Minority Government and Coalition Conflict
Minority governments hold fewer than 50 per cent plus one seat in parliament. Based on data from the ParlGov database (Doring and Manow 2011), I code the parliamentary seat share of cabinet parties and distinguish minority from majority governments at the beginning of each reform opportunity.
Coalition conflict can take place along a left-right or a pro-anti integration dimension. As discussed both types of conflict could be relevant for EU-related institutional reforms in member state parliaments. Operationally, one could measure coalition conflict in different ways. One option would be to take the range of the ideological positions on either one of the two dimensions. However, doing so would indicate that the only relevant coalition disagreements lie between the two most extreme members of the cabinet, even if, for instance, one party is extremely small compared to the other, or holds positions far away from what is otherwise an ideologically cohesive alliance. An alternative measure could be to take the standard deviation, which takes into account each party's deviation from the mean cabinet position. The standard deviation does not take into account that parties contribute different amounts of parliamentary seats to the legislative majority. It seems likely, however, that conflict in the coalition is more difficult to manage, and thus more likely to encourage the formation of additional monitoring mechanisms in the parliament, if large parties disagree with the rest as opposed to small parties. Consequently, coalition conflict will be measured as the seat-weighted standard deviation, respectively from the average left-right and pro-anti integration cabinet position.
Information on party positions on the pro-anti integration dimension comes from the CH-LR surveys discussed above. These surveys also ask experts to indicate the 'position of the party in 2010 in terms of its overall ideological stance', where 0 is 'extreme left' and 10 is 'extreme right'. Even though the CH-LR surveys also contain partisan left-right positions, they cover only the time period from 1999 onwards. In order to be able to analyse parliamentary reform opportunities over a longer time period, I use the left- right scale provided by the Comparative Manifesto Project (CMP) (Volkens etal. 2012; Budge etal. 2001). The CMP scale measures the share of negative or positive sentences parties include in their election manifestos (out of all sentences) on a range of topics of relevance to the left-right dimension of political conflict.