Assessing Energy Poverty and Environmental Indicators in Sub‐Saharan Africa Leveraging a New Advanced Spatiotemporal Artificial Intelligence Regression Technique
Year: 2026
Type: article
Abstract: ABSTRACT Energy poverty, associated with SDG 7, remains pervasive in sub‐Saharan Africa (SSA) and is closely linked to sustainability (SDG 13), yet prior evidence is fragmented and often limited to a single indicator. This study re‐examines the energy poverty–environment nexus adopting three different environmental metrics and allowing impacts to vary across space and time. Using an SSA panel (1990–2023 and 30 countries), the Geographically Neural Network Weighted Regression (GNNWR) is applied to model load capacity factor (LCF), CO 2 emissions, and ecological footprint (EF) as functions of energy poverty (POV), GDP, urbanization, globalization, renewable energy use, and natural‐resource rents; robustness is assessed with Geographically and Temporally Neural Network Weighted Regression (GTNNWR). Diagnostic tests indicate a strong fit ( R 2 ≈0.81–0.93) and spatiotemporal non‐stationarity. POV is associated with higher CO 2 emissions, while it significantly reduces LCF, implying weaker ecological carrying capacity. Conversely, POV is linked to a lower EF, consistent with constrained access dampening consumption‐based pressures even as pollution and biocapacity dynamics worsen. Renewable energy consumption improves environmental quality by lowering EF and increasing LCF. GDP growth increases CO 2 and EF, whereas urbanization shows mixed effects (higher CO 2 and LCF but lower EF). Globalization raises CO 2 and EF but has a weak effect on LCF. Spatial diagnostics show that adverse POV effects cluster in Southern and Western SSA. GTNNWR confirms the main results. Policy should treat POV as an environmental risk and couple electrification and clean‐cooking access with accelerated deployment of renewables and region‐targeted measures where impacts are strongest.
Source: Sustainable Development
DOI: 10.1002/sd.71867
Language: English
Subfield: Economics and Econometrics
Domain: Social Sciences
Open Access status: bronze
OpenAlex ID: W7225348248
