Accountability in Coalition Governments

Abstract

Most theoretical studies of coalition politics have focused on selection, rather than accountability: Coalition partners are selected according to the proximity of their positions in a Euclidean policy space. This proximity, together with institutional attributes of the party systems or the coalitions, serves also to explain the duration of the coalition. Empirical studies of retrospective voting, often with little connection to accountability theory, have generally concluded that the political survival of coalitions is considerably independent from elections. Such results, however, refer to governments as a whole. In this work, voters allocate rewards and punishments for past outcomes focusing on the prime ministers and their parties. If differences in clarity of responsibility exist, they do not seem to produce greater economic accountability of single-party governments—it is similarly limited under both coalition and single-party governments. Coalitions, however, increase the risks of losing office due to political crises, rather than elections. Prime ministers can respond to challenges by reshuffling the government or the coalition. Because such crises are launched under economic conditions that improve the welfare of citizens, coalitions may undermine democratic accountability.