Chasing the crowd

What we seek when we turn to crowd-driven fact-checking

Hey y’all. Faked Up’s word count has been creeping dangerously upwards. That’s mostly because I’m doing extra research (not just bloviating more) but it’s still a lot to read in one sitting. So I’ve decided that whenever an individual item gets longer than ~700 words, I’ll run it as a standalone piece. The digest on Wednesday will include a summary with a link to the whole thing. Let me know if that’s a welcome development.

---

February 2014. Katy Perry’s Dark Horse is topping the charts. For the first time in history, EU parties are contesting parliamentary elections headed by named candidates. Standalone fact-checking is entering a boom phase.

It is against this backdrop that my friends and I launched the pan-European fact-checking website FactCheckEU. Our plan to debunk, in six languages, the fibs of Europe’s politicians was fueled by the belief that we could build a community of volunteer fact-checkers from all over the continent.

The website invited users to do one of four things: propose a claim for fact-checking by others, fact-check the statement themselves, translate an existing fact check into one of the other languages, and rate existing fact checks.

14 months later, FactCheckEU was kaput. The site had its fans in the Euro-bubble and penned a partnership with POLITICO Europe. But even as its parent website Pagella Politica continues to publish more than a decade later, the EU effort was too clunky and too niche to stay online. After we forgot to renew the domain, factcheckeu[.]org was snatched up by a dubious business in the concrete industry.

The problem statement

Fact-checking is slow. Lying is fast.

This fundamental imbalance has haunted fact-checking from the get-go. And new technology is of little help, because liars can use its multiplying force just as effectively as truth-seekers.

Yet the aspiration of crowd-checking is bigger than just scale – by democratizing the fact-checking process, it promises greater legitimacy of outputs. This isn’t the lamestream media, Soros-backed nonprofits or college-educated elites telling you that your favorite politician is lying – it’s ordinary citizens like you (whatever that even means).

For the past decade, fact-checkers and platforms have given crowdsourcing a shot. Yet probably no project outside Wikipedia has succeeded at delivering a broadly legitimate fact-generating crowdsourced institution at the scale necessary for the modern information ecosystem. And when Wikipedia’s founder Jimmy Wales tried to recreate the magic formula with a project that was explicitly aimed at targeting fake news, WikiTribune, he failed.

The magic formula

It seems to me that there are a handful of variables that matter in the success of a crowd-checking platform.

• WANTED: scale (# of fact checks) and legitimacy

• NEEDED: scale (# of contributors), quality, efficacy, procedural integrity, collective identity, and some leftover variable ε

I’m going to review these necessary inputs in the context of a handful of different efforts: FactCheckEU, X’s Community Notes, the French fact-checker Science Feedback and the Taiwanese platform CoFacts.