Electoral geographies, multi-level politics and ideology: New data sets to explore the role of ideology in multi-level politics

The study of multi-level politics is key for understanding the development of democratic systems and policy-making. Although the political science of multi-level and territorial politics alike has widely advanced in previous years (think of Hoogh, Marks, Schakel and others), we still lack the data to study the behaviour of parties across different regional levels as a function of their ideological positions. However, without such data, we are not able to study, for example, how polarization unfolds across regions over time, how it might spill over from one region to another, or how territorial conflict might spark due to polarization between different regions within countries. We propose a new measure of ideology for regional party systems based on national party positions and their electoral results at statewide and regional elections: the Ideological Centers of Gravity (COG). This measurement is based on two newly created datasets: the Regionally Disaggregated State-wide Elections (RD|SED) and the Regional Elections (RED). With this data, we cover 337 statewide elections on the regional level, 1,005 regional elections and 557 regional cabinets in 365 regions of 19 countries from 1944 to 2019 while accounting for 800 political parties and their ideological positions. In terms of quality, our data can compete with previously collected data on elections at the regional level (e.g. by Arjan Schakel), but provides extra information on the ideological composition of regional parliaments and cabinets and extends the country selection to countries such as Australia, Mexico or Turkey. Using this novel dataset, we provide first insides into how polarization unfolds across regions within countries, claiming that studies of polarization should focus more on regional trends within nation states.