|
The 39th WORKSHOP |
|
Subject:
MA in International Development, GRIPS-FASID Joint Program |
|
Summary
To begin with the presentation, Ms. Hien made a brief introduction about definitions of such important terms as FDI, infrastructure, and ODA, and then the roles of ODA and FDI in economic growth and poverty reduction in developing countries. She also illustrated how ODA in infrastructure could influence FDI inflows in developing countries. It was shown that ODA might have effects on FDI inflows via various channels, such as accumulating human capital through improving education and health care systems, or increasing market size though generating higher income. Ms. Hien also argued that the relationship between ODA in infrastructure and FDI needed to be reviewed at different levels and aspects, and she then reviewed some key studies in the past about the relation. According to her, there have been no studies discussing such relation in Vietnam, and thus her research would be the first challenge for the case of Vietnam.
Ms. Hien went further with detailed discussion on data and methodology. In the paper, she used data collected from various sources in the period 2002-2004. ODA data showed the disbursed amount, while FDI data indicated the registered amount. All the data of ODA and FDI were provincial data, as the main aim of the paper was to see how ODA in infrastructure in each province could impact on FDI inflows in the province itself. The presenter started her research with two null hypotheses: (1) H01 (ODA in infrastructure had no positive impact on FDI inflows) vs. H11 (ODA in infrastructure had positive impact on FDI inflows); and (2) H02 (ODA of one donor had no positive and strong impact on FDI from that particular country) vs. H12 (ODA of one donor had positive and strong impact on FDI from that particular country). To avoid some econometric problems, such as endogeneity, possible correlation between explanatory variables and error terms, and omitted variables, Ms. Hien used two-stage least square (2SLS), fixed effects and random effects models. A number of variables were considered in the models, such as provincial per capita GDP as proxy for wage rate in each province, and average tax rate presenting the treatment of the local government to attract FDI.
The estimated results provided many interesting findings. For instance, ODA in infrastructure was not significantly higher in the provinces with special difficulties (i.e. mountainous and remote provinces), and it was less in the provinces which had better public services such as health care system. The results from 2SLS model indicated that the direct short-term impacts of ODA in infrastructure on provincial FDI inflows were not substantial. However, fixed effects model controlling provincial factor showed that ODA in infrastructure had positive impact on FDI inflows in the studied provinces. In other words, the first null hypothesis was rejected. Also, regarding the second null hypothesis, the paper considered the cases of Japan and EU as the two biggest ODA donors of Vietnam. The estimates showed contradictory findings for Japan and EU, in which ODA from Japan had a positive and significant impact on the Japanese FDI in the studied provinces, while that did not occur with EU’s ODA and FDI. This meant that the second null hypothesis could not be rejected. From these findings, Ms. Hien proposed some policy implications to attract ODA and FDI, including promoting ODA in infrastructure; creating concrete social and economic plans towards increasing returns of infrastructure development; and investing more on industrial human resources via vocational training.
[For further information of models and findings, please refer to her paper and presentation at the end of this summary]
The discussion session was really interesting with many questions and comments on data quality and policy implications. Prof. Kenichi Ohno (GRIPS & VDF) commented that FDI in different industries, such as manufacturing and hotels, were clearly different, and thus had different relations with ODA in infrastructure, and therefore considering FDI in aggregate might not be appropriate. In addition, he said that some ODA in mega projects, such as My Thuan bridge, would provide benefits to many provinces in the region, so that it might be difficult to precisely conduct research on such expected relation. In responses, Ms. Hien acknowledged these difficulties, because it was impossible for her to categorize the available data of ODA and FDI to pursue such analysis.
Also about data, Mr. Vu Tuan Khai (YNU & VDF-Tokyo) and Mr. Giang Thanh Long (GRIPS & VDF-Tokyo) shared a similar question of why FDI data indicated registered amount, while that of ODA was disbursed amount. According to them, it was clear in the case of Vietnam that registered FDI was significantly different from implemented FDI, and thus studying on implemented FDI and disbursed ODA would be more appropriate under the purpose of fact findings. Ms. Hien answered that her research aimed to see how ODA in infrastructure would affect the FDI attraction in different provinces, and thus she used data for registered FDI. As Ms. Hien said that she used flow data for both FDI and ODA, Prof. Ohno commented that for infrastructure, stock data would be more important because they would show the long-term expectation of infrastructure development, and thus environment for possible investment in the future.
Regarding hypotheses and models, Prof. Ohno said that the second hypothesis might be not encouraging, because FDI investors might not care about whether infrastructure in one province was built by Asian Development Bank (ADB), World Bank, or Japan International Cooperation Agency (JICA). Moreover, he said that the instrumental variable “mountainous province” might not appropriate since ODA and FDI might have causal relationship in some cases. In her reply, Ms. Hien said that the model would consider only the relation between ODA from a particular country, e.g. Japan, and FDI from that country. She also admitted that it would be difficult to evaluate the impacts if the infrastructure was built by many donors. For the second question, she did agree with Prof. Ohno’s comment.
Discussing about per capita GDP as the proxy for provincial wage rate, Mr. Long commented that such proxy would be biased when considering different income levels between provinces, meaning that it would be underestimated for the rich provinces, while it would be overestimated for the poor provinces. Prof. Ohno gave an example that Binh Duong and Dong Nai had high income levels but their wage rates for a large proportion of labor force, which was mainly composed of migrant labor, were much lower. Regarding the proxy for education, Mr. Long and Ms. Tran Binh Minh (IDE Advanced School & CIEM) had the same comment that it would be better to use completion rate of secondary school than the enrolment rate, because such the former would, to some extent, reflect the actual average level of education for the labor force in each province.
Ms. Le Thi Thuy Duong (Waseda University) wondered whether the term “infrastructure” also included human infrastructure, as she thought that such human factor would have significant influences on inflows of both ODA and FDI. Ms. Hien said that her research considered only physical infrastructure, and she did agree that human capital played a crucial role in attracting further ODA and FDI, even her estimates also indicated this comment.
At 16:00 we close the presentation session, and we then had 1-hour informal meeting. In this workshop we also welcomed some new participants from other research institutions in Japan and the UK. A lot of information about Vietnam’s development was excitingly exchanged between participants.
(By Giang Thanh Long) |
