Understanding International Law: Framework and Challenges

The Reagan administration saw the verdict coming and withdrew from the International Court of Justice's compulsory jurisdiction before the case concluded, declaring that the Court had no authority over what Washington considered national security questions. The Court ruled in June 1986 regardless.

It found that the United States had violated international law by mining Nicaragua's harbors and by training, equipping, and directing the Contra forces fighting to overthrow the Sandinista government,^[7] by 14 votes to 1 on the mining and 12 votes to 3 on the non-intervention count. The United States then vetoed the Security Council resolution calling on all states to comply.

Nicaragua never saw a dollar of compensation. Nobody was punished.

Cite that sequence and you have the standard case for treating international law as a fiction: powerful states ignore rulings they dislike, and nothing follows. Then look at what Washington actually spent its time on. It disputed the Court's jurisdiction, contested the standing of Nicaragua's claims, and offered alternative characterizations of its own conduct, none of which it needed to do.

A government that genuinely thought the proceedings meaningless would have let them proceed and said nothing. Going to the trouble of looking compliant while defying the ruling is itself evidence that the norms bite,^[3] and pure power politics would not generate that kind of effort.

Both the enthusiasts and the cynics are describing something that does not exist. There is no global rule of law, and Nicaragua's empty hands prove it. But there is also no empty pretension, or Washington would not have bothered with the lawyers. What there is instead sits awkwardly between the two: a legal order that almost nobody enforces and most states follow anyway, that rules out some options before anyone considers them, and that makes violation expensive in diplomatic and reputational currency rather than in penalties a court can collect.

It is the most serious attempt anyone has made to govern relations between sovereign states by something other than raw power, and it works unevenly enough that both camps can keep citing it. Reading the rest of this guide means keeping the achievements and the limits in the same frame.

"Almost all nations observe almost all principles of international law and almost all of their treaty obligations almost all of the time."

Louis Henkin, How Nations Behave (1979)^[6]

Key Definitions

International law is the body of rules, principles, and norms governing how states, international organizations, and increasingly non-state actors behave toward one another.^[2] It has no centralized enforcement authority, and it still generates substantial compliance through reciprocity, reputation, and institutional mechanisms.^[8] Everything below is a piece of how that happens.

Two sources do most of the work, and they bind states in different ways. Treaty law covers binding written agreements between states, governed by the Vienna Convention on the Law of Treaties (1969), and it rests on pacta sunt servanda: the principle codified in Article 26 of that Convention that every treaty in force binds its parties and must be performed in good faith.^[13]Customary international law works without anyone signing anything.

It emerges from consistent and general state practice combined with opinio juris, the belief among states that the practice is legally required rather than merely habitual. Those two elements together are what establishes a custom, and once established it binds every state, including ones that took no part in creating it, subject only to the narrow "persistent objector" exception.

Above both sits a category that consent cannot touch. Jus cogens norms are peremptory: no derogation is permitted from them even by treaty, and they bind all states whatever any government has agreed to. The prohibitions on genocide, slavery, torture, and aggression belong here.

Two further terms describe how institutions divide the labor. Complementarity governs International Criminal Court jurisdiction, letting the Court act only where national courts are "unable or unwilling" to genuinely investigate and prosecute, which keeps state primacy over criminal jurisdiction while providing a backstop against impunity. Chapter VII is the section of the UN Charter that authorizes the Security Council to take binding measures, force included, against threats to international peace and security.

The veto held by its five permanent members is what critically limits how far that reaches.

Is International Law Real Law? The Austin Objection

John Austin settled the question to his own satisfaction in 1832. Law, he argued in The Province of Jurisprudence Determined, is the command of a sovereign backed by a credible threat of sanction.^[1] International law has neither. No sovereign sits above states issuing commands to them, and nothing resembling domestic police and courts reliably enforces anything.

So what states observe between themselves is not law on this account but "positive morality": norms they may follow out of custom, convenience, or conscience, carrying no binding legal force at all.

Give Austin his due, because the structural point is correct and has not gone away. International law still has no centralized legislature, no hierarchical court system with compulsory jurisdiction over everyone it binds, and no executive to enforce its judgments. Violate it and no international police force arrests your leaders, no sanction triggers automatically, and no institutional consequence is guaranteed to follow at all.

