formation, but they do find support to suggest that upper chambers are more likely to
have surplus majorities in order to obtain a government majority in the lower chamber. I
expect similar outcomes when this comparison is made in a future iteration of this
research.
The second hypothesis considers the role of an additional chamber and how it
increases the life of a cabinet. The data from the upper chamber compositions and status
will be included in a model of government survival (Warwick 1992, 1994). DT includes a
dummy variable to the model which indicates the presence or absence of a government
majority of seats in the upper chamber (2002: 765). Other variables in the model include
majority status in the lower chamber, post-election status of the cabinet, investiture vote
requirement, returnability, left-right diversity, clerical-diversity, and regime-support
diversity. The model will be tested with and without the dummy variable in order to see
how Central and Eastern European cabinets conform to the previous research on Western
Europe in addition to testing the bicameralism effect. I expect the results to reflect the
findings of Druckman and Thies: the length of a government increases dramatically when
a government has a majority in both chambers. The final hypothesis tests whether
bicameral systems are different from unicameral systems in terms of the number of
surplus majority governments. I expect surplus majorities to be more frequent in
bicameral systems. I plan on analyzing the data using Cox’s partial likelihood method to
determine how concurrent majorities impact the hazard rate of governments in Central
and Eastern Europe. Finally, I will test the third hypothesis by seeing if there is a
statistically significant difference in means between the percentages of surplus majorities
in bicameral versus unicameral parliamentary systems. The expected results from these
tests are provided in the next section.
19
Convention