A multilevel behavioural and structural determinants of energy consumption in student housing in Ghana

Purpose Conventional energy simulations often overlook the specific behavioural and building-related factors. In Ghana, where electricity reliability is limited and environmental values primarily influence student decisions, such models lack predictive reliability. This study assesses the behavioural and structural determinants of energy consumption in Ghana. Design/methodology/approach This research employs a multilevel linear regression model to investigate the behavioural, structural and institutional factors influencing electricity use across student residences in Ghana. Drawing from eight universities and 8050 student residence, the study employs a three-tiered analytical model based on the theory of planned behaviour, incorporating data from behavioural surveys, structural audits and institutional policy records. Findings The model accounts for 62% of the observed variation in electricity use. Structural and contextual determinants, including tariff responsiveness, occupancy density and policy implementation, exert stronger influence than general conservation attitudes. Behavioural variables, such as peak-hour presence and device use, show a significant correlation with consumption. Practical implications Results support institutional approaches that focus on adapting behavioural routines, optimizing infrastructure and providing real-time energy usage feedback. In addition, effective energy reduction can be achieved through both physical retrofits and strategic enforcement of behavioural and occupancy norms. Originality/value This study presents a culturally adapted simulation model for Sub-Saharan higher education settings, addressing the global bias in conventional energy frameworks and offering a transferable method for behavioural integration.