TRADE, FOREIGN DIRECT INVESTMENT AND SPILLOVER EFFECT: AN EMPIRICAL ANALYSIS ON FDI AND IMPORT FROM G7 TO CHINA

The International Journal of Economic Policy Studies
Volume 1


October 2006

Article 5

TRADE, FOREIGN DIRECT INVESTMENT AND SPILLOVER EFFECT: AN
EMPIRICAL ANALYSIS ON FDI AND IMPORT FROM G7 TO CHINA





Mei JI
3-1-1?Nakagaito, Daito-City
Osaka, 574-8530, JAPAN
E-MAIL:jimei@eco.Osaka-sandai.ac.jp



ABSTRACT


The recent theories of economic growth indicated a country’s productivity depends not only on the
domestic R&D but also on Foreign R&D capital. Especially, the developing countries can benefit from
R&D that is performed in the industrial countries by trading with the industrial countries or by
receiving FDI from the industrial countries. The purpose of this paper is to test the spillover effect
through import and FDI from the developed countries to China. The empirical results, support the
beneficial spillover effect both through import and FDI.


Key words: Trade, Foreign Direct Investment, G7, China, Spillover Effect
JEL Classification: F2, O5



















83

---

The International Journal of Economic Policy Studies
TRADE, FOREIGN DIRECT INVESTMENT AND SPILLOVER EFFECT: AN
EMPIRICAL ANALYSIS ON FDI AND IMPORT FROM G7 TO CHINA *


1. INTRODUCTION

The recent theories of economic growth indicated innovation effort is a major engine of
technological progress and productivity growth. The R&D process is essentially a knowledge
generation process in which one utilizes resources (scientists, engineers, technicians, research
equipment, and so on) to create new knowledge. Innovation feeds on knowledge that results from
cumulative R&D experience and contributes to this stock of knowledge. The innovative activities of
firms not only lead to new products (whose benefits the firm can appropriate), but also contribute to a
general stock of knowledge upon which subsequent innovators can be built. So the benefit of
innovation accrues not only to the innovators, but spillover to other firms by raising the level of
knowledge upon which new innovations can be based. This is referred to as “knowledge spillover”.
Some studies have measured the extent to which growth in total factor productivity in a country
depends not only on the domestic R&D capital stocks but also on the foreign R&D capital stocks.
In a world with international trade in goods and services, foreign direct investment, and an
international exchange of information and dissemination of knowledge, a country’s productivity
depends on its own R&D as well as on the R&D effects of its transaction partners.1. As important
channels for knowledge spillover, trade and inward FDI boost domestic productivity by making
products available with the use of foreign knowledge and information that would otherwise be costly
to acquire.
Some of the theoretical and empirical studies highlight the importance of trade as a vehicle for
technological spillovers. (Coe and Helpman (1995), Coe, Helpman, Hoffmaister(1997),
Branstetter(2001)). In particular by trading with industrial countries, the less developed countries can
benefit from the industrial countries’ R&D efforts. Coe, Helpman and Hoffmaister (1997) based on
data for 77 developing countries suggest that the total factor productivity in developing countries is
positively and significantly related to R&D in their industrial country trade partners.
Hejazi and Safarian (1999) considered foreign direct investment to foreign trade as diffusion
channels linking total factor productivity, and their research diffusion effect for G6 to the OECD
countries showed that the coefficient estimates for FDI are higher than those for trade.

This is a revised version of a paper presented at COE/JPEA Joint International Conference 2005 in
Awaji Kobe. The author would like to thank Professor Reisi Maruya(Kobe University ), Koji Shinjo
( Kwansei Gakuin University ), Yuko ARAYAMA(Nagoya University), Hikari Ban (Kobe Gakuin
University),and other discussants and participants for their valuable comments. The author appreciates
the detailed and helpful comments from the journal editorial board and anonymous referees who
contributed to the improvement of earlier drafts.
1 See,Coe and Helpman (1995)
84

