weaken a precedent without overturning it completely. Thus a party with an interest in overturning an inefficient precedent would also have to consider the possibility that litigation could strengthen as well as weaken the precedent. This consideration greatly weakens the evolutionary pressures for efficiency. Parsons(1983)combined Priest's oint( that inefficient precedents would lead to increasing litigation) with the Landes and Posner point(that precedents that were litigated might become entrenched) to argue that here is a tendency for the common law to become"reckless"to favor rules that inefficiently lead to increased accidents. Cooter and Kornhauser(1980) present a complex evolutionary model in which there are some tendencies towards efficiency, but in which both efficient and inefficient rules will be observed at any time. This model and alternative definitions and implications of efficiency, are discussed in Kornhauser (1980). Von Wagenheim(1993) presents a model with similar results Hirshleifer(1982), building on my discussion of inefficiency when stakes in precedent are asymmetric, provided what may be one of the most useful and influential criticisms of the evolutionary models. Recall that in my original model and in some others, including Goodman(1979)and Landes and Posner(1979), evolutionary forces moved the law towards efficiency only if the party with an interest in efficiency had an ongoing interest in the form of he law. Hirshleifer generalized this point to show that the law could come to favor whichever party could most easily organize and mobilize resources for litigation of unfavorable precedents. This movement would be independent of efficiency. I used this point to argue that common law was more like statute law than many want to admit interest groups could use either common or statute law to achieveweaken a precedent without overturning it completely. Thus a party with an interest in overturning an inefficient precedent would also have to consider the possibility that litigation could strengthen as well as weaken the precedent. This consideration greatly weakens the evolutionary pressures for efficiency. Parsons (1983) combined Priest's point (that inefficient precedents would lead to increasing litigation) with the Landes and Posner point (that precedents that were litigated might become entrenched) to argue that there is a tendency for the common law to become “reckless”—to favor rules that inefficiently lead to increased accidents. Cooter and Kornhauser (1980) present a complex evolutionary model in which there are some tendencies towards efficiency, but in which both efficient and inefficient rules will be observed at any time. This model, and alternative definitions and implications of efficiency, are discussed in Kornhauser (1980). Von Wagenheim (1993) presents a model with similar results. Hirshleifer (1982), building on my discussion of inefficiency when stakes in precedent are asymmetric, provided what may be one of the most useful and influential criticisms of the evolutionary models. Recall that in my original model and in some others, including Goodman (1979) and Landes and Posner (1979), evolutionary forces moved the law towards efficiency only if the party with an interest in efficiency had an ongoing interest in the form of he law. Hirshleifer generalized this point to show that the law could come to favor whichever party could most easily organize and mobilize resources for litigation of unfavorable precedents. This movement would be independent of efficiency. I used this point to argue that common law was more like statute law than many want to admit: interest groups could use either common or statute law to achieve 5