Problems in measuring price dispersion in e-commerce

Until recently, Internet was considered as technology that will make the trade in goods frictionless. Online retailers' margins were to fall to zero and prices - according to theory of economics - were to equalize as a result of buyers comparing prices more easily (e.g. using shop bots). Empirical research performed so far has not proven these expectations right. Studies in many countries show that online prices vary significantly (sometimes price dispersion in the Internet is higher than that in traditional trade). The purpose of this article is to present a critical view on the methods of measuring price dispersion in e-commerce. Researchers of this area use different measures of price differentials, include shipping costs or not, use the proposed price or try to determine transaction prices, reject part of the data considered as outliers that may indicate a hidden heterogeneity of a product. Some scientists also try to justify price dispersion with the reputation of a vendor, and also additional features of the sellers such as the amount of information presented in the offer, convenience of shopping, userfriendly interface, etc. All these factors are problematic for the research due to lack of a clear (and proper) way of measuring the mentioned attributes. Most of the previous studies also ignored the pricing strategy of vendors, which is a very important factor for price dispersion - it may involve reduction in prices of several products in order to attract customers to the store to buy other products with a much higher margin.