Listen

News 88.9 KNPR
Classical 89.7 KCNV
'Jazz'

member station

NPR
National

Social Media Sites Under Pressure To Prep For Census Trolls And Interference

774944142_2142557876.jpg

Lawmakers and civil rights groups are pressuring tech companies to come up with detailed policies about how to combat potential misinformation and disinformation about the 2020 census.
Denis Charlet, AFP via Getty Images

Lawmakers and civil rights groups are pressuring tech companies to come up with detailed policies about how to combat potential misinformation and disinformation about the 2020 census.

The first U.S. census to allow all households to participate online is facing another unprecedented challenge — the looming threat of disinformation through social media.

As 2020 draws closer, federal officials fear foreign governments and Internet trolls could use Facebook, Twitter and other platforms to spread rumors and propaganda to derail the constitutionally mandated count with at least 10 years' worth of implications on elections around the country.

Results from the head count of every person living in the U.S. are used not only to determine the distribution of congressional seats and Electoral College votes among the states but also to redraw state and local voting districts beginning in 2021.

"This is a new age. The last decennial census didn't really face this type of challenge," Census Bureau Director Steven Dillingham acknowledged in April during a U.S. Senate hearing where Sen. Jack Reed, D-R.I., pressed for answers on how the bureau is preparing to combat disinformation campaigns and any other online attempts to disrupt the count.

"We're working with the Facebooks. We're working with the Twitters. We're working with Microsoft," explained Dillingham, who said the bureau is also coordinating with the Department of Homeland Security and intelligence agencies to determine potential sources of disruptions to the 2020 census.

Support comes from

Pressure to "do more and do better"

Lawmakers and civil rights groups have been ratcheting up the pressure on tech companies not to overlook the census as they roll out policies for dealing with the potential onslaught of disinformation about the 2020 election. Last week, Twitter announced it would stop running political ads later this month.

"We believe social media platforms must do more and do better than was done in 2016 to deter and mitigate interference in our upcoming count," wrote 57 House Democrats, led by Rep. Bill Pascrell of New Jersey, in a letter to Twitter's CEO, Jack Dorsey, that was released Monday.

This week, members of the Congressional Asian Pacific American Caucus plan to send a letter to Facebook's chief operating officer, Sheryl Sandberg, asking for a briefing on the company's census plans, according to the press office of Rep. Judy Chu, D-Calif.

"We are deeply concerned that the [Asian American and Pacific Islander] community, which is already vulnerable to being undercounted, will be susceptible to misinformation and dissuasion tactics by nefarious actors who wish to undermine our democracy," write the lawmakers in a draft shared with NPR.

Facebook, in particular, has come under fire recently for its decision to not block or fact-check what politicians say on its platform. During a U.S. House hearing last month, CEO Mark Zuckerberg said that Facebook believes that "in a democracy, it is important that people can see for themselves what politicians are saying."

In an open letter to Zuckerberg released Tuesday, however, the head of the Lawyers' Committee for Civil Rights Under Law, Kristen Clarke, pushed back on that argument.

"We know that your decision to allow unchecked false statements by politicians will increase voter and census disinformation campaigns and increase activity exposing Africans Americans and other people of color to harm," Clarke wrote.

The wait for specifics

With just over four months until the head count is fully underway, members of Congress and other census advocates are still waiting for some tech companies to release policies designed specifically to counter census disinformation.

Zuckerberg appeared mealy-mouthed during the House hearing when Rep. Vicente Gonzalez, D-Texas, tried to gauge the limits of Facebook's policy.

Federal law prohibits the release of census information identifying individuals until 72 years after it's collected. But Gonzalez asked hypothetically if Facebook would take down an ad that falsely stated information about immigrants who participate in the census would be shared with U.S. Immigration and Customs Enforcement.

"Okay. So congressman, where we are right now is we have in place a voter suppression policy," Zuckerberg replied, "and we're working on finalizing that to extend that to a census suppression policy as well."

"This isn't voter suppression," Gonzalez countered.

"I agree. I'm sorry. I'm trying to answer your question," Zuckerberg said, before adding that he expects more specifics on the company's census plans to be announced "in the coming weeks."

"The proof will be in the pudding"

Sen. Brian Schatz, D-Hawaii, has been pressing tech companies for months to spell out their plans. In letters sent to the heads of Facebook, Twitter, Google and Reddit in July, Schatz asked the companies to provide "clear and specific" policies by Oct. 31.

In their responses to Schatz, details were laid out by Reddit and Twitter (whose updated policy categorizes the U.S. census as a "civic event"), while Google and Facebook deferred to an unspecified date.

"I think it's fair to say that some of them are further along than others in terms of developing the policies," Schatz told NPR.

The senator noted that for now, tech companies are "saying the right things about being on top of it."

"But the proof will be in the pudding over the next several months to see whether they're actually doing it," Schatz added.

Google spokesperson Nu Wexler did not answer NPR's question about when the company plans to release its policy. But in a written statement, Wexler said, "Google and YouTube are committed to combating disinformation and fraudulent activities to help protect the integrity of the 2020 Census."

The danger of a "data void"

Civil rights organizations are particularly concerned about the potential impact of census rumors and propaganda on communities of color and immigrant groups, who may not trust government authorities, especially after the more than year-long legal battle over the citizenship question the Trump administration failed to add to 2020 census forms.

"That fear can really be exacerbated and exploited when inaccurate information is put out in these communities," said Vanita Gupta, president and CEO of the Leadership Conference on Civil and Human Rights and a former Justice Department official during the Obama administration.

Some disinformation researchers have highlighted the relative dearth of results when searching online for content about the census. It sets up what researchers at Microsoft call a "data void," ready for Internet trolls to exploit.

"When there's misinformation or disinformation put out, it can fill the void of information," Gupta said.

Fiction spreads "more quickly" than the truth

The polarized rhetoric drummed up by the citizenship question debate has also turned the census into a "weak" area of U.S. society that malicious actors will want to target, according to Young Mie Kim, a professor at the University of Wisconsin-Madison who led a study of Facebook ads released in the run-up to the 2016 elections.

Kim said another potential challenge is the spread of misinformation or disinformation about the census by U.S. politicians, whom social media platforms may not want to cross for fear of appearing partisan.

"They don't want to deal with a political situation," Kim said. "There's no incentive for tech platforms to do that."

Regardless of where the disinformation is coming from, a key factor in fighting it off will be how fast a social media platform can take action, explained Dipayan Ghosh, a former adviser to Facebook on privacy and public policy.

Facebook, for example, "has to develop a tremendously sophisticated artificial intelligence to understand the content and take it down in a timely manner and be courageous to take it down in a timely manner, even when it knows that it might be violating freedom of political expression," said Ghosh, who now co-directs the Digital Platforms and Democracy Project at Harvard Kennedy School's Shorenstein Center.

In a study about stories shared on Twitter, MIT researchers found that falsehoods are more novel and that novel information is more likely to be retweeted.

"It's actually spread around by people more quickly than the truth," Ghosh said. "That's a deeply concerning fact when we consider that we want a fair census."

Copyright 2019 NPR. To see more, visit https://www.npr.org.

You won’t find a paywall here. Come as often as you like — we’re not counting. You’ve found a like-minded tribe that cherishes what a free press stands for.  If you can spend another couple of minutes making a pledge of as little as $5, you’ll feel like a superhero defending democracy for less than the cost of a month of Netflix.