Housing prices and price endogeneity in tenure and dwelling type choice models

• Develop a highly disaggregated model to predict dwelling prices.
• Build the dwelling prices into a joint dwelling type and tenure choice model.
• Test for price endogeneity, an issue hardly investigated in the literature on housing choices.
• Show the way in which these models are integrated into a transport and land use model system.

The application of a strategic transport–land use model usually requires dwelling price data for both the chosen and non-chosen alternatives. Previous studies have used average dwelling prices either externally sourced from Census and commercial data or internally generated by hedonic models. The use of average prices may be reasonable if the spatial resolution is fine enough to ensure the absence of price endogeneity – a well known issue but hardly investigated in the literature on housing choices. Using data collected from a revealed preference survey conducted in Sydney in 2013, this paper presents a hedonic price model to obtain price data for each dwelling as a function of dwelling and location characteristics. The data are then used as input into a tenure and dwelling type choice model and a test undertaken to investigate the possible presence of price endogeneity, which we show to not exist. Price elasticities of demand for different dwelling types are derived and the application of this model within a broader and very general framework of the new version of the ‘Transportation and Environment Strategy Impact Simulator’ (New-TRESIS) is discussed.