What Austin could not have predicted is how states actually behave. Treaty compliance studies keep finding the same thing: states honor their commitments the overwhelming majority of the time, and not because somebody enforces each instance. They do it because reputations as reliable actors are worth having in a system where everyone needs future agreements.

Henkin's line about almost all nations observing almost all principles almost all of the time flattens real variation, but it gets the important part right. The default is compliance. Defection is the exception that gets written about.

Four mechanisms do the work Austin thought required a sovereign. Reciprocity means a violation invites retaliation in kind. Reputation means a serial violator pays more to negotiate anything later. Domestic politics turns international commitments into expectations that constituencies at home will hold a government to.

And international institutions make non-compliance visible, which raises its price. None of the four holds when a powerful state has a strong short-term interest in breaking a rule,^[5] which is exactly when you would most want them to. Together, though, they produce real behavioral constraint without any enforcing sovereign anywhere in the picture.

Sources of International Law: Article 38 and Its Framework

Where does a rule of international law actually come from? The definitive answer is Article 38(1) of the Statute of the International Court of Justice, which directs the Court to apply:^[11]

1. International conventions (treaties), whether general or particular, establishing rules expressly recognized by the contesting states;
2. International custom, as evidence of a general practice accepted as law;
3. The general principles of law recognized by civilized nations;
4. Subject to Article 59, judicial decisions and the teachings of the most highly qualified publicists, as subsidiary means for the determination of rules of law.

That list runs roughly from the most deliberate source to the least. Treaty law sits at the top because nothing about it is accidental: a treaty requires affirmative state consent, signature followed by ratification through whatever domestic constitutional process applies. The Vienna Convention on the Law of Treaties (1969), itself a treaty and now with 116 parties, codifies how treaties are concluded, interpreted, and terminated.

Its Article 26 carries the foundational rule, that "Every treaty in force is binding upon the parties to it and must be performed by them in good faith", while Article 31 handles interpretation: good faith, the ordinary meaning of the terms in their context, and the object and purpose of the thing.

Customary international law is messier, because nobody signs it. A custom exists where consistent state practice meets opinio juris, the belief that the practice is legally required rather than merely conventional, and identifying one is closer to an art than a science. Take the prohibition on torture.

It binds all states as custom, not only the 173 parties to the Convention Against Torture. The practice side is evidenced in domestic legislation, diplomatic communications, military manuals, and court decisions; the opinio juris side shows up in the fact that states formally characterize torture as illegal even while practicing it covertly, which is its own kind of proof. Put those together and you get why international tribunals and most scholars treat the torture prohibition as more than customary: it is jus cogens, a peremptory norm admitting no derogation at all.

The last two sources fill gaps rather than create rules. General principles of law supply what treaties and custom leave uncovered, drawing on principles recognized across major legal systems: good faith, res judicata, procedural fairness, the prohibition on unjust enrichment. Judicial decisions and scholarly writings are subsidiary, and they matter most for establishing what the primary sources actually provide in areas where state practice is sparse or contested.

International Humanitarian Law

The laws of war, also called international humanitarian law or the law of armed conflict, are among the oldest and most developed parts of international law, and their codified framework is effectively universal. The core instruments are the four Geneva Conventions of 1949 and the Additional Protocols of 1977.^[4] The Conventions have 196 states parties, which is more than the United Nations has members: a rare case of a treaty that genuinely everyone has ratified.

The four Conventions divide the protected up by circumstance. Convention I covers wounded and sick combatants on land, Convention II wounded, sick, and shipwrecked naval personnel, Convention III prisoners of war, and Convention IV civilians in occupied territory and during armed conflict. Common Article 3, which appears identically in all four, sets minimum protections for non-international armed conflicts, meaning civil wars and insurgencies rather than only wars between states.

The 1977 Additional Protocols pushed the protections further, especially for those internal conflicts, and wrote into treaty form a good deal of what had until then been customary law.

Three cardinal principles govern how fighting may be conducted, and each one is a limit on what a commander may order. Distinction comes first: a party to a conflict has to tell civilians from combatants, and civilian objects from military objectives, at all times, and may aim only at the latter. Proportionality then constrains even a lawful target, because an attack is forbidden when the incidental civilian casualties or damage it is expected to cause run out of proportion to the concrete and direct military advantage anticipated. Military necessity narrows things again: force is permitted only so far as a legitimate military objective actually requires, so violence past that point is unlawful whoever it was aimed at.

