Varying levels of information and the embedding problem in contingent valuation: the case of Canadian wilderness
This study examines the effect of providing varying amounts of information and embedding to contingent valuation respondents in the context of wilderness reservation. Contingent valuation is a technique developed to assess the monetary value of nonmarket goods by asking survey respondents how much they would be willing to pay (or accept) for an increase (or decrease) in the level of provision of such goods. Embedding refers to obtaining willingness-to-pay measures for a good when it is valued as part of a larger good. When subjects were provided with more complete information to make evaluations, their willingness-to-pay responses were reasonably consistent across levels of embedding. In contrast, when little background information was provided, willingness-to-pay amounts were inconsistent and showed significantly more variability and greater signs of embedding. In addition, low-information subjects volunteered more protest zeros and reported feeling slightly less confident of their responses. Perhaps the most interesting finding was the difference in mean valuations depending on how much background information subjects received. This issue of information provision can be a problematic challenge for contingent-valuation practitioners.