Essays on spatial price competition

In this thesis, I examine the importance of spatial interaction amongst retail outlets in demand system estimation. In Chapter 1, I discuss why and how a spatial linear demand system model can be examined. The focus of this chapter is primarily to illustrate how the various parameters of a linear demand system model developed for store level demand analysis should be interpreted. In addition, I discuss how the aggregate market level analysis is different from store level analysis. The second chapter builds on the framework and results of first chapter to empirically demonstrate the effect of spatial interaction amongst retailers in the Singapore market for Cola category. The study uses a reduced form model inspired by almost ideal demand system model and spatially weighted price. In addition, own store-own brand and cross brand prices are used as explanatory variables. The findings show that spatial coefficients are significant and positive. An important implication of our findings is that consumers can switch either brand or store in case of increase in price of a specific brand. Store-substitution is more pronounced for the brand with higher market share and price premium (market leader) owing to higher percentage of brand loyal consumers. Additional results demonstrate that intra and inter-channel store substitution effects are asymmetric. The third chapter uses a reduced form spatial econometric model on characteristic space for demand system equation. This approach is more suited to accommodate multiple brands within a category. I apply this model to investigate the spatial effects amongst retailers in the carbonated soft drinks category in Singapore. The significance of spatial coefficients in final results lends support to the importance of spatial price in demand estimation and existence of spatial spillover effects in elasticity estimates.Doctor of Philosoph