A GAME THEORETIC MODEL OF REFORM IN LATIN-AMERICAN DEMOCRACIES

In this article I develop a simple game-theoretic model of administrative reform in Latin American democracies. The model, which is based on the incentives facing the politicians who must initiate reforms if any are to occur, yields two predictions: (1) reforms are more likely to pass the legislative hurdle when patronage is evenly distributed among the strongest parties, and (2) initial reforms are more likely to be followed by further extensions of reform where the electoral weight of the top parties remains relatively even and stable. Attention to the incentives facing legislators and party leaders also results in the expectation that certain political institutions, such as open list proportional representation and electoral rules that minimize party control over candidate lists, reduce the probability of reform. I test these predictions and expectations on a set of Latin American democracies and find them consistent with historical events. © 1991, American Political Science Association. All rights reserved.