WhatsApp Groups and Misinformation Are a Threat to Fragile Democracies
(Bloomberg Businessweek) -- For Stela Wanda Pereira da Silva, the breaking point came when her father posted a video of a woman getting assassinated to the family’s private WhatsApp group, calling it an example of the violence that would ensue if leftist Workers’ Party candidate Fernando Haddad prevailed in Brazil’s presidential election.
Da Silva, a 22-year-old resident of the coastal city of Salvador and a Haddad supporter, did some digging and discovered that the woman in the video was the victim of a robbery gone bad and not a politically motivated hit, as her father maintained. When she showed her family that the post was fake news—from Venezuela, yet—a civil war broke out, with half the group’s members defending her and the other half taking her father’s side.
“Our family was totally divided because of this election, so I had to leave the group,” says da Silva, who acknowledges that her relationship with her father has always been turbulent. Her experience on the platform isn’t unique, she says: “I have many friends who would prefer to leave their family WhatsApp group than deal with the unhealthy environment they create.”
Brazilians are among the world’s top users of social media, leaving them especially exposed to fake news and political influence campaigns online. Social media forums have replaced traditional media, which for decades were controlled largely by a single Brazilian conglomerate, Globo Group. Facebook Inc.-owned WhatsApp, in particular, has become the main vehicle for the internecine spats that happen elsewhere on Twitter or Facebook. Brazil is WhatsApp’s second-biggest market, with more than half of its 208 million people counted as users. They cluster in family or affinity groups whose typical fare is quotidian—holiday plans, an upcoming volleyball match, dinner Thursday night. But the groups also serve as virtual propulsion jets for political news, both real and fake.
“Brazil is dealing with a very powerful combination right now,” says Maurício Santoro, a political scientist at Rio de Janeiro State University. “It’s a combination of a lack of confidence in traditional media and easy access to alternative social media outlets.”
This dynamic has played out against the dramatic backdrop of the October presidential election, one of the most critical votes anywhere this year. Jair Bolsonaro, the far-right head of the Social Liberal Party and a former Army captain, won the Oct. 28 runoff with 55.1 percent of the total, besting Haddad, a substitute candidate for the Workers’ Party, whose former leader, Luiz Inácio “Lula” da Silva, is in jail on corruption charges. Bolsonaro’s tough-on-crime message resonated in a nation where 63,880 people were murdered last year, but some of his other rhetoric—including praise of a notorious torturer from the two decades of military rule that ended in 1985—rattled observers worried about the future of Brazil’s democracy.
The fake news flood during the campaign also prompted concerns that this wasn’t a fair fight. A bombshell investigation published a week and a half before the runoff by Folha de São Paulo, one of Brazil’s most respected newspapers, revealed that a group of entrepreneurs had paid influencers to spread anti-Haddad content from their private WhatsApp groups. The report sent Workers’ Party representatives running to the country’s electoral court claiming fraud, arguing that the actions amounted to illegal campaign donations. The court opened an investigation, but no determination has been made.
It’s impossible to quantify how much of a lift Bolsonaro got from fake news, and his supporters say such claims are overstated. However big the bump, the spread of misinformation on social media could pose a long-term threat to democratic norms and institutions. Politics in Latin America’s biggest economy have always been fragmented—no fewer than 13 parties contested the presidency—but it’s difficult to recall a time when they’ve been this polarized.
Brazil has more internet users than any country in Latin America and a long tradition of early social media adoption. Remember Orkut? In the U.S., the Google-owned social network was quickly eclipsed after its 2004 launch by rivals such as Myspace and Facebook. But it took off in Brazil; by late 2007 there were 40 million registered Orkut accounts there. In a book on the country’s democracy, Georgetown University history professor Bryan McCann attributed the surge to what he termed the “Orkut rule,” which held that Brazilians would use digital media for social purposes at every opportunity. They came to dominate Orkut so much that Google turned over operations of the entire network to its office there in 2008.
Since that time, Brazilians’ affinity for digital media has only grown, spurred by an economic boom from about 2008 to 2011. The subsequent crash, though, made iPhones and mobile data harder to afford. WhatsApp, which can run on any platform and whose data usage isn’t always counted against data quotas by carriers, became the de facto messaging system nationwide. As of July, the service had 120 million active users in Brazil, only 7 million shy of Facebook’s tally.
This predilection has given political debate on Brazilian social media a distinctive set of characteristics. Unlike Facebook’s timeline feature, which shares a user’s posts and activity with her contacts and sometimes with the general public, WhatsApp is simply a set of private group conversations. Its chat rooms are further shielded by encryption, a measure intended to increase user security that also has the effect of blocking monitoring efforts by outsiders.
In the runup to the election, WhatsApp’s privacy features turned it into an impenetrable candyland of misinformation. When the 24 Brazilian newsrooms participating in Comprova—a project of the Shorenstein Center on Media, Politics and Public Policy, funded by Google and Facebook—asked users to flag WhatsApp messages that appeared to be fake news, they received more than 60,000 submissions in two months.
“What we have here that is different from the U.S. is an election mediated by WhatsApp, a closed platform, where we can’t measure exactly the volume or the content of the fake news that is being spread, much less who is responsible for the disinformation,” says Tai Nalon, executive director and co-founder of Aos Fatos (“To the Facts”), a Brazilian fact-checking group. Her organization tracks viral fake news items and debunks them in real time, spreading factual accounts of the same events instead.
