Advances and Challenges in Party Positioning

Abstract

Proper estimation of parties’ policy positions is a key prerequisite before issues such as congruence and party responsiveness (Ibenskas & Polk, 2022) can be successfully addressed. The research findings frequently depend on the sources used for position data (O’Grady & Abou-Chadi 2019), and unfortunately different sources frequently diverge (Ecker, Jenny, Müller & Praprotnik 2021). Accordingly, researchers constantly seek to improve existing methods and explore new solutions in this area. Moreover, advances in machine learning, and especially in natural language processing, create novel opportunities for using big data to improve party positioning. The aim of the panel is to review different approaches to estimating party positions (e.g. party manifestos, expert and party elite surveys), analyze their strengths and limitations, and discuss how they can be improved or complemented. We propose to focus on the following issues: ▪️ Where do the existing methods of party positioning diverge and what are the underlying causes of such divergence? In particular, does it reflect a deeper conceptual challenge in our understanding of what it means for two parties to be similar? ▪️ Can the divergences be overcome -- and empirical models improved by reducing bias and measurement errors -- by integrating data from multiple methods? ▪️ What data party researchers currently need, and how existing databases (e.g. MARPOR, CHES) respond to this demand? ▪️ How can the methods of big data analysis, machine learning, and natural language processing be incorporated into the party positioning toolbox?

Title	Details
Anti-Immigration Rhetoric in Flux : The Case of Turkey	View Paper Details
Measuring Party Similarity: A Comparative Evaluation of Metrics	View Paper Details
The best of both worlds - Improving human coding with automatic classification models	View Paper Details
New party programs, policy, personnel or electorate: What does ‘a new party’ really mean?	View Paper Details