Skepticism about international law abounds these days. A commentator in a national newsmagazine probably spoke for many when he wrote that international law is to law as professional wrestling is to wrestling: No one over the age of nine mistakes it for the real thing. International law has long had its critics, but in recent years they have seemed more numerous and included not only laypersons but specialists and diplomats. Meanwhile, its supporters express growing concern about its lack of clout. French president Jacques Chirac, for example, fears that the “law of the jungle” now prevails, and United Nations secretary general Kofi Annan has warned that we are “living through a crisis of the international system.” In an extraordinary news conference on July 30, 2003, Annan wondered aloud “whether the institutions and methods we are accustomed to are really adequate to deal with all the stresses of the last couple of years.” “What are the rules?” he asked.

Can it be that, 355 years after the Peace of Westphalia ended the Thirty Years’ War and established the principle of the sovereign equality of nations, the “rules” of the international system are still in doubt? In fact, most of the rules are not in doubt, and for the most part the international legal system functions effectively, regulating air travel, telecommunications, and the like. The problem, rather, is that the two categories of rules that are in doubt—rules about rules, and rules regarding security—are vitally important.

Rules about rules—so-called metarules—are foundational and shape the content of every legal system. They specify what qualifies as a “rule”—how the rules that govern day-to-day conduct are made and unmade. The rest of a legal system depends for its vitality and coherence on the strength of its metarules, and three particular metarules of international law provide especially weak support. These rules relate to the issues of consent, obligation, and causation.

First, consent. It’s commonly said that the international legal system is voluntarist, that is, that its rules are based on the consent of individual states. A state is not bound by any rule it does not accept. Thus, the system is grounded, ultimately, on self-restraint. Unless a state voluntarily restrains itself by consenting to be bound by a rule, it remains free to act in violation of the rule. This arrangement contrasts with the operation of domestic legal systems, which are based not on consent but on coercion. One can hardly decide that one will no longer be bound by the rule prohibiting bank robbery. A domestic legal system is voluntarist only in the sense that one can always leave it and relocate to a state with more congenial laws. In the international system, there’s no overarching authority. All states have an equal right to accept or reject rules. It’s sometimes claimed that this right of rejection exists only when a rule is first proposed, while it is in an inchoate state. But the whole logic of voluntarism undercuts this contention, for the notion of a consent-based system is meaningless if consent cannot be withdrawn in the same way it’s given. States have not consented to the elimination of their consent.

But a system grounded on self-restraint creates serious problems—to the point of raising doubts as to whether it can accurately be described as “law.” A leading international jurist, Judge Hersch Lauterpacht of the International Court of Justice, addressed the question in a narrower context in a 1957 case involving the validity of a state’s acceptance of a treaty subject to an unusual reservation. The reservation in question would have rendered the treaty applicable only when the reserving state desired it to be applicable. In Judge Lauter­pacht’s words, it would have left to the reserving state “the right to determine the extent and the very existence of its obligation,” with the result that the state would have “undertaken an obligation to the extent to which it, and it alone, consider[ed] that it had done so.” And this would have meant, the judge concluded, that the reserving state had “undertaken no obligation,” for an “instrument in which a party is entitled to determine the existence of its obligation is not a valid and enforceable legal instrument.” The treaty as modified would have lacked an “essential condition of validity of a legal instrument.”

Judge Lauterpacht would no doubt be surprised to find that his logic in this one case could be extended to apply to the entire international legal system. But because the system is consent based, every state maintains the right to determine “the very existence of its obligation.” The judge’s reasoning suggests, therefore, that all international legal “obligations” undertaken by states are illusory because an “essential condition” of law is missing. Absent genuine obligation rather than mere self-restraint, it’s hard to make the case that international law is really law.

U.S. domestic law rejects the notion that self-restraints are binding law. In constitutional law, a branch of the federal government cannot impose binding obligations on itself. For example, an executive order issued by President Gerald Ford, and still in effect, prohibits officials of the executive branch from engaging in assassination. Yet despite that executive order, President Bill Clinton ordered the assassination of Osama bin Laden. Though the earlier order had never been repealed, the later order simply superseded it. Self-restraints are not binding law.

This suggests a second systemic weakness of international law, deriving from the notion of obligation. The “glue that holds the system together,” it’s often said, is the rule that a state is bound to carry out treaties to which it is a party. But where does this rule to comply with treaties come from? In a consent-based system, from the states themselves. There’s no alternative. So states can reject this rule just as they can reject any other rule. Yet if states can turn their backs on the rule that requires compliance with all rules, where does that leave the system?

Again, to respond that states may not withdraw their consent from the rule requiring compliance with treaties would be to reject the voluntarist foundation on which the whole system is based and to necessitate some alternative, transcendent source of obligation—“some brooding omnipresence in the sky,” in the disparaging words of Oliver Wendell Holmes, Jr. Such an obligation would be moral, not legal, and its source would be unclear. Whether there exists a moral obligation to obey laws of human making is an important question—can a city council, for example, create a moral obligation to cross streets only in crosswalks?—but the question is moral, not legal.

