The politics of Japan-China trade and the role of the world trade system
Shiro Patrick Armstrong
Crawford School of Economics and Government
Australian National University
Abstract
The large and rapidly growing trade relationship between Japan and China has occurred
against a backdrop of political tensions. This study measures performance of the trade
relationship, benchmarking it against other trade flows worldwide, and examines the impact
of the politics on bilateral trade performance. In order to do this a frontier gravity model is
estimated using core determinants of trade. This gives a measure of trade performance, which
is explained using resistances to trade. While the economic relationship is not independent
from the politics an important conclusion is that trade has not been diminished or disturbed
by politics to a significant extent. China’s commitment to the global trading system from the
mid 1980s and its accession to the WTO in 2001 have meant that tensions in the political
relationship with Japan from time to time have not derailed, but rather have increasingly
come to be dominated, by the economic relationship.
Keywords: International trade; Sino-Japan relations; Frontier gravity model; Politics and
trade.
Background
The Japan-China relationship is one of the most important bilateral relationships in the world
in terms of its impact on economic welfare, security and peace in the East Asian region and
the world more broadly. The economic relationship between the two countries is large and
significant and has seen unprecedented growth in trade and foreign direct investment (FDI) in
the last three decades. Starting from a very low base in 1978, by 2007 trade between Japan
and China was the third largest bilateral merchandise trade relationship in the world, in terms
of exports and imports together, behind the United States-Canada and United States-China
trade relationships. However, the Japan-China economic relationship is also one that is
complicated by political and historical tensions1.
From Japan’s perspective, the economic engagement with China has been important in recent
years for recovery from the economic downturn of the 1990s, with China a major factor in
Japan’s large-scale externally-driven growth. But the antecedents of the relationship lie in the
deep complementarity in the structure of the two countries’ economies and how it has
changed over time. The relationship with Japan has been important for China’s ‘catching up’
to the frontier of industrial technology and its transition from a low to a higher income
country through the access that it has provided to industrial technology and investment. The
bilateral relationship also has implications for the rest of the region, because of the extensive
production networks in which both economies are involved, and for the broader global
economy, as these are the second and third largest economies in the world.
The growing trade between the two economies has occurred despite the political sensitivities
that are primarily due to unresolved historical issues but also because of regional rivalry.
1
See Whiting (1992) for a discussion of the coexistence of historical tensions and the economic relationship.
2
Politics can severely affect economic relations and history is littered with large-scale conflicts
between trading nations which damage or even lead to the cessation of economic
transactions. Contemporary examples are the India-Pakistan, United States-Cuba and North
and South Korea relationships. This paper asks whether the vagaries of political distance have
affected the economic relationship between Japan and China and discusses how the economic
relationship has been able to prosper despite the political distance. Political distance is a
measure of how ‘close’ two countries are politically or geopolitically and how well they get
along.
In order to answer the question the paper estimates performance of the relationship by
estimating a counterfactual. This is done by estimating trade potential, or a trade frontier,
which allows a measure of the economic distance, or unmeasurable costs, in trade
relationships.
The next section discusses how political relations can affect trade. Then the frontier gravity
model is introduced and estimated. The paper then proceeds to discuss how, in the
relationship between Japan and China, the economics has managed to dominate the politics,
and not vice versa.
How politics can affect economics
An important part of the story of the Japan-China relationship is the uneasy and sometimes
uncertain geopolitical relationship which cannot be ignored in an analysis of the economic
relationship. These tensions have at times looked as if they might threaten to derail the
economic ties. From 2001 to 2006 leadership visits, or state visits, were suspended between
3
the two because of political tensions. This period included large scale protests in China
against Japan and Japanese interests in 2005, including a boycott of Japanese products2.
Two countries which are political and security allies can be described as being close in terms
of political distance, whereas two nations that are political rivals can be described as
politically distant. Between these two extremes there is a wide range of degrees of closeness
and distance in the relationships between traders and investors. There is a well established
literature that examines how politics affects trade (see Hirschman, 1945; Polacheck, 1980).
This study includes a measure of bilateral political distance in extending the gravity model to
examine the impact of political relations on economic relations.
The literature that focuses on the interaction between political distance and trade has shown
that it is a complicated interaction. While the debate has raged for centuries, the empirical
findings show that causality runs both ways between trade and politics (Polacheck, 1980;
Pollins, 1989a; 1989b), the lag lengths of interactions vary depending on the country pairs
involved (Gasiorowski and Polachek, 1982; Reuveny and Kang, 1996), the interactions are
country and country pair specific and that these relationships can change over time
(Mansfield and Pollins, 2003).
Political distance between countries causes uncertainty to increase and acts as a resistance to
trade or investment. Although there are no extreme events between Japan and China in the
contemporary period analysed, such as war or the establishment of a security alliance, there
were significant and prolonged low intensity conflicts as well as positive political
2
While the situation is much more complex with protests not solely aimed at or motivated by Japan, this study
does not focus on the politics but instead takes it as a given and analyses the effect on economics.
4
developments fostering cooperation in the relationship that affected the economic
relationship.
The political distance variable is set out in Figure 1 and shows Japanese ‘sentiment’ towards
China, separated into both positive and negative elements, based on event coding from
newspaper articles. The data are from King (2003) and are updated to 2004 data in the new
IDEA dataset. The scale on the vertical axis is an index (and a relative measure of conflict
and cooperation in political relations). There is an established literature which employs, tests
and develops such event data (Mansfield and Pollins, 2003).
The net effect of the political positives and negatives are difficult to determine from Figure 1
alone. As in many conflict and cooperation event data, events are weighted according to a
scale to reflect the severity and significance of events. Using the Goldstein (1992) weighting
of events, in these data from King (2003), for example, Figure 1 shows China’s WTO
accession in 2002 offsetting rising negative political sentiment with positive economic
sentiment. The positive news dominates negative political events in the period following
WTO accession.
