This study provides information for water planners in their decision-making process for allocation of water among competing uses. In this case the benefit of interest is a non-market good, water left flowing in a river, and its value to white water boaters. A Contingent Valuation survey is used to elicit respondents' willingness to pay for four alternative flow regimes. In order to take into consideration possible within-individual correlation, a random effects probit model is used to estimate benefits and marginal impacts of changes in flow levels based on the repeated dichotomous-choice responses. An alternative non-parametric approach is also used to analyze the dichotomous choice responses and a comparison is made between the two methods. Finally, estimates based on responses to the open-ended valuation questions are compared to those based on the dichotomous choice responses and differences are explored. |