Abstract
By making use of a recently released panel dataset that covers 61 provinces of Vietnam from 1996–2005, this study examines the link between foreign direct investment and economic growth. Our analysis, which is based on a simultaneous equations model, reveals that in overall terms a mutually reinforcing two-way linkage between FDI and economic growth exists in Vietnam. However, this is not the case for each and every region of Vietnam. The results presented in this study suggest that the impact of foreign direct investment on economic growth in Vietnam will be larger if more resources are invested in education and training, financial market development and in reducing the technology gap between the foreign and local firms.
Acknowledgements
The authors are grateful to five anonymous referees for very useful comments and suggestions. An earlier version of this study was presented to the 2007 Australasian Econometric Society Meeting held in Queensland, Australia. The authors are grateful to the participants of the meeting for helpful comments and suggestions. The usual disclaimer applies.
Notes
1. See Keller (Citation2004), Meyer (2003, 2004), Lipsey (Citation2002) and Marino (Citation2002) for an interesting survey.
2. From 1 January 2008 to 22 April 2008, Vietnam has already attracted FDI in the amount of US$7.22 billion (GSO Citation2008).
3. Prior to this, Findlay (Citation1978) has argued that FDI increases the rate of technological progress in host economies through a ‘contagion’ effect arising from the introduction of more advanced technology, management practices, and so forth.
4. See Nguyen and Nguyen (Citation2007) and references therein.
5. It is well-known that the GMM method provides consistent and efficient estimates in the presence of arbitrary heteroskedasticity (Greene Citation2008). Moreover, most of the diagnostic tests discussed in this study can be cast in a GMM framework. Hansen's J-test was used to test for over-identification of GMM (that is the null hypothesis of correct model specification and valid over-identifying restrictions was tested and the results were found to be satisfactory).