Lecturer
Department of Political Science & Statistics
Yale University
Studying radicalism, populism, polarisation & democracy with survey experiments & LLMs | Amateur drummer & cyclist off the clock.
alberto.stefanelli@yale.edu
Institution for Social and Policy Studies, 77 Prospect Street
New Haven, CT 06511
USA
With both affective polarization and populism on the rise in several countries, many have proposed a link between the two phenomena. Yet, research offers little direct evidence on whether populist individuals are more polarized than their mainstream conunterparts. This paper aims to fill this gap by using data from 37 elections in 31 countries to provide a comparative account of the relationship between populism and affective polarization at the individual level. Contrary to common assumptions, resuts show no significant difference in affective polarization between more and less populist individuals. Instead, we identify a curvilinear relationship wherein both individauls who strongly endorse or reject populist ideas exhibit high levels of polarization, with substantial variations across countries. Furthermore, we find that populism is associated with a general disdain for all political parties rather than just rival parties. These findings challenge the prevailing assumption that populism is asymmetrically associated with affective polarization. We also provide evidence that ideological extremism, rather than populism, is more consistently and strongly associated with affective polarization in a large number of advanced democracies.
Populists often cast themselves as unwavering defenders of free speech. In practice, however, they tend to apply free speech instrumentally, denying it to those who hold opposing political views. While existing research has primarily focused on populist individuals’ general commitment to democracy, little direct evidence exists on how they concretely understand and apply freedom of speech. Using an original survey question and a survey-embedded experiment, this paper examines how support for freedom of speech among populist individuals oscillates between normative principles and instrumental considerations. The findings indicate that those with stronger populist attitudes tend to support free speech in normative terms but are simultaneously inclined to restrict it when their ideological interests are at stake. This suggests that populists view democratic norms not as universal rights but as instrumental values that can be traded for political gain
In recent years there has been an increasing interest in whether populism is related to opinion extremity. Yet, research on the topic offers little direct evidence on whether and under which conditions populist ideas at the individual level are related to policy extremity and inter-party dislike. This paper aims to fill this gap by focusing on the reasons populist individuals hold more or less extreme opinions. Using data from the 2016 American National Electoral Study, I find that populist attitudes are a strong correlate of both ideological extremity and affective polarization, yet this association is conditional on respondent’s party affiliation. Populism is related to higher levels of ideological extremity among Democrats and stronger negative leader evaluations among Republicans. This finding indicates that the relationship between populism and citizens’ political judgements varies depending on the ability of populist leaders to make certain dimensions of the competition salient (i.e. ideological or affective) and exploit pre-existing ideological and partisan rivalries (i.e. party identity).
Over the past decades, scholars have extensively studied the ideological positions of radical individuals. Yet, most research focuses on vote choice and attitudes towards particular policy issues. This paper sets out to identify the abstract and general elements (i.e., antagonism, dogmatism and populism) that constitute the ontological core of radical belief systems and assess their prevalence among the general population. Using data from the 2019 Belgian National Electoral Study, we employ a person-centred approach to classify individuals into ideological profiles. Contrary to the widespread assumption that radical individuals are a more "radicalised" version of mainstream voters, we provide evidence that they have a markedly different understanding of politics and society, and that political radicalism is not monolithic as there exist multiple, alternative, and more nuanced, non-mainstream ideologies. We also found that, compared to voters of the radical right, radical left voters are less likely to endorse a black-and-white understanding of politics despite showing high levels of populism and antagonism. Education, perceived social class, and trust in political institutions influence the degree to which individuals subscribe to different belief systems.
Objectives: To investigate country-specific drivers and barriers of positive COVID-19 vaccine intentions in the Federation of Bosnia and Herzegovina (FBiH), one of the two entities comprising Bosnia and Herzegovina.
Methods: A cross-sectional study design was used, using an online behavioural insights survey tool adapted to the context of FBiH. Three survey waves, each including approximately 1,000 adults, were conducted in July, September and December 2020. Fixed-effects regression analysis was used to explore the drivers, barriers and attitudes towards accepting a future COVID-19 vaccine.
Results: COVID-19 risk perception, trust in health institutions and negative affect were positive predictors of positive COVID-19 vaccine intentions, as were living in urban areas and having a college education (versus having primary or secondary education). Conversely, being female and feeling that the pandemic was overhyped by the media and the country of vaccine production were negative predictors.
Conclusion: This study provided snapshots on the state of attitudes regarding future COVID-19 vaccine acceptance and hesitancy in 2020. These findings provided useful insights into the efforts to introduce and roll out the COVID-19 vaccines in FBiH. Further efforts should focus on better understanding the demographic, cultural, and behavioral contexts of COVID-related vaccination perceptions in FBiH.
The article explores the reasons for the success of the Five Star Movement (FSM), one of the most electorally successful yet peculiar populist parties in the European context. I first analyze the main characteristics of the FSM in relation to the principal conceptual dimensions of European populism. In particular, I briefly examine the FSM profile and organization, underlining the key elements that make the FSM unique. Secondly, I review the electoral history of Grillo's movement. I analyze the results of the main national and local elections, looking at the transformations in its organization and strategy. Thirdly, I investigate the consolidation of the FSM at the local level using three indicators: the FSM's electoral performance, the consolidation of its organization, and the use of preferential voting by its voters. I conclude by examining the transformation of the FSM electorate in relation to the socio-demographic profile of its voters.
