Decomposing Regional Health Disparities Under Heatwave Exposure: Evidence from South Korea

Abstract Heatwaves are increasingly linked to health risks, yet evidence remains limited on how heat exposure relates to health disparities and whether patterns differ between subjective and objective health outcomes. This study compared individual health outcomes among the residents of high- and low-heat exposure regions in South Korea in 2010, 2015, and 2019. Regions were classified each year into high- and low-heat exposure groups using k-means clustering based on heatwave and tropical night counts, avoiding arbitrary absolute thresholds and reflecting year-specific exposure distributions. Using individual-level data, we estimated logistic regression models for subjective and objective health indicators, controlling for demographic and socioeconomic covariates, including urban status. We then used Blinder-Oaxaca decomposition to quantify the exposure-group health gap and partition it into an explained component and an unexplained component. The results indicate that health disparities existed between high- and low-exposure regions, but the magnitude and direction varied by health indicator. Moreover, a meaningful share of the gap was associated with differences in observed individual characteristics. Overall, the findings highlight the value of jointly examining subjective and objective health and provide a structured account of spatial health disparities under relative heat exposure.