Abstract: In urban context, urban nodes such as amenity or
hazard will certainly affect house price, while classic hedonic analysis
will employ distance variables measured from each urban nodes.
However, effects from distances to facilities on house prices generally
do not represent the true price of the property. Distance variables
measured on the same surface are suffering a problem called
multicollinearity, which is usually presented as magnitude variance
and mean value in regression, errors caused by instability. In this paper,
we provided a theoretical framework to identify and gather the data
with less bias, and also provided specific sampling method on locating
the sample region to avoid the spatial multicollinerity problem in three
distance variable’s case.
Abstract: In urban area, several landmarks may affect housing
price and rents, and hedonic analysis should employ distance variables
corresponding to each landmarks. Unfortunately, the effects of
distances to landmarks on housing prices are generally not consistent
with the true price. These distance variables may cause magnitude
error in regression, pointing a problem of spatial multicollinearity. In
this paper, we provided some approaches for getting the samples with
less bias and method on locating the specific sampling area to avoid
the multicollinerity problem in two specific landmarks case.
Abstract: The main purpose of this research is the calculation of implicit prices of the environmental level of air quality in the city of Moscow on the basis of housing property prices. The database used contains records of approximately 20 thousand apartments and has been provided by a leading real estate agency operating in Russia. The explanatory variables include physical characteristics of the houses, environmental (industry emissions), neighbourhood sociodemographic and geographic data: GPS coordinates of each house. The hedonic regression results for ecological variables show «negative» prices while increasing the level of air contamination from such substances as carbon monoxide, nitrogen dioxide, sulphur dioxide, and particles (CO, NO2, SO2, TSP). The marginal willingness to pay for higher environmental quality is presented for linear and log-log models.