Breach any of the three and the word for what you have done is war crime.

Monitoring works oddly here, and deliberately so. The International Committee of the Red Cross holds a treaty-based right of access to prisoners of war and detained civilians, and it uses that right through confidential visits and private reports to the detaining authority instead of public condemnation. The reasoning is practical rather than timid: a state embarrassed in public revokes access, and access is what actually improves how detainees are treated.

Prosecution is the weak link. Domestic cases remain rare, and the international criminal tribunals have reached only a small fraction of the serious violations committed.

Enforcement Mechanisms: Reciprocity, Reputation, and Institutions

International law has plenty of enforcement mechanisms. None of them is adequate on its own, and collectively they still cannot reliably compel a powerful state. The domestic analogy is what misleads people here: enforcement in this system is distributed, indirect, and ultimately dependent on what states themselves want.

Reciprocity is the foundation. States keep to treaties and customary rules partly because they expect everyone else to, and because defecting licenses the other side to defect too. Trade agreements run almost entirely on the expectation that a violation invites countermeasures.

The laws of war hold partly because each party to a conflict expects its own combatants and civilians to be treated by the same rules it applies. Where this breaks down is predictable: dramatically asymmetric power, or a violation whose benefits plainly exceed what retaliation would cost.

Reputation effects pick up some of the slack. A state that violates international law systematically pays diplomatic costs, finds future agreements harder to negotiate, and can lose access to preferential arrangements or international financing. Real costs, and still not enough to deter a government with a strong enough short-term interest.

UN Security Council Chapter VII authorization is as close as the system gets to centralized enforcement. The Council can determine that a threat to international peace exists and authorize binding measures up to and including military force, as it did in Korea in 1950, Iraq in 1990-1991, Somalia, and Libya in 2011. Then there is the veto.

Russia invaded Ukraine in 2022 in straightforward violation of Article 2(4) of the UN Charter, and no Chapter VII response was possible because Russia could veto any resolution proposing one. Structurally, the Council cannot authorize enforcement against a permanent member or its close allies, and the gravity of the violation makes no difference to that.

WTO dispute settlement is the strongest thing in the system for trade obligations. The Dispute Settlement Understanding lets any member complain against any other, requires the loser to bring its measures into conformity, and authorizes the winner to suspend equivalent concessions if that does not happen. Over 600 disputes have gone through it since 1995.

Since 2019 the US has blocked appointments to the Appellate Body, leaving it short of a quorum, and the appellate mechanism has been substantially weaker since.

Human rights treaty bodies, the committees of independent experts that monitor the major conventions, have no enforcement power at all. What they produce is authoritative interpretation and public documentation.^[12] Whether that matters is an empirical question, and Beth Simmons answered it in Mobilizing for Human Rights (2009), still the most comprehensive quantitative study of treaty effects.

Ratification does predict improved outcomes, but only where civil society is active, courts are relatively independent, and the country is in democratic transition.^[10] In a stable autocracy, ratifying without domestic mobilization achieves very little. That result cuts against the optimists and the skeptics in roughly equal measure.

The International Criminal Court

The Rome Statute was adopted in July 1998 and entered into force in July 2002, creating the first permanent international tribunal with jurisdiction over individuals rather than states: genocide, crimes against humanity, war crimes, and, from 2018, the crime of aggression.^[9] The Court sits in The Hague and had 124 states parties as of 2023. The United States, China, and Russia are not among them.

American policy on the Court reversed within two years. President Clinton signed the Rome Statute in December 2000. The Bush administration formally withdrew that signature in May 2002 and, the same month, enacted the American Servicemembers' Protection Act, which authorizes military force to release any American the ICC detains.

It then went state by state negotiating bilateral immunity agreements, the so-called "Article 98 agreements", threatening to cut military assistance to governments that refused. The Court's geographic reach has been substantially narrower ever since.

