Allies Yet Rivals : Input Joint Ventures and Their Competitive Effects

Sammanfattning: Essay I explores the competitive effects of input joint ventures, more precisely the connection between horizontal product differentiation and competition. Four quantity-setting downstream firms jointly and symmetrically own two upstream firms in pairs. Final goods are homogeneous if originating from the same upstream firm, differentiated horizontally if not, leading to two varieties. The equilibrium output prices are lower than under both vertical separation (the downstream firms do not own the upstream firms) and merger (the part-owners of each upstream firm have merged) if varieties are substitutes, and lower than under vertical integration (each downstream firm has its own input facility) if varieties are close substitutes, higher if not.Essay II investigates, in a two-networks setting, the consequences of having jointly owned network infrastructures. The model is formally that of Essay I extended to arbitrarily asymmetric equity shares. Owners bargain to determine the price of infrastructure access. As compared with separately owned networks, social welfare is higher. The degree of horizontal product differentiation between services of different networks is crucial to the ability to agree on a high access charge: if they are not close substitutes, there is a distinct collusive region; if they are close substitutes, the collusive region is not only absent but access can even be priced at marginal cost.Essay III develops a fairly general model of two competing network operators in the electronic-communications industry. Investments in the network infrastructure enhance a firm’s product quality while giving rise to demand-side spillovers. The quality-setting incentives under noncooperation (separate network infrastructures) and NIJV (network infrastructure joint venture) are compared. When the savings in physical equipment made possible under NIJV are large, NIJV leads to an increase in the level of quality. When they are small, NIJV leads to an increase in it if there are Cournot competition and large positive spillovers, or Bertrand competition and either positive or small negative spillovers, but to a decrease if there are Cournot competition and either negative or small positive spillovers – and probably also if there are Bertrand competition and large negative spillovers.