Examining impacts of transportation on residential property values using geographical information system: a hedonic price model approach

Abstract

Transportation system has a great impact on the land uses of an urban area. Many transportation researchers used hedonic price models to understand the relationship between transportation and land uses. Hedonic price model studies are commonly focused on the impact of transportation system on land uses, particularly on land values. This study was an attempt following hedonic price model approach to provide empirical evidence regarding the nature and magnitude of accessibility impacts of transportation system on asking rental price of residential properties. It also investigated and compared impacts of structural attributes and neighborhood attributes on residential property values with that of transportation system. Accessibility to transportation infrastructure, central business district, wholesale markets, shopping centers and primary schools were considered as the transportation attributes that might have significant impact on residential property values at Rajshahi City. Both linear and semi-log model specifications were estimated based on data collected by field visual inspection of advertisements for rent during May 2004. The study used Geographical Information System (GIS) as a tool for the determination of hedonic models. It was facilitated by a large GIS dataset collected by the Rajshahi Master Plan Project. The study revealed that the semi-log specification provides better results than simple linear hedonic models. Again, omission of transportation attributes in the hedonic price models caused four percent reduction in explaining the overall variability of the property values. In the study, stepwise regression technique was used to determine best-reduced model. Moran's J and Geary's c were estimated to identify the effects of spatial autocorrelation in the residuals of predicted property values. The final model explained 58.8 percent of the variability in the residential asking rental prices. Structural attributes are the most significant determinants for property value variations at Rajshahi. Number of bedrooms proved to be the highest contributor as a predictor with a standardized coefficient of 0.427. One of the most interesting findings was that increase in network distance to CBD from the residential properties increases property values in the study area, as the corresponding explanatory variable constituted a positive coefficient of 0.34. However, Accessibility effects of the transportation infrastructure, which was measured by the network distance from the properties to the major arterials, found to be negative coefficient indicating that the rent of house decreases with the increase of distance to the major roads. The estimated hedonic model suggested that moving 0.5-mile closer to the major arterials adds 585.858 Taka premiums, all else held equal, on residential asking rental price at Rajshahi City.