Even where it does reach, the Court is designed to stay out of the way. Its complementarity principle lets it act only when national courts are genuinely unable or unwilling to prosecute, which makes it a backstop for failed domestic accountability rather than a general criminal tribunal. States keep sovereignty over their own criminal justice; what they get in exchange is accountability of last resort.

What that looks like in practice became clear in March 2023, when the ICC issued arrest warrants for Russian President Vladimir Putin and Russia's Commissioner for Children's Rights, Maria Lvova-Belova, over the alleged war crime of unlawfully deporting Ukrainian children during Russia's invasion. Putin became the first sitting head of state of a permanent UN Security Council member to face one.

Russia is not a Rome Statute party, will not surrender him, and can veto any Security Council action, so in the obvious sense nothing happened.

In a less obvious sense something did. Putin cannot now enter any of the 124 member states without risking arrest, which is a real constraint on where a head of state can go. South Africa, an ICC member, found itself under acute legal and political pressure when it hosted the 2023 BRICS summit, and Putin attended by video rather than in person. The warrant shows the Court's authority and the distance between a legal determination and anyone actually enforcing it, both at once.

Can Powerful Countries Ignore International Law?

More easily than weaker states, and not for free. Enforcement in this system runs asymmetrically, and a permanent Security Council member has far more room to avoid formal accountability than anyone else does.

Three cases make the point. The United States mined Nicaragua's harbors, was found by the ICJ to have violated international law, ignored the ruling, and faced no direct legal consequence. Russia invaded Ukraine in 2022; the UN General Assembly called it a violation of the Charter's prohibition on the use of force in a resolution adopted 141 to 5, and Russia's Security Council veto stopped any binding response.

China rejected the 2016 UNCLOS Arbitral Tribunal ruling that found its South China Sea claims inconsistent with the law of the sea, kept building and militarizing artificial islands, and has met no binding enforcement mechanism since. That is impunity, three times over.

Stopping the analysis there is the mistake, because the costs were not zero in any of the three. They came as sustained diplomatic effort to manage the fallout, as economic damage (the sanctions on Russia after 2022 were enormous, though no court had ordered them), as erosion of credibility as a treaty partner, and as the precedential cost of having helped normalize a violation others can now cite.

Notice also where the effort went. Powerful states offer legal arguments, seek Security Council authorization when they can get it, and negotiate treaty exemptions rather than simply disregarding treaties. That is not what pure power calculation looks like.

So the system is not impartial. It generates different costs for different actors depending on power, on relationships with whoever controls enforcement, and on what exactly was violated. The useful question was never whether international law stops a determined great power, because it frequently does not.

It is whether the law raises the costs, narrows the options, and supplies a legitimacy framework that makes certain conduct systematically harder and more expensive than it would otherwise be. On that question the evidence says yes.

Sources

1. Austin, J. (1832). The Province of Jurisprudence Determined. J. Murray.
2. Crawford, J. (2012). Brownlie's Principles of Public International Law (8th ed.). Oxford University Press.
3. Franck, T. M. (1990). The Power of Legitimacy Among Nations. Oxford University Press.
4. Geneva Conventions of August 12, 1949 and Additional Protocols of 1977. International Committee of the Red Cross.
5. Goldsmith, J. L., & Posner, E. A. (2005). The Limits of International Law. Oxford University Press.
6. Henkin, L. (1979). How Nations Behave: Law and Foreign Policy (2nd ed.). Columbia University Press.
7. International Court of Justice. (1986). Military and Paramilitary Activities in and Against Nicaragua (Nicaragua v. United States of America). ICJ Reports.
8. Koh, H. H. (1997). Why do nations obey international law? Yale Law Journal, 106(8), 2599-2659.
9. Rome Statute of the International Criminal Court. (1998). UN Doc. A/CONF.183/9.
10. Simmons, B. A. (2009). Mobilizing for Human Rights: International Law in Domestic Politics. Cambridge University Press. doi:10.1017/CBO9780511811340
11. Statute of the International Court of Justice. (1945). Article 38. Annex to the UN Charter.
12. Tomuschat, C. (2014). Human Rights: Between Idealism and Realism (3rd ed.). Oxford University Press.
13. Vienna Convention on the Law of Treaties. (1969). UN Doc. A/CONF.39/27.

Further Reading

• Slaughter, A.-M. (2004). A New World Order. Princeton University Press.