---

Vol. 1 Oct. 2006
There are a great number of studies on the transfer of technology from FDI to host countries. Most
of the studies find that there are positive effects from FDI flow to host country firms in advanced
economies. But the result of the case of FDI flow to developing economies is mixed2. In particular, a
number of studies for developing countries document that a foreign investment presents higher in host
country sectors while other studies point out to limited or no significant efficiency spillovers.3
As China’s Economic growth has been remarkable since the reform started in 1978, the empirical
literature on FDI in China is growing rapidly. Most studies conclude FDI has played a positive role in
promoting trade, economic growth. Recently some studies investigate whether FDI generates
technology spillover from foreign-investment firms to local ones. Liu(2002), using the industrial data
for 93-98 in Shengzhen special economic Zone of China, finds that FDI has large and significant
spillover effects in raising the productivity of manufacturing industries.
This paper, which examines the technology spillover effect from G7 countries 4 to China,
contributes to the existing literature by providing international spillovers measures for both trade and
FDI from developed countries to less developed countries.
The panel analysis is based on provincial–level data in China for the period from 1990-20025. This
studying support that the developing countries can benefit from R&D that is performed in the
industrial countries by trading with the industrial countries or by receiving FDI from the industrial
countries.
This paper proceeds as follows. Section 2 highlights the general trend and characteristics of FDI and
Import from G7 countries to China. In section 3, we set out our econometric models that are based on
a production function theory. The data are discussed briefly in section 4. Estimation results are
presented and discussed in section 5, the estimates underline the importance of the interaction between
the international trade, FDI and the foreign R&D. The Last section contains some concluding remarks.

2. AN OVERVIEW OF FOREIGN DIRECT INVESTMENT AND IMPORT FROM G7
TO CHINA

In this Section we briefly introduce the general trend and characteristics of FDI and international
trade in China.

2 For review of the relevant literature, see Blomstr m, Globerman and Ari Kokko (2000),
3 See Keller(2004) for recent review on FDI spillovers.
4 Almost the entire R&D activity in the world economy is concentrated in the industrial countries,
especially in the seven largest countries. For example, in 1990, the industrial countries accounted for
96% of total world R&D expenditures(UNESCO,1993), and G7 countries account 92% of R&D in
1991. (see Coe, Helpman and Hoffmaister, 1997).
5 It is since the 1990s that inward FDI to China has been consistently on a large scale (see fig1). So we
choose the time period of 1990-2002 in this analysis.
85

---

The International Journal of Economic Policy Studies
Figure 1 presents realized FDI6 flows in China. In the early 1990s FDI took off with rapid economy
growth. The increase in 1992 and 1993 resulted from Deng Xiaoping’s tour of Southern China when
he reaffirmed the open-door policy and encouraged the faster reform. Realized FDI continued to
increase at 10 percent growth rate. But in 1999, Realized FDI became quite flat, partly because of the
East Asian currency crisis. Despite the slowdown, Realized FDI in 2002 is 15 times the level of 1990.
The FDI from G7 countries was 23% in total FDI flows averagely during 1990s.
The distribution of FDI by region is uneven as presented. Most FDIS is located in the south and
coastal areas despite efforts by the government to diversify the locations of foreign direct investment
and lure FDI inland and toward the central and western regions. In 2002, more than 87 percent of the
total stock of realized FDI was concentrated in the eastern part of China. (See Figure2)
Figure 3 shows the total import and the Import from G7 countries to China. The average share of
G7’s import in total import is around 40%. Moreover, in 2002, among the G7 countries, Japan is the
biggest source country for China’s imports with a share of almost 50% followed by the United States.

3. THE MODEL

Based on the recent theoretical models of innovation–driven growth, R&D activity is one of the
major engines of technological progress and productivity growth. (Helpman, 1992). Consequently,
cumulative domestic R&D is an important determinant of productivity. In a world of international
trade and foreign investment, a country’s productivity depends on its own R&D as well as on the R&D
efforts of its foreign partners. Especially developing countries can benefit from R&D that is performed
in the industrial sector by trading with industrial countries or by attracting foreign investment.
As is well recognized in the literature, there are several important channels through which import
can benefit the importer country. In summary, there are three broad ways in which international trade
can boost domestic productivity. Firstly, by International trade, the import country can employ a large
variety of intermediate products and capital equipment. Secondly, international trade provides
channels of communication that stimulate cross-border learning of production methods, product design,
organizational methods and market conditions. Thirdly, International trade will indirectly affect the
domestic country’s productivity through demonstration effect, by imitating the import products, or
copying foreign technologies and adjusting them to domestic use.
FDI, as another channel, may facilitate technology spillover. For example, local firms may increase
their productivity by observing or modifying foreign firms or becoming their suppliers or customers or
attracting employees to move from foreign firms to local ones.
We estimate equations in which variations in TFP are explained by variables in both domestic and
foreign R&D capital stocks.

