Interviewed for La Presse

5 minute read


I was recently interviewed by La Presse about our recent pre-print (OSF/Twitter thread).

Find the piece here (in French). The English version (translated by Google Translate) can be found here.

Transcript of the original interview is as follows:

Question: Were you surprised when you realized that politics isn’t the main driver of the US Twitterverse?

Answer: It’s somewhat reassuring to see that people on Twitter aren’t that different from the norm in the United States in being indifferent to politics. Twitter “drives” content on social media in different ways, but individual choices to follow accounts remains an important way for people to control the content they see. What is instead surprising is that this effect isn’t reflected often in the other research which explores filter bubbles and polarisation on Twitter. By focusing solely on politics on Twitter, we have perhaps been blindsided by how much non-politics is really out there.

Question: So basically the debate about the political impact of Elon Musk buying Twitter is overstated since most people don’t go on Twitter to follow politics?

Answer: What could the political impact be – will we see the amount of political content increase on Twitter? We can’t say – we don’t expect that people will suddenly take up discussing politics. Will we see the amplification of political content decrease – that’s a separate question, and one that Musk does not speak to.

While we can’t say what the political impact of Elon Musk buying Twitter will be, we do think that it will have to do more with the changes in platform policies he introduces than in increasing the polarization or the politicization of the platform. But many of his talking points are possibly based on speculation that Twitter has a liberal bias – whereas in some studies, the findings suggest the opposite. And in another study of ours that is currently under peer review at a journal, we find that Twitter shadowbans political tweets by downtiering political content from both the left and the right. Therefore, it will perhaps be best if Twitter makes evidence-based policy decisions.

What we can hope for is a more open-sourced Twitter, with greater transparency and accountability in the decisions it makes.

Question: It’s interesting to note that journalists writing about Twitter see it through the political lens since they themselves follow politics on Twitter. Is it a case of, to a hammer, every problem looks like a nail?

Answer: Yes, it is. There’s this phenomenon that scholars call “the streetlight effect” where a drunkard looks for his lost keys under the streetlight because that’s where the light is. Similarly, journalists writing about Twitter tend to use what they readily see (their political feeds, and others like them, tweeting about politics) to inform their opinions and draw (often unfounded) inferences about the platform at large. There’s scholarly work that shows how Washington journalists operate within insular microbubbles, which are cut off from the rest of the American public. If they step out of these bubbles they’ll perhaps see the politically disinterested reality that most of Twitter is. In this way, Twitter mirrors most of America, which is also politically disinterested. In fact, as political scientist Yanna Krupnikov writes, the real divide in the US, is between a politically engaged minority and a politically disinterested majority. We suspect that Twitter would be similar as well. Again, it’s important to consider that their – and scholarly – observations are often based on the content and trends that Twitter drives and perhaps not what people encounter day-to-day in their feeds, which doesn’t make them wrong but just not representative of the norm.

Question: So since politics is not the main driver of Twitter engagement in the US, should we all chill out a bit about what’s doing on on this social network in terms of polarisation?

Answer: We should probably worry about who is controlling free speech and how, rather than about polarisation per se And we’d like to remind people that free speech isn’t only to do with politics. In our study, we find that most of the information on Twitter is being produced by centrists or moderates. In fact, a lot of the popular narrative of polarisation on Twitter is unscientific and anecdotal. Even if we look at studies that claim to find evidence of polarisation, chances are, they’re focusing on political tweets or political elites only, and not Twitter at large. If we do look at Twitter at large, we find that the non-political content is overwhelmingly more than the political, partisan stuff.

In recent work that’s under review, we tested who is shadowbanned on Twitter and what they talk about. We suggest that those may be the important questions to ask and answer as they affect the credibility of Twitter as a platform for social activism and truly free speech. The important thing is to not use our personal Twitter experiences to draw inferences about what’s going on in the world out there (or even on Twitter more generally).