The Effective Number of Parties At Four Scales

Abstract

Different scholars count `the number of parties' in different ways, partly because they examine different parts of the political process and for different reasons. Sartori's qualitative approach focused mainly on competition for government, but is now largely supplanted by the quantitative `effective number of parties' index, which deals with votes and seats. But some quantitative research requires us to look beyond votes and seats. For example, testing proportional and pluralitarian models of democracy requires us to count parties not just in terms of votes and seats but also in terms of legislative and cabinet power. I estimate party shares of legislative and cabinet power to test democratic norms at British and German elections and parliaments. So, the effective number of parties index needs to be, and can be, extended beyond votes and seats alone.