Linking discrete choice to continuous demand within the framework of a computable general equilibrium model

Discrete choice (DC) models are commonly used as basic building blocks in ‘bottom-up’ models which seek to describe consumer and producer behaviour at a disaggregate level, in contrast to continuous demand (CD) models which are used to describe behaviour at a more aggregate level. At a disaggregate level, choice behaviour is defined in terms of commodities differentiated by qualities or attributes. In contrast, aggregate demand behaviour is defined in terms of broadly defined and generically different commodities. In a DC model, the main focus of analysis is not the total quantity of demand, but rather the relative shares or substitution between the choice alternatives, in contrast to a continuous demand model where the focus is on the aggregate substitution between groups of commodities as well as on the income effects. Seen in this way, there is scope for complementary usage of DC and CD models within the framework of a CGE model where DC models are used to describe the preferences for a narrowly defined set of commodities belonging to a particular sector of an economy, and CD models are used to describe the interactions between these sectors. In this paper, we describe how DC and CD models can be used in such an integrated fashion in a spatial computable general equilibrium model to inquire into the wider economic impacts of a transport investment project in the Sydney Metropolitan Area.

► A methodology for integrating discrete choice (DC) models with continuous demand (CD) models in a CGE framework.
► DC models are used to determine the heterogeneous preferences of consumers for quality-differentiated product.
► CD models are used to link the aggregate demand for different groups of commodities together in the CGE model.
► Integrated models are used to estimate conventional users benefits and wider economic impacts.