Estimating the Impact of R&D Intensity on Manufacturing Firm Performance: An Instrumental Variable (TSLS) Analysis

This study investigates the causal impact of R&D intensity (RDI) on Chinese manufacturing firms' performance (ROA) using a panel instrumental variable approach. By leveraging enterprise life cycle stages (ELC) and industry competition intensity (DIC) as exogenous instruments, two-stage least squares (TSLS) estimates reveal a significant positive effect of RDI on ROA (0.41, p < 0.01), contrasting with biased OLS results (-0.001). Growth-stage firms and moderate industry competition amplify RDI's benefits, while debt ratios negatively affect performance. Robustness checks confirm instrument validity (Cragg-Donald F = 48.51 > 19.93) and panel data superiority. Policy implications advocate targeted R&D subsidies for growth-phase firms and innovation alliances in competitive sectors. This work resolves the "R&D paradox" by contextualizing innovation impacts, offering actionable insights for industrial upgrading.