Spatial process for housing prices in Seoul using spatiotemporal local G statistics

The housing market in Korea was unstable for a period of 5 years from 2003 to 2007, during which time housing prices in certain areas greatly increased in comparison to housing prices in other areas in a relatively short period. The purpose of this study is to explore temporal trends in spatial patterns of housing prices in the housing market in Seoul, Korea. We utilized apartment locations based on monthly housing prices for the sub-administration areas (dongs) of the 25 local governments (gu) in Seoul from January 2004 to December 2007, and applied spatiotemporal local G statistics. The major findings of this study are as follows. First, housing prices are highly spatially and temporally correlated in certain areas, such as Gangnam and the new towns, and housing price hotspots are sufficiently detectable in terms of spatiotemporal autocorrelation. Secondly, government housing policies affect the spatiotemporal patterns of housing prices. These results indicate that we are able to monitor spatiotemporal patterns of housing prices in a housing market, and use this approach to effectively support the decisions of housing policy makers.