Figure 1 Japanese political distance towards China, 1990-2004
5
Note: Measurements of political distance are from King (2003). Negative events are a 6 month moving average
and positive events are a 12 month moving average.
Source: King (2003)
The negative events are subtracted from the positive events to obtain a measure of net
political closeness, a measure commonly used in the literature (Schneider et al., 2003). As in
utility theory the assumption here is that positive events cancel out negative events. Therefore
a movement in a positive direction for the political variable indicates political closeness and a
change in a negative direction indicates widening political distance. A movement in a
positive direction implies a narrowing political distance. Political distance from both Japan
and China’s perspectives will be used in the model described next.
The frontier gravity model
In order to assess whether a bilateral trade relationship is affected by politics, it is necessary
to measure its performance and then estimate the effect of politics alongside other resistances
to trade. A traditional gravity model will estimate the mean effects of all determinants and
give a measure of whether an observation is performing above or below what the model
predicts. This method, while useful, will not identify whether it is the politics or another
factor that is the reason for underperformance in a particular relationship. However, with a
two stage estimation using stochastic frontier analysis, the first stage can estimate a frontier
and give a measure of performance while the second stage can explain that performance. In
the first stage, core determinants of scale (GDPs), geographic distance (including common
border measures) and a measure of complementarity are used, including controlling for
multilateral resistances. In the second stage, all other resistance variables that are commonly
6
found in gravity models, such as regional trade agreements, a measure of risk, tariff and
language similarity variables are included. To that list we add a measure of political distance.
The logic is that if a trade flow is high relative to its potential, as defined by the frontier, its
performance is high given the size of the trading countries, the distance they are apart, the
complementarities of their economic structures and controlling for the influence of third
countries. What explains the high levels of trade and investment performance, then, may be
similarity of language, membership of the same FTA or RTA, low border barriers, and other
factors.
The performance of trade relationships can also be thought of as a measure of economic
distance where, since geographic distance is already controlled for, high performance shows
low resistance to goods flows between countries. The most liberal and free flowing trade
relationships are characterised by low economic distance and help to define the frontier.
Stochastic frontier analysis combined with the gravity model of trade is a little used method
but has been shown to produce sensible results (Drysdale, Huang and Kalirajan, 2000) and its
use has recently been given conceptual justification (Armstrong, 2007; Kalirajan, 2008).
Stochastic frontier analysis was first used for analysing agricultural output (Aigner et al.
1977; Meeusen and van den Broeck, 1977). Its application is also prevalent in social
efficiency literature where a government’s ability to deliver social services to those most in
need are analysed.3 Applying stochastic frontier analysis to the gravity model may be more
applicable than its application to measuring aggregate social efficiency, as empirical trade
gravity models have theoretical foundations whereas the models of social indicators do not. It
3
See Ravallion (2003) for a review of stochastic frontier analysis used in measuring aggregate social efficiency.
7
may therefore be more justifiable to impose and extract a latent ‘inefficiency’ feature from a
theoretical gravity model than from a social indicator model.
It is understood in the literature that most trade costs are difficult to quantify or measure
(Anderson and van Wincoop, 2004). Anderson (1979) and Anderson and van Wincoop
(2003) show that in a properly specified gravity model with theoretical underpinnings, it is
economic distance that determines trade, not geographic distance. In Anderson van Wincoop
(2003), and many other gravity models, all trade costs, or economic distance, are proxied by
geographical distance or, as Baldwin and Taglioni (2006) put it, distance and ‘other stuff’.
The frontier gravity model separates out the ‘other stuff’ into a one sided error term that
captures resistances from reaching potential.
The trade frontier is estimated using the following model.
ln xijt = ln β0 + β1 ln y it + β2 ln y jt + β3 ln rDist
(1)
ij
+ β4 Border
ij
+ β5 COMP ijt + vijt − u ijt
Table 1 includes detail about these variables and the sources of data for their measurement.
Table 1 Variable description and data sources
Variable
xijt
Description
Data source
Notes
trade from i to j at time t
IMF’s Direction of Trade
Calculated from imports
Statistics (various years) and
instead of exports for
gaps in the data are filled in
accuracy*.
from the International Economic
Databank (IEDB)
yit
ij
rDist
Country i’s size (GDP) at
World Development Indicators
time t
(WDI) and at current prices
Relative distance from i to
Great circle distance between
j
capital cities of each country
rDist ij =
Dist ij
∑Dist
k≠j
ik
∑Dist
k ≠i
8
jk
was collected from the Chemical
Ecology of Insects website:
http://www.chemicalecology.net/
Border
ij
Variable that takes on the
value of one if i and j
share a common land
border, zero otherwise.
ijt
COMP
Complementarity index of
International Economic
i’s trade with j at time t.
Databank (IEDB), Australian
 X ik M w − M i M kj 

C ij = ∑ 
⋅ k
⋅
k


X
M
M
−
M
k 
i
j 
w
i
National University
[see below]
Notes: * Importers have less incentive to under-report and imports are a more accurate reflection of trade flow values than reported
exports. The exception is European trade where there is tax incentive to under report imports due to the value added tax structure but import
flows were used for consistency. This is common practice.
