Energy poverty’s association with health: Dynamics and persistence patterns in spain

Does energy poverty make a person more likely to report poor health? Using Spanish panel data for 2016–2021 and a dynamic random-effects probit with explicit treatment of initial conditions, we examine associations between income, housing, education, several measures of energy poverty, and self-assessed health. Lower income, certain housing types, and lower education increase the probability of poor health. Health shows strong persistence: the coefficient on lagged health lies between 0.30 and 0.45 and declines only modestly over time. Energy poverty is associated with worse health both contemporaneously and dynamically. For a residual-income measure, the coefficient on the current indicator is about 0.092, while the one-period lag is 0.036–0.048, suggesting that 40–60 % of the contemporaneous association carries into the next year. Similar lagged effects emerge when energy poverty is defined relative to the national median energy-spending share, 0.11 to 0.22 or by a ten-percent-of-income threshold, 0.13 to 0.19. The pattern is stronger among respondents with poor or very poor health, where lagged associations are 1.3–1.5 times larger than in the full sample, indicating that energy poverty most severely affects those already in fragile health. This heterogeneity shows that individuals already in poor health experience a compounded vulnerability to energy deprivation, making them a priority target for policy intervention. Recognising these health-based disparities is essential for designing equitable energy and social support programmes. The persistence patterns documented here may inform policy prioritisation, while causal evaluation requires designs with exogenous variation.