The R&D literature framed in a strategic context shows two unpleasant outcomes for the public goods nature of knowledge: 1) the private R&D activity results in under-investment (with no information leakage – no spillovers) or over-investment (with information leakage – positive spillovers) compared to the social optimum because of appropriability, and 2) the R&D outcome shared by each firm is lower than full disclosure, as innovators are not rewarded for disseminating information. This article departs from De Bondt et al. (1992), who consider the cost-reducing (process) innovation duopoly à la d’Aspremont and Jacquemin (1988, 1990) with non-network goods showing that the (second-best) social optimum requires partial disclosure if products are homogeneous. Unlike these studies, this work finds that, in a network industry, full disclosure becomes optimal depending on the extent of the network externality. Results offer clear policy implications.

战略背景下的研发文献显示了知识的公共产品性质的两个令人不快的结果：1）私人研发活动导致投资不足（无信息泄漏——无溢出）或投资过度（有信息泄漏——正）由于专有性，与社会最优相比，溢出效应）；2）每个公司共享的研发成果低于完全披露，因为创新者不会因传播信息而获得奖励。本文源自 De Bondt 等人。(1992)，他们考虑了 d'Aspremont 和 Jacquemin (1988，1990) 与非网络商品的降低成本（过程）创新双头垄断，表明如果产品是同质的，（第二佳）社会最优需要部分披露。与这些研究不同，这项工作发现，在网络行业中，根据网络外部性的程度，完全披露变得最优。结果提供了明确的政策含义。