The complementarity index used here is from Drysdale (1967) and Drysdale and Garnaut
(1982):
 X ik M w − M i M kj 

C ij = ∑ 
⋅ k
⋅
k


X
M
Mw − Mi
k 
i
j 
where X is exports, M is imports, subscripts denote country (i, j and world) and superscript k
implies commodity k. The index is calculated at the three digit level from the Australian
National University’s International Economic Databank4 for all combinations of countries
and years. The index captures the complementarity of trade structures between countries and
a higher index implies a higher degree of complementarity.
vijt is an independently and identically distributed normal variable with mean zero and
variance σv2 and uijt is an independent and identically distributed non-negative variable which
4
http://iedb.anu.edu.au/
9
usually has a half normal, truncated normal or exponential distribution (Kumbhakar and
Lovell, 2000).
The disturbance term vijt accounts for random variation in trade similar to the disturbance
term in the standard OLS model. The non-negative (or one sided) disturbance term, uijt,
measures the difference between potential trade and actual trade. More precisely, it is the
amount of trade that falls short of the frontier for trade from country i to j at time t.5
The trade model is estimated for an unbalanced panel between 1980 and 2006. The data
includes a representative sample of world trade with the bilateral trade flows of 65 countries
by 65 countries. The countries are listed in the Appendix. Some bilateral flows are missing
for some years due to the unavailability of data but given the large numbers of observations,
the data represent a relatively complete and balanced panel.
The stochastic frontier method is one way of accounting for multilateral resistances which
were shown by Anderson and van Wincoop (2003) to be crucial to the model being consistent
with theory and a necessary explanator of trade. While the literature uses dummy variables
and fixed effects estimation across varying dimensions (for example, country pair specific
fixed effects or country-time specific fixed effects) (Baldwin and Taglioni, 2006), Kalirajan
(2008) explains that stochastic frontier analysis is one method of controlling for the
heterogeneity often found in gravity models and often partially controlled for with the fixed
effects. Indeed, Greene (2004) shows that heterogeneity across the dependent variable
observations, in this case country pairs, is captured in the inefficiency term, uijt. Fixed effects
estimation using panel data is a trivial extension of stochastic frontier models (Greene, 2005).
The difficulty with the fixed effects approach is that depending on the dimensions that need
5
For a detailed technical description of the estimation procedure see Coelli (1996) and Kumbhakar and Lovell
(2000).
10
to be fixed and the larger the data set, the more computationally demanding it becomes with
large numbers of dummy variables. Also, with simpler models the fixed effects are absorbed
in the intercept term so that the result is a firm (or dependent variable unit) specific intercept6
(Kumbhakar and Lovell, 2000).
Instead of using fixed effects (dummy variables), with the stochastic frontier method, the
variables of interest such as the effect of China’s WTO accession on a bilateral relationship
(in this case the Japan-China relationship) will be ‘forced into the inefficiency term’ (Greene,
2005). Greene’s hint (2004; 2005) that the unobserved variation across the dependent
variable unit is captured in the inefficiency term allows estimation of the frontier variables
without bias stemming from multilateral resistances. The choice then becomes whether the
dependent variable unit is across the country pair dimension or the time-varying country pair
dimension. This study applies time varying country pair fixed effects (µijt) to account for
multilateral resistances.
Table 1 shows results for ordinary least squares estimation in column (1). Column (2) is the
model estimated over the time-invariant country pair dimension and Column (3) is the same
model as Column (2) with additional time dummy variables which are not presented in the
results (to save space).
All coefficients are statistically significant at the 1 per cent level and the signs are as would
be expected. The larger two countries are, the more they trade and the further they are apart,
the less they trade. A complementary trade structure with a partner helps explain an increase
6
Another problem with this approach is that it requires one country (or firm in the original literature) to lie on
the frontier (efficiency of 100 per cent) and the rest of the observations are calculated from that benchmark
(Kumbhakar and Lovell, 2000).
11
in trade as does sharing a border. The OLS coefficients on the GDP variables are unity which
is a result consistent with the gravity model literature.
12
Table 2 OLS and MLE stochastic frontier estimation results
(1)
OLS
(2)
Frontier
country
pair
(3)
Frontier w
country pair &
time dummies
Constant
-33.37***
(0.2303)
-18.49***
(0.1423)
-20.39***
(0.1552)
lnGDPit
0.99***
(0.0066)
0.70***
(0.0041)
0.74***
(0.0049)
lnGDPjt
1.00***
(0.0061)
0.78***
(0.0038)
0.80***
(0.0039)
rDistij
-3.59***
(0.0486)
-2.28***
(0.0278)
-2.28***
(0.0444)
2.48***
2.10***
2.06***
(0.0265)
(0.0210)
(0.0331)
0.92***
(0.0669)
0.98***
(0.0403)
0.98***
(0.0373)
11.04
52.04***
(0.298)
49.01***
(0.3795)
0.976***
0.98***
-14.26
-13.85
-244612
-212727
-211240
93382
93382
93382
Compijt
Borderij
sigma-squared
Gamma
Mu
log likelihood
function
Number of
observations
Standard errors in parentheses; * p < 0.05, ** p < 0.01, *** p < 0.001
Table 3 presents trade performance results, or actual trade as a ratio of potential trade. High
trade performance (a high ratio of actual to potential trade) is associated with low trade
resistances. Conversely, low trade performance reveals high resistances to trade.
