Search

Observing the destructive impact of the seventeenth century European wars of religion, Thomas Hobbes concluded that man by himself is incapable of preserving peace and order. He assumed that peace [...]

A widely accepted argument among scholars of Latin American presidential regimes is that interbranch cooperation is impaired when the president's party falls short of a majority of seats in the legislature. This argument fails to consider three factors that affect the performance of minority presidents: the policy position of the president's party, the president's capacity to sustain a veto, and the legislative status of the parties included in the cabinet. This article argues that the greatest potential for conflict in a presidential regime occurs when the president's party lacks the support of both the median and the veto legislator and no cabinet coalition holding a majority of legislative seats is formed. This hypothesis is supported using data on executive-legislative conflicts and on interrupted presidencies in Latin America during the period 1978-200 [ABSTRACT FROM AUTHOR]

Students of presidential regimes claim that while the combination of plurality rule for presidential elections and concurrent electoral cycles favors bipartism, majority rule for electing presidents favors multipartism. I argue that a reverse causality also affects the relationship between party systems and electoral systems. Using a bargaining model of institutional change, I propose that while dominant and large parties are likely to choose plurality rule and concurrent elections, small parties are likely to choose majority rule. I also argue that military rulers and military-civilian coalitions tend to follow the logic of electoral choice of small parties. These hypotheses are supported by a statistical analysis of the determinants of electoral choice in 49 cases of constitutional change in Latin America. Mechanisms of choice are analyzed in several episodes of electoral reform, including a negative case that suggests explanations of electoral choice not covered by the model presented in this paper. [ABSTRACT FROM AUTHOR]