This article was downloaded by: [University Of Pittsburgh]

On: 05 November 2014, At: 17:36

Publisher: Routledge
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,
37-41 Mortimer Street, London W1T 3JH, UK

The International Trade Journal

Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/uitj20

ECONOMIC INTEGRATION AMONG THE ASIA-PACIFIC

ECONOMIC COOPERATION COUNTRIES: Linder Effect on
Developed and Developing Countries (1985–1999)
Donny Tang
Published online: 21 Jun 2010.

To cite this article: Donny Tang (2003) ECONOMIC INTEGRATION AMONG THE ASIA-PACIFIC ECONOMIC COOPERATION
COUNTRIES: Linder Effect on Developed and Developing Countries (1985–1999) , The International Trade Journal, 17:1, 19-49,
DOI: 10.1080/08853900390152791

To link to this article: http://dx.doi.org/10.1080/08853900390152791

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained
in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no
representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the
Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and
are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and
should be independently verified with primary sources of information. Taylor and Francis shall not be liable for
any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever
or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of
the Content.

This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any
form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://
www.tandfonline.com/page/terms-and-conditions
 ITJ 17(1) #6421

ECONOMIC INTEGRATION
AMONG THE ASIA-PACIFIC
ECONOMIC COOPERATION
COUNTRIES: Linder Effect on
Developed and Developing
Countries (1985–1999)
Downloaded by [University Of Pittsburgh] at 17:36 05 November 2014

Donny Tang

This study tests the income similarity effect developed by Lin-

der on the developed and developing APEC countries. A modified
gravity model is estimated for fourteen APEC countries based on
trade data from 1985 to 1999. The result confirms early prediction
that the income similarity effect is stronger for developed APEC
countries. The developed APEC countries with similar per capita
incomes tend to trade more with each other during the study pe-
riod. In addition, the level of trade between the APEC countries
can be predicted by their economic characteristics. The result re-
veals that trade between the APEC countries would increase as
their economies become more developed.

* * * * *

Donny Tang is Business Economics Instructor at the School of

Continuing Studies, University of Toronto, Canada.
ISSN: 0885-3908. THE INTERNATIONAL TRADE JOURNAL, Volume XVII, No. 1, Spring 2003 19
DOI: 10.1080/08853900390152791
 20 THE INTERNATIONAL TRADE JOURNAL

I. INTRODUCTION
For the past decade, we have witnessed major developments
in economic integration among all three continents (North Amer-
ica, Europe, and Asia-Pacific). In North America, substantial
trade liberalization has been made through the implementation
of the North American Free Trade Agreement (NAFTA) among
the United States, Canada, and Mexico. In Europe, the fifteen
Downloaded by [University Of Pittsburgh] at 17:36 05 November 2014

European countries have achieved the highest form of integra-

tion through the formation of the European Union (EU). In the
Asia-Pacific region, a formal regional trade arrangement—the
Asia-Pacific Economic Cooperation (APEC) grouping—has been
created to promote trade and investment liberalizations among
the fourteen member countries in 1989. The APEC grouping con-
sists of the major trading nations from three regions: America
(United States, Canada, and Mexico), Asia (Japan, South Korea,
Taiwan, Hong Kong, Singapore, Malaysia, Thailand, Indonesia,
and the Philippines), and Pacific (Australia and New Zealand).
In 1994, the APEC countries have finally agreed to form the Free
Trade Area (FTA) by 2010. Given the inclusion of the world’s
largest trading nations, the APEC grouping has the full poten-
tial to become the largest trade bloc after the EU in the next
decade.
There are two main objectives in this article. First, this
study will test the Linder hypothesis on the APEC developed
and developing countries for 1985–1999. As stated by the Linder
hypothesis, countries with similar per capita incomes would
trade more with each other due to similar demand patterns
(Linder, 1961). Previous studies found that income similarity
effect is more applicable to developed countries, as they mainly
produce tradable manufactured goods. However, the level of
trade increase among the developing countries has exceeded
that among the developed countries in recent years. This study
 Economic Integration Among . . . 21

will investigate whether the income similarity effect may be

equally applicable to both country groups. Second, this study
will examine the effect of the FTA on bilateral trade flows
between the member countries. In particular, the results would
suggest whether the implementation of the FTA has promoted
the income similarity effect among these countries. In fact, their
bilateral trade has substantially increased after some of the FTA
Downloaded by [University Of Pittsburgh] at 17:36 05 November 2014

provisions have become effective since the mid-1990s. It is likely

that the APEC FTA may boost the income similarity effect by
increasing trade between the member countries.
The contributions of this article are two-fold. First, while nu-
merous studies have tested the validity of the Linder hypothesis
on trade in general, there seems to be a lack of similar stud-
ies on the APEC countries in particular. Since its inception in
1989, the APEC has become one of the largest trade blocs in the
world. Consistent with their trade-oriented growth policies, most
of the member countries have maintained high economic growth
in the past decade. The implementation of the FTA would prob-
ably lead to higher trade interdependence among these countries.
While a few recent studies have focused on the trade enhancing
effect of the FTA, none of them has examined the income simi-
larity effect on the member countries. This article will contribute
to the literature in this area by testing the Linder hypothesis on
the APEC developed and developing countries respectively.
Second, in contrast to previous work, the estimation model
in this study will analyze the income similarity effect on trade
for different time periods. There has been substantial trade
expansion within and outside the APEC grouping since the
FTA implementation in 1990. As their share of world trade
has steadily increased from 39 percent in 1990 to 47 percent
in 1999, their overall trade patterns would have important
implications for other countries. In recent years, there is some
evidence to suggest that the level of trade increase among the
 22 THE INTERNATIONAL TRADE JOURNAL

