Researchers Fight for Data Access under the DSA

After years of debate, the European Union has found a way to comprehensively regulate very large online platforms such as Instagram, Twitter, Google or even Wikipedia. The Digital Services Act will force large tech companies to a wide range of commitments and measures, from removing illegal content to specific user rights and restrictions on targeted advertising. One important aspect of the DSA is that those designated Very Large Online Platforms must grant auditors, authorities and researchers access to their internal data.

The European Commission has recently started a Call for Evidence on how this access to platform data should be granted, what kind of data should be accessed, who can access it and how the process should be managed. Jakob Ohme, who researches digital news dynamics at the Weizenbaum Institute, and Ulrike Klinger, one of our Associate Researchers and Professor for Digital Democracy at the European New School of Digital Studies, have answered that call. You can read their full response here.

Jakob Ohme, why is the DSA so important for digital researchers?

Digital platforms affect the lives of many people in Germany, the EU and around the world. The job of researchers is to study these effects, but so far, this has been a tedious task because researchers had little access to the actual platform data. And without that access it is very difficult to know how people use these platforms, what information they see there, and how this affects their personal life and society in general. It’s as if medical researchers want to study the human body, but never have access to one. The DSA will change this. It will force platforms to provide certain data and implement procedures through which researchers will get access to it, creating a better data basis for platform research. This EU law is one of the first, and so far, the most far-reaching to guarantee researchers the right to access platform data.

The law was passed last October, what is there still left to debate now?

The law, which, by the way, includes many other aspects than research data access, was indeed passed a while ago, which is in itself a big achievement. The research data access is mostly specified in Article 40 of the DSA and while it describes a clear direction, it leaves many questions unanswered, especially on the practical implementation: how researchers get to the data and who supervises this process. For these questions, the EU Commission is preparing a so called Delegated Act, which can be understood as a specification of Article 40. For this act, the Commission has sought input from stakeholders all around Europe. In our initiative by the Weizenbaum Institute and European New School, we have responded to the Call for Evidence by the commission, preparing a document that details on how researchers would benefit most from getting access to platform data.

What kinds of data do researchers need access to?

In summary: Data on who is connected to whom on platforms, who the platform actors are (e.g. politicians, NGOs, media organizations, or citizens), and what content is circulated and viewed on platforms. It’s important that users privacy rights are respected at all times, so researchers will not have access to private communications, such as chat protocols, for example. But this is not necessary. Most research questions can be answered with data that is publicly available.

What will you be able to research with this data? 

Here it helps to use the old formula by political scientist Harold Dwight Lasswell, explaining acts of communication: “Who”, “Says What”, “In What Channel”, “To Whom”, and “With What Effect”. All these acts are important to study, especially because DSA Article 40 specifically talks about studying systemic risks and ways to mitigate them. Examples here are to study how algorithmic decisions impact people’s political opinions, who are the actors targeting users with misinformation, or to what extent the use of certain platforms affects the well-being of users. Specifically for my research, I hope to get a better understanding on the content that users receive on platforms, especially news and political information. So far, we know very little about this important part of people's platform media diet.

What challenges remain in getting this data access? 

Many. I personally see the technological pre-conditions that researchers and their institutions need to fulfill to handle platform data as one big caveat. It is also an open question, if and to what extent platforms will cooperate with the Digital Service Coordinators, who are responsible in each country, to make this data access possible. Lastly, the modes of access will keep us busy for a longer time. In some instances, an API access like the one of the academic Twitter API would be a good solution, and we hope to see this materializing fast for all platforms. In other cases, and for more sensitive data, the technical access will remain a challenge.

What’s next? 

After the Commission has gathered the input with a deadline on May 23, 2023, they will prepare the Delegated Act. It is expected that it becomes effective in early 2024. Then we will know more specifically how the process of data access requests for researchers will look like. In the meantime, researchers should establish networks to combine and further prepare strategies to make access requests a success. We at the Weizenbaum Institute together with the European News School of Digital Studies are hosting a workshop in June for the German research community, with the goal to also connect with German policymakers and the newly established Digital Service Coordinator in Germany. Public officials told us they would value the support and input from researchers in this process, this is why we are getting organized and will try to deliver input, to make sure Article 40 of the DSA works for the research community in Europe and beyond.

Thank you for the interview!

Jakob Ohme is Research Group Lead of the “Digital News Dynamics” group. His research centers around the impact of digital and mobile communication on news exposure and political behavior in digital democracies, and in developing digital methods in political communication and journalism research.

He was interviewed by Leonie Dorn.