As first year law students unhappily discover, the meaning of "law" is frustratingly protean, shifting by usage and user. Depending on whom you ask, law is a system of rules, a body of precedents, a legislative enactment, a collection of norms, a process by which social goals are pursued, or some dynamic mixture of these. Law's principal purpose is to define and protect individual rights, to ensure public order, to resolve disputes, to redistribute wealth, to dispense justice, to prevent or compensate for injury, to optimize economic efficiency, or perhaps to do something else. And yet one thing is irreducibly clear: at its most basic, every legal system exists to effect some change in human behavior. That is, law is a lever for moving human behavior. The very obviousness of this proposition obscures its significance. The principal implication is this: law depends on a behavioral model as a lever depends on a fulcrum. Only a behavioral model, which purports to explain why people behave as they do, can suggest that if law moves this way behavior will move that way. This means that the success of every legal system necessarily depends, in part, on the solidity-that is, the accuracy and predictive power-of the behavioral model on which it rests.