This paper measures cross-price elasticity for online and retail personal computers (PCs) across several years. The results indicate that cross-price elasticity for 1996 and 1997 was a statistical 0, while in 1998 it was large and statistically significant (approximately 3.0). I consider both demand-side and supply-side explanations for this change. My demand-side explanations argue that a change in the composition of households shopping in online and retail markets is responsible for the switch. The empirical results suggest that these explanations are not correct. My supply-side explanations argue that changes in business models late in 1997 were responsible instead. The first supply-side explanation says that expansion by firms from single to multiple distribution channels (e.g., expansion from online only to online and retail) drove the new competition. Both Apple and Gateway made such expansions late in 1997, and the data provide some evidence that this contributed to online/retail competition. The second explanation says that the increase in the number of brands offering both physical inspection (through a physical location) and customizability drove the new competition. These changes marked directed efforts to target high-end buyers by both types of suppliers. The data also support this explanation. While I am unable to quantify which has the greater impact, it is clear that online/retail competition began largely because of these two supply-side changes.
Prince, Jeff (2007), “The Beginning of Online/Retail Competition and Its Origins: An Application to Personal Computers,” International Journal of Industrial Organization, Vol. 25, No. 1, pp. 139-156.