Simulating housing prices in Lyon with UrbanSim: predictive capacity and sensitivity analysis

Abstract

Housing prices in the Lyon Urban Area are simulated with the land use framework UrbanSim interacting with transportation model. We focus on the Real Estate Price Model of the UrbanSim framework. This OLS regression model of housing prices is calibrated using a nine-year back-casting period. The calibrated model, applied in simulation, provides price dynamics similar to actual one in the centre of Lyon. Sensitivity analysis demonstrates the model’s ability to capture changes in employment accessibility on price dynamics. We conclude that the calibrated residential Real Estate Price Model from the UrbanSim application in Lyon is sensitive to changes in accessibility and provides good predictive capacity in the city centre, but underestimates prices in other areas.