end. This concession, however, would be admitting defeat. After all, the anarchist’s position is that Joe has an obligation to obey the law when, and only when, there are good authority-independent reasons to do so.
Oddly, Wellman never discusses legal obedience when there is no authority-independent reason to do so. But perhaps we can cobble together a response on his behalf. To explain why citizens may not do their fair share of rescuing others from the state of nature by contributing to Oxfam, or to a foreign government, Wellman points out that “(1) political instability creates a coordination problem and (2) discretion is a good.” Footnote 19 Let us grant that achieving political stability is a coordination problem: we can avoid the state of nature only if a critical mass of citizens obeys the law. Of course, that does not explain why Joe must obey. If a critical mass does its part, Joe’s disobedience is unlikely to make much difference. So, sensibly, Wellman turns to the idea that discretion is a good: everyone would like the freedom to choose between, say, supporting Oxfam or supporting their government; Joe can choose Oxfam without hastening the state of nature only because a critical mass of his compatriots does not. Thus, by supporting Oxfam, Joe has “helped [himself] to more than [his] fair share of discretion.” Footnote 20
Perhaps Wellman might say the same thing about Joe’s disobedience of laws that serve no authority-independent point. Everyone would like the freedom to pick and choose which laws to follow; Joe can pick and choose only because his compatriots don’t. So, perhaps, by disobeying a law—even when there is no authority-independent reason for him to obey—Joe helps himself to more than his fair share of discretion.
The problem with this argument, should Wellman appeal to it, is that we are ex hypothesi concerned with cases in which there is no authority-independent reason to obey. Widespread (even universal) disobedience in such cases would not hasten the state of nature. Footnote 21 Thus, by disobeying, Joe is not helping himself to anything that others cannot have in equal measure. He is not presuming the freedom to pick and choose which laws to follow; he is presuming the freedom to disobey when nothing at all hangs on his obedience.
Now, it is perhaps likely that Joe will sometimes err, disobeying when there actually is a good authority-independent reason to obey. But this, at best, is a reason for Joe to adopt legal obedience as a decision-making heuristic. And the fact that legal obedience is a good heuristic is not, on its own, a reason to think that Joe actually has an obligation to obey when there are no good authority-independent reasons for him to do so. As J.C.C. Smart said in a structurally similar debate, the fact that a nautical almanac (another heuristic) provides good guidance 99 percent of the time is no reason to think that one should do as it directs on the rare occasions when it errs. Footnote 22
To sum up: Wellman justifies state coercion by citing the principle of Samaritan Coercion. But that principle is a ‘natural permission’ and renders de jure authority superfluous. He defends the duty to obey the law by combining Samaritanism with fairness, but his account succumbs to a dilemma: when there are good authority-independent reasons to obey, we should explain the duty to obey by citing those reasons, not by invoking a special normative status on behalf of the law. When there are no such reasons, denying the duty to obey is just as plausible as affirming it.
3.2 Estlund’s Theory of Normative Consent
Like Wellman, Estlund takes the state of nature as the starting point for his empirical argument. Unlike Wellman, Estlund relies on a small-scale replica rather than dealing with the state of nature directly: a plane crashes and most of the passengers suffer terrible injuries. Luckily, the flight attendant is unscathed and begins to mount a rescue operation. The task is too big for her alone, so she turns to one of the few uninjured passengers (a man named Joe) and says, “You! I need you to do as I say!” Footnote 23
Joe’s position vis-à-vis the flight attendant is a model for a citizen’s position vis-à-vis her government. Both confront a morally important task: Joe must save the passengers; a citizen must rescue her compatriots from the state of nature, where, echoing Thomas Hobbes, Estlund insists that life would be solitary, poor, nasty, brutish, and short. Neither can complete their task alone: Joe must work with his fellow passengers to save the injured; a citizen must cooperate with her compatriots. Finally, both situations call for a competent decision-making agent: in Joe’s case, the flight attendant is apt for this role; in the citizen’s case, the government is supposed to be the obvious choice. Given these similarities, Estlund hopes that establishing the flight attendant’s authority over Joe will suffice to establish a good democratic government’s authority over citizens.
The normative half of Estlund’s view relies not on Samaritanism, but on a theory of normative consent. It is widely accepted that consent can be null—if someone promises you her life savings because you’ve conned her, you’re not entitled to a dime. Estlund’s novel idea is that non-consent can also be null. Footnote 24 When it is wrong to withhold one’s consent, according to Estlund, the normative situation is as if one had given valid consent. More pithily: ‘null non-consent = valid consent.’
So: the flight attendant turns to Joe and says, “You, I need you to do as I say!” Suppose Joe withholds his consent. Still, given the situation, Joe has an obligation to do what the attendant says. What explains that obligation?
The first explanation—the one Estlund prefers—is authority-based. Joe’s non-consent is null, and thus equivalent to valid consent. When the attendant tells Joe to, say, fetch bandages, Joe is obligated to obey because the attendant has authority over him.
But there is an authority-free explanation too: Joe must do what the attendant says because it is important to help the injured passengers. On its face, the authority-free explanation is just as plausible as the one Estlund prefers—arguably more so, as the obligation to help vulnerable people in such emergencies is a philosophical commonplace. According to the authority-free explanation, Joe rightly regards the flight attendant exactly as Wolff regards the ship captain, or as you regard Foghorn—an agent positioned to coordinate action towards a desperately important end, not as the bearer of a special and controversial normative status.
