As the president of Argentina took to the podium to speak to Congress on March 1, local fact-checkers Chequeado were biting their nails.
In December, when he spoke after being sworn into power, Mauricio Macri only made one checkable claim — a big change from outgoing president Cristina Fernández de Kirchner, whose speeches went on for hours and were packed with claims fact-checkers could scrutinize.
But Chequeado was prepared to fact-check the event live, undeterred by the possibility of another speech devoid of facts.
I was sitting in as an observer for Full Fact, the U.K.’s leading fact-checking organization. With a growing community of international fact-checkers, there is plenty to learn from observing how other organizations are approaching similar challenges in different countries.
Chequeado set up shop in a lecture theater at Di Tella University in Buenos Aires. The room was large, and it needed to be: Chequeado’s staff was joined by a group of 40 students, journalists and experts who volunteered to help.
Volunteers were assigned to teams, each led by one of Chequeado’s editors. At the back were security and narcotics, political institutions, and economics. At the front were social protection and health, and education.
A hush fell over the room as Macri started speaking. His first checkable phrase after only a few minutes was greeted with relieved laughter and a “thank you” from Chequeado director Laura Zommer: There were going to be fact checks to write after all.
As the claims piled up, Zommer and editorial coordinator Olivia Sohr made pages and pages of notes. Separately, editor Ariel Riera used a Google Doc to document checkable phrases that were tweeted out along with any previous checks on the same claim.
The scene was recognizable but different from our work at Full Fact. Our ideal tweet contains a conclusion about the claim, plus a link to the source. If we think the research will take more than a few minutes, we might put the claim on hold and return to it in a post-debate roundup.
There were echoes of the U.K. political debate in both the topics of Macri’s speech — inflation, employment, debt and deficit, crime statistics and teachers’ pay — and in the kind of tweets Chequeado received: “how ignorant are you Macri!” “is that compared with the changing purchasing power of salaries and pensions?” “Recommended: follow the excellent work of the Chequeado team on the opening of Congress”
Occasionally, the president’s speech was interrupted by interference on the broadcast stream or chanting from other politicians.
The disruptions were similar to those we get in the UK’s House of Commons, but there’s more chanting and singing. As Chequeado’s new web developer says, “It’s like a football match.”
A burst of excited discussion erupted when #MacriChequeado, the hashtag Chequeado was using to share its content, became a top-trending topic on Argentine Twitter. This quickly subsided as fact-checkers returned to monitoring Macri’s every word.
During the speech, Sohr went from team to team, discussing which claims could be tackled first and assigning claims to subject specialists. Chequeado’s editors coordinated and wrote up their team’s research, checked links and data and made sure judgments are consistent and fair.
I asked Sohr about the learning curve for new fact-checking volunteers. She says that while there aren’t particularly common pitfalls, sometimes volunteers’ conclusions are harsher than those Chequeado’s own staff would give.
Once drafted, the editors sent their fact checks to Zommer for a final edit. She looked for sources to support factual statements and make sure that those sources are credible. Zommer also decides what order the fact checks should be published in, considering not only what is ready to go, but also neutrality. Four fact checks concluding that a claim was incorrect would not be published at the same time, but spaced out between fact checks with less critical conclusions.
Full Fact’s system is a bit different. Researchers edit and check each other’s work, including the clarity of the writing, calculations, sources and phrasing that might be read as biased. The director doesn’t usually edit articles unless the topic is very controversial. Because we use a “claim and conclusion” summary, rather than a rating system, we don’t have to worry so much about what to publish first.
The team identified 95 checkable claims in Macri’s speech. Even before he was done speaking, Zommer was editing an article about employment. Five hours after that 15 fact checks had been published on topics ranging from inflation and the deficit to children’s early years development and Argentina’s global transparency rating.
Full Fact doesn’t publish separate fact checks for one speech or political event, nor do we check so many claims. All the fact checks of an election debate or a politician’s speech would be published as part of one ‘roundup’ article, analyzing a maximum of seven or eight claims.
When it was all over, I wondered whether Full Fact could involve more volunteers in our research and publication process. The obvious benefit is that you can publish more, faster, and so have a better chance of stopping inaccurate claims before they take hold in public debate.
But perhaps a more lasting impact is the result of face-to-face contact and direct feedback from experienced staff: enthusiastic fact-checkers who feel confident enough to assess a claim without needing to take anyone else’s word for it but their own.
