Peace-builder and Ashoka Fellow Helena Puig Larrauri co-founded Construct As much as remodel battle within the digital age–in locations from the U.S. to Iraq. With the exponential development of viral polarizing content material on social media, a key systemic query emerged for her: What if we made platforms pay for the harms they produce? What if we imagined a tax on polarization, akin to a carbon tax? A dialog concerning the root causes of on-line polarization, and why platforms must be held accountable for the detrimental externalities they trigger.

Konstanze Frischen: Helena, does know-how assist or hurt democracy?

Helena Puig Larrauri: It relies upon. There’s nice potential for digital applied sciences to incorporate extra folks in peace processes and democratic processes. We work on battle transformation in lots of areas throughout the globe, and know-how can actually assist embody extra folks. In Yemen, for example, it may be very tough to include girls’s viewpoints into the peace course of. So we labored with the UN to make use of WhatsApp, a quite simple know-how, to succeed in out to girls and have their voices heard, avoiding safety and logistical challenges. That is one instance of the potential. On the flip aspect, digital applied sciences result in immense challenges – from surveillance to manipulation. And right here, our work is to know how digital applied sciences are impacting battle escalation, and what could be carried out to mitigate that.

Frischen: You may have workers working in international locations like Yemen, Kenya, Germany and the US. How does it present up when digital media escalates battle?

Puig Larrauri: Right here is an instance: We labored with companions in northeast Iraq, analyzing how conversations occur on Fb, and it shortly confirmed that what folks mentioned and the way they positioned themselves needed to do with how they spoke about their sectarian id, whether or not they mentioned they have been Arabic or Kurdish. However what was taking place at a deeper degree is that customers began to affiliate an individual’s opinion with their id – which implies that ultimately, what issues just isn’t a lot what’s being mentioned, however who’s saying it: your individual folks, or different folks. And it meant that the conversations on Fb have been extraordinarily polarized. And never in a wholesome means, however by id. All of us should be capable of disagree on points in a democratic course of, in a peace course of. However when identities or teams begin opposing one another, that is what we name affective polarization. And what meaning is that it doesn’t matter what you truly say, I will disagree with you due to the group that you just belong to. Or, the flip aspect, it doesn’t matter what you say, I will agree with you due to the group that you just belong to. When a debate is at that state, you then’re in a scenario the place battle could be very prone to be harmful. And escalate to violence.

Frischen: Are you saying social media makes your work tougher as a result of it drives affective polarization?

Puig Larrauri: Sure, it definitely seems like the chances are stacked in opposition to our work. Offline, there could also be area, however on-line, it usually seems like there is no means that we will begin a peaceable dialog. I keep in mind a dialog with the chief of our work in Africa, Caleb. He mentioned to me in the course of the latest election cycle in Kenya “once I stroll the streets, I really feel like that is going to be a peaceable election. However once I learn social media, it’s a struggle zone.” I keep in mind this as a result of even for us, who’re professionals within the area, it’s unsettling.

Frischen: The usual means for platforms to react to hate speech is content material moderation — detecting it, labeling it, relying on the jurisdiction, maybe eradicating it. You say that’s not sufficient. Why?

Puig Larrauri: Content material moderation helps in very particular conditions – it helps with hate speech, which is in some ways the tip of the iceberg. However affective polarization is usually expressed in different methods, for instance via concern. Worry speech just isn’t the identical as hate speech. It may possibly’t be so simply recognized. It most likely will not violate the phrases of service. But we all know that concern speech can be utilized to incite violence. Nevertheless it would not fall foul of the content material moderation pointers of platforms. That’s only one instance, the purpose is that content material moderation will solely ever catch a small a part of the content material that’s amplifying divisions. Maria Ressa, the Nobel Prize Winner and Filipino journalist, mentioned that not too long ago so nicely. She mentioned one thing alongside the traces that the problem with content material moderation is it’s such as you fetch a cup of water from a polluted river, clear the water, however then put it again into the river. So I say we have to construct a water filtration plant.

Frischen: Let’s discuss that – the foundation trigger. What has that underlying structure of social media platforms to do with the proliferation of polarization?

Puig Larrauri: There’s truly two explanation why polarization thrives on social media. One is that it invitations folks to govern others and to deploy harassment on mass. Troll armies, Cambridge Analytica – we’ve all heard these tales, let’s put that apart for a second. The opposite side, which I feel deserves much more consideration, is the way in which wherein social media algorithms are constructed: They’re seeking to serve you up with content material that’s participating. And we all know that affective polarizing content material, that positions teams in opposition to one another, could be very emotive, and really participating. In consequence, the algorithms serve it up extra. So what meaning is that social media platforms present incentives to provide content material that’s polarizing, as a result of it will likely be extra participating, which is incentivizing folks to provide extra content material like that, which makes it extra participating, and so forth. It is a vicious circle.

