Facebook is sharing data to figure out how it messes with democracy

Facebook is handing over data to academic researchers to help them figure out how social media impacts elections and democracy. Is it a bit of a PR stunt? Probably. But it could also provide some important insights into how the internet shapes the way we interact with the world around us.

Facebook is handing over data to academic researchers to help them figure out how social media impacts elections and democracy. Is it a bit of a PR stunt? Probably. But it could also provide some important insights into how the internet shapes the way we interact with the world around us.

This week, Facebook announced that more than 60 researchers from 11 countries and 30 academic institutions had gotten the go-ahead to access some of its user data for studies on how the social network influences politics around the globe. It’s part of an initiative in tandem with the Social Science Research Council (SSRC), a research nonprofit, and Social Science One, a data project to help bridge the gap between academia and the private sector. Facebook announced the project in April 2018, and Social Science One put out a request for proposals that July that lasted through January 2019.

These sorts of projects can take time — not only to choose the researchers but also to figure out what sort of data to hand over and how to do it. And it’s still not clear what the end goal is in all of this once the findings are out.

The initial crop of selected researchers plans to delve into a wide range of arenas, including the effects of peer-shared fake news and polarized news consumption, differences between how mainstream and misleading news are engaged with online, and disinformation and hyperpartisan social media campaigns in places such as Germany, Chile, and Italy.

Initially, researchers will get access to the API for CrowdTangle, a tool that tracks how popular news items and public posts are on social media platforms. Facebook bought CrowdTangle in 2016. Through that, they will get public data from Facebook and Instagram and posts from public pages, public groups, and verified profiles.

Later, Facebook will also give researchers access to its library of political ads in various countries, including the UK, the US, Brazil, India, Ukraine, Israel, and the EU. It’s a way for Facebook to try to demonstrate that it’s taking political advertising on its platform seriously and is making an effort to find out where ads are coming from and who is paying for them. It had to work out kinks in the system on that front — last year, for example, Vice was able to place ads that said they were paid for by Vice President Mike Pence and ISIS.

Eventually, researchers will also be able to use a new tool to access URLs that have been shared on Facebook.

“This has been a lot of work,” Chaya Nayak, strategic initiatives manager at Facebook, told me. “It took us longer than we hoped, because we really wanted to invest in privacy to do this the right way.”

Facebook didn’t get a hand in picking the researchers or projects selected — SSRC and Social Science One did that. It won’t get a say in future projects, either. (This is just the first cohort.) The Menlo Park, California-based company will get to review the results of the research for legal compliance before they become public, but it says it won’t be able to prevent publication.

Facebook said in a blog post that it hopes the initiative “will deepen public understanding of the role social media has on elections and democracy and help Facebook and other companies improve their products and practices.” It’s not clear, once the research is out, what sort of action will come of it — it will be up to Facebook, other social media companies, or policymakers to act, and none of those parties has proven particularly apt at making significant improvements.

This is an endeavor in privacy … and PR

One major issue in all of this is privacy, because Facebook is opening up its data to 60 researchers and has to make sure the data isn’t misused. The company says this time it’s safe, but it has failed in the past. The Cambridge Analytica scandal was the result of a Cambridge University researcher selling data to Cambridge Analytica that he gathered from millions of people on Facebook. The scandal has led to an enormous amount of fallout for Facebook, including congressional hearings, a potential blockbuster fine from the Federal Trade Commission, and increased discussion of regulation of Facebook and other social media companies. Users have also become increasingly skeptical of the platform.

The leads behind this project are well aware of that inglorious history.

“Part of what I think is important to understand is that while it’s seen as a Facebook scandal, it’s also an academic scandal,” said Nate Persily, co-chair of Social Science One and a professor at Stanford University. He said some platforms have “really clammed up” in the past two years in light of Cambridge Analytica, making it harder for academics to analyze information. “The problem is that everybody understands that we need to protect privacy, and we [also] need to answer some of these great societal questions,” he said.

The privacy question is perhaps most important when it comes to the URLs. Facebook will only include in the data set it eventually provides to researchers URLs that have been shared by at least 100 users, the idea being they wouldn’t be identifiable. It will include with the URLs a summary of the content, engagement statistics, and information from Facebook’s third-party fact checkers if they have rated the post. Facebook will start training researchers on the URL tool in June to make sure they’re using it correctly.

There is no timeline on the studies, meaning the results of this endeavor could be months, or even years, away. SSRC president Alondra Nelson said there is “real awareness on the part of the scholarly community that some of this research is on some real and serious issues,” and that they want to take their time to get it right but also “there’s a real sense of urgency.”

As The Verge notes, the project has already hit some road bumps: In July 2018, Facebook suspended the analytics firm Crimson Hexagon over concerns it was harvesting data from the platform. That company was co-founded by Social Science One co-chair Gary King. He told the Wall Street Journal at the time that he was “never involved in the firm’s day-to-day operations.”

Facebook, and everyone behind the initiative, insists that this time, it will get things right on privacy.

This is yet another step in Facebook’s efforts to improve its image and practices after a cascade of scandals and to reassure users that it is taking these issues seriously. Much of this is a public relations exercise — Facebook probably already knows a lot about the misinformation being spread on its platform, including when it comes to elections and politics. And the platform has been stuck in a cycle of making mistakes, apologizing, and then saying it’s getting better for years. At Facebook’s F8 developer conference, CEO Mark Zuckerberg emphasized the importance of privacy and said that this time, his company is really taking things seriously.

“The future is private,” he said at the conference on Tuesday.

Facebook has fallen short again and again on privacy, despite continually promising improvement. In giving out data to researchers, it is betting it can deliver on making sure the information is used right and trying to show it’s getting better. Of course, getting information to academics is only part of the equation — we’ll also have to see, if and when the findings come out, what Facebook does with them.


Recode and innerdaily have joined forces to uncover and explain how our digital world is changing — and changing us. Subscribe to Recode podcasts to hear Kara Swisher and Peter Kafka lead the tough conversations the technology industry needs today.

Leave a Reply

Your email address will not be published. Required fields are marked *