Political obligation and the authority of the state

Do we have a moral duty to obey the law simply because it is the law? The question of political obligation is one of the oldest in political philosophy — it is the question Socrates answers, fatally, in the Crito — and one of the hardest to answer well. It must be kept distinct from the related question of the state's authority or legitimacy: a state might have the right to rule (legitimacy) without it following that each citizen has a duty to obey every law (obligation), and the gap between the two is where much of the modern debate lives.

A successful theory of political obligation, most writers agree, must meet several demands at once. The obligation should be general (binding on more or less all citizens, not just the politically active); it should be content-independent (a reason to obey because the law commands, not merely because the act commanded is independently right); and it should satisfy the particularity requirement (it must bind me to my own state, not to just any tolerably just state). No leading theory comfortably meets all three, which is why a strand of serious philosophers — the philosophical anarchists — conclude that there is no general obligation to obey at all.

This note works through the main candidates: consent (Locke, and Hume's devastating reply); the principle of fair play (Hart, Rawls, and Nozick's objection); the natural duty to support just institutions (Rawls, and the particularity problem); associative obligation (Dworkin, Gilbert); and Joseph Raz's service conception of authority, with its normal justification, dependence, and pre-emption theses. It closes with the anarchist challenge. The Razian material connects directly to your jurisprudence work on rights and Raz, and the debate runs alongside Rawls and Nozick.

The problem is ancient. In Plato's Crito, the imprisoned Socrates refuses to escape, imagining the Laws of Athens arguing that he has, by a lifetime's residence and acceptance of their benefits, entered into an agreement to obey them, and that to flee would be to wrong the city that nurtured him — an argument blending consent, gratitude, and fair play that already contains, in embryo, most of the moves the later tradition makes explicit.

The modern treatment begins with the social contract theorists. Hobbes grounds the duty to obey in the rational escape from the state of nature; Locke, more influentially for this topic, grounds it in consent, and is driven to the doctrine of tacit consent to explain how those who never expressly agreed are nonetheless bound. The contract tradition's vulnerability — that hardly anyone has actually consented to anything — was exposed once and for all by David Hume, whose essay 'Of the Original Contract' (1748) is the hinge of the whole subject.

After a long period in which the question was thought either settled or uninteresting, political obligation was revived as a sharp analytic problem in the mid-twentieth century, above all by H.L.A. Hart's 1955 paper on the principle of fairness and by John Rawls. The revival's most striking product, however, was sceptical: A.J. Simmons's Moral Principles and Political Obligations (1979) argued, theory by theory, that none succeeds, reviving a respectable philosophical anarchism. Running through and beyond this debate is Joseph Raz's reorientation of the question away from a single obligation owed by everyone and toward an analysis of authority itself — what it is for one person's say-so to make a difference to what another ought to do.

What a theory of political obligation must deliver

Before surveying the theories, fix the target. A political obligation, as standardly understood, is a moral requirement to obey the law that is:

• general — it binds (more or less) all citizens of a legitimate state, not merely those who hold office or are politically engaged;
• content-independent — it gives a reason to do what the law requires because the law requires it, over and above any independent reason the act may have (you should not murder anyway; the interesting cases are parking rules and tax);
• particular — it ties each person to their own state, explaining why I owe allegiance here rather than to any other equally just state.

Much of the debate is the discovery that theories which secure one of these tend to lose another.

Consent

The most intuitive theory: I am bound because I have agreed to be. Express consent (an oath, naturalisation) plainly binds, but covers almost no one. Locke therefore invokes tacit consent — by residing in a territory, using its highways, or enjoying its protection, one tacitly agrees to its laws. The strength of consent theory is that it delivers content-independence and particularity at a stroke; its weakness is empirical, and fatal, as the next section shows.

Fair play (the principle of fairness)

Developed by H.L.A. Hart ('Are There Any Natural Rights?', 1955) and John Rawls, the principle of fair play holds that when a number of people cooperate in a mutually beneficial scheme, restraining their liberty under its rules, one who has accepted the benefits of the scheme owes it to the others to bear a fair share of the burdens — to free-ride is to treat the cooperators unfairly. Applied to the state, the law is such a scheme (defence, order, public goods), and citizens who enjoy its benefits owe obedience. The theory grounds obligation without fictional consent — but faces Nozick's objection that benefits one merely receives, rather than voluntarily accepts, cannot generate obligations.

