Purpose
This paper aims to develop a house price forecasting model to investigate the impact of neighbourhood effect on property value.
Design/methodology/approach
Multi-level modelling (MLM) method is used to develop the house price forecasting models. The neighbourhood effects, that is, socio-economic conditions that exist in various locations, are included in this study. Data from the local government area in Greater Sydney, Australia, has been collected to test the developed model.
Findings
Results show that the multi-level models can account for the neighbourhood effects and provide accurate forecasting results.
Research limitations/implications
It is believed that the impacts on specific households may be different because of the price differences in various geographic areas. The “neighbourhood” is an important consideration in housing purchase decisions.
Practical implications
While increasing housing supply provisions to match the housing demand, governments may consider improving the quality of neighbourhood conditions such as transportation, surrounding environment and public space security.
Originality/value
The demand and supply of housing in different locations have not behaved uniformly over time, that is, they demonstrate spatial heterogeneity. The use of MLM extends the standard hedonic model to incorporate physical characteristics and socio-economic variables to estimate dwelling prices.