Analyzing the link between market structure and market performance is of central importance in the field of industrial organization. The aim is to understand the role of market structure, i.e., the number of firms in a market, their sizes and the products they offer, in determining the extent of market competition and market performance. In particular, antitrust and regulatory authorities are interested in knowing how many firms it takes to sustain competition in a market. For example, the expected relation between the number of firms in the market and market outcomes such as prices or qualities is at the core of merger assessments. These questions are of central importance to society and to the consumers, because a minimum level of competition ensures both sufficient provision of the good and reasonable prices for the consumer. In this paper, we investigate the interaction between market structure and market performance in e-commerce for consumer electronics. We use data from Austria's largest online site for price comparisons combined with retail-data on whole sale prices provided by a major hardware producer. We observe firms' prices as well as their input prices, and all their moves in the entry and the pricing game. With this information we can analyze how sellers’ markups over the producer’s wholesale price react to the number of firms that compete in the market. We also look at the impact of market structure on market performance over the product life cycle, as other studies focusing on market structure find that entry has, especially at the beginning of the life cycle, a significant impact on prices. An important contribution of this paper stems from a novel way of dealing with the phenomenon that, it is extremely easy for e-commerce-shops, to add and remove items from their product portfolio. This makes the analysis very difficult, because usually the number of shops will be related to how attractive an item is to sell, and will thus depend on the variables we wish to explain. This situation is an example of the "endogeneity problem", which can pose a threat to the validity of empirical estimates. We are dealing with this issue by using the information on how many shops typically listed earlier cameras at a particular stage of the life-cycle (in the past). This variable will capture overarching factors in the listing decision (e.g. distribution patterns), that cannot so easily be changed and is hence immune to such issues as whether an item is currently en vogue or not. We find a very short lifecycle of usually less than a year and a highly significant and strong effect of the number of firms on markups. Ten additional competitors in the market reduce the markup of the cheapest firm by more than 1.5 percentage points on average. We also find that this effect is strongest in the first month, but also in the end of the lifecycle. Interestingly, markups were found to be lowest in months 2-4, but at the same time, the number of firms seems to be less relevant to pushing down prices in that stage of the lifecycle. For the consumer this means that by waiting three more weeks she will get the same price reduction she would get by going to a market with one additional firm.
Hackl, Franz, Michael Kummer, Rudolf Winter-Ebmer and Christine Zulehner (2011), Market Structure and Market Performance in E-Commerce, ZEW Discussion Paper No. 11-084, Mannheim.