Incorporating Discrete Choice Experiments into Policy Decisions: Case of Designing Public Long-Term Care Insurance

Discrete choice experiments (DCEs) have been widely used to elicit preferences in the health economics field but recent reviews found that DCE results are rarely incorporated into health policy decisions. We conjecture that one reason is most health policy practitioners only focus on estimating marginal willingness to pay (MWTP), the measure that is not directly applicable for policy-related questions. We show that when designing a new program, translating preference information into the demand for packages and benefits of alternative schemes (the choices made available) can make the DCE results more policy relevant. This concept is illustrated using data collected to evaluate the benefits of introducing a public long-term care insurance program to a middle-income country, Thailand. We find that preferences are very heterogeneous, implying that no one-size-fits-all solution exists. The estimates from the preferred model are then used to calculate benefits and losses (based on the consumer surplus measure) for plausible implementation scenarios such as different universal schemes, multiple-tier schemes, and schemes in which premium are subsidized for low-income households.