Paper for the 16th International Conference on Input-Output Techniques
July 2-6, 2007, Istanbul, Turkey
Import Dependence of Foreign Trade: A case of China1
Yang Cuihong, Pei Jiansuo
Academy of Mathematics and Systems Science, the Chinese Academy of Sciences
Abstract: Since reform and opening-up to the outside world, China’s foreign trade
develops very rapidly, especially after China’s entry to the WTO in 2001. Foreign trade
dependency ratio of China (FTDR), which is the ratio of total volume of foreign trade to
GDP, has increased year by year, especially after 2001, from 38.7% in 1995 to 69.1% in
2006. While most of China’s exports are processing exports, intermediate materials of
which including raw materials, accessories and components are mostly imported. So
China’s exports have a high dependency ratio on imports, especially in Foreign-invested
Enterprises (FIE). In processing trade, the value-added China could obtain from exports
is much smaller than the total exports value would show.
In this paper, we adopted the definition of vertical specialization proposed by
Hummels, Ishii and Yi (2001) to depict the dependency ratio on imports of China’s
exports. Based on China’s non-competitive input-output table capturing processing
1
Yang Cuihong, Academy of Mathematics and Systems Science, the Chinese Academy of Sciences, No.
55, Zhongguancun Dong Lu, Beijing 100080, P. R. China. Email: chyang@iss.ac.cn; Pei Jiansuo,
Academy of Mathematics and Systems Science, the Chinese Academy of Sciences, No. 55,
Zhongguancun Dong Lu, Beijing 100080, P. R. China. Email: jspei@amss.ac.cn. The authors would like
to express their sincere thanks to late Professor Christian DeBresson for his valuable comments and
important suggestions in our collaboration on the book China’s Re-entry into the Global Economy:
Innovation, Learning, and Catching-up he edited, we drafted the chapter “China Foreign Trade
Evolution and its Share in World Trade”. His wisdom and patience will be fully remembered by us.
trade we compiled2, we give a modified version of Hummels, Ishii and Yi model (2001)
on vertical specialization. We then estimate the direct and total VS share both of
processing exports and non-processing export of 2002 and also make a brief comparison
with that of the United States.
The results show that, in 2002, the VS shares of processing exports of China are
very high, direct VS share and total VS share amounting to 66.6% and 71.4%
respectively. By contrast, those of non-processing export are much lower, with direct
VS share and total VS share at 24.4% and 36.7%. While compared to other countries,
China’s export have a high dependency ratio on imports, for example, the direct and
total direct VS share of the United States are only 6.7% and 10.3% in 2002. By sectors,
telecommunication equipment, computer and other electronic equipment sector,
petroleum processing sector are highly dependent on imported intermediate inputs.
Keywords: Vertical Specialization, Processing trade, China’s non-competitive
input-output table capturing China’s processing trade.
1. Introduction
Since reform and opening-up to outside world, China’s foreign trade develops very
rapidly, especially after China’s entry to the WTO in 2001. At the same time, foreign
trade dependency ratio of China (FTDR) has been increasing year by year. The
2
In a joint project by financial support from the Chinese University of Hongkong, with the collaboration
of the Chinese University of Hongkong, Hongkong University of Science and Technology, University of
California at Santa Cruz, the AMSS (Academy of Mathematics and Systems Science, the Chinese
Academy of Sciences) team headed by Professor Chen Xikang compiled non-competitive input-output
table of China for 2002. For comparison, we also constructed non-competitive input-output table of the
United States for 2002.
1
traditional definition of FTDR is the ratio of gross value of foreign trade to GDP, by this
method, China’s FTDR increase from 38.7% in 1995 to 69.1% in 2006.
Although the FTDR of China is seemingly high, this index cannot truly measure the
dependency ratio of Chinese economy on foreign countries, because FTDR of China has
been overestimated. One of the most important reasons is that processing trade account
for more than 50% of China’s foreign trade, for example, 57.4% in 2006. Since
processing trade mainly use imported raw material, spare and accessory parts to
manufacturing products and then export, processing trade doesn’t relate so closely to
domestic economy as non-processing trade. According to our estimate, for year 2002,
Chinese direct and total value-added induced by processing exports are 0.58 and 0.38
times that of non-processing exports, respectively (Lau, Chen, et.al 2006). The main
limitation of FTDR is that it is not necessarily related to policy- a country can distort
trade heavily, and still have a high trade dependency ratio (Edwards, 1998). For the case
