This study examines whether network capability continues to facilitate corporate digital innovation amid rapid technology iteration, thereby addressing a significant theoretical paradox.
Based on firm-year observations of Chinese A-share listed companies during 2012–2022, we construct a measure of digital innovation from annual-report MD&A texts using a dictionary-based text-mining procedure. Fixed-effects models, robustness tests and an instrumental-variable approach are used to examine the hypotheses.
Results reveal that network capability exerts an overall inhibitory effect on digital innovation, representing a contextual shift from traditionally assumed benefits to negative consequences. This inhibitory effect is weakened by TMT heterogeneity and by stocks of granted patents, including invention, utility model and design patents.
This study extends embeddedness theory by showing that network capability can shift from an innovation-enabling resource to an innovation-constraining rigidity. It also identifies internal cognitive supplementation and external codified knowledge supplementation as boundary conditions that help firms maintain knowledge flexibility.
