For best experience please turn on javascript and use a modern browser!
uva.nl

Naked exclusion: towards a behavioral approach to exclusive dealing

Detail Summary
Date 1 December 2008
Time 11:45 - 13:15
Location Oudemanhuispoort

Naked exclusion: towards a behavioral approach to exclusive dealing

Abstract

We present the results of an experiment designed to test the empirical relevance of recent models on exclusive contracts. These models feature an incumbent seller, a more efficient entrant and multiple buyers. Due to economies of scale (caused, for example, by fixed entry costs), the entrant needs to sell to a sufficiently high number of buyers to enter the market profitably. In the models it is shown that an incumbent firm may deter entry of a more efficient rival by writing exclusive contracts with buyers in the market. If an incumbent is able to exploit the problems that buyers have when coordinating whether to accept or decline the payments offered by the incumbent for signing an exclusive contract, entry is successfully deterred. One prominent theoretical result is that such exploitation, and thus entry deterrence, is guaranteed if the written contracts are discriminatory in the sense that the incumbent can offer different payments to different buyers. Moreover, if contracts are private information such that buyers cannot observe each other's contracts, exclusion is obtained at a negligible cost. Exclusion is also obtained at a negligible cost when the incumbent writes the contracts with buyers sequentially.

Our main results can be summarized as follows. First, in contrast to what the existing models predict, exclusion is mostly not guaranteed in the case of discriminatory contracts. In fact, overall, the exclusion rate does not increase significantly compared to the one observed in the case of non-discriminatory contracts. Only if contracts are written sequentially and are private information, the exclusion rate increases. Second, costs of exclusion are quite substantial, even when predicted to be negligible, and are mostly not lower than in the case of non-discriminatory contracts.

We suggest that the main reason why behavioural evidence provides no clear support for the existing models' predictions is that the models do not take into account that the probability that buyers sign an exclusive contract with an incumbent is positively related to the size of the payment. In order to account for this fact, we propose to characterize buyers' behaviour by a logistic function (which is consistent with, e.g., noisy decision-making or a perturbation to buyers' payoffs). We show that this approach organizes behaviour better and does not give predictions as "extreme" as those obtained in the existing literature.

Oudemanhuispoort
Oudemanhuispoort

Oudemanhuispoort 4-6
1012 CN Amsterdam