What Estlund needs is an argument that the authority-free explanation is mistaken. He offers what we might call the ‘argument from sub-optimal commands.’
Suppose that [the attendant] were to order Joe to grab the bandages from … the overhead compartment. Joe correctly believes that it would be wiser to secure whatever fresh water can be found first. Does this exempt Joe from the duty to obey her command? On the contrary, unless the stakes were especially high, it would be wrong for Joe to decline to obey on that ground …. This is a characteristic of authority, and different from merely following the leader when and only when she is leading correctly. Footnote 25
In the sense relevant here, a sub-optimal command (a) doesn’t seriously imperil anyone; (b) comes from someone who is effectively coordinating our actions towards an important goal; and (c) serves that goal slightly worse than a different order would. The argument is that Joe is obligated to obey the attendant’s sub-optimal commands, and that only the attendant’s authority can explain that. After all, if Joe is correct to see her as Wolff sees the ship captain, or as you regard Foghorn, wouldn’t he also be correct to disregard sub-optimal commands? If the attendant’s orders are a mere means to saving the passengers, thinks Estlund, there cannot possibly be any other reason to obey them.
Unfortunately, Estlund’s argument from sub-optimal commands just calls for another iteration of the Occam-inspired test. We again must distinguish two kinds of cases: those in which there are authority-independent reasons for Joe to obey, and those in which there are not.
When there are authority-independent reasons for Joe to do as he’s told, it is, of course, plausible that Joe has an obligation to obey the attendant. The problem for Estlund is that the relevant authority-independent reasons are no worse an explanation than the attendant’s authority. Suppose, to make the example vivid, that Joe can read the rising panic on the faces of his fellow passengers. He thinks that publicly obeying the attendant’s sub-optimal command will set a precedent, thereby pre-empting a costly chaos in which each panicked passenger strikes out on her own. Why must Joe obey? On its face, ‘Because it is important to avoid costly chaos’ is no worse than ‘Because his non-consent was null and thus established the flight attendant’s authority.’ As the duty to avoid costly chaos is another philosophical commonplace, the authority-free explanation might even be better.
But what if there is no authority-independent reason to obey the attendant’s sub-optimal commands? If Joe actually had an obligation to obey in such cases, Estlund would win the argument. But two considerations combine to make denying Joe’s obligation just as plausible as affirming it. First, Joe’s disobedience would have no ill consequence whatsoever—any ill consequence, after all, would be an authority-independent reason for Joe to obey. Second, the attendant’s command is sub-optimal, so Joe’s disobedience would better serve the wounded. Joe’s disobedience would be akin to driving slightly faster than the posted limit when (a) there is no reason not to—outside the commander’s authority, which is precisely what’s at issue—and (b) the traffic patterns make mild speeding slightly safer for everyone on the road.
To sum up, Estlund’s overall view fails in roughly the same way as Wellman’s. We must distinguish between cases in which there are good authority-independent reasons to obey, and cases in which there are not. In the former, authority is otiose. In the latter, denying the existence of Joe’s obligation to obey is just as plausible as affirming it.
3.3 The Argument Generalized
It is now time to generalize beyond Estlund and Wellman. Let E be an end (or a portfolio of ends) such that (i) securing E is important, and (ii) some sort of centralized government is helpful, or necessary, for securing E. If citizens place themselves under their government’s (or the law’s) authority by performing some sort of voluntary action—as the citizens of Consentia did—the authority-independent value of E will not fully explain why citizens should obey. A full catalogue of their error must include the fact that they made their government (or the law) normatively special and nonetheless failed to give that special normative status appropriate uptake. Otherwise, the need for authority disappears. When there are authority-independent reasons for citizens to obey, de jure authority is otiose; when there are no such reasons, the claim that citizens may disobey is no less plausible than the claim that they must obey.
This argument, it is important to remember, is not at all weakened by the fact that governments solve coordination problems. Estlund’s flight attendant joins Wolff’s ship captain and Foghorn as a plausible model for thinking about governments. They are instrumentally useful for solving vast coordination problems across huge territories and among disparate groups of people. Governments, like Foghorn, can ‘speak loudly.’ But we can explain our duty to obey (when it is uncontroversial that we have one) by citing authority-independent reasons. When there are no such reasons—when our disobedience would have no ill consequence whatsoever—it is plausible to deny that we must obey at all.
Conclusion
It is widely accepted that we should not invoke authority to explain our obligations when there is an equally plausible, authority-free explanation available. And that, I have argued, means that Estlund’s flight attendant stands alongside both Wolff’s ship captain and Foghorn as a plausible model for thinking about just laws and just governments. They are nothing more (or less) than helpful instruments.
Statists often portray philosophical anarchism as the enemy of common sense. Footnote 26 But common sense might not be consistent. The intuition that we may disobey the law when obedience would serve no authority-independent point is arguably just as strong as the intuition that authorities are normatively special.
Acknowledgements:
I would like to thank Henry Richardson, Mark Murphy, Mark Lance, Jason Matteson, Russ Pryba, and Jona Vance.