13
Table 3 Trade performance results, 1987-2006
1981-86 1987-91 1992-96 1997-01 2002-06
0.42
0.33
0.35
0.38
0.43
0.36
0.35
0.38
0.38
0.40
Exporter
China
Japan
Importer
Japan
China
China
Japan
France
Germany
Hong Kong
United Kingdom
United States
World
World
World
World
World
World
World
0.30
0.44
0.36
0.42
0.47
0.46
0.43
0.32
0.38
0.32
0.39
0.45
0.38
0.36
0.35
0.34
0.30
0.36
0.42
0.36
0.35
0.38
0.35
0.32
0.37
0.39
0.35
0.35
0.43
0.33
0.29
0.37
0.36
0.31
0.30
World
World
World
World
World
World
World
World
China
Japan
France
Germany
Hong Kong
United Kingdom
United States
World
0.32
0.37
0.36
0.40
0.38
0.41
0.39
0.35
0.31
0.26
0.32
0.37
0.41
0.38
0.36
0.31
0.28
0.21
0.29
0.33
0.40
0.36
0.33
0.29
0.25
0.23
0.30
0.35
0.38
0.35
0.34
0.29
0.32
0.22
0.27
0.34
0.40
0.33
0.34
0.28
Open economies that are close to large markets, such as are the Netherlands, Singapore and
Hong Kong, perform better as expected and it is their trade characteristics, or trade
technologies, that help define the frontier.
The world average trade performance is stable over the period 1987 to 2006 and is relatively
low. The low average is common in stochastic frontier models and is a result of there being
significant variation in the data. Given the reductions in transportation and communications
costs and the reduction of barriers to trade, both at the border and beyond the border,
reflected in rapidly increasing world trade values, one might expect mean trade performance
(realisation of potential) to be increasing. But the best performers push the elasticities higher
and the frontier shifts outwards (an improvement in ‘trade technology’) meaning the average
14
trade relationship has to keep up with the best performers for average to grow. This
observation is consistent with the findings of Dowrick and DeLong (2003) who show that in
the second half of the 20th century there has been divergence in growth between those
countries that have been at the global table and those that have not, and that there has been
convergence among those economies that have opened their economies. The increased
variation in the sample over time and the outlier bilateral trade flows that push the frontier
outward means that most countries see performance fall over time as the increased trade does
not keep up with the frontier.
Table 3 shows the remarkable achievement of China in realising its trade potential, an
achievement that is more significant given the steady world average explained above.
Trade between Japan and China is clearly above world average for the period 1990 to 2006.
This included periods of significant movements in political distance between the two
countries (Figure 1).
The variables that are included in explaining performance are regional and multilateral
grouping variables, the Economic Freedom Index of the Fraser Institute, language similarity
and tariff levels. Indexes of economic freedom are included for both exporter (i) and importer
(j) and are from the Fraser Institute (Gwartney et al., 2008). The average tariff rate for each
exporter and importer is also extracted from the index of economic freedom. There is
assumed to be very little correlation between the economic freedom index and the tariff rate,
as the economic freedom index is calculated from many variables of which the tariff rate is
only one.
15
The language index is based on Boisso and Ferrantino (1997) and is an index that takes a
value of 0 if none of the population of country i speaks the same language as in country j and
a value of 10,000 if all of the population in both countries speak the same language. Due to
the lack of data availability, the language similarity index does not change over the 27 year
period.
Variables for China’s WTO membership (both in relation to Chinese exports and Chinese
imports) are included. The WTO variable for Chinese outward trade takes the value of one
for Chinese export flows after 2001 and zero otherwise. The import equivalent is constructed
similarly. APEC, MERCOSUR, ANDEAN, ASEAN, EU and NAFTA regional agreement
variables are included. For each of these arrangements, two variables are included: one to
account for any trade diversion for non-members and another to measure the increase in trade
for members of those regional groupings7. The first variable takes the value of one if, within a
bilateral trade flow, one country is a member and the other is not. The other regional
agreement variable takes the value of one if both are members and zero otherwise.
A positive coefficient on the variable with both countries members of the regional agreement
would mean the policy helps explain why there is higher trade (and lower resistances to)
along that bilateral route bringing actual closer to potential trade. A negative coefficient on
the variable to measure the extent of additional trade between a grouping member and nonmember would show trade diversion or a corresponding increase in relative resistances
between non-partner and partner country trade.
Finally, political distance, as explained before, is included. It is included from the importer’s
perspective, the exporter’s perspective and as a sum of both, for each trade flow. It is also
7
A priori it is not expected that APEC will have any trade diversion effects as it is not preferential in nature.
16
reasonable to expect that negative news events or incidents do not impact on trade
immediately as contracts take time to cancel and events accumulate over time to show trends.
A lagged political distance variable is therefore included, with a one year lag consistent with
other findings. A one year lag is considered more appropriate than a two year lag (Reuveny
and Kang, 2003). This formulation has the added benefit of avoiding causality running from
economic distance to political distance8. There is evidence that changes in economic relations
influence political relations (Polachek, 1980; Mansfield and Pollins, 2003). Improvements in
a political relationship would not be expected to impact on the economic relationship, and
vice versa, immediately, but there is likely to be an effect after a lag, as economic agents and
foreign policy stances adjust.
8
Other studies estimate simultaneous equations or use Granger causality type tests that account for, and often
find evidence of bi-directional causality (Reuveny and Kang, 2003; Mansfield and Pollins, 2003; Armstrong,
2009).
