The Optimal Network Contract Under Partial Bypass in Oligopolistic Network Industries

The rapid widespread technological change and concomitant deregulation of network industries has engendered a burgeoning demand for connection between technologically similar as well as technologically dissimilar networks. The processes by which interconnection contracts are reached and the nature of these contracts is important for the performance of these industries. The basis for public policy concern about these contracts stems from the perception that there remain natural monopoly elements in networks. Wherethese elements are absent or equivalently if networks are open to economic bypass interconnection contracts will generally not pose special competition concerns. The purpose of this paper is to examine the effect of credible potential bypass of a network on the economic efficiency of a privately chosen network contract. There is no regulation excepting the requirement that there must be no non-price discrimination against a firm that wishes to utilise the network.The set up is one in which there are two downstream retail firms that require a network to do business. They supply different but possibly very similar services and they compete as a duopoly for retail customers. They each face the same contract for the use of the network. There is a vertically integrated firm (the incumbent) that owns the network and one retail firm. Two organisational structures are considered. In one the vertically integrated firm is a conglomerate that chooses the network contract and its retail firm's output subject only to the reaction of the other retail firm. The second is a divisional organisational structure where the incumbent's retail firm takes the network contract as given and acts independently to maximise profits subject to interaction with the other retail firm. In the divisional structure the network contract is designed to maximise the profits of the total firm but subject to divisional autonomy. The divisional structure yields relatively more efficient contracts than does the conglomerate but total incumbent profits are lower. As a practical matter this represents a factor to be considered in choosing an organisational structure because there canbe organisational performance gains in decentralised structures.The second retail firm may build its own network to bypass the incumbent's network. There is no regulatory intervention and because of the threat of bypass the vertically integrated firm does not price the other firm off the network. The vertically integrated firm has the inherent natural monopoly characteristic of taking leadership in the design of its two-part tariffcontract for the use of its network. It therefore designs the network contract taking cognisance of the ability of the second retail firm to construct a bypass network and its strategic output reaction to actions of the incumbent.The results indicate that when the full network has to be bypassed if at all the vertically integrated firm has very considerable latitude to raise the variable component of the network charge to restrict the other retail firm's output. This inefficient situation is tantamount to bypass not being possible.The position is different when the second retail firm can partially bypass the incumbent's network and use a combination of its own and the incumbent's network to service customers. Assuming that if the second retail firm did partially bypass it would do so by bypassing targeted network segments that are relatively densely populated by potential subscribers (e.g. the central business district) partial bypass is such a threat to the incumbent that it designs a much more efficient contract. The vertically integrated firm designs a contract that raises the variable component of the network contract as much as possible to reduce the outside firm's output while ensuring that the outside retail firm does not bypass much of the network. This threat of bypass forces in the range of examples described in the paper a relatively efficientoutcome and a contract that approximates the incumbent's network cost function.Because the network contracts are relatively economically efficient the profit performance of the divisional and conglomerate structures are so similar that a choice between them could be made on grounds of organisational performance. This is not considered in the paper.Although the paper's assumption that there is one contract for all segments of the network is restrictive the paper does illustrate the strong competitive influence of potential bypass on an unregulated firm's choice of a network contract.