When pandemics meet climate risk: An agent-based model of non-linear macroeconomic dynamics under compound stress
构建了包含流行病和气候风险的基于主体模型，模拟复合风险下宏观经济、金融和环境的非线性动态，发现复合风险产生非加性、路径依赖效应，对政策制定者评估综合风险有参考价值。
This paper examines how pandemic and physical climate risks can co-evolve endogenously and jointly shape macroeconomic, financial, and environmental outcomes. It introduces CliMaPan Lab, an agent-based framework that incorporates epidemiological dynamics and a channel linking emissions, atmospheric concentrations, and temperature within a macroeconomic-financial environment featuring heterogeneous households and firms, as well as adaptive behavior. The framework is designed to capture non-linear propagation and state dependence under alternative public-health interventions. Simulation results show that stronger contact suppression is associated with faster infection control, but also with larger contemporaneous contractions in GDP and consumption, alongside weaker credit activity. Vaccination-centered regimes, and to a lesser extent social distancing, are associated with fewer severe outcomes and comparatively smoother macroeconomic and credit trajectories when protection accumulates sufficiently early. When physical climate risk is activated, cumulative damages weaken economic performance and increase persistence, with larger disruptions under compound scenarios where pandemic-related labor-market frictions interact with climate-related production losses. Emissions decline during pandemic disruptions but rebound as activity normalizes. Under compound risk, more persistent emissions reductions primarily reflect prolonged economic weakness rather than an endogenous decarbonization transition. Overall, the findings indicate that compound risks can generate non-additive, path-dependent dynamics that are not captured by single-risk analyses.