Global Migration Patterns: Community Detection and Network Analysis of Bangladesh Subgraphs

Abstract:

Bangladesh is recognized as one of the most vulnerable nations to impacts from climate change, with many environmental factors driving migration. Network analysis measures can illustrate relationships between countries in terms of ongoing migration patterns, such as out-migration from Bangladesh. Using annual dyadic migrant stock data from 1990–2017, a community detection algorithm was applied to partition global migration stocks into subgraphs. The community-detection algorithm resulted in quasi-regional subgraphs that remain relatively stable over time. Hierarchical clustering and blockmodeling analyses were performed on the global migration subgraph in which Bangladesh is situated. The blockmodeling revealed patterns of countries within the subgraph network which are structurally equivalent, allowing for further inference and correlation with factors such as economics, policy, and environmental events.