the custom of"free ships, free goods, " whereby all property on neutral ships is immune from seizure (including enemy property ) is at times illustrative--powerful states may espect the principle because the seizure of neutral ships to capture enemy property is not worth the bother, while weaker states may respect the principle for fear of retaliation by powerful states. The third possible reason for convergence arises when a common practice represents the solution to an iterated Prisoners Dilemma, which can be sustained over time by states with open-ended time horizons and sufficiently low discount rates They again offer ambassadorial immunity as a possible illustration, suggesting that an exc ange of ambassadors amounts to an exchange of hostages, and that the prospect o retaliatory acts against ones own ambassador can dissuade any temptation to interfere with the ambassadors of others. Finally, convergence may arise in the face of a pure coordination problem or otherwise where a"focal point is useful. They suggest that the convergence on a three-mile limit for territorial waters is an example here. For a variety of reasons including security, nations have an interest in claiming dominion over waters along their coast but the exact limits of territorial waters is to a degree a matter of indifference--a three mile limit supplies a focal point that all nations can accept In short, Goldsmith and Posner argue that convergence in state practice occurs for reasons of pure national self interest, albeit not the same reason every time. They further suggest that continued adherence to customary practice happens because the self- interested reasons for convergence remain in place, not because of any independent sense of legal obligation. Opinio juris, they suggest, is a fiction, and what legal scholars refer to law"" is really no more than a descriptive account of certain regularities in the behavior of states To bolster this latter claim goldsmith and posner document how ostensible rules of customary law are frequently violated when states have an interest in deviating. They further illustrate how rogue states, which they suggest have shorter time horizons and higher discount rates, are more likely to deviate than others. Because historical violations and breakdowns of custom can be linked to self-interested reasons for them Goldsmith and Posner find anecdotal empirical support for the claim that customary practices are mere regularities of self-interest, and that customary law per se exerts no tug on state behavior The proposition that customary international law'" generally emerges from the self-interested interaction of states, and that it promotes with their mutual interest for one reason or another, seems rather unremarkable. It would indeed be odd if a customary practice emerged on a large scale that made its adherents worse off over an extended period of time. While this aspect of Goldsmith and Posner's analysis seems compelling,athe custom of “free ships, free goods,” whereby all property on neutral ships is immune from seizure (including enemy property), is at times illustrative—powerful states may respect the principle because the seizure of neutral ships to capture enemy property is not worth the bother, while weaker states may respect the principle for fear of retaliation by powerful states. The third possible reason for convergence arises when a common practice represents the solution to an iterated Prisoner’s Dilemma, which can be sustained over time by states with open-ended time horizons and sufficiently low discount rates. They again offer ambassadorial immunity as a possible illustration, suggesting that an exchange of ambassadors amounts to an exchange of hostages, and that the prospect of retaliatory acts against ones own ambassador can dissuade any temptation to interfere with the ambassadors of others. Finally, convergence may arise in the face of a pure coordination problem or otherwise where a “focal point” is useful. They suggest that the convergence on a three-mile limit for territorial waters is an example here. For a variety of reasons including security, nations have an interest in claiming dominion over waters along their coast, but the exact limits of territorial waters is to a degree a matter of indifference—a three mile limit supplies a focal point that all nations can accept. In short, Goldsmith and Posner argue that convergence in state practice occurs for reasons of pure national self interest, albeit not the same reason every time. They further suggest that continued adherence to customary practice happens because the self￾interested reasons for convergence remain in place, not because of any independent sense of legal obligation. Opinio juris, they suggest, is a fiction, and what legal scholars refer to as customary “law” is really no more than a descriptive account of certain regularities in the behavior of states. To bolster this latter claim, Goldsmith and Posner document how ostensible rules of customary law are frequently violated when states have an interest in deviating. They further illustrate how rogue states, which they suggest have shorter time horizons and higher discount rates, are more likely to deviate than others. Because historical violations and breakdowns of custom can be linked to self-interested reasons for them, Goldsmith and Posner find anecdotal empirical support for the claim that customary practices are mere regularities of self-interest, and that customary law per se exerts no tug on state behavior. The proposition that “customary international law” generally emerges from the self-interested interaction of states, and that it promotes with their mutual interest for one reason or another, seems rather unremarkable. It would indeed be odd if a customary practice emerged on a large scale that made its adherents worse off over an extended period of time. While this aspect of Goldsmith and Posner’s analysis seems compelling, a