This thesis investigates the relationship between demand functions and discrete choice models, the specific objective being to infer ordinary demand elasticities from estimated choice elasticities. A necessary intermediate step is to uncover the relationship between conditional demand systems whereby non-transport goods are pre-allocated and conditional share systems whereby transport goods are pre-allocated. A conditional share system is applicable to the demands for a set of competing alternatives where there is a constraint on the consumption quantity for the aggregate group demand. This describes the demand implications for a population of individuals making discrete choices. The inference is appropriate to a stand alone revealed and stated preference choice model, where the aggregated choice elasticities are used to extrapolate to market share elasticities. The missing component for measuring demand responses to be used in a policy application is the generation elasticity: the marginal increase or decrease in the volume of transport activity in response to a pricing decision. The thesis develops a generation elasticity that preserves the theoretical properties of demand elasticities. The application of the developed theory is suitable for settings where the transport group elasticity is not measured by the survey instrument. Stand alone revealed and stated preference surveys only measure the switching behaviour between modes. The complete relationship under weak separability is developed for the case where elasticity of demand for the transport group is known. Otherwise to implement the theory this thesis makes use of the theoretical properties of block-additive utility functions. Frisch’s money flexibility term is interpreted as the average elasticity of substitution for broad groups of commodities, including transport.
|Qualification||Doctor of Philosophy|
|Publication status||Unpublished - 2011|