of China, China has been encouraging the production of processing trade by many ways,
including free import taxes of intermediate inputs, while from this kind of trade, China
can only get a small portion of processing fee. Simply calculated FTDR obviously
exaggerates China’s dependency ratio on international market.
As mentioned above, processing trade is the main body of China’s foreign trade,
intermediate materials of which including raw materials, accessories and components
are mostly imported. So China’s exports have a high dependency ratio on imports,
especially in Foreign-invested Enterprises (FIE). In other words, international
production fragmentation has great impact on China’s foreign trade. This is one of the
2
most important characteristics of international trade - increasing interconnectedness of
production processes in a vertical trading chain that stretches across many countries,
with each country specializing in particular stages of a good’s production sequence
(Hummels, et. al, 2001). This phenomenon is labeled as several different terms by
different researchers (Feenstra and Hanson, 1996, 1997; Feenstra, 1998; Deardorff, A.,
1998; Hummels, et. al, 2001; Grossman and Helpman, 2005), for example, vertical
specialization, international fragmentation of production, international disintegration of
production, international outsourcing. China is now a major international outsourcing
destination, importing raw materials for intermediate inputs and exporting to other
countries after manufacturing in China, which constitutes a main part of China’s export.
In this paper, we adopt vertical specialization (VS) to illustrate the import dependency
of China’s foreign trade.
Balassa (1967) is probably the first researcher who use the term “VS”, then Findlay
(1978). Sanyal and Jones (1982), Dixit and Grossman (1982) established theoretic
models for VS and trade in the framework of relative factor endowment, emphasizing
that comparative advantages are the determinant of specialization. Regarding the
measurement of VS, Campa and Goldberg (1997) investigated increases in external
orientation in terms of industry export shares, import penetration, and imported input
use in production by input-output technique for the United States, Canada, the United
Kingdom and Japan. Hummels (2001) define the VS strictly in an input-output
framework and by using input–output tables, calculate the vertical specialization for 10
OECD countries and four emerging market countries. Dean, Fung and Wang (2007)
3
measure Chinese vertical specialization by employing China’s input-output table of
1997 and 2002 by sector, export destination, and input source. Liu and Wu (2005),
CCER(2006), Zhang and Sun(2006) investigate the degree of China’s vertical
specialization by adopting Hummel’s VS index and input-output technique or
econometric models.
Yet, recent literature on fragmentation has focused on developing and testing
various theories of the firm’s decision to fragment production across borders (Dean,
Fung and Wang, 2007). As mentioned above, a few studies attempted to measure
China’s vertical specialization by input-output techniques, but mainly are based on
competitive input-output tables and use proportional calculation to estimate import
matrix---assume that each commodity requires imports in proportion to their usage of
inputs. In addition, these work do not differentiate vertical specialization of China’s
processing trade and non-processing trade. Since production structure and source of
intermediate inputs for processing trade and non-processing trade are quite different,
their dependency ratio on imported materials differs too. In this paper, based on China’s
non-competitive input-output tables capturing processing trade for 2002 we compiled,
we study the import dependency of China’s total trade, processing trade and
non-processing trade, mainly by employing Hummels, Ishii and Yi (2001) measure of
vertical specialization.
The paper is organized as follows. The following section gives an overview of
China’s import dependency of foreign trade. The third section introduces basic
methodology and data preparation. The section after shows main results of this paper
4
and discusses their interpretation. The fifth section concludes.
2. Overview of China’s Import Dependency of Foreign Trade---the Case of FIEs
As mentioned in section 1, processing trade is the main body of China’s foreign
trade, especially in FIE the processing export of which account for a little less than 80%
of total processing export. Take processing export with imported materials (PIM) in FIE
as an example1(table 1), the dependency ratio on import during 1990 and 1992 was
usually over 90%. Since 1995, it decreases gradually but still higher than 60%, for
example, in 2004, the dependency ratio on import of PIM export in FIE was 66.43%.
From the perspective of commodities, more capital-intensive the commodities are,
