Cabinet Duration in Latin American Presidential Democracies
Unformatted Document Text:
shift their supporting parties.
It varies from 5 to 80.7, with a mean of 30.2.
The parliament features theorists argue that high fragmenatation, polarization and
electoral volatility reflect a complicated “bargaining environment,” and that the unstable
environment, in turn, often causes the dissolution of cabinets . However, the causality is not
evident because it is the institutional features of cabinets that directly affect cabinet duration,
while parliamentary features constitute the environment from which the cabinet is formed.
Fractionalized parliaments are likely to lead to multi-party cabinets that may increase the
instability of the cabinets. Moreover, polarized parliaments do not necessarily generate polarized
cabinets, and high electoral volatility can bring about stable coalitions due to the uncertainty
about the outcomes of the next election .
Number of parties in cabinet. Among cabinet parties only those who have seats in the
parliament are counted. Since the president needs the support of parliamentary parties to get his
bills passed, non-parliamentary parties can hardly be significant. According to Tsebelis (2002),
the more coalition partners the president has, the more difficult for him to respond to varying
policy demands. When the people’s dissatisfaction over policy immobilism increases, some
coalition partners may find an incentive to leave the cabinet. The number varies from 0 to 8, with
a mean of 2.4.
Ideological distance between cabinet parties is a unidimensional five-item scale (0-1
In addition to the fragmentation and polarization in the parliament, the electoral volatility is often used by the
parliament features theorists to estimate the effects of complex and unstable bargaining environment on cabinet duration .
Grofman and Roozendaal (1997) suggest a hypothesis that under the great uncertainty larger parties are more
likely to stay in the current coalition whereas smaller parties would have an incentive to get out and try their luck.
All the information on the cabinet attributes except for party ideology is obtained from the Acir Almeida’s dataset.
The information about the ideology of Latin American parties is obtained from Coppedge (1997).