Innovation and survival of new firms in Chinese manufacturing, 2000-2006

Using a large dataset of over 100,000 Chinese firms created between 2000 and 2006, we explore whether there is a link between innovation effort (R&D) or innovation output (the share of innovative sales) and the firm's duration of survival. We estimate a complementary log-log model with time-varying explanatory variables controlling for individual heterogeneity. We find that innovative firms tend to survive longer, more so because of R&D than because of introducing new products. There seems to be an inverted-U relationship between R&D or innovation output and long-term survival, suggesting that too much R&D or product innovation can cause firms to die, perhaps because of excessive risk. Survival has a cyclical behaviour, and it varies across provinces. It also varies with ownership. State-owned firms have a higher hazard rate than privately-owned firms, which have a higher hazard rate than foreign-owned firms.

Key words: firm survival, complementary log-log duration models, China, innovation

JEL codes: L25, O38