Capturing Preference Heterogeneity in Stated Choice Models: A Random Parameter Logit Model of the Demand for GM Food by

Analyses of data from random utility models of choice data have typically used fixed parameter representations, with consumer heterogeneity introduced by including factors such as the age, gender etc of the respondent. However, there is a class of models that assume that the underlying parameters of the estimated model (and hence preferences) are different for each individual within the sample, and that choices can be explained by identifying the parameters of the distribution from which they are drawn. Such a random parameter model is applied to stated choice data from the UK, and the results compared with standard fixed parameter models. The results provide new evidence of preferences for various aspects of the UK food system, particularly in relation to different forms of GM food but other environmental and technical aspects also. The results are also discussed in terms of their implications for further development and refinement of random parameter models.