Title
Author(s)
Citation
Issue Date
Type
Structural Changes in World Trade Flows
Yamazawa, Ippei
Hitotsubashi Journal of Economics, 11(2): 11-21
1971-02
Departmental Bulletin Paper
Text Version publisher
URL
http://hdl.handle.net/10086/8032
Right
Hitotsubashi University Repository
STRUCTURAL CHANGES IN WORLD TRADE FLOWS*
By IPPEI YAMAZAWA*
I. Intensities of World Trede Flows
The purpose of this paper is to find out the structure of international trades between
pairs of countries and its overtime changes and to analyze factors affecting them. It may be
asserted that one should consider the structure of a country's export to or import from the
world as a whole but not that of its trade with any particular partner since the latter is
determined more or less by chance and is volatile every year. In reality, on the countrary,
there is found a distinct structure of high or low intensity of trade between countries and it
has been fairly stable over time. This implies some underlying determinants affecting the
trade intensity among countries.
The intensity index of country i's export trade with country j is defined by
Xij / X,j (1)
lij~-Xir/ X..
where Xij, Xi.(~~Xij), X.J(~~Xij) and X..(=~~Xij) represents country i's export to
country j, total export of country i, total import of country j, ~nd the total volume of world
trade respectively. The index measures the share of country j in the total exports of country
i relative to its share in the world imports. Iij equals unity when country i exports to
country j in proportion to the latter's share in world imports. It exceeds unity when country
i trades with country j more intensively than on the average, while its falls short of unity
when country i trades with country j less intensively than on the average.
\Vhat do the world trade flows look like when lij equals unity for all pairs of i and j?
This is an extreme situation where multilateral trade dominates the world in the sense that
each country trades with others at equal intensity. It is, however, far from the case. The
values of I*J calculated from the actual trade flows are high above unity among some groups
of countries and low below unity among other groups of countries. '
A number of factors cause intensity of trade between countries diverge from unity. The
complimentarity in comparative advantage structure of two trade partners should be counted
first. Intensity of trade is likely to be high between a pair of industrial exporter and primary
* Lecturer (K6shi) of Economics. Hitotsubashi University,
1 This is an extended version of the author's previous study. Yamazawa, Ippei, "Intensity Analysis of
World Trade Flow", Hitotsubashi ,Journal ofEconomics, Vol. 10, No. 2, February 1970. He has benefited
from comments by Professors Kiyoshi Kojima, Yoichi Shinkai and Hlroshi Niida and other participants in
the staff seminar of Hitotsubashi University and the 8th Econometric Conference at Rokko in July, 1970.
He is also indebted to Misses Masako Kohno and Shoko Fujii of Hitotsubashi University for computa-
tional works. Computation was carried out by FACOM 230-25 system at Hitotsubashi University.
12 HITOTSUBASHI JOURNAL OF ECONOMICS [Februar
goods supplier where strong complementarity exists in their structure of comparative advantage,
while intensity will be low either between two industrial exportors or between two primary
goods suppliers where their comparative advantage structure is less complementary or competitive with each other.
Transportation costs, discriminatory commercial policies of various types, the flow of
capital and economic aids from developed to developing countries, should be added to the
list of possible factors causing the divergence of trade intensity from unity.
In the previous paper the intensity of trade is calculated for fifteen regions at three-)'ear
periods 1955-57, 60-6_,), 65-67.2 It is reproduced in Table 2 in which we can easily confirm
the above statement. High values of intensity are found between some pairs of countries,
such as that of 8.5 in E EUR-CHN MX trade, 7.5 in intra-E EUR trade, 5.0 in CAN-US
trade. On the other hands, very low intensity is found between US and CHN MX (0.0),
between J and UK or EEC (O._1)-O.3). It should be noted that the structure of high and lovy
intensity among countries has been fairly stable over the three periods.
At the same time it may weil be indicated that there exists such a steady change over
the recent decade that I.J below unity tends to rise whereas lij above unity tends to fall.
This tendency of lij toward unity is statistically verified. Out of 219 pairs of i and j, L] is
found to tend toward unity in 137 pairs. The ratio of which is high enough to reject at 1%
significance level the hypothesis that there exists no particular tendency of change in lij'
Furthermore, the variance of loglij around O declines steadily from 0.3167 in 1955-57, to
0.2512 in 1960-62, and 0.2291 in 1965-67. The tendency of liJ toward unity implies the
multilateralization of world trade flows.3
The rest of this paper is devoted to identify quantitatively the effects of l'arious factors
mentioned above on the actual structure of world trade flows and their changes over the
recent decade. In the following section a measure of complementarity in comparative advantage
between trade partners is devised and the structure of complementarity among the fifteen
regions is analyzed. In the third section each of other important determinants is specified
in quantitative terms, and in the final section it is attempted to regress intensity of world
trade flows on its various determinants and to assess the relative importance of each deter'
minant in explaining the structure of world trade flows and its overtime changes.4
2 Fifteen regions are USA, Canada, l_atin America, EEC, UK, EFTA excludmg UK, Other Western
Europe, E(Istern Europe including USSR, South African Republlc, Other Africa, Japan, West Asia (Asian
part of Mid East), Other' Asia, Mainland China and other Asian Socialist countries, Australia and New
Zealand. Trade statistics were taken from U. N., Monthly Bulletin of Statistics.
