Academic Researchers Need Access to the Facebook Papers

November 4, 2021  ·   Commentary

With access to these documents, scholars could support the media, public, and policymakers in identifying where Facebook’s internal research is conclusive, what inferences can be drawn, which topics require more evidence and future research, and what that research should be.

A hand holding a smart phone with a Facebook logo on the screen.

Credit: Pexels

This article was originally published at Slate.

The Facebook Papers offer a chilling account of how the company’s internal research revealed potential harms caused by its platforms. Reports underscore Facebook’s role in the diffusion of misinformation, the prevalence of hate speech, and the coordination of the Capitol insurrection.

And that’s just in the U.S.  The problems documented across the globe are even worse.

While the company conducted this research over many years—the findings of which often contradicted public claims by executives—it took a courageous whistleblower and incisive reporting to bring the documents into public view. A key lesson of the Facebook Papers is the tremendous social cost of siloed information: Facebook had a monopoly on relevant data to study the social, political, and psychological effects of its platforms and kept the internal research hidden, even as we and others argued for the importance of sharing it.

This lesson makes the current situation so alarming: Although recent reporting has publicly revealed important new information about Facebook’s global impact, the Facebook Papers themselves remain inaccessible to all but 17 news outlets and government officials. Most notably, scholars who study the effects of social media on society still can’t see the documents.

Not only can we not see the Facebook Papers, we aren’t even sure what they are. Based on  the few examples released so far, most appear to be a combination of slide-deck summaries and employee discussions on Facebook’s Workplace message board, which is basically a staff-only version of Facebook. Are there any more detailed analyses, complete with research designs and statistical appendices, of these experiments—or is that all?

Scholars are trained in—and the academic process is structured around—being able to make inferences from data. This is true of both the quantitative (e.g. exposure or engagement metrics) and qualitative (e.g. interviews with social media users) kinds of data included in the Facebook Papers. With access to these documents, scholars could support the media, public, and policymakers in identifying where Facebook’s internal research is conclusive, what inferences can be drawn, which topics require more evidence and future research, and what that research should be.

For example, many reports have examined Facebook’s efforts to fight misinformation. Ahead of the 2020 election, the company’s civic integrity team implemented several measures to curb misinformation and hate speech on the platform. Afterward, employees felt they had largely succeeded. “It was like we could take a victory lap,” one former employee said. “There was a lot of the feeling of high-fiving in the office.” But then, reports indicate, Facebook took its foot off the gas, rolling back key measures that had reduced harmful content and largely disbanding the team responsible for them. In the following months, as momentum picked up around the Stop the Steal campaign, Facebook struggled to address it, and employees blamed the platform for helping foment the Jan. 6 riot.

So what direct role did Facebook play leading up to Jan. 6? Based on these reports, it’s hard to know for sure. While Facebook’s central role warrants extensive scrutiny, we shouldn’t forget that cable television maintains large audiences, and networks such as Fox News, OAN, and Newsmax spread false claims of voter fraud. Indeed, recent research from Harvard’s Berkman Klein Center on the development of the fraud narrative throughout the election emphasized that the disinformation campaign was an “elite-driven, mass-media led process,” with social media playing a secondary role.

This isn’t a defense of Facebook’s inadequate response to the rise of Stop the Steal groups. But people’s media diets are diverse, and political movements (even anti-democratic ones) are complex. A comprehensive analysis of the insurrection—much like studies of the influences on teenage girls’ well-being—should take account of the myriad forces that coalesce and contribute to what we observe. Having access to the Facebook Papers—and any other internal research underlying the documents—would help scholars study the extent to which social media, traditional media, and political leaders each played a part in the violence.

To be sure, Facebook has partnered with independent researchers on some data-sharing research initiatives, including the Social Science One and 2020 Election projects (which one of the authors co-leads). These are important and necessary. But overall, since 2016, academic access to platform data has been largely eliminated, or at least greatly restricted, by the company, leaving us with far too little academic research analyzing Facebook’s impact in the U.S. and abroad.

Now, as the company’s effects on society are deservedly interrogated, scholars are once again shut out.

Imagine if a pharmaceutical company ran experiments to test the efficacy of a new treatment, and internal documents were released by a whistleblower. Rather than experts from medicine, biology, and chemistry participating in analysis of the research, only journalists and government officials were given the documents. The question is not which groups should have access, but why can’t all three?

So what can we do about it?

First, the Facebook Papers should be shared widely with academic researchers as quickly as possible. Universities have well-established protocols and institutional review boards to ensure data security and ethical compliance, such as protecting the privacy of study participants (although, again, without access we don’t know if there are privacy concerns). Either way, these frameworks can be leveraged to enable academic researchers’ access.

Second, journalists and scholars should work together more closely to assess what we can learn from the Facebook Papers. Journalists have done an incredible job combing through the documents in a short period of time and drawing out potentially important findings from Facebook’s research. As the story continues to unfold, media and scholars working together can go deeper to further interpret and contextualize the research, especially when analyzing the technical work included in the documents.

Finally, and most importantly, scholarly access to data needs to be made more robust. The current model of relying on leaked corporate research to inform what we know about social media platforms’ impact on society is unsustainable. Not only is there evidence that Facebook is now limiting access to internal message boards in an effort to protect against future whistleblowers, but external researchers should also be empowered to design their own studies. There are current proposals to create such data sharing systems. In recent weeks, Congress’ numerous hearings with scholars have demonstrated the value it places on independent research. It would do well to consider policies to enable even more studies in this domain.

A precondition for effective public policy and informed public discourse is a deep and nuanced understanding of the issues. Journalists, government officials, and academic researchers each play integral roles, but currently only two groups have access to the largest and richest leak of platform research. More than seven weeks after the first published story using the Facebook Papers, it’s time to change that.

Future Tense is a partnership of SlateNew America, and Arizona State University that examines emerging technologies, public policy, and society.