PANEL 1.1. Changing Rules & Norms in Welfare States Friday, 23.09.2022
10.45 – 12.15
Chair: Christel Koop
1Automating Welfare - Algorithmic Infrastructures for Human Flourishing in Europe


In a world where automation is thought to increase productivity and efficiency with less effort and at lower costs, what happens to human flourishing when this logic is deployed to support decisions in the welfare sector? AUTO-WELF investigates the extensive implementation of automated decision-making in the welfare sector across Europe. It is the first to provide a comparative analysis of automated welfare provision across European welfare regimes to examine the implications of algorithms and artificial intelligence for the future of European citizens and societies. Data-based infrastructures for public administration are shaping not only welfare provision, but also state-citizen relations and prompt questions of human agency in relation to complex socio-technical systems, ethics and accountability, as well as biases and inequalities. The project foregrounds the perspective of people implicated in the automation process including software engineers, case workers and citizens. Implementing a multi-method, interdisciplinary and cross-country comparative approach, the project will develop groundbreaking knowledge on the consequences of automating welfare in two domains: a) core welfare service and b) communal welfare infrastructures. These domains will be explored across eight European countries (Austria, Denmark, Estonia, Germany, Italy, Poland, Portugal, Sweden) representing four types of the welfare state and its different stages of automated decision-making. The project provides an in-depth and cutting-edge understanding of the process of automating welfare from a European perspective producing highly relevant insights into how automated decision-making can support but also harm human flourishing.
2Re-connecting access to benefits? Administrative burdens and welfare to work

Martin Lodge (London School of Economics and Political Science) & Alice Moore (London School of Economics and Political Science) (ReConnect)

The notion of the ‘enabling state’ has often referred to unemployment regimes emphasising the importance of generating ‘welfare to work’ incentives. At the same time, the ‘enabling state’ is said to suffer from a legitimacy issue in that citizens are said to be increasingly ‘disconnected’ from the state. While the literature has largely focused on the policy content of these welfare reforms, less attention has been paid to the ways in which welfare states have sought to ‘connect’ to potential welfare recipients. This paper focuses therefore in particular on the ‘administrative’ side of the enabling state, pointing to a central source of ‘citizen disconnect’ as diagnosed in the wider public administration literature. Drawing on cross-national research, this paper explores how understandings of administrative burdens have shaped the design and operation of unemployment regimes in the UK, Netherlands, Norway and Spain. A particular focus will be paid to the UK with its ambitious welfare reform, namely the move towards ‘Universal Credit’. Drawing on novel data, the paper explores the extent to which changes in the design and practices of welfare state reform across countries have been informed by concerns regarding administrative burdens and how feedback mechanisms have been used to inform ongoing administrative calibration.
3Penalization versus participation? Some thoughts on criminalization as mode of democratic governance

Beate Binder (Humboldt University of Berlin) & Friederike Faust (Humboldt University of Berlin) (CrimScapes)

In recent decades, criminalization as a mode of governing social problems and insecurity has expanded in many European democracies. Taking this observation as a starting point, the research network “CrimScapes: Navigating citizenship through European landscapes of criminalization” explores the expanding application of criminal law, crime control measures and imaginaries of (il)legality as both responses to, and producers of, the perceived turbulences and uncertainties that are currently shaping policy and law making in the European region. An overarching goal of the CrimScapes project is to reflect on the implications criminalization has on concepts of social justice and the participatory nature of democratic societies from a cultural anthropological perspective. In our presentation, we provide first theoretical reflections for discussion. Therefore, we will introduce the concept crimscapes, which grasps the shifting, historically situated and context-specific assemblages of laws, discourses, institutions, imaginaries and practices evolving and materializing in specific fields of policy. These landscapes of criminalization are highly sensitive to power formations, distributing rights and the possibility to participate unequally. Against this starting point, we will discuss three implications of criminalization for democratic governance that result from the synopsis of the case studies pursued in the project: First, we ask about the inherent tension between democratic processes and ever-expanding legal regulations, and link our observation to the current debate in the social sciences about juridification. Second, we are interested in what it means to look at these developments from a social justice perspective. Third, we tentatively sketch four democratic dilemmas that emerge from the expanding use of criminalization to, on the one hand, govern threat and insecurity, and on the other hand, to protect and empower marginalized social groups. These preliminary theoretical reflections are based upon qualitative ethnographic research conducted within seven different fields by members of the CrimScapes project.
4Do they feel like they don’t matter? The rural-urban divide in external political efficacy

Rubén García del Horno (Autonomous University of Barcelona), Guillem Rico (Autonomous University of Barcelona) & Enrique Hernández (Autonomous University of Barcelona) (RUDE)

Recent events in multiple advanced democracies have led to renewed interest in one of the classic political divides, seemingly dormant for decades: the rural-urban divide. The acceleration of the processes linked to globalization, and more specifically to the consolidation of the knowledge economy, has accentuated a gap between the countryside and the city that has extended beyond demographic and economic aspects. The relevant question is whether this contrast is consistent with a division in the political attitudes of their respective inhabitants. The attention devoted to this issue is still scarce and the first studies have focused on examining the differences in satisfaction with democracy between urban and rural habitats.
Rural areas have often been labelled in the literature as “left behind” or “places that don’t matter”; This could probably explain the outstanding anti-system reactions that have emerged in the rural world and their lower levels of satisfaction with democracy. In this article we ask whether these processes may be related to differences in political efficacy between rural and urban dwellers. Using data from rounds 8 and 9 of the European Social Survey, we examine whether residents of rural areas feel less effective politically. In order to better understand this division, we analyse the role that the sociodemographic characteristics of the inhabitants and their levels of satisfaction with the provision of public services can play.
Furthermore, we assess whether the political efficacy gap between urban and rural areas is smaller in those countries with higher levels of electoral malapportionment leading to over-representation of rural areas in national parliaments.
PANEL 1.2. Unequal Distribution, Divides and Polarization Friday, 23.09.2022
14.30 – 15.45
Chair: Christina Eckes
1Mapping the Rural-Urban Divide in Europe: The role of socio-political clusters for issue polarization and vote choice

Antonia Lang (Goethe University Frankfurt) & Sigrid Roßteutscher (Goethe University Frankfurt) (RUDE)

In recent debates, the rise of right-wing populism and political conflict is often related to the (re-) emergence of a divide between the “left-behinds” in rural areas and the educational elite in metropolitan areas. Based upon the concept of a political cleavage we contend that a cleavage is present when certain social groups share common values and attitudes and when these are mobilized by particular parties. We argue that shared social characteristics amplify potential geographic divides. If people in certain types of places are socially similar to each other and very different from people in other types of places, we expect greater in-group homogeneity and larger out-group heterogeneity in issue preferences and value orientations. Such a bi-polarity based on geography and the related similarity in social characteristics can fuel issue polarization and, subsequently, differences in voting behaviour. Using the last wave of the European Social Survey (ESS 2018), we apply cluster analyses and multilevel regression techniques to examine i) how place of living is linked to similarity in social attributes and ii) how this place based social similarity contributes to issue polarization and vote choice.
2Political effects of educational conflict and polarization in European countries

Jochem van Noord (University of Groningen) (UNDPOLAR)

In recent political events and trends we see that conflict does not revolve per se around economic indicators (e.g. income, occupation, class), but rather education. This can be explained in different ways such as cultural capital or political interest/sophistication, but there is also evidence that education can be a basis for status and identity as previous research has demonstrated both that education can be a basis for social identification, but also that this identification is relevant for political processes. As such, such political conflict could (to an extent) be organized around education-based groups. This paper addresses the question to what extent this conflict expresses itself in education bias, or affective polarization between educational groups, and further, to what extent these affect important political outcomes such as political trust, satisfaction with democracy, and populism. We investigate this across nine different European countries (N = 11217). Results show education is strongly related to affective polarization (with education groups significantly preferring their own group). Further, we see that education bias towards the higher educated is negatively related to political trust, satisfaction with democracy, and populism. Populism is investigated in four ways: populist voting, populist attitudes (Akkerman et al., 2013, scale), affective polarization towards populist parties over mainstream parties, and identification with ‘the people’. These relationships hold while controlling for income and identification people with a similar income level. This shows that, across Europe, education is a relevant factor for politics, not just in the form of resources, skills, and network position, but also as a basis for identity and affective polarization.
work, visibility, and inclusivity of LGBTQIA+ employees of multinational companies under populist governments in Eastern Europe

Magdo Chuchracka (Goethe University Frankfurt) (POPBACK)

The dynamics of discrimination and inclusion are correlated with the so-called anti- politics. The anti-EU, anti-genderism, and anti-LGBTQIA+ sentiments of the populist governments (Chojnicka 2015) breed foundation for a prejudiced treatment of LGBTQIA+ individuals in populist societies. Gender and sexual minorities appear to be an element of a political game between companies and governments.
Companies strive for the consumer’s approval by engaging with LGBTQIA+ groups (Bielska and Tamborska 2018), while governments see the acceptance of cultural diversity and the notion of traditional nationhood as mutually exclusive (Chojnicka 2015). This research intends to uncover the dualism of political and economic approach to LGBTQIA+ persons in Eastern European countries. The objective is to present how populist tactics of othering combined with capitalist rainbow washing are contributing to the struggle of gender and sexual minorities. This inquiry presents an innovative angle of approaching rainbow washing in Eastern European countries under populist governments by employing in-person interviews with managers and employees of multinationals, as well as LGBTQIA+ individuals living in regions of heightened homophobic tensions, like Poland and Hungary. A part of the study will entail an analysis of official social media accounts of multinational companies. Especially during pride month, when queer narratives are put into work for capitalist organizations to accelerate popularity and financial gains. New media, and especially social media, is seen here as one of the factors leading to potential aggravation of discrimination. The forms of digital communication can have considerable effects on the potential consumer and the voter. The work attempted here has insightful implications for the future research on the dynamics between LGBTQIA+ communities and exclusionary governments. Thus, by employing a de- Westernized approach to queerness and an interdisciplinary perspective, it may contribute to solving the problem of corporate and political exploitation of queer individuals and their narratives.
4Drug use, pandemic, and state-citizen relations. A case study from Poland

