Sequential Innovation, Network Effects and the Choice of Compatibility

It is commonly argued that when innovation is sequential, and the product has network externalities, incumbents build a large network that inefficiently blocks the entry of future incompatible innovators. This paper shows that when intellectual property rights permit some degree of compatibility between the technologies of the incumbent and entrant, increased network size is not necessarily a deterrent to entry. In some cases an incumbent might prefer to construct a small network in order to make its capture by an entrant less appealing. Furthermore, the threat of entry may induce an incumbent to underinvest in the quality of his product. Finally, we show that property rights that fail to protect the incumbent are suboptimal, especially when the cost of future research is high. Moreover, weak property rights decrease current welfare by reducing the incumbent's incentives to create a big network in the first place. In most cases, a subsidy of the cost of R\&D of future innovators turns out to be more efficient than allowing entrants to use highly compatible technologies