nohaadil/Modelling-Internal-Displacement-in-Sudan

# Modelling Internal Displacement in Sudan: A Two-Stage Bayesian Framework for Displacement Occurrence and Destination Patterns This repository is the reproduction companion to a thesis submitted in partial fulfilment of the requirements for the degree of MSc in Social Data Science at the Oxford Internet Institute, University of Oxford, August 2026. It contains the analytical data and the code that produced every model, table and figure reported in the thesis. --- ## Overview Since the outbreak of war between the Sudanese Armed Forces and the Rapid Support Forces in April 2023, Sudan has become the site of the world's largest internal displacement crisis, with approximately 11.5 million people internally displaced by January 2025. The dynamics of this displacement remain largely unexamined in the quantitative forced migration literature. The study provides one of the first systematic statistical accounts of internal displacement during the conflict, examining both the conditions under which populations leave their localities and the factors shaping where they move. It employs a two-stage Bayesian framework on locality-level data spanning September 2023 to December 2024, integrating IOM Displacement Tracking Matrix (DTM) records with conflict events (ACLED), food insecurity classifications (IPC), and baseline geodemographic data (WorldPop, OCHA). Note: All estimates are associations. The observational design does not support causal interpretation. --- ## The two stages | | **Stage I** — occurrence | **Stage II** — destination choice | |---|---|---| | **Question** | Which conditions at origin accompany new displacement | How the displaced distribute across destinations | | **Unit** | Locality by reporting round | Origin–destination corridor by round | | **Model** | Bayesian multilevel logistic regression, locality varying intercept | Hierarchical negative binomial gravity model, round varying intercept | | **Notebooks** | `analysis/01`, `analysis/02` | `analy …