Despite the importance of power analysis for survey-experimental techniques, power considerations are often disregarded in the design of conjoint experiments. The main goal of this article is to provide rigorous guidance on how the number of experimental conditions, trials, and subjects impact the statistical power of conjoint experiments. To this end, we first conducted an extensive literature review to gauge the experimental designs typically employed in conjoint studies and the plausible effect sizes discovered in the literature. Using the information gathered from the literature review, we explore the statistical properties of a wide range of commonly employed designs using a simulation-based framework that employs a flexible data-generating model. Results show that—even with a relatively large sample size and the number of trials—conjoint experiments are not well suited to draw inferences for designs with large numbers of experimental conditions (> 15) and relatively small effect sizes (< 1000). Based on our simulation results, we develop a web application that can be used by researches to perform a priori power analysis and hence achieve adequate design for future conjoint experiments.
The goal of this study is to examine heterogeneity in citizens’ commitment to democratic principles by their i) left, centre, or right political orientation; ii) populist or mainstream political tendencies. We use a candidate-choice survey experiment fielded in 7 different Latin American countries. Results show that individuals trade off democracy for differ- ent policy preferences and priorities (i.e. the left for leftist policies; the right for rightist policies; the centre for centrist policies).
We test whether the entrance of populist parties in parliament changes the way citizens think about immigration and European Integration. Making use of exogenous variation created through differences in electoral thresholds, we causally assess that the representation of populist parties is associated with changes in attitudes. The results suggest the presence of attitude polarisation across the L-R ideological spectrum: left-wing individuals move towards more progressive positions while right-wing voters become more conservative.
We scraped all the parliamentary interrogations and collected all television news broadcasts concerning LGBT issues over a thirty-year time period (1994-2022) in the Dutch-speaking part of Belgium. Results show (1) a growing polarization on how LGBT issues are framed and (2) a strong agenda-setting potential of TV news on MPs’ behaviour in parliament.
In this paper, we propose that corruption is a crucial—hitherto overlooked—factor that could help better explain which structural conditions create a more fertile ground for partisan affective polarization to emerge and increase. We combine observational and experimental data from six different European countries to examine the extent to which perceptions of corruption amplify negative emotions toward political opponents. Our findings suggest that corruption intensifies citizens’ animosity and mistrust toward out-parties. Crucially, this effect is most pronounced among those with strong populist leanings, who are more likely to see the out-party as a self-serving political elite.
Citizens are generally worried about climate change. Yet, environmental concern rarely translates into electoral support. In this paper, we argue that voters’ environmental choices depend both on the type of climate policy and the monetary or non-monetary grievances generated by the green transition. To test this, we rely on an adaptive conjoint experiment that tests four types of climate policies, each generating either (1) direct material costs on individuals (e.g., imposing building renovations), (2) broader costs for people’s communities (e.g., a carbon tax on industries), (3) no material costs but high behavioral costs (e.g., the reduction of speed limits), or (4) no direct material costs but potential cultural grievances (e.g., the development of vegetarian menus in schools).
This study seeks to improve the external validity of survey experiments by exploring how they can better reflect real-world voting behavior. It develops a typology of costs that help explain why citizens may discount corrupt behavior in experimental settings: (1) cognitive costs associated with updating beliefs about a real and familiar candidate; (2) personal and ideological costs stemming from the risk of increasing the out-party’s chances of winning; (3) social desirability costs related to revealing support for corrupt politicians to survey researchers; and (4) the absence of social pressure that typically accompanies voting in real-world contexts.
Using the simulation framework developed in Stefanelli, A., & Lukac, M. (2020), this R package provides an extensive set of functions to simulate conjoint data, calculate power (1−Beta), Type S, and Type M error rates for forced-choice conjoint experiments based on the design proposed by Hainmueller, J., Hopkins, D., & Yamamoto, T. (2014).
The course explores how data science is being used to design winning campaigns. Students gain an understanding of what data is available to campaigns, how campaigns use this data to identify supporters, and the use of experiments in campaigns. This course provides students with an introduction to political campaigns, an introduction to data science tools necessary for studying politics.
The course explores fundamental forces that determine election outcomes in the United States. Students practice data analytics using real election results and historical data, and ultimately build a predictive model of the 2024 U.S. election.
Estimating causal effects is a central aim of quantitative empirical analysis in social sciences. Recently, Conjoint Analysis and Choice-Based Conjoint Experiments have gained interest among social scientists to understand and predict people's preferences in a multi-dimensional and multi-choice environment. This course offers an applied introduction to Choice-Based Conjoint, along with hands-on experience in lab sessions.
The module introduces the logic and the statistical theory of structural equation modeling. The emphasis is on the practical use of models and methods as research tools in the social and behavioral sciences. In particular, this module covers the estimation of such models for normal, non-normal, and ordinal variables using R. This includes maximum likelihood, robust maximum likelihood, and various least-squares methods.
I received the 2021 Berg-Schlosser Award for outstanding pedagogy at the ECPR Methods School for my role of Teaching Assistant for a variety of advanced statistical courses. This work included creating lab sections for a several advanced methods courses, including Structural Equation Modelling (SEM), Multilevel SEM, and Advanced Multilevel Regression Modelling.
You are welcome to use my teaching material, but please make sure to attribute it to me.
About this website: This website uses code from Minimal Mistakes and is based on a simplified version of Shiro Kuriwaki's website.