3 This finding seems to support Harberler's analysis of the tendency toward multilateral trade in the
post World War 11 era. Harberler. G., "Integration and Growth of the World Economy in Historica]
Perspective", American Econonlic Revt'ew, March 1964.
4 Several l~'riters have analyzed trade flows by means of trade intensity indexes. See, Brown. A. J.,
Applied Economics, Aspects of the World Economy tn Wa' and Peace. London, 1947. Kojima, K. Sekai
Kdzai to Nihon Boeki (World Economy and Japan's Foreign Trade), Tokyo, 1962. Drysdale, D. P.,
"Japan, Australia, New Zealand: The Prospect for Western Pacific Economic Integration", Econo'ntc Record, 45, (lll), Sept. 1969, pp. 321-42. But this paper is the first attempt to quantify the determinants
of structure of trade intensity by econometric method.
The concept of trade intensity is used for the prediction of trade flow in Uribe, P., de Leeuw. C. G.,
and Theil, H., "The Information Approach to the Prediction of Interregional Trade Flows", Rev. oj
Econ. Sted., Vol, 33, 1966. The analysis of world trade flows in alternative approaches have been under'
taken by Dutch and Finish economists. Tinbergen, J., Shaplng the World Econo"ry, New York, 1962.
Pdyhonen, Pentti, "A Tentative Model for the Volume of Trade between Countries". TVeltwirtshaftliches
Archiv, Band 10, Heft 1, 1963, pp. 93-99. Linneman. H., An Econometric Stndy of Internotional Trede
Flow, Amsterdam, 1966.
197l]
STRUCTURAL CHANGES IN WORLD TRADE FLOWS
II. Determinants of Trede Intensity .' Comparative Advantage
Comparative advantage has been thought much of in the theory of international trade.
Trade would be more intensive between a pair of countries whose comparative advantage
structure is complementary with each other, other conditions being equal, and it would be
less intensive between countries of comparative advantage structure competitive with each
other. But how can we measure the degree of complementarity in comparative advantage
slructure between trade partners ? Even if we could devise such a measure in theory, can it
be imp]emented with existing statistics of international trade?
Recent studies of international trade patterns 1lave become to list up various determinants
of comparative advantage, instead of simplc capital-Iabor ratio. In addition to introducing
more factors, traditional trade models of two countries by tv~'o commodities should be applied
to the world of many countries and many commodities. There still remain many problems
unsolved in applying the theoretical model of comparative advantage to actual world trade
flow-s. Instead we would take a short cut and assume, according to Bela Balassa, that
country's export and import performance reveals its structure of comparative advantage.5
The index of export specialization of country i, with respect to commodity h, the index
of "revealed compairative advantage", is defined by
X•,h h
X.
Sth~~~
s_'
Xi . X~. .
\lihere A'hi. is the value of i's export of commodity h to the world and X'~., the value of
\vorld total tradc of the same commodity. It compares the share of commodity h in co_untry
i'~'- ex~orts with that in world total trade. It exceeds un[ty when country i specializes more
intensively in the export of commodity h than on the average, it falls short of unity when
specializes less intensively and it equal unity in the case of average specialization. The index
of import specialization is defined similarly and measures the country's comparative disadvantage structure.
'~ / '~- (3)
R h- X.j X..
J=
,j X..
Both indexes take values above or below unity, which is the weig'hted aritlumetic mean of
both Shl and R'~j.
X.
:fR,
,LX.
'~)Si
~( '~:.~5
~(
f~ =: I (4)
If we define the covariance of Sns and R'}j
as
=,~ )
X.
~( .:
cov (Si, Rj)
X' (Si/L_ 1)(Rjt- 1) (5)
it would be positive when the structure of S'Ls and that of R'~J are similar, that is, the struc-
ture of comparative advantage of the two countries is complementary with each other. It
would be negative when the structure is competitive between the two. ,
Let us define the degree of complementarity in comparative advantage structure between
Balassa, B., 'Trade Liberalization and "Revea]ed" Comparative Advantage', The Manchester School
of Ec0'10"tic and Social Studies, Vol. 33, No. 2, May 1965, pDl . 91-123.
