How Many Seats Does Your Party Have? It Depends on the Type of Voter

In most democracies, particularly in parliamentary systems like Italy, multiparty government is the norm. One of the central characteristics of this type of government is that it requires bargaining and compromise among parties with distinct electoral platforms. Coalition partners must agree on a common policy and must distribute important ministries and other perquisites of office. What is the nature of these bargains? How successful are parties in securing what they (and their voters) value, and what factors shape their ability to do so? These questions are important not only for academics and political analysts, but also for citizens as they make voting decisions, gauge the performance of their elected representatives, and evaluate the quality of governance.

In a recent study, "What You See is Not Always What You Get: Bargaining Before an Audience Under Multiparty Government," I and my coauthor Georg Vanberg (Duke University) contend that what parties achieve in coalition bargaining is partly a function of the political sophistication of their supporters. Party elites naturally worry about how any bargain they agree to will be perceived, and they are especially concerned about receiving payoffs from bargaining that are less than what supporters would see as the party's "fair share." Some of these supporters (like highly interested voters and lobby groups) follow politics closely, and are aware of the details of coalition negotiations. Others (including the majority of voters) are politically less knowledgeable, and are only aware of the most salient and easily observable aspects of coalition bargains.

We argue that-since most party supporters are not politically knowledgeable-parties face incentives to negotiate outcomes that reflect supporter expectations primarily on the most visible goods distributed in a bargain. One example of a visible good is the share of cabinet ministries received by parties. Recent work in political science has shown that voters are generally aware of this bargaining outcome and that they expect that each coalition party will receive a share of ministries that is roughly proportional to the number of legislative seats they contribute to the coalition. Thus, we predict that, since this good is visible to the bulk of voters and is easily compared against their ''yardstick'' expectations, parties will insist on receiving a share of ministries that is proportional to their relative size.

At the same time, such an outcome will typically not align with the underlying bargaining strength of the parties. As a result, we argue, party elites will use "side payments" of less visible goods to reconcile the overall coalition bargain with relative bargaining power. That is, parties that are "undercompensated" (relative to their bargaining strength) with respect to the numerical share of ministries will be remunerated by their "overcompensated" partners through concessions on more complex aspects of the bargain. We develop an empirical measure of a complex bargaining outcome that corresponds to the ''policy risk'' associated with a particular allocation of ministries. This risk arises because ministers can use their considerable agenda powers to implement policies in their jurisdictions that violate the coalition agreement. The policy risk for a party goes up if the ministries that are not under its control go to coalition partners with very different policy positions in issue areas the party cares a lot about. Our findings, based on an analysis of 308 bargaining situations in 16 democracies, provide support for our argument. To attain the numerical proportionality expected by politically unsophisticated supporters, an "overcompensated" party is generally willing to allow its undercompensated partners to control ministries on important dimensions on which the preferences of the parties significantly diverge.