How data is playing a key role in American security

Last week, Kenyans went to the polls in what was a hotly contested and potentially violent presidential election. On the day of the election, observers from the U.S. fanned out across the country, but U.S. government preparations to help this key ally avoid unrest started months before.

Some of the earliest steps leveraged big data analytics — a powerful tool for an evidence-based approach to preventing and responding to conflict and violent extremism. But it is a tool that works best when the data is put in context and informed by insights from diplomats on the ground.

{mosads}This is the approach we take in the State Department’s Bureau of Conflict and Stabilization Operations. It is precisely how we’ve approached our role helping the Department and our partners in the U.S. Agency for International Development (USAID) and the Defense Department prevent conflict, counter violent extremism and promote stability. Since 2013, the bureau has combined predictive data analytics with insights from analysts trained in social science methodologies that we deploy overseas. The combined analysis helps focus U.S. government strategies on the factors driving a specific conflict.

This is big data and social science working together. By combining big data with proven analytic methods rooted in social science, the bureau has been able to complement the extensive political and economic reporting carried out by U.S. diplomats from around the world and identify previously unseen or unconfirmed patterns and associations. 

The bureau routinely compares data that documents the who, what, where, when and how of political and extremist violence, with other data on public perceptions, income, demographics and more in order to identify correlations between events and historical context. We have done this for various facets of extremism, stabilization, elections violence and atrocities. 

After identifying patterns, we can assess and compare risks and pinpoint hotspots to focus our efforts. Using these methods, we have been able to show that abuses by state security forces are a key driver of violent extremism and that states with records of abuse are at greater risk.

With big data and social science methods, we are also able to map networks of militias, insurgent groups and violent extremist organizations. We can measure degrees of connectivity and levels of influence to identify strong and weak points in these networks.

Similarly, by using computer modeling to test informal and formal negotiations, we can determine most likely outcomes, potential coalitions of actors, key influencers, potential areas of give and take, and most likely pathways to outcomes that support U.S. goals and objectives. 

Of course, without good interagency coordination, data analytics isn’t going to do much to prevent conflict. It is critical we leverage the combined elements of U.S. power defense, diplomacy and development. 

This is exactly why the bureau established a novel Regional Support Team for East Africa early last year to aid implementation of counter terrorism and violent extremism in the region. This seven-person team — operating from Djibouti, Kenya, Tanzania and at home in Washington — has taken analytics to the field and brings together expertise from USAID missions, embassies, and U.S. Special Forces.

We promote the sharing of survey data on public perceptions between these U.S. government agencies, which avoids duplication of effort. Where gaps do exist, often at the local or community level, we lead assessments and invest in targeted surveys, focus groups and locally led research in hotspots.

The Regional Support Team also serves as the anchor point for a web of programs to monitor key indicators across the region on violent extremism and implementation of the Partnership for Regional East Africa Counter Terrorism.

For the bureau, data analytics has been a game changer — like healthcare advancements of moving from x-rays to more granular MRIs and CAT scans — and consequently earlier and better diagnoses. This is crucial, because it’s more cost effective to prevent conflict than it is to support military intervention after conflict erupts and extremism takes hold. As Secretary of Defense James Mattis put it while still in uniform, “the more that we put into the State Department’s diplomacy, hopefully the less we have to put into a military budget.” 

Data analytics also helps the department systematically compensate for cognitive biases everyone tends to have when weighing information, like being over-reliant on the first (or last) piece of information we get on a topic or placing too much importance on one event. But data analytics works best when coupled with forward deployed field analysts that can interact with communities to see the grievances driving conflict as the communities themselves see them. Evidence and perspective are the crucial ingredients the Bureau of Conflict and Stabilization Operations uses for more effective and efficient conflict prevention.

Tom Hushek is the acting Assistant Secretary of State for Conflict and Stabilization Operations. He is a career Foreign Service officer whose previous overseas postings include the U.S. Mission to the International Organizations in Vienna, Micronesia, Tajikistan, Pakistan, Afghanistan, Sudan and Russia.

Brian Keane is a director of Analysis, Planning, Programs and Learning at the Bureau of Conflict and Stabilization Operations and helped to establish its data analytics capability.

The views expressed by contributors are their own and are not the views of The Hill.