Measuring the “polarization footprint” of social media

From Syria to Yemen, Burkina Faso and Myanmar to Mali and the United States – conflict has migrated to online spaces. Recognizing this important shift, peacebuilder and Ashoka Fellow Helena Puig Larrauri co-founded Build Up to transform how we address conflict in the digital age.  

Ashoka’s Konstanze Frischen explored novel ideas with Helena on how to measure and reduce the amount of viral polarizing content spread through social media platforms (i.e., their “polarization footprint”) – a persistent challenge for peacebuilders in particular. You can watch the full conversation here. Here are a few of the highlights:  

Tech for peace and digital conflict 

Working with NGOs, UN agencies, community leaders, and multilateral organizations, Build Up develops innovative tech solutions to conflict. This means using technology to, for example, increase citizen participation in the context of peacebuilding processes, making the outcomes more optimal and longer lasting. And it also means examining the ways digital technologies exacerbate conflict. Helena illustrates this with a couple of examples. 

Why content moderation doesn’t help much

Over the years, tackling the exponential growth of viral polarizing content spread on social media has become a pressing challenge for peacebuilders. Helena is careful to distinguish between differences of opinions, which are essential for any democracy, and the rise of identity-based polarization, which the Build Up team refers to as “affective polarization.”

Affective polarization exacerbates offline conflict and is known to affect electoral and peace processes, which is why Helena and her team are increasingly focused on finding ways to decrease it. Here’s why content moderation is appropriate to deal with hate speech, but not the solution Build Up is focused on to reduce polarization. 

How maximizing engagement has led us here

If content moderation alone is not the solution, then what could be? To find answers, Helena says we have to look at the underlying business model of social media platforms. In a nutshell, engagement is the key metric of success for social media algorithms because it’s what allows platforms to target us with ads and extract more data from us – both of which constitute their main revenue streams. And it is widely documented that polarizing content is very engaging.

Measuring and taxing social media platforms’ polarization footprint

While social media platforms may not set out to pollute our lives with polarization by prioritizing clicks above all else, it is one of the negative consequences of their business model. What if we were able to measure the amount of viral polarizing content spread on social media - i.e. their polarization footprint – and press for compensation for the harm they produce? Imagine a tax on polarization, akin to a carbon tax. By putting a price on polarization, it would finally create an incentive for companies to clean up the data pollution they produce.  

Designing better social media algorithms

Let’s not forget that optimizing for engagement is only one possible business model. Platforms could design algorithms that optimize for other things, such as up-ranking content that bridges divides or de-escalates conflict. But Helena says, those alternatives may only gain steam if platforms are incentivized to change their business model. And that’s where a tax on polarization comes in handy!  

.    .    . 

This conversation is part of a collaboration under Ashoka's Tech & Humanity Impact Labs, mitigating the societal and environmental harms of the data economy.