A Structural Analysis of Joint Decision-Making*

This paper defines an econometric model with a coordination game to structurally analyze joint decision-making. The model solves a simultaneity bias occurring when individuals care about others’ decisions when making theirs. The model is also useful for evaluating the implications of marketing to groups of customers. We illustrate how the model measures the importance of demand complementarities within groups of customers and show that targeting one member of a group typically increases the returns to targeting other members. For some groups, these effects favor targeting all members, while for other groups, it is optimal to only target a single member. We estimate the model using hierarchical Bayesian MCMC to include a rich heterogeneity structure and obtain posterior estimates of individuallevel parameters for the purpose of targeting. To estimate heterogeneity within groups of customers, we devise an approach for imposing restrictions in the covariance matrix in Bayesian MCMC models.