Modeling the Household Purchasing Process Using A Panel Data tobit Model
Dong, Diansheng; Schmit, Todd M.; Kaiser, Harry M.
A panel data Tobit model is developed to examine the household purchase process for a frequently purchased commodity. The proposed model accounts not only for censoring or sample selectivity, but also the temporal dependence of the purchasing process using household panel data. The flexible error structure in the model accounts for both state dependence and household heterogeneity. Empirical findings show that purchase habits of milk persist across households over time, and most of them come from the household heterogeneity in preferences. Results also show that advertising increases the purchase quantity and purchase frequency simultaneously.
Charles H. Dyson School of Applied Economics and Management, Cornell University