13
14 HITOTSUBASHI JOURNAL OF ECONOMICS [Februar
two countries Cij by the covariance of country i's export specialization and country j's import
specialization added by one
CijE cov(Si, Rj) + I (6)
Thus Cv above unity indicates strong complementarity in comparative advantage structure
between two countries, while Cij below unity indicates weak complementarity or competitiveness between the tw0.6
In my previous study the values of S'~*, R'~j and Cij were calculated for fifteen regions
and three periods. It may be appropriate to summarize the result of the previous study.T
The fifteen regions are classified into two groups according to the structure of Shi. For
industrial group (US, UK, EEC and J) export specialization is high for manufactures and
low for primary goods, whereas the structure is reversed for primary goods supplying group
(the rest of the regions, except OTH EFTA and E EUR).8 Since the structure of Rhj is
contrary to that of Sht in general,9 the covariance is likely to be positive between a pair of
countries belonging to different groups and negative between a pair belonging to the same
group. Therefore the value of Ctj will be high above unity between the former pair of
countries and low below unity between the latter pair.
In addition to this general pattern, Cij tends to be higher or lower for such countries
as J, UK, ANZ, S AFR, whose structure of export and import specialization is concentrated
in some particular commodities. For example CUK.J=.5295 and CANZ.J=1.8652 'in 1955-57
period. It should be noted that the structure of Cij among fiftteen regions has been fairly
stable over the last decade. But none the less there appears a tendency of steady change,
such that Cij below unity rises and Cij above unity falls. This tendency of CtJ toward unity
is found in the cases of 77.33~, which is significant at 1~ Ievel. The variance of CiJ around
O declines steadily from .0147 in 1955-57 to .0095 in 1960-62, and to 0090 m 1965 67
How has this tendency been brought about during the recent decade ? It is explained
by the world wide change in the structure of trade and specialization. Industrialization of
primary goods supplying regions has raised steadily their export specialization and lowered
their import specialization in manufactures, especially that in SITC 6 & 8 category, accompanied by the contrary movements on the side of industrial regions in the same categories.
This worldwide change has tended to cause Cij toward unity through lower correlation, both
positive and negative, in the structure of export and import specialization on one hand and
more diversified structure of them both in industrial and primary goods supplying regions.
6 It was proved in the prevlous study that Ctj rs the hypothetlcal intensity of trade in the world of
exchange of homogeneous commodities with neither transport costs nor any artific[al trade impediments.
See Yamazawa, op. cit., pp. 63-64.
1 All commodities are classlfied into six categories (Standard International Trade Classification Sections
O & 1, 2 & 4, 3, 5, 7, 6 & 8), the first three of which are primary commodities and the last three are
manufactures. More detailed analyses of the export and import specialization of the fifteen regions and
the complementarity among them were done in the previous study. See Yamazawa, op. cit., pp. 73-77.
8 These two regions are classified as the first group according to the structure of their export speciali.
zation but as the second group according to that of their import specialization.
9 Structures of import specialization are contrary to those of their export specialization and are gener-
ally more diversified, which seems to be explained by the similarity of demand patterns in comparison
with skewness of resource endowments among countries on one hand, and protective commercial policies
on the other. See Yamazawa, op. cit., pp. 74-75.
STRUCTURAL CHANGES IN WORLD TRADE FLOWS
197l]
15
III. Other Determinants of Trede Intensity
It is generally presumed that trade becomes more intensive between countries located in
proximity than at a distance. This is partly because it would cost more to transport the
distance and partly because the horizon of traders does not always cover the whole world
but is restricted by the intensity of historical and cultural exchange and traditional channels
of market'ng The h 'c I distance between two countries is not certainly the best measure
l . p ysl a
of the social and economic proximity between them, but it will stand proxy for it.
Trade intensity seems to be increased not in reverse proportion to the absolute distance
itself' (dij) but to the absolute distance relative to the average. The measure of relotive
distance Dij is defined by
Dij=
4~/(~i dij)(~:dij)j
and it takes values around unity.ro
hen we defined measures of export and import specialization and the complementarity
in comparative advantage structure, we have implicity assumed homogenity within each commodity category. In rea]ity, however, each commodity category consists of a number of
~
differentiated products and country i's product may be preferred in country j's market to that
of others in the same commodity category. Assuming aside the product differentiation, Cij
may give a biased effect on trade intensity, unless it is supplemented by a variable representing the effect of product differentiation. Lind9r suggested that two countries with similar level
of per capita income will have the "representative demand" of similar pattern and trade of
differentiated products will become intensive between the tw0.11
Linder's hypothesis is tested by introducing to our model the dtfference ofper capita
income between two countries.
yi j = yji =
_ min(yt, y/)
max(yt, yJ)
where yi rnd yj are per capita income of country i and j and it may well be assumed that
the effect is symmetrical, that is, it has the same effect regardless of wherther the exporting
or importing country has a higher level of per capita income. yij takes positive values up
to unity at which there is no difference in per capita income.
