Endnotes

¹See Saggi (2002), Keller (2004), Görg & Greenaway (2004), and Lipsey & Sjöholm (2005) for excellent surveys on this issue.

²Jaffe, Trajtenberg, & Henderson (1993) and Keller (2002) have shown that due to the tacit nature of knowledge, face-to-face interaction is probably the most effective way of transferring knowledge, and thus knowledge diffusion tends to be geographically localized within regions and countries.

³This result is consistent with Todo, Zhang, & Zhou (2006).

⁴China has long instituted a strict hu kou (household registration) system to regulate the mobility of people across localities. A person without a hu kou in a particular locality is not entitled to social benefits (e.g., schooling and housing subsidies) in that locality.

⁵The calculation is based on simple comparison of the average TFP between MNEs and domestic firms, without controlling for any firm characteristic. The size of the TFP differential varies across industries. For example, in the food processing, the chemical, and the pharmaceuticals industry, the MNE average TFP is more than 80% higher than that of domestic firms, whereas in the plastic products, the non-metallic mineral products, and the general equipment industries, the MNE premium is 60-70%. By contrast, in the coke and refined petroleum product, rubber, other basic metals, and transport equipment industries, in which the number of firms is relatively small, the average TFP is higher for domestic firms than for MNEs.

⁶The details of observations from the interviews are presented in Cai, Todo, and Zhou (2007).

⁷Alternatively, we used MNE capital stock and sales totals to represent the extent of foreign firms' production activities. The results from using these alternative variables are very similar to the benchmark results from using total MNE labor.

⁸Buckley, Clegg, and Wang (2007) find that the effect of the percentage of foreign-owned firms on labor productivity of locally-owned firms (not in logs) is nonlinear, using industry-level cross-section data for the PRC. Since our estimation uses logs as variables, non-linearity is incorporated to some extent. Further, when we included the square of the natural logarithm of FDI as an additional independent variable, we found that its effect is always insignificant.

⁹It should be noted that since our data set covers only firms in the Z-Park rather than all firms in the market, SHARE in our regression is the market share of the firm in the science park. However, as we explained earlier, the Z-Park is very large in size and we therefore assume that the market share in the science park reflects the market share in the whole market.

¹⁰See Li, Zhang, and Zhou (2005) for a more detailed description of the Z-Park. Note that data on intermediate goods and R&D expenditures are not available in 1999 or earlier.

¹¹Huang (2003) claims that a considerable proportion of FDI in the PRC from Hong Kong, China is round-trip FDI, i.e., the investment actually originated from domestic sources and was undertaken to take advantage of the tax, and other, benefits enjoyed by foreign-funded firms. Using the same data set as in this paper, Todo, Zhang, and Zhou (2006) find no spillovers from FDI in R&D from Hong Kong, China; Macau; or Taipei,China. Using the framework in this paper, we found that the effect that MNEs from Hong Kong, China; Macau; or Taipei,China had on local productivity is always insignificant. In addition, using industry-level data, Buckley, Clegg, and Wang (2002) find no productivity spillovers from FDI from Hong Kong, China; Macau; or Taipei,China firms, and Buckley, Wang, and Clegg (2007) find that this is particularly the case for technologyintensive industries which are the focus of this paper. Todo, Zhang, and Zhou (2006) and Ge and Chen (2008) find that MNEs from Hong Kong, China; Macau; or Taipei,China are less productive than other MNEs. This productivity differential may be the reason for the absence of knowledge spillovers from overseas Chinese MNEs from Hong Kong, China; Macau, China; or Taipei,China. Since a comparison between Hong Kong, China; Macau, China; or Taipei,China MNEs and those of other economies is not the main focus in this paper, we completely exclude Hong Kong, China; Macau, China; or Taipei,China MNEs from our definition of MNEs.

¹²These are (1) the petroleum, chemical, and pharmaceuticals industries, (2) the non-metal, metal, machinery, and transportation equipment industries, (3) the electric equipment, communication equipment, computer, and precision machinery industries, and (4) other industries.

¹³A similar procedure is employed in other studies, e.g., Basant and Fikkert (1996).

¹⁴Moreover, modifying this assumption does not materially change the benchmark results.

Download this Paper [ PDF 194.5KB| 27 pages ].

Comment(s)

There are [0] comment(s) for this entry. Post a comment.