Mobile Phone Data for Crisis-Induced Migration and Mobility

Forced displacement has reached unprecedented levels globally, with 1.5 percent of the world’s population forcibly displaced in 2023, a sharp increase from a decade earlier. This growing scale underscores the need for timely, evidence-based approaches to monitor, analyze, and anticipate migration flows. Traditional migration data sources face substantial limitations, including temporal delays and spatial constraints. This PhD dissertation addresses a critical research gap by examining how Mobile Phone Data (MPD) can be used to measure and analyze crisis-induced migration and mobility. The central research question explores how MPD can be effectively utilized for migration and mobility research in crisis contexts. Chapter 2 describes how three MPD sources—call detail records (CDR), extended detail records (xDR), and airtime top-up transfers (ATT)—were ethically prepared and accessed. Chapter 3 reviews existing practices, opportunities, and challenges in using MPD for anticipating displacement. The remaining chapters present empirical studies introducing novel methods and proxies for measuring migration and mobility. The first research question investigates how different MPD sources can support new measurement approaches for internal displacement and cross-border mobility that complement conventional data sources. Chapter 4 demonstrates that international ATT counts can serve as proxies for migrant stocks and provide insights into migrant groups missing from official statistics. Chapter 5 introduces an xDR-based indicator of cross-border mobility, termed the “lost subscriber.” Using this indicator alongside antenna-level signals, we estimated border crossings during the 2020 Turkish–European border opening, a highly politicized and speculative event. Chapter 6 presents a novel disaster-induced displacement measurement methodology, the Activity Space Approach (ASA), which improves the detection of displacement origins and destinations at finer spatial granularity. The second research question examines how publicly available data collected by governments and international organizations, as well as other big data sources, can triangulate and contextualize MPD-based migration indicators during crises. Across Chapters 4, 5, and 6, the dissertation demonstrates systematic data fusion approaches that combine MPD with official statistics, social media data, and administrative records to validate results and support contextual interpretation. The third research question explores how MPD indicators can be leveraged to analyze displacement dynamics before, during, and after crises. The empirical chapters demonstrate MPD’s utility for crisis response and impact assessment. Chapter 4 assesses the impact of COVID-19 on ATT flows using gravity analysis, while Chapter 6 shows that ASA-based measurements capture individual-level variation in displacement timing, urban–rural transitions, movement away from damaged areas, and trajectories toward borders and other areas of interest. Finally, the fourth research question examines MPD’s potential for analyzing differential mobility impacts on vulnerable populations. Chapter 6 reveals contrasts between the displacement patterns of Syrian refugees and the local Turkish population following the 2023 Türkiye–Syria earthquake, showing that Syrians experienced more spatially and temporally constrained mobility. Overall, this dissertation contributes novel measurement methodologies and data sources, systematic data fusion frameworks, and empirical evidence on crisis-induced mobility. It establishes MPD as a valuable complementary tool for migration research and humanitarian response, advancing ethical and privacy-preserving approaches for real-time crisis monitoring in an era of increasing environmental and political instability.