Finite Mixture Approach to Contingent Valuation Methods: An Application to the Mangrove Forest in the Gulf of Guayaquil, Ecuador

Abstract

There are exhaustive methods to model the Willingness-to-Pay for non-market goods and services, such as environmental services and amenities. By far the most popular is the contingent valuation method; however, failure to capture unobserved heterogeneity in the underlying distribution of survey responses can result in biased estimates of the value individuals place on nature conservation or policy interventions. In this paper we use a latent class approach to obtain a Type-Consistent measure of the Willingness-to-Pay, by means of a finite mixture of distributions. We test the model against traditional methods in applied environmental valuation and show statistical advantages. In our application, we concentrate on the mangrove forest in Ecuador which has been under heavy pressure due to the development of the aquaculture industry and urban expansion. Our results show that allowing for heterogeneous preferences, through a two-types finite mixture, reduces potential bias; while enabling differentiated Willingness-to-Pay values suitable for policy purposes.