Integration of spatial, labour and demographic heterogeneity in a CGE to model the distributional impacts from a disaster

The United Nations highlights the need to protect underrepresented and vulnerable groups from disasters. Computable General Equilibrium (CGE) models commonly assess disaster impacts but typically aggregate agents, limiting the identification of who bears the greatest burdens. Leveraging advances in spatial hazard modelling and synthetic microdata, we integrate spatial and demographic heterogeneity into a multi-regional, recursive-dynamic CGE framework to evaluate disaster impacts on representative household types. We estimate sectoral damage using hazard layers and a business operability model, and we extend the model with a synthetic population to capture differentiated household effects. Using a modeled eruption of Mount Taranaki (New Zealand) as a case study, we find a sharp regional GDP decline in the eruption year, with manufacturing, food, and construction in Stratford, South Taranaki, and New Plymouth most severely affected. Reconstruction drives labour inflows (notably professionals and administrators) that restore output quickly, but household welfare remains persistently suppressed because a putty–clay capital specification renders much of the capital stock unproductive and collapses capital income. Households reliant on capital income—one-person households (largely retirees) and couples with children—suffer the largest and longest-lasting welfare losses; single-parent households are primarily affected by eruption-year price rises. These results underscore a critical divergence between economic activity and household well-being: GDP recovery can mask persistent welfare declines when shocks undermine the capital income base of vulnerable households. We conclude that resilience strategies must target household income losses (including capital-income compensation and strengthened safety nets) and adopt labour-market interventions informed by household heterogeneity and interregional spillovers. • Spatial microdata expand SAM’s spatiodemographic heterogeneity. • Multi-regional CGE simulates volcanic impacts on diverse households. • Eruption causes sharp regional GDP drop despite swift output recovery. • Impacts concentrate on specific occupations, sectors and districts. • Capital-income households suffer largest, persistent welfare losses.