There exist various economic relationship among countries other than commodity trade,
and they will affect more or less trade intensity among countries. It is often suggested that
long-term capital movements, especially direct investment by multi-national enterprises will
possibly changb the pattern of international trade. It seems to exist a more direct link
ro The distance between two regions is measured by the average navigation miles between major ports
of each regions. More than two ports are sellected for such broad regions as LAT AM, E EUR, OTH
AFR, etc. Figures are taken from Kaijo-Hoan-cho, Suiro-bu, Kyori-hyo (Distance Table), Tokyo, Japan,
1964.
In principle, Iand transportation from major port to centers of individual countries in each region is
neglected as well as the well-developed land transportation between regions such as those in Europe.
This measure of distance can not be helped being arbitrary, but it will serve at least to distinguish short,
medium, or long distance between and within regions.
11 Linder, S. B., An Essay on Trade and Transformation, Stockholm, 1961.
(8)
16 HITOTSUBASHI JOURNAL OF EcoNoMlcs [February
between the flow of economic aid from a developed to a developing country and trade flow
between the two since the economic aid oft t k the form of capital loan tied with rhe
commodity export of the donor country to the receiving country. Data is available only for
flow of inter-governmental loan for 1960 and 1965.12 The index of intensity ofaid fi:ow from
country j. Aij, is defined by
A,j= a~a:/ /~L (g)
where aij the value of inter-governmental loan from region i to region j. at,(=~atj) the
total aid donated by region i, a,j(=~aij) the total aid received by region j, a.. tjhe world
total of the aid flows within a particular year.
Various types of economic integration both among developed and developing countries
and traditional trade blocs such as British Common Wealth and French Community seem to
explain considerable parts of the structure of trade intensity. Since it is difficult to quantify
them, dummy variables are introduced to stand for the effects of these discriminatory commercial policies.
1, if both countries are members of the same economic integration among developed
Pijl = countriesl8
O, otherwise
1, if both countries are members of the same economic integration among developing
Pi f = countriesl4
={O, otherwise
Pi,
s 1, if both countries join the same traditional trade blocl5
O, otherwise
The difference in economic system between capitalist and socialist countries will have a
discriminatory effect on trade relationship similar to discriminatory commercial policies. Two
dumy variables are introduced to represent trade increasing effect between socialist countries
and trade reducing effect between countries with different economic system.
-{
_ 1, if both are socialist countriesl6
Pi j
fO, otherwise
1, if one is capitalist and the other is socialist country
Pij5=10, otherwise
To sum up, trade intensity is represented as a function of nine variables
lij=flCij, Dij, y,/. Aij, Pijl, Ptf, Pij3, Ptf, Pif) (10)
12 The data source is OECD, Geographical Distribution ofFinancial Flows to Less Developid Countries,
1968, which records the flow of inter-governmental loan from OECD member countries (US, CAN, EEC,
UK, OTH EFTA, J and ANZ) to developing regions (LAT AM, O W EUR, OTH AFR, W ASIA,
OTH ASIA). Therefore Aij takes value of O for other pairs of i and j.
IB It includes EEC, EFTA for 1960-62 and 1965-67 and Australia New Zealand free trade agreement
for 1965-67.
l' It includes Latin American Free Trade Area (LAFTA), Central American Common Market (CACM),
Arab Common Market, and Maghreb Economic Community, all for 1965-67.
15 It includes British Common Wealth and French Community for the three periods.
16 It is applied to trades between E EUR and CHN MX and within E EUR. On the otherhand Ps
takes value of unity for trades between one of the two regions and the rest of the regions.
/
STRUCTURAL CHANGES IN WORLD TRADE FLOWS
197l]
17
The first four explanatory variables are normalized to take values around unity in correspondence to lij and the last five are dummy variables, it may well assumed that the effect of
individual variables are multiplicative with each other.
The following equation of log-linear form is estimated with cross-region data (with
sample size of 219) for each of three periods.
log lij =ao +al log Cij +a2 Iog Dij +(r3 Iog yiJ + a4 Iog Ai,
+a5 Iog Pijl + a6 Iog Ptj2 + al log PiJs +a8 Iog P*f + ag log Pij5+ vij (11)
As is clear from the above arguments, the signs of al, as~a8 are expected to be positive,
while those of a2 and crg negative.
rv. Interpretotion of the Results of Estimation
The results of estimation of equation (11) are shown in Table 1. For each period alter
native estimation is attempted to exclude some variables which turned out to be statistically
insignificant and of wrong sign. Five variables, C, D, P3-P5 have proved to be significant
and of'proper signs over three periods. The coefiicients are fairly stable for alternative
estimations. Two variables for economic integration P1 and P2 are not significant statisticaily
but of proper signs. Neither y nor A is significant statistically and both are of wrong signs.