higher proportion of imported material in processing trade.
Table 1 is here
Due to lack of data for some industries, here we can only discuss the reliance of FIE
processing trade on imported materials from domestic accessory ratio aspect. Import
dependency ratio of FIE high-tech enterprises is very high. Most of the raw materials,
accessories and components of FIE have to be imported. For example, among FIE
high-tech enterprises in Beijing, 73.9% of them mainly rely on imported materials; for
FIE high-tech enterprises in Shanghai, imported accessories and components from
overseas account for 56% of total; similar situation exists in Suzhou, and moreover
those domestic accessories are mainly low-grade or middle-grade products. The main
reasons of low domestic accessory ratio are that, on one hand, the technology and
1
Processing trade includes two types: processing with imported materials (PIM) and processing with
customer's materials (PCM), in which PIM is the main form, accounting for more than 70% of processing trade
value (for example in 2002, 73%).
5
quality of domestic products could not meet the requirements of FIE, and globalization
trend of high-tech industry also affects the adoption of domestic products of China. On
the other hand, current tax policies of China also go against domestic accessories.
Imported materials for processing trade are tax free, but if the enterprises buy domestic
products, they cannot get reimbursement of the tax included, thus FIE has no enough
enthusiasm to use domestic accessories. Some FIEs point out that for part of the
accessories they require, both the quality and price of China domestic products have
been very competitive. If tax reimbursement for export policy can be expanded to
intermediate inputs, the domestic accessory ratio will increase and at the same time
reduce the cost the FIE exported commodities.
Before the middle of 1990s, FIE regarded China as a low-grade technological
processing and assembly base, domestic accessory ratio was then very low. In recent
years, however, domestic accessory ratio of FIE increases notably. Especially after 2000,
percentage of home-made parts in FIE high-tech industry increase gradually. Take
mobile communication equipments for example (Jiang, 2002a), Motorola Company has
over 700 suppliers in China, purchased 7.5 billion RMB of domestic accessories and
components in 2000 and more than 12 billion RMB in 2001; In Ericsson Company,
localization degree of some mobile phones has been over 60%, accessories like charger
and battery have been completely localized in China, furthermore many accessories are
exported to overseas from China. In wireless base station, import substitution ratio
increased to 51% in 2000 from 5% in 1999; Nokia Company purchased over 1 billion
RMB domestic products of China.
6
Along with the increase of FDI, FIE has developed the production of accessories in
China through further investment, therefore increase the localization degree and
self-producing proportion of intermediate inputs in China. For example, in Japan-capital
enterprises, domestic accessory ratio in China was only 20% in 1992, while in
Southeastern countries up to 47% then. Till 2002, the domestic accessory ratio of
Japan-capital enterprises in China reached 49.6% (Jiang, 2002b), while at the same time
that of Southeastern countries do not change much.
3. Methodology and Data
3.1 Measure of Vertical Specialization by Hummels, Ishii and Yi (2001)
Hummels et. al(2001) proposed a method by using input-output technique to
measure vertical specialization, which has been widely accepted by international
economic circle. In their paper, vertical specialization (VS) is depicted by two indexes:
VS value and VS share.3 VS value is defined as the amount of imported intermediate
inputs in producing exported products, VS share is the ratio of imported intermediate
inputs to total exports value.
3.1.1 Direct VS share
For country k and sector i, VS value is defined as follows4:
 IM 
VSki   ki  Eki
 X ki 
(1)
Where, for country k and sector i, IM ki stands for imported intermediate inputs,
3
Hummels, Ishii and Yi (2001) give strict definition of vertical specialization (VS), emphasizing imported
intermediate inputs to produce exported products. That is to say, VS share is the ratio of imported intermediate inputs
to total exports value, not to total output value.
4 Formula (1)-(4) are cited and modified from Hummels, Ishii and Yi (2001)
7
Eki denotes exports, X ki stands for total output value. The first term of the right side
means contribution ratio of imported intermediate inputs to total output value, while
multiplying exports we obtain the imported content in exports.
VS for country k is simply the sum of VS across all sector i, VSk  VSki
i
VS share of total exports is as follows by formula:
VS
VSSk  k 
Ek
VS
E
ki
i
ki
(2)
i
Where VSS k stands for VS share of total exports. VSS k can be expressed as:
VSS k
VS

