Why does regional innovative capability vary so substantially in China? The role of regional innovation systems

Abstract

This paper empirically investigates the influence of a regional innovation system (RIS) on regional innovative capability in China. Differing from previous studies, we propose a two-step framework that firstly identifies the determinants of R&D and then tests the impact of RIS on regional innovative based on four hypotheses. The empirical results show that a region with higher ratios of private R&D and foreign-owned enterprise production, as well as closer interactions among industry, universities and academia, experiences superior innovative capability. By contrast, a higher ratio of state-owned enterprise R&D to total R&D lowers a region's innovative capability, due to the inefficiency of the state-owned enterprise R&D.

Keywords:

• regional innovation system
• innovative capability
• patent
• R&D
• foreign direct investment

Notes

For China's innovation system prior to the economic reforms, please refer to Xue (1997) for a comprehensive discussion.

Furthermore, the local governments can establish some specific R&D policies and play an important role in stimulating regional innovative activity.

For sales, employees, and numbers of enterprises based on various types of ownership in China, please refer to the China Statistics Yearbook.

China's official R&D statistics do not clearly indicate whether it contains FDI companies (including joint ventures). Thus, a higher SOE ratio of expenditure (Hypothesis 3) may mean less R&D expenditure by FDI companies, causing less regional patents. To lower the bias, we use the ratio of FDI in industrial structure rather than R&D to total regional manufacturing R&D in Hypothesis 4.

Methodological background of LISA can be found in Anselin (1995). As the variable is not the key concern in this study along with saving space, the construction of the local Moran's I statistic is not shown in the content.

In general, a successful application for an invention may take about two years, whereas the utility model or its design may take approximately one year in China.

For the detail, please see Funke and Niebuhr (2005).

Because most universities are public-financed universities in China, suggesting that both universities and research institutes belong to the public sector. Therefore, this variable can capture the interactions among the three actors to a certain degree.

Tibet is excluded in this study due to the limited information on technological activities and regional characteristics.

This positive indirect effect depends on the ability to absorb the knowledge embodied in imported technologies. That is, whether or not imported technologies can contribute indirectly to patenting heavily depends on regional absorptive ability in terms of R&D or human capital.

We also adopted two-year-ahead patents as the dependent variable to implement the estimations and obtained similar results.