A good economic institution will have an efficient equilibrium and, equally important, will promote rapid convergence to that equilibrium. In this paper we introduce a general approach--landscape learning--for comparing convergence properties across institutions.

We consider three different liability rules: negligence with contributory negligence, comparative negligence and no-fault. We demonstrate theoretically and then corroborate experimentally that comparative negligence promotes the fastest rate of convergence. Along the way we also test various hypotheses regarding learning and other adjustment dynamics. Thus our paper extends the notion of institutional choice from a mere static one -- liability rules with efficient equilibria are chosen ó to a more dynamic perspective --rules that rapidly achieve efficient equilibria are chosen.
 
 

 In recent elections, voters supported initiatives to limit the number of terms that their representatives may serve, yet at the same time, overwhelmingly re-elected their incumbents. We provide a theoretical explanation for this and other puzzles associated with voting on term limitations. The pattern of voting on term limits can be explained by the desire to redistribute power from one party to another, from one branch of government to another, and from districts with long-term incumbents to districts whose representatives have served only for a short time span. We test these hypotheses by looking at voting patterns on California Proposition 140 and the vote on the 22nd Amendment with generally positive results.
 

D72, H10
 

Back to home page