E
ki
i
ki
i
 VS E

E
ki
ki
i
ki
i
E
 ki

 E   VS  
   ki   ki  
i  Ek   Eki  
(3)
That is to say, VS share of total exports for country k is equal to the export-weighted
average of VS share of all sector exports for country k.
In input-output framework, VSS k can be expressed as formula (4) in matrix
notation:
VSSk  uAM E / Ek
(4)
where u is a 1n vector of 1’s. AM  (aijm )nn is the matrix of direct imported
input coefficients, aijm represents direct imports coefficient, i.e., imported products of
sector i that are directly consumed by one unit output of sector j. E stands for a n1
vector of exports. Ek stands for total exports for country k.
3.1.2 Total VS share
Before introducing total VS share, we firstly give a brief introduction of direct
imports coefficient and total imports coefficient (Lau, Chen and Yang et.al, 2006). Take
8
steel products for exports for example, to illustrate the concept of direct imported inputs
and total imported inputs of an exported product. In producing steel for exports, both
imported intermediate inputs and domestically produced intermediate inputs (e.g., steel)
are used. Direct imported inputs in steel production are pig iron, coke and so on. During
the production of pig iron and coke, except domestically produced iron ore and coal,
etc., imported iron ore and coal are consumed too, which is the first round of indirect
imported inputs. The consumption of imported materials in the production of iron ore
and coal is the second round of indirect imported products by steel production. There
are infinitely many rounds of indirect imports.
Denote B M as matrix of total imports coefficient, B M  (bijm )nn
n
bijm  aijm   bikm a kjd
(i, j=1, 2, …, n)
(5)
k 1
where aijd is domestic products of sector i that are directly consumed by one unit
output of sector j. Denote AD as matrix of direct domestic input coefficient. Equation
(5) can be expressed as matrix notation:
B M  AM  B M AD
(6)
B M  AM ( I  AD )1
Direct VS share of total exports illustrate the dependency ratio on imported
intermediate inputs directly used to produce one unit export of a certain sector. Total VS
share of total exports is dependency ratio on imported intermediate inputs directly and
indirectly used to produce one unit export of a certain sector. Based on equation (6), we
can obtain total VS share of total exports, TVSSk , as follows:
9
TVSSk  uAM (I  AD )1 E / Ek
(7)
3.2 Data Source
The primary data sources are China’s non-competitive extended input-output tables
capturing processing trade (abbreviated as China’s non-competitive EIOP table
hereinafter) for 2002 (Lau, Chen and Yang et.al, 2006). The table form of China’s
non-competitive input-output tables capturing processing trade is listed in table 2. For
comparison, we also constructed non-competitive input-output table of the United States
for 2002. From Bureau of Economic Analysis (BEA), U. S Department of Commerce,
we obtained U.S Make Table, Use Table and import matrix for 2002, based on which
and industry technology assumption we compiled non-competitive input-output table of
the U. S.
In China’s non-competitive EIO table, in order to capture China’s special production
structure of its extensive participation in global production, we divide China’s total
production into three parts(Lau, Chen and Yang et.al, 2006). The first one is production
for domestic use in China (D); the second one is processing exports production (P); the
third one is non-processing exports production and other production of FIEs (N). As
mentioned earlier, in China much of the exports are produced by FIEs. Most of the
products by FIEs are directly or indirectly exported, while some of their products are
also used for domestic consumption, investment and intermediate inputs in China,
which are listed as ‘other production of FIEs’. In our tables, we do not put FIEs’
products that are used domestically into production for domestic use D, because most of
other production of FIEs is composed of components, parts, materials of the exported
10
commodities and its input structure is similar to that of exports production.
Table 2 is here
3.3 New Formula to Estimate Total VS Share of Total Exports
In China’s non-competitive EIOP table, domestic intermediate inputs are divided
into nine sub-matrix: X DD , X DP , X DN ; X PD , X PP , X PN ; X ND , X NP and X NN ,
where X DD , X DP , X DN stand for matrices of products of D used as intermediate
inputs by D, P and N, respectively; X ND , X NP and X NN denote matrices of products of
N used as intermediate inputs by D, P and N, respectively5. The matrices of products of
P used as intermediate inputs by D, P and N, X PD , X PP , X PN are assumed zero.
Thus AD in traditional non-competitive input-output table should be transformed
as follows:
 ADD

A  0
 AND

_
D
ADP
0
ANP
ADN 

0 
ANN 
where ADD  [ AijDD ]  [ X ijDD / X Dj ] , ADP  [ AijDP ]  [ X ijDP / X Pj ] ,
ADN  [ AijDN ]  [ X ijDN / X jN ] , AND  [ AijND ]  [ X ijND / X Dj ] ,
ANP  [ AijNP ]  [ X ijNP / X Pj ] , ANN  [ AijNN ]  [ X ijNN / X jN ] ,
X ijDD , X ijDP , X ijDN are delivery of domestic products of sector i of D to sector j of D,
P and N, respectively; X ijND , X ijNP , X ijNN are delivery of domestic products of sector i of
N to sector j of D, P and N, respectively. X Dj , X Pj , X jN are the gross output value of
sector j of D, P and N, respectively.
5
For simplicity, we then use D, P and N to express Production for domestic use, Processing exports, Non- processing
exports and other production of FIEs in the following text.
11

Correspondingly, AM   AMD , AMP , AMN  , the element of which AMD , AMP , AMN
are direct imports coefficient of D, P and N, respectively; AMD  [ AijMD ]  [ X ijMD / X Dj ] ,
AMP  [ AijMP ]  [ X ijMP / X Pj ] , AMN  [ AijMN ]  [ X ijMN / X jN ]
X ijMD , X ijMP , X ijMN are the delivery of imported products of sector i to sector j of D, P
and N, respectively.
The direct VS share of D, P, N, i.e., VSS kD , VSSkP , VSS kN , can be obtained from
formula (8):



