In the past three months, the AUTHLIB (Neo-authoritarianisms in Europe and the Liberal Democratic Response) project has delved deeper into the intricate dynamics of illiberalism. From analyzing citizens’ attitudes and parties’ ideologies to mapping the transnational networks of illiberal actors, our interdisciplinary team has been busy uncovering the multifaceted challenges to liberal democracy. With data pouring in from surveys, experiments, and advanced text analysis, we are beginning to see the contours of illiberal trends across our seven key countries (Austria, Czechia, France, Hungary, Italy, Poland, and the United Kingdom). Our initial findings not only highlight the influence of political, emotional, and historical factors but also offer critical insights into how illiberalism shapes and is shaped by social, cultural, and geopolitical dynamics.
In this quarterly update, you will find fresh insights on our recent survey results and how AUTHLIB uses cutting-edge AI tools to decode illiberal rhetoric. You will also learn about our historians’ reflections on the deep roots of illiberalism—which you can also watch online on our YouTube channel—and our progress in developing interactive tools such as an ideological map of illiberalism, and collaborative efforts to foster deliberative dialogue through mini-publics. As the project unfolds in the coming year, we will continue to bring the pieces of the puzzle together, building a clearer picture of how illiberalism manifests itself and operates in Europe, and what this means for the future of democracy.
The challenges to liberal democracy are complex but we work hard to connect the dots. Dive into the details below and stay tuned for our upcoming findings!
Survey results are in! Let’s see how illiberal voters are.
The SWPS University team has successfully concluded the data collection and preliminary analysis of its public opinion survey and conjoint experiments that examined citizens’ liberal and illiberal attitudes on representative samples of adult populations (minimum of 1,000 respondents per country) in the seven European countries studied by AUTHLIB. The surveys were conducted in the early summer of 2024 (in France and in the United Kingdom, right after their respective parliamentary elections).
Initial findings from the survey highlight interesting patterns in public opinion across the seven countries. For example, the research indicates that party affiliation plays a major role in shaping views on liberal democracy. Supporters of radical right-wing parties tend to exhibit stronger illiberal and authoritarian tendencies, while liberal and left-wing voters show more resistance to these ideologies.
Mean levels of populist, illiberal and authoritarian attitudes by age (7 countries)
Age and education also emerge as key factors in determining people’s political views. Younger age groups generally show less support for illiberal and populist ideas, whereas older ones tend to be more susceptible to these attitudes. Additionally, those with lower levels of education are more likely to hold authoritarian views, highlighting the complex relationship between education, political beliefs, and democratic values. This work emphasizes the need to understand anti-liberal ideologies in their specific socio-political contexts.
As the SWPS team—in cooperations with researchers across the consortium—continues its analysis of the collected data to refine the findings and contribute to publications in peer-reviewed journals, the team at the Transatlantic Foundation is starting to explore how these findings can contribute to broader discussions and inform policy decisions about the future of democracy in Europe. Stay tuned for a summary of our early findings to come out on the AUTHLIB Blog early next year!
Illiberal parties are illiberal in different ways…
Over the past months, the work package analyzing the configurations of liberal democratic and neo-authoritarian ideological proposals in Europe, led by Sciences Po, was heavily involved in data preparation and processing. First, the team trained a multilingual DistilBERT model (distilbert-base-multilingual-cased) in a supervised way in order to identify illiberal speech in party manifestos, political speeches, and tweets. To arrive at a better model, the creation of a training labelled dataset was then triangulated between ChatGPT and a number of human coders. This was a highly labor-intensive task that took on a number of iterations. The human coders would identify text as belonging to one or more of our seven pre-identified categories capturing various aspects of illiberalism: “power concentration”, “closed society”, “partisan state”, “anti-immigration”, “censorship”, “economic protectionism”, and “euroscepticism”. The researchers would then review all these codes with the coders, potentially adjusting the definitions and understandings, in order to arrive at a common coding scheme. Finally, the coders coded a number of training data. This data was then provided to the BERT models that are currently classifying all of our data—speeches, manifestos, and tweets—to identify text that includes our seven concepts of illiberalism.
In the end, we also used a more sophisticated multilingual XLM RoBERTa model to classify all our data using a technique called Natural Language Inference. This technique makes us able to harness the power of BERT models without creating a training dataset, while being able to understand the context efficiently. This model performed much better than the supervised one in providing consistent classifications of speeches, so we decided to use this one for the final version of our dataset. Preliminary results of this data collection have already provided us with measures of the ideological space and ideological configurations of liberal democracy and neo-authoritarianism in Europe.
The team is also in the process of finalizing data collection in the framework of the Chapel Hill Expert Survey that will be used as an additional data source on party positions and as an anchor to our text-as-data data-generation and measurement approach in the seven countries we study. The Chapel Hill Expert Survey is currently in the field for all countries of the EU as well as Switzerland and the United Kingdom.
