Online platforms should become more transparent. That is one of the goals of the Digital Services Act (DSA), the most ambitious piece of EU platform regulation legislation to date. It will be fully applicable from February 2024 and is eagerly awaited by researchers. This is because the DSA provides for the mandatory introduction of so-called "research data access" as a central measure for providers of very large online platforms (45 million active users per month). This is a potential milestone for research on central phenomena in the context of digital media - for example, disinformation, social polarisation, or digital right-wing extremism.
"The legislator assumes that large online platforms pose certain risks to democracy, which he wants to minimise with stronger regulation," says legal expert Vincent Hofmann. "Risks" mean, for example, the dissemination of illegal content, digital violence, violence against minorities or the endangerment of children and young people. "This is also where research data access comes in. Everything that researchers want to find out with platform data must relate to these risks in order to better understand and to better minimise them in the long run."
So, research with platform data must contribute to a specific, democratic goal. Other research projects, such as market research with data from online platforms, will therefore not be possible.
Researchers are eagerly awaiting the new regulation. "Especially in my field, research of digital right-wing extremism, the DSA could change a lot," Jan Rau tells us. "Until now, it has sometimes been frustrating as a researcher to know that you could answer incredibly interesting research questions if you could just get hold of the data. The data was always there, you just couldn't get it because the platforms didn't cooperate."
For research data access, as prescribed by the DSA, not to degenerate into a paper tiger, but to be of optimal use to research, a few things need to be taken into account. "It starts with getting an overview of what data exists in the first place," says Jan Rau. "It would be helpful to have a public repository that lists which data can be successfully queried and which cannot. In addition, the volume of data that can be queried should not be limited or should not be severely limited. As far as the specific data to be queried is concerned, it would be helpful if we researchers could gain insight into, for example, the reach of certain postings and the moderation decisions of the platform. It would also be important to know whether and how the platform pushes certain content and on what basis."
Vincent Hofmann
Jan Rau
Johanna Sebauer
Leibniz Institute for Media Research | Hans-Bredow-Institut
Hamburg, 25 May 2023