Frischen: So the unfold of divisive content material is sort of a aspect impact of this enterprise mannequin that makes cash off participating content material.

Puig Larrauri: Sure, that is the way in which that social media platforms are designed in the mean time: to have interaction folks with content material, any sort of content material, we do not care what that content material is, until it is hate speech or one thing else that violates a slim coverage, proper, wherein case, we are going to take it down, however basically, what we would like is extra engagement on something. And that’s constructed into their enterprise mannequin. Extra engagement permits them to promote extra adverts, it permits them to gather extra knowledge. They need folks to spend extra time on the platform. So engagement is the important thing metric. It isn’t the one metric, however it’s the important thing metric that algorithms are optimizing for.

Frischen: What framework might drive social media firms to alter this mannequin?

Puig Larrauri: Nice query, however to know what I’m about to suggest, let me say first that the principle factor to know is that social media is altering the way in which that we perceive ourselves and different teams. It’s creating divisions in society, and amplifying politically present divisions. That is the distinction between specializing in hate speech, and specializing in this concept of polarization. Hate speech and harassment is about what the person expertise of being on social media is, which is essential. However after we take into consideration polarization, we’re speaking concerning the affect social media is having on society as an entire, no matter whether or not I am being personally harassed. I’m nonetheless being impacted by the truth that I am residing in a extra polarized society. It’s a societal detrimental externality. There’s one thing that affects all of us, no matter whether or not we’re individually affected by one thing.

Frischen: Detrimental externality is an economics time period that – I’m simplifying – describes that in a manufacturing or consumption course of, there’s a price being generated, a detrimental affect, which isn’t captured by the market mechanisms, and it’s harming another person.

Puig Larrauri: Sure, and the important thing right here is that that value just isn’t included within the manufacturing prices. Let’s take air air pollution. Historically, in industrial capitalism, folks have been producing issues like vehicles and machines, within the strategy of which additionally they produced environmental air pollution. However first, no person needed to pay for the air pollution. It was as if that value did not exist, though it was truly a detrimental value to society, however it simply wasn’t being priced by the market. One thing very related is occurring with social media platforms proper now. Their revenue mannequin is not to create polarization, they simply have an incentive to create content material that’s participating, no matter whether or not it is polarizing or not, however polarization occurs as a by-product, and there is no incentive to wash it up, similar to there was no incentive to wash up air pollution. And that is why polarization is a detrimental externality of this platform enterprise mannequin.

Frischen: And what are you proposing we do about that?

Puig Larrauri: Make social media firms pay for it. By bringing the societal air pollution they trigger into the market mechanism. That’s in impact what we did with environmental air pollution – we mentioned it must be taxed, there must be carbon taxes or another mechanism like cap and commerce that make firms pay for the detrimental externality they create. And for that to occur, we needed to measure issues like CO2 output, or carbon footprints. So my query is: Might we do one thing related with polarization? Might we are saying that social media platforms or maybe any platform that’s pushed by an algorithm must be taxed for his or her polarization footprint?

Frischen: Taxation of polarization is such a artistic, novel means to consider forcing platforms to alter their enterprise mannequin. I wish to acknowledge there are others on the market – within the U.S., there’s a dialogue concerning the reform of part 230 that at the moment shields social media platforms from legal responsibility, and….

Puig Larrauri: Sure, and there is additionally a really large debate, which I am very supportive of, and a part of, about design social media platforms in a different way by making algorithms optimize for one thing aside from engagement, one thing that may be much less polluting, and produce much less polarization. That is an extremely vital debate. The query I’ve, nonetheless, is how can we incentivize firms to truly take that on? How can we incentivize them to say, Sure, I will make these modifications, I am not going to make use of this straightforward engagement metric anymore, I will tackle these design modifications within the underlying structure. And I feel the way in which to try this is to primarily present a monetary disincentive to not doing it, which is why I am so on this thought of a tax.

Frischen: How would you guarantee taxing content material just isn’t seen as undermining protections of free speech? An enormous argument, particularly within the U.S., the place you possibly can unfold disinformation and hate speech beneath this umbrella.

Puig Larrauri: I do not assume {that a} polarization footprint essentially wants to have a look at speech. It may possibly have a look at metrics that must do with the design of the platform. It may possibly have a look at, for instance, the connection between belonging to a bunch and solely seeing sure varieties of content material. So it would not have to get into problems with hate speech or free speech and the talk round censorship that comes with that. It may possibly look merely at design selections round engagement. As I mentioned earlier than, I truly do not assume that content material moderation and censorship is what is going on to work significantly nicely to deal with polarization on platforms. What we now have to do is to set to work to measure this polarization footprint, and discover the appropriate metrics that may be utilized throughout platforms.

For extra comply with Helena Puig and Construct Up.