Justyna Struzik (Jagiellonian University) (CrimScapes)

When examining the criminalisation of identities and social practices, we often look at groups and communities who, finding themselves in a criminalised context, are exposed to chronic insecurity, restrictions on their civil rights and systemic violence. The negative consequences of criminalisation can be further reinforced by rising authoritarianisms, economic and political crises and unpredictable events that particularly affect the most vulnerable groups and populations. Taking the SARS- CoV2 pandemic and political responses to it as a starting point, in my presentation I want to look at the experiences of people who use drugs in Poland, in the context of restrictive drug laws, public health measures introduced in response to the pandemic and changing state-citizen relations. Drawing on qualitative research conducted as part of the project Crimscapes: Navigating citizenship through European landscapes of criminalisation, I want to address how the pandemic has affected drug users' experiences of their interactions with state institutions (e.g. access to healthcare, treatment facilities, substitution). I want to look at these experiences through the lens of the category of citizenship, asking questions about state-citizen relations in democracies that use criminalization to govern society. The voices of people who use drugs and their strategies for navigating relationships with state institutions figure prominently in the analysis.
PANEL 1.3 Challenges to Democratic Governance and European Economies Saturday, 24.09.2022
10.30 – 11.45
Chair: Birgit Sauer
1Translating democracy into finance? Pension funds and the role of member preferences for investment decisions

Philipp Golka (Leiden University) & Natascha van der Zwan (Leiden University) (DEEPEN)

Scholarship on financialization has investigated the multifarious ways how financial market actors’ growing power threatens or even undermines democracy. However, less is known about the conflicted space between finance and democracy where powerful financial actors operate in highly regulated contexts that mandate democratic participation. Pension funds in the Netherlands are a case in point here as pension assets far exceeding the national GDP are embedded in participatory governance models that legally guarantee equal voice to employees, employers and pensioners. Based on qualitative case studies, we compare how pension funds using different forms of participatory governance translate member voice into financial investments. In particular, we investigate pension funds’ power as they are sandwiched in between members and powerful asset managers and financial consultants. We argue that the extent to which pension funds are able to translate participant preferences into extant financial chains is key to the balance between finance and democracy.
2Can Government Policies Moderate Political Backlash to Technological Change?

Reto Bürgisser (University of Zurich), Silja Häusermann (University of Zurich), Thomas Kurer (University of Konstanz/University of Zurich) & Tobias Zumbühl (University of Zurich) (TECHNO)

A rapidly growing literature suggests that economic uncertainty created by structural transformation and technological change contribute to political dissatisfaction and the recent surge in populist voting. This project addresses a natural — but so far largely unresolved — follow-up question: can governments moderate the political backlash to economic modernization through appropriate policy interventions? While existing work suggests that spending cuts and austerity are electorally harmful for governments, we know surprisingly little about the presence of the reverse mechanism. We theorize the conditions under which expansive government policies may increase political support among those affected by technological change and present a research design to test our hypotheses empirically. Rather than examining a set of cross-nationally comparable but often somewhat abstract policies, we zoom in on a carefully chosen and financially significant intervention: the French professional security contract (CSP), which was introduced in 2011 with the explicit aim to support workers hit by structural economic change. In a first step, we examine whether the number of local CSP recipients affects regional-level turnout rates and election outcomes. In a second step, we complement this aggregate-level analysis with an original survey to learn more about underlying mechanisms and to test various potential explanation of why exactly a sizeable and targeted intervention like CSP does or does not moderate political responses to structural economic change.
3A friend in need: How crisis affects the CJEU’s inter-institutional relations

Allison Östlund (University of Gothenburg) (SepaRope)

Most legal orders make provision for times of crisis and emergency, particularly by suspending normal safeguards concerning the checks and balances between governmental actors. In particular, the executive powers of the government tend to expand in times of crisis. At the same time, crisis increases political controversy.
For the judiciary, this entails, on the one hand, a heightened need for scrutiny to counter the risk of judicial overreach, and on the other, increased deference to the elected branches due to both political salience and need for decisive action. For a European Union in a state of successive or almost perpetual crisis, it becomes imperative to uphold an institutional balance ensuring that executive and legislative actors are subject to judicial control. The purpose of this paper is to examine whether and how crisis affects the CJEU’s relations to other institutional actors.
Using migration law as a case study, the paper relies on a combination of empirical and doctrinal methods. First, the paper relies on doctrinal methods of legal analysis to conduct an in-depth analysis of key rulings concerning the exercise of emergency powers by the EU institutions. Secondly, using a unique dataset of 283 legal issues decided by the CJEU in the field of migration between 2010 and 2020, the paper compares the Court’s rulings and reasoning before and after the unfolding of the migration crisis of 2015. Finally, the paper discusses the implications of the findings for the separation of powers in the EU.
4More detecting and effecting? The changing ability of consumers to interact and engage with the regulatory state

Christel Koop (King's College London) & Alena Pivavarava (King's College London) (ReConnect)

In the 1980s and especially the 1990s, governments and parliaments in established democracies delegated far-reaching competences to independent regulatory agencies. Insulated from electoral pressure and party politics, these agency would possess limited input-based democratic legitimacy, but would compensate this by producing better outcomes for the economy and society. Yet, over the past decade, the performance of independent regulators has increasingly been questioned – especially in terms of what it has delivered for (ordinary and vulnerable) consumers, and how well it has protected consumers against corporate power. These questions have been translated into (public and political) pressure on the agencies to become more consumer-oriented. Yet, although anecdotal evidence suggests that these pressure have been followed by consumer-oriented reforms of regulatory systems, there are no systemic studies of these changes. Focusing on regulatory agencies in four European countries with different political-economic and welfare systems (the Netherlands, Norway, Spain, and the United Kingdom), we seek to fill this gap in the literature by analysing consumer-oriented changes in the statutes and practices of regulators (from 2000 to 2020). More specifically, building on Hood’s (1986) NATO framework, we look at (changes in) mechanisms that allow citizens to learn and gather information about regulatory processes (“detectors”), and mechanisms that enable citizens to actively participate in, and influence, these processes (“effectors”). Our study also explores what has driven changes in these mechanisms. By answering these questions, we seek to shed light on the changing ability of consumers to shape, and engage with, the regulatory state.
5Who is open for ideas? Comparing and explaining variation in public consultation practices of regulatory agencies in 30 OECD countries

Moritz Müller (Leiden University), Caelesta Braun (Leiden University) & Bert Fraussen (Leiden University) (ReConnect)

Regulatory agencies rely on consultations with external stakeholders to provide them with unique and often technical information that would be difficult to attain otherwise. While targeted (closed) consultations with specific stakeholders might help agencies to act more efficiently, open consultations, typically conducted online, allow agencies to engage with citizens and increase their perceived legitimacy. In this study, we assess what types of regulatory agencies make use of open consultation and examine how and why open consultation practices differ across national contexts. We map the use of open consultations by 310 regulatory agencies in 5 regulatory domains across 30 OECD countries. Drawing on insights from previous work on stakeholder engagement in regulatory governance, we test various explanations of the differing use of open consultations. In this way, we clarify how variation across national context and policy domains shapes the extent to which, and how, regulatory agencies apply open consultation tools. Findings help us to understand better potential motives behind employing open consultations in regulatory governance.
PANEL 2.1 Understanding Political Polarisation across Europe Friday, 23.09.2022
14.30 – 15.45
Chair: Agata Dziuban
1Visual Persuasion in a Transforming Europe: the affective and polarizing power of visual content in online political discourse

Alexandra Segerberg (PolarVis; CHANSE VISITOR)

This project examines the mechanisms through which visual content becomes a beacon of political aggregation and polarization in political movements, and assesses the outcomes of this process. Visual content is powerful in political communication, and it is increasingly salient in the digital age. Various types of visual content, such as images, videos, and memes, are persuasive, emotive, and affective, and they are widely shared in online environments. Visual content plays an important symbolic, emotional and (dis)connecting role in political movements, where it becomes the focus of social and algorithmic negotiation. Even as movement and countermovement actors use visual content to mobilize and frame issues, citizens play a part by modifying and sharing it, and platforms mediate these processes. The project focuses on a controversial issue of cross-generational concern, climate change, and specifically the youth climate movement. The project deploys a methodological approach that combines qualitative, computational, textual, and network analysis of visual content to address the issue in four research areas. The project will investigate 1) how visual content enters movements’ repertoire of communication; 2) the characteristics of visual content, its framing and how it can support the emergence of visual narratives; 3) how online audiences react to visual content and how counter-publics and delegitimization processes emerge; and 4) how visual content propagates online. Building on a unique combination of methods the project will forge a unique interdisciplinary framework from visual framing and content analysis, interview methodology, and state of the art in computational methods that will not only produce scientific advancement but, through an extensive activity of knowledge exchange and public outreach, it will directly benefit the many stakeholders in the area of climate change.
2Leveraging politicization? National governments’ success in defining Council position for legislative negotiations.

