Monday, June 8
13.20–14.20
Room: Antarktis
Chaired by
Marina Nord
Postdoctor
University of Gothenburg
Presenter
Robin Gut
Doctoral Student, Centre for Democracy Studies Aarau (ZDA)
Title
Measuring the Quality of Referendums – A Proposal for a Composite Index
Referendums come in many varieties and take place in different contexts. In some cases, referendums increase the legitimacy of political decisions and empower citizens. In other cases, they are instrumentalized as democratic window dressing by those in power.
In this paper, we conceptualize the quality of referendums and develop an analytical framework with which to assess them. We argue that to be of high quality, a referendum needs to take place in a democratic context, be conducted according to an established legal and procedural framework, and allow citizens to freely form and express their opinion.
For the empirical analysis, we distinguish three dimensions of a referendum – context, instrument, and popular vote. The context dimension encompasses the political environment in which a referendum takes place. The instrument dimension comprises the institutional setting of the referendum, and the popular vote dimension its implementation. Each dimension is operationalized with several indicators which are then aggregated into the Referendum Quality Score.
We test the Referendum Quality Score on five referendums in 2025 in Ecuador, Guinea, Hungary, Italy, and Taiwan. We find empirical variance in all three dimensions and vast differences in referendum quality. We conclude that the RQS allows for differentiated assessment and discussion of referendum quality.
Presenter
Alexander Hudson
Senior Adviser, International IDEA
Co-Author:
Seema Shah, International IDEA
Title
The Second Boundary Problem in Democracy Measurements: How Do We Account for External Influence?
More Information
Global State of Democracy Indices | The Global State of Democracy
Among the contested aspects of democracy is the boundary problem: who is included in ‘the people’ who rule. But there is a second boundary problem for democracy, namely, that democratic performance is subject to transborder influences. Each state has some level of influence on the democratic performance of others and is itself empowered or constrained by external forces. This point surfaces when we observe that countries that are reasonably democratic internally engage in actions externally that harm democracy, such as in warfare, support for separatist or rebel groups, and various other activities short of war that seek to undermine a foreign government or subvert its democratic processes. At present, there are some indicators that relate to these dynamics, including V-Dem’s measure of ‘Government dissemination of false information abroad’, the UCDP External Support Dataset, and Transnational Repression data from Freedom House. This paper elaborates some of the theoretical challenges in a concept such as ‘external democracy’ and maps the indicators that are most relevant to such a concept. The paper evaluates the utility of an external democracy concept and the extent to which it can be measured.
Presenter
Melina Liethmann
Data Manager, University of Gothenburg
Co-Authors:
Stefan Dahlberg, Mid Sweden University
Dennis Andersson, University of Gothenburg
Maximilian Hornung, University of Gothenburg
Title
A Latent Measure of Satisfaction with Democracy in Europe, 1989–2024
Satisfaction with Democracy (SwD) is one of the most widely used indicators of democratic legitimacy and regime performance. Although measured in numerous cross-national survey projects over several decades, differences in question wording, response scales, survey designs, and country coverage complicate the construction of comparable longitudinal estimates. Existing harmonization approaches either harmonize survey instruments directly or estimate latent public opinion from multiple indicators. Yet SwD presents a distinctive challenge because it is typically measured using only a single survey item. We ask whether latent-variable harmonization can produce valid long-run trends for a concept that is commonly measured by a single item. Bayesian item response theory (IRT) models are usually applied to concepts measured by multiple items. This paper bridges both challenges .
The paper contributes to the literature on public opinion measurement by examining whether ordinal latent-variable harmonization remains feasible when the latent construct is observed primarily through single-item indicators rather than multiple survey items. Using 14 European survey projects from 1989 to 2024, we construct crosswalks for SwD items and estimate Bayesian ordinal latent-variable models of country-year SwD. If this proves successful, we compare these estimates with existing harmonization strategies for single-item indicators, such as linear rescaling, observed score equating, and assess construct validity by comparing SwD trends with trends in trust in electoral and judicial institutions.
Presenter
Svend-Erik Skaaning
Professor, Aarhus University
Title
On the Conceptualization and Measurement of Democratic Crisis
Claims about a global crisis of democracy have become increasingly prominent, yet existing attempts of conceptualization and measurement often rest on vague and broad definitions and sets of indicators as well as ad hoc aggregation strategies. As a result, many diagnoses conflate minor democratic malaise or challenges to other features of the political system with situations in which democracy as a regime form is genuinely at risk. This paper presents a more disciplined framework for conceptualizing and measuring democratic crisis, understood as a high-risk inflection point threatening the core institutions of electoral democracy: free and fair elections and the protection of political liberties. The framework distinguishes between three components, treated as distinct diagnostic dimensions: Level captures the fulfilment of constitutive democratic institutions and reflects the underlying “immune strength” of a political regime. Hazard captures contemporary behavioral pressures that strain these institutions, such as violations of institutional restraint, erosion of mutual toleration, and political violence. Trend captures the recent trajectory of these pressures, indicating whether democratic strain is intensifying or receding. We illustrate the framework using comparative behavioral indicators from various sources, such as V-Dem and UCDP, applied to OECD countries in recent years.
Demscore Conference 2025