The Treaty

A river can cross a border without acquiring a second source of water. The actions of the upstream state reach people whom its government does not represent, while the downstream state cannot settle the problem merely by writing a stronger domestic rule. Trade, communication, pollution, and shared resources repeatedly create this distance between the reach of an action and the reach of a government. The earlier Stranger encountered it at the edge of the commons. A treaty is one way of building an institution across that distance: states specify obligations, procedures, and expectations that their separate powers could not supply alone. The difficulty is to make the common arrangement hold while the authority to perform it remains distributed among its participants.

An international promise can have legal force even where enforcement lacks the concentrated form familiar within a state. The Vienna Convention on the Law of Treaties states that treaties in force bind their parties and must be performed in good faith. It also distinguishes those parties from third states, for which a treaty does not ordinarily create obligations or rights without consent; rules of customary international law remain a separate source of obligation. The scope of the instrument, the parties, and other applicable law therefore matter. A treaty's force cannot be inferred merely from the presence of a signature on any international document, and its obligation cannot be erased merely by observing that no world police officer stands beside the table. Legal commitment and the practical means of securing performance are connected questions that require separate answers.¹

Continuing benefits can help sustain cooperation when no superior government supplies the whole sanction. Imagine two states that each receive five units every period when both honor an arrangement. A state can gain eight in the current period by defecting while the other performs, but thereafter both receive two under an assumed permanent breakdown. With a discount factor of three-fifths, continuing cooperation is worth twelve and a half; the deviation and subsequent breakdown are worth eleven. At a discount factor of two-fifths, cooperation is worth about eight and a third, while deviation yields about nine and a third. The example isolates the value of the future. Its conclusion depends on observing defection, expecting the arrangement to continue, and finding the threatened response credible; the arithmetic alone supplies none of those conditions.²

Monitoring and procedures can make a promise usable by giving the parties a common account of what happened. A missed target may reflect deliberate evasion, faulty measurement, or a shock beyond the government's immediate control. An agreement that cannot distinguish these possibilities can turn an accident into an escalating dispute. Shared definitions, reporting, inspections where agreed, and a forum for testing evidence can reduce that uncertainty. A secretariat or tribunal may organize the process without possessing the powers of a domestic sovereign. Its authority, resources, and the consequences of its findings depend on the arrangement that establishes it. These institutions matter because a rule can fail through disagreement about performance as well as through a calculated refusal to perform.³

The gains from cooperation may need to be divided before cooperation is attractive to every indispensable participant. Suppose a joint arrangement gives one state a net benefit of ten and imposes a net cost of four on another, before any transfer. The combined gain is six, but the second state has no immediate material reason to join. A credible transfer of five changes the respective positions to gains of five and one. Finance, technical assistance, differentiated obligations, or a longer transition can serve related functions in real negotiations, though their effects depend on their design. The existence of a positive aggregate gain does not write the bargain. Someone must identify the participants, distribute the costs, and make the promised adjustment dependable enough to enter each side's decision.⁴

The state at the table contains people who may experience the agreement very differently. An export industry can gain while workers in another sector lose; an environmental benefit can arrive later than the expenditure needed to obtain it. Domestic authorization, public explanation, and implementation therefore connect the treaty to the Compact's question of legitimacy. Continuity also needs rules for amendment and exit. Under the Vienna Convention's framework, withdrawal and termination depend on the treaty's provisions or the Convention's applicable rules; a new government cannot simply treat its preference as the complete legal answer. An agreement that anticipates changed circumstances may preserve cooperation by providing a lawful path for revision. The practical strength of that path depends on whether the parties can use it before frustration destroys the arrangement they meant to improve.⁵

The treaty is an institution made between powers that remain distinct. It can supply common definitions, reciprocal commitments, information, adjudication, and a division of gains without becoming a single government. Some arrangements hold because they continually serve the parties; others depend on monitoring, domestic institutions, or the costs of losing wider relationships. Many use several of these supports together. The inquiry must follow the particular design instead of declaring international cooperation either effortless or impossible. The last plate moves to another place where rules can travel beyond the reach of a familiar public author: the software that ranks, admits, excludes, and executes. There the command may operate automatically, while the question of who must answer for it returns with undiminished force.