… and they play on your emotions
The Oxford team continues its work on the core empirical components of its work package that focuses on emotional rhetoric used by parties and politicians across the seven countries studied by AUTHLIB. This exploration involves further analysis of the emotional context of texts drawn from the tweets of political parties active in AUTHLIB’s seven countries, as well as expanding a laboratory experiment—so far only piloted in Hungary and the United Kingdom—that is designed to test the effects of emotional framing on participants’ propensity to engage in illiberal actions in the context of the lab game (not in real life!). Furthermore, the team is preparing an additional survey experiment on the effects of emotional framing of key political issues across all the AUTHLIB states plus Germany (where parliamentary elections will take place with a clear illiberal challenge in early 2025), and the United States (where the recent presidential election was arguably won by an illiberal challenger). All these analyses provide input into the combined dataset-building and validating a multidimensional map of illiberalism (see below).
Let’s pull it all together!
As our various teams are diving into identifying the patterns in the AUTHLIB-generated data across the seven countries, a team of researchers led Dean Schafer of the CEU Democracy Institute is working across partner institutions to synthesize all data that has been collected in AUTHLIB in order to construct an interactive map that will be published on the project website in the form of a Shiny app early next year. Some preliminary findings that synthesized AUTHLIB survey data with expert data were summarized in a paper authored by Zsolt Enyedi, Bálint Mikola, and Dean Schafer titled “The non-economic varieties of illiberalism—A study of European parties through expert judgments and the attitudes of party constituencies” and was presented by Zsolt Enyedi at the Annual Conference of the Serbian Political Science Association on October 26-27 in Belgrade and at the Annual Conference of the Austrian Political Science Association on November 28-29 in Vienna.
Illiberals network with increasing success
In the final quarter of 2024, the work package exploring the international cooperation of illiberals and the diffusion of their frames reached important milestones. The team at the Scuola Normale Superiore (SNS) initiated a series of in-depth interviews with pivotal figures from illiberal actors, including political parties, social movements, and nongovernmental organizations. While the data-collection process is ongoing, the team anticipates that the findings will yield significant insights into the emotional, cognitive, and relational mechanisms of the transnationalization of illiberalism in Europe. Moreover, the SNS team developed a comprehensive report based on an event-chronology methodology to identify patterns of transnationalization of illiberal politics across the seven countries studied in AUTHLIB. The report, which employs an innovative approach that combines qualitative and quantitative methods, will be published imminently, but its initial findings have already been showcased at the online discussion titled “The Transnationalization of the Illiberal Right” available for viewing on AUTHLIB’s YouTube channel.
Furthermore, building on research conducted at the Central European University and at the Transatlantic Foundation, AUTHLIB organized in December a discussion on the interactions among Russia and Europe’s illiberal actors that highlighted how parties in Austria and Hungary on the one hand and the Kremlin on the other use each other for mutual legitimation.
Watch the video on YouTube and read the related papers by Franziska Wagner, Liliia Sablina and Bálint Mikola (CEU) and by Dorka Takácsy (TF).
It did not start today
While most work packages have been studying the varieties of illiberalism in the contemporary context, AUTHLIB’s team of historians also organized a successful workshop to compare the illiberal legacies and ideological morphologies of illiberalism, authoritarianism, and populism. The workshop brought together experts from the AUTHLIB countries, including several members of the consortium and a broader network of researchers, to discuss the historical antecedents of illiberalism in November in Budapest.
The presentations touching on examples from Austria, Czechia, France, and Hungary as well as comparative cases from Europe are now available to watch on AUTHLIB’s YouTube channel. The full day of discussions was followed by a book-launch event that presented various new publications addressing new tendencies in writing the intellectual histories of East Central Europe.
Mini-publics incoming
In the past two months, the mini-publics team at Charles University has concentrated on planning and preparing AUTHLIB’s Ideological Opponents’ Forum in Liberec in Czechia, which aims to facilitate deliberative discussions and develop policy recommendations. The team is finalizing the topics best suited for the town, with close cooperation with the town council, while analyzing the literature about deliberative public talks and developing its own methodology. Aleš Michal and Jonáš Suchánek from the Charles University team have started collaborating with the team in charge of building and validating AUTHLIB’s multi-dimensional map of illiberalism, pulling data from across the project’s work packages in order to feed these findings into the mini-publics too.
A special issue in the making
Three members of the AUTHLIB team (Zsolt Enyedi, Dean Schafer, and Bálint Mikola) have worked with Petra Guasti from the Charles University on reviewing manuscripts that were submitted to our special issue on “Illiberal Politics in Europe” to be published in the open-access journal Politics and Governance next year.
While you are waiting for it to come out, do not forget to check out the article by Franziska Wagner, Liliia Sablina, and Bálint Mikola on “Russia and Central European illiberal actors: mutual legitimation amid the full-scale invasion of Ukraine” that has been published open-access in the East European Politics journal, and Dorka Takácsy’s working on “Cross-Border Propaganda Networks: Examining the Relationship Between the Russian and Hungarian Regimes” mentioned above. Watch our discussions on the transnationalization of illiberal right, on the connections between illiberal parties and the Kremlin, and the many interesting presentations on the roots of illiberalism across Europe!
To make sure you don’t miss our upcoming events and publications, follow us on Bluesky, X, Facebook, LinkedIn and YouTube.
Photo: Jaroslav Bílek (Charles University) presenting at the “Compering Illiberal Legacies and Ideological Morphologies” workshop on November 14, 2024, at the CEU Democracy Institute in Budapest, Hungary.