Anastasia Ershova (Leiden University), Aleksandra Khokhlova (Leiden University), Nikoleta Yordanova (Leiden University), Christine Sylvester (University of Strathclyde) & Zachary Greene (University of Strathclyde) (EUINACTION)

Despite a rich literature on the dynamics and bargaining strategies used during legislative trilogue negotiations in the European Union (EU), little is known about the extent to which the initial positions of EU legislators are defined by domestic politics of the member states. Here, we investigate the conditions which enable national governments represented in the Council of Ministers to shape the overall Council position at the start of trilogue negotiations. We posit that a successful translation of the national preferences over policy outcome is mediated by the degree to which the respective policy area is politicized on the domestic level. Growing salience and contestation characterizing EU politics in the member states enable political actors to defend their stance over policy on the supranational level, and steer the overall Council’s position closer to their preferred outcome.
Furthermore, we anticipate that such factors as national elections and party polarization serve as additional leverage used by the Council members when framing the negotiating mandate prior to the trilogues. Our results add to the growing literature unpacking the relationship between national political dynamics and behavior of the institutions in the multilayered system of the EU. To test our expectations, we rely on a semi-supervised machine learning scaling of a new political debates dataset to map the positions of the national governments across the Union and to examine the content of the negotiation mandates. We draw on the Eurobarometer indicators to capture the domestic politicization of the EU issues across the societies and policy areas.
3Beyond the culture wars? Assessing the socio-economic determinants of affective polarization

Kamil Bernaerts (Vrije Universiteit Brussel), Joke Matthieu (Vrije Universiteit Brussel), Nino Junius (Vrije Universiteit Brussel) & Didier Caluwaerts (Vrije Universiteit Brussel) (UNDPOLAR)

This paper contributes to the emerging field on affective polarisation by exploring its macroeconomic and socioeconomic antecedents, which have hitherto largely escaped academic scrutiny (except for Boxell et al., 2020; Gidron et al., 2018, 2020). This study looks at both individual-level and macro-level determinants to explain affective polarisation at the individual level. Building on the CSES dataset for a large-N, global comparative analysis of the economic antecedents of affective polarisation, we argue that the economy does play a role in predicting affective polarisation, but almost exclusively so at the individual level. On the one hand, individual socioeconomic status is negatively correlated with affective polarisation, meaning that lower-educated individuals and individuals with lower income are more likely to be polarised. On the other hand, contrary to our second hypothesis, objective macroeconomic indicators, with the sole exception of inflation, have little explanatory power in predicting individual levels of affective polarisation. Hence, affective polarisation seems to be driven by how individuals feel about economic performance, rather than by how the economy is actually doing, as this relationship holds when controlling for objective variables about the state of the economy.
4Public Opinion Polarization Across Rural-Urban Europe

Sascha Göbel (Goethe University Frankfurt) & Richard Traunmüller (University of Mannheim) (RUDE)

Media portrayals of rifts in issue attitudes between rural and urban areas abound. Yet empirical evidence on the geographic polarization of mass opinion in Europe remains scarce. We address this gap by analyzing millions of responses to issue questions and self-reports of rural-urban location from established public opinion surveys fielded in 36 countries over four decades. Acknowledging the multifaceted nature of the concept of public opinion polarization and its many measurements, we take a comprehensive perspective that considers (1) the average divergence of issue preferences across rural-urban subgroups, (2) the heterogeneity or radicalization of attitudes within subgroups, and (3) the consistency or alignment of subgroup- preferences across issue domains. To measure corresponding quantities of interest from our data, i.e., the mean and variance of subgroup preferences as well as the inter-issue domain correlation of individual preference estimates, we build on recent advances in latent variable modeling to develop a Bayesian cross-country heteroscedastic graded response model. Against the background of media narratives and the existing literature on rural-urban divides, we employ this model to assess the evolution of rural-urban economic, immigration, moral, and family value conservatism in Europe since the 1980s. We demonstrate the robustness of our estimates to the exclusion of specific item-subsets and low-reliability items and assess potential repercussion of using self-reported as opposed to objective measures of rural-urban location. Results from our analyses contribute unprecedented insights into the temporal and regional dynamics of public opinion polarization across rural-urban Europe.
PANEL 2.2 Populism and the Far-Right Friday, 23.09.2022
10.45 – 12.15
Chair: Karolina Koc-Michalska
1Who’s afraid of ‘the people’? An empirical study into perceived threat by the people and anti-people elitism in four countries

Bram Spruyt (Vrije Universiteit Brussel), Didier Caluwaerts (Vrije Universiteit Brussel), Céline Darnon (University of Clermont Auvergne), Matthew Easterbrook (University of Sussex), L. Kavadias, Rebekka Kesberg (University of Sussex), Toon Kuppens (University of Groningen), Antony Manstead (Cardiff University), Lien Smets (Vrije Universiteit Brussel), & Jochem Van Noord (University of Groningen) (UNDPOLAR)

Populist politics have gained momentum in many Western countries during recent decades. Apart from the continued electoral success of populist political parties, events like Brexit and the storming of the Capitol building show that populist forces can have significant and tangible effects. In response to such events, research has examined who supports populist politics, and under what circumstances. Less is known, however, about the impact of these developments on voters who do not support these populist politics. This paper follows scholars who argue that the success of populist and radical politics has increased the salience of the category ‘the people’ in contemporary politics, and assesses whether as a response to this increased salience some voters feel threatened by ‘the people’ and plea for a view on politics that reduced their direct impact on politics (i.e., anti-people elitism).
Based on pre-registered analyses of newly gathered survey data from four countries (i.e., the UK, The Netherlands, Denmark and The Netherlands) we contribute to the literature in two ways. First, we engage with an emerging literature that has started to study more elitist attitudes towards politics. More specifically, we zoom in on a form of elitism that explicitly denies a central role of ‘the people’ in politics and as such rejects a central element of populist rhetoric (i.e., anti-people elitism).
Secondly, our paper contributes to a growing number of studies that argues for integrating an intergroup perspective and the associated identity measures into the study of political opinion in political science. More specifically, we study whether educational and income differences in perceived threat of the people and anti-people elitism are moderated by education- and income-based social identities.
2The antecedents of MNC political risk under right-wing populist governments: A research agenda based on evidence from East-Central Europe

Daniel Kinderman (University of Delaware), Andreas Nölke (Goethe University Frankfurt), Dorottya Sallai (LSE) & Gerhard Schnyder (Loughborough University London) (POPBACK)

Right-wing populist parties who obtain governmental power rely on ethno- nationalist mobilization for domestic legitimacy and may therefore adopt policies that explicitly seek to disadvantage foreign multinational enterprises (MNEs).
Consequently, they constitute particularly revealing cases for scholars of international business policy and political risk. Yet, scholarship on the various ways in which populist business policies affect MNEs remains scarce. Understanding what makes a foreign MNE particularly exposed to adverse action by right-wing populists is a theoretically and practically important yet understudied question for the field of international business policy. In this paper we investigate the hitherto largely understudied context of right-wing populism in post-socialist members states of the European Union, which constitute extreme cases of right-wing populist government power. We adopt a case study-based theory-building approach. We draw on a unique set of interviews and secondary literature from a range of disciplines to developed propositions that form a research agenda on this important question.
3Theorising and mapping media ownership networks in authoritarian-populist contexts: a comparative analysis of Austria, Slovenia, Hungary and Turkey

Burçe Celik (Loughborough University London), Melek Küçükuzun (Loughborough University London), Gerhard Schnyder (Loughborough University London), Fanni Toth (Loughborough University London), Mojca Pajnik (Peace Institute, Ljubljana), Marko Ribač (Peace Institute, Ljubljana), Iztok Šori (Peace Institute, Ljubljana), Tjaša Turnšek (Peace Institute, Ljubljana), Lana Zdravkovič (Peace Institute, Ljubljana), Marlene Radl (University of Vienna) & Birgit Sauer (University of Vienna) (POPBACK)

This paper explores the changes in the structure of media ownership in selected Central and Eastern European countries in the context of the rise of authoritarian populism and anti-democratic backlash. Recent media research suggests that populism can be understood in part as a response to the failure of tackling concentrated media ownership (Freedman, 2021). Expanding this work, our paper uses social network analysis (SNA) to investigate changes to media ownership looking at both ownership structures as well as concentration dynamics. The paper covers developments in Austria, Slovenia, Hungary and Turkey over the past two decades, during which authoritarian tendencies were evident in all four countries, albeit to varying degrees. While contextualising the authoritarian-populist transformations in the countries under study, we explore how national media spaces have changed accordingly in structural and economic terms. Contrary to studies that focus on the discursive aspects of populist communications treating populism as a performance, style or rhetoric that articulates the people against the elite (Mudde, 2004; Laclau, 2005), this paper emphasizes the substantive, political-economic and institutional aspects of populist politics (see Tugal, 2021). Hence, the authors strive to shift the discussion on populist communications from ‘what populists say through media’ to ‘what populists and populism do to media’. Understanding the transformation of media ownership structures as intrinsic to the political economic re-organization of authoritarian populism, the paper provides evidence that the rise of populism is accompanied by extensive changes to media ownership networks, suggesting that such changes may be instrumental in consolidating and maintaining populist regimes. Going beyond conventional market concentration indices, SNA has recently been used as a useful tool to avoid a medium-specific approach to measuring media ownership concentration (Birkinbine & Gomez, 2020). Combined with a comprehensive dataset indicating the owners of high-reach news media outlets in the print, TV, and online sector for the years 2000, 2010 and 2020, the method allows us to visualize conspicuous changes in the network structures and determine the centrality and density of the ownership networks over time.
Preliminary results indicate that far-reaching changes have taken place over the past two decades, pointing to the emergence of dominant media actors and increasing polarisation between large, dominant media groups and more marginal players in all four countries. The paper offers a novel way of addressing the question of populist control over media systems: It adds valuable cross-national ownership data to the discussion of media-populism-relations, expands the possibilities of SNA to study media ownership concentration and derives propositions from these findings, which shed light on the political-economic restructuration of media spaces within authoritarian-populist contexts.
4Grievance and Hate Speech in U.S. Far-right Online Ecosystem in the Wake of the 2021 Capitol Insurrection

Aaron Rudkin (Trinity College Dublin) (ExId)

