Though it is widely believed that the spreading of false information changes people’s perceptions, it has so far been very difficult to prove the role of disinformation in influencing elections. Civitates has funded the development of a toolkit by Democracy Reporting International and MEMO98 in order to assess how online public discourse is impacted by disinformation, hate speech and other phenomena.

Rafael Goldzweig, research coordinator at Democracy Reporting International, monitors the impact of social media on elections around the world.

How big is the problem of elections being influenced by false information?
Despite many discussions on the effects of political advertisements, disinformation and hate speech in the electoral context, it is hard to quantify the impact of online messages in the offline world. We do however know that many people believe that the false information spread during elections is true, and that its goal is not only to misinform but also to suppress voting among specific groups.

Can you tell us more about the toolkit you’re developing?
The toolkit is a result of a methodology we’ve developed to understand how social media affects elections. Genuine political discourse is impossible if the information at its basis is manipulated. Together with partners in Portugal, Austria, Poland, Romania and Croatia we will incorporate the lessons learned from elections in these countries into a user-friendly online toolkit. This resource can eventually be scaled up with results from other elections.

Any results yet?
Our partner in Portugal discovered that the biggest online narrative related to disinformation was about corruption. So we organised an event at the EU Representation in Lisbon to discuss challenges and solutions: from individual actions by social media companies to regulation at the national and EU levels. Although the regulation of transparency is a sensitive topic, we need a mandate to carry out in-depth research about who is behind certain accounts and to ultimately remove actors that threaten to damage our democratic values.

What is the biggest challenge within the project?
Online manipulation is constantly evolving. In 2016 for example, Facebook decided that political ad campaigns could only be run by citizens based in the country of the campaign. During the presidential elections in Ukraine this year, Russian actors rented Ukrainian accounts in order to circumvent this regulation. People will always find creative ways to overcome actions taken to counter disinformation. It’s important to remember that as the problem evolves, so should the solution.

How did you get involved in this work?
I am Brazilian by origin, and started my career as a political consultant. When I covered the Brazilian elections in 2014, I saw the increasing impact of social media as a tool: not only for political campaigning, but also as a space for people to discuss their ideas. During my master’s degree I looked deeper into the intersection of technology and politics, and my thesis compared the use of Twitter by both extremist and establishment candidates in four different elections around the world. One of my findings was that extremists benefit more from Twitter due to the language they use. After graduating, I became a Google fellow and worked on understanding the impact of social media during elections. Joining this project was the next ­­­­logical step.

What is your dream with regard to the transparency of social media?
My dream is to generate more accountability. Real change will only come if social media providers change their business models; they are currently trying to lessen the problem without attacking it at the roots.