The small number of cross-border displacements that were identified appears to support the assumption that while numbers may in some cases be significant, they are relatively small compared with those for people displaced internally. Factors that drive people to cross borders include the extent of the damage wrought by a disaster, poor access to basic services and recovery assistance and the proximity and porosity of the nearest borders. The severe food crisis or famine brought on by recurrent drought against a backdrop of poverty and insecurity in some parts of the Horn of Africa illustrate both the importance of understanding that cross-border displacement can be key to the survival and resilience of vulnerable populations, and the limited nature of the data and information available. This is discussed further in the spotlight on South Sudan. Even when cross-border disaster displacement is monitored or quantified, common frameworks and methodologies for doing so are lacking. People who flee beyond their own countries for reasons other than conflict, generalised violence or persecution do not fit the legal definition of a refugee or asylum seeker. The fact that someone displaced across a border by a disaster does not exist as a legal category makes it less likely that they will be systematically recorded or identifiable in official data and statistics. As with other types of displacement, nor are there any universally recognised criteria to determine whether a person’s flight across a border as a result of disaster should be characterised as forced or voluntary. For the purpose of providing protection and assistance, the Nansen Initiative suggests criteria that include the directness and seriousness of a disaster’s impact on the person in question, and their pre-existing vulnerabilities. It also suggests benchmarks that consider solidarity with an affected country temporarily unable to assist and protect all of those in need, whether international humanitarian access is possible or not, and the extent of the people’s ties with family or diaspora communities in the destination country.171 An alternative approach might be to consider whether obliging people to return would be realistic, legal or morally responsible.172 54 GRID 2017 Before any such criteria can be applied and priority given to those in greatest need, the systematic collection and sharing of data must overcome conceptual, technical and political obstacles to monitoring and reporting.173 Addressing data and knowledge gaps is the first of three priorities identified in the Nansen Initiative’s protection agenda for people displaced across borders in the context of disasters and climate change. This is reflected in the 2016 to 2019 work plan for the Platform on Disaster Displacement, a state-led multi-stakeholder initiative taking forward the agenda’s implementation.174 In support of the protection agenda and work plan, and in order to better quantify and understand displacement associated with disasters, IDMC is gradually broadening the scope of its global monitoring to capture data and build knowledge about both internal and crossborder flows. Its work also supports the agenda’s comprehensive approach, which recognises the need “to reduce vulnerability and build resilience to disaster displacement risk, facilitate migration out of hazardous areas, conduct planned relocation and respond to the needs of internally displaced persons.”175