Study
|
Politicians, parties, polls: Online Disinformation and Information Manipulation Targeting Elections in Germany, Spain and Slovakia
A study commissioned by the Greens/EFA Rights and Democracy Cluster
A study authored by Christian Schwieter & Richard Kuchta
Executive Summary
Disinformation has emerged as a significant threat during democratic processes and in particular elections. This threat is driven in part by the proliferation of social media platforms that facilitate the rapid dissemination of deliberately fabricated and misleading information for political or economic gains. Consequently, the negative implications of disinformation for electoral processes are identified as a major systemic risk by European policymakers in the Digital Services Act, the new European legislation governing online services.
This study examines instances of disinformation during the 2021 German federal election, the 2023 Spanish general election, and the 2023 Slovak parliamentary election using novel fact-checking datasets comprising 150 different election-related pieces of dis- or misinformation observed across 410 posts on social media. It employs the ABCDE approach to analysing Foreign Influence and Information Manipulation (FIMI) first introduced by Camille François and later expanded on by Alexandre Alaphilippe and James Pamment, providing insights into the actors, behaviours, content, distribution, and effects observed in the context of disinformation campaigns targeting European elections. The key findings of this study are as follows:
-
A variety of actors engaged in disinformation activities, and only a small portion of those can be confidently linked to foreign influence campaigns, based on the available data. While the specific actors differed across Germany, Spain and Slovakia, election-related disinformation was frequently spread by smaller private social media accounts as well as larger influential accounts that were frequently linked to conspiracy communities or well-known far-right groups, many of which consistently shared pro-Russian disinformation. Other common actors included self-described journalists or “alternative” media, some of which could easily be mistaken for credible news sources, often featuring a blue checkmark next to their username on X. Lastly, in all countries, politicians affiliated with predominantly right-wing political parties were found to share election-related disinformation.
- The behaviour of these disinformation actors was comparable across Germany, Spain and Slovakia, with the most common disinformation strategy observed being the decontextualisation of real photos or videos to make unfounded claims, or the omission of important contextual information to mislead audiences. Another common behaviour with more technical sophistication was the fabrication or manipulation of media to create so-called imposter content.These media frequently purported to show screenshots of alleged articles published by trusted newspapers, fake social media posts by political opponents, or manipulated campaign posters. While in Germany and Spain this behaviour mainly relied on standard photo manipulation techniques, Slovakia also saw the deployment of AI technology for disinformation purposes by imitating the voices of politicians and journalists.
- The specific content of election-related disinformation similarly converged around common themes, with the majority of disinformation targeting either individual Green or left-wing politicians, liberal and progressive parties, or the electoral process more generally. In Germany, disinformation most commonly targeted female Green politicians and the Green party in general, but disinformation targeting the integrity of the election integrity was also pronounced. In Spain, the majority of disinformation sought to undermine the legitimacy of the election itself, with actors alleging a “pucherazo” based mainly on falsehoods shared on social media about the postal and expat vote. Similar claims were popular in Slovakia, with additional evidence of minorities such as immigrants and the LGBTQ+ community being a major target in the electoral campaign.
-
In terms of distribution, Facebook was found to still be the most important platform for disinformation actors in Germany and Slovakia, whereas X was most commonly used in Spain, based on the fact-check datasets. Telegram was another key platform used for the dissemination of election-related disinformation across all countries, with content targeting the validity of the election viewed over one million times in Germany alone. Substantial evidence of election-related disinformation on Instagram, TikTok and YouTube was only found in the Spanish dataset. Notably, disinformation on Facebook was mostly accompanied by a fact-check warning label, which was usually not the case for posts found on X, YouTube or Tik-Tok. On the contrary, on X, Spanish disinformation actors frequently featured a blue checkmark in a likely attempt to increase their credibility and reach.
-
The common targets, actors and behaviours observed, combined with secondary survey data, suggest that the large number of falsehoods circulating on social media alleging “Wahlbetrug”, “pucherazo” and “zmanipulované voľby” are part of larger cross-national campaigns that seeks to undermine the legitimacy of democratic institutions, and coincides with declining trust in established media observed across all cases. Besides declining faith in democratic institutions, the finding that progressive female politicians and LGBTQ+ communities in particular are targeted bydisinformation also threatens another key requisite of functioning democracies, namely political participation. Those members of society disproportionately subjected to personal attacks online may be less willing to speak up, let alone run for public office.
The data available for this study does not allow for confident conclusions on the actual effect of disinformation spread in the context of elections. However, while the findings of this report are based on only a subset of disinformation content circulating online, they can nevertheless lay the foundation for both country-specific risk assessments as well as EU-wide measures to combat election-related disinformation. The commonalities and divergences between disinformation campaigns targeting Germany, Spain and Slovakia identified in this report highlight what type of actors, behaviours, targets and platforms to look out for as the EU gears up for the European Parliament elections in June 2024.