In one memorable instance, Aos Fatos singled out a candidate’s claim that 400 million Brazilians, almost twice the country’s population, live in extreme poverty. During the weekend of the first round of voting in the presidential election, the group debunked 12 pieces of viral fake news that had been shared a collective 1.2 million times on Facebook. One of these items—a story saying Haddad was promoting a “gay kit” that would spread homosexual “ideologies” to public schoolchildren—was shared 400,000 times, by Aos Fatos’ count.
Attempts to limit disinformation by tech companies and Brazil’s electoral court, the TSE, a seven-person body that hears all cases related to electoral fraud, only scratch the surface, Nalon says. Even after the court prohibited the spread of the “gay kit” story, she points out, it was shared 100,000 times over the next week.
In June the court asked each political party to sign an agreement not to spread fake news, but not all of them did. Compliance was effectively a token gesture, given the challenges of tracing fake news to its source. Once the campaign was in full swing, the TSE invited representatives from WhatsApp and Facebook to discuss methods to prevent fake news distribution; Facebook ultimately set up what it called a “war room” to combat fake news, while WhatsApp banned hundreds of thousands of accounts between the first round of the election and the runoff. In an email, a spokesman for WhatsApp wrote, “We’ve set strict limits on how messages can be sent, including placing a limit on forwarding ahead of the election period in Brazil for all users. In addition, we sponsored a broad education campaign and worked with Comprova to provide them our new WhatsApp Business solution, which helped make it possible for them to fact check rumors from tens of thousands of users.”
But the sense among most Brazilians was that all this was too little, too late.
The risk for a country so attached to social media is that, absent effective moderation or regulation, digital politics will become a permanent version of the American Thanksgiving family dinner, marked by outraged rants that swirl a rainbow of perspectives into a sludgy mess. Pedro Abreu, a 37-year-old graphic designer in Salvador, has become an outspoken critic of closed WhatsApp groups. “I’ve been enraged. I’ve even ended up crying when I hear stories of how my friends felt attacked by their own family members,” he says. When his own mother left his family’s group because she was offended by what he describes as the “candid and raw” tenor of the political debate, he started a group for Brazilians who’ve left their family groups, as one might storm away from the Thanksgiving table.
Historically, Brazilian media consumers have relied on Globo, a company founded in 1925 and owned by the family of founder Irineu Marinho, for news. But over the years, as it grew in size and power, its influence came under scrutiny. Globo’s support for the military dictatorship set the tone for many Brazilians, who continued to suspect, even after military rule ended in 1985, that the company’s TV and radio shows conspired to attack candidates it didn’t like and gloss over corruption accusations against candidates it did, with the ultimate goal of ensuring that the wealthiest members of society stayed that way. (The company disavowed its support of the dictatorship in a 2013 editorial and has said it considers democracy “an absolute value.”) Santoro, the political scientist, says that in this election Globo appears to have had virtually no role in steering the electorate’s opinion—the first time that’s happened in modern Brazilian history.
The shift to social media has left McCann, the professor who once praised Brazil’s aggressive adoption of Orkut and its descendants, reeling. He did, he says, predict the rise of the right in Brazil and the broader increase in nontraditional media consumption, but he missed the way social media, especially WhatsApp, would be used to distort the information reaching the electorate. “This is a disaster,” he says. “I feel like I was naive in not seeing this coming earlier.”
Should the pattern evident in this year’s elections become entrenched, the implications for Brazil—and other countries with social media addictions and wobbly democratic institutions—will be far-reaching. Michael Patrick Lynch, a philosophy professor at the University of Connecticut who studies fake news, says that as people’s views become distorted by false information, “they start to not know what to think or who to trust.” Over time, this breakdown of what he calls “epistemic,” or knowledge-based, trust can threaten democracy itself. When people start to believe that all information is biased, Lynch says, they tend to either double down on preexisting beliefs or opt out. The electorate is then divided between those who dominate the discourse with information that supports only their personal views and those who tune out politics entirely.
Evidence of the former group is abundant in Brazil, but there are signs of the latter, too. Voting is obligatory by law, which in theory ensures participation in the political process. But Brazilians are allowed to cast “null” and “blank” ballots, which are viewed as protest votes and don’t enter into the final tally. Some can also abstain because they’ve moved away from their registered residence. In the runoff, just over 30 percent of the electorate fell into one of these three categories, amounting to 42.5 million nonvotes in an election that was decided by 10.8 million ballots.
The twin trends of polarization and apathy will place a heavy burden on company-sponsored initiatives such as Comprova and civil society groups like Aos Fatos. “For 2022, we’ll need a much better way to deal with fake news,” says Aos Fatos’ Nalon, referring to the next presidential election. “We’ll need new technology, and we’ll need WhatsApp and Facebook to invest more in managing the information on their platforms.” Her donor-funded group has expanded from two to nine employees and has begun monitoring campaigns at other levels of government, such as the gubernatorial election in the state of São Paulo. The people behind Comprova, meanwhile, hope to keep the project going, and are working to establish similar initiatives in India, Indonesia, and Nigeria.
A lot could change in Brazil by 2022, though. Bolsonaro’s antidemocratic rhetoric augurs ill—not long after winning the runoff, he attacked Folha de São Paulo, saying, “This newspaper is over.” With him in power, McCann predicts, the country “will become more partisan and atomized.” Brazilians’ shared sense of national identity survived decades of dictatorship, but now, he says, “the election shows that that doesn’t really exist anymore, that people have totally opposed ideas of what it means to be Brazilian.”
To contact the editor responsible for this story: Jeremy Keehn at email@example.com, Eric Gelman
©2018 Bloomberg L.P.