The coefficient of determination is 0.52 at best, but we can not expect very high R2 with our
data of broad regional classification.
The estimates of each coefficient and their over time changes seem to throw light on the
structure of world trade fiows. The elasticity of C,j Is in the range of 1.3 and 1.6, and that of
D•
is about 0.5. Both do not seem to have changed ove
tional trade blocs is still strong, while those of recent economic integration are weak and not
*,
significant. But over the decade the value of P3 has declined steadily, whereas that of P1
has increased.
The difference :in economic system has proved to have the significant efiect on the
structure of world trade flows. Trade intensity is increased much more between socialist
countries than within traditional trade blocs. But there appeared an asymmetry that the
intensity of East-West trade is reduced by a quarter time as much as the trade within the
East is increased.17 It should be noted that the estimates both of P4 and P5 decreased in
absolute values steadily over the last decade. '
Unsuccessful result of introducing both Yij and Aij deserves some comments. As to
the former of the two, we should not test the Linder's hypothesis regardless of the absolute
levels of per capita income of individual regions or of whether they are industrial regions or
primary goods supplying regions. As to the latter our measure of Aij do not include the
aid flow donated by socialist countries and it should also be combined with capital flow by
private enterprises. Thus we should improve these shortages in data before we conclude
neither Yij nor AiJ has significant effect on trade intensity.
The tendency of lij toward unity pointed out in Section 11 is related to two factors. One
is concerned with C,j and the other with P3, P4 and Pb. The coefflcient of Cij has not
IT In the regression estimation of log-linear form, dummy variables are adjusted to take values of ten
and unity in stead of unity and zero respectively. Similarly very low values of Itj are uniformly adjusted
to take 0.1.
18 HITOTSUBASHI JOURNAL OF EOONOMICS [February
changed significantly over the decade but the value of Ct! itself tends toward unity. On
the otherhand the coefficients of the three dummy variables have declined steady in absolute
values over the decade. Both factors have caused lij move toward unity. The former is
associated with industrialization in primary producing regions whereas the latter with the decline
of discriminatory effects of both traditional trade blocs and of different economic system.
To conclude we have succeeded to find out several factors affecting the structure of
world trade fiows and its overtime changes. But nearly a half of the variance of Itj still
remains unexplained. The pattern of unexplained residuals of lij shows such distinct and
stable biases as high intensity among US, CAN, and LAT AM, or among Asian-Pacific
regions. This seems to be attributed not merely to the geographical proximity but to such
historical factors as the process of the expansion of world trade flows. In this respect we
may have to analyse not so much the existing structure of world trade flows as its overtime
changes.
L
STRUCTURAL CHANGES lN WORLt) TRADE FLOWS
197l]
H
tO
cll,
C\l
Lt)
~
~
CJO
C\]
LO
~ LO~~
OH
~~
C\]O
~tlL!)
OOH
CO~
(NO
,~
C1,0
y:)O
,:)C\I
~
~
~,~
~(:)
t!)(:)
':~rl
v
~3
~
oo~
c~]H
~
hco
Loo
'bOo
;=1
rH~~
C'bCC
~CO
LOO
~ o~
Heo
cOH
~~~;
c\lo
•~
c~]oo
Hy:)
coa)
~H
,,
cs)q,)
c~o
~
~~
~11~
O)
t~H
v
~,~
~ •:rH
~O~
I~~O
(:)CO
LOO
\l
~
~~
o
~
~
~~
H
~
~r
~0
~r
~'
cYD~o
co
o,
\l
~3
l~
CO~
CO
~:)
C~~O
~r
~
S
l~
c')c')
I~H
~
Ho
,,
~tle
C,tO
q:)O
LOCh~
~~~~
~r
C\lCO
~'Q LOCO
~O
~
Olr~
~:~
~
~
'Ht~
C7)1~
H_
u'q'~
eq~
(7)O
~:)~
'~ $S
:'O
8e,1:1
:5
~{~
HO *o
I
'~*
C!).LO
(:)(7)
COO
t!)H
u)~
eO O~)
edC:, ~'
~1e,a~
0~}
.~i
e,q'
o~
~(:1
o(1)
> > ~ •~L
Q'e'
lO1;
'~ :e
~~)
coH
bOf:~~0$:1 F1~r~~~
OO *C!'HC::
~~
**
,~~
•~,~i
l:1 f:~ O~~
Ee
H~
H~ ~~
'~ '~
t!)co
Lf)CO
O O -C!)^
'* '*U
T':!~:l~:
eOei
~
~~
t~~
''
q55:~Q'(U
'~ '~ .~ ~~ ~
O O.': q' Q'
~ro
~~1
H~
Cf)H
~O
CeC9
-O
hl'HG'O'u'
~O ~O~1
~~~~~~~
OU
ch~H
F~ F11'~!1'
OOCES!S~$
~~~~~
~;
C~]Cf)
f~
C\10
C~aCO
c~lco
*~~
~
Cf)O
OH
f:~e:~
<
Q~
H
~
O
u,
-~,
u
Ha
C)H
,J
oH
OH
,J
oc\]
~~~~~c~
'*
co
~
OH
\l
H
~s
:9
:,
~
~
CO~
C,)C\l
~~;
~F~
Lf,~
~
~;
l~
~O~r
~eC~g
LOt~
l~
c\s6~
CY)
~c)(N
H
Hq'ee
c,)
HO Qq~
Ho
C\t~
OO OO
~l
oo
Z
~s
~
1;,~'
>
u,
u,
~]O~
c~
p~:
LOa
~~r
~~
LOO
~
eOL!)
