This article was originally published on the Humanitarian Coalition’s Relief to Recovery blog, here.
One in every 122 humans is either a refugee, internally displaced or seeking asylum. The world’s attention is focusing on Europe as desperate refugees flee violence and persecution. The UNHCR’s estimate is that 42,500 people are newly displaced each day, a migration not witnessed at this magnitude in 70 years, since the end of World War II. Yet the response has been purely reactionary – politicians struggling to combat rising tensions from the strain on infrastructure and overwhelmed aid organizations with insufficient resources. But, instead of being reactive, could we have predicted this and any other refugee and humanitarian crisis?
Much has changed since WWII. Technology is now ubiquitous – by the end of 2015, there will be over seven billion mobile cell subscriptions, and 95% of the world’s population will be covered by at least a 2G mobile-cellular network – and we have a tremendous ability to aggregate and monitor data. Can these not be combined to improve our ability to prepare for and cope with an influx of refugees? Says Rana Novack, founder of the Refugee Admissions Network Alliance, “We have the technology – right here, right now – to create a new, agile, insightful model that will predict mass migrations and help us better serve displaced families even before they are displaced.”
Gathering Characteristics of Vulnerable Populations
Flowminder, a Swedish nonprofit, has developed a technology that uses position data from SIM cards to track the movement of people. With a focus on assisting vulnerable low and middle-income countries at scale, the organization collects, aggregates and analyzes anonymous mobile operator data – through cooperation with mobile companies – and data from satellites and household surveys.
In doing so, they take billions of data points and produce information on the distribution and characteristics of vulnerable populations. In advance, they can extrapolate from general population mobility to predict where people will go in an emergency situation.
First used in Tanzania in 2008 to determine if malaria could be eliminated from Zanzibar, they estimated the movements of those infected with malaria from the mainland to Zanzibar. In Haiti, in the wake of the 2010 earthquake and cholera outbreak, they tracked population displacements to assist relief agencies in orienting their work. Contrary to their estimates, the resultant mass exodus from Port-au Prince was surprisingly highly predictable and largely influenced by historical behaviour and their social bonds.
Most recently, Flowminder has been used with the Ebola outbreak, to predict the spread of the disease and determine where to focus preventive efforts, and in the aftermath of the Nepalese earthquakes.
The use of mobile operator data does raise some concerns, particularly of privacy, convincing mobile phone providers to provide access to their data and the possibility the data could be used negatively, to close off borders. It is also one thing to know and another to act – there must be a political will to acknowledge and prepare for the predictions, and humanitarian organizations must be open to outside technology and to change.
Nevertheless, the technology behind Flowminder, in collaboration with humanitarian organizations, has tremendous potential. The organization is largely focused on improving public health and welfare, but similar technology could be applied to map the flow of refugees.
Humanitarian crises do not emerge from thin air, but instead are the result of specific sets of circumstances. If we approach them proactively by analyzing and predicting trends, we can use that time and knowledge to be better prepared and scale up responses. We can prepare contingency plans, allocate resources and streamline responses.
However, we will also not have the luxury of saying we had no way of knowing. Next time, even more so than this time, we have an obligation to be better prepared.