Application of Spatial Auto-Regressive Model for Determining Urban Land Market
Rapid conversion of agricultural, flood flow and sub-flood flow lands into urban uses at the periphery of Dhaka city increases the demand for developing useful supply, demand, and pricing models. Nevertheless, pricing land offers a challenging task. During high economic growth time, the conventional belief was that land values should soar with increasing demand and decreasing vacancies. Conversely, accurate anticipation of these values is problematic since spatial factors are often influential ones along with other characteristics of land. Moreover, land represents not only a bundle of physical characteristics but also sets of location-specific, transport, urban amenity, socio-economic and environmental characteristics. The extent of such characteristics on land market can be estimated by applying Spatial Autoregression model. The objective of this paper is to carry out an empirical analysis to examine the effects of different land characteristics on land prices in Savar municipality by using Spatial Auto-Regression (SAR) model. In particular, spatial econometric procedures (spatial autocorrelation) and spatial auto-regression analysis are applied to determine spatial dependence among the land value of plots and to evaluate the impact of land characteristics on land prices in the land market of Savar municipality respectively. The findings of this paper reveal that transport attributes have most significiant relationship with the land value and other explanatory variables like municipal amenities, physical characteristics of the plot, employment attraction, and commercial services are also influential factors of land price.
Journal of Bangladesh Institute of Planners Vol. 2, December 2009, pp. 107-115