6 Contracted Investment refers to the amount of investment committed in signed contracts. Realized
investment refers to the total amount of FDI that actually materializes and arrives in China.
86

---

Vol. 1 Oct. 2006
Our simplest equation has the following specification

d
f
ln F = ? + ? ln S + ? ln S + ? ???
i,t
i
d
i,t
ft
i,t
i,t
where i is a province index, lnF is the log of total factor productivity (equal to lnY-blnK-(1-
b)lnL),
d
S represents the domestic R&D capital stock, and f
S represents the foreign R&D capital
i,t
i,t
stocks of trade partners or investment countries.
Two variables are utilized, namely trade weighted and FDI weighted stocks of foreign R&D.
In detail, when we analyze the relationship between the R&D and international trade,
f
S is defined
i,t
as the import-share-weighted average of the domestic R&D capital stocks of trader partners.
f TRADE
,
d
S
=
,
?m S
i t
hi,t
h
h

here, m
is the imports weight, these weights are fractions and add up to one.
hi ,t
So we get the equation (2 )as following.
d
f TRADE
ln F = ? + ? ln S
+ ? ln S ,
+ ? (2)
i,t
i
d
i,t
ft
i,t
i,t
We assume the country is more open to world economy may be benefit more from the foreign R&D.
so we add the import shares in equation (2),and estimate
d
f TRADE
ln F = ? + ? ln S
+ ?
m
ln(
? S ,
) + ? ?3?
i,t
i
d
i,t
ft
i,t
i,t
i,t
where m stands for the fraction of imports in GDP.
On the other hand, when we analyze the R&D and FDI,
f
S is defined as follows
i,t
f ,FDI
d
S
=
,
? f S
i t
hi,t
h
h

f
is the FDI weight from investor countries, these weights are fractions and add up to one.
hi,t
So we can change equation (1) into equation (4) as following.

d
f ,FDI
ln F = ? + ? ln S
+ ? ln S
?
+
(4)
i,t
i
d
i,t
ff
i,t
i,t
Provinces attracting more FDI are supposed to benefit more from foreign R&D, so we add f to
i,t
equation (4) and get equation (5)
d
f ,FDI
ln F = ? + ? ln S
+ ? *ln( f * S
) + ? ?5?
i,t
i
d
i,t
ff
i,t
i,t
it
where f is the foreign investment share in gross fixed capital formation.
i,t
Finally, taking into account FDI and trade as channel of technology transfer, the regression are as
follows.
d
f ,TRADE
f ,FDI
ln F = ? + ? ln S + ? ln S
+ ? ln S
?
+
(6)
i,t
i
d
i,t
ft
i,t
ff
i,t
i,t
87

---

The International Journal of Economic Policy Studies
d
f ,TRADE
f , FDI
ln F = ? + ? ln S + ? ln(m * S
) + ?? * ln( f * S
) + ? (7)
i,t
i
d
i,t
ft
i,t
i,t
ff
i,t
i,t
it

. DAT
4
A

The data covers three municipalities (Beijing, Shanghai, Tianjin)7and 26 provinces for the period
from 1990-2002.Tibet is excluded in our analysis because most of the import and FDI data for it is
either not available or zero during the time period examined.

4-1 Total Factor Productivity

Total factor productivity is defined as the log of output minus a weighted average of the logs of
labor and capital inputs, where the weights equal factor shares. LNF=lnY-blnK-(1-b)lnL. The
coefficient b, which is the share of capital income in GDP, is set to 0.4.( following Coe, Helpman
&Hoffmaister(1997)).
Y is the provincial real
he GDP
GDP. T
data and implicated GDP deflator for 1990-2002 come from
the China Statistics Yearbooks.
K is the capital stock. We estimate the real capital stock using the standard perpetual inventory
approach. The investment series used is the total social fixed asset investment from the China statistics
Yearbooks for 1991-2003.Since 1991,the Chinese statistical sources began to report the price indices
for fixed assets investment .So we used that deflate to construct the time series of physical capital
stock from investment flows. We adopt an overall depreciation rate of 10%.
L is the labor from “Total labor force of society by the end of year” which is reported annually in
the China statistics Yearbooks. We used the annual average total labor force of society.
As show in Table 1,total factor productivity (TFP) increased over the period of 1990 to 2002 in all
provinces. Jiangsu, Zhejiang, Fujian, Shangdong, Hainan and Shangdong experienced the greater
increase in productivity of more than 20%, respectively, and all of these provinces are located in south
and coastal region of China. While Guizhou’s TFP increase rate is the lowest. Almost all the provinces
had the intermediate values, and averagely in the western region the TFP increase rate is a little lower
than that in other regions.

4-2 R&D Capital Stock from G7 Countries

The R&D expenditures data are from the OECD’s “Main Science and Technology Indicators”. We
used the perpetual inventory approach to calculate the R&D capital stocks .The depreciation rate is

7 Chongqing was established in 1997, so it is included in Sichuan province for the consistency of the data.
88

---

Vol. 1 Oct. 2006
defined at 10% a year. As shown in Table 1, overall changes in foreign R&D capital stocks were not
so dramatic.
Table 1 also provides data on import shares in GDP and FDI shares in total fixed investment. There
exist substantial differences in import shares and FDI shares between the coastal provinces and the
inland provinces, which indicated the coastal region is more open to the foreign countries and depends
more on the foreign economy.

