The model was estimated using a simulated maximum likelihood estimator based on 500 Halton draws. Due to scale heterogeneity, comparing coefficient estimates across models of tourists and residents is not appropriate. Comparison between the marginal values of the attributes, which we interpret as the stated WTP to obtain an attribute level, was calculated by dividing the attribute level’s coefficient estimate by the payment vehicle’s coefficient estimate (the marginal utility of income), which we assumed to be constant across all respondents. This calculation removed the impact of the scale factor associated with a logit-type model.