Based on choice experiment (CE), evaluating the public's heterogeneous preferences and willingness to pay (WTP) for air pollution treatment policies can provide useful social views for the reasonable formulation of treatment schemes. However, the application of CE contains an implicit assumption that respondents understand their real preferences and can make choices with complete certainty. In reality, for a variety of reasons, not all respondents are absolutely certain about their responses, this assumption distinctly is hard to be consistent with reality. To explore the impact of respondent uncertainty on the public's WTP and heterogeneous preferences for air pollution treatment policies, this study introduces the critical point and exponential weighting methods to deal with this uncertainty in the context of CE and conducts comparative analysis based on the random parameters logit (RPL) and latent classes models (LCM). The results show that, ignoring uncertainty leads to distortions in the public's WTP and preference characteristics. In the RPL models, on average, the WTP for attributes is overstated by 32.10%. Our results also reveal that, whether to consider uncertainty does not affect the ranking of the implicit prices of these attributes. After incorporating uncertainty into the analysis, respondents were divided into two potential groups with different preferences, namely the environment-focused group (79.44%) and the price-focused group (20.56%), which is quite distinct from research results of ignoring the uncertainty. Contribution of this study is not only to provide theoretical insights for exploring the effects of uncertainty on public preferences based on CE, but also to provide valuable guidance for policy makers to formulate more accurate and effective treatment measures.
Read full abstract