developing countries has exceeded that among the developed

countries. Hence, this article will extend the Linder hypothesis
by comparing the changes in income similarity effect for three
sub-periods.
The remainder of this article is organized as follows: Sections
2 and 3 describe the APEC development and review the literature
on the Linder hypothesis and the APEC FTA. Section 4 outlines
Downloaded by [University Of Pittsburgh] at 17:36 05 November 2014

the specification of the modified gravity model and its data

source. Section 5 presents the empirical results and discusses
the economic and statistical significance of these results. The
summary of major findings and suggestions for future research
will be presented in Section 6.

II. APEC DEVELOPMENT

In 1989, the APEC was formed initially as a trade discussion
forum among the Asia-Pacific countries. As an informal trade fa-
cilitation group, finance ministers from the APEC countries hold
annual meetings to discuss trade and investment liberalization is-
sues. The APEC includes the major trading nations from three
continents: North and South America (United States, Canada,
Mexico, Chile, and Peru), Pacific (Australia, New Zealand, and
Papua New Guinea), and Asia (Japan, South Korea, Singapore,
Taiwan, Hong Kong, Malaysia, Thailand, Indonesia, the Philip-
pines, China, Russia, and Vietnam). Because of its global trade
orientation, the APEC has continued to admit new member
countries from different regions. With the inclusion of the lead-
ing trading nations, the APEC has the full potential to become
the world’s largest trading bloc second only to the EU.
Since its inception, the APEC adopts the open regionalism
rather than preferential trade approach to promote trade lib-
eralization with both member and non-member countries. The
APEC would extend the tariff reduction measures to the mem-
 Economic Integration Among . . . 23

ber and non-member countries unconditionally. Hence, the open

trade policy would help to facilitate higher trade liberalization
not only for the member countries but also for the non-member
countries from around the world (Cheong, 1998). On the other
hand, the APEC has progressed very rapidly from an informal
trade forum to a formal regional trade arrangement. Within five
years since its formation, the APEC has agreed to the formation
Downloaded by [University Of Pittsburgh] at 17:36 05 November 2014

of APEC FTA in the 1994 summit meeting. The full-scale FTA

will be implemented for the developed countries by 2010 and for
the developing countries by 2020 (Krueger, 1999).
The APEC countries have shown their commitments in trade
integration in the three summit meetings during 1994–1996. In
1994, the APEC clearly outlines the main objectives and time
frame of launching the FTA. At the subsequent summit in 1995,
they specify more detailed procedures necessary to implement
the objectives of the FTA. The main provisions under the FTA
consist of two components: trade liberalization and economic and
technical cooperation. The trade liberalization process includes
a comprehensive coverage of border and domestic measures. In
particular, it involves policy measures on tariff and non-tariff re-
ductions such as custom procedures, rules of origin, standards
and conformance, and competition policy. The subsequent pro-
visions include the procedures to complete the implementation
of the full-scale FTA by the year 2010. In sum, the direct con-
sequence of the APEC FTA would not only limit to trade and
welfare gains from trade barrier elimination, but also involve har-
monization of existing trade policies among the member coun-
tries (Ethier, 1998). The FTA would provide member countries
a strong incentive to trade more, as it would replace the various
trade preferences currently in use with only one set of uniformed
trade liberalization measures.
In addition to the inclusion of major trading nations, the
APEC has tried to develop closer economic ties with the EU. In
 24 THE INTERNATIONAL TRADE JOURNAL

fact, the APEC countries especially the East Asian and Pacific
member countries have started to hold summit meetings with
the EU to promote higher trade liberalization. If the APEC and
EU groupings would indeed progress to higher trade integration,
it will be the largest trade bloc in the world given their total
trade accounting for more than 80 percent of the world trade
(Langhammer, 1999). The APEC which adopts open regionalism
Downloaded by [University Of Pittsburgh] at 17:36 05 November 2014

principle would eventually pursue global trade integration in the

next several decades.

III. LITERATURE REVIEW

This literature review will focus on the Linder hypothesis
and the APEC integration respectively. First of all, there has
been a lack of empirical studies to test the Linder hypothesis
on the APEC countries. Nevertheless, previous studies which
include some of the APEC countries as samples can shed some
light on this issue. Recent studies have examined whether the
income similarity effect would be equally applicable to both
developed and developing countries. Using the bilateral trade
for 35 countries, the research by Arnon et al. (1998) provides
clear evidence that the income similarity effect may apply
to both developed and developