Migrant Data Goes Global

Lindsay Stark speaks at the United Nations on how science is helping us understand child separation during emergencies—with and without the help of technology

September 27, 2016

Last week, the United Nations hosted an unprecedented summit on refugees and migrants. UN Secretary General Ban Ki-moon told representatives of more than 193 member states, “We are facing a crisis of epic proportions,” referring to the more than 65 million people displaced from their homes around the world.

Beyond the daunting scale of this situation, among the biggest challenges for humanitarians and academics is finding reliable information about the movements and needs of migrants. At a Thursday meeting following the summit, Lindsay Stark, associate professor in Population and Family Health, joined a panel of experts from government and civil society to discuss the state of migrant data.

Today there are more potential sources of data on migrants than ever before—text messages, GPS, satellite images, social media. Yet while they each have potential to help, members of the panel said the glut of information brings new challenges, from integrating disparate data types to coordinating between various public and private data repositories and ensuring privacy and human rights protections (for some humanitarians, “big data” has connotations of “Big Brother”). The overarching goal is to analyze all this information and disseminate it so support can be directed where it’s most needed—not least of all to migrants themselves who may be looking for safe passage, documentation, or primary care services.

The Science of Child Separation

As the sole representative of the research community on the panel, Lindsay Stark shared her experience collecting data on child separation through the Measuring Separation in Emergencies Project, a collaborative effort funded by the USAID Office of Foreign Disaster Assistance, and implemented by the Mailman School, Save the Children, and several other partners.

Traditionally, humanitarian groups perform family tracing and reunification. They find children who are separated from their families then find their caregivers or make alternate arrangements. “This is what we in the science world call ‘numerator-driven practice,’” said Stark, who wrote on the topic for the Huffington Post. “For years, the community has been moving forward without any sense of the true magnitude of child separation after an emergency has hit.’”

Even a low-tech method can provide valuable information. Stark was involved with a survey in the Democratic Republic of Congo that used baseball card-size pictorial images to represent members of a family; respondents would use the cards to depict who lived in their homes before and after an emergency event—in this case, an attack by the M23 rebel military group—and whether any children had come or gone on their own. Prior to the study, a humanitarian rule of thumb said 3 to 5 percent of children are separated during conflicts; however, Stark found that close to one in ten children were separated within communities, and another 5 percent of children had left the area.

In another study, the researchers distributed cell phones to elected representatives of communities who were charged with monitoring 150 households in their area for instances of child separation, reporting back by text on the age and sex of the child, whether they had arrived or departed from the community, and reasons behind the separation. According to Stark, the method allowed for rapid collection and distillation of information. “And it allowed us to be in areas we maybe wouldn’t have access to otherwise.”  

“There are a wealth of technologies and methodologies at our disposal,” she said. The goal now is to use the resources available to piece together information on populations who have “fallen off the statistical maps,” by virtue of not being counted through household surveys or other methods. “We are really just at the early stages of figuring out ways to do this research.”

Joining Forces

William L. Swing, director general of the International Organization for Migration (IOM), co-sponsor of the event with the Missions of the Kingdom of Belgium and Republic of Mali, had harsh words for the big picture on migrant data, calling it fragmented and largely “a mess.” Yet he was optimistic about efforts to correct the situation, particularly through an ambitious program called the Displacement Tracking Matrix. Conceived in 2004 in Iraq, DTM now collects data on more than 14 million displaced people in 30 countries.

Recognizing that companies like Facebook and Google now control much of the world’s data, Claire Melamed, director of the Global Partnership for Sustainable Development Data, spoke to the importance of bringing the private sector to the table. IOM too is broadening its horizons through partnerships with Gallup and the Economist magazine.

Appropriate to the United Nations setting, Stark emphasized that nothing can happen without the cooperation of nation states. “Getting this kind of data is dependent on political will and policies that aren’t going to drive migrating populations underground, making them less visible and making our job that much harder.”