The Moment You Realise Some Bodies Actually Have Teeth
Somewhere in a trade ministry, a lawyer pulls up a ruling from a panel of three arbitrators sitting in Geneva. The ruling is twelve years old, names a country that no longer has the same government, and yet it is being cited as binding precedent in a dispute worth several hundred million dollars. Nobody voted on this. No army enforced it. And yet the minister's first instinct is compliance, not defiance.
That is what genuine institutional authority looks like. Not a headquarters building. Not a secretary-general giving a press conference. The real question, the one political scientists have circled for decades, is why some international institutions accumulate that kind of gravitational pull while others produce documents that get filed and forgotten.
The short answer: enforcement architecture, credible exit costs, and what scholars sometimes call the "density of commitments." The short answer is almost useless without the mechanics.
The Trap States Walk Into (Willingly)
The single most reliable predictor of whether an international body develops real authority is whether it can raise the cost of defection over time. This is not about punishment in any dramatic sense. It is about entanglement.
Take the World Trade Organization's dispute settlement mechanism. When a country joins, it accepts a system in which trade grievances are adjudicated by panels whose rulings, once passed by the Appellate Body, carry automatic standing unless every member votes to reject them. That inversion, requiring consensus to block rather than consensus to adopt, was a deliberate design choice in the 1994 Marrakesh Agreement. It sounds procedural. It is actually the load-bearing wall.
Over time, countries bring thousands of disputes. They win some, lose some. They cite prior rulings in new cases. A body of case law accumulates that no single government wrote but that every government now has to reckon with. Defying one ruling means implicitly threatening the rulings you won. The trap is elegant: states walk into it because the early benefits (market access, predictable rules) are large and immediate, while the sovereignty costs compound slowly and feel abstract until they don't.
Contrast that with the United Nations General Assembly. One country, one vote, resolutions passed by simple majority, no binding enforcement mechanism on substantive matters. The result is an institution that can generate extraordinary moral pressure in specific circumstances, but whose formal outputs, the resolutions themselves, are routinely ignored by powerful states without consequence. Syria, Myanmar, Western Sahara: the General Assembly has passed resolutions on all of them. The situations have not resolved because of those resolutions. The body is not worthless. It is a barometer and a legitimating forum. But it is not a power centre.
Two Countries, Same Treaty, Different Outcomes
Here is a concrete illustration of how this divergence plays out. Imagine two mid-sized economies, call them Arvenia and Belkora, that both ratified the same investment protection treaty in the same decade. The treaty includes an investor-state dispute mechanism allowing foreign companies to sue governments before international arbitration panels.
Arvenia, over fifteen years, loses three significant arbitrations. Each time, the government pays. Not because tanks appeared, but because Arvenia wants to keep issuing sovereign bonds, and bond markets price political risk partly through sovereign compliance scores tracked by agencies that monitor arbitration outcomes. The reputational cost of non-payment is not theoretical. It shows up in basis points on the next debt issuance. Arvenia has, in effect, been enrolled into a system it can only exit by paying a price steeper than compliance.
Belkora, by contrast, refuses to pay after its first loss, citing constitutional constraints. International arbitration panels have no bailiffs. The winning company lobbies its home government, which raises the matter in bilateral trade talks, which stalls a preferential tariff negotiation Belkora wanted. The linkage is not automatic, not clean, and takes four years to bite. But it bites. Belkora eventually settles.
The institution did not enforce either outcome directly. What it did was structure the incentives so that every actor with a stake in Belkora's behaviour, bond investors, trading partners, foreign companies considering future investment, had a shared interest in compliance. That is the real enforcement mechanism of effective international institutions: distributed, interest-aligned pressure, not centralised coercion.
What People Get Wrong About "Toothless" Bodies
The standard critique of international institutions, that they are talking shops dressed in marble, misses something important. And it misses it badly. Ceremonial institutions are not failures. They are often the product of deliberate design by states that wanted the appearance of cooperation without its substance.
The League of Nations is the textbook case. The unanimity requirement for Council decisions meant that any major power could veto action against itself. This was not an oversight. It was the price of American, British, and French participation, each of which wanted a forum for managing disputes short of war while retaining the right to ignore inconvenient outcomes. When Japan invaded Manchuria and the Council could not act, the League did not malfunction. It functioned exactly as designed.
So ask yourself: how many institutions celebrated as breakthroughs are, in fact, this? Dressed-up non-commitments that great powers can admire from a safe distance?
The better question is not "why do some institutions fail?" but "who benefits from institutional weakness?" The answer is usually the states powerful enough to get their preferred design built in at the founding. Great powers tend to prefer bodies where their veto is structural. Smaller states, paradoxically, often push for stronger formal mechanisms precisely because those mechanisms constrain the behaviour of larger players who would otherwise simply impose outcomes bilaterally.
The European Court of Justice is perhaps the most striking counter-example in modern history. Member states created a court that could, and eventually did, assert the supremacy of European law over national law, a development that several founding governments almost certainly did not fully anticipate. The court accumulated authority incrementally, each ruling building on the last, enlisting national courts as partners in enforcement, until the architecture of constraint was so dense that unwinding it became unthinkable. The EU is exceptional precisely because it achieved that density. It is also, not coincidentally, the institution that largest-state defection has cost the most.
The Density Problem
Density is the concept most guides skip. An institution becomes genuinely authoritative not when it passes its first binding resolution but when enough separate commitments, legal, financial, reputational, diplomatic, become interlocked that a state choosing to defect faces costs across multiple domains simultaneously.
Think of it less like a chain and more like a root system. Pull one root, and the others hold. The International Criminal Court struggles with authority in part because its enforcement depends on a single mechanism: state cooperation in arresting indicted individuals. States that decline to cooperate face diplomatic censure but rarely anything that bites harder. The root system is shallow.
Shallow, and by design.
The Basel Committee on Banking Supervision, by contrast, sets capital adequacy standards that are technically non-binding. No treaty. No enforcement panel. And yet major banks in non-member countries largely comply, because their access to correspondent banking relationships, dollar clearing, and foreign capital markets depends on being seen as regulated to international standards. The density is provided not by the committee itself but by the financial system it sits inside. The committee is the label on the jar. The jar is the global banking architecture, and nobody sane smashes that.
Checking whether an institution has that kind of density is more useful than counting its enforcement clauses.
States that want accountability from international bodies should ask a harder question than "does this institution have a court?" The right question is: what happens to a powerful state, specifically a powerful state, when it defects? If the answer is mostly embarrassment, the institution is decorative. If the answer involves costs that compound across trade, finance, and diplomacy simultaneously, you are looking at something with real weight. The difference is almost never accidental, and it is almost never neutral. Someone chose it.