Do spillover benefits grow with rising foreign direct investment? An empirical examination of the case of China

Abstract

Using data from the Chinese manufacturing industry for 2001, this article examines the impacts of foreign presence on the performance of locally owned Chinese firms. Our key result supports a curvilinear functional form. Foreign penetration rates in excess of just about two–thirds of industrial capital are associated with declining spillover benefits, indicating the dominance of negative spillovers. The curvilinear relationship is found to be particularly strong in labour-intensive industries, contrasting with a standard linear relationship in technology-intensive sectors. The finding of the complexity of spillover effects challenges the laissez-faire view that ‘the more inward foreign direct investment (FDI), the better’ and that inward FDI into all types of domestic industry is equally valuable, in terms of performance benefits. Our findings argue for policy measures to strengthen domestically owned Chinese industry, to provide effective competition to foreign firms and to absorb the benefits from spillovers more effectively.

Notes

¹ Chakraborty and Basu (2002) however found that inward FDI in the host country produces a labour-displacing effect. This arises because the technology transfer brought in by FDI causes an excess supply of labour, creating downward pressure on unit labour costs. It seems that further evidence on the impact of FDI on labour costs is still needed.

² The association between FDI and LOEs’ performance may be a result of MNEs entering industries with higher performance, rather than of performance being raised by FDI (Meyer, 2004). While we make optimal use of the data that are available by adopting a lag structure, panel data would be better able to address the issue of causality.

³ This measure of managerial input is recorded in the Chinese statistical source and refers to the salaries of managers plus other fees that facilitate transactions.

⁴ The Report is not publicly published and is produced primarily for internal use within the SSB. It bears similarities with the widely-cited Third Industrial Census of China published by the SSB (1997) in contents and indeed it is produced especially for the years for which the industrial census is not conducted.

⁵ According to this principle, we excluded industries in which the government prohibits the operation of nonstate-owned enterprises and this naturally includes foreign firms. These state monopoly industries include ‘Oil extraction’, ‘Gas extraction’, ‘Petrolo-shale mining’, ‘Other black metal ore mining’, ‘Precious metal ore mining’, ‘Salt production’, ‘Tobacco production’, ‘Other Tobacco processing’ and other industries relating to the production and supply of electricity, gas and water. Also excluded are industries where the government is the monopoly or major purchaser of products and where nonstate firms are largely denied from entering. These industries include ‘Locomotive manufacturing’, ‘Far ocean ships’, ‘Aerospace crafts’, ‘Telecommunication transmission equipment’, ‘TV and broadcasting equipment’ ‘Radar’.

⁶ These include ‘Other nonmetal ore mining’, ‘Other ore mining’, ‘Logging of timber’, and ‘Logging of bamboo’, ‘Salt processing’, ‘Tramcar manufacturing’, ‘Instrument repairing’, ‘Arts and crafts’, ‘Daily groceries’, ‘Other groceries’.

⁷ The results of the normality test are not highly satisfactory for some models and caution needs to be exerted when interpreting results. It is generally accepted that this test has no properly defined alternatives and so has limited power.

⁸ The literature often links the magnitude of FDI spillovers with technology intensity of the industries involved (e.g. Liu et al . 2000).