Given the ubiquity of multiparty government in parliamentary democracies, the study of coalition bargaining has attracted considerable scholarly attention. Most work in this tradition models coalition negotiations as a bargaining game among party elites.
In the current paper, we move beyond conventional models to highlight the fact that coalition bargaining takes place before an audience: parties must consider the fact that the bargain they strike will be evaluated by their supporters, and that they may be punished or rewarded according to voters’ perceptions of each party’s efficacy in representing their interests.
The key feature of the theoretical model we develop is that the ability of voters to evaluate the coalition bargain differs systematically across different kinds of “goods” over which parties bargain.
The central implication of our argument is that easily observable and quantifiable outcomes of coalition bargaining (such as the number of ministries government parties receive) are likely to correspond to the proportion of legislative seats parties contribute to the coalition - consistent with the well-known empirical regularity known as Gamson’s Law - but that outcomes that are more difficult to observe and quantify (such as the “quality” of ministries parties receive, or the policy of the government implied by the distribution of ministries) will typically be less proportional to seat share.
We evaluate our argument empirically by examining the distribution of government ministries in 257 situations of coalition bargaining in 16 parliamentary democracies.