Dynamics of Environmental Change, Livelihoods, and Migration in Bangladesh: An Agent-Based Modeling Approach

Abstract

Environmental change interacts with human mobility in complex ways that depend on interactions between impacts on individual households and on communities. These coupled individual-collective dynamics make agent-based simulations useful for studying environmental migration. I present an original agent-based model (ABM) that simulates environment-migration dynamics in terms of the impacts of natural hazards on labor markets in rural communities. I use a pattern-oriented approach that seeks to reproduce observed patterns of environmentally-driven migration in Bangladesh. In my first study, I find that a simple economic model robustly reproduces the patterns of interest, and that correctly representing economic inequality in land ownership is essential to matching the observed patterns. I also demonstrate a machine learning method for calibrating model parameters in a pattern-oriented approach. In my second study, I expand my ABM to include more complex decision-making based on the psychological Theory of Planned Behavior (TPB). I find that the simple utility maximization method performed better than the more psychologically complex method at reproducing the migration patterns of interest. As with the utility maximization method, I find that the distribution of land ownership is critically important. Next, I investigate the role of social networks by testing several different network structures and sizes within the model. I find that the network structure does not have a significant effect on the patterns of migration outcomes, but that increasing network size increases overall rates of migration. Finally, I investigate the extreme heat in Bangladesh under future climate change. Wet-bulb temperature (Twb) is an important measure of the heat stress extreme heat poses to people’s health, especially to those who work outdoors and who do not have access to air conditioning. I find that Twb is expected to increase in Bangladesh in the future under moderate and severe global warming scenarios, which will likely affect health and also migration. Future work will continue to expand on incorporating behaviorally complex decision methods into my model as well as incorporate participatory methods.