Statistical Analysis of the Relationship between Public Transport Accessibility and Flat Prices in Riga

A relationship between public transport accessibility and residential land value is a point of interest of many recent researches. A hedonic price regression model, widely used in this research area, has one very important shortcoming – it calculates an "average" influence of factors on land value in the analysing area. Usually spatial effects present in data, and the influence of public transport accessibility can be distributed over the area non-uniformly. In this study we apply a comparatively new modification of the regression model – geographically weighted regression – to examine the relationship between public transport accessibility and residential land value (in a form of rent and sell prices) in Riga. The proposed method allows taking into account spatial effects present in the relationship. We use information about geographical locations of urban public transport stops and routes to calculate a level of transport accessibility. Together with the transport accessibility level and a common set of property-specific parameters (floor area, number of rooms, etc.) we consider additional hedonic properties of a flat location such as distances to supermarkets, higher schools and natural attractors like large parks, the river, and the seaside.