The natural duty of justice

Rawls, in A Theory of Justice, grounds the political bond of most citizens not in consent or benefit but in a natural duty — which, in Rawls's own technical vocabulary, is distinct from an 'obligation' (in his usage, obligations arise only from voluntary acts, through the principle of fairness). This natural duty binds independently of any voluntary act, requiring us 'to support and to comply with just institutions that exist and apply to us'. This neatly secures generality (it binds everyone, regardless of what they have done or accepted). Its difficulty is the particularity requirement: a duty to support just institutions as such seems to bind me equally to every just state, and so fails to explain my special tie to my own.

Raz: authority as a service

Joseph Raz reframes the question. Rather than ask what single obligation everyone owes, he analyses what it is for an authority to be legitimate, through the service conception: authority is justified when it helps its subjects act better on the reasons that already apply to them. Its three components — the dependence, normal justification, and pre-emption theses — are set out in §05 and §06. A striking consequence is that legitimate authority is piecemeal: it holds only where the justification is met, which may vary from person to person and law to law — so there may be no general obligation to obey at all.

Locke, Second Treatise of Government (1689)

The classic statement of consent theory and the source of the doctrine of tacit consent (§§119–122): a person who enjoys the protection of a government's laws thereby gives tacit consent to obey, 'whether this his possession be of land… or a lodging only for a week, or whether it be barely travelling freely on the highway'.

Hume, 'Of the Original Contract' (1748)

The decisive critique of consent theory, and a set essay. Hume argues that the social contract is both historically false and morally idle: rulers everywhere claim to rule by divine right or conquest, not contract, and the supposed tacit consent of the governed is a fiction, since most people 'have never dreamed of' consenting and could not in any case leave.

Rawls, A Theory of Justice (1971)

For this topic, read the sections on the natural duty of justice and the principle of fairness (§§18–19, 51–52). Rawls restricts the principle of fairness to those who voluntarily accept the benefits of, or take advantage of opportunities offered by, just institutions, and rests the political bond of ordinary citizens on the natural duty to support just institutions instead. See also his earlier 'Legal Obligation and the Duty of Fair Play' (1964).

Simmons, Moral Principles and Political Obligations (1979)

The most rigorous modern treatment, and the locus classicus of contemporary philosophical anarchism: a theory-by-theory demonstration that consent, fair play, gratitude, and natural duty all fail to ground a general obligation meeting the generality and particularity requirements.

Raz, The Authority of Law (1979) and The Morality of Freedom (1986)

Raz's account of authority. The Authority of Law introduces exclusionary reasons and argues there is no general obligation to obey; The Morality of Freedom (ch. 2–4) develops the service conception — the dependence, normal justification, and pre-emption theses.

1. Hume's critique of tacit consent

Hume's argument is the one every essay must engage. Tacit consent, he says, cannot be inferred from mere residence, because genuine consent requires a real alternative, and most people have none. In his famous image, we may as well say that a man who is carried aboard a ship while asleep 'freely consents to the dominion of the master' because he could, in principle, leap into the ocean and drown. Where the only 'choice' is to abandon one's language, livelihood, and home, continued residence signifies no agreement at all. Consent theory thus secures the right kind of obligation but cannot show that anyone has actually incurred it.

2. Nozick's objection to fair play

Robert Nozick (Anarchy, State, and Utopia, 1974) attacks the move from benefit to obligation. Imagine your neighbourhood runs a public-address entertainment system, assigning each resident a day to broadcast; you never agreed, but you have enjoyed others' broadcasts. Are you now obliged to give up your day? Nozick says plainly not: 'one cannot, whatever one's purposes, just act so as to give people benefits and then demand (or seize) payment'. Benefits merely received — even welcome ones — do not bind; only benefits one has voluntarily accepted can. Simmons sharpens the point into a general requirement of voluntary acceptance, which most citizens never meet for the state's goods.

