efficiency in an attempt to solve this problem(Rubin, 1977). The evolutionary models are attempts to explain judicial behavior without resort to preferences or utility functions Initially, these models aimed at explaining Posner's putative observation that the common law was efficient. It is fair to say that the models failed in this endeavor, perhaps because the law is not so efficient as posner argued. nonetheless these models have had an important impact on the literature because they have called attention to forces other than judicial preferences in explaining the law. The evolutionary models are ultimately based on the model of the litigation process first set forth by Landes(1971) In the first paper applying an evolutionary model to the common law, I argued that most cases are settled, rather than litigated, and that it is only litigated cases that can lead to legal change. Cases are settled when the expected value to the plaintiff of a case is less than the expected cost to the defendant, which is generally true if stakes are symmetric. However, inefficient laws can sometimes create asymmetric stakes because the inefficiency means that there are deadweight losses than cannot be bargained away in the settlement process. That is, an inefficient rule creates a loss to one party that is greater than the gain to the other because of future stakes in similar type cases. Thus, litigation becomes more likely when rules are inefficient, and so inefficient rules are subject to greater selection pressure, and more likely to be overturned. (Note that this model, like many of its successors, depends on parties having ongoing interests in disputes of a certain sort, rather than merely in the matter at hand. Following this initial contribution were several extensions and modifications Priest(1977)argued that inefficient rules generated larger stakes and so were more likelyefficiency in an attempt to solve this problem (Rubin, 1977). The evolutionary models are attempts to explain judicial behavior without resort to preferences or utility functions. Initially, these models aimed at explaining Posner's putative observation that the common law was efficient. It is fair to say that the models failed in this endeavor, perhaps because the law is not so efficient as Posner argued. Nonetheless, these models have had an important impact on the literature because they have called attention to forces other than judicial preferences in explaining the law. The evolutionary models are ultimately based on the model of the litigation process first set forth by Landes (1971). In the first paper applying an evolutionary model to the common law, I argued that most cases are settled, rather than litigated, and that it is only litigated cases that can lead to legal change. Cases are settled when the expected value to the plaintiff of a case is less than the expected cost to the defendant, which is generally true if stakes are symmetric. However, inefficient laws can sometimes create asymmetric stakes because the inefficiency means that there are deadweight losses than cannot be bargained away in the settlement process. That is, an inefficient rule creates a loss to one party that is greater than the gain to the other because of future stakes in similar type cases. Thus, litigation becomes more likely when rules are inefficient, and so inefficient rules are subject to greater selection pressure, and more likely to be overturned. (Note that this model, like many of its successors, depends on parties having ongoing interests in disputes of a certain sort, rather than merely in the matter at hand.) Following this initial contribution were several extensions and modifications. Priest (1977) argued that inefficient rules generated larger stakes and so were more likely 3