Feelings of collective victimhood have been demonstrated to have a strong effect on intergroup bias, outgroup hostility and support for violence. The use of narratives stirring these feelings in far-right communications is especially concerning given their inclusion in the manifestos of several mass killers across Europe and North America. However, scholars still have little knowledge on the reach of such narratives as well as the extent to which major salient events increase attention to collective victimhood messaging among far-right followers. To address these gaps, we analyze the use of collective victimhood narratives on the popular secure instant messaging service, Telegram, which has exploded in popularity in response to mainstream platforms attempts to moderate speech. we develop a supervised machine learning algorithm to predict the presence of these discourses in text from over 18.5 million messages that were extracted from 1,870 far-right Telegram channels. We then use these data to test what impact the George Floyd protests and the storming of the U.S. Capitol had on the frequency of collective narrative discussions on far-right Telegram. Our findings suggest that both events coincided with a significant increase in the use of victimhood narratives, thus providing insight into the radicalization process of far-right communities online.
PANEL 2.3 A Crisis of Trust and Solidarity Saturday, 24.09.2022
09.00 – 10.15
Chair: Hayley James
1Criminal Figures in the landscape of solidarity towards migrant

Jérémy Geeraert (University of Paris-Saclay) (CrimScapes)

One recent development in the ever-expanding crackdown on migration and the implementation of a hostile environment for migrants with precarious legal status in the EU has been the criminalization of migrant solidarity (Fekete, 2018). By means of various legal tools, EU governments have been trying to hinder solidarity actions from the civil society toward migrants. Additionally, a narrative depicting civilians and groups helping migrants as criminals has been developed by governmental and supranational organizations and strengthened by far-right groups. In reaction, a counter-narrative has been strengthened and spread by pro-migrant groups and media, which presents concerned criminalized activists as modern criminal heroes and their action as of civilian disobedience (Cusamano&Villa, 2020). Based on a critical discourse analysis of academic texts, press articles and reports from governmental and non governmental organizations, this article examines two criminal figures coming from the field of civilian search and rescue: the ‘smuggler’ and its counterpart the ‘criminal hero’. By analyzing the genealogy of these two antagonistic figures and their narratives, the paper contributes to a better understanding of the moral order underpinning the criminalized landscape of solidarity towards migrants. On the one hand, the figure of the ‘smuggler’ is part of a narrative depicting migration as a threat to political, social and economic stability in the EU. Our argument is that, in this sense, the narrative is part of a moral order, which aims to promote a governance based on fear and threat, and which has been developed in recent years. On the other hand, the figure of the modern criminal hero tends to develop a counter-narrative to this moral order based on the idea of solidarity and the defense of basic human rights.
2The construction of Islam as threat in France from the 80's to the present day

Dorra Mameri-Chaambi (School for Advanced Studies in the Social Sciences - EHESS) (QUEST)
4The Rural Urban Divide in France and the democratic crisis in France

Kevin Brookes (Sciences Po Grenoble - Université Grenoble Alpes) & Sonja Zmerli (Sciences Po Grenoble - Université Grenoble Alpes) (RUDE)

The democratic crisis refers to a set of phenomena that weaken the democratic system: the growth in popularity of radical anti-system parties, the decline in political participation, the decline in attachment to democracy. This crisis is evident in most European countries and is well documented, but the literature takes little account of the spatial dimension to explain it. The aim of this paper is to present results from a French pilot survey (N=523) that measures the magnitude of the differences in values and lifestyles between urban, peri-urban and rural people. To do this, we measure citizens' identification with a territory, their attachment to it, and their resentment towards people living elsewhere. We also take into account their satisfaction with local services and their means of transportation. Do people experience a spatial cleavage? How does it express itself? Does it have consequences on their relationship with democracy and should it be taken into account in order to explain the democratic crisis in France? Our paper suggests that people living in rural places feel a strong resentment towards other outgroups related to political power, ressources and respect of their way of living. These spatial resentments seem to have impact political behaviour.
5No relief. The links between criminalization and medicalization of abortion.

Agata Chełstowska (Jagiellonian University) (CrimScapes)

The criminalization of abortion in Poland is inextricably weaved with its medicalization, a process of establishing abortion as a social event belonging solely to the medical field, with doctors and hospitals as the only actors capable of ensuring a safe abortion. This genealogy of abortion medicalization is traced by historians to the post-war state-socialist times in Poland (Ignaciuk, Markowska- Kuźma). The newest developments in abortion activism in Poland, however, have targeted this aspect of criminalization, and brought about deliberate efforts of demedicalization of abortion through the use of abortion pills and practice of at- home abortion. Furthermore, the 2020 tightening of the abortion ban, excluding fetal malformation from the list of conditions qualifying for a legal abortion, resulted in at least three known cases of pregnant women dying in hospitals due to treatable pregnancy complications. Their death are being contributed to the hesitation of doctors to evacuate a fetus, presumably in connection with the new version of the abortion legislation. That has, in turn, eroded the public trust for doctors and hospitals, and resulted in questions about what is being banned, protected and prioritized in the medical staff’s interpretation of the abortion ban. In this paper I aim to shed some light on the practices and discourses of doctors; the erosion of public trust for hospitals and medical profession; and the development and growth of complex international activist networks, facilitating at-home, demedicalized abortion. I link data on actual criminalization of medical staff with theory on law, implementation and cultural norms, and provide context for the power relations inherent in the doctor-patient relationship. The paper is based on an ongoing, multi-sited ethnographic study conducted as part of the CrimScapes project.
PANEL 3.1 Digital Media and AI as Political Tools Friday, 23.09.2022
10.45 – 12.15
Chair: Peter Van Aelst
1Smart Borders are Watching You! Automated Decision-Making and AI at the EU Border

Paulina Jo Pesch (Karlsruhe Institute of Technology) & Franziska Boehm (Karlsruhe Institute of Technology) (INDIGO)

The text-as-data PolSci community increasingly relies on supervised text classification models -- that is, machine learning models trained to predict discrete categories assigned to texts -- for various tasks, e.g., predicting policy areas of legislative text or predicting topics/themes in manifestos of political parties or speeches delivered in parliaments. Training such machine learning models requires non-negligible amounts of labeled data (i.e., texts manually coded with category labels by domain experts), which is, in general, both expensive and time-consuming. Multilingual use cases, in which one needs to predict the categories for texts written in two or more languages, pose an additional challenge that has, to a large extent, been unaddressed in this community. For many tasks, the manually coded texts exist only in English or a handful of (major) languages: without the help of multilingual representation learning and cross-lingual transfer, such models cannot be transferred to languages unseen during training, that is, they cannot make any predictions on texts from languages for which there are few or no coded texts. In this talk, we will present state-of-the-art methods for multilingual text representation
-- based on pretrained multilingual language models – that enable zero-shot (i.e., when there are no coded texts in the target language(s) at all) and few-shot (when there is only a handful of labeled text instances in target language(s)) cross-lingual transfer of text classification models. We will then present the results achieved by these approaches in use cases from the EUINACTION project: (i) policy area classification and (ii) identification of European relevance for political texts -- parliamentary speeches and legislative proposals.
2On or off topic? Understanding the effects of political targeted ads based on issues

Xiaotong Chu (University of Amsterdam), Lukas Otto (University of Amsterdam), Rens Vliegenthart (University of Amsterdam), Sophie Lecheler (University of Vienna), Claes de Vreese (University of Amsterdam) & Sanne Kruikemeier (Wageningen University & Research) (DATADRIVEN)

Whilst targeting strategies are allegedly prevalent in political campaigns, evidence regarding their actual effectiveness is scarce. One of the most common uses of targeting strategies is with policy appeals, where online users are targeted with an issue aligned with their own interests. From a user perspective, this study investigates: (a) the effect of issue congruency on vote choice, (b) to what extent ad perception and party evaluation mediate this process, and (c) the moderating role of party identification. The primary impediment to understanding the real-life effectiveness of online advertising is researchers’ limited access to advertising data, as its usage often happens covertly in a tangled manner. This study adopted a combined methodology of: the mobile experience sampling method (mESM) with an event-contingent sampling design, a four-wave panel survey, and content analysis. The combined approach allowed us to capture voters’ online advertising exposure, momentary responses, and final vote decision, and thus effectively study the effects of issue congruency within the mobile and cross-platform environment. To test the effect of issue congruency on the outcome variables, two multilevel analyses using the R package Lavaan (version 0.6-11). The results showed that citizens find an online political ad with a topic related to their own interest more interesting, informative, and persuasive regardless of their partisanship. This ad perception fully mediates the relationship between issue congruency and vote choice. Besides, a congruent ad can lead to a more positive party evaluation only when the ad promotes a more favorable party but not when the ad promotes a less favorable party. Taken together, this study provides insights on efficiently allocating data- driven strategies in political campaigns.
3Preferences over policies related to workplace digitalization

Nicolas Bicchi (Barcelona Institute for International Studies), Aina Gallego (Barcelona Institute for International Studies) & Alexander Kuo (Oxford University) (TECHNO)

What policies do citizens prefer in response to the ongoing processes of digitalization and automation in the workplace? To what extent does “occupational risk” condition such preferences? In this paper, we distinguish among compensation policies, retraining policies, and protectionist policies intended to slow down or prevent technological change. Using alternative measurements of automation risk and new survey evidence from Spain, we find little evidence that automation risks matter for compensation or retraining but that they condition support for technological protectionism. These results have implications for the measurement of both what features of a worker matter most for predicting policy preferences, as well as what policies that parties or governments might propose to gain political traction in response to technological change and the overall decreasing cost of tools to substitute workers.
4Data-driven campaigning: a literature review and a model to understand the effects

Sanne Kruikemeier (Wageningen University & Research), Xiaotong Chu (University of Amsterdam), Marlis Stubenvoll (University of Vienna), Selina Noetzel (University of Vienna), Lukas P. Otto (University of Amsterdam), Jörg Matthes (University of Vienna) & Sophie Lecheler (University of Vienna) (DATADRIVEN)

