Combining stated and revealed choice research to inform energy system simulation models: the case of hybrid electric vehicles.

This study estimated dynamic behavioural parameters for an energy-economy model (called CIMS) using discrete choice modelling techniques, focusing on hybrid-electric vehicles (HEVs) . An online survey collected stated preference (SP) data from Canadian and Californian vehicle owners under different hypothetical market conditions. Revealed preference (RP) data was collected by eliciting the year, make and model of recent vehicle purchases. SP and RP data were combined in a ‘joint’ multinomial logit modelling technique, yielding choice models that were more realistic and useful than models estimated from SP or RP data alone. Dynamic behavioural parameters were estimated from the joint choice models and integrated into CIMS, significantly altering HEV adoption forecasts. Policy simulations with the improved model demonstrate the potential efficiencies of policies that induce technological change by reducing a technology’s non-financial costs (e.g. vehicle emissions standard), as opposed to targeting capital costs (e.g. subsidies) or fuel costs (e.g. gasoline tax).