Understanding disaster displacement and assessing future risk

Recording displacement in the context of disasters and estimating the risk of future displacements is challenging. Evacuation numbers, both for planned and spontaneous evacuations, and arrivals in shelters are recorded on an ad hoc basis during or immediately after a disaster event. However, just days after even a major event, very few data are collected on people remaining in shelters, returning or moving on to other locations. Weeks and months after a disaster, this lack of data is even more pronounced, making protracted displacement all but invisible. Most of the data used to record disaster displacement are published by government authorities, drawing on loss accounting systems or humanitarian situation reports and PDNAs. In several instances, evacuation numbers cannot be properly established at all. In these cases, data on the destruction of housing can act as a proxy for displacement (International Organization for Migration and IDMC, 2023). Compared to evacuation data, which include people who may return home only hours after an event, data on the destruction of data can also provide insights into how long people are likely to remain displaced, i.e. unable to return home. In situations of largescale housing destruction, longer-term displacement can be assumed, even if it must be noted that displacement dynamics are complex, and linking potential displacement to a single variable such as housing destruction is a simplification. Nonetheless, housing destruction and displacement are related

In turn, monitoring disaster displacement via housing destruction and other indicators, such as loss of agricultural production or infrastructure damage, collected through national disaster loss accounting, contribute to modelling potential future impacts and can inform preparedness for shelter provision as well as for looming longer-term displacement crises. In Indonesia, data collected and stored in the loss accounting system of the National Disaster Management Agency provides a clear picture of the primary triggers of displacement in the country and specific locations that are most affected by them. Analysis of the data shows that flood-related displacements mainly occur in December and January and are largely concentrated in Central and West Java and Aceh (IDMC, 2024). Combined with data on population growth and urbanization, this analysis provides a critical source of information for prioritizing contingency planning and response preparation for recurring events. In addition, collecting data on displaced populations throughout their displacement offers an insight into the types of support they need to avoid long-term displacement. In the Philippines, the national Disaster Response Operations Monitoring and Information Center assesses the number of people in evacuation centres as well as those hosted by family and friends every 12 hours and often for months after an event. Data from the Disaster Response Operations Monitoring and Information Center show that the duration of displacement in evacuation centres tends to be longer, i.e. that those staying with family return home earlier. Such insights have important implications for shelter planning in disasters, as well as the support provided to those displaced within communities (ASEAN Coordinating Centre for Humanitarian Assistance on Disaster Management, n.d.). Disaster losses and damages data can also inform displacement risk assessments with a longer perspective into the future. IDMC has developed a displacement risk model, building on the UNDRR-led probabilistic approach employed for the Global Assessment Reports (see Figure 24).

The model uses housing destruction as a proxy for displacement and relies on relevant data collected during and after disaster events to calibrate and validate the model. Recognizing that housing destruction alone is a limited way of estimating displacement, and as people's level of vulnerability and exposure to hazards change over time, IDMC and its partners continue to develop new and further improved methods to assess displacement risk (see Figures 25 and 26). For this, disaster loss data on the loss of agricultural production, infrastructure and services are important sources of information, as they help conceptualize and estimate displacement linked to loss of income and employment or the lack of access to basic services.