THREAD: So, a little while ago, @RVAwonk and I saw a tweet that baffled us. That's right folks, I'm talking about that time Theo Fleury said @gmbutts is sitting on $23 million. Which is extremely false! So where did it come from? nationalobserver.com/2019/07/16/new… #cdnpoli

One thing was clear right away: this false info didn't start with Theo. He was just the most prominent person to tweet about it. As far as we could tell, it originated on a fake news/clickbait site in 2016, as part of a listicle about the "richest politicians" in Canada.

There were a LOT of problems with this listicle. The grammar & spelling are atrocious. It seems to pull biographical details from Wikipedia and pair them with random, absurdly high net worths. It doesn't cite sources. It has no byline. The site is called TopRichests. Like, c'mon

The article's marked 2019, but archived versions go back to 2016 (source code shows it was updated this year). tbh, it reads like someone Googled 'Canadian politicians' in early 2016. Lots of Harper appointees in there (his name is misspelled as "Stephan"). Also Tom Mulcair!

So, the listicle is fairly obvious disinformation. @TheHillTimes wrote a story about it a few years ago, and it's been publicly refuted by plenty of people featured in it (fun fact: one of them is @jkenney!) So why is it still circulating? Why are people still believing it?

Hint: it's not because of the original listicle alone. The issue is that people copied and cross-posted this information all over the internet. We found it in odd places! A children's book blog, Facebook, YouTube, Twitter, even Quora, the Q&A site used by teens doing homework.

Each time the false information was removed from its original source and repeated, it started to look a little more credible and pick up a little more steam. There are a few psychological phenomenons at play there, which @RVAwonk explains very well in the story...

But another important thing is that the false claim was picked up by Twitter accounts that appear to be bots. To verify this, we used botcheck.me, which picked up 7...

... we also saw tens of other accounts with bot-like characteristics repeating this claim (Botcheck didn't detect these, but the tool isn't 100%). A lot had sus handles (first name + lots of numbers), spammy tweets and stock art profile photos.

We don't know exactly how Theo Fleury came across this information, because he wouldn't talk to us about it. But we do know that when claims are repeated often in several different places, they look more credible.

We can't tell who's behind the listicle. Was it carelessness? Nefariousness? As @mckelveyf explains, it could be either. Low-quality clickbait is somewhat profitable thanks to online ads, so lots of sites do it. This is bad. Creates disinfo that can be used for partisan purposes.

We also can't tell who was behind this apparent bot effort to amplify the disinformation. Anyone can get bot accounts anonymously. But we did talk to @HeidiTworek, who explained that this kind of thing is usually about stoking division.

What does that division accomplish? With the heavy caveat that no one piece of disinformation will turn someone into an extremist, it does encourage extreme views. It could also encourage distrust in democracy -- in this case, it implies that public officials are all corrupt.

It's not really clear how effective this attempt was, but we thought it was an interesting case study. There's lots more to this, so please read the full story! It's a wild ride! nationalobserver.com/2019/07/16/new… #cdnpoli

This story is part of @NatObserver's Election Integrity Reporting Project, and we couldn't have done it without support, time and resources. If you appreciate this work and you're able, please consider subscribing so we can do more of it. nationalobserver.com/subscribe