Innovation Networks and R&D Allocation

We study the cross-sector allocation of R&D resources in an innovation network, where one sector’s past innovations may benefit other sectors’ future innovations. We solve for the optimal R&D allocation and show that a planner valuing long-term growth should allocate more R&D toward network-central sectors, but this incentive is muted for countries receiving more foreign knowledge spillovers. We derive sufficient statistics for measuring the welfare impact of improving cross-sector R&D allocation. Empirically, we build the global innovation network based on patent citations and evaluate R&D allocative efficiency across countries. For the US, improving R&D allocation could generate 8% welfare gains.