Author(s)

B. Smith & D. Olaru

Abstract

A survey of factors affecting residual relocation choices for families who had recently moved onto the Mandurah Railway corridor (Perth, Western Australia) has been undertaken using stated preference methods. While stated preference models are appropriate tools for valuating attributes of choice, there remains the concern of realism in an experimentally controlled environment. In this paper we validate stated preference models using hedonic pricing regressions based on observed real estate prices for three Perth-Mandurah Railway precincts. The results of the regressions are compared to those of the discrete choice models by examining the levels of significance of housing or neighbourhood characteristics in each modelling paradigm, as well as a comparison of the closeness of the valuations for these characteristics. The findings indicate that factors affecting residential location are consistent whether observed by experimental data or by revealed choices in the market. From a planning perspective, the results indicate that not only housing features, but also neighbourhood characteristics such as proximity to public transport hubs or local schools affect residential property values. Keywords: discrete choice modelling, hedonic regressions, housing attribute valuation, transit-oriented development. 1 Introduction This research investigates similarities and differences between hedonic pricing and discrete choice modelling for housing valuation along transit-oriented

Keywords

discrete choice modelling, hedonic regressions, housing attribute valuation, transit-oriented development