PARTYPRESS v2
Coming soon.
A large-scale, cross-national corpus of party press releases—with LLM-assisted tools to explore who responds to whom, and a public web app for researchers, journalists, and citizens.
What’s coming
- ~30 countries, 2010–2025 — Standardized, multilingual corpus of party press releases from major parties across the EU, UK, Norway, and Switzerland.
- Inter-party responsiveness — Validated LLM-assisted methods to detect who responds to whom, on what issue, and with what stance.
- Open by design — Data, code, and documentation under open licenses; replication and extension in mind.
- Web application — Explore the dataset, visualize party communication patterns, and run custom analyses without coding.
PARTYPRESS v1
PARTYPRESS v2 builds on the PARTYPRESS Database and related work:
The PARTYPRESS Database
Research and Politics, 2023
Large-scale collection of party press releases from European parties—the basis for studying agenda-setting and responsiveness.
@article{erfort_stoetzer_kluver_2023,
author = {Erfort, Cornelius and Stoetzer, Lukas F. and Kl{\"u}ver, Heike},
title = {The {PARTYPRESS} Database},
journal = {Research and Politics},
year = {2023},
volume = {10},
number = {3},
doi = {10.1177/20531680231183512},
url = {https://doi.org/10.1177/20531680231183512}
}
Party data & identifiers
We use ParlGov for party identifiers (party IDs) and party-level information such as party names. When citing PARTYPRESS, please also cite ParlGov as the source of these codes and metadata.
Team
Postdoctoral Fellow in Quantitative Methods at LSE. His research focuses on how distributional conflicts over resources and policy shape political behaviour, with work on right-wing populism, environmental policy, and LLM-based text analysis.
Postdoctoral researcher in computational social science and comparative politics. He developed the PARTYPRESS Database (v1) together with Lukas F. Stoetzer and Heike Klüver, and builds data-driven tools for election research and political communication.
With support from