LOt~
O11O
LOO
,~
H~11
C)cr)
C~]~
Lr)O
~
C,)CO
~rC~
C~1~
LOO
,~
C\a~~
t!)
~..:O
O
C06~
H~r
l!)~
LOO
~r~
C,)~f
~~~
LO,:,
~
Lr)Q
COO
C;)~
t!)O
q' d .
>
'~:S ~~
cs
Oo
H
:J1:~
a5
~ ,:~
~~ .~ ~:'_
H
~pQ
e~
H
C!)
•~
O)C\]
CO~tt
C~~CO
r4~
cOH
o'~
,~,
<
<
oH
oH
~l<
~)
Fq
~ l~
Hco
c\]
~Jpc7)
~CO
C~]
Hc~
eo~
~j
~r~~
~o~
~r':)
~
f:1Cl~a;q,
H
~Q
~~'~
~
-$1
~:S
co~~~
~,
c~l~
oo
~F
c\l
~r
cr)
o
cot:~
v
cr)
cf)
o
~
'~
cf)
lg
cl)1~
~)
oco
Hc')
LOG~
CtDt~
clrDH
OCr)
CT)H
IH~-
H\-/
l~~
H~)
t!)~
c,)
(,)
c~o
~r~
c\;eq
cr)o
l
~
a)6~
COH
C~1~r
COC~I
cl)~r
Hco
o)~r
c\]o
c\]e
c~o
~r~~
~oc~]
H'~/
H'v
cr)~
~e~~:
cJ:)~
OO~
cqo
~;
H
t~
l~
C\]
.-h
14
Lr),
C\l
~dD
l
u:)
t!)
C:)
,:)
8
~r~~
cv)c,)
toce
coc~
Loco
Lr)c~
c~o
~
~
~
H~ CO~r
H~
HOO
Cr,CF)
LO~~
LOC~l
H(~
coco
o~l
cv)o
l
co~
tpco
Q~r
cf)o
~c')
H\-'
el
oS
~~Ho~
'~~1'*~~8
'~
t~~
~~:)
Si~Sf::q'
f~ u) '~~'
G';~ 8 o~
ce
~~
tr)Qq
coc~~
~a'q'f:1
c~qc:)
~
-C~'~:!
