The (dis)connection between R&D and productivity in China: Policy implications of R&D tax credits

We use Chinese firm-level data from 2001 to 2007 to estimate the (dis)connection between firms' R&D and productivity and find that the productivity effect of R&D investment is very low. We conjecture that firms conduct/report unproductive R&D in order to obtain policy benefits. To explore this plausible misconduct, we investigate the effects of China's 2003 R&D tax reform on firms' R&D investment. The reform generates exogenous treatment variations across firm ownerships for causal identification. We find that the reform has statistically significant and positive effects on firms' R&D investments. Quantitatively, the reform raises firms' R&D investments by 6.68 % and the estimated R&D elasticity of tax deduction for private firms is 0.9147, which is comparable to other countries. However, our further empirical results indicate that the policyinduced R&D is less efficient in promoting firms' productivity than the spontaneous R&D. We provide evidence of firms' relabeling non-R&D expenses to R&D expenses, which partly explains the inefficiency of R&D investment.