May 13, 2020
Mollie Zapata is the Research Manager of the Simon-Skjodt Center for the Prevention of Genocide’s Early Warning Project. She has co-authored early warning reports on Mali and Côte d’Ivoire, and regularly briefs policymakers on how to most effectively use early warning information in decision making.
Today, policymakers are increasingly embracing data science and new technologies, recognizing the critical role they can play in promoting peace and improving security. As the adoption of these tools accelerates, renewed attention is needed to the challenges of communication, security, and ethics.
These recurring themes were prominent at a recent workshop, Data for Peace and Security: leveraging the opportunities of data-driven innovation and technological developments, focused on using data and technology to prevent conflict and promote peace worldwide. Hosted by the Netherlands Ministry of Foreign Affairs in February, the two-day workshop brought together representatives from governments, major international organizations, NGOs, private companies, and universities.
At the workshop, I presented the Early Warning Project, the Simon-Skjodt Center’s data-driven effort, which uses statistical modeling and crowd forecasting techniques to identify countries at risk for mass killing. Given that one of the Project’s goals is advancing the science of early warning, I attended several presentations to share our work and learn from others. Through the presentations and discussions that accompanied them, I noted some common themes and challenges faced by those who analyze and use data for peace and security.
Growing Acceptance of the Practical Applications of Data and Technology
In the past, discussions of data and technology were concentrated in the academic and intelligence spaces or amongst designated “data people” within organizations. Today, however, there appears to be a broadening acceptance of the importance and utility of data amongst policymakers. This is good news; policymakers should be able to leverage all available information, and it behooves analysts creating quantitative or tech-based systems to understand the needs and preferences of their users.
This trend is reflected in the Simon-Skjodt Center’s integration of the Early Warning Project’s quantitative assessments into our daily work. We use them to select countries for additional research and as an additional source of information on countries we are tracking. We have also noted that recently our target audiences—policymakers, NGOs, and academics—seem more open to discussing and engaging with our data. For the past two years, we have been invited to launch our annual report on Capitol Hill, and have presented the results to policymakers in Washington, New York City, Brussels, and Addis Ababa.
Importance of Communicating Data-Driven Findings to Policy Audiences
Multiple sessions and discussions focused on how to “bridge the gap” between data and policy. For example, one session was framed around challenges in “Transforming Early Warning into Early Action.” Participants discussed the need for analysts to frequently engage directly with policymakers to build trust in the data and other tech-based solutions, the importance of communicating findings responsibly, being clear on what data and tech tools can and cannot be used for, and the need to include local voices in presentations of data to make the numbers meaningful and personal.
We have learned that talking through our Early Warning Project analysis with key stakeholders is critical, and that explicitly stating what we want people to do with the results—how to think about them, how to use them, and what other tools might complement our analysis—is an important step towards improving their utility. To this end, we have a page on our website about how to use our risk assessments and include key questions users should ask in our annual report.
Technological Advances in the Field
One of the most interesting and exciting parts of these types of workshops is learning about innovations in the field. Technological developments enable better and faster data collection, analysis, and presentation of results. For example, one organization uses remote sensing technology—literally “listening” to the types of airplanes approaching—to detect imminent airstrikes in Syria and sound the alarm for civilians to take cover. Others, including our Statistical Risk Assessment, benefit from advances in machine learning algorithms to identify what factors contribute to risk. When it was created in 2015, our Early Warning Project used theory-based models to produce risk assessments. But today, we and others use machine learning techniques to produce assessments, and some are experimenting with such tools to collect and analyze “big data”—social media analysis, geospatial data, and crowd-sourced information. These types of algorithms, along with newly available data, can help us better understand things like the role of climate change in conflict, to take one example presented by the German Ministry of Foreign Affairs.
Ethical and Security Considerations
The final theme I noted was the ethical considerations around collecting and using data in situations where peoples’ security is at stake. This theme reflects a broadening consensus that as our abilities to collect, synthesize, and analyze information advance, so too must our consideration of the risks posed to those we seek to protect. When considering data collection, organizations must take responsibility for protecting all personally identifiable information (PII) while it’s being collected, as well as for its long-term storage, and should make informed decisions about what data is necessary to collect.
Some participants raised concerns a step beyond individual privacy, to focus on and consider group-level identification risks—could collecting any data about a specific group put them at risk for targeting, now or in the future? Through a real-time survey, we learned that most organizations represented did not have an expert responsible for data security, making it all of our jobs to consider these risks. Furthermore, security is never a completed task; as technology continues to advance, so too must our systems.
While in-person workshops may be temporarily on hold, engaging with this community of analysts, scholars, and policymakers will ensure that our Early Warning Project continues to contribute to the data for peace and security field, and offer lessons for us on how to conduct our work in a way that will help us achieve our ultimate goals: to prevent genocide and mass atrocities worldwide.
PREVIOUS POST: Civilians at Risk in Burundi Ahead of May Elections