Three Essays on Exclusionary Bundling

Anil Caliskan

Truland Building, 400-R
December 03, 2006, 07:00 PM to 07:00 PM

Abstract:

The first essay examines the welfare implications of exclusionary bundling using experimental posted-offer markets. The bundle contains two products, A and B, whose valuations have perfect positive correlation across the buyers to control for the price discrimination motivation to bundle. The efficiency motivation for bundling is also absent. A monopolist is the sole supplier of A and also sells B in an imperfectly competitive market identified with constrained capacities. Using two different demand schedules in the B market, the essay shows that the monopolist bundles and bundling deters entry to the B market. However, exclusionary bundling does not have any statistically significant effect on the consumer and total surplus. Only when the monopolist pure-bundles, refusing to sell any independent A, does bundling harm the consumers and the total surplus. This, however, depends on how the demand schedule in the B market is constructed. The second essay shows that a fringe seller in the A market increases the consumer surplus while decreasing the producer surplus. However, the fringe does not affect the consumer surplus extracted from the bundle despite a decrease in the bundle transaction price. The consumer surplus gains generated by the fringe seller erode if the dominant seller has a lower fixed cost in the A market. The third essay finds that moving from a mixed-strategy Bertrand-Edgeworth structure in the B market to a pure-strategy competitive environment decreases the number of sellers simultaneously present in the market and increases the transaction price. The price increase is a result of having only one seller in the B market in most of the observations, which helps the seller to better understand the characteristics of the demand schedule. The decrease in the number of sellers under unconstrained capacities also increases the frequency of signaling as well as its effect on increasing the next period price, but the latter effect is not statistically significant.