17
Table 4 Explaining trade performance
(1)
OLS
(2)
OLS with
political dist
(3)
FE
(4)
FE with political
dist
constant
0.115***
(0.004)
0.0425***
(0.007)
0.261***
(0.005)
0.252***
(0.008)
WTO Ch exp
0.144***
(0.007)
0.102***
(0.0076)
0.111***
(0.0062)
0.0517***
(0.0065)
WTO Ch imp
0.0375***
(0.006)
-0.00855
(0.008)
0.0508***
(0.0063)
0.0461***
(0.0069)
APEC one
0.0221***
(0.0016)
0.0242***
(0.002)
0.00704***
(0.0015)
0.0123***
(0.00307)
APEC both
0.0965***
(0.002)
0.0904***
(0.00296)
0.0164***
(0.00293)
0.0302***
(0.00472)
Freedom exp
0.0149***
(0.0005)
0.0126***
(0.0008)
-0.00167*
(0.00065)
-0.00591***
(0.00108)
Freedom imp
0.0217***
(0.0005)
0.0192***
(0.00065)
0.00822***
(0.00064)
0.00874***
(0.00095)
Tariff exp
0.00286***
(0.000193)
0.00131***
(0.00024)
0.00101***
(0.00017)
0.000376
(0.00023)
Tariff imp
-0.000185
(0.000198)
-0.00298***
(0.00026)
0.000701***
(0.00017)
0.000490*
(0.00023)
ASEAN one
0.0625***
(0.00174)
0.0661***
(0.0024)
0.168***
(0.00547)
0.0920***
(0.0087)
ASEAN both
0.0868***
(0.00614)
0.0560***
(0.0082)
0.209***
(0.0155)
0.111***
(0.0227)
NAFTA one
-0.0285***
(0.00201)
-0.0361***
(0.0026)
-0.00305
(0.00270)
0.00966
(0.006)
NAFTA both
-0.0119
(0.00849)
0.0533***
(0.0065)
0.0822***
(0.0174)
0.0920**
(0.03)
EU one
0.0236***
(0.00118)
0.0160***
(0.0017)
-0.0195***
(0.00164)
-0.0114***
(0.0022)
EU both
0.0496***
(0.00216)
0.0544***
(0.003)
0.0186***
(0.00331)
0.0170***
(0.0045)
ANDEAN one
-0.0148***
(0.00176)
-0.0224***
(0.002)
0.0376***
(0.00391)
ANDEAN both
0.107***
(0.00730)
0.104***
(0.0119)
0.00847
(0.0119)
MERCOSUR
one
-0.0142***
(0.00184)
-0.0153***
(0.0022)
-0.000227
(0.00226)
0.00204
(0.00447)
MERCOSUR
both
0.0995***
(0.00789)
0.126***
(0.0085)
-0.0462***
(0.0113)
-0.0488**
(0.0171)
Languageij
0.000012***
(0.0000003)
0.0000131***
(0.0000005)
Time trend
-0.0069***
(0.0001)
-0.00341***
(0.0001)
-0.00112***
(0.00016)
Political dist
exp to imp
-0.0000593
(0.0000566)
Political dist
imp to exp
0.0000971*
(0.000056)
Political dist sum
0.0000319*
(0.0000193)
0.163
42162
0.164
R squared
85172
85172
observvations
Standard errors in parentheses. * p < 0.05, ** p < 0.01, *** p < 0.001
42162
18
The results in Table 4 explain trade performance and show that regional grouping variables,
language similarity and economic freedom help explain trade performance and trade
resistances. The variables of particular interest here are RTA and regional and multilateral
trade grouping variables.
The effect of politics on trade is not strong but there is evidence that political distance does
affects trade. In this model, the sum of directional political distance is statistically significant
at the 5 per cent level. Political distance from the importer’s perspective (events reported in
the importing country which are related to the exporting country) is also statistically
significant while political distance from the perspective of an exporter is not.
This asymmetry in the impact of political distance on trade is interesting and plausible. More
policy and political control is likely to be exercised over import activities than the activities
of exporters. On average it appears that political distance does not matter for an exporter and
increased uncertainty in dealing with a buyer in another country does not significantly affect
the volume of trade. Trade seems to be affected by how the importing country perceives the
source country of the goods.
Take the case of China’s accession to the WTO. China’s entry to the WTO increased both its
export and import performance by reducing resistances (except for imports in column (2) of
Table 4). These two variables have the largest estimated parameters and their effects on trade
performance are the largest among the trade policy related variables.
19
All regional groupings or RTA variables have a positive coefficient when both countries are
members, showing, as would be expected, that an RTA increases trade (and trade
performance) between members holding all else (such as distance, for example) constant. The
story diverges for regions when looking at the variables estimating the effect on trade of an
RTA or region when one county is a member and the other is not. The negative coefficients
on ‘NAFTA one’, ‘EU one’, ‘ANDEAN one’ and ‘MERCOSUR one’ mean that the RTA
membership or formation is trade diversionary.
A positive coefficient on the variables ‘ASEAN one’ and ‘APEC one’ means, for example,
that trade between an APEC member and a non-APEC country increased following the
formation of APEC or its membership of APEC.
The results show there is trade diversion from discriminatory regional trade blocs such as
NAFTA, MERCOSUR, ANDEAN, and the EU. APEC and ASEAN, on the other hand, show
increased trade among members as well as between members and non-members. There is no
evidence of trade diversion in the latter arrangements (Table 4).
The low R-squared of 16 per cent for these estimates confirms the consensus in the trade
literature that most of the frictions that limit the realisation of trade potential cannot be easily
measured. The trade performance results from Table 3 are a measure of all resistances that
limit the achievement of trade potential. The resistances identified in the analysis here
quantify roughly 16 per cent of their effect.
20
Economics over politics
While adverse politics may have affected the economic relationships of other countries
significantly, for Japan and China the market appears to have dominated over politics in the
development of their economic relationship. Table 4 supports the evidence of a large body of
literature that shows politics between countries does affect the economics (Polachek, 1980;
Mansfield and Pollins, 2003). Yet importantly, for Japan-China, it has not affected the
performance or realisation of the potential of the trade relationship significantly. Trade
between Japan and China outperforms the world average, Japan’s average export and import
performance, as well as China’s average import performance (Table 3).
There is no bilateral agreement between China and Japan that has underpinned the
development of their economic relationship in recent times. The Long Term Trade
Agreement of 1978 was directed to another purpose in another era before China had fully
committed to marketisation. Nor has the Bilateral Investment Treaty of 1988 been the main
driver on the recent surge of Japanese FDI into China. Political distance could be the reason
bilateral agreements have not been renovated but that does not appear to have affected the
achievement of trade potential..
