Improving drinking water quality in S. Korea: A choice experiment with hypothetical bias treatments

Increased pollution leads to a constant decrease of drinking water quality worldwide. Due to safety concerns, unpleasant taste and odour only about 3% of the population in S. Korea is drinking untreated tap water. The present study uses choice experiments and an extensive cost-benefit analysis (CBA) to investigate the feasibility of installing two advanced water treatments in Cheongju waterworks in S. Korea. The waterworks is situated in the middle of the country and is providing more than half a million people with drinking water. 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 that they would 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. This is innovative and important as hypothetical bias impedes the reliability of survey results. The lower bound of the median WTP for installing a new advanced water treatment system is about $2 US/month, which is similar to the average expenditures for bottled water per household in S. 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. Scenarios under which the instalment of the advanced water treatments is feasible are discussed together with environmental solutions in the long-run.