2020 Volume 49 Issue 2 Pages 47-69
In this study, we analyze the spatial models to fit the annually recorded Tokyo official land prices data from 1997 to 2018. For the covariates in the spatial models, we use 13 explanatory variables, which include indexes for access to central Tokyo, walking distances to nearest stations, building coverage ratios, acreage, and so on. A summary of findings is as follows. First, the land prices become lower as further away from a central area of Tokyo year by year. Second, the obtained generalized least squares (GLS) estimates together with its time series variations reveals us that the land price differences become larger between west and east side of Tokyo. Finally, the nugget effects of the residuals in spatial models are becoming larger and larger for non-central region, whereas those decreasing in central area of Tokyo. This can be explained by the individual factors, for example, daily life noises of a local area, daylight conditions, and so on, that cause spatial variation in land prices to become large. As a result, we found that the bipolarization of land prices in Tokyo is getting larger and larger.
