More random or more deterministic choices? The effects of information on preferences for biodiversity conservation

For many years, stated preference researchers have been interested in the effects of information on willingness to pay for environmental goods. Within the random utility model, information about an environmental good might impact on preferences and on scale (error variance), both between and within samples of choices. In this paper, we extend the G-MNL model to investigate the effects of different information sets on choices over the management of biodiversity in the UK, looking specifically at moorlands managed for red grouse shooting. Specifically, we make the individual scale parameter a function of observable (dataset-specific) characteristics. Our results show that changing information sets results in significant differences in the mean scale between datasets, and in the variance of scale. Respondents are more deterministic in their choices and show lower within-sample scale heterogeneity in the alternative information treatment. Changes in information provision also effect willingness to pay estimates, reducing the value people place on the conservation of two iconic birds of prey. The methods used will also be of interest to researchers who need to combine choice experiment data sets.