A latent segmentation approach to a Kuhn–Tucker model: An application to recreation demand

In this paper, we extend the latent segmentation approach to the Kuhn–Tucker (KT) model. The proposed approach models heterogeneity in preferences for recreational behavior, using a utility theoretical framework to simultaneously model participation and site selection decisions. Estimation of the latent segmentation KT model with standard maximum likelihood techniques is numerically difficult because of the large number of parameters in the segment membership functions and the utility function for each latent segment. To address this problem, we propose the expectation-maximization (EM) algorithm to estimate the model. In the empirical section, we implement the EM latent segmentation KT approach to analyze a Southern California beach recreation data set. Our empirical analysis suggests that three groups exist in the sample. Using the model to analyze two hypothetical beach management policy scenarios illustrates different welfare impacts across groups.