Citizens in Five Democracies Uneasy About Parties Using Their Data

Voters in Australia, the United States, the United Kingdom, Canada and Germany share a deep unease about political parties harvesting their personal data for election campaigns, but they disagree sharply about whether governments should step in and regulate the practice. That is the central finding of one of the first cross-national surveys to examine citizen attitudes toward data-driven campaigning, or DDC, and the regulatory responses voters would like to see introduced. The study, based on 5,040 respondents across the five advanced democracies, reveals a striking paradox at the heart of public opinion: German respondents are the most opposed to parties using their data, yet also among the most opposed to regulating political parties.

The research team, led by Glenn Kefford, Andrew Barclay of the University of Sheffield and Simon Kruschinski of GESS in Cologne, fielded an online survey distributed by the panel provider Lucid between 4 and 7 September 2021. Each country sample contained roughly 1,000 respondents, with 1,003 in Australia, 1,002 in Canada, 1,012 in Germany, 1,023 in the United Kingdom and 1,000 in the United States. The data were stratified by age, gender and education level to mirror each country’s general population, and the survey achieved a response rate of 70 percent. Respondents answered items on five-point Likert scales measuring their perceptions of what parties do with data, the acceptability of specific data practices, and the importance they attached to five different regulatory measures.

The five countries were deliberately chosen because they span the full spectrum of regulatory environments for campaign data. The United States and Australia sit at the weakly regulated end: Australian parties are almost entirely exempt from the country’s privacy laws, while American campaigns operate under First Amendment protections and data held by third parties falls outside privacy law. Canadian parties are similarly unconstrained in practice. The United Kingdom has adopted the General Data Protection Regulation almost in its entirety but exempts political parties from restrictions on targeting with personal data. Germany goes furthest of all, applying the GDPR framework together with significant national restrictions on parties without exemptions.

The survey’s descriptive results paint a picture of widespread public skepticism. Respondents across all five countries considered it relatively likely that parties use their data to understand, target and even manipulate voters, with the perceived likelihood of manipulation scoring particularly high, something the authors attribute to years of media coverage following the Cambridge Analytica scandal. When asked which specific types of voter information parties may legitimately use, basic demographics such as age and gender were rated most acceptable, followed by political views, relationship status and ethnicity. Religious views, household income, sexual orientation and, above all, online browsing history were judged the most unacceptable. German respondents recorded the lowest acceptability scores for nearly every category.

Attitudes toward where data comes from revealed a similar pattern. Citizens found data provided directly by voters themselves and data drawn from public authorities to be the most acceptable sources, while data obtained from social media companies or purchased from commercial vendors was widely rejected. The rejection of social media data echoes the earlier finding on browsing history and, the authors argue, reflects genuine citizen discomfort with how organizations track people’s movements online. Country-level differences were limited in this domain, with one notable exception: data obtained from public authorities was seen as significantly more unacceptable in Germany than elsewhere.

On regulation itself, the picture is broadly favorable but nuanced. Stronger regulation was the modal response for every measure tested, yet substantial minorities believed political campaigning should remain unregulated on free-speech grounds. The most popular options in every country were transparency requirements, ensuring parties are open about how they campaign, and mandatory fact-checking of campaign messages to prevent misleading claims. Less popular, though still supported by substantial numbers, were financial spending limits on political advertising and outright bans on parties buying voters’ data. The authors note that measures requiring parties to fundamentally change how they campaign attract less support than less interventionist transparency measures. Australian and Canadian respondents proved the most regulation-friendly, while American and, especially, German respondents were the least inclined to advocate regulation.

The multivariate analysis, a series of ordinary least squares regressions predicting support for each regulatory measure, confirmed and extended these descriptive patterns. Women were significantly more supportive of regulation than men across every measure, with effect sizes broadly similar across models and slightly larger for banning data purchases. Older respondents consistently favored regulation more than younger ones. Ideological self-placement, measured on a left-right scale, mattered in the expected direction: right-leaning citizens were consistently more opposed to regulation, on both free-speech grounds and specific regulatory proposals. The authors caution that their outcome variables are strictly ordinal, but cite prior methodological work supporting the use of parametric statistics with Likert scales of this kind.

Country of residence produced the study’s most intriguing results. Compared with Australia, the reference category, respondents in every other country were more opposed to regulating campaigns on free-speech grounds, with the effect particularly pronounced in the United States and Germany. American and German respondents were also consistently more opposed to each specific regulatory measure. When the researchers interacted ideology with country, they found the relationship between ideology and support for regulation to be broadly linear everywhere, but considerably weaker in Germany and the United States than in the other three countries. The divergence was driven almost entirely by left-leaning voters: there were few differences among the most right-wing respondents across countries, but large and highly significant differences among those furthest left.

Interpreting the German and American skepticism, the authors point to cultural and historical context. Germany’s robust data protection tradition emphasizes personal integrity and informational self-determination, yet simultaneously breeds skepticism about state intervention in speech and political activity, a wariness the authors link, speculatively, to the country’s history of autocratic rule. In the United States, privacy is often framed as an individual right in tension with free-expression values, within a laissez-faire regulatory culture. The authors themselves stress that any explanation of the German pattern must remain speculative without further research, since their data cannot definitively identify the mechanism behind it.

The study carries important limitations that the authors acknowledge openly. Their models omitted education, an increasingly significant divide in Western democracies’ political attitudes, as well as political trust, both of which might shape regulatory preferences. Respondents who strongly opposed all regulation in principle had attitudes toward specific measures imputed, an imperfect solution. Because the survey used an online panel, responses may understate support for regulation relative to the broader population, given panel participants’ likely higher digital literacy. The September 2021 fieldwork also captured a specific moment, when Cambridge Analytica-era concerns dominated but artificial intelligence had not yet reshaped campaign technology, and the authors caution that results from a survey conducted today might differ. They frame the study as a starting point rather than a conclusion, arguing that regulators should tailor rules to national contexts and that parties using these techniques owe citizens a clearer explanation of why and how they use their data.

Subject of Research: A cross-national survey of public attitudes toward data-driven election campaigning and its regulation in five democracies.

Article Title: Mapping public attitudes toward the regulation of data-driven campaigning across five democracies

Article References: Kefford, G., Barclay, A., & Kruschinski, S. (2026). Mapping public attitudes toward the regulation of data-driven campaigning across five democracies. Frontiers in Political Science, 8, Article 1919560. https://doi.org/10.3389/fpos.2026.1919560

Image Credits: AI Generated

DOI: 10.3389/fpos.2026.1919560

Keywords: data-driven campaigning, political microtargeting, election regulation, public opinion, survey research, data privacy, political parties, Germany, United States, GDPR, voter surveillance, cross-national comparison