Abstract

When modeling consumers’ forward-looking behavior using choice data on frequently purchased products, the common approach assumes that consumers have rational expectations about future promotions. Previous studies modeled such expectations using a first-order Markov (FOM) process. However, empirical evidence from several categories suggest that inter-promotion intervals can last several weeks implying that a FOM process that conditions future expectations of prices only on current-period prices can be limiting. We utilize a Proportional Hazard model (PHM) to characterize consumers’ rational expectation of future price promotion. We first show that estimating a dynamic structural model that uses a FOM specification for rational expectations can bias estimates of promotion effects with both simulation analysis and scanner panel data from four consumer packaged goods product categories. Secondly, we empirically show that a structural model employing a PHM specification for promotion expectations fits the data better than ones that assume only a FOM price or promotion expectation. Lastly, we show using an analysis of promotion policy changes that a structural model with a FOM expectation can lead to suboptimal managerial decisions.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call