What has rather seemed to be more decisive is both countries’ commitment to the rules and
norms of the international institutional system embodied in the WTO. These commitments
constrain the effects of bilateral political tensions and have provided the foundations for the
huge growth of the bilateral relationship between Japan and China within the multilateral
trading system.
21
China’s accession to the WTO in 2001 after 15 years of negotiation was a policy initiative
unlikely to be matched in the foreseeable future in terms of gains in international trade
(Drysdale and Song, 2000). The effect on Japanese trade with, and investment in, China was
profound as Japanese investors and business saw China as a real market for the first time
(Armstrong, 2009). Table 4 shows how large the impact of WTO accession was on Chinese
trade. But it was not only the event in 2001 that was important. The lead up to accession
shaped the way in which Japanese business dealt with China. In the lead up to accession,
China’s commitment to the global trading system and ultimately to a rules-based institution
was the significant factor.
China achieved a higher level of openness in trade before WTO accession than any other
large developing country and than is often understood (Lin et al., 2003; Branstetter and
Lardy, 2006). China’s trade liberalisation and market oriented reforms in the 1990s cannot be
viewed independently of its GATT, and later WTO, accession bid. The unilateral
liberalisations that were undertaken in the 1990s gave Chinese policy makers credibility and
confidence in the international trading system that was reciprocated by the international
economic community (Drysdale, 2000; Garnaut and Huang, 2000). The average statutory
tariff rate in 1982 was 56 per cent, which was reduced to 43 per cent in 1985 and then stayed
constant at that level until 1992. The period following 1985 saw the average tariff rate decline
steadily to 15 per cent by the time of WTO accession (see Figure 1) (Lardy, 2002). The 1990s
was a period of rapid liberalisation at the border but also involved significant domestic
reforms. Tariff and NTB liberalisation cannot occur in isolation of other domestic reforms
and the 1990s saw the establishment of the legal basis for trade and deep reform of both
import and export systems (Shengming, 2000). Unilateral trade liberalisation and market
oriented reforms from 1986 and throughout the 1990s (including through the APEC process
22
in 1995) meant that Japanese traders and investors could engage China confidently and
commitment to liberalisation and accession to the WTO provided insurance for Japan against
the effects of bad political relations upon its trade with China.
Figure 1 China’s trade dependence and average tariff rate, per cent, 1990 – 2004
% 70
Total Trade/GDP
WTO
60
50
Avg Tariff rate
40
30
20
10
0
1990
1992
1994
1996
1998
2000
2002
2004
Source: Tariff data from MOFCOM website; China.org.cn and trade data from UN Comtrade database and GDP
data from World Development Indicators (www.worldbank.org/data).
The gains to the Chinese economy and the international economic community from the
liberalisations undertaken leading up to accession in 2001, as well as the accession
23
commitments to further tariff liberalisation and removal of NTBs, are well documented9.
Why then did China’s trade (both imports and exports) increase so sharply after WTO
accession when most of the tariff liberalisation had taken place prior to 2001 (Figure 1)? And
more importantly, in the present context, how do we understand the surge in trade after the
WTO accession in the context of Japan-China bilateral trade and political relations?
Studies of China’s accession to the WTO usually focus on two aspects. One is the tariff
liberalisations, market access concessions through reduction of NTBs and other at-border
measures to integrate China into the global economic system. The other aspect is the reform
of Chinese institutions. Banking, legal and even political institutional reforms were agreed to
in accession commitments. This includes transparency issues related to legal and
administrative policies and the resolution of domestic commercial and trade disputes by
foreign firms that also affect domestic firms (Yu et al., 2003; Farah, 2006). These
institutional reforms are very important for continued economic reforms in China (Bhattasali
et al., 2004). Both aspects of accession affected China’s economic interaction with the
international economic community, including Japan.
China's pro-reform leaders used the external institution of the WTO to increase the pace of,
and lock in, reforms. The reforms were wide-ranging and importantly secured financial, legal
and economic institution reform. In fact, no other member joining the WTO has given so
many concessions on the way to accession (Drysdale, 2000; Brandt et al., 2007). The
comprehensive reforms towards a more market-oriented system and commitments to
transparency constrained Chinese policy makers from political intervention in international
commerce across a very wide range of business activity (Garnaut and Huang, 2000).
9
See, for example, studies by Rumbaugh and Blancher (2004) and Martin et al. (2009).
24
While the direct benefits from market access concessions and continued impact of reforms
are important, the significance of China's joining the WTO extended beyond its simple
impact on the change in relative prices it wrought between China and the rest of the world.
Perhaps in no other country in recent history has WTO accession effected such deep
economic, legal and political changes as it did with China (Cross, 2004). It meant the world
community recognised China as an equal partner and one that had committed to continued
progress towards economic globalisation (Chow, 2003). More fundamentally, it signalled
China’s commitment to a rules-based global trading system, illustrated by its willingness to
place constraints on Chinese policy makers towards foreign traders who were party to the
same commitments (Garnaut and Huang, 2000).
China’s membership in the WTO also affected the behaviour of China’s major trading
partners, including Japan, towards China. It changed Chinese policy behaviour and the
environment in which foreign and domestic firms operate, and the behaviour of foreign and
domestic firms operating in China (Drysdale, 2000). It managed to do this because of the
commitments that were made and the credibility of those commitments which China had
earned in the process leading up to accession. The commitment to the global trading system
prior to WTO accession should not be overlooked as it demonstrated that China was serious
in its long term intentions of moving towards a market economy.
