Getting volatility right

Abstract

Abstract The electoral volatility index is a workhorse in studies of party system change, yet volatility scores for individual elections vary widely across existing data sets. This chapter shows how candidate data can improve the calculation of the index. Most traditional approaches to volatility require that electoral slates (i.e. electoral formations in one election) are linked to a single successor or predecessor and entering and exiting parties are strictly defined. However, slates with partial novelty or significant links to more than one successor or predecessor are very common. Building on the concepts and measures introduced in the previous chapters, this chapter presents a novel approach to electoral volatility that fully accounts for these issues. The proposed split-by-congruence index allows for an orderly linkage of slates to multiple predecessors or successors, distinguishes genuinely novel parties from seemingly novel ones, and accounts for party change reflected in high candidate turnover among continuing parties. The case studies of individual elections demonstrate how the new volatility index provides reliable measures based on a set of clearly defined and standardized rules. The index is tested with an analysis of the effect of economic growth that confirms the presence of an effect that has been reported for other regions but has, so far, been elusive in Central and Eastern Europe.