As technology develops, data-driven targeting techniques are widely employed in political campaigns, and it is hotly debated among policymakers and scholars regarding the virtuous and vicious consequences. Following a top-down approach, this study aims to provide a literature overview of what has been studied in existing research. We first introduce the Data-Driven Campaigning Effects Model (DDCEM) to gain a thorough understanding of the individuallevel effects of data-driven campaigning. A literature review was then conducted. The results indicate that unintended effects have been understudied and we also suggest future studies to zoom in on immediate responses toward data-driven campaigning. A research agenda is also developed, where we call for further methodological development, such as comparative and dynamic research designs.
PANEL 3.2 Citizens vs. Online Hate, Misinformation, and Inequality Saturday, 24.09.2022
10.30 – 11.45
Chair: Sascha Göbel
1Researching Europe, Digitalisation, and Conspiracy Theories

Michael Butter (REDACT; CHANSE Visitor)

Digital communication technologies enable the exchange, adaptation, and adoption of conspiracy theories at an unprecedented speed and scale, facilitating the creation of counter-publics joined by a propensity for mal-information. This is due not only to the ease of online micro publishing, but also to the infrastructural design of platforms that algorithmically identify potential audiences of niche content.
Moreover, such platforms are based on data extractive business models that require a certain amount of content agnosticism and benefit from adversarial modes of communication. Most research on conspiracy theories has focused on the US and, to a lesser extent, Western Europe. The REDACT project will analyse how digitalisation shapes the form, content, and consequences of conspiracy theories, including online sociality and offline actions and effects. Rather than see digitalisation as a process that has universal outcomes, REDACT considers online conspiracy theories and counter-publics in different European regions (Western Europe, Central Europe, the Baltics, and the Balkans) in order to make robust and nuanced recommendations about conspiracy theories—a particularly durable form of mal-information—for policy makers, media regulators, fact-checking and extremism-monitoring organisations, as well as the internet companies themselves. This will make a significant contribution to scholarship on conspiracy theories, first by displacing the focus on the US as the default, and second by situating digital communication technologies, platform affordances, and online sociality, at the centre of the enquiry. With a richer understanding of how conspiracy theories operate in the European online ecosystem, as well as monitoring their offline effects, we will be able to make informed recommendations to civil society stakeholders on how to deal with a problem that threatens to undermine trust in democracy, science, and even truth itself.
2Causes, consequences, and solutions. Findings from a qualitative study on disinformation perception across generations in five countries (Germany, Poland, Romania, Spain, and the UK)

Denis Halagiera (Adam Mickiewicz University), Patrick van Erkel (University of Antwerp), Luisa Gehle (Johannes Gutenberg University of Mainz), Christine E. Meltzer (Johannes Gutenberg University of Mainz) &Ludovic Terren (Open University of Catalonia) (THREATPIE)

Concerns about a declining supply and quality of the news and public affairs coverage have been raised throughout the years (de Vreese et al., 2017). Recently, disinformation has been perceived as an acute threat to democratic societies (Bennett & Livingston, 2018). Despite a significant number of scientific research on the identification and characterization of the dynamics of the dissemination of false information on social media (inter alia: Allcott & Genzkow, 2017; Vosoughi, Roy & Aral, 2018), less is known about citizens' perspectives on this phenomenon and the way they deal with this threat. The main aim of this paper is to share recent findings on disinformation perceptions. We address the following research questions: (RQ1) According to citizens, what are the causes of spreading disinformation, and what are the sources of false information? (RQ2) In citizens’ view, what would be the consequences of disinformation (at personal and social levels)? (RQ3) What solutions to the spread of false information are offered by regular citizens? (RQ4) How often and using what source of information do citizens come across news that they suspect is false, and how do they react? To provide answers to these questions we conducted both qualitative (focus groups) and quantitative (survey research) studies in five European democracies with different political and media systems (Germany – democratic corporatist democracy; Poland and Romania – post- communist democracy; Spain – polarized pluralist democracy; the UK – democracy in turmoil). This allowed us to get an insight into perceptions, opinions, and experiences of two specific groups: younger (18-25 years old) and older (55+) media users. Additionally, we focus on differences across democracies with different political heritages, democratic traditions, media systems (Brüggemann et al., 2014), and different news consumption habits (Shehata & Strömbäck, 2011).
3The Usage of Social Media in the Fight for Queer Rights in Poland

Magdo Chuchracka (Goethe University Frankfurt) (POPBACK)

This Master Thesis is an experimental study reviewing, contrasting, and bringing together theoretical positions of queer studies, postcolonialism, and the binarity of the “East vs. West” divide. This is a contribution to the theoretical discussion of postcolonial studies at intersections with gender studies and queer studies, pointing to desiderata of these theoretical discussions and suggesting contributions to the same. Through a lens critical of the westeronormative tools and ideas of Gender Studies and Queer Theory, the thesis aims to redirect the tendencies of contemporary academia. The Butlerian thought is questioned and repurposed to highlight the and the illegitimization of gender identity using the example of the discrimination of LGBTQIA+ individuals in Poland. Thes study revolves around the LGBTQIA+ community in Poland. The author presents which political and juridical restrictions, illegitimizations and relations of violence the LGBTQIA+ movement is increasingly exposed to. The recent historical as well as socio-political context for the LGBTQIA+ movement in Poland are presented as a prelude to the discussion of gender and power relations. It thus investigates how queer identities are represented online. The author proposes, that social media and progressive digitalization enable a continuation of the activist movement in and with the new media. Qualitative interviews of 45 to 60 min length with Polish activists of the LGBQIA+ movement supplement the theoretical premises and scope with regard to the results obtained. One of the objectives of this thesis is to further dismantle the Western-centered approach to sexuality, gender, and media. Dealing with the duality of the West vs.
East divide, it aims to recognize the patterns of compulsory binarity of gender. How they uplift patriarchy and the colonial approach to research. The author pledges that “There is social power that lays in the presentation and perception of LGBTQIA+”.
4Acts of Citizenship in Cleansing the Internet of Online Hate

Todd Sekuler (Humboldt University of Berlin) (CrimScapes)

The notion of citizenship has been reframed in the social sciences and humanities from a formalised legal or political status to include practices and narratives of claim-making and socio-political inclusion and exclusion, often described through the lens of participation, belonging, and subjectification. For various reasons, landscapes of criminalisation (i.e. crimscapes) offer uniquely complex fields through which to appreciate the multiple and contradictory dynamics of contemporary citizenship forms and practices. In my ongoing research about the criminalisation of online hate, I interviewed content moderators engaged in moderating the increasingly criminalised landscape of hateful postings on the internet. While journalistic accounts of their work have focused primarily on the implicated scars and traumas that follow from moderation, this presentation points to the fact that these moderators also often talk about the necessary and principled nature of their jobs. If citizenship dynamics entail taking on certain responsibilities to secure the preservation and protection of rights and entitlements, I argue that the work of content moderators might be categorised both as protector of rights and as the further sacrificing of personal sovereignty. Moreover, I show how moderators can come to develop creative ways to enact their own forms of resistance and claim- making, thereby infusing acts of citizenship with their own political interests and dispositions. While criminalisation can involve the direct or indirect revocation of one’s rights, the crimscape of hate speech evidences that it can also involve the proliferation of nonstate, and thus not-elected, actors, such as non-governmental
not-for-profit or for-profit organisations and corporations, and their integration into regulatory or extra-legal processes out of moral, economic or political interests.
5Good Targeting, Bad Targeting? How Perceptions of Scientific and Subversive Data-Driven Campaigning Influence Negative Affect toward Targeted Political Campaigns

Marlis Stubenvoll (University of Vienna), Selina Noetzel (University of Vienna), Alice Binder (University of Klagenfurt), Melanie Hirsch (University of Vienna) & Jörg Matthes (University of Vienna) (DATADRIVEN)

Scholars have argued that data-driven campaigning (DDC) has fundamentally changed political parties’ capacity to persuade and mobilize voters (Zuiderveen Borgesius et al., 2018). However, with the emergence of sophisticated targeting strategies, the electorate has grown more concerned over the use of DDC (Auxier, 2020; Turow et al., 2012). The current study investigates how strongly German citizens perceive the use of two types of DDC strategies by politicians in their country, and how these perceptions influence affective responses to DDC. Following Roemmele and Gibson (2020), we distinguish between perceptions of scientific and subversive DDC. Scientific DDC uses data and scientific methods to optimize campaign outcomes, for instance by using forecasting models. Conversely, subversive DDC is characterized by “deliberate attempts to divide, demobilize, and disinform citizens” (Roemmele & Gibson, 2020, p. 607). While both forms of targeting might elicit negative affect toward DDC, we argue from a social contract perspective (Miyazaki, 2008) that perceptions of subversive DDC might do so to a greater extent.
PANEL 3.3 Technology vs. Democracy Friday, 23.09.2022
14.30 – 15.45
Chair: Henning Finseraas
1"Why am I seeing this regulation?": Exploring underlying issues from the proposed Political Advertising Regulation

Miikka Hiltunen (University of Helsinki - Erik Castren Institute) & Sam Wrigley (University of Helsinki) (INDIGO)

Following the Cambridge Analytica scandal, online political advertising has come under intense public scrutiny and much has been written about its potential to damage both democracy and society writ large. It is therefore unsurprising that, in 2021, the European Commission introduced a proposal to regulate the transparency of political advertising and their targeted delivery online. While the legislative process is still ongoing, the way the Union understands political advertising—and, importantly, its problems—is already visible in the proposal. This article looks at three important discussions which underlie the EU's approach. These concern, first, the distinction between the political and the commercial; secondly, the distinction between "online" and "offline"; and, thirdly, the distinction between the autonomous and socially embedded view of the citizen, which informs how political influencing is understood in the proposal. In examining these distinctions, this article seeks to explore and encourage reflection on their consequences for the legal and regulatory picture, and particularly on how the law should tackle such a complex topic as online political advertising.
2The Bright Side of Data-Driven Campaigning? How Targeting can be Beneficial for Society

