This paper investigates the nonlinear relationship between CO2 emission and economic development using a newly developed functional coefficient panel model. In contrast to the existing literature, which suggests that the income elasticity of CO2 emission is parametrically modeled as a function of income, the income coefficient of CO2 emission is set as a function of both income and time. Then, we estimate the income elasticity in a nonparametric way using the country panel data covering 1971–2017. By doing so, we impose richer dynamics to the income elasticity not only over income but also over time. Our empirical results indicate that the income elasticity has decreased over time for high-income countries, whereas it has increased over time for low-income countries.
This study proposes a nonlinear cointegrating regression model based on the well-known energy balance climate model. Specifically, I investigate the nonlinear cointegrating regression of the mean of temperature anomaly distributions on total radiative forcing using estimated spatial distributions of temperature anomalies for the Globe, Northern Hemisphere, and Southern Hemisphere. Further, I provide two types of nonlinear response functions that map the total radiative forcing level to mean temperature anomalies. The proposed statistical model provides a climatological implication that spatially heterogenous warming effects play a significant role in identifying nonlinear climate sensitivity. Cointegration and specification tests are provided that support the existence of nonlinear effects of total radiative forcing.
We develop a distributional regression framework to quantify heterogeneous sectoral impacts of climate change by linking spatial temperature shocks to cross-sectional growth outcomes. Starting from a climate-augmented regional growth model, we show that aggregating heterogeneous regional responses into scalar averages obscures economically relevant distributional mechanisms, motivating an approach that treats temperature anomalies and sectoral growth as distribution-valued objects. Using global data on land temperature anomalies and subnational sectoral output, we estimate sector-specific distributional response functions that map climate perturbations into reallocations of probability mass across the growth distribution. We find that temperature shocks systematically shift mass toward adverse growth realizations, with effects concentrated in the lower tail and strongest in agriculture, intermediate in manufacturing, and attenuated in services. Decomposing shocks further shows that sustained global warming induces broad-based distributional degradation, whereas episodic El Niño events primarily amplify downside tail risk, highlighting the limitations of mean-based damage assessments.