A Panel-Data Based Method for Merging Joint Space and Market Response Function Estimation

Multidimensional Scaling (MDS) techniques have proven to be useful in understanding market structures at a single point in time. In addition to examining snapshots of markets, it has been suggested that the construction of longitudinal joint spaces in conjunction with econometric modeling techniques would provide useful insights into the dynamic structure of a market. However, several pragmatic factors related to data collection have limited the amount of research in this area. In this paper, a method is proposed that enables researchers to use commonly available consumer diary panel data to construct a time series of joint space maps and then to integrate these maps into market response models. Two illustrations using longitudinal maps demonstrate successful applications of this methodology.