Researchers Consider the Impact of Meta's CrowdTangle Shutdown

Meta's decision to discontinue CrowdTangle as of August 14 has sparked widespread concern among researchers, journalists, and policymakers. The company's proposed replacement, the Content Library, is still in development and lacks many of CrowdTangle's valuable features. According to researchers and advocates, the transition has been poorly communicated, leaving users uncertain about their future access to Meta's data.

The timing of the shutdown is particularly problematic given that we are in a year of global elections. Despite widespread criticism from various stakeholders, including open letters from the likes of Mozilla Foundation and the Center for American Progress, inquiries from the European Commission, and a bipartisan appeal from US lawmakers, Meta has not altered its decision.

To understand how Meta’s decision would affect journalists and researchers, the Coalition for Independent Researchers (CITR), of which Tech Policy Press is a member, surveyed key stakeholders and conducted unstructured interviews to gain deeper insights into the impact of Meta’s decision to shut down CrowdTangle. The survey was conducted in May 2024 and “garnered 36 responses from 34 different organizations, including major news outlets, academic institutions, unaffiliated researchers, and civil society organizations.”

The survey sought to gauge how CrowdTangle’s shutdown will impact the work of academics, journalists, and civil society researchers. Respondents were also asked how they felt about the Content Library and whether Meta took their concerns into account when transitioning from CrowdTangle to the Content Library.

Key Findings:

1. Widespread Impact: 88% (32 out of 36) of respondents expressed concern that CrowdTangle's shutdown will hinder their work. Many projects will need to be revised or scrapped entirely, with some researchers considering stopping their work altogether. One academic institution told the researchers, “We know we don’t have real exploitable access to relevant Facebook data anymore. This means we will probably stop working with Facebook data and reorient our research towards other data sources.” CrowdTangle’s demise will require changes to methodologies and study design for studies focused on Meta platforms that do persist, leading to further delays and labor costs.  
2. Insufficient Replacement: According to the respondents, Meta's proposed substitute, the Content Library, falls short in several crucial areas:
   • The Content Library does not allow researchers to track individual accounts of influential public figures, which is especially problematic in cases where researchers want to monitor public figures using Meta’s platform to engage in hate speech or incite violence.
   • The Content Library restricts the amount of data that can be downloaded. According to a civil society organization interviewed as part of this study, the library “also has limitations on data downloads/use that prevent analysts from using other tools to interrogate the data. These limitations will both slow down the speed we're able to research emerging trends, while also driving up the costs for doing so.” Along with these limitations, the Content Library delays access to real-time data and has a complex user interface which makes it harder to use compared to CrowdTangle.
3. Limited Access: Only a small fraction of CrowdTangle's tens of thousands of users have been granted access to the Content Library. For instance, 11 respondents indicated that they had applied to access the Content Library, but only four of them were granted access. Notably, commercial media outlets are banned from access which makes it harder for newsrooms to do data-driven stories on issues at the intersection of technology and democracy.

Conclusion

The timing of this shutdown is particularly concerning. In a year when half the world's population heads to the polls, Meta’s decision to revoke access to a valuable data resource is troubling. Monitoring and addressing hate speech and disinformation is already complicated, and discontinuing CrowdTangle makes this job much harder. While Meta has made some improvements to the Content Library in response to feedback, the research community maintains that it's not enough. The researchers made some important recommendations that Meta should immediately implement:

• Keep CrowdTangle running at least through the end of 2024.
• Improve the Content Library's functionalities.
• Streamline the application process.
• Allow all current CrowdTangle users, including newsrooms, access to the new platform.

While these recommendations might help improve data access for research and journalists, there has to be a concerted effort across civil society, government, and technology companies to address digital threats to democracy. Meta’s unilateral decision to discontinue CrowdTangle in an election year without adequate public consultation sets a dangerous precedent. As the researchers note, “Meta has been celebrated as a leader for data access and transparency compared to its peers; now is the moment for the company to live up to its principles and protect the public’s right to know.”