A Choice Model with Systematic Structures in Decision Weights

This article introduces a discrete choice model which incorporates a nonlinear structural adjustment to the standard utility coefficients or decision weights. The proposed model is theoretically and empirically appealing when compared to several alternative approaches, and it can be estimated by conventional maximum likelihood. Application of the proposed model in a case study shows that it outperforms two competing approaches in model fit. Given its simplicity, this model is also capable of revealing consumers' heterogeneous choices. It is shown that based on consumers' different characteristics, their product choice and its welfare implications are also potentially different.