Nieman Foundation at Harvard
On Election Day, the nonprofit investigative outlet led Electionland, a coalition of more than 1,100 people that covered voting problems across the country, and in a tweetstorm written by senior reporting fellow Jessica Huseman, ProPublica said it “saw no evidence” that the election was rigged. Here are the first six tweets in a chain of 15:
We had 1,100 people monitoring the vote on Election Day. We saw no evidence the election was “rigged” no matter what Stein or Trump say. /1
— ProPublica (@ProPublica) November 27, 2016
Electionland had huge amounts of data. 600 ppl monitored social media. We had @LawyersComm call logs. We had 120,000 people texting us. /2
— ProPublica (@ProPublica) November 27, 2016
We had 400 partner reporters across the country, including three of the largest news organizations in the U.S. /3 https://t.co/jWJYpqyHXJ
— ProPublica (@ProPublica) November 27, 2016
We had voting experts in the room with us and election sources all over the country. /4
— ProPublica (@ProPublica) November 27, 2016
We saw plenty of problems: Long lines, broken voting machines, incorrect poll books, confusion abt voter ID laws /5 https://t.co/IWzZgtscmV
— ProPublica (@ProPublica) November 27, 2016
But we saw no reason to doubt the results. There 👏🏼 was 👏🏼 no 👏🏼 widespread 👏🏼 voter 👏🏼 fraud. /6 https://t.co/ulzmUb9onU
— ProPublica (@ProPublica) November 27, 2016
ProPublica is still tallying the number of stories that originated from Electionland’s reporting (and more stories are still being written) but it said that, so far, it’s found about 300 stories that appeared either in local or national outlets or on Electionland’s own site.
“Donald Trump [said] over the weekend that several million people voted illegally,” ProPublica deputy managing editor Scott Klein said. “We have data that shows it’s simply not true. If anything happened at the scale that he’s implying, we would’ve seen it. We were there. We had a real-time look at the nation and how things were going.”
Though ProPublica spearheaded the effort, six other organizations took part in organizing Electionland. In total, 450 journalists from about 250 news organizations took part in the project.
Using tips and tools from the First Draft Coalition, students from 14 journalism schools scoured social media looking for reports of issues at the polls. Electionland also worked with the Google News Lab to analyze search trends. It had access to reports submitted to Election Protection, an initiative run by the Lawyers’ Committee for Civil Rights Under Law that lets people report voting problems by phone.
Additionally, WNYC developed a tool that enabled Electionland to interact with voters through SMS text messages to report issues. In total, about 100,000 people received texts from Electionland.
All of the data and reports were then sent out to the local news organizations that participated in Electionland.
The collaboration with local news organizations was critical to Electionland’s success, said Celeste LeCompte, ProPublica’s director of business development and a former Nieman Fellow.
For instance, there was a lot of talk leading up to Election Day about possible voter intimidation in Philadelphia. On Nov. 8, Electionland noticed a spike in search traffic around the topic. While there were some one-off confrontations, they were able to use on-the-ground reporting to determine that there weren’t widespread intimidation issues. The spike, it turned out, was likely from people searching whether there was intimidation going on.
“A lot of the reporters we were working with are familiar with looking at the data in their own jurisdictions,” LeCompte said. “We know what the social media looks like for our area, but having that national context was incredibly helpful for knowing they weren’t missing something. We were there saying, ‘We’re seeing the same thing you are. We’re seeing no more real reports than you might be if you’d spent all day doing this. You’re not seeing something anomalous to the national trend.’”
Electionland’s main newsroom was located at the CUNY Graduate School of Journalism in New York. More than 100 students and journalists worked out of the space.
To communicate with reporters in New York and around the country, Electionland had a dedicated Slack. It also used Landslide, a custom-built dashboard that organized all of the incoming information.
Electionland used the early voting period leading up to Election Day to test its systems out, but Klein said the processes worked smoothly on Election Day itself.
“We expected a lot of improvisation during the day and a lot of big moves,” Klein said. “We had a couple of contingency plans in place for moving people around, in case we had guessed wrong, but it went remarkably smoothly from our vantage point. We proved that social media verification is a potent newsgathering technique that can function at scale.”
ProPublica put together state-by-state background resources that included information such as how elections are run, how many voters are in each county, and problems that were reported in the past. But Klein said that, in retrospect, he and LeCompte would have liked to provide more resources to the participants to prepare them for what they would likely see on Election Day.
“If we had to do it over again, we’d spend some time writing and teaching what we knew after months of reporting on election administration,” he said. “I think that would’ve caused even more stories to be published.”
Klein said ProPublica will continue to cover voting rights issues closely over the coming years of the Trump administration. But the organization is still working out how it could continue to utilize the Electionland infrastructure that it built up.
They are especially interested in ways to pair local and national coverage. That’s something ProPublica already does with its reporting, but LeCompte said they’re already thinking about ways to further their partnerships.
“There are an unlimited number of places where you see really strong overlaps between the ways national stories and local stories are told,” LeCompte said. “Local reporters…know their areas better than we do, but being able to help provide context, time, and resources to help make those local stories part of a larger conversation is so important, and there’s so much potential to do things like this beyond social media projects.”
