Home › Principles of Public Law › Part 5 · Customary international law

Part 5 · Customary international law

Custom is the unwritten half of international law: it binds states because of what they do and what they believe, not because of what they sign. The exam question is always the same, prove the rule or break it, and both jobs run through two elements and one exception.

Ideas and cases at a glance

Law without a legislature

International law has no parliament, no compulsory court and no police. It is a horizontal order of sovereign equals in which the rules are made by the very states they bind, on the basis of consent. Lotus states the foundation: the rules of law binding upon states emanate from their own free will. Yet the system holds, because compliance rests on reciprocity, reputation and legitimacy rather than enforcement. Before you can prove any rule, you need to know where rules come from, and the answer sits in one provision.

The list carries no formal hierarchy. In practice a treaty applies first between its parties as the more specific rule, general principles fill gaps, and conflicts are managed by the ordinary tools of specificity and time, always subject to jus cogens. Custom is the source that does the heavy lifting when no treaty binds the state in front of you, because custom binds all states.

The two-element test

Everything in this module funnels into a single test, and the discipline is to run its limbs separately.

Element one: state practice

Practice means what states actually do: physical acts and abstentions, national legislation, diplomatic statements and correspondence, and conduct in international organisations. Three qualities matter. Consistency: the practice must be reasonably settled, requiring substantial but not complete uniformity. Generality: the practice must be widely shared, with broad geographic spread, and it must include the states whose interests are specially affected, so a maritime rule without the major maritime states is going nowhere. Duration is flexible: a short period is no bar, but if the time is short the practice must be both extensive and virtually uniform (North Sea [74]). Local or regional custom can also exist, but it is harder to prove: a constant and uniform usage accepted as law by the state against whom the rule is invoked (Asylum Case).

Element two: opinio juris

Opinio juris is the psychological element: states must follow the practice because they recognise themselves as legally obliged to do so. This is what separates law from habit. Acts of ceremony, courtesy, convenience or tradition fail, however uniform they are (North Sea [77]), and even abstention counts only if the state refrains because it is conscious of a duty to refrain (North Sea [78], applying Lotus). Evidence is usually inferred: explicit statements about the law, positions taken on General Assembly resolutions, and, as the International Law Commission accepts, a failure to react over time to a practice, provided states were in a position to react and the circumstances called for it.

Imperfect practice

States break rules, and the test does not demand perfection. Conduct need not conform rigorously with the rule: it is enough that state conduct is generally consistent with it and that inconsistent conduct has generally been treated as a breach of the rule, not as the seed of a new one (Nicaragua [186]). The point cuts sharply in argument: a state that defends its violation by appealing to the rule's own exceptions has confirmed the rule, not weakened it.

Objectors, new states and the peremptory floor

Once both elements are satisfied, the rule binds all states. Consent then survives in one narrow doorway.

The doctrine preserves the consent principle from Lotus, but its limits matter as much as its existence. New states cannot pick and choose: they are born into the existing corpus of custom and cannot object their way out of it. A subsequent objector can still change the law the slow way, if other states acquiesce in its deviation over time. And no objection, however early or persistent, exempts a state from jus cogens, because peremptory norms permit no derogation. If the question involves genocide, slavery, torture or the use of force, persistent objection is a dead end, and saying so crisply earns marks.

Treaties and custom

Treaties and custom are separate sources that constantly interact. A treaty may codify custom that already exists, crystallise a rule that was emerging, or generate new custom as non-parties adopt its rule in practice. The customary rule keeps its separate existence even between treaty parties (Nicaragua), which is why leaving a treaty, or never joining it, does not escape the parallel customary obligation. The hard question is when a treaty rule becomes general custom, and North Sea Continental Shelf answers it.

The ICJ applied its own test strictly. The equidistance rule in Art 6 of the Continental Shelf Convention failed at every step: the provision was not norm-creating, since it was secondary to agreement and carried a faculty of reservation, which adds considerably to the difficulty ([72]); ratifications were insufficient; and the practice of parties proved nothing, because states acting under a treaty may be doing no more than performing it ([76]). That last point is the one students miss: 174 ratifications are equivocal evidence of custom, since a party's compliance may show treaty obligation rather than legal conviction outside the treaty.

How custom changes

Custom changes by the same two elements that made it. A state deviates, and the deviation is, at that moment, necessarily a breach of the existing rule. What matters is the response: if other states object, the old rule holds; if they acquiesce and fresh opinio juris gathers around the new practice, yesterday's breach becomes tomorrow's rule. That is why the reaction of other states is always the fact to hunt for in a problem about changing custom.

The attack plan

Worked example

An Indigenous elder's licence to run a cultural workshop is cancelled by a local council after a ten minute debate. She argues that customary international law gives Indigenous peoples a right to participate in decisions affecting their cultural rights, relying on legislation from six states dating back to the 1970s, a 1984 GA resolution adopted by consensus, a 2017 statement by Iceland in the Human Rights Council endorsed by three delegations, and a 2019 book by a renowned expert. Advise her.

Critical pitfalls and counter-arguments

• Collapsing practice into opinio juris. The elements are cumulative and separately proved. Where practice is thin, the psychological element cannot do the work of both.
• Treating a breach as disproof. Violations generally confirm the rule when treated as breaches, and a state pleading the rule's exceptions has admitted the rule (Nicaragua [186]).
• Mistiming the objector. Persistent objection must begin during formation and never lapse. Late objection binds; and no objection survives jus cogens.
• Counting ratifications as custom. Treaty participation is equivocal: parties may act from treaty obligation, not legal conviction (North Sea [76]). Non-party practice is the better evidence.
• Overstating consent. The policy counter-argument runs the other way: custom binds states that never actively consented, and new states inherit the corpus, a real tension with Lotus voluntarism worth naming in an essay.

Take it to the practice bank

The Principles of Public Law practice bank drills this module across twenty customary international law problems, each with a Pass answer, an H1 answer and the examiner's comments. Start with Problem 1 on proving practice and opinio juris, then Problem 2 on regional custom and the persistent objector and Problem 14 on the subsequent objector. Problems 4, 5 and 11 work the treaty and custom relationship, Problem 6 tests GA resolutions as evidence, Problems 7 and 17 run jus cogens against consent, and Problems 9 and 20 turn on specially affected states and imperfect practice.

Practice prompts

1. Take the worked example and rewrite the practice paragraph twice: once where the six legislating states include those with the largest Indigenous populations, once where they do not. Watch the specially affected states requirement change the conclusion on its own.
2. Write the paragraph proving that a rule survives widespread violation, using Nicaragua [186] in two sentences: one for the principle, one applying it to a state that justified its breach under the rule's self-defence exception.
3. Draft the persistent objector paragraph you would write for a state that began objecting five years after the rule crystallised, and conclude honestly against your own client before explaining the acquiescence route by which the law might still change.