VSSk  VSSkD , VSSkP , VSSkN   u AM E / Ek
(8)

where E is a diagonal matrix, the element of the diagonal E D , E P , E N , are n1

vector of exports of D, P and N, respectively; Ek is a diagonal matrix, the element of
the diagonal EkD , E kP , EkN , are total exports of D, P and N, respectively. Thus:
VSSkD  uAMD E D / EkD
VSSkP  uAMP E P / EkP
(9)
VSSkN  uAMN E N / EkN
_
We then obtain (I  AD )1 as shown in formula (10)6.
( I  ADD )

(I  A )  
0
  AND

_
D 1
 ADP
I
A
NP
  B DD
  PD
0
  B
NN 
( I  A )   B ND
 ADN
B DP
B PP
B
NP
B DN 

B PN 
B NN 
where:
B DD  ( I  ADD )1  ( I  ADD )1 ADN B NN AND ( I  ADD )1
B DP  ( I  ADD )1 ADP  ( I  ADD )1 ADN B NN [ ANP  AND ( I  ADD )1 ADP ]
B DP  ( I  ADD )1 ADP  ( I  ADD )1 ADN B NN ANP
6
Please see Lau, Chen and Yang et. al (2007) for detailed proof.
12
(10)
B DN  ( I  ADD )1 ADN B NN
B PD  0 ,
B PP  I , B PN  0
B ND  B NN AND ( I  ADD )1
B NP  B NN [ ANP  AND ( I  ADD )1 ADP ]
B NN  [ I  ANN  AND ( I  ADD )1 ADN ]1
B DD , B DP and
B DN denote matrices of total requirement coefficient of production
D per unit of final demands of D, P and N, respectively. B PD , B PP and
B PN denote
matrices of total requirement coefficient of production P per unit of final demands of D,
P and N, respectively. B ND , B NP and
B NN denote matrices of total requirement
coefficient of production N per unit of final demands of D, P and N, respectively.
Then we obtain the new formulae of total VS share of total exports based on China’s
non-competitive EIO table, shown in formula (11)
_
_


TVSSk  u AM (I  AD )1 E / Ek
(11)
where TVSSk  TVSSkD , TVSSkP , TVSSkN  , in which TVSS kD , TVSS kP , TVSS kN are the
total VS share of D, P, and N.
From equation (11), we get:
TVSSkD , TVSSkP , TVSSkN  
u( AMD B DD  AMN B ND ) E D / EkD , u( AMD B DP  AMP  AMN B NP ) E P / EkP ,
N
u( AM D B D 
N