> q.*~
~ r:
\l
\J
~~~~::q'
~;
rl
~
~
~o
cJ:)
H
Lf)C~
o
~!~5
CJ '~~ .~
Lo~
o
H
l~
cr)
\l
O qJ'~ *
~)P~(:)FS
c,,
o
~~ ~)C~ 5~~
~5
HITOTSUBASHI JOURNAL OF ECONOMICS
20
[February
TABLE 2 TRADE INTENSITY
Importer
Exporter
US
CAN
4. 1528
4. 3911
4. 9228
US
EEC
UK
2. 7942
2. 6890
2. 6975
O. 7303
O. 8378
O. 7400
O. 5647
O, 6943
O. 8292
O. 5161
O. 5422
O. 5104
O. 8958
O. 5325
O. 6057
O. 6738
O. 3629
O. 3401
O. 2608
1. 7611
1. 5814
O. 3063
O. 2800
O. 2097
O, 1067
O. 1645
O. 1673
LAT AM
OTH EFTA O W EUR
O. 9502
l . 0448
CAN
5. 0091
5. 0571
5. 1588
LAT AM
3. 7104
3, 4132
2. 7492
O. 3385
O, 5533
O. 6140
1, 1516
1. 2015
2. 0021
O. 8261
O. 8238
O. 8147
O. 8532
O. 9455
O. 8300
O. 5242
O. 4082
O. 4300
O. 5979
O. 5174
O, 7362
O. 5605
O. 6399
O. 6427
O. 1970
O. 2290
O. 2262
O. 8418
O. 8430'
O. 6900
1. 5757
1. 6143
1. 7599
O 6610
O 6616
O 6801
2. 2970
2. 0577
1. 7521
1 . 4381
EEC
l. 3327
l. 3113
UK
O. 6035
O. 8017
O. 9901
l. 1503
1. 3455
O. 9643
O. 5897
O. 7301
O. 6519
O. 6756
O. 7534
O. 7657
1. 4282
1. 4287
1. 6212
2. 0961
1. 9451
1* 7756
OTH EFll'A
O. 6274
O. 5721
O. 6563
O 1856
O 2383
O 2736
O. 8242
O. 7269
O. 7003
1. 7126
1. 5278
1. 9601
1. 9888
2. 4124
1. 1858
1. 4384
1 . 2538
1. 6449
1. 5690
1_ 8095
1 . 5929
O W EUR
O. 7841
O. 7411
O. 7855
O 0704
O 1113
O 1220
O 6668
O 5922
O 8796
1. 3452
1. 1832
1. 1069
2. 4664
2. 6216
2. 5659
1. 0384
l. 0188
1. 1559
1. 0199
O. 8333
O. 6921
E EUR
O. 0576
O. 0489
O. 0658
O 0231
O 0293
O 0633
O 1642
O 5117
O 6685
O. 3122
O. 3150
O. 3284
O. 3664
O. 3473
O. 4177
O. 5678
O. 4459
O. 4393
1. 3781
O. 9473
1. 0727
S AFR
O. 6244
l. 1725
O. 8734
O 1208
O 2405
O 3585
O 0812
O 0909
O 0864
O. 9538
O. 8241
O. 7338
3. 2916
3. 4962
4. 3876
O. 2227
O. 2422
O. 2918
O. 1864
O. 3209
O. 8257
O 8019
O 7886
O 6274
O 1236
O Il50
O 1849
O. 0426
O. 1174
O. 1063
2 3531
2 0138
1 8052
2 3541
2 0759
1 9753
o. 5855
OTH AFR
o. 5575
O. 4145
O. 7918
O. 6971
1 8337
2 4173
2 5212
O 5119
O 6602
O 5904
O. 9006
1. 1361
O. 9549
o 2213
o 2192
o 2293
O 2895
O 3801
O 3472
o. 3009
o. 3522
o. 4280
O. 7017
O. 4992
O. 3781
O 6258
O 6356
O 4236
O 1793
O 5317
O 2789
O. 21SO
O. 2321
O. 2954
l 39eo
1 1895
l 1673
1 3097
1 7083
1 6989
o. 5066
o. 3eo5
o. 4129
O. 9620
1. 0899
O. 9785
1 . 2678
J
W ASIA
1 . 8620
o. 49n
OTH ASIA
1. 4059
1. 5718
O 2698
O 3429
O 3561
O. 2656
O. 2905
O. 2236
O 6215
O 4600
O 4392
1. 4245
1. 5299
1. 2393
O. 1780
O. 1791
O. 2039
O. 2824
O. 3626
O. 3106
CHN MX
O. OOOO
O. 0133
O..0132
O 0648
O 0590
O 1923
O. 0221
O. 5879
O. 8582
O 2290
O 2451
O 3877
O. 2199
O. 4477
O. 5101
O. 1888
O, 1840
O. 2764
O. 1101
O. 0727
O. 1682
ANZ
O. 6227
l. 0621
1. 0611
o. 3049
o. 3922
o. 3540
O. 0679
O. 1038
O, 2084
O 9811
O 7243
O 5615
4. 2413
3. 3SOO
3. 1096
O. 1412
O. 0919
O. 1201
O, 1716
O. 4361
O, 3485
This table is reproduced from Yamazawa, I
Intensity Analysis of World Trede Flow, this
STRUCTURAL CHANGES lN WORLD TRADE FLOWS
197l]
