We convene scholars who use AI tools, particularly LLMs, to study political elites and their networks. AI promises greater data-gathering efficiency and thereby enables researchers to study a larger set of elites, compare networks across countries and regime types, and examine networks with different types of connections. The Workshop will address technical and conceptual challenges of studying political elite networks: identifying relevant network ties across cultural contexts, selecting optimal data sources and LLMs, ensuring replicability, and validating automated outputs. These advances will benefit both academic research on informal politics and public concerns about corruption and nepotism among powerful political actors.
Political elite and network studies have a long history. In the past, armies of research assistants had to manually collect the necessary data, which has prevented systematic testing of important theoretical questions: How do elite networks differ across regime types? What role do negative ties and enmities play? How does a broader set of elites beyond top leaders influence politics?
Recent decades have seen expanded data collection efforts (Archigos, Global Leadership Project, WhoGov, World Elite Database), providing names and basic information on political elites worldwide. However, detailed background information—particularly complete career paths and affiliations crucial for network construction—remains limited to specific countries. Research on elites in China in particular has demonstrated the value of social network analysis based on this data. The emergence of AI tools for data-gathering and analysis thus provides a unique opportunity.
The Workshop will allow the field to discuss and evaluate the different approaches taken so far (see bibliography). Some gather elite career paths to construct networks using overlap rules (e.g., working at the same institution simultaneously), offering systematic coverage but potentially inferring connections that don't exist. Others search for actors' co-occurrences in unstructured texts like newspaper articles. This allows them to extract richer relationship information including type and sentiment, but may suffer from bias toward well-known elites. We bring together scholars to discuss optimal approaches, data sources, and validation strategies. The goal is to enable large-scale comparative projects, and expand analysis beyond narrowly defined elites, and from simple to multiplex or signed networks.
Best, H., & Higley, J. (2017). The Palgrave handbook of political elites: Introduction. In The Palgrave handbook of political elites (pp. 1-6). London: Palgrave Macmillan UK.
Bro, N. (2025). A frustratingly easy way of extracting political networks from text. PloS one, 20(1), e0313149.
Bühlmann, F., Christesen, C. A., Cousin, B., Denord, F., Ellersgaard, C. H., Lagneau‐Ymonet, P., ... & Yu, X. (2025). Varieties of economic elites? Preliminary results from the World Elite Database (WED). The British Journal of Sociology, (76), 663–673.
Del Río, Adrián (2025): Elites under pressure? How parliaments can undo dictatorships. Proposal presented at the WhoGov Workshop on political elites 2025 (September 1-2, 2025), University of Oslo, Norway.
George, Julie, Keller, Barbara Franziska, Radnitz, Scott (2025). Networks and Contingency in Hybrid Regimes: Understanding Party Defections and Coalitions during Georgia’s Colored Revolution. Working paper. WhoGov Workshop on political elites 2025 (September 1-2, 2025), University of Oslo, Norway.
Goemans, H. E., Gleditsch, K. S., & Chiozza, G. (2009). Introducing Archigos: A dataset of political leaders. Journal of Peace research, 46(2), 269-283.
González-Bustamante, B. (2025). Machine Learning and Political Events: Application of a Semi-supervised Approach to Produce a Dataset on Presidential Cabinets. Social Science Computer Review, 08944393251315917.
Guo, Y., Humphreys, M., Naumann, L., & Garbe L. (2025). Map out Elite Networks: The Case of German Political-Economic Elites (2013-2023). Presentation at the American Political Science Association’s Annual Meeting (September 11, 205), Vancouver, Canada.
Higley, J., Hoffmann-Lange, U., Kadushin, C., & Moore, G. (1991). Elite integration in stable democracies: a reconsideration. European Sociological Review, 7(1), 35-53.
Keller, F. B. (2016). Moving beyond factions: using social network analysis to uncover patronage networks among Chinese elites. Journal of East Asian Studies, 16(1), 17-41.
Keller, F. B. (2017). Analyses of elite networks. In The Palgrave handbook of political elites (pp. 135-152). London: Palgrave Macmillan UK.
Lee, J., & Shih, V. C. (2023). Machine-learning analysis of leadership formation in China to parse the roles of loyalty and institutional norms. Proceedings of the National Academy of Sciences, 120(45), e2305143120.
Mahdavi, P. (2019). Scraping public co-occurrences for statistical network analysis of political elites. Political Science Research and Methods, 7(2), 385-392.
Melnikov, Kirill (2025): Performance, patronage, or control? Determinants of elite mobility in personalist autocracies: the case of Russian governors. Working paper. WhoGov Workshop on political elites 2025 (September 1-2, 2025), University of Oslo, Norway.
Nagawa, Maria and Romero, Diego (2025). Pre-Analysis Plan for Managing Elite Defection: Coalition Diversity and Punishment Severity in Authoritarian Regimes. WhoGov Workshop on political elites 2025 (September 1-2, 2025), University of Oslo, Norway.
Ngo, Eric (2025). Webs of Power: Unveiling Autocratic Elites Networks and Their Influence on Leader Constraints Using Large Language Models (LLMs). Working paper. WhoGov Workshop on political elites 2025 (September 1-2, 2025), University of Oslo, Norway.
Nyrup, J., & Bramwell, S. (2020). Who governs? A new global dataset on members of cabinets. American Political Science Review, 114(4), 1366-1374.
Nyrup, J., Knutsen, C. H., Langsæther, P. E., & Kristiansen, I. L. (2023). Paths to power: A new dataset on the social profile of governments. Available at SSRN 4631225.
Paustyan, E. (2025). Control vs. Electoral Performance: Why Personalist Rulers Choose Local Governors. Publius: The Journal of Federalism, pjaf053.
Rohr, B. (2025). Elite cohesion in the American administrative state, 1898–1998. Social Science History, 1-28.
Shih, V., Adolph, C., & Liu, M. (2012). Getting ahead in the communist party: explaining the advancement of central committee members in China. American political science review, 106(1), 166-187.
Shih, V., & Lee, J. (2020). Locking in fair weather friends: Assessing the fate of Chinese communist elite when their patrons fall from power. Party Politics, 26(5), 628-639.
Traag, V. A., Reinanda, R., & van Klinken, G. (2015). Elite co-occurrence in the media. Asian Journal of Social Science, 43(5), 588-612.
Trinh, D. (2025). The Factional Logic of Political Protection in Authoritarian Regimes. Comparative Political Studies, 58(1), 155-189.
1: What LLMs and prompting frameworks are best suited to extracting information on political elites and their networks?
2: What are the best data sources for extracting different types of political networks in specific countries?
3: What are the best data sources and approaches for extracting political networks across different countries?
4: How to best test the external validity of the networks constructed, i.e. what sources for ground truth are there?
5: How can we ensure replicability when using, in particular proprietary, AI-tools and text data?
1: Open source LLMs vs. proprietary LLMs: advantages and disadvantages
2: Using LLMs and other AI-tools to gather CV data
3: Using LLMs and other AI-tools to identify political entities in unstructured text
4: Using LLMs and other AI-tools to identify different types and valence of relationships
5: Sources for ground truth data and other ways to verify political network information
6: The problem of comparative network research: what types of networks matter in which context?
7: Signed and multiplex networks: theory and analysis of political alliances and enemies
8: Expanding beyond the inner circle: gathering and analysing data on connections to broader society
