their goals(Rubin, 1982). I argued that the apparent efficiency of the common law was because most common law was developed at a time when organization of interest groups was expensive, and that more recently both common and statute law have been subject to interest group pressures. Crew and Twight(1990)expanded on this point and found common law less subject to rent seeking than statute law. Rowley and Brough(1987) and Barzel (2000) find that contract and property might be expected to be efficient, but not tort Martin bailey and i extended this theme in a formal model of the influence of interest groups on the law Bailey and rubin, 1994), and we have shown that plaintiffs attorneys have been responsible for the shape of modern tort law, using an evolutionary mechanism to shape the law Rubin and Bailey(1994). In Rubin, Curran and Curran (2001)we modeled an interest group deciding whether to use litigation or lobbying as a method of rent seeking: Osborne(2002)presents a similar model. Fon and Parisi(2003) provide an additional mechanism to explain expansion of tort(mainly product liability) law: since plaintiffs chose courts in which to file, judges who are in favor of expansive law will see more cases and have more influence than more conservative judges Although shaped by evolutionary forces, this law is socially quite inefficient An interesting set of hypotheses regarding legal evolution is in Roe(1996). Roe argues that the notion of evolution towards efficiency is an important determinant of lega form, but it is not the only determinant. He incorporates three subsidiary notions into an efficiency framework- the importance of initial conditions(borrowed from chaos theory ) path dependence, and evolutionary accidents. The result of these processestheir goals (Rubin, 1982). I argued that the apparent efficiency of the common law was because most common law was developed at a time when organization of interest groups was expensive, and that more recently both common and statute law have been subject to interest group pressures. Crew and Twight (1990) expanded on this point and found common law less subject to rent seeking than statute law. Rowley and Brough (1987) and Barzel (2000) find that contract and property might be expected to be efficient, but not tort. Martin Bailey and I extended this theme in a formal model of the influence of interest groups on the law (Bailey and Rubin, 1994), and we have shown that plaintiffs' attorneys have been responsible for the shape of modern tort law, using an evolutionary mechanism to shape the law Rubin and Bailey (1994). In Rubin, Curran and Curran (2001) we modeled an interest group deciding whether to use litigation or lobbying as a method of rent seeking; Osborne (2002) presents a similar model. Fon and Parisi (2003) provide an additional mechanism to explain expansion of tort (mainly product liability) law: since plaintiffs chose courts in which to file, judges who are in favor of expansive law will see more cases and have more influence than more conservative judges. Although shaped by evolutionary forces, this law is socially quite inefficient. An interesting set of hypotheses regarding legal evolution is in Roe (1996). Roe argues that the notion of evolution towards efficiency is an important determinant of legal form, but it is not the only determinant. He incorporates three subsidiary notions into an efficiency framework - the importance of initial conditions (borrowed from chaos theory), path dependence, and evolutionary accidents. The result of these processes 6