A M NB )N E
/ kN E
N
Thus, we can obtain the formulae of total VS share for D, P and N shown as
13
formulae (12)-(14):7
TVSSkD  u ( AMD B DD  AMN B ND ) E D / EkD
(12)
TVSSkP  u( AMD B DP  AMP  AMN B NP ) E P / EkP
(13)
TVSSkN  u( AMD B DN  AMN B NN ) E N / EkN
(14)
4. Empirical Analysis
Based on section 3.3 and China’s non-competitive EIOP table for 2002, we calculate
the China’s VS share for both processing exports and non-processing exports. The
results are listed in table 3.
Table 3 is here
4.1 Aggregate VS Share
From Table 3, we can see that the direct VS share of processing export is quite high,
the ratio of intermediate import content to China’s processing export is 66.6%. Total
intermediate import content reached 71.4% of total export, a little higher than its direct
share. Regarding non-processing export, the direct VS share is 24.4%, much lower than
that of processing export. Unlike processing export, however, its total VS share is 1.5
times its direct VS share. Non-processing export and others use more domestically
produced products, which in turn consume lots of rounds of imported intermediates.
Processing exports, however, consume only a low ratio of domestically produced
products, most of their intermediate input are imported from other countries, the chain
effect of which do not occur in China but in other countries, this can partly explain why
7
Because the export of the production for domestic use is zero, so the total VS share of D is equal to zero.
14
the total VS share of processing export is only a little larger than its direct VS share.
4.2 VS Share of Processing Exports by Sector
Table 3 shows that sixteen manufacturing sectors have VS share higher than average
both for direct and total (figure 1).
As mentioned earlier, telecommunication equipment sector of processing export has
very high import dependency ratio, reaching 82.2% of the value of total exports. That of
petroleum processing ranks the second (73%). Seven of the sixteen manufacturing
sectors have direct VS share larger than average, the other five are: metals smelting and
pressing (69.2%), metal products (73.8%), common and special equipment (69.4%),
transport equipment (69.0%) and, electric equipment and machinery (72.4%). Most of
these sectors belong to electromechanical products, which have a relatively higher
value-added ratio. From figure 1, we can see that the intermediate inputs for high
value-added products still rely heavily on imports. Besides manufacturing sectors, other
sectors with relatively high direct and total VS shares included textile goods, chemicals,
other manufacturing products, construction, renting and commercial service.
Figure 1 is here
4.3 VS Share of Non-processing Exports by Sector
Figure 2 is the chart of direct and total VS share of non-processing export by sector.
Petroleum processing has the highest VS share (both direct and total).8
8
Although China's petroleum import is just nearly 6% of the world's trade volume, China’s dependency ratio of
imported petroleum in the past several years keep increasing, the ratio of imported petroleum to China’s total
petroleum consumption is only 29.1% in 2001, while has risen to 47.3% in 2006.
15
Telecommunication equipment, computer and other electronic equipment sector follows
(with direct VS share 51.1% and total VS share 62%). The other sectors with higher
total VS share include transport equipment (53.6%), common and special equipment
(52.8%), electric equipment and machinery (51.8%), chemicals (51%), metals smelting
and pressing (47.1%), metal products (45.8%) and, renting and commercial service
(43.5%). Different from processing exports, several sectors almost do not directly use or
a very small portion of imported materials in their production, while their total VS are
relatively high, for example, scientific research, health service, social guarantee and
social welfare and so forth.
Figure 2 is here
Since processing trade accounts for more than a half of China’s foreign trade,
China’s import content of exports is quite high compared with other countries. Based on
the non-competitive table of the United States for 2002 we compiled, we estimate that
the direct VS share of the U.S. in 2002 is 6.7% and 10.3% for total VS share.
5. Conclusion
Different from developed countries and most developing countries, processing trade
plays a very important role in China’s trade development. Because most of the
intermediate inputs for processing exports are imported, China’s export has a high
dependency ratio on imports. Although some researchers addressed the issue of import
dependency ratio of China’s export by using econometric model and input-output model
for example VS share based on Hummels, Ishii and Yi theory (2001), but only consider
16
the VS share of China’s total export. In this paper, based on China’s non-competitive
input-output table capturing processing trade, we give a modified version of Hummels,
Ishii and Yi model on vertical specialization. We then estimate the direct and total VS
share both of processing exports and non-processing export of 2002 and also make a
brief comparison with that of the United States.
The results show that, in 2002, the VS shares of processing exports of China are
very high, direct VS share and total VS share amounting to 66.6% and 71.4%
respectively. By contrast, those of non-processing export are much lower, with direct
