Negative externalities of long-term vacant homes: Evidence from Japan

Abstract

Employing parcel-level data on vacant houses in a depopulating city in the Tokyo metropolitan area, we provide the evidence of negative externalities of long-term vacant houses that persist for several years. Ownership of a vacant house continues, whereas the level of maintenance declines; thus, the Japanese context enables us to simply capture the long-run impact of neglect, separate from the foreclosure stigma in the US context. We find that the externality spills over to approximately 50 m and that one additional long-term vacant house within 50 m pushes property transaction prices down by approximately 3%. We argue that the externalities come at least partly from the disamenity channel, since they are not observed for vacancies that disappear within one or two years. Externalities are observed from three years after the houses become vacant, which are, often not put on the housing market, and the externalities diminish around the time the vacancies finally disappear. We also show that externalities exist only in areas where nearby long-term vacant houses are not yet common, as their existence stands out in those areas. These results imply that reducing the number of long-term vacant houses will help mitigate disamenities in areas that have not yet experienced severe declines.