The accession commitments were a blueprint for China’s institutional changes for at least a
decade from 2001, even if the timetable has not been strictly met against an ambitious
timetable (Chow, 2003). The process of WTO accession provided the international
community and China with a rare opportunity to set out principles governing trade relations
between China and the rest of the world (Garnaut and Huang, 2000). WTO accession and the
25
policy credibility and predictability that arose from accession protocol commitments limited
the likelihood of distortionary interventions in trade and in respect of FDI in China (Drysdale,
2000). It accorded to China’s trading partners the same treatment to which they are also
bound by the rules that govern the world trading system when dealing with China.
The effects of political distance on trade changed once China became a member of the WTO
because of the confidence that all players developed in China's moving towards a market
economy and the constraints the agreement put upon on Chinese policy makers. Sudden
reversals in regulations and policies became much more difficult in most sectors, behaviour
of policy makers that affects foreign enterprises became more predictable and there was
increased transparency in China's policy making processes. The deepening of reforms
towards industrialisation and an efficient market economy also meant that there is less state
involvement in the economy and that, even if political distance widens, less scope for
government intervention in trade dealings. Japanese enterprises were able to continue to
engage in operations in China when political distance widened.
China’s commitment to, and responsibilities within, the global trading system was the critical
reason why political distance has not significantly affected the economic relationship
between Japan and China over the past decade. In fact, the economic relationship, with
commitment to GATT, and later WTO entry, not only helped to insulate against political
tensions on the economic relationship, but allowed the economic relationship to come to
constrain and shape the political relationship.
Armstrong (2009) provides empirical support for the conclusions that the economic
relationship drives the political relationship. It can be seen quite simply, however, in the
26
robust and extraordinary growth of the trade both ways and investment from Japan over
periods of political tension, and the improving of political relations after a period of
accelerated economic dealings immediately after China’s membership of the WTO.
Conclusions
The results of the modelling used here suggest that trade between Japan and China
performed closer to its potential than trade flows worldwide after controlling for distance,
size, endowment structures and multilateral effects. Moreover, the performance of the JapanChina bilateral relationship compares well with the performance of each country's trade with
all their other trading partners generally. The proximity of Japan and China, their
complementary economic structures and the multilateral settings in which both are deeply
integrated have been more powerful drivers of trade than the ups and downs of politics.
While economic relationships are not independent from politics, as is also shown, the JapanChina trade relationship does not appear to have been diminished or disturbed by political
tensions to any significant extent. While political relations can derail other economic
relationships, or act as barriers to trade, for Japan and China they would seem to matter only
at the margins. This is the case despite the noteworthy ups and downs in the political
relationship over the period reviewed here.
China’s commitment to the global trading system from the mid 1980s and its accession to the
WTO in 2001 have meant that tensions in the political relationship have not derailed, but
rather have increasingly come to be dominated by, the economic relationship. This
conclusion deserves further and forensic examination both for what it tells us about the
importance of multilateral economic institutions (other circumstances remaining the same)
27
and how they constrain the political impact of the conduct of bilateral economic affairs, as
well as what it might suggest about the future course of political relations between these two
major economic partners (other things remaining equal).
References
Aigner, D., Knox-Lovell, C. A. and Schmidt, P., 1977. 'Formulation and estimation of
stochastic frontier production function models', Journal of Econometrics, 6(1):21–37.
Anderson, J.E., 1979. 'A theoretical foundation for the gravity equation', American Economic
Review, 69(1):106–16.
Anderson, J.E. and van Wincoop, E., 2003. 'Gravity with gravitas: a solution to the border
puzzle', American Economic Review, 93(1):170–92.
— —, 2004. ‘Trade costs’, Journal of Economic Literature, 42(3):691–751.
Armstrong, S.P., 2007. ‘Measuring trade and trade performance: a survey’, Asia Pacific
Economic Papers, No. 368.
— —, 2009. The Japan-China economic relationship: distance, institutions and politics,
Canberra, Australian National University. Ph D. thesis.
Baldwin, R.E. and Taglioni, D., 2006. 'Gravity for dummies and dummies for gravity
equations', NBER Working Paper No. W12516, National Bureau of Economic
Research.
Bhattasali, D, Li, S. and Martin, W., 2004. China and the WTO: Accession, Policy Reform
and Poverty Reduction Strategies, World Bank Publications, Washington DC.
Boisso, D., Ferrantino, M., 1997. ‘Economic distance, cultural distance, and openness in
international trade: Empirical puzzles’, Journal of Economic Integration, 12(4):456484.
28
Brandt, L., Rawski, T. and Zhu, X., 2007. ‘International dimensions of China’s long boom:
trends, prospects and implications, in China and the Balance of Influence in Asia, in
W. Keller and T. Rawski (eds), University of Pittsburgh Press, Pittsburgh.
Branstetter, L. and Lardy, N., 2006. ‘China’s embrace of globalisation’, NBER working
paper No. 12373, National Bureau of Economic Research.
Chow, G.C., 2003. ‘Impact of joining the WTO on China’s economic, legal and political
institutions’, Pacific Economic Review, 8(2):105-115.
Coelli, T.J., 1996. 'A guide to FRONTIER Version 4.1: a computer program for stochastic
frontier production and cost function estimation', CEPA Working Paper 96/07,
University of New England.
Cross, K.H., 2004. ‘China's WTO accession: economic, legal, and political implications’,
Boston College International and Comparative Law Review, 27(2).
Dowrick, S. and J. DeLong, B., 2003. ‘Globalisation and convergence’, in M.D. Bordo, A.M.
Taylor and J.G. Williamson (eds), Globalization in Historical Perspective, University
of Chicago Press, Chicago.
