Measuring Two-Partyness: An Overview and a New Index with Minimum Data Requirements

Due to their practical importance and major theoretical implications, two-party configurations are central in research on elections and party systems. However, the measurement of two-partyness in cross-national political research is often substituted with the measurement of party system concentration/fragmentation, which leads to undesirable information losses. This study overviews the three measures of two-partyness already offered, albeit rarely used, in political science, and develops a new index that minimizes data demands, which is important in the field where the problem of data incompleteness is endemic. At the same time, the new index is transparent in its mathematical build-up, and it satisfies all theoretical requirements that have to be met in the measurement of two-partyness. The properties of the indices of two-partyness, both old and new, are tested on a hypothetical dataset comprising party constellations with widely variable structural characteristics, and on a real-life dataset that embraces elections held in 75 democratic polities of the world in 2001–2023. The analysis shows the utility of the new index in comparative research, at the same time confirming the validity of the previously developed measures.