Kate Dommett (University of Sheffield), Rachel Gibson (University of Manchester), Esmeralda Bon (University of Manchester), Sanne Kruikemeier (Wageningen University & Research) & Sophie Lecheler (University of Vienna) (DATADRIVEN)

To date, coverage of data-driven campaigning has tended to focus on its negative consequences for democracy and has sparked numerous calls for platforms and policymakers to limit or ban this activity. Less attention has, however, been paid to the potential for (online) political microtargeting to enhance rather than undermine democracy. In our paper, we focus on the positive outcomes that can be advanced through political microtargeting. We argue, for instance, that targeting can be used to reach and inform citizens who are less engaged in politics or more likely to tune out from traditional media. Moreover, it provides a democratic linkage between politicians and the electorate. In the second half of the paper, we take a closer look at the conditions under which more positive endeavors of political targeting can be regarded as desirable from a citizen perspective. To put these claims into perspective, we will present data on public attitudes towards the use of different data sources for targeting to investigate the boundary conditions of targeting.
Results indicate that all data should not be viewed as equal and that certain types of data are appropriate for advancing positive goals. The paper will offer more nuanced insights into the democratic role of targeting by moving the debate of targeting in a different light by focusing on its potential.
3The Regulatory Ecosystem of Data Driven Campaigning

Andrew Barclay (University of Manchester), Rachel Gibson (University of Manchester) & Kate Dommett (University of Sheffield) (DATADRIVEN)

Recent elections have increasingly been described as data-driven, with parties using large quantities of voter data to target their campaign messages in ever more granular ways, particularly online. These practices have increasingly been facing calls for greater regulation, however data-driven campaigning encompasses a range of different practices which may in turn come under the remit of several different regulatory actors, as well as government and social media bodies. As a result, we know little about which institution(s) are currently most active in their oversight of campaigns’ use of data, as well as which factors best explain their approach to regulation in this space. Our paper addresses this by using interviews and an elite survey of election, media and data protection regulators to map the regulatory ecosystem of data-driven campaigning in British elections. We find that the scope and powers of each regulatory body are limited by legislation to some extent, but that this is less true for election regulators who are the dominant actor in regulating data-driven campaigning practices. We also find that those bodies with the most regulatory power tend to attach greater importance to ensuring that campaigns are being transparent in how they use data to target voters. Finally, we also highlight that social media platforms themselves are key actors within this regulatory landscape, as their policies determine the access that parties have to their users’ data, and thus their ability to target campaign materials at them.
4When do fact-checks work best? A cross-national experiment on the influence of the source and degree of politicization of fact-checks on the effectiveness to counter misinformation

Patrick van Erkel (University of Antwerp), Peter Van Aelst (University of Antwerp), Claes De Vreese (University of Amsterdam) & David Hopmann (University of Southern Denmark) (THREATPIE)

The past decennium has seen a strong surge of mis- and disinformation. One solution that is often suggested to counter this trend is the use of fact-checks. While many studies have demonstrated that these can be effective tools in countering misinformation, less is known about the conditions under which fact-checks are effective, which is what we aim to examine with this study. Concretely, we look at two factors. First, we are interested to what extent the source of the fact-check matters, comparing fact-checks coming from traditional media with fact-checks from independent organizations. We expect that the former are more effective, as citizens are more familiar with this news source and hence may place more trust in it.
However, this may be moderated by trust in the media, both at the individual and country level. Second, we expect that the effectiveness of fact-checks depends on how politicized they are. Our expectation is that more politicized fact-checks are less effective, as they may prime partisan identity and strengthen partisan motivated reasoning. We expect this difference to be more prominent in polarized countries. To test our hypotheses we conducted an experiment across 16 European countries. In all countries respondents read a misinformation story with a national (right-wing) political actor making a false claim about European immigration numbers.
Subsequently, respondents read a fact-check debunking this misinformation. The fact-check differs on two dimensions, namely 1) whether it comes from the public broadcaster or from an independent fact-check organization, and 2) whether it is politicized or depoliticized. The control group only reads the misinformation article.
Afterwards, respondent are asked to what extent they accept the false claim, and find the misinformation article credible. Additionally, we study how country-level variables such as levels of media trust and polarization moderate the effects.
5A qualitative examination of citizens’ (political) media diets across generations in five countries

David Hopmann (University of Southern Denmark), Agnieszka Stępińska (Adam Mickiewicz University), James Stanyer (Loughborough University London), Denis Halagiera (Adam Mickiewicz University), Ludovic Terren (Open University of Catalonia), Luisa Gehle (Johannes Gutenberg University of Mainz), Christine Meltzer (Johannes Gutenberg University of Mainz), Raluca Buturoiu (National University of Political Studies and Public Administration), Nicoleta Corbu (National University of Political Studies and Public Administration), Ana S. Cardenal (Open University of Catalonia) & Christian Schemer (Johannes Gutenberg University of Mainz) (THREATPIE)

Recent research has raised several concerns about citizens in contemporary media rich democracies, particularly younger citizens (e.g., Hills, 2019). These concerns include: citizens lacking discernment in the media they consume; citizens avoiding political news and information; or selecting information that confirms their prejudices (e.g., Van Aelst et al., 2017). While the volume of research on these topics has grown, it is however notable that studies exploring these issues rarely discuss in-depth these matters with citizens. Prior findings tell us that using the diet analogy is a particularly effective way for citizens to make sense of their existing practices and to identify healthy normative practices (Marcu et al., 2015). Although the concept of citizens’ “media diets” has received increasing attention (Young & Anderson, 2017), it remains vague and not fully developed. The aim of this study is threefold. First, we are interested in mapping perceptions of the main types of media content citizens, across countries and age levels, can choose from within the current media landscape (perception of the first-person media use). Second, we identify citizens’ perceptions of ‘healthy’ media content that citizens should consume to develop sound attitudes and behaviors (normative perspective). Third, we explore how citizens perceive the others’ consumption patterns of various types of media content within the current media landscape (perception on the third-person media consumption). To achieve these goals, we conducted a series of focus group interviews (between April and July 2021) with younger adults (18-25 years old) and older adults (above 55) in five European countries (Germany, Spain, Poland, Romania, and UK). These countries vary on several key contextual factors relevant for the study of the political information environments, including “young” and “new” democracies with different political heritages, democratic traditions, and media systems and different news consumption habits. Focus group interviews as a method provided us with the data needed for an in-depth assessment of people’s perceptions. The results revealed a clear cross-country generational difference (with some country variations) in perception of the first-person media use. At the same time, we found similarities across generations and countries in normative dimensions (description of the healthy media diet). Finally, while tracing the perceptions of the dietary performance of other citizen, we noticed a third-person effect: each generation believed that the other had a poorer media diet: the youth considered that the elderly are more prone to selective exposure, disinformation, and a less diverse media diet, while older citizens talked about the almost total lack of interest for (political) news of younger generations.
PANEL 4.1 New Perspectives on Political and Civic Engagement Saturday, 24.09.2022
09.00 – 10.15
Chair: Gerhard Schnyder
1Your efficacy or mine? Determinants of support for democratic innovations and the willingness to engage in them

Lien Smets (Vrije Universiteit Brussel), Didier Caluwaerts (Vrije Universiteit Brussel), Bram Spruyt (Vrije Universiteit Brussel), Rebekka Kesberg (University of Sussex) & Matthew Easterbrook (University of Sussex) (UNDPOLAR)

From the beginning of the 1990s onwards, Western European countries witnessed the contours of a widespread crisis of democracy. It is in this turbulent period that the idea of democratic innovations was coined. Instead of entrusting political power exclusively to elected officials, advocates of democratic innovations argued that governance structures needed to increasingly rely on citizen involvement to generate legitimacy and accountability. These democratic innovations have mushroomed in policy making, and much attention has hitherto been devoted (1) to the transformative effects of these innovations, i.e. the question whether they can lead to opinion, identity and attitude changes, (2) to their policy impact, (3) to their legitimacy, and (4) to the question who supports these innovations. However, the question that has only received scant attention (and mostly from qualitative researchers) is why citizens would be willing to engage in such innovations themselves. After all, it is one thing to support democratic innovations ideationally, but is it quite another to be willing to engage with them. Based on a Norface UNDPOLAR population survey (N=6.371) conducted in five countries (Denmark, Greece, the Netherlands, the UK and Poland), we will argue that the drivers of support for DIs and the willingness to engage in DIs are quite distinct. Support for DIs is driven mainly by perceptions of other citizens’ political competence; willingness to engage with DIs is driven mainly by perceptions of ones’ own political competence.
2Voices of the Vulnerable: Long-run Effects of Routine. Task Intensity on Union and Political Participation

Harald Dale-Olsen (Institute of Labor Economics) & Henning Finseraas
(Norwegian University of Science and Technology) (TECHNO)

Automation and technological changes have had important and well-documented effects on the labour market, and recent work has started to document effects on political preferences. A related question is to what extent automation creates \left- long-run effects of vulnerability to automation on the political voice of exposed workers, and ask to what extent a weaker socioeconomic position due to automation is reflected in political participation. We leverage Norwegian population-wide administrative data that allow us to measure the routine task intensity (at the four digit level) of employed workers. We follow workers over more than a decade and explore how routine task intensity is related to their political voice, as measured by administrative data on union membership and voter turnout. Furthermore, we will explore labour market outcomes as representing the mediator from routine task intensity to political voice. To improve causal interpretation, we will use (double) negative controls to adjust for unmeasured confounding.
3Voters’ understanding and evaluation of data-driven political advertising

