Sheldon Himelfarb, director of USIP’s Center of Innovation: Science, Technology, and Peacebuilding, on December 12 spoke at the Council on Foreign Relations workshop on “Social Media and Conflict Prevention.” Himelfarb discussed the latest from USIP’s ongoing Blogs & Bullets initiative, showing how lessons from existing social media-based efforts can be used to improve the impact of future efforts in this regard, especially through the use of more nuanced, data-based research.

Here, he outlines the insights this initiative is generating about this emerging field.

What recent trends are we seeing in the use of social media tools to manage conflict?

While much of the attention in Washington has focused on debating the role of social media in the Arab Spring for the last two years, we’ve watched a quieter revolution happening around the globe in the use of social media at the community level to try and mitigate causes of violence. The extent to which we see this at work in the field every day wherever cell phones are found cannot be overstated. Some of the major trends in this field include:

Fostering inter-ethnic dialogue: We see social media being used to bridge divides between adversaries, especially by youth. This is not just a “feel-good” tactic. It’s aimed at getting ahead of cycles of conflict. There is a very solid conflict resolution curriculum and evaluation platform underpinning the Salam Shabab online youth network in Iraq, showing shifting attitudes about ethnic diversity. And  a less rigorous, but larger, 200,000-member YaLa-Young Leaders network is taking shape in the Middle East between Israelis and Palestinians, creating a growing lobby against violent conflict.

Managing elections: In places like Kenya and Sudan, voting is actively monitored for everything from violence to fraud using a range of social media platforms. And both countries recently held referenda that were considered successes in violence prevention by international observers

Preventing gang violence: Twitter penetration in Brazil is among the highest on the planet, and communities in Brazilian favelas, or shantytowns, say it has been a game-changer in helping bring down both gang and police violence. We’re seeing similar programs in places like Ciudad Juarez in Mexico.

Preventing resource disputes: Early-warning networks like CEWARN in sub-Saharan Africa are being set up to leverage social media, along with other forms of satellite information and traditional media reports, to prevent conflict over land, water and other resources.

Constitution-building: We’ve seen efforts to use social media in transitional nations like Egypt to help build constitutions with public input through social media, though it wasn’t very successful there.  But there, as in Morocco and Iceland, we are learning a lot about crowdsourcing input on constitution-writing, and I can see it continuing down the road.

Protesting violence: By now, most have heard of Oscar Morales, the young Colombian engineer who used Facebook to rally people all across Colombia, and the world, against the violent tactics of the guerrilla group commonly known as the FARC in 2008. That was just the beginning of a trend in the use of social networks to bring people together to protest against violence.

Fundamentally, I can’t think of a single issue we work on in the conflict management field – preventing election violence, preventing inter-ethnic killings, preventing violence over land – in which we haven’t seen an effort to use social media and mobile phones in order to inflect the causes of violence.  Social media is rapidly becoming an almost ubiquitous tool everywhere we work, and people are using this tool, with mixed success, to prevent violence.

How do we ensure that we go from “mixed” successes to more consistent impact?

That’s the question we are asking ourselves in an initiative we began three years ago, called Blogs and Bullets, and I’m still not sure we have too many  good answers yet because these are early days. We’ve been working with experts from George Washington University, Harvard University’s Berkman Center and others. Here are a few key insights we’ve come up with in our work:

Impact analysis on five levels: The first big “a-ha” was more about process than about tactics. As we tried to unpack all of the ways social media was being used in conflict settings, we realized we needed a better analytical framework than simply to simply ask, “What’s the role of social media in conflict prevention or conflict promotion?”  And our research indicated that the impact of social media can be better understood through a framework that considers five levels of analysis: individuals, collective action among individuals, intergroup action, regime activity and external audiences.

So in asking whether the shifting dynamics of the war in Colombia was a result of the Facebook campaign, or whether the last referendum in Kenya was relatively peaceful as a result of the early warning networks or if, at a more local level, prevention efforts in a specific county using social media were successful – it is instructive to have these five levels of analysis for our assessment. And when we use this framework, I think we begin to see social media as an invaluable tool in bridging and enabling successful bottom up-top down violence prevention campaigns.

Measurement matters: Another takeaway has been around the importance of measurement. For example, the George Washington University team did an analysis around milestone events in the Arab Spring. They mined millions of data-points provided by Bit.ly, the link shortener used extensively within Twitter and other social networks, and found evidence that new media informs international audiences and mainstream media reporting more than it plays a direct role in informing local audiences. Researchers sourced the location of users who clicked links related to the Arab Spring movements in four countries, and the vast majority of clicks came from outside the country. This data suggests that social media was more important in drawing external attention to the protests than in local collective action, creating a kind of megaphone effect.

Certainly, this is not conventional wisdom.  Using these technologies, so the contention goes, people interested in democracy can organize political action with a speed and scale never seen before. But the data throws some doubt on those sweeping claims.

So again, the interesting question is not, “Was social media the key ingredient?”, but rather, “How did it matter?” And measurement is essential to producing answers. Will it tell the full story? Absolutely not. As our report indicates, mere output --  or numbers of tweets for example – does not equate to impact. But I am highlighting the importance of measurement because, our field struggles mightily with data-driven analysis and decision-making. I think most of my colleagues would agree with this.

Invest in People. Empower Communities:  Once peacekeeping equated to international forces, but today, as I observed earlier, local citizens often take violence prevention into their own hands, thanks in part to social media. A book of case studies for these recently came out from MIT Press, titled “The Technology of Non Violence: Social Media and Violence Prevention.” It analyzes a handful of social media-based violence prevention initiatives with specific outcomes. We recently invited the author, Professor Bock to USIP and it was validating to see how close his conclusions are to our own based on a handful of local programs we are supporting in  Nigeria, Haiti, the slums of Kenya and elsewhere. Like us, Dr. Bock also observes that it is still early days for the use of social media in conflict management, leaving us with more questions than answers at this stage, but on this we’re both sure: the technology is a key enabler, and that it is most effective when used by people in local communities who have the training or the talent, and preferably both.

This, by the way, also applies to countering the harmful uses of social media – for disinformation, for inciting violence, for hate speech, etc. The most effective strategy for dealing with these things is investing in the individuals and communities closest to the conflict. No amount of technology can replace that kind of preparedness and training which, by the way, is getting ridiculously small dollars compared with the billions that is going into the “predictive analytics” that we hear so much about. But what happens when you have a good, actionable prediction? It is the local communities that are the key to effective action – and preparing them needs to get more of our focus.