Network Externalities, Mutuality, and Compatibility

Positive network externalities can arise when consumers benefit from theconsumption of compatible products by other consumers (user-positiveconsumption externalities) or, alternatively, when they incur costsfrom the consumption of incompatible products by other consumers(nonuser-negative consumption externalities). But whereas user-positiveexternalities are typically mutually imposed and imply mutual benefitbecause they relate to interoperability, with nonuser-negativeexternalities the costs of incompatibility may be imposed unilaterallyand borne asymmetrically. For example, increased risks of death andinjury on the roads due to the co-existence of large and small vehiclesare imposed exclusively by the owners of the large vehicles and borneexclusively by the occupants of the small vehicles. This paper comparesthe social optimality of incentives for compatibility under regimesinvolving user-positive and nonuser-negative externalities. Earlierwork with respect to user-positive externalities (e.g., Katz andShapiro, 1985) suggests that firms with relatively small networks orweak reputations tend to be biased in favor of compatibility, whileindividual firms’ incentives for compatibility are suboptimalwhen their networks are closely matched in size. Meanwhile, intuitionsuggests that with nonuser-negative externalities incentives forincompatibility should always be excessive, reflecting the notion thatactivities involving unilaterally imposed negative externalities willalways be overprovided by the market (in the absence of regulation orCoaseian mitigation). Using a "location" model ofdifferentiated products, we find that, under both regimes, incentivesfor compatibility tend to be suboptimal when firms' networks are closein size, and excessive for the small firm when the networks differgreatly in size. Surprising public policy implications with respect toexternalities are discussed.