21
1955-56 60-62 65-67
OTH AFR
J
W ASIA OTH ASIA CHN MX ANZ
EEUR
S AFR
O. 0235
O. 0694
O. 0599
1. 1218
1. 1055
1. 2001
O. 3660
O. 6370
O. 6696
1. 9052
2. 3019
l. 9524
O 7929
l 0876
l 2217
O 9904
l 4263
1 6463
O OO11
O OOOO
O OOOO
O. 5780
l. 0673
l. 5380
O. 0889
O. 1130
O. 2916
O. 8299
O. 6946
O. 6476
O. 1031
O. 1213
O. 1005
O 9617
1 1087
1 0268
O Il50
O 1600
O 1202
O 2234
O 2515
O 3156
O. 0227
l. 2653
l. I143
O. 6715
l. 0060
O. 8991
O. 2019
O. 4592
O. 6250
O. 1020
O. 1006
O. 1468
O. 1221
O. 1415
O. 1400
1 1422
1 0541
1 1323
O 1002
O 1028
O 1852
O 0450
O 0759
O 1018
O. 0311
O. 7836
1. 2462
O. 0562
O. 1088
O. 0560
O. 3127
O. 3254
O. 3351
O. 7329
O. 8253
O. 8762
2. 1895
l. 7817
1. 36S4
O 2432
O 2603
O 2067
l. 1751
O 7029
O 4831
O 4197
O. 2766
O_ 3940
O. 5730
O. 4316
O. 4268
O. 4259
O. 1855
O. 2492
O. 2902
3. 8888
3. 9648
4. 3718
1. 9670
1. 8415
2. 2093
2. 0543
l . 6618
O. 2582
O. 3113
O. 3218
1 . 9476
1 6196
1 4455
1 1243
O 2165
O 3826
O 5742
5. SO47
4. 4358
3. 9019
O. 5712
O. 5110
O. 5120
O. 7196
O. 7836
O. 7304
O. 8155
O. 6901
O. 7840
O. 1808
O. 2455
O. 2522
O. 8100
O. 7706
O. 7877
O 4059
O 3777
O 3350
O 5555
O 2356
O 4196
O. 5993
O. 5835
O. 5379
1. 9243
l, 2874
l. 7662
O. 14SO
O. 1392
O. 2514
O 4244
O 5049
O 7158
O. 1640
O. 1388
O. 1962
O . 9020
1 . 9470
O. 1047
O. 2644
O. 2594
O 3097
O 1542
O 2400
O 2397
O 2167
O 2878
7. 4529
6, 0834
6. 1659
O. 0686
O. 0507
O. OOOO
O 3286
O 4498
O 7290
O. 0378
O. 2415
O. 3531
O. 4645
O. 2086
O. 3287
O. 4133
8 5673
5 2533
3 9_107
O 0289
O 0336
O 0432
3 7750
3 3419
3 8370
O. 7040
1. 7899
2. 2574
O. 1735
O. 2359
O. 4338
O. 2400
O OOOO
O 2202
O OOOO
O 4243
O 7155
O 5305
O. 4646
O. 4432
O. 6956
O. 8823
O. 7646
O. 4421
O. 4290
O. 3185
O 5279
O 7533
O 9620
O. 0236
O, 1681
O. 1731
4. 4858
4. 2970
3. 9208
1. 3847
o. eol6
2. 7577
O, 9198
l. 6440
2. 0445
1. 0056
O. 9734
O. 9062
O. 2072
O. 2503
O. 2700
2, 0180
1, 5416
l. 3690
1. 2446
1. 3365
1. 3848
l . 5603
o. 1802
o. 06eo
o. 0159
1 . 0589
O. 9642
O. 8155
O . 4752
O. 6165
O . 2862
O. 0898
O. 4757
O. 5063
O. 5789
l. 7200
1. 4477
l. 3051
O. 9266
1. 0656
1. 6997
O. 0542
O. 2218
O. 2518
l. 0740
1. 2313
l . 2485
2. 5452
l. 4224
l. 6558
O. 2361
O. 1871
o. Ieo8
l. 7978
l. 6205
l. 3006
O. 8773
O. 9979
O. 8877
2. 1351
2. 4566
3. 8962
O. 2178
O. 4642
O. 5947
O. 7094
O. 7915
O. 4961
O. 5404
O. 6015
O. 7923
3. 1311
3. 0161
3. 3134
o. 9967
1 . 0408
3. 9129
3. 7462
3. 5429
8. 2507
5. 6970
2. 5955
O. 0248
O. 0709
O. OOOO
O. 3867
O. 4968
l. 2963
l. 9508
O. 7363
3. 0153
o. 1622
o. 2839
o. 9861
2. 2017
2. 6945
4. 4956
O. 2826
O. 2423
O. 2417
O. 2337
O. 6805
O. 9024
O. 0099
O. 1667
O. 2204
3. 7468
4. 1138
3. 9729
o. 3eo8
o. 5202
O. 7022
O. 8995
Journal, Vol. 10, No. 2. Table 3 (p. 82-83)
O . 6720
1. 6894
1 . 4239
1 . 3643
5. 1311
3 6891
3. 1574
l . 0699
l . OI04
l 2992
1. 6065
l. 4946
O. 1747
O. 2794
O. 6522
o 3849
2 7200
3 5096
2. 3003
2. 6471
2. 7599