26/4/2024
Starting in August of this year, CrowdTangle, a Meta tool that for years has served civil society, academia and journalists to investigate the dissemination of disinformation narratives and problematic discourses on Facebook and Instagram, will cease to operate. The news, announced by the company, draws attention to the resources available to analyze the digital public debate and the position of the research community regarding access to data in the middle of an election year.
CrowdTangle was acquired by Meta in 2016, when it had proven its effectiveness in measuring the virality of content on the company's two main social networks. The announcement of its closure was the confirmation of a suspicion that had been alerted for some time, as Meta had given enough signals to imply that retaining this product was not one of its priorities. In mid-2021, the CrowdTangle team was disbanded, and the following year they stopped accepting new users.
Meta claims that the closure of CrowdTangle can be offset by the opening of its Content Library, a tool to provide near real-time access to public content from pages, posts, groups and events on Facebook, as well as content creator and business accounts on Instagram. According to the company, this data can be accessed through a graphical interface or API.
Meta developed this resource as part of its obligations under the European Union's Digital Services Act, which provides that large Internet companies must offer access to data to independent researchers to detect and understand systemic risks in that territory and the effectiveness of measures taken by the platforms to mitigate them.
Although the rules of this new law apply in principle only to the European Union, some platforms have provided this information on a global scale, which is a step forward in terms of access to data that was not previously available. (You can learn more about these transparency initiatives and their geographic scope at this link).
However, in the case of Meta, there is every reason to believe that, at least for now, this may be a setback, as the Meta Content Library is not yet a true replacement for CrowdTangle.
First, the company sets stricter selection criteria: To qualify for the Library, researchers must be associated with an academic institution or non-profit organization whose primary activity is scientific research or the public interest. Also eligible are members of Meta's fact-checking program, a network of media and fact-checking organizations that detect and flag fake content on their platforms.
These criteria leave out, for example, journalists who are not part of this program, and make access difficult for independent researchers. In addition, the application process is lengthy, as it requires support to demonstrate the suitability of the organizations that aspire to access and the evaluation of documents and requirements is done manually.
Beyond these criteria, the Library's functionalities also do not fully supplement the data that could be obtained through CrowdTangle. Although the Library offers much more data, changes are needed to make the best use of this information.
As Brandon Silverman, co-founder of CrowdTangle, cautions, the transparency of the platforms actually has little to do with the amount of data offered, but rather the usefulness of that data and the features available to analyze it and draw conclusions. For example, the Library does not yet allow you to aggregate data to evaluate it at the account or topic level, do phrase searches and not just keyword searches, or easily export data via email notifications or search engine extensions.
In addition, the new arrangement of data will require greater technical skills to analyze it. To exploit the Library's information, researchers need to know how to use the API and program, which means that civil society organizations that want to use this information will have to compete for human talent that is highly valued in the labor market.
It is likely that in the future Meta will tweak and add more functionality to the Library, however, the fact that CrowdTangle is shutting down in the biggest election year in history has been a cause for concern. For Silverman, there are still no guarantees for election observation groups to monitor elections, for fact-checkers to respond to disinformation narratives, or for activists to study human rights violations online.
"What we need right now are effective tools to monitor candidates and political narratives, and protect people before, during and immediately after elections, at times when false and confusing narratives spread like wildfires and when misinformation and hate speech can lead to physical harm," the Mozilla Foundation recently warned in a letter signed by researchers, journalists and academics around the world regarding Meta's announcement.
The closing of CrowdTangle exposes the difficulties of conducting digital research and especially the power of technology companies' decisions to increase or limit access to relevant data. These changes force civil society organizations to innovate in their research methodologies, but also underscore the need for independent mechanisms to access information to measure the pulse of the digital public debate.