(Page 9 of 29) - A Combination of Methods. The Way Forward in Coalition Research authored by B�ck, Hanna. and Dumont, Patrick.

their ideological range (see e.g. Leiserson 1966; De Swaan 1973). Yet another
policy-based prediction can be derived from the median voter theorem. Applied to
coalition bargaining this theorem suggests that the party that controls the median
legislator position will have increased bargaining power, since there is no policy
position that can be implemented that is preferred by a majority of legislators to the
median legislator’s ideal policy position (assuming that the policy space is one-
dimensional). From this argument, we derive the prediction that potential
governments are more likely to form if they include the median party.

A more recent account about coalition formation is the idea that incumbent

cabinets are favored in negotiations since they represent “the reversion point in the
event the other parties fail to agree on an alternative” (Strøm et al. 1994: 311).
Alternative arguments for why we should expect incumbent cabinets to be
advantaged have recently been presented in the coalition literature. For example,
some authors argue that it is less costly to keep the same partners, for instance
since parties who have cooperated before should have established routines for
governing together. Changing partners implies governing with the opposition,
which may have been harsh in the previous term, and a change of partners may
even imply governing with newcomers, thereby increasing uncertainty with regard
to trust. Transaction costs are thus likely to increase when parties change partners,
which implies that incumbent parties are more likely to be in government (see e.g.
Bäck 2003; Dumont 2004).

From this idea we draw the hypothesis that potential

governments are more likely to form if they are incumbent cabinets.

Data and operationalizations

We will in this paper use a data set consisting of a large number of formation
opportunities in Western Europe between 1970 and 2000. There are a number of
reasons for concentrating on this post-1970 period. First, as Budge and Bara (2002:
53) argue, “the second half of the 1960s formed the great watershed of the post-war
era in terms of society and economics as well as (party) politics. All recent political
developments emerged around or after this date.” During this time, dimensions of
political competition were altered, as the New Politics and post-materialism theses
(Inglehart 1977, 1990, 1997) argue, reducing the relevance or transforming the
meaning of the left-right opposition, and allowing the electoral success of new
actors. Aside from such substantive reasons, restricting our sample to post-1970s
formation situations enables intensive research of any case present in our data set,
as we maximize the chances of finding actors to interview or rely on multiple
media reports. Hence, from a methodological point of view, as the large-n analysis
is our starting point for selecting cases to be intensively studied, these cases have to
be present in the data set constructed for this large-n analysis in the first place.

The countries and years included in our data set are: Austria 1970–2000,

Belgium 1972–1999, Denmark 1971–1998, Finland 1970–1999, Germany 1972–
1998, Iceland 1971–1999, Ireland 1973–1997, Italy 1972–1993, Luxembourg
1974–1999, Netherlands 1971–1998, Norway 1973–1997, and Sweden 1970–