Using elicitation mechanisms to estimate the demand for nutritious maize: Evidence from experiments in rural Ghana

In this paper we assess (a) consumers’ willingness to pay (WTP) for a recently developed variety of maize that is high in provitamin A in the context of a public health intervention and (b) the performance of three elicitation mechanisms in estimating WTP in a field experiment in Ghana. The mechanisms that we used for elicitation are the Becker-DeGroot-Marschak (BDM) mechanism, kth price auction, and choice experiment. The basic design of the experiment involved random allocation of consumers to one of three elicitation methods. This was augmented to include treatment arms to address the effect of (1) participation fees and (2) nutrition information on WTP. Estimation of BDM and kth price auction models that account for censoring of bids at the market price for maize (kenkey) and estimation of a conditional logit (CLM) model for the choice experiment that accounts for lexicographic preferences yield estimates of average WTP that are similar in magnitude across the three elicitation mechanisms. Variation in participation fee has no effect on estimated WTP in the two mechanisms that varied participation fee, suggesting that people did not have a higher propensity to spend out of windfall income. In the absence of information on the nutrient density of the new maize variety, subjects are willing to pay less for it than the existing varieties; however, nutrition information transforms this discount into a substantial premium.