Siting Versus Capitalization of Disamenity and Amenity Facilities in Housing and Land Prices Using Integrated Multi-Source Spatial Data in South Korea

Disamenity and amenity facilities pull property values in opposite directions, yet prior studies have treated disamenities as a homogeneous category and have been constrained by data scattered across agencies and by mismatched spatial scales. This study integrates multi-source spatial data, comprising 3426 sub-districts (eup-myeon-dong) nationwide, a 500 m population grid of 376,668 cells, and the actual point locations of individual facilities, into a single spatial framework and applies a multi-method design that combines a spatial Durbin model (SDM), explainable machine learning (XGBoost–SHAP), an environmental-justice analysis, and a micro-siting analysis. Four findings emerge. First, not all disamenities are associated with lower prices: only pollution-emitting facilities with a clear emission signature, namely, sewage treatment plants and incinerators, capitalize robustly as cross-sectional associations, into both housing prices and officially assessed land values, with the effect extending into neighboring sub-districts rather than staying purely local. The other disamenities show no robust total effect on either price; their estimates are small or unstable across specifications. Second, capitalization of pollution-emitting facilities is pronounced in the urban sample and disappears in the rural one, and the coefficient structure differs overall between urban and rural areas (global Chow test, p < 10−21). Third, micro-siting analysis shows that where a facility is placed (siting) and what it leaves behind in prices (capitalization) are distinct layers, and that capitalization tracks the population a facility keeps nearby rather than its nuisance type: pollution-emitting facilities retain enough nearby housing for their burden to register, whereas funerary facilities sited far from people do not. Fourth, the environmental inequity of pollution exposure runs along income rather than age and concentrates in cities rather than in rural areas: the exposure rate among residents of low-income urban sub-districts (49%) is 2.5 times that of the urban population as a whole (19%), while the greater Seoul metropolitan area remains comparatively insulated. These layered findings were possible only once the dispersed spatial data had been integrated and standardized, and the study demonstrates that fusing and efficiently managing spatial data is a precondition for evidence-based land-use and environmental policy.