Voting, political and immigrant attitudes, populism and demographic data:
European Social Survey (ESS): https://ess-search.nsd.no/
The European Social Survey (ESS) is a cross-national survey conducted through face-to-face interviews with representative samples, aimed at tracking changes in public attitudes, values, and behaviors across European countries over time. In our project, we use ESS data from 34 countries, covering ten rounds (2002–2021), to build a harmonized longitudinal dataset. Each round was processed individually and merged using standardized country and regional codes based on the NUTS classification (levels 0, 1, and 2), enabling flexible regional analysis. Variable names were cleaned and standardized for consistency across datasets. As the ESS forms an unbalanced panel—with irregular country participation and regional data availability—we address these limitations through econometric methods, including fixed effects and robustness checks using balanced panel subsets.
Ιntegrated Values Survey (IVS) : https://www.worldvaluessurvey.org/WVSEVStrend.jsp
The Integrated Values Surveys (IVS) dataset, covering 1981 to 2022, combines data from the European Values Study (EVS) and the World Values Survey (WVS), resulting in 452 surveys across 115 countries and territories. For our project, we aggregated all EVS and WVS waves—EVS Rounds 1 to 5 and WVS Rounds 1 to 7—into a harmonized longitudinal dataset with standardized country and regional codes. Integration was conducted using the European NUTS classification at NUTS 0, 1, and 2 levels, allowing for regional analysis and consistency with other datasets. The IVS offers insights into a wide range of topics, including life perceptions, family, work, politics, religion, and national identity. As with other survey-based sources, the IVS forms an unbalanced panel due to uneven country participation and varying availability of regional data over time.
Regional data from Eurostat was collected alongside national-level data within each relevant statistical domain, typically representing a subset of the national variables. The data were merged using the European NUTS classification system at three spatial levels—NUTS 0 (country), NUTS 1, and NUTS 2—ensuring compatibility with other regional datasets and allowing for flexible regional analysis. These regional statistics are used as supplementary control variables in our empirical work and align with similar variables from the WDI. Covering EU member states, EFTA, and enlargement countries, the data span a wide range of topics, including agriculture, business, demography, education, health, labor markets, poverty, and technology.
The Populist Dataset: https://popu-list.org/
The PopuList dataset compiles information on political parties in 31 European countries that are classified as populist, far left, and/or far right, and indicates whether these parties are Eurosceptic. Covering all national elections from January 1989 to December 2022, the dataset enables analysis of the political landscape over time. For integration with other datasets in our project, we aligned the data at the NUTS 0 (country) level using the European NUTS classification. Key variables include the presence and classification of populist, far-left, far-right, and Eurosceptic parties across countries and election years.
Authoritarian Populism Index Data: https://populismindex.com/report/
The Authoritarian Populism Index provides a detailed account of ideological trends in European politics, focusing on parties promoting illiberal and authoritarian views. Covering all national elections from 1945 to 2023 across 31 European countries, it offers historical and comparative insights into the rise of populism and its implications for liberal democracy and market-based systems. For consistency with other datasets in our project, the data was integrated at the NUTS 0 (country) level using the European NUTS classification. Key variables include election results for all parties, vote shares for populist, radical right, and radical left parties, the presence of populists in government, and annual ideological classifications by country.
Manifesto Dataset: https://manifesto-project.wzb.eu/datasets
The Manifesto project relies on primary sources such as political party manifestos and speeches, and to corroborate the validity of the resulting populist classification, he also asks a pool of country experts to validate or reject it by answering an ad hoc questionnaire. The dataset covers 67 countries and includes 1,387 relevant political parties spanning the period from 1945 to 2023. We merged the Manifesto data with the European Union's NUTS classification system at the NUTS 0 level (country level) to use it with the other datasets in our analysis. For integration with other datasets in our empirical analysis, we aligned the Manifesto data with the European Union’s NUTS classification system at the NUTS 0 level, corresponding to the country level.
Quality of Governance Data (Qog): https://www.gu.se/en/quality-government
The QoG EU Regional dataset offers a comprehensive time-series collection of over 350 variables across three administrative levels of European regions—NUTS 0 (national), NUTS 1, and NUTS 2—covering the years 1960 to 2019. With the region-year as the unit of analysis, it provides rich data on demographics, economy, education, health, labor markets, environment, digital society, poverty, quality of government, science and technology, tourism, and transport, making it a valuable resource for regional-level empirical research within Europe.
