Using Bayesian belief networks to analyse social-ecological conditions for migration in the Sahel

Abstract In order to understand the impact of climatic and environmental changes as well as socio-economic drivers on human migration, it remains a challenging task to find a method to analyse the knowledge from different scientific disciplines in an integrated way. The Sahel region with its high ecological dynamic has a long history of migratory movements. Within this work, we integrate and analyse socio- and natural-scientific data from two Sahelian study areas in Mali and Senegal using Bayesian belief networks. The core of the network’s structure is formed by four main motives to migrate which are education, family, visit and curiosity and sustenance and employment. It is assumed that these motives determine the spatial and temporal patterns of migration. On the basis of submodels for each migration motive, we identify the decisive factors that constitute the socio-economic and ecological conditions with a combination of sensitivity analyses and train-and-test validation method. In combining these factors, the model avoids implying monocausal dependencies and allows an analytical view on the likely consequences of different settings of social-ecological conditions on migration. Furthermore, we use the model to estimate the consequences of alternative future developments in contrasting scenarios. The results show that changing environmental conditions lead to changing patterns of migration, regarding its duration and destination. These patterns can be very specific for different motives and their underlying factors. One principal result of the analysis is that uncertainty in the main income sources correlates with an increase of short-term migrations in order to increase the households' possibilities for income generation. Nevertheless, socio-economic conditions show a greater impact on the people's decision to migrate than environmental conditions.