Calculating willingness-to-pay with discrete cost and random coefficients in discrete choice experiments

Area of Science:

• Health Economics
• Behavioral Economics
• Econometrics

Background:

• Willingness-to-pay (WTP) is a key metric in economic evaluations.
• Discrete choice experiments (DCEs) are widely used to elicit WTP.
• Traditional DCE models often assume continuous cost, which may misrepresent discrete costs.

Purpose of the Study:

• To provide guidance on calculating WTP in DCEs with discrete costs.
• To highlight limitations of linear cost disutility assumptions.
• To propose and demonstrate methods for accurate WTP estimation with discrete costs.

Main Methods:

• Five mixed-logit models were employed.
• Log-normal distributions were used for cost parameters.
• A piecewise linear utility function was developed for discrete cost WTP calculation.
• Individual-level simulations were conducted to compare median and mean WTP.
• A case study on colorectal cancer screening preferences was used for demonstration.

Main Results:

• Discrete cost models showed higher disutility at low costs and lower disutility at high costs compared to continuous models.
• Continuous cost models tended to overestimate WTP at low costs and underestimate at high costs.
• Quadratic cost terms offered only partial improvement, failing to capture sharp preference changes.
• Policy recommendations differed between continuous and discrete cost models.

Conclusions:

• Accurate WTP derivation necessitates incorporating discrete costs in DCEs.
• Appropriate distribution assumptions for cost parameters are crucial for precise WTP estimation.
• The choice of cost representation significantly impacts WTP results and policy implications.