The impact of foreign direct investment on domestic supplier industries within Chinese provinces

Abstract

We investigate empirically whether the presence of foreign enterprises in downstream industries has increased intra-provincial backward linkages in China, and what characteristics of foreign direct investment (FDI) contribute to the generation of backward linkages. We conclude that FDI in downstream industries generally has failed to induce higher backward linkages except for the electronic and telecommunications equipment industry. Across industries, those characteristics of FDI, such as the labour productivity gap between indigenous and foreign companies, whether FDI is from Hong Kong/Macau/Taiwan or other countries and whether FDI is wholly owned or jointly owned, appear less influential on the backward linkages than the activities of foreign enterprises (e.g. value-added activities and import and export activities).

Keywords:

• FDI
• backward linkages
• China

JEL classifications:

• F23
• O10
• O53

Acknowledgements

We acknowledge the valuable comments and suggestions made by the anonymous referees that helped improve the paper in many ways.

Notes

1. This is because there is no inter-provincial input-output table available in China.

2. In 1998, the GDP of the industrial sector accounted for 49.3% of China's total GDP. In 2004, this grew to 52.9% (NBSC 1999, 2005).

3. The motivation of including Y₁ is that backward linkages constructed in this paper are essentially a measurement of output in the upstream industries, which is determined by the demand and supply factors including the environmental variables and the augmented foreign-related variables defined in Table 1.

4. We regard it as appropriate to treat the FDI presence variable in our model as exogenous. The reasons are, first, in the literature on FDI locational determinants, backward linkages to supplier industries are rarely a significant factor that draws FDI inflow, particularly in the context of China, a developing economy where backward linkages between industries are weaker than in developed countries. Second, the measurement of backward linkages in this study only includes domestically owned suppliers. This measurement ensures that FDI is still not endogenous if it is more likely to flow in regions/industries with more foreign-invested suppliers. Third, the measurement of FDI presence in this study is a relative measure, which is the percentage of industry output (employment) produced (employed) by foreign-invested enterprises. This avoids possible endogeneity of an FDI-level variable that might be the result of a feedback effect from the dependent variable.

5. Note that we deliberately exclude the textiles industry itself from the n sectors supplying the textiles industry, as we are only interested in the inter-industrial demand created for the other industries, and not in the intra-industrial demand created for itself.

6. The manufacturing sector has attracted a significantly higher amount of FDI than have the other two sectors, agricultural and service. In 1998, manufacturing FDI accounted for about 56.3% of the total actually utilised FDI flow in China (NBSC 1999), while in 2004, this number grew to 71.0% (NBSC 2005).

7. In 2003, the capital-to-labour ratios ranged from 25,403 to 132,049 RMB for the labour-intensive industries and from 152,866 to 272,467 RMB for the capital-intensive industries, for the 10 sample industries in this study.

8. Correlation coefficients and VIFs are not reported but are available upon request.

9. The correlations between output share and employment share in the 10 industries indicate there is no strong relation between the two measures. The highest correlation is 0.26 in the Medical and pharmaceutical industry and the lowest is 0.13 in the Electronic and telecommunications equipment industry. The average of output share is also consistently higher than the average of employment share in every industry.

10. The average labour productivity of domestic firms in this industry grew significantly between 1998 and 2004, and in 2003 and 2004 it exceeded that of foreign firms from HMT. The labour productivity gap between domestic firms and other foreign firms also reduced during this period.

11. When examining the summary statistics (not reported in the paper but available upon request) of this variable, we find the largest gaps occurred between domestic and foreign firms in the Textiles and Machinery industries, at 13.29 and 14.51, respectively. This means that the labour productivity of foreign firms is 13.29 and 14.51 times higher than that of domestic firms, respectively. For many of the other industries, the largest gap, at least 17.19, occurred between domestic and foreign firms.

12. The result of the 10-industry aggregated model is not reported but is available upon request.

13. As China has been open to FDI for more than 30 years, the trial and error stage of utilising FDI has already passed. It is now time for the government to consider backward linkages as a main benefit instead of the earlier focus on foreign exchange earnings and export generation.

14. See the detailed summary in World Investment Report 2001: Promoting Linkages (UNCTAD 2001).

15. A memorandum of the State Taxation Administration on 14 September 2006 stated that new tax rebate rates would apply to exports from China, reflecting the government's intention to encourage exports of large machinery, medical and pharmaceutical, and high-technology commodities. The new tax rebate rate for these commodities was raised from 13% to 17%, effective from 15 September 2006.