Improving Drinking Water Quality in South Korea: A Choice Experiment with Hypothetical Bias Treatments

The objective of this present study is to use choice experiments and an extensive cost-benefit analysis (CBA) to investigate the feasibility of installing two advanced water treatments in Cheongju waterworks in South Korea. The study uses latent class attribute non-attendance models in a choice experiment setting in order to estimate the benefits of the two water treatments. Moreover, it explores strategies to mitigate potential hypothetical bias as this has been the strongest criticism brought to stated preference methods to date. Hypothetical bias is the difference between what people state in a survey they would be willing to pay and what they would actually pay in a real situation. The study employs cheap talk with a budget constraint reminder and honesty priming with the latter showing more evidence of reducing potential hypothetical bias. The lower bound of the median WTP (willingness to pay) for installing a new advanced water treatment system is approximately $2 US/month, similar to the average expenditures for bottled water per household in South Korea. These lower bounds were found using bootstrapping and simulations. The CBA shows that one of the two treatments, granular activated carbon is more robust to sensitivity analyses, making this the recommendation of the study.