3. The particularity problem

The natural-duty theory faces a different difficulty. Grant that we have a natural duty to support just institutions. Why does this bind me to Britain's institutions rather than to every just state's? A duty owed to justice-in-general seems not to single out any particular polity as the one I must obey — yet political obligation is paradigmatically a special tie to one's own community. Jeremy Waldron ('Special Ties and Natural Duties', 1993) replies that just institutions legitimately claim the allegiance of those whose conduct they actually coordinate and to whom they apply, so the duty does fasten on the institutions that 'range over' me — but whether this fully answers the objection is contested.

4. Wolff's autonomy argument for anarchism

R.P. Wolff (In Defense of Anarchism, 1970) presses an a priori case. Moral autonomy — the duty to act on one's own reasoned judgement — can never be surrendered. But to accept the authority of the state is precisely to take its commands as binding whatever one's own judgement of their merits. Authority and autonomy are therefore incompatible, and legitimate authority is impossible. The argument is widely thought too strong (Raz's analysis is in part a reply), but it sets the bar that any account of authority must clear.

5. Raz's normal justification thesis

Raz's central claim, the normal justification thesis (NJT), states that the normal way to establish that a person has authority over another is to show that the subject is more likely to act in accordance with the reasons that already apply to them if they accept the authority's directives as binding and try to follow them, than if they try to weigh and act on those reasons directly. Authority, so understood, is a service: it mediates between people and the right reasons, and is legitimate just where it performs that service. (You met this in jurisprudence; the same NJT, dependence, and pre-emption theses are developed there in the context of law's claim to authority.)

Raz's service conception in full

Raz's account has three linked theses. The dependence thesis: authoritative directives should be based on reasons that already, independently, apply to the subjects of the directive (the dependent reasons). The normal justification thesis (above): authority is justified where following it makes one conform better to those dependent reasons than deciding for oneself. The pre-emption thesis: a valid authoritative directive is not merely another reason to be weighed in the balance, but a pre-emptive (exclusionary) reason that replaces some of the reasons it is based on — one is meant to act on the directive instead of re-deliberating, otherwise the point of having authority (to settle what to do) is lost.

Two consequences matter. First, authority is piecemeal and content-sensitive: whether the NJT is satisfied depends on the person, the domain, and the expertise involved, so a state may have authority over me in some matters and not others. Second, and strikingly for this topic, Raz concludes that there is probably no general obligation to obey the law even in a just state, because the NJT is unlikely to hold across the board for every citizen and every law. Authority and obligation come apart.

Associative obligations

A different post-Rawlsian move denies that political obligation needs a voluntary or instrumental ground at all. Ronald Dworkin (Law's Empire, 1986) argues that political obligation is an associative or 'fraternal' obligation — like the non-chosen obligations of family or friendship — that arises from membership of a genuine community whose practices treat members with equal concern. Margaret Gilbert (A Theory of Political Obligation, 2006) grounds it in joint commitment: members of a 'plural subject' are bound to one another by their participation in a shared political life, independently of consent or benefit. Critics (Simmons again) doubt that the analogy with family holds, and ask why mere membership of a political community, which one did not choose, should generate genuine moral requirements.

Fairness revisited

George Klosko (The Principle of Fairness and Political Obligation, 1992) tries to rescue fair play from Nozick's objection by restricting it to 'presumptive goods' — goods such as physical security and the rule of law that any rational person must want and that are non-excludable. For such goods, he argues, the receipt/acceptance distinction loses its grip: it is not open to anyone reasonably to decline national defence, so the benefit binds even without an act of acceptance.

The generality–particularity squeeze

The organising debate is whether any theory delivers an obligation that is at once general, content-independent, and particular. Consent has the right shape but binds almost no one (Hume). Fair play binds only those who voluntarily accept benefits, which excludes most citizens for most public goods (Nozick, Simmons) — unless Klosko's presumptive-goods move succeeds, which critics say either overreaches or quietly abandons the voluntariness that made fairness attractive. The natural duty binds everyone but struggles with particularity (Waldron replies, inconclusively). Associative theories deliver particularity and generality but at the cost of content-independence and of the worry that unchosen membership cannot ground genuine obligations. The pattern is what drives Simmons's conclusion.

