Essays on Climate Governance in the Global South: State Capacity, Just Energy Transitions, and AI Applications

Abstract

This thesis investigates the challenges and opportunities of climate governance in the Global South. The first essay investigates how climate change shapes citizens’ tax compliance across Africa by combining geocoded public opinion surveys with satellite remote sensing data. The findings show that exposure to weather extremes increases tax noncompliance by reducing citizens’ capacity to pay and weakening institutional legitimacy, underscoring how climate change can undermine state capacity. However, strong climate governance, including adaptation and mitigation efforts, can mitigate these adverse effects. The second essay examines the potential of electrification powered by renewable energy to drive socioeconomic transformation in the Global South, with a particular focus on women’s empowerment. Drawing on subnational panel data from 447 regions in 52 developing countries, complemented by an individual-level panel case study from Malawi, the analysis shows that renewable energy–sourced electrification promotes shifts in gender roles and improves women’s well-being. These findings suggest that investment in renewable energy and electrification can simultaneously support climate mitigation and drive structural transformation in the Global South. The third essay evaluates the potential of AI in climate governance, focusing on the application of Large Language Models (LLMs) to identify low-carbon transition risks. By comparing LLM outputs with over 500 municipal transition risk survey responses, we find that LLMs can capture the broad thematic structure of locally identified risks. However, they also exhibit limitations, including representational biases and a tendency to flatten contextual variation. These findings highlight the potential of LLMs to support risk identification in low-resource regions, while emphasizing the importance of human–machine collaboration. By integrating interdisciplinary insights from climate science, social science, and computer science, the thesis provides empirically grounded evidence to inform the design of climate governance frameworks that are effective, equitable, and responsible to the realities of the Global South.

Date of Award	2025
Original language	English
Awarding Institution	The Hong Kong University of Science and Technology
Supervisor	Lin YANG (Supervisor) & Jin WANG (Supervisor)