The issue of obligation suggests a third systemic weakness, relating to causation. Inter­national law scholars have long been concerned about distinguishing what states do as a matter of legal obligation from what states do for other reasons—motivated, for example, by considerations of comity, courtesy, or simple self-interest. In assessing whether a given practice constitutes a norm of customary international law, therefore, international law has insisted upon some evidence that states have followed the practice in question because they have believed such conduct to be legally required. Traditional analysis, in other words, requires both a consistent state practice and a belief on the part of the state that the practice is obligatory as a matter of law. The belief must cause the conduct.

But the difficulty here is obvious. States, like individuals, seldom if ever act from a single intent. Conduct almost always flows from a tangled web of motives. Some international lawyers resolve this problem by assuming that if a rule exists and conduct consistent with the rule also exists, the rule must be the cause of the conduct. But such an inference is manifestly unjustified. If a city council adopted an ordinance requiring residents to brush their teeth daily, would it be accurate to ascribe the practice of daily toothbrushing to the new requirement imposed by law? In fact, it’s often impossible to separate self-interested behavior from behavior caused by legal requirements.

The International Court of Justice took a new crack at this conundrum in Nicaragua v. U.S.A (1986). The case arose after the United States mined Nicaragua’s harbors and otherwise provided support to the so-called contras, who were attempting to overthrow the Nicaraguan government. In the course of rejecting arguments that the conduct of the United States was lawful, the court considered the status of the underlying rule. “If a State acts in a way prima facie incompatible with a recognized rule,” the court said, “but defends its conduct by appealing to exceptions or justifications contained within the rule itself, then whether or not the State’s conduct is in fact justifiable on that basis, the significance of that attitude is to confirm rather than to weaken the rule.” Unfortunately, the court’s new approach is circular and self-referential. Some­times a breaching state may indeed agree with a rule that it violates. But again, there may be many reasons why a state appeals to “exceptions or justifications” contained within the rule other than an intent to confirm the rule. For example, the state may wholly object to a rule but appeal to an exception merely to avoid retribution. The assumption that the state’s intent is necessarily to “confirm” the rule is arbitrary. If the state has engaged in a prima facie violation of a rule, it’s more sensible to conclude that the state disagrees with the rule, not that it wants the rule strengthened.

These conceptual problems arise primarily in connection with customary international law, but they can also infect the application of treaty rules, for obligations imposed by treaties and customary international law often overlap. Consider once again the practice of assassination, which is commonly said to violate not only customary international law but also Article 2(4) of the UN Charter, prohibiting any use or threat of force against the territorial integrity or political independence of a state. States rarely engage in assassination, but what’s the proper inference to draw from their behavior? That assassination is legally prohibited? It’s possible that states forgo assassination for reasons related entirely to self-interest: Many may believe that the risks of retaliatory assassination are too great. The source of the rule may be treaty or custom, then, but it’s impossible to know whether the behavior in question represents compliance or coincidental conformance with the rule.

So is everything up for grabs in the international legal world? Hardly. As Columbia University law professor Louis Henkin has famously observed, “It is probably the case that almost all nations observe almost all principles of international law and almost all of their obligations almost all of the time.” My point is simply that the international legal order is far more fragile than most domestic legal regimes because it rests on a foundation of problematic metarules. Most of the time, the system works well enough because most states derive greater benefit from honoring day-to-day rules than from breaching them. Issues concerning the metarules do not arise, and international life proceeds. States that deviate from expected patterns of practice face reprisals. Sometimes, the consequences of divergence take the form of immediate diplomatic, economic, or military sanctions, and sometimes they’re reputational, with penalties long-term and indirect. Either way, violators suffer costs, even though those costs are imposed horizontally, at the hands of other actors within the system, rather than vertically, at the hands of some supranational authority.

Whether this is law, meaning a proper legal system, is, in many ways, beside the point. The real question is whether it works—whether the international legal system fulfills the functions that it’s intended to serve. And here the record is decidedly mixed. Some rules work much better than others. As Georgetown University professor Anthony Arend has pointed out, legal rules have a stronger impact on state behavior in areas of “low politics” that “do not strike at the core security concerns of states”—international trade, communication, and transit—than they do in the realm of “high politics,” where issues do touch on states’ core security concerns. On issues of high politics, consensus is much harder to obtain, and legal regulation is correspondingly more difficult. Accordingly, states are more apt to rely on themselves than on international institutions, for often their very survival is at stake. The determinants of state behavior in the realm of high politics tend to be the cultural, historical, and power-related factors that affect states’ calculations of their nerve-center security interests. In this realm, international rules are epiphenomenal, more effect than cause. So while it’s important to know that most states observe most rules most of the time, it’s equally important to realize that when some states violate some rules some of the time, those states are likely to be among the most powerful states, the rules are likely to be extraordinarily significant rules, and violations are likely to be highly visible and historically significant. Hence, the recent burst of skepticism about international law.

By their very effectiveness, the enormous body of international legal rules governing the quotidian dealings of states and nonstate actors—rules affecting such matters as finance and trade—have spun an increasingly tight web of interdependence and made globalization possible. But the fact that planes land, packages are delivered, and phone calls go through does not mean that the international legal order is operating as it should. The risks flowing from the failure of security rules are not lessened because many less important rules work. Though rules governing the use of force constitute only a small part of the international regulatory scheme, their dramatic collapse has overshadowed international law’s many small successes—and understandably so, for the stakes could hardly be greater. Until international law does a better job of tackling the large issues, doubts about it will persist.