Drysdale, P.D., 1967. Australian-Japanese trade, Canberra, Australian National University.
Ph D. thesis.
— —, 2000. ‘The implications of China’s membership of the WTO for industrial
transformation’, in P.D. Drysdale and L. Song (eds), China's Entry to the WTO:
Strategic Issues and Quantitative Assessments, Routledge, London.
Drysdale, P.D. and Garnaut, R., 1982. ‘Trade intensities and the analysis of bilateral trade
flows in a many-country world: a survey’, Hitotsubashi Journal of Economics,
22(2):62-84.
29
Drysdale, P.D., Huang, Y. and Kalirajan, K.P., 2000. 'China's trade efficiency: measurement
and determinants', in P. Drysdale, Y. Zhang and L. Song (eds), APEC and
liberalisation of the Chinese economy, Asia Pacific Press, Canberra.
Drysdale, P.D. and Song, L., 2000. China's Entry to the WTO: Strategic Issues and
Quantitative Assessments, Routledge, London.
Farah, P., 2006. ‘Five years of China WTO membership: EU and US perspectives about
China's compliance with transparency commitments and the transitional review
mechanism. Legal issues of economic integration’, Kluwer Law International,
33(3):263-304.
Garnaut, R. and Huang, Y., 2000. ‘China and the future of the international trading system’,
in P.D. Drysdale and L. Song (eds), China's Entry to the WTO: Strategic Issues and
Quantitative Assessments, Routledge, London.
Gasiorowski, M. and Polachek, S., 1982. ‘Conflict and interdependence: East West trade and
linkages in the era of détente’, Journal of Conflict Resolution, 26:709-730.
Goldstein, J. S. 1992. ‘A conflict-cooperation scale for WEIS event data’, Journal of Conflict
Resolution, 36:369-85.
Greene, W., 2004. ‘Distinguishing between heterogeneity and inefficiency: stochastic frontier
analysis of the World Health Organization's panel data on national health care
systems,’ Health Economics, John Wiley & Sons, Ltd., 13(10):959-980.
— —, 2005. ‘Reconsidering heterogeneity in panel data estimators of the stochastic frontier
model’, Journal of Econometrics, Elsevier, 126(2):269-303, June.
Gwartney, J., Lawson, R. and Norton, S., 2008. Economic Freedom of the World: 2008
Annual Report, The Fraser Institute, Data retrieved from www.freetheworld.com.
Hirschman, A. [(1945)(1980)]. National Power and the Struggle of Foreign Trade,
University of California Press, Berkeley.
30
Kalirajan, K.P., 2008. ‘Gravity model specification and estimation: revisited’, Applied
Economics Letters, 15(13):1037-39.
King, G., 2003. ‘10 million international dyadic events’, hdl:1902.1/FYXLAWZRIA
UNF:3:um06qkr/1tAwpS4roUqAiw== Murray Research Archive [Distributor]
http://gking.harvard.edu/homepage.html
Kumbhakar, S.C. and Knox-Lovell, C.A., 2000. Stochastic Frontier Analysis, Cambridge
University Press, New York and Melbourne.
Lardy, N., 2002. Integrating China into the Global Economy, Brookings Institution Press,
Washington DC.
Lin , J.Y., F. Cai, Z. Li, 2003. The China miracle: development strategy and economic
reform, Chinese University Press, Hong Kong.
Mansfield, E.D. and Pollins, B.M., 2003. Economic Interdependence and International
Conflict: New Perspectives on an Enduring Debate (eds), University of Michigan
Press, Michigan.
Martin, W., Anderson, K. and Pham, C.S., 2009. ‘Effects of GATT/WTO on Asia’s trade
performance’, in S.P. Armstrong and V.T. Thanh (eds) International Institutions and
Asian Development, Routledge, London.
Messerlin, P., 2007. ‘China in the WTO: Antidumping and safeguards’, in D. Bhattasali , S.
Li and W. Martin (eds), China and the WTO: Accession, policy reform and poverty
reduction strategies, World Bank Publications, Washington DC.
Polachek, S.W., 1980. ‘Conflict and trade’, Journal of Conflict Resolution, 24:55-78.
Pollins, B. 1989a. ‘Conflict, cooperation, and commerce’, American Journal of Political
Science, 33:737–61.
— —, 1989b. ‘Does trade still follow the flag?’, American Political Science Review, 83:465–
80.
31
Ravallion, M., 2003. 'On measuring aggregate ‘social efficiency’', World Bank Policy
Research Working Paper, No. 3166.
Reuveny, R. and Kang, H., 1996. ‘International trade, political conflict/cooperation, and
Granger Causality’, American Journal of Political Science, 40:943-970.
— —, 2003. ‘A simultaneous-equations model of trade, conflict, and cooperation’, Review of
International Economics, 11:279-95.
Rumbaugh, T. and Blancher, N., 2004. ‘China: international trade and WTO accession’, IMF
Working Paper No. 04/36, March.
Schneider, G., Barbieri, L. and Gleditsch, N.P., 2003. ‘Does globalization contribute to
peace? A critical survey of the literature’, in G.Schneider, L. Barbieri, and N.P.
Gleditsch (eds), Globalization and Armed Conflict, Rowman and Littlefield, Lanham,
Maryland.
Shengming, Y., 2000. ‘The WTO and China’s trade strategies in the 1990s’, in P.D. Drysdale
and L. Song (eds), China's Entry to the WTO: Strategic Issues and Quantitative
Assessments, Routledge, London.
Yu, P.K., Chang, G.G., Cohen, J., Economy, E.C., Hom, S. and Li, A.Q., 2003. ‘China and
the WTO: progress, perils and prospects’, Columbia Journal of Asian Law, 1(1).
32