4-3 China’s Domestic R&D

We used patent application to stand for the domestic R&D stock in China. The data are taken from
the China’s statistical yearbook and China’s Statistical Yearbooks on Science and Technology from
1991 to 2003.
From 1990 to 2002, the total of patent applications increased significantly in most provinces. The
coastal provinces, which take amount of 75% of the total of patent applications in 2002, have the
fastest growing. In Gouangdong, Shanghai and Fujian, in particular, it increased by a factor of 17, 13
and 12 respectively. Comparing with the coastal provinces, the central and western provinces
experienced the slower expansion of patent applications about four fold larger between 1990 and 2002.
The distribution of patent applications across provinces, as shown in Table2, exhibits a large
discrepancy between the coastal region and the other inland (central and western) provinces.


5. RESULTS AND DISCUSSION

The model developed in section 3 was estimated for a panel of data for 29 provinces over the
period of 1990 to 2002.
The regression equations are shown in Table 3. Both domestic and foreign R&D variables are
statistically significant determinants to the domestic TFP .The coefficient on the trade weighted
foreign R&D variable
f TRADE
,
S
in eq(2)is smaller than the coefficient on
f ,FDI
S
in eq(4), and the R2 is
i,t
i,t
larger when f FDI
S ,
is used. Furthermore, in eq(6), when both variables are used , the point of
i,t
f FDI
S ,
is 0.193, larger than the point of f TRADE
S ,
which(0.075).We used an F-test for an upper one
i,t
i,t
tailed test8 to examine the statistical significance of the difference. The null-hypothesis was rejected,
which confirmed the estimation result.

8 The F hypothesis test is defined as: H0: ? = ? , H1: ? > ? (for an upper one tail test) Using
ff
ft
ff
ft
the full sample and the sample of eastern region, the hypothesis was rejected at p=0.001.

89

---

The International Journal of Economic Policy Studies
f*
f FDI
S ,
and m*
f TRADE
S ,
are foreign R&D variables weighted by openness. In Table 3, the
i,t
i,t
f ,FDI
f TRADE
,
estimation coefficient of f* S
and m* S
in equation (7) are positive and significant. This
i,t
i,t
suggests the higher the ratio of FDI in the fixed capital formation, or the higher the import share in
GDP the greater the technology transfer effect can be expected.
Since the FDI and
port is concentrated in the coast
im
al region in China, and the coastal region is the
most open area to the foreign countries, we would expect to observe greater spillover effect through
FDI and trade. So we estimated the equations reported in table 3 using only the observations of the
eastern 12 provinces.
The results presented in Table4 are similar to those in Table 3.The statistic analysis is the same and
the results provide some support for our expectation. The coefficient on f FDI
S ,
is quite larger than
i,t
that of
f TRADE
S ,
, which implies the FDI channel identifies larger spillover.
i,t
In particular, the estimated coefficient on f*
f FDI
S ,
and m*
f TRADE
S ,
are positive and statistically

i,t
i,t
significant, which implies that if an economy is more open to trade and FDI, greater technology
transfer effect can be expected.











6. CONCLUDING REMARS

Using the provincial data for the period of 1990 to 2002, we estimated the spillover effect through
Import and FDI. We found generally significant and positive spillover effect of foreign R&D stocks
through trade and FDI. So this study empirically supports that both FDI and Import generates
externalities in the form of technology transfer. In addition, the coefficient estimates for FDI are larger
than those of trade, which suggest comparing with import, FDI is supposed to be the more important
spillover channel in transferring the foreign R&D stocks. Furthermore, according to the estimation
the region with a high openness to the trade and FDI
results on eastern region observations,
demonstrates high spillover effect from the foreign R&D. In line with the findings of this study, it
90

---

Vol. 1 Oct. 2006
implied that China might be better off if the government devises policies to attract more investments
from technologically advanced countries.
This study is based on the macroeconomic data. If detailed and publicly available industrial data are
vailable, we
a
like to expand this study and to see whether these conclusions are further confirmed.


























FIGURES AND TABLES
91

---

The International Journal of Economic Policy Studies
Fig1 Realized FDI flows in China,1990-2002
600
500
Other countries
400
Italy
France
Germany
300
United Kingdom
Japan
USD 100million 200
Canada
United States
100
0
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002

Source: SSS, China statistics Yearbook,1991-2003


Fig2 FDI to China by Location(Realized Value,2002)
western region
3%
central region
10%
eastern region
87%

Source: SSS, China statistics Yearbook,2003



92

---

Document Outline

• ??
  • ??

---