Publications

Publications

Preprints

Elite Polarization in European Parliamentary Speeches: a Novel Measurement Approach Using Large Language Models (ArXiv, 2025)
This project introduces a new measure of elite polarization via actor and subject detection using artificial intelligence. I identify when politicians mention one another in parliamentary speeches, note who is speaking and who is being addressed, and assess the emotional temperature behind these evaluations. This maps how elites evaluate their various out-parties, allowing us to create an index of mutual out-party hostility, that is, elite polarization.

Journal articles

Preconditions for Pacted Transitions (European Political Science Review, 2023)
Draws on a new dataset of 45 negotiation attempts to show that successful democratising pacts arise where oppositions possess strong organisational capacity, typically via nationwide trade unions or the Catholic Church.

From Partnerships to Bureaucracies: The Constitutional Evolution of Russian Universities (Voprosy Obrazovaniya / Educational Studies Moscow, 2018, with M. Sokolov & S. Lopatina)
Analyses 400 university charters to map three constitutional axes—founder independence, collegiality, and federalisation—and shows how their combinations generate distinct intra‑university regime types.

Articles under review

Modernization Theory of Democratic Stability is Over: Structural Conditions of Pre‑ and Post‑2008 Democratic Backsliding

(under review, Swiss Political Science Review)
Global panel evidence (1986–2022) reveals that classic modernization shields (wealth, regime age) predicted backsliding before 2008, but unemployment became the dominant driver afterwards; a three‑step pattern—affective polarization → party‑system volatility → executive aggrandizement—emerges across waves.

From Pact to Bloodshed: How Protest‑Movement Structures Shaped Two of Ukraine’s Revolutions (under review, Europe‑Asia Studies)
Compares Ukraine’s 2004 Orange Revolution with the 2014 Euromaidan and shows that a hierarchical, party‑led movement enabled a successful pacted transition in 2004, whereas the decentralized 2014 network structure was not beneficial to negotiations.

Dissertation

With Whom Should They Make the Pact? Preconditions for Pacted Transitions (PhD thesis, European University Institute, 2021)
Employs a mixed‑methods design to explain when negotiations between autocrats and oppositions end in democratisation. Identifies organisationally strong oppositions as the decisive enabling factor and illustrates paths to success and failure through paired case studies of the 2004 and 2014 Ukrainian revolutions.


Syllabi & Teaching materials


Using Large Language Models in Political Science (2025)

4-week intensive seminar
Syllabus

Course description Recent breakthroughs in NLP and the public release of powerful LLMs such as GPT-4 opened fresh avenues for automated text analysis in political science. Over four weeks we cover theory, hands-on applications, and ethical considerations, equipping students to design their own LLM-driven projects.

Learning outcomes By the end of the course students will be able to:

  1. Explain the architecture of modern LLMs and key NLP concepts.
  2. Distinguish major LLM techniques relevant to political text (classification, sentiment, topic modelling, entity recognition).
  3. Apply LLM tools to real-world corpora such as speeches or social-media posts.
  4. Critically evaluate reliability, validity, bias, and ethical risks of LLM output.
  5. Connect model results to substantive political‑science theory.
  6. Design a research project that leverages an LLM to answer a social‑science question.

Low-Entry Barrier LLM Research Tools (2025)

Curated resource list for self-study
Cheat-Sheet

This one-page guide collects high‑yield primers, tutorials, and “vibe-coding” walkthroughs that help students jump‑start LLM use without heavy coding prerequisites. Highlights include the Financial Times interactive Generative AI explainer, 3Blue1Brown’s neural‑network visualisation, and chain‑of‑density prompt‑engineering examples.

Regime Change: Democratization and Democratic Backsliding (2020)

4-day crash-course (15 × 90-min sessions)
Syllabus

Course description. Why do some countries democratise while others slide back into authoritarianism? This intensive seminar surveys classic and cutting-edge scholarship on regime types, transitions, hybrid regimes, and the contemporary wave of backsliding.

Learning outcomes. Participants will be able to:

  • recognise ideal-typical regimes and diagnose empirical cases;
  • apply theoretical frameworks to real-world country trajectories;
  • analyse the drivers of autocratic stability and democratic erosion;
  • critically engage with regime-change literature;
  • estimate prospects for democratisation or backsliding.