Philosophical anarchism

A.J. Simmons defends a posteriori philosophical anarchism: not the a priori claim (Wolff) that authority is impossible, but the conclusion, reached after surveying the theories, that as things stand no general obligation to obey has been shown to exist. Crucially, this is anarchism about obligation, not about action: Simmons does not say the state is illegitimate in every sense, nor that we should disobey — there remain content-dependent moral reasons to do most of what the law requires (not to kill, not to steal), and prudential reasons besides. What collapses is only the special, content-independent duty to obey qua law. Distinguishing this carefully from revolutionary anarchism is the single most important move in an exam on this topic.

Legitimacy without obligation

Raz's piecemeal authority, and Simmons's separation of legitimacy from obligation, converge on a now-influential thought: a state may be legitimate — it may have the standing to make and enforce law, and citizens may have good reason not to interfere with it — without there being a general obligation on each citizen to obey each of its laws. Whether this is a comfortable resting place or an abandonment of the original question is itself debated: some (Klosko, the associative theorists) regard a general obligation as a datum that any adequate theory must vindicate, while others (Simmons, and in a different way Raz) are content to let it go.

It helps to see the theories as trading off the three desiderata against one another.

Consent vs natural duty. These are near-mirror images. Consent theory secures content-independence and particularity (I am bound to the state I agreed to) but fails on generality, since hardly anyone has consented. The natural duty secures generality (it binds everyone independently of what they have done) but fails on particularity. The history of the subject is largely the search for a theory that captures both virtues at once.

Fair play vs Raz. Fair play locates obligation in a relationship among co-operators — a horizontal tie between citizens who share the burdens of a scheme. Raz locates legitimacy in a vertical relation between subject and authority, justified by the authority's service in tracking the subject's own reasons. The two can be complementary: fair play might explain why I owe my fellow citizens, while the NJT explains when the state's directives bind. But Raz's piecemeal, content-sensitive authority pulls against fair play's aspiration to a general obligation.

Associative obligation vs the liberal theories. Dworkin and Gilbert break with the whole liberal-voluntarist framework, treating political obligation as a non-chosen relational duty. Against Rawls and Nozick, who in their different ways demand that obligations be earned (by acceptance, by agreement), the associative theorists insist that the deepest obligations — to family, friends, country — are precisely the ones we never chose. The dispute is, at bottom, about whether obligation must be voluntarist, and it connects back to the communitarian critique of the Rawlsian unencumbered self.

Question: 'Is there a general obligation to obey the law?'

Introduction

The question asks whether citizens of a reasonably just state have a moral duty to obey its laws that is general (binding on nearly all of them), content-independent (a reason to obey because the law commands, not merely because the act is independently right), and particular (tying each to their own state). This essay surveys the leading attempts to ground such an obligation — consent, fair play, natural duty, associative membership, and Raz's service conception — and argues that none establishes an obligation meeting all three conditions. The right conclusion is a qualified philosophical anarchism of Simmons's kind: there is no general content-independent duty to obey qua law, though this is fully compatible with there being ample moral reason to do most of what the law in fact requires, and with the state's being legitimate in other senses.

Consent

Consent theory has exactly the right shape: one who has agreed to obey is bound, content-independently and to a particular state. The trouble is that almost no one has expressly consented, and the doctrine of tacit consent, invoked to fill the gap, collapses under Hume's critique: continued residence cannot signify agreement where leaving means abandoning one's home, language, and livelihood — the sleeper carried aboard ship does not 'consent' to the captain merely because he could drown himself by jumping overboard. Hypothetical consent fares no better: as Dworkin observes, 'a hypothetical contract is not simply a pale form of an actual contract; it is no contract at all', and the work is then done by whatever makes the agreement rational, not by consent. Consent thus fails on generality.

