The Conditional Climate Effect: Understanding When and Where Environmental Stress Drives Migration

Empirical research on climate-related migration has produced highly heterogeneous findings. While many studies identify correlations between climatic shocks and migration flows, results vary substantially across regions, time periods, and model specifications. This heterogeneity largely reflects the continued use of linear and additive frameworks that conceptualize climate change as an isolated driver of mobility, overlooking its interaction with broader economic and social conditions. In reality, environmental stress operates through complex interdependencies involving labor demand, development levels, and adaptive capacity, which jointly shape whether individuals move, remain, or adapt in place.

This study proposes a dynamic and nonlinear empirical framework to re-examine the climate–migration nexus through an integrated lens. Building on the aspirations–capabilities approach (de Haas, 2021), it conceptualizes migration not as a direct response to climate shocks but as a conditional outcome of intersecting environmental and socio-economic forces. Using publicly available country-level panel data (possibly, for 1990-2025), the empirical strategy combines two-way fixed-effects panel regressions with nonlinear specifications - including quadratic and interaction terms between climate, labor demand, and development indicators - to allow the marginal effects of climate variability to differ across contexts. To uncover threshold and non-monotonic relationships, Generalized Additive Models (GAMs) will flexibly estimate nonlinear climate - migration responses. A dynamic panel extension based on the Arellano - Bond GMM estimator will incorporate lagged migration and climate terms to account for persistence, adaptation, and potential endogeneity.

The article aims to identify thresholds and context-dependent mechanisms under which climate variability translates into increased or reduced migration. By combining nonlinear, interactive, and dynamic modelling within a theoretically grounded framework, it contributes both conceptually and methodologically to a more nuanced understanding of the climate–migration relationship.

*This abstract was written by the author; AI tools were used solely for language editing and proofreading, while all ideas, analyses, and conceptual content are entirely the author’s own.