VS share and total VS share at 24.4% and 36.7% respectively. While compared to other
countries, China’s export have a relatively high dependency ratio on imports, for
example, the direct and total direct VS share of the United States are only 6.7% and
10.3% in 2002. By sectors, telecommunication equipment, computer and other
electronic equipment sector, petroleum processing sector are highly dependent on
imported intermediate inputs. Most of the sectors with high VS share of exports are
basically electromechanical products, which are usually considered as high value-added
products, showing that China is still just an assembly base, not an independent producer,
of so-called high-technology products.
Due to data limit, in this paper, we only estimate the VS share for processing exports
and non-processing exports of 2002. We are planning to compile China’s
non-competitive input-output tables both for 2005 and 2007, and based on which
estimate the VS share for processing exports and non-processing exports over time.
17
References
Balassa, B. (1967) Trade Liberalization Among Industrial Countries(New York,
McGraw-Hill).
Campa Jose and Goldberg Linda S.(1997) The Evolving External Orientation of
Manufacturing Industries: Evidence from Four Countries, No 5919, NBER Working
Papers from National Bureau of Economic Research
CCER Team(2006) Vertical Specialization of China’s Export and Sino-US Trade, The
Journal of World Economy (Shi Jie Jing Ji), 29(4), 3-11 (in Chinese)
Deardoff, Alan (1998) Fragmentation in Simple Trade Models, Discussion Paper no.422,
Research Seminar in International Economics, University of Michigan, January 7.
Dean, Judith M., Fung, K.C. Fung and Wang, Zhi (2007) Measuring the Vertical
Specialization in Chinese Trade, Working Paper, U.S. International Trade Commission
Dixit, Avinash K, Gene M. Grossman(1982). Trade and Protection with Multistage
Production, Review of Economic Studies, 49(4), 583-594.
Edwards S. (1998), Openness, Productivity and Growth: What do we really know?
Economic Journal 108, March, 383-398.
Feenstra, R.C., Hanson, G.H.(1996). Foreign investment, outsourcing and relative wages.
In: Feenstra, R.C., Grossman, G.M., Irwin, D.A. (eds.), The Political Economy of
Trade Policy. Papers in Honor of Jagdish Bhagwati (Cambridge, MA, MIT Press)
89–127.
18
Feenstra, R.C., Hanson, G.H.(1997). Foreign direct investment and relative wages:
evidence from Mexico’s maquiladoras. Journal of International Economics 42,
371–394.
Findlay, R.(1978). An Austrian model of international trade and interest rate
equalization. Journal of Political Economy, 86, 989–1008.
Grossman, G.M., Helpman, E.(2005). Outsourcing in the global economy, Review of
Economic Studies 72, 135–159
Hummels, David, Jun Ishii and Kei-Mu Yi (2001) The Nature and Growth of Vertical
Specialization in World Trade. Journal of International Economics, 54, 75-96.
Jiang Xiaojuan (2002a), China: Transfer from processing and assemble base to
manufacturing base, International Trade (Guo Ji Mao Yi ), No. 3, 44-47(in Chinese)
Jiang Xiaojuan (2002b), Export growth and structural change: contribution of FIE,
Nankai Economic Research (Nan Kai Jing Ji Yan Jiu), No. 2, 30-34(in Chinese)
Lau, L. J., Chen Xikang, Cheng, L. K., Fung, K. C., Pei Jiansuo, Sung, Y.-W., Tang
Zhipeng, Xiong Yanyan, Yang Cuihong and Zhu Kunfu (2006). The Estimation of
Domestic Value-Added and Employment Generated by U.S.-China Trade, Working
Paper No. 295, Stanford Center for International Development, Stanford
University
Lau, Lawrence J., Chen Xikang, Yang Cuihong, Cheng, Leonard K., Fung, K.C., Sung,
Y. W,
Zhu Kunfu, Pei Jiansuo and Tang Zhipeng( 2007) Extended Input-Output Model
19
with Assets in Non-competitive Imports Type capturing
China’s Processing Trade and
Applications---Perspective of U. S.-China Trade Balances, Social Sciences in China
(Chinese version), forthcoming in No.4.
Liu Zhibiao, Wu Fuxiang (2005) Disintegration of Production in the Globalized
Economy—An Empirical Study Based on Input-output Tables of Jiangsu Province,
Zhong Guo Gong Ye Jing Ji, No. 4, 12-19(in Chinese)
Sanyal, Kalyan K, Ronald W. Jones (1982) The Theory of Trade in Middle Products,
American Economic Review, 72(1), 16-31.
Wang Hongqing, Lv Likang (2005), Inward Foreign Direct Investment and China’s
Processing Trade, Reform (Gai Ge), No. 7, 77-81(in Chinese)
Zhang Xiaodi, Sun Jingwei (2006). Analysis on China’s International Competitiveness
based on Vertical Specialization, Journal of World Economy (Shi Jie Jing Ji), 29(5),
3-11 (in Chinese).
20
Table 1 Processing trade with imported materials by FIE (million USD, %)
Year
Export of processing
trade with imported
materials
Import of processing
trade with imported
materials
Dependency
ratio on
import*
1990
1991
1992
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
6368
9879
14446
39177
48595
57663
61958
64161
84100
92252
117963
169219
236415
5741
9143
13279
34409
37757
42898
43211
45283
58894
59541
81523
115697
l57039
90.15
92.55
91.92
87.83
77.70
74.39
69.74
70.58
70.03
64.54
69.11
68.37
66.43
Value of
domestic
accessories**
533
626
992
4053
9212
12550
15935
16046
21425
27804
30974
45494
67470
Source: Calculated by the author according to Wang & Lv, 2005
* Dependency ratio on import=Import of processing trade with imported materials / Export of processing trade with
imported materials
