Bias and Efficiency of Uniform Bid Design in Contingent Valuation

While contingent valuation (CV) methods have experienced growing popularity for estimating the willingness to pay for nonmarket goods and services, optimal bid designs for CV that provide guidance in bid point placement often render themselves impractical by relying on pretest or prior information about the true distribution for willingness to pay. We investigate the use of a practical alternative to existing optimal or robust bid designs called the uniform design. Uniform design randomly draws bid points from a predetermined uniform distribution. Analytics and simulations show that the uniform design has higher low-bound of relative efficiency at 84 percent of D-optimum than a robust design. Simulations also demonstrate that uniform design outperforms other optimal designs when initial information about true parameters is poor and even outperforms robust designs when the true values of parameters are known.