Willingness to pay for physician services: Comparing estimates from a discrete choice experiment and contingent valuation

Discrete choice experiments (DCE) and contingent valuation (CV) are often applied to value health care benefits. However, whether the two techniques yield converging willingness-to-pay (WTP) estimates is not studied well. This study aims to compare at a disaggregated level WTP estimates for physician services obtained from DCE and CV estimates. We study the consistency between the estimates and whether there are systematic differences between the two. The analysis is based on data from a household survey in Ukraine that includes 303 respondents and is taken to be representative of the Ukrainian population. The respondents participated in both DCE (16 choice tasks) and CV (4 valuation scenarios) in a form of payment scale followed by open-ended questions about the exact maximum WTP. We find that DCE produces higher WTP estimates than CV does, and the estimates are not consistent across the two techniques. A difference between the WTP estimates from DCE over those derived from the CV technique is found (i) for respondents who do not discriminate well between the profiles, and (ii) for an increase in the presented attribute level changes. The implications for achieving better convergence between the WTP estimates from the two techniques are discussed.