Sophie Minihold (University of Vienna/University of Amsterdam), Sophie Lecheler (University of Vienna), Claes de Vreese (University of Amsterdam), Rachel Gibson (University of Manchester) & Sanne Kruikemeier (Wageningen University & Research) (DATADRIVEN)

Data-driven political advertising (DDPA) has received much attention from scholars. Despite this attention, we lack empirical findings on the role of voters in data-driven election campaigns. As presented at the IJPP conference 2021, voters lack understanding of DDPA (Minihold et al, 2021). This lack of understanding is problematic as it likely affects the assessment of implications that DDPA can have for voters. We argue that voters require extended literacies to count as competent in navigating the changed political landscape in which DDPA is embedded.
Conceptually we follow advertising literacy scholars (Boerman et al., 2018; Rozendaal et al., 2016) and offer a voter-centric approach by investigating their conceptual understanding (CU), and evaluative perceptions (EP) of DDPA. This paper offers 1) a voter typology along different levels of CU and EP, and 2) investigates to what extent voters with different levels of CU and EP differ in their engagement with and avoidance of political ads. Our research relies on data from a multi-wave panel survey study conducted during the German Federal Election in August and September 2021(NW1= 1914, NW3 = 1303). Latent Class Analysis guides us in the creation of a voter typology. So far, results indicate that understanding and evaluation of DDPA are related. In other words, understanding data-driven tactics and strategies of political actors matter for evaluating DDPA, and vice versa. Furthermore, CU matters for engaging with political campaign ads, but is not related to ad avoidance. However, we find that a negative evaluation of DDPA, such as perceiving DDPA as creepy, is related to actively avoiding data- driven campaign ads. Our findings imply that understanding and evaluating data- driven campaigning techniques and implications matter for voters’ campaign (dis)engagement. With this paper we thereby contribute to a long-neglected discourse about voter agency and empowerment in increasingly digital campaign practices and democratic governance.
PANEL 4.2 “Common” Values and Democratic Behaviours Across Europe Saturday, 24.09.2022
10.30 – 11.45
Chair: James Stanyer
1Belief systems in an international landscape: Exploring the varieties of belief systems across 23 European countries

Jochem van Noord (University of Groningen) (UNDPOLAR)

We investigate the structure of political belief systems across Europe. We answer three questions: First, are political belief system structures similar across Europe? Second, which demographic groups are likely to have similar belief systems within countries? Third, how are belief systems related to voting behavior? We do this in 23 European countries using the 2016 wave of the European Social Survey.
Correlational Class Analysis results indicate that a wide variety of belief systems exist in Europe (2-5 per country), but that these can be summarized into two groups. Although one group is considerably more homogenous than the other, on average, the groups do not differ in constraint (i.e., strength of association between beliefs). Belief systems of both groups contained cultural and economic belief dimensions, which were related positively in the first, but negatively in the second group. Belief systems of the first group were more likely to be from Western European countries and its members more likely to be higher educated compared to the second group. Membership in the second group was associated with more anti-establishment voting and vote abstention. Thus, by comparing and contrasting belief systems we find systematic variation in belief systems across demographic groups and countries.
2How Human Values and the Rural-Urban Divide Interact to Shape Voting Behaviour

Alina Zumbrunn (University of Bern) & Sonja Zmerli (Sciences Po Grenoble - Université Grenoble Alpes) (RUDE)

While the relevance of values and value orientations have for long been at the forefront of political sociology, Shalom Schwartz’ encompassing conceptualisation of basic human values and their interrelatedness has only recently started to inspire political science research. Recent empirical studies demonstrate, for instance, that some of Schwartz’ values are politically more consequential than others and that citizens are as much influenced by their value dispositions as are elected officials. To be sure, the extent to which human values will contribute to a better understanding of political involvement still needs to be explored. In this vein, this paper investigates the role a set of pre-selected human values, i.e. power, security, tradition and universalism, plays in voting for political parties along the ideological spectrum by combining the Manifesto Project’s content analysis of over 1’000 political party programmes with individual-level voting data from all nine waves of the European Social Survey. To advance this realm of research even further, we also examine the relevance of the re-emerging rural-urban divide in democratic societies and scrutinize, in particular, the manifold interactions between our set of human values and place of residence in a comparative perspective. At first, our comparative analyses with around 130’000 observations from 14 countries reveal a clear trend to vote for left-leaning parties in the aftermath of the financial and economic crisis with a right-wing uplift in 2018 and a tendency for rural areas to favour conservative parties. Second, human values depict only minor changes over time, yet are fairly differently distributed over urban and rural zones, with power and universalism showing the smallest and security and tradition demonstrating the largest differences across places of residence. Third, the analyses, which control for country- and time-fixed effects, strongly suggest the impact of place, with living in big cities and in the countryside on the opposite ends of the dependent variable, and human values with a particularly strong impact of universalism. Finally, the follow- up inspection of interaction effects provides a more nuanced picture which underlines the importance of place for human values to affect voting behaviour, yet simultaneously calls for caution not to embrace pre-emptively a seemingly simplistic rural-urban divide in this regard.
3The Relationship between Political Entertainment Media Use and Political Trust: A Comparative Study in 17 Countries

Christian Schemer (Johannes Gutenberg University of Mainz), Christine Meltzer (Johannes Gutenberg University of Mainz), Luisa Gehle (Johannes Gutenberg University of Mainz), Peter Van Aelst (University of Antwerp), Yannis Theocharis (Technical University of Munich), Jesper Strömbäck (University of Gothenburg), Václav Štětka (Loughborough University), Agnieszka Stępińska (Adam Mickiewicz University), James Stanyer (Loughborough University London), Sergio Splendore (University of Milan), Tamir Sheafer (Hebrew University of Jerusalem), Jörg Matthes (University of Vienna), Karolina Koc-Michlska (Audencia Business School), David Hopmann (University of Southern Denmark), Frank Esser (University of Zurich), Claes de Vreese (University of Amsterdam), Nicoleta Corbu (National University of Political Studies and Public Administration), Laia Castro (University of Zurich), Ana S. Cardenal (Open University of Catalonia), Toril Aalberg (Norwegian University of Science and Technology), Patrick van Erkel (University of Antwerp), Ludovic Terren (Open University of Catalonia) & Denis Halagiera (Adam Mickiewicz University) (THREATPIE)

Triggered by the changes in the supply (e.g., diversification of the news environment, Van Aelst et al., 2017) and demand for political information (e.g., increase in news avoidance Prior, 2007), understanding the effects of soft or hybrid forms of political communication have begun to grow (Baym & Holbert, 2020).
Outlets and shows that vary in fictionality and entertainment value can also convey political messages and affect people’s attitude toward political actors and the political system. For instance, frequency of exposure to talk shows can reduce political trust (Guggenheim et al., 2011; Mutz & Reeves, 2005), and political satire and comedy can affect political attitudes negatively (Burgers & Brugman, 2021).
Even fictional media have been shown to affect political attitudes in the audience (Jones & Paris, 2018; Morgan & Shanahan, 2017). There are two major shortcomings in extant research, however. First, most studies focus on single genres, e.g., how the frequency of comedy use is related to political outcomes. Second, most research is not comparative in design, but carried out in a U.S. context. The question therefore is whether findings from a liberal media system with weak public service broadcasting and high political polarization can be generalized to other media and political contexts. The present study addresses both limitations by examining the relationship between the use of three hybrid genres (i.e., political talk shows, political comedy, and political dramas) and political trust in 17 developed democracies. A survey study fielded in May 2020 assessed people’s media diet controlling for a host of news media use measures (for details, see Castro et al., 2021). The findings show that frequency of exposure to hybrids of entertainment and information is positively related to trust in government in most countries.
Specifically, talk show exposure and use of political comedy outlets is positively related to trust in government in most cases. In other countries, there is no relationship at all. In a few countries, exposure to fictional political dramas is also associated with higher trust in government. Unlike prior research, the study did not find any negative relations between hybrid political content and trust in government. Thus, the general concern that exposure to politainment is detrimental to political attitudes in the public is not warranted. The data were collected at the beginning of the COVID-19 pandemic. Therefore, the positive relationship of politainment exposure and trust in government may be due to an increase in information exposure and increase in trust in most countries at the beginning of the pandemic. A replication study is currently in the field and will help interpret the findings from the survey that was conducted in the first phase of the pandemic. Based on these findings we can parse out the effect that the pandemic may have played.
PANEL 5.1 Evolving Democratic Legitimacy Friday, 23.09.2022
14.30 – 15.45
Chair: Allison Östlund
1Variation In Democratic Legitimacy Of Occupational Pensions

Tobias Wiß (Johannes Kepler University Linz) & Thomas Mayer (Johannes Kepler University Linz) (DEEPEN)

In times of a growing importance of asset-based welfare such as funded occupational pension schemes, questions about their democratic legitimation arise due to financial market downturn and negative or ultra-low investment returns.
However, we lack knowledge about democratic input-legitimacy of funded pensions and its variation across countries. Therefore, we map the linkages between different occupational pension schemes and their democratic input-legitimacy. We measure input-legitimacy by distinguishing individual and collective-representative inclusion in decision-making processes of funded occupational pension schemes. We hypothesise that democratic legitimacy increases with the importance and the degree of compulsion of occupational pensions. To test our hypothesis, we compare voluntary occupational pension schemes in Austria, Spain and Ireland with quasi- mandatory schemes in the Netherlands and Denmark. Germany serves as a test case as occupational pensions are voluntary, but recent reforms increased their importance and coverage rates (e.g. individual entitlement to occupational pension plans).
2The adoption of digital regulatory practices for consumer empowerment: do country and sector boundaries matter for digital regulatory innovation?

