Willingness to Pay for Beef Quality Attributes: Combining Mixed Logit and Latent Segmentation Approach

This study develops an alternative way of identifying the heterogeneity of WTP estimates for product attributes. The alternative approach combines mixed logit models with a latent segmentation method (MLS). First, the combined MLS approach estimates a mixed logit model and retrieve individual specific parameter estimates based on individual specific posterior distribution derived from the individual sequence of observed choices in the conjoint experiments. Second, WTPs of each choice variables are computed for each individual using the marginal rate of substitution between economic and choice variables. Finally, latent segmentation is conducted based on WTPs, individual perceptions on quality attributes, and demographic characteristics, and WTPs are compared among segments. The MLS approach is applied for estimating willingness to pay for beef quality attributes and country-of-origin in the Korea market. The MLS approach allows us to identify various clusters based on people’s preferences and demographics, and provide estimates of willingness to pay for beef quality attributes and country-of-origin for these clusters.