Fair play and natural duty

The principle of fair play (Hart, Rawls) grounds obligation in the unfairness of free-riding on a cooperative scheme whose benefits one has accepted. But Nozick's public-address example shows that benefits merely received do not bind; only benefits voluntarily accepted do, and most citizens never accept the state's public goods in the required sense. Klosko's restriction to 'presumptive goods' (security, the rule of law) is the best reply, but it arguably purchases generality by dropping the voluntariness that gave fairness its moral force. The natural duty of justice (Rawls) secures generality directly — we owe support to just institutions independently of any voluntary act — but founders on particularity: a duty to justice-in-general does not obviously bind me to my state. Waldron's reply, that just institutions claim the allegiance of those they actually coordinate, mitigates but does not clearly dissolve the problem.

Associative obligation and Raz

Two further routes remain. Associative theories (Dworkin, Gilbert) treat political obligation as a non-chosen relational duty of membership, like familial obligation; they capture particularity and generality, but at the cost of content-independence and of the unanswered worry that unchosen membership of a coercive community is disanalogous to the loving relationships from which the intuition is drawn. Raz's service conception, finally, reframes the issue: authority is legitimate where following it helps one conform to the reasons that already apply (the normal justification thesis), and its directives pre-empt rather than merely outweigh one's own reasoning. But precisely because the NJT is satisfied only piecemeal — varying with person, law, and domain — Raz himself concludes that there is no general obligation to obey, even in a just state. The most sophisticated theory of authority thus yields a sceptical answer to our question.

Judgment

Every candidate secures some of the desiderata at the expense of another, and the most rigorous surveys (Simmons) and the most developed theory of authority (Raz) both reach the sceptical conclusion. I therefore answer the question 'no': there is no general, content-independent obligation to obey the law qua law. Two qualifications keep this from being alarming. First, the conclusion is anarchism about obligation, not about action: we retain strong content-dependent reasons to comply with most laws (against violence, fraud, and the like) and prudential reasons besides, so near-universal obedience is rational even without a duty to obey as such. Second, the legitimacy of the state — its standing to make and enforce law — can survive the denial of a general obligation, as Raz's piecemeal authority shows. The honest position is that the state may rightly govern, and we may have every reason to obey, without there being the single, tidy, content-independent duty that the tradition set out to find.

1. Conflating obligation with legitimacy/authority. The state's right to rule (legitimacy) and each citizen's duty to obey (obligation) are distinct; a leading modern view (Raz, Simmons) is that you can have the former without a general version of the latter. Keep them apart.

2. Reading philosophical anarchism as a call to revolution. Simmons's anarchism is about obligation, not action: it denies a content-independent duty to obey qua law, while allowing that we have ample other (content-dependent, prudential) reasons to do most of what the law requires. Confusing it with bomb-throwing anarchism is the classic blunder.

3. Forgetting the particularity requirement. A theory can secure generality and content-independence yet still fail because it does not bind me to my own state (the natural-duty theory's weak point). Always test a candidate against all three: general, content-independent, particular.

4. Treating tacit consent as obviously sufficient. Locke's tacit consent is precisely what Hume destroys; do not invoke it without the critique. Mere residence or benefit-enjoyment is not agreement where there is no real exit.

5. Mixing up 'receiving' and 'accepting' benefits. Nozick's objection to fair play turns on this distinction. Benefits thrust upon you do not bind; only benefits you voluntarily take up do. Klosko's 'presumptive goods' is the move that tries to get around it.

6. Misstating Raz. The normal justification thesis is about helping subjects conform to reasons that already apply to them; pre-emptive reasons replace rather than merely outweigh first-order reasons. And note Raz's surprising upshot: legitimate authority does not entail a general obligation to obey.

7. Ignoring the autonomy/anarchism a priori argument. Wolff's claim that authority is incompatible with moral autonomy is the strong, a priori anarchism that any account of authority (Raz's especially) must answer; distinguish it from Simmons's a posteriori version.

The practice questions for this topic — with model answers and mark schemes across foundation, standard, and challenge tiers — are set out below. Attempt each under timed conditions before revealing the model answer, then mark your work against the scheme.

The primary texts and the leading literature are listed below. Read Hume's essay and Simmons first; then Raz (The Morality of Freedom, chs. 2–4) for the service conception, and Dworkin (Law's Empire, ch. 6) for the associative alternative.