Alena Pivavarava (King's College London) & Christel Koop (King's College London) (ReConnect)

This study bridges two emerging areas in the literature of economic regulation: the adoption of digital practices by national regulatory agencies (NRAs) and the empowerment of consumers for e-participation in regulatory processes through digitally enhanced consumer-facing platforms created by NRAs. With an aim to re- connect the regulatory state with citizens, we designed a novel framework, which comprehensively captures the digital capacity of NRAs and disentangles the multiple channels for consumer digital engagement with NRAs across the key regulatory procedures, including information provision, communication, education, and decision-making. To measure the extent of digitalisation of regulatory functions, we derived a composite index of digital adoption for 236 individual NRAs from a newly collected cross-sectional dataset, which spans across 42 geographically disperse OECD and non-OECD countries with distinct regulatory and institutional pressures and public proactiveness in digitalisation. With the applied multi-level analysis, we uncover and visualise the cross-country and cross-sectoral differences in the adoption of digital regulatory practices among the agencies with sole- and multi- sectoral competencies in five economic markets from utilities to financial services. Through modelling the impact of organisational-, industry-, and country-level factors on the digital NRA indices, we address an overlooked question of whether the borderless nature of digital technologies allows to overcome the gap in adoption of consumer-oriented regulatory practices. With these insights, this study offers implications for both improving the effectiveness of regulatory procedures through consumer-oriented digital transformation, as well as government initiatives for enhancing digital trust and e-participation in economic regulation among consumers.
3The complexities of representation in funded pension schemes in Ireland

Karen Anderson (University College Dublin) & Hayley James (University College Dublin) (DEEPEN)

Representation is an important aspect of democratic governance in pension schemes. In Ireland, workplace pensions operate at the employer level as trusts operated in the members’ interests. Yet, collective representation of members in trusts in Ireland is not obligatory and covers less than half of trustees (the remainder being employer-nominated). The lack of compulsion means there is little centralised reporting on pension scheme representation. Moreover, the role of the trustees (both employer- and member-nominated) is changing: first, due to the shift from DB to DC, changing the trustee’s focus from delivering good outcomes to good administration, and second, following the implementation of additional governance requirements under IORPSII. We investigate representation in the landscape of workplace pensions in Ireland through documentary analysis and expert interviews with stakeholders in Irish pensions. We find that representation occurs in different ways, split between minimalist and maximalist approaches. At the minimalist end of the scale, where DC schemes are predominant, collective representation appears to be overtaken by notions of individual choice. Yet, despite little struggle to maintain collective representation, very few members make active choices about their pensions. Members are reliant on scheme defaults, which are designed by trustees based on professional advice in an arm’s length manner. At the other end, some boards are engaged in understanding the needs of their workers and member representatives play a key role in ensuring scheme design suits the specific context. Additionally, there are examples of different forms of collective representation, such as works councils, especially in Master Trusts (where there is no provision for trustee representation) or quasi-state schemes (where decisions are ultimately taken by the government). Yet this good practice is limited to specific contexts and its extension may be hampered by the burden of regulatory changes, such as IORPSII and automatic enrolment.
4The Norwegian Administrative State: Towards a Crossroads?

Nick Sitter (BI Norwegian Business School) & Are Vegard Haug (BI Norwegian Business School) (ReConnect)

In the international literature on welfare states, the rule of law and democracy, the Nordic model is typically emphasized as a success. Both the European Union and the World Economic Forum have lauded the Nordic examples, and these countries stand out in the Human Developmental Index, the Global Competitiveness Index, the Social Mobility Index, and global liberal democracy indexes (LDI /V-dem). But is the Nordic model at a crossroad? Liberal democracy, the rule of law and the administrative state is under attack across the west. Citizens express disillusionment with politics, populist parties are on the rise, and party systems have fragmented.
The populist left and right question some of the core pillars of liberal democracy: the rule of law, separation of powers, and citizen rights. The central question in this paper is whether we see an erosion of the democratic, social and economic values that constitute the cornerstones of the Nordic model? Focusing primarily on Norway and on current reforms and debates, the ambition of this article is threefold. First, it identifies the key components of the Nordic model, as well as its types and preconditions. Secondly, the status of what the Economist called the ‘supermodel’ is examined in light of changes on two Stein Rokkan’s core institutional channels: numerical democracy and corporatism. Today, both channels have similar challenges – a comprehensive weakening since its heyday in the 1960s and 1970s.
The paper analyses the key challenges or dilemmas that must be addressed. The third part of the paper explores responses to these challenges, both in terms of party politics and reforms of the administrative state. The conclusion returns to the question of whether we are at a possible crossroads in the Nordic model, and what lessons can be drawn from Norwegian efforts to reconnect the administrative state with citizens and maintain citizens’ trust and confidence?
PANEL 5.2 Understanding Global and EU Institutions Saturday, 24.09.2022
09.00 – 10.15
Chair: Kamil Bernaerts
1Transfer Learning for Multilingual Political Text Classification

Fabian David Schmidt (University of Mannheim) & Goran Glavaš (University of Mannheim) (EUINACTION)

The text-as-data PolSci community increasingly relies on supervised text classification models -- that is, machine learning models trained to predict discrete categories assigned to texts -- for various tasks, e.g., predicting policy areas of legislative text or predicting topics/themes in manifestos of political parties or speeches delivered in parliaments. Training such machine learning models requires non-negligible amounts of labeled data (i.e., texts manually coded with category labels by domain experts), which is, in general, both expensive and time-consuming. Multilingual use cases, in which one needs to predict the categories for texts written in two or more languages, pose an additional challenge that has, to a large extent, been unaddressed in this community. For many tasks, the manually coded texts exist only in English or a handful of (major) languages: without the help of multilingual representation learning and cross-lingual transfer, such models cannot be transferred to languages unseen during training, that is, they cannot make any predictions on texts from languages for which there are few or no coded texts. In this talk, we will present state-of-the-art methods for multilingual text representation
-- based on pretrained multilingual language models – that enable zero-shot (i.e., when there are no coded texts in the target language(s) at all) and few-shot (when there is only a handful of labeled text instances in target language(s)) cross-lingual transfer of text classification models. We will then present the results achieved by these approaches in use cases from the EUINACTION project: (i) policy area classification and (ii) identification of European relevance for political texts -- parliamentary speeches and legislative proposals.
2EU Treaty-Making and the Practice of ‘Common Accord’: Creating Unity within a Composite of Separated Powers

Christina Eckes (University of Amsterdam and Amsterdam Centre for European Law and Governance) & Anne Thies (University of Reading) (SepaRope)

‘[C]ertain tensions are […] inherent in the design of the European Union.’ BVerfG, PSPP ruling, 2020 Inherent tensions between the powers of the EU and of the Member States have become particularly apparent in post-Lisbon external relations. The request for increased international capacity and political autonomy of the EU continues to clash with Member States’ political struggles for visibility and control on the global stage. Interinstitutional disputes have popped up like mushrooms, supported by large numbers of Member States. An increasing number of requests for opinions by the Court of Justice reveals further disagreement in the context of international treaty-making. The Court of Justice has thus been increasingly asked to further clarify how best to protect the functionality of the EU as a global actor under EU law, while recognising the legal and political need to accommodate Member States' interests, standing and obligations as 27 separate global actors.
This paper exposes the legal difficulties arising when attempting to ensure the EU’s capacity as global actor in the context of international treaty-making. The paper distinguishes in its analysis between EU law framing the internal decision-making related to EU external action as well as the EU’s capacity on the global stage on the one hand, and the international law context in which both the EU and its Member States operate on the other. The paper unpacks such legal difficulties by distinguishing between the implications of the construction of power relations within the EU, EU law and EU membership for (1) Member State discretion/scope for manoeuvre in international treaty making (‘constraints on MS'), (2) the limits to the EU’s capacity to contribute to treaty-making (‘constraints on EU’), and (3) the coordination between the EU and the Member States in the context of international treaty-making both within and outside the EU’s institutional setting, which is framed also by national and international law (‘scope for coordination’). The paper argues that the balancing act between the requirements of international law and the complex internal power structures of the EU is a political one and cannot be reduced to a legal question. Moreover, even though the CJEU has recognised the significance of the ‘perceived legitimacy and effectiveness’ (Opinion 1/19) on the international stage of what could be called the composite of the EU and its Member States as (legally) separate global actors, the protection of such cannot be guaranteed by EU law alone but requires the political commitment of the EU institutions and Member States to act in unity.
3A review review

Allison Östlund (University of Gothenburg) (SepaRope)

The present article explores the judicial review carried out by the CJEU, based on a total selection of CJEU rulings within three fields: trade, migration and EMU, collected by the Gothenburg Team of the SepaRope project. The article explores findings emanating from a subset of variables pertaining to legal basis, legislative competence review, treaty violation findings, the striking down of measures, intensity of judicial review and the CJEU's measure of deference to other non- judicial decision-making actors. The analysis relies on Bell's three accountability modes; the fiduciary model, the representative-elector model and principal-agent theory. Of these, the first trust-based model, describing the reviewed decision-maker as split in its loyalties between an electoral public and an instructing principal, is the most useful for approaching judicial review in the EU law context. With this as starting point, the dataset's findings are sorted by Widdershoven's typology for standards and intensity of review. Although the results are thus far preliminary, the study indicates that the CJEU's review in direct and indirect validity cases leads to a striking down of contested acts in almost half of cases brought within the subject- areas covered by the study. Among these, lack of competence with the adopting actor was the most prevalent invalidity ground. The review carried out was, in Widdershoven's terminology, predominantly substantive and limited, followed by procedural and least commonly strict and substantive. That substantive and procedural review both figured is consistent with Bell's trust model insofar that – in my extrapolation of it – accountability towards the represented group would emphasize substantive review whereas accountability towards a principal would be able to include both substantive and procedural review (depending on how the mandate is formulated).