Why Cross-Level Data Matters
Most political systems operate across multiple territorial layers, but empirical datasets frequently sit at just one level—national, regional, or local—encouraging methodological nationalism, regionalism, or localism. Leonce Röth, Daniel Saldivia Gonzatti, Lea Kaftan, and André Kaiser respond to that gap by building integrated, cross-level resources that make it easier to study how elections, parties, institutions, and governments interact across levels of government.
What the Datasets Include
The authors present three linked datasets—RD|CED, RED, and RPSD—that combine electoral results, institutional features, party ideological positions, and government (cabinet) composition at both country and regional levels. Together the files track parties and their positions while linking elections to subsequent regional cabinets, allowing researchers to observe vertical political dynamics rather than single-level snapshots.
Scope and Coverage
Key coverage details are provided up front:
These counts illustrate the datasets' temporal depth and cross-national/subnational breadth.
How the Data Were Built
Röth et al. harmonize multiple data types across territorial tiers to create a coherent cross-level panel. The integrated files match party identities and ideological measures across national and regional contests and record cabinet composition at the regional level. The authors position these resources as complementary extensions to existing single-level datasets, enabling systematic comparison of party systems, policy representation, and government formation across levels.
Where to Access the Data
All datasets are publicly available; the authors provide downloads via http://multi-level-cross-level-politics.eu/ and through the Harvard Dataverse repository.
What This Enables for Research
By making cross-level linkages explicit and machine-readable, the new resources lower the barrier to studying vertical party-system congruence, subnational government formation, multi-level institutional effects, and other questions about political interaction across scales. Röth and colleagues close with a research agenda highlighting promising cross-level investigations that these data can support.
| Studying Multi-level Systems with Cross-level Data was authored by Leonce Röth, Daniel Saldivia Gonzatti, Lea Kaftan and André Kaiser. It was published by Cambridge in BJPS in 2025. |
