Mapping (A)Ideology
A Taxonomy of European Parties Using Generative LLMs as Zero-Shot Learners
Bibliographic Data
| ID | 7971100 |
|---|---|
| Authors | Riccardo Di Leo (0000-0003-4113-7954, European University Institute), Chen Zeng (0000-0003-0554-945X, Department of Political and Social Sciences, European University Institute King’s College London), Elias Dinas (0000-0003-2153-6077, European University Institute), Reda Tamtam (0009-0005-0548-0992, Department of Political and Social Sciences, European University Institute Princeton University) |
| Year | 2025 |
| Volume | 33 |
| Issue | 4 |
| Pages | 456-463 |
| Publication date | 2025-10-01 |
| Peer Reviewed | Yes |
| Open Access | Yes |
| Type | ARTICLE |
| Venue | Political Analysis (JOURNAL) |
| Journal identifiers | ISSN: 1047-1987 • E-ISSN: 1476-4989 |
| Publisher | Cambridge University Press (CUP) (PUBLISHER) |
| DOI | 10.1017/pan.2025.7 |
| OpenAlex | W4409407912 |
| Language | EN |
| Citations received | 7 |
| References cited | 29 |
We perform the first mapping of the ideological positions of European parties using generative Artificial Intelligence (AI) as a “zero-shot” learner. We ask OpenAI’s Generative Pre-trained Transformer 3.5 (GPT-3.5) to identify the more “right-wing” option across all possible duplets of European parties at a given point in time, solely based on their names and country of origin, and combine this information via a Bradley–Terry model to create an ideological ranking. A cross-validation employing widely-used expert-, manifesto- and poll-based estimates reveals that the ideological scores produced by Large Language Models (LLMs) closely map those obtained through the expert-based evaluation, i.e. , CHES. Given the high cost of scaling parties via trained coders, and the scarcity of expert data before the 1990s, our finding that generative AI produces estimates of comparable quality to CHES supports its usage in political science on the grounds of replicability, agility, and affordability
Biology · Generative grammar · Ideology · Linguistics · Mathematics education · Political science · Politics · Shot (pellet · Taxonomy (biology · Zero (linguistics · Chemistry · Computational and Text Analysis Methods · Computer Science · Electoral Systems and Political Participation · Law · Philosophy · Psychology · Social Media and Politics · Artificial Intelligence · Ecology
Revisiting business associations in trade politics
Piercing a Methodological Bubble
Fake news as a rhetorical weapon
News under pressure? Politicization of scientific expertise in German mainstream news during Covid-19
China’s Security Signaling in the Taiwan Strait
Scaling Open-Ended Survey Responses Using LLM-Paired Comparisons
The 2024 Chapel Hill Expert Survey on political party positioning in Europe
Semantics derived automatically from language corpora contain human-like biases
ChatGPT outperforms crowd workers for text-annotation tasks
Large Language Models Outperform Expert Coders and Supervised Classifiers at Annotating Political Social Media Messages
Measuring party orientations towards European integration
Policy Adjustment by Parties in Response to Rival Parties’ Policy Shifts
How to train your stochastic parrot
Out of One, Many
Positioning Political Texts with Large Language Models by Asking and Averaging
Coder Reliability and Misclassification in the Human Coding of Party Manifestos
Measuring Parties’ Ideological Positions With Manifesto Data
Locating Political Parties in Policy Space
Scaling Policy Preferences from Coded Political Texts
Reliability and validity of the 2002 and 2006 Chapel Hill expert surveys on party positioning
A problem with empirical studies of party policy shifts
Putting Parties in Their Place
Can Large Language Models Transform Computational Social Science
Testing the power of arguments in referendums
Chapel Hill Expert Survey trend file, 1999-2019
Crowd-sourced Text Analysis
"Trump and the Shifting Meaning of "Conservative
Extracting Policy Positions from Political Texts Using Words as Data
| Unique citing works | 7 |
|---|---|
| Citations per year | 7 |
| Citation span | 2025 - 2026 (2) |
| Citation velocity | current |
| Highly cited | No |
| Citation types | Neutral: 7 |