Scope Effects in Contingent Valuation: Does the Assumed Statistical Distribution of WTP Matter?
本文通过瑞士森林保护区案例，检验支付意愿的统计分布假设对范围效应检测的影响，发现分布假设影响结果，建议在样本量小时使用非参数或尖峰模型。
Economic theory assumes that willingness to pay (WTP) increases with the quantity of the consumed good. This implies that there should be a scope effect in contingent valuation studies. However, in previous issues of Ecological Economics, several authors criticized the contingent valuation (CV) method for the absence of such effect or its inadequacy. In this paper, we contribute to this ongoing debate by proposing to systematically apply several WTP statistical distribution assumptions to test for scope effects and check its plausibility, following Whitehead’s (2016) recent recommendations. We perform this approach using data from a Swiss case study assessing the WTP for an increased surface of forest reserves. We find that both mean WTP and scope effects are sensitive to the statistical distribution assumption. Regarding plausibility, scope elasticities provide mixed result and also depend on the assumed statistical distribution of WTP. For small sample size CV studies, a non-parametric analysis, a spike model or an open-ended format can thus be better suited to reveal scope effects than the classical parametric dichotomous choice analysis. We thus recommend to systematically apply several statistical distribution assumptions of WTP to test for scope effects and their plausibility.