From social media polarization to online deliberation
What we’re learning in Kenya and Sudan about moving difficult conversations from toxic to constructive digital spaces.
In July last year, protests against the Finance Bill — a finance instrument proposed by the government to increase taxes on essential commodities — erupted across Kenya. What were initially peaceful youth-led protests — that would later be referred to as the “Gen-Z protests” due to the overwhelming turnout of youth — eventually turned violent and deadly. Ultimately, President Ruto declined to sign the Finance Bill in response to the continued demonstrations, but deliberation on the core issues it addressed remains highly polarized.
This polarization was often blamed on the fact that Gen-Z protesters were organizing and communicating in a networked, “leaderless” way rather than through appointed spokespeople and on official channels. The Government insisted that it was hard to enter into dialogue because no trusted “representative” was there to share the movement’s position on issues. This was compounded by a sense that because the movement was communicating on social media, there were no adequate avenues to harness what was being said on policy.
That is an unfair characterization: the medium through which most protesters were talking is a fragmented medium prone to polarization and manipulation — X, Facebook, TikTok — but when the same conversations happened on the radio, on popular programs like The Situation Room, it was very different. The challenge was not in the plurality of voices that made up the Gen-Z protests, but in the spaces where they were showing up, and the processes and principles that guide these spaces.
What the Kenyan Government and other brokers of dialogue needed to consider was what kind of digital spaces would foster deliberation that led to a consensus that could more easily connect to policy processes.
This is not a Kenyan problem: social media spaces are fragmented and prone to polarization and manipulation, and this is the result of their business models as we argued a while ago here. In a nutshell: social media platforms are in the business of capturing our attention spans so they can sell ads to businesses or otherwise monetize the data they collect from how we interact with content online. Our brains like content that is emotional, easy and (critically) from people who we know or identify with. Social media platforms serve up this content because it keeps our attention, and when we indicate we like it (by watching, by giving it a heart or thumbs up, by sharing), the algorithms that prioritize what content we see amplify these likes. In some ways, this may seem like they are doing us a favour: giving us more of what we like. But as our friend Ravi Iyer aptly points out, if you put a plate of donuts in front of me, I’m going to eat it, but I don’t aspire to eat more donuts, I aspire to a diverse and balanced diet.
So why don’t social media platforms give people a diverse and balanced information diet, or at a minimum the capacity to decide more consciously what diet they want?
Some are, and we’ll come back to these hopeful examples later. But most don’t for a simple reason: it’s cheaper not to. It’s not that platforms want to be polarizing, it’s not their objective, but it is an involuntary consequence of their business model. Think of a factory that manufactures, say, chairs. The factory’s purpose is to make a profit from chair-making, and it’s cheaper to make chairs while polluting, so it has a disincentive to produce chairs cleanly. In economics, this is called a negative externality. Similarly, it’s cheaper for social media platforms to retain our attention while polarizing. Polarization is their pollution, a negative externality of their business model — it is not a necessity of their design nor what is best for individual users.
Back to Kenya, if we know this about social media platforms, then one avenue for fostering more constructive (digital) engagement at a time of critical societal conflict is to create a digital space devoid of the polarizing influence of social media algorithms; one that is structured to identify areas of consensus. There is a growing movement of “deliberative technologies” that are variously built to foster constructive dialogue by prioritizing content that is liked by people who otherwise disagree or like different things. Taiwan’s Minister of Digital Affairs is perhaps the most well-known proponent of how these new spaces can foster better, more inclusive governance that returns citizen trust to institutions.
Inspired by this, we formed a collaboration with Siasa Place (a grassroots youth NGO) and The Situation Room (a popular national radio program) to set about moving the public policy conversations sparked by the Gen-Z protests from a toxic social media environent to a constructive deliberative tech environment. Over four months, we curated public policy conversations with thousands of young Kenyans by connecting two analogue spaces (in-person youth assemblies (barazas) and a dedicated radio program) with a digital deliberative ecosystem made up of a WhatsApp bot that youth could subscribe to (which provided updates and avenues to contribute), an instance of Talk to the City (where young people could share opinions via voicenote), a series of conversations on pol.is (to hone in on policy proposals with the highest consensus socres), and a meet-up on Remesh (to bring together all the strands). Together, all these spaces are zKE, a new space for sustained national engagement by and for Kenya’s youth that we are continuing to build.
We built zKE because we could see how toxic and polarizing social media spaces were turning what could be a constructive conflict in a democratic nation to destructive conflict. But we didn’t turn away from social media. In fact, we used a variety of social media assets — TikTok explainer videos, X discussion spaces, Facebook and Instagram posts — to bring people to zKE. We did that because social media is also a space of expression for GenZ. There is a world in which social media itself could be designed with pro-social goals in mind, but while the current business models of social platform stand in the way of that, zKE uses the virality of social media to popularise deliberative digital spaces.
Our hope is zKE will make Kenya stronger and more deliberative. But the harm inflicted by toxic and polarizing social media discourse in the context of armed conflict is even more alarming. We saw this very clearly when we started analysing social media discourse in Sudan, with a particular focus on how pro-democratic, civilian actors were being attacked and maligned. Attempts to shift conversations away from division, to build a constructive narrative of unity against the war, were met with coordinated campaigns of misinformation and vilification that made any public debate on social media impossible.
Even in active conflict, deliberative technologies can help shift what is happening on social media.
Last year, we supported Tagadom (a coalition of democratic and civilian forces, now known as Sumoud) to create a digital space for the inclusion of a diversity of voices in defining what vision and positions they should represent during peace talks. We worked with Tagadom to launch a public WhatsApp channel where thousands of Sudanese could receive information about the peace process, and where Tagadom could post questions for discussion. These questions were asked via pol.is, the same deliberative platform we used in zKE, to help Tagadom understand what views had the most consensus and which were most divisive — and critically, to understand this away from the toxic cacophony that results from social media’s ranking algorithms.
Notice that these new digital spaces don’t eliminate conflict, they just turn it from destructive to constructive.
This is critical because a peaceful society is not one void of conflict, but one where all voices can negotiate a shared understanding of peace. Often at critical moments in peace processes, such as when a ceasefire is being brokered or a political concession is being made, it seems expedient to cut corners and include only key power brokers rather than a plurality of voices. Broad consultative processes complicate the situation, but they also make dialogue more real, more lasting. Constructive conflict leaves room for a critical question: who decides what peace we are working to build?
The same is true about any digital spaces that complicate peace processes. Social media can seem cacophonous during a peace process, but like any digital space that brings in more voices, this can be important. It’s a matter of balance, and what we need to ask is: how much fragmentation and polarization is it worth risking to get to meaningful bridge building and consensus?
How can digital spaces be designed so they are both spaces for a plurality of differing views and places that foster connection and social cohesion?
There is a vibrant, hopeful conversation right now about pro-social technology design governance that is attempting to bring the learnings from deliberative technologies to social media platforms and other digital spaces. If pro-social technology design becomes ubiquitous, peacebuilders and other brokers of dialogue will have less of an uphill battle in the fight to turn the violent conflict that destroys societies into the constructive deliberation that builds a plural, sustainable peace.
