Heterogeneous Preferences, Perceived Quality And Drinking Water Choices In France: A Scaled Multinomial Logit Approach

We investigate the determinants of individual drinking water choices in France. Among these determinants, we focus on the perceived quality of tap water measured by the stated satisfaction level of individuals. As this variable raises a potential endogeneity issue, we propose a novel methodology, which combines the estimation of a scaled multinomial Logit model with the two-stage residual inclusion method to address the endogeneity issue. Using a unique cross-section dataset of about 3,500 individuals, we show that the level of satisfaction is clearly a determinant of drinking water choices in France. We also show that addressing the endogeneity issue significantly changes the results regarding the sources of individuals' drinking water choices heterogeneity.