Internal migration as climate adaptation: A gravity model approach in seven latin american countries

Among the multiple drivers for human migration, climate and environmental risks are receiving much attention in migration research, predominantly linked to out-migration. In this sense, migration is understood as a response and an adaptation mechanism to climate and environmental stressors. Climate plays an important role in the decision-making process to migrate, for and against (Mueller, et al, 2019; Thiede, et al, 2022). Heterogeneity in migration and climate research leads us to ask whether and to what extent migration can and should be viewed in blanket terms as climate adaptation or whether this question can be reframed or even abandoned. This paper employs gravity models to analyze the relation between the flows of migration-sending and migrant-receiving level 2 administrative boundaries and climate vulnerability scores in seven latin american and caribbean countries: Bolivia, Chile, Costa Rica, Cuba, Ecuador, Guatemala and Mexico,  while controlling for demographic characteristics, assets, migration prevalence and proximity among boundaries.  For this analysis we use Admin2 Level data from IPUMS International and the Physical Vulnerability to Climate Change (PVCCI).  IPUMS international databases provide harmonized micro census data which provide place of residence five years prior to data collection.  The PVCCI is a composite indicator measuring climatic shocks and countries' exposure to these shocks at subnational administrative units that range from 0 to 100. We calculate the quartile values for each country and after merging the micro census data and the PVCCI indicator, we estimate the flow of migrants among boundaries.

Our hypothesis is that migration flows will show movement of people from more to less climate vulnerable places. The initial results, however, show no consistent pattern in migration flows in relation to PVCCI quartiles in origin or score in destination, but there are clear and significant positive results in some countries. Further routes of analysis are required to better understand the differences among countries.