At the time of writing the proposal in 2019 unauthorized migration from West Africa to Europe was at high rates and accompanied by extreme danger and death in the Sahara, North Africa and Mediterranean.  Migration was also catalyzing political instability in Europe as right wing parties and many in the population advocated for harsh anti-migration policies.  The drivers of increased migration were not well known but the media repeatedly implied that climate change was undermining traditional ways of life in West Africa and leading desperate people to take their chances on a perilous migration to a better life in Europe.  However, social, political and economic factors could be just as important.  The purpose of this project was to disentangle the social, political, economic and environmental drivers of migration in West Africa and from West Africa to Europe with a convergence research team that included climate scientists, geographers, political scientists and economists and using qualitative, quantitative and mixed methods.  These scientists come from disciplines that traditionally attempt to explain migration using very different methodologies and assumptions, often implicit and hidden. To resolve these tensions, the work proceeded according to relentless questioning and interrogation of methods used and assumptions made and how they might bias results in terms of implied causes of migration.   Success in explaining the drivers of migration would be demonstrated if we could develop computational models that quantitatively explained the variation in international migration across countries of origin and destination and from year to year.  

The team engaged widely with a large number of migration researchers in academia and non-governmental organizations, including the United Nations, deliberately selecting people who were outside of our networks and with whom we did not have prior association, to bring in as many perspectives on the character and drivers of migration as possible.  Our early in-person (pre-pandemic) meetings in 2019 and 2020 led to a major reconception of our project and the work to be performed.  Though we thought our proposal and proposed work were written to place the influence of climate variability and change in the proper context of social, political and economic concerns, our visitors thought it places too much emphasis on climate and the conflict led to new directions in the subsequent work that attempted to resolve the tension.

One strand was to think hard about what explicit and hidden assumptions are contained in work on drivers of migration.  This highlighted the problem with arguing that climate events - e.g. a drought or flood that damages crops and livelihoods - drive migration given that that framing does not examine why farming families are so vulnerable to such events.  Shifting the frame identifies the social, political, and economic (e.g. corruption, representation, markets) realities that make farmers vulnerable to weather and climate events. Such a line of thinking identifies drivers of migration as long-term socially determined vulnerability and an international economic order rigged against small producers.  

A second strand dealt with how the lessons learned from engaging with a broad community of migration researchers and practitioners could be translated into quantitative models of migration for testing hypotheses of drivers.  This was begun with a dive into the uses and development of conceptual modeling via a series of virtual workshops with researchers, practitioners and students in West Africa and then an in-person workshop drawing on the same demographics held in Senegal.  Conceptual models can provide the frameworks used to build computational models and clarify what processes must be included and how.  For example, these can identify nonlinear processes - e.g. low income as a financial barrier to migration but high income reducing the aspiration to migrate - or thresholds in lived experience - that must be included within models. We pursued Agent-Based Modeling as one approach to include such behavior but recognize it needs to be rigorously quantitatively validated.

A  third strand used established statistical modeling approaches to predict reported migrant numbers with proposed drivers (GDP, rainfall, corruption, diaspora etc.). This related variations in internal West Africa migration to variability of cash and subsistence crop prices and rainfall.  It is now being used in attempts to predict variations in migration to Europe across countries and years.  This approach can be used to infer associations but cannot provide explanations that place migration decisions in the context of proximate causes (e.g. changes in crop yields) and underlying causes (e.g.crop prices controlled by traders and international markets).   

Our work indicates the complexity of drivers of migration, and that environmental drivers, if operative, are intertwined with social drivers making quantitative modeling and prediction challenging.  The research points the way for developing quantitative methods for predicting future migration.  As a convergence research success, the team still works together, continuing the research under different Federal funding and continues to engage the network of collaborators and advisors. 

Last Modified: 02/28/2024
Modified by: Richard Seager