** Value of domestic accessories =( Export of processing trade with imported materials－Import of processing trade
with imported materials)- Profit - Cost of non-raw materials; Profit=( Export of processing trade with imported
materials－Import of processing trade with imported materials)*5%; Cost of non-raw materials=( Export of
processing trade with imported materials－Import of processing trade with imported materials)*l0%.
21
Table 2: China’s non-competitive input-output tables capturing processing trade
Intermediate Use
Final Use
Gross
Production
for
Non- processing
Processing
Gross
Exports and
Exports
Total
Other Production
Domestic
Capital
Production for
Domestic Use (D)
Domestic
Processing
-ally
Exports
Interme-d
(P)
diate
Non-processing
Inputs Exports and Other
Production of
FIEs
Exports Others
Formation
of FIEs
Output
Final
or
Use
Imports
ption
(P)
Use (D)
Total
Consum-
(N)
1,2,…, n
1,2,…, n
1,2,…, n
X DD
X DP
X DN
F DC
F DI
0
FD
XD
0
0
0
0
0
F PE
FP
XP
X ND
X NP
X NN
F NC
F NI
F NE
FN
XN
X MD
X MP
X MN
F MC
F MI
FM
XM
VD
VP
VN
XD
LD
XP
LP
XN
LN
(N)
Intermediate Inputs from
Imports
1
:
n
Total Intermediate Inputs
Value-added
Total inputs
Employees
NOTE: The superscript D, P, N and M denote domestic products, processing exports, non-processing exports and
imports respectively. The superscript DD stands for domestic products used by domestic use, DP domestic products
used by processing exports, DN means domestic products used by non-processing exports and others, and so forth.
22
Table 3. VS Share based on China’s Non-competitive Extended Input-output Table
capturing Processing Trade for 2002
Processing Exports
(P)
IO
code
01
02
03
04
05
06
07
08
09
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
Agriculture
Coal mining, washing and processing
Crude petroleum and natural gas products
Metal ore mining
Non-ferrous mineral mining
Manufacture of food products and tobacco
processing
Textile goods
Wearing apparel, leather, furs, down and related
products
Sawmills and furniture
Paper and products, printing and record medium
reproduction
Petroleum processing, coking and nuclear fuel
processing
Chemicals
Nonmetal mineral products
Metals smelting and pressing
Metal products
Common and special equipment
Transport equipment
Electric equipment and machinery
Telecommunication equipment, computer and other
electronic equipment
Instruments, meters, cultural and office machinery
Other manufacturing products
Scrap and waste
Electricity and heating power production and supply
Gas production and supply
Water production and supply
Construction
Transport and warehousing
Post
Information communication, computer service and
software
Wholesale and retail trade
Accommodation, eating and drinking places
Finance and insurance
23
Non-processing
Exports (P)
Direct
Total
Direct
Total
0.439
0.449
0.451
0.542
0.489
0.495
0.514
0.498
0.611
0.567
0.094
0.152
0.129
0.255
0.242
0.174
0.273
0.246
0.409
0.367
0.507
0.566
0.111
0.190
0.639
0.697
0.199
0.272
0.593
0.661
0.198
0.282
0.583
0.651
0.180
0.316
0.540
0.614
0.206
0.340
0.730
0.775
0.684
0.732
0.628
0.551
0.692
0.738
0.694
0.690
0.724
0.711
0.628
0.738
0.777
0.747
0.755
0.772
0.349
0.248
0.281
0.232
0.371
0.326
0.344
0.510
0.395
0.471
0.458
0.528
0.536
0.518
0.822
0.842
0.511
0.620
0.606
0.620
0.000
0.367
0.358
0.431
0.659
0.460
0.407
0.641
0.678
0.000
0.452
0.482
0.521
0.720
0.538
0.508
0.363
0.187
0.000
0.000
0.222
0.208
0.252
0.164
0.193
0.510
0.327
0.000
0.262
0.373
0.330
0.426
0.334
0.322
0.571
0.604
0.262
0.365
0.410
0.433
0.265
0.484
0.501
0.351
0.172
0.100
0.000
0.266
0.193
0.188
33
34
35
36
37
38
39
40
41
42
Real estate
Renting and commercial service
Tourism
Scientific research
General technical services
Other social services
Education
Health service, social guarantee and social welfare
Culture, sports and amusements
Public management and social administration
0.000
0.632
0.054
0.000
0.000
0.501
0.330
0.000
0.385
0.295
0.001
0.670
0.188
0.001
0.001
0.569
0.418
0.001
0.461
0.383
0.000
0.327
0.086
0.000
0.000
0.200
0.113
0.000
0.178
0.110
0.175
0.435
0.157
0.412
0.246
0.319
0.222
0.406
0.285
0.212
Total
0.666
0.714
0.244
0.367
24
Textile goods
Wearing apparel,
leather, furs, down
Sawmills and
furniture
Paper and products,
printing and record
Petroleum
processing, coking
Chemicals
Nonmetal mineral
products
Metals smelting and
pressing
Metal products
Common and
special equipment
Transport
equipment
Electric equipment
and machinery
Telecommunication
equipment,
Instruments,
meters, cultural and
Other
manufacturing
0.2
Manufacture of
food products and
Figure 1 Direct and total VS share of processing trade by manufacturing sectors
DVSS
T VSS
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
06
07
08
09
10
11
12
13
14
15
16
17
18
19
20
21
25
0.0
Agriculture
Coal mining,
Crude petroleum
Metal ore mining
Non-ferrous
Manufacture of
Textile goods
Wearing apparel,
Sawmills and
Paper and products,
Petroleum
Chemicals
Nonmetal mineral
Metals smelting and
Metal products
Common and
Transport
Electric equipment
Telecommunication
Instruments,
Other
Electricity and
Gas production and
Water production
Construction
Transport and
Post
Information
Wholesale and
Accommodation,
Finance and
Real estate
Renting and
Tourism
Scientific research
General technical
Other social
Education
Health service,
Culture, sports and
Public management
Figure 2 Direct and total VS share of non-processing trade by sectors
0.8
DVSS
26
TVSS
0.7
0.6
0.5
0.4
0.3
0.2
0.1
01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42