Does executive autonomy reduce second-order election effects?

Abstract

The second-order election (SOE) model expects voters to punish parties innational government and reward opposition, small and new parties becausethere is ‘less at stake’ in an SOE. One key assumption that is rarely studied iswhether SOE-effects depend on the extent to which voters can impact theselection of the executive in an SOE. This article argues that executiveautonomy – i.e. the extent to which executives are independent fromthe parliament regarding their formation, termination and execution of theircompetences – increases the impact of authority. Executive autonomyreduces SOE-effects when authority is high but increases SOE-effects whenauthority is low. An empirical analysis of 41,603 vote share swings for 4733parties competing in 2665 elections held in 282 regions in 14 Europeancountries between 1945 and 2019 confirms the hypotheses. These resultshave important implications for electoral democracy and party competitionat the regional, national and European level.