The Contingent Valuation Method (CVM) is widely used to estimate the value of non-market goods. As a stated preference technique, CVM operates through a constructed hypothetical market that enables individuals to directly express their preferences for a particular good. However, the reliance on a hypothetical market has led to criticism that CVM estimates may be affected by various biases. Despite these concerns, the method continues to occupy a prominent place in valuation literature, largely because of its unique ability to capture both use and non-use values. Against this backdrop, the present study attempts to examine the validity of CVM by testing for two important criticisms commonly associated with it, i.e., information bias and free-riding behaviour in the context of a valuation programme for Dibru Saikhowa National Park in Assam. In the presence of such biases, the true willingness to pay (WTP) may not be revealed and could either be overestimated or underestimated, leading to potentially misleading policy decisions. The study also investigates whether these anomalies, if present, can be corrected through appropriate methodological precautions. Adopting a completely non-parametric approach, the analysis reveals that free-riding behaviour is altogether absent, while information bias, although initially observed, disappears once hypothetical bias is adequately addressed. The alternative tests performed in the study yield broadly consistent results, thereby strengthening the robustness of these findings. The results therefore suggest that it is not inevitable for WTP responses obtained through CVM to be influenced by information bias or free-riding behaviour. Rather, the adoption of appropriate corrective measures, both ex-ante and ex-post, at different stages of a CVM study can substantially reduce, if not entirely eliminate, such biases. The findings of the present study further indicate that correcting for hypothetical bias may serve as an important step in mitigating several other anomalies that are often attributed to the hypothetical nature of CVM.
Konwar et al. (Tue,) studied this question.