본문 바로가기
  • Home

A Bayesian Spatio-temporal Analysis of Land Price

  • Korea Real Estate Review
  • 2014, 24(4), pp.119-135
  • Publisher : korea real estate research institute
  • Research Area : Social Science > Law > Law of Special Parts > Law of Real Estate

Lee, Changro 1 Eum Young-seob 1 Park, Key Ho 1

1서울대학교

Accredited

ABSTRACT

Land price, one of spatial phenomena, is not generated independently, but affected by the price of nearby land. ‘the price of nearby land’ means the price of neighboring lots which were sold recently. Using government announced land price data accumulated for more than 20 years, this study examines the consequences of explicitly allowing for the presence of spatial correlation as well as temporal one. A hierarchical bayesian estimation method is utilized in order to deal with both the temporal and spatial aspects of the data, and a plausible implementation method to overcome the 'big-n' problem is also considered in the study. The results indicate that submarkets of commercial land in Seoul are delineated more narrowly than those of residential land. As for the spatial pattern of land price distribution, Gangnam was clearly depicted as the highest land price area, and several isolated areas whose land price levels were similar to that of Gangnam were also created distinctly in 1996. These isolated areas, however, seem to be incorporated into a broader market area now. In addition, residential land shows stronger temporal correlation than commercial one does, which is due to the fact that the price of commercial land is more sensitive to a general economic cycle rather than the previous-year price level.

Citation status

* References for papers published after 2023 are currently being built.