not commit on the basis of what they will do in the future (Barron 1991). If a coalition
party behaves contrary to the terms of the government agenda after the vote of
confidence, then there is not a mechanism to punish the party until the next formation of
a government.
Laver and Shepsle (1996, 1998) provide the most extensive analysis using non-
cooperative formal models. They model how executives are created by parties and,
subsequently, control policy. The policy outcomes associated with any government can
be predicted based on policy preferences of those in cabinet positions (Laver and Shepsle
1990). The credibility of a potential government rests of the proposed ministerial
portfolios. Their approach focuses on the role of a strong party and its capacity to
allocate portfolios. Strong parties are viable players in the government formation
process. A merely strong party is in a position to dominate the cabinet making process
because no alternative to a single party cabinet exists to which they cannot veto. A very
strong party can form a cabinet on its own, whether or not it commands a parliamentary
majority. In general, strong parties are those that have ideal points which are central in
relation to those of other parties. Very strong parties sit at the dimension-by-dimension
median in a multi-dimensional policy space, whereas merely strong parties sit at the
median on at least one policy space. The existence of strong parties in the government
formation process, due to their veto power, allows for a more structured policy voting
process.
Institutional Influences on Coalition Politics
Recently scholars have turned their attention to how institutions influence
outcomes in parliamentary systems. These institutional models examine how the formal
9
Convention