Eurobarometer: https://europa.eu/eurobarometer/surveys/browse/a-z
The Eurobarometer provides harmonized, individual-level data on political attitudes, trust in EU institutions, and perceptions of European integration, available at multiple regional levels (NUTS 0, NUTS 1, NUTS 2, and NUTS 3), though coverage varies by year and region. For our empirical analysis, we use Eurobarometer data from 1960 to 2020, merged using the NUTS classification system, to test the external validity of our findings and examine additional dimensions of public sentiment toward the EU. Specifically, we include variables capturing trust in EU institutions (European Parliament, Commission, ECB), support for further integration, EU citizenship identity, optimism about the EU’s future, and attitudes toward EU-level decision-making—all of which are coded as dummy variables and used as dependent variables in our models.
IMF Adjustment Programs Dataset (IMF) : IMF Dataset
The International Monetary Fund (IMF) has evolved into a near universal financial institution, growing from 44 member states in 1946 to today. It supports international economic stability and often intervenes during crises through adjustment programs that promote structural reforms, including privatization, deregulation, and labor market liberalization. In our analysis we investigate whether lifetime exposure to IMF programs affects EU perceptions. Using IMF data, we identify whether and when each EU country participated in an adjustment program since 1956. A notable recent example is Greece’s 2010 request for IMF assistance during the euro crisis. To capture variation more precisely, we also consider the number of episodes of the IMF program each country experienced during that period.
Chapel Hill Expert Survey on classification of Eurosceptic parties: https://www.chesdata.eu/ches-europe
The Chapel Hill Expert Surveys (CHES) estimate party positioning on European integration, ideology and policy issues for national parties in a variety of European countries. The first survey was conducted in 1999, with subsequent waves in 2002, 2006, 2010, 2014, 2019. The number of countries increased from 14 Western European countries in 1999 to 24 current or prospective EU members in 2006 to 32 countries today. In this time, the number of national parties grew from 143 to 277. Using these data, we classify each party as Eurosceptic or not and construct a binary variable indicating whether a respondent voted for a Eurosceptic party.
Aging data:
United Nations Population Projections (UN)-Predicted Aging Data : https://ess-search.nsd.no/
The United Nations Population Projections (UN) dataset provides population estimates and projections produced by the UN Population Division, covering a 150-year span from 1950 to 2100. It combines historical data (1950–2020) with projections (2020–2100), using the cohort-component method based on fertility, mortality, and migration trends. For our project, we focus on 37 European countries from 2001 to 2050, integrating the data at the NUTS 0 level (country level) using the European NUTS classification. The dataset includes key demographic indicators—population, fertility, mortality, and net international migration—and is built on national census data, supplemented where necessary with UN estimates for consistency and completeness.
Macroeconomic data:
Barro-Ursúa Macroeconomic Data: barro.scholars.harvard.edu/data_sets
The Barro Dataset offers historical economic growth and consumption data for 41 countries from 1790 to 2009, including 161 observations on factors influencing cross-country GDP growth rates. While data coverage varies across countries and years, we aligned the dataset with the European Union’s NUTS classification at the NUTS 0 (country) level to ensure consistency and enable integration with other project datasets.
World Development Indicators: databank.worldbank.org/source/world-development-indicators/preview/on
The World Development Indicators (WDI) is a comprehensive database of internationally comparable statistics on global development, covering key aspects such as poverty, population, health, education, environment, and economic performance. It includes over 1,400 time series indicators for up to 266 countries from 1960 to 2021. For our project, we integrated WDI data at the NUTS 0 (country) level using the European NUTS classification system. While the dataset provides extensive historical coverage, country-level data availability varies across years. The indicators are grouped into broad themes, including poverty and income, demographics, environment, economic growth, trade, and migration, offering a rich source for cross-country and temporal analysis.
Globalization data:
KOF Globalization Index: https://kof.ethz.ch/en/forecasts-and-indicators/indicators/kof-globalisation-index.html
The KOF Index of Globalization measures the extent of globalization in countries worldwide across three key dimensions: economic, social, and political. These dimensions encompass factors such as trade flows, information exchange, personal contact, and cultural proximity. For our analysis, we compiled country-level data from 1970 to 2018 into a unified dataset. Since not all countries have complete data for every year, we addressed missing values through linear interpolation and substitution using the nearest available observations. To ensure compatibility with other datasets in the project, the data was structured using the European NUTS classification at the NUTS 0 (country) level.
Corruption data:
V-DEM Corruption Data: https://v-dem.net/data/the-v-dem-dataset/
V-Dem offers a detailed and multidimensional dataset capturing the complexity of democracy beyond just elections, measuring five key principles: electoral, liberal, participatory, deliberative, and egalitarian democracy. Covering countries worldwide, it includes 531 indicators, 245 indices, and 60 additional measures from other sources. For integration with other data in our project, V-Dem data was merged using the European NUTS classification system at the NUTS 0 (country) level. Its variables span democracy indices, electoral processes, political institutions, civil liberties, media, political equality, and background factors such as education, economy, and conflict, providing a comprehensive tool for analyzing democratic systems.