Individual Versus Intra-household Decision Making: Who Really Wears the Trousers?
S Hess, N Beharry-Borg, V Gibson, D Campbell, ITS University of Leeds, UK
The field of choice modelling has evolved dramatically over recent years, with numerous developments that aim to realign modelled behaviour with real world behaviour.
The field of choice modelling has evolved dramatically over recent years, with numerous developments that aim to realign modelled behaviour with real world behaviour. Examples include the representation of variations across respondents in their sensitivities as well as their choice processes, the incorporation of latent attitudes, and the recognition that many real life decisions are made not by a single person, but in consultation with other actors. It is these last two developments that this paper is concerned with.
Studies for joint decision making have typically been based on studies where individual and joint choices are modelled at the same time, with the joint choices being a driven by a weighted average of the sensitivities of the individual decision makers. The hypothesis put forward in this paper is that just as with individual choice processes, joint decisions are similarly driven in part by unobserved attitudes. Different possibilities may arise. The attitudes of the different decision makers may all play a role, or the attitudes of one decision maker may be dominant. Similarly, one decision maker may know the attitudes of another decision maker and either try to take them into account, or act against them.
The theoretical part of the paper thus presents a framework for the joint modelling of latent attitudes and decision processes within the context of a multi-agent decision environment. The paper then includes an empirical application making use of stated choice data looking at the choice between different beach sites for marine recreation in Tobago. Each beach differed based on nine attributes representing coastal water quality improvements. Individual as well as joint choices were collected. Additionally, as part of this questionnaire, attitudinal data on payment for improvements was collected from each member of the couple. This variable was used to represent the degree of opposition to paying for these improvements. Results suggest that significant improvements in model fit can be obtained through accounting for this latent variable, also leading to important changes in the implied willingness to pay measures. Finally, our modelling approach also reveals the interesting dynamics which can arise during the joint decision-making process of a couple.
Association for European Transport