The Recreational Value of Coral Reefs: Classical and Bayesian Meta-Analytic Approaches to Benefit Transfer

Date of Completion

January 2010


Economics, Environmental|Economics, Agricultural|Recreation




This dissertation provides a robust assessment of value surfaces and the potential for benefit transfer (BT) in the international context of coral reef. Environmental values are considered as useful tools to improve management of these declining ecosystems. Nevertheless, primary valuation is often limited by difficulties in rigorous applications of valuation surveys, and budgetary and time restrictions. Acknowledging that, Spurgeon (2001) suggested BT as an alternative to primary valuation of coral reefs. Nine years later, Brander, Van Beukering and Cesar (2007) and this dissertation are, to the knowledge of the author, the only meta-analytic BT (MA-BT) studies in the context of coral reefs. ^ The dissertation seeks to provide a comprehensive assessment of the suitability of international MA-BT on the recreational value of reefs. The motivation for this assessment comes from an increasing demand for welfare measures at policy sites and from limitations in existing estimates. These limitations include both the sparseness of original studies and the presumed poor potential for BT on coral reefs (Brander et al., 2007). From a meta-analytic perspective, these limitations are related to a common issue: spare and highly heterogeneous data. This dissertation addresses this issue at two levels: the generation of new meta-data with improved commensurability of observations, and the implementation of a computational approach that allows tailoring of meta-regression models (MRMs) to more specific policy contexts. ^ First, the generation of a new meta-data imposes commodity and welfare consistency, and it also incorporates variables that improve physio-economic linkages in the modeling. Second, the implementation of both classical and Bayesian approaches allows for a comparison of the performance of alternative statistical methods and models in presence of small and heterogeneous samples. Bayesian methods add flexibility to: (i) diversify sources of information by incorporating prior distributions, and (ii) better address small-sample estimations. ^ This research provides valuable insight into the capacity of classical and Bayesian MRMs to capture heterogeneity of the data, explain the variation in willingness to pay, generate robust assessment of value surfaces, and enhance MRMs to improve the potential for BT, based on in-sample validity testing. Recommendations for further primary and meta-analytic research are also provided. ^