Tom Bearpark

Quantifying and projecting climate migration is a critical challenge for policymakers and researchers. While causal inference, agent-based, gravity, and general equilibrium models have been employed widely to estimate the climate-migration relationship, their results often diverge, and they are typically used in isolation and at different spatio-temporal scales. Here, we aim to bridge this gap by introducing a unified notation and conceptual framework, designed to compare and evaluate the assumptions and outputs of these diverse models at a common scale. We illustrate our comparison using an empirically relevant case study: estimating and projecting the relationship between annual average temperature and internal state-to-state migration in the United States. We analyse results from hundreds of model permutations, illustrating how methodological choices and researcher degrees of freedom shape model outcomes. Our findings show that variations in modelling approach introduce uncertainty comparable in magnitude to statistical uncertainty. By integrating policy-relevant insights across methods, we highlight the value of model intercomparison in improving projections and informing policy decisions.

Empirical climate impacts research is dominated by studies that use panel fixed-effects regressions to estimate the causal effect of weather on outcomes. However, the functional form of time controls in these studies, i.e., the specification of time trends, lacks a clear theoretical foundation or data-driven justification. Moreover, these choices can substantially influence the magnitude of estimated results in important applications. In this paper, we elucidate a framework for choosing time controls in climate impacts studies. We propose a two-equation model which aligns with the reasoning used in climate impacts studies to justify identification. In our framework, competing models are evaluated according to their ability to isolate plausibly random variation in the treatment variable. We provide simulations to illustrate our proposal, and apply our framework to an open and policy relevant agenda in the climate impacts literature; estimating the effect of temperature on GDP growth rates. 

Projected temperature changes are uncertain in both their magnitude and geography. This paper studies how nonlinear damages and general equilibrium forces filter this climate risk across time and space. To do so, we build a tractable dynamic spatial model linking countries through trade and migration, and derive analytical first- and second-order welfare approximations that decompose the mean and variance of welfare changes into damage function, trade, and migration components. Using an ensemble of temperature projections from CMIP-6 and an empirically estimated damage function, we show that climate change-induced welfare risk is large. The standard deviation of country-level projected welfare loss across temperature projections is on average over 8% – compared to an average welfare loss of 13% – and is spatially unequal. Climate risk inequality is half as large as global income inequality. We show that spatial linkages reshape not only the level of climate damages, but also the spatial distribution of climate risk: accounting for trade and migration can reduce the standard deviation of welfare changes by up to 40% in low-income, internationally integrated nations that can diversify their exposure to local shocks.