Strategic Decision of Network Intensity: Evidence from U.S. Interfirm R&D Collaboration Network
Research Poster Arts & Humanities 2025 Graduate ExhibitionPresentation by Sinjeong Kim
Exhibition Number 93
Abstract
Firm-to-firm R&D collaborations have become more and more common, and at the same time, firms have been assigning increasing numbers of inventors to these collaborative projects. Despite this trend, previous research on R&D subsidy has mainly focused on the existence of collabora tions without considering varying intensities of collaborations. However, ignoring collaboration intensity, R&D subsidy can misleading favor firms with many weak collaborations over firms with fewer, yet potentially more impactful collaborations. In this paper, we develop a model for firms’ choices of R&D collaborators and varying investment levels for collaborations. We then estimate our model using U.S. patent registration and accounting data for publicly listed firms, utilizing the number of inventors as a proxy for network intensity. Our findings indicate that increasing collaboration intensity by one level results in a 107% increase in technology spillovers but incurs 76% higher costs. In our counterfactual analysis, we will examine optimal R&D subsidy design. We will redo this analysis while wrongly assuming all collaborations have the same intensity, to show the importance of accounting for collaboration intensity in R&D subsidy design.
Importance
The U.S. government invests over 200 billion in R&D subsidies, yet current policies often overlook how firms allocate resources across different R&D collaborations. My project sheds light on these previously unexplored aspects of R&D policy, potentially suggesting more effective R&D subsidy designs.