Essays on Climate Variability, Energy Deprivation, and Adaptive Capacity to Climate Change

Unofficial AI-generated summary based on the public title and abstract. Not an official translation.

🔬研究者:This paper provides robust empirical evidence on the climate-energy-poverty nexus using panel econometric methods, useful for further academic research on climate impacts in developing countries.

🏢実務担当者:Corporate sustainability teams in developing country contexts can use these findings to understand energy access risks and inform supply chain resilience strategies.

🏛政策担当者:Policymakers in LMICs should consider climate impacts on energy poverty and deforestation when designing adaptation policies, while international donors can leverage these insights for targeted climate finance.

Changing weather patterns have made the last decade the warmest on record, with increasingly frequent wildfires, droughts, floods, and storms threatening lives and livelihoods worldwide (United Nations, 2023). These climate-related risks are especially severe in low- and middle-income countries (LMICs), where limited infrastructure, weaker institutions, and constrained adaptive capacity heighten vulnerability. A growing literature shows that climate variability can slow economic growth, increase mortality and morbidity, accelerate migration, and contribute to political instability (Burke et al., 2015; Dell et al., 2014). An equally pressing concern lies in the impact on energy systems: higher temperatures surge the demand for energy, while climate shocks reduce household incomes and agricultural productivity, weakening the affordability and reliability of modern energy access. This raises fundamental questions about whether and how climate variability deepens energy poverty. Beyond this form of deprivation, climate change also shapes disparities in access to energy within countries. Urban areas typically benefit from more reliable energy infrastructure, whereas rural households often depend on traditional fuels, placing greater pressure on natural resources – particularly forests – as biomass becomes a substitute for modern energy. At the same time, the consequences of climate stress extend beyond material hardship to social and psychological wellbeing. As environmental instability erodes economic security and heightens livelihood risk, especially where adaptation systems are limited, concerns have intensified about its effects on public health, including suicide mortality. Against this backdrop, my thesis examines these interconnected yet underexplored issues across LMICs in Africa, Asia, and Latin America and the Caribbean. This thesis comprises three empirical essays on climate variability, energy deprivation, and adaptive capacity to climate change in LMICs. Each chapter is self-contained and examines questions specific to its topic. In particular, the first chapter assesses whether short-term weather shocks and longer-term climate change influence energy poverty, and how these climatic conditions operate through different pathways. Energy poverty is measured along four dimensions: energy availability, energy affordability, energy efficiency, and environmental health. The second chapter investigates the impact of energy inequality on deforestation, focusing on urban-rural disparities in access to modern energy as a determinant of environmental degradation. The third chapter turns to public health by analysing whether adaptive capacity reduces suicide mortality and how this relationship may operate through mediating mechanisms. A detailed summary of the findings from each chapter is provided below. Chapter 1 constructs a novel energy poverty index using the entropy method to empirically assess the impacts of short-run weather shocks and longer-run climate changes in 107 LMICs from 1991 to 2020. I primarily employ a traditional fixed effects approach and then extend the analysis using a panel error correction model. Driscoll-Kraay standard errors are used to obtain robust inference that accounts for cross-sectional dependence, while the error correction specification addresses potential non-stationarity and cointegration when evaluating the effects of short-run weather shocks on energy poverty. Additionally, I utilize a long-difference model within a cross-sectional framework to investigate the consequences of longer-run climate change on energy poverty. I find robust evidence that higher temperatures cause an increase in energy poverty in both the short and longer run, while precipitation shows no significant effects. LMICs in Africa and Asia are more obviously affected by changing climate conditions. Longer-run climate change significantly intensifies energy poverty across all income and agricultural categories, with a more pronounced impact observed in low-income and agrarian countries. Furthermore, compelling evidence suggests that longer-run climate change reduces crop yields, thereby exacerbating energy poverty in LMICs. Chapter 2 provides new empirical insights into the impact of within-country urban-rural energy inequality as a significant contributor to deforestation in LMICs. While deforestation is frequently linked to agricultural expansion and resource extraction, the effect of unequal access to modern energy between urban and rural areas remains largely unexplored. By constructing a composite measure of energy inequality using the Theil index – based on access to electricity and clean cooking fuels – this study examines its effect on deforestation across 105 LMICs from 1991 to 2022. I use heterogeneous panel data estimators, with a focus on explicitly accounting for unobserved common factors, non-stationarity, and cointegrated relationships. The results reveal a quantitatively and statistically significant positive relationship between urban-rural energy inequality and deforestation, suggesting that limited access to modern energy in rural areas drives greater reliance on forests for fuel, thus accelerating deforestation. Extensive robustness checks further support these findings. Additionally, heterogeneity analysis shows that the effect is particularly strong in Africa and among lower-middle and low-income countries, where weaker institutions and regulatory capacities may intensify this impact. Chapter 3 examines whether and how adaptive capacity to climate change affects suicide mortality in LMICs. Using annual panel data for 112 countries from 1995 to 2023 and a long-differenced local projections approach, I estimate impulse response functions of suicide mortality in response to adaptive capacity. The findings provide new empirical evidence that adaptive capacity significantly reduces suicide mortality, with the effect materializing after two years and persisting through the medium term. Extensive robustness analyses, including extended time horizons, alternative lag and lead structures, different measures of the key explanatory variable, additional potential confounders, the inclusion of historical and geographic controls, the exclusion of outliers, and an alternative empirical specification, support these findings. Heterogeneity analysis indicates that this negative effect is especially pronounced for males, older age groups, lower-middle- and low-income countries, stable or transitioned democracies, and countries with lower conflict intensity. Mechanism analysis further shows that the relationship operates at least partly through the poverty channel, as adaptive capacity mitigates climate-induced losses in income, assets, and employment, thereby lowering suicide mortality. The use of different econometric approaches across the three chapters reflects differences in data structure, identification requirements, and the specific nature of the research questions. In Chapter 1, limited within-country variation in climate variables, together with standard practice in the literature, motivates the use of a fixed effects model. To additionally capture long-run equilibrium relationships and adjustment dynamics, an error correction framework is employed. Longer-term climate change effects are further examined using a long-difference specification in a cross-sectional setting. In Chapter 2, sufficient cross-sectional and temporal variation in the data motivates the use of heterogeneous panel estimators that allow for slope heterogeneity and unobserved common factors. In Chapter 3, panel local projections are employed to flexibly trace dynamic responses of suicide mortality to adaptive capacity without imposing restrictive functional form assumptions, making them well suited for analysing medium-term adjustment paths. Overall, the three approaches are complementary and aligned with the empirical structure and objectives of each chapter.

• openalex https://doi.org/10.82348/our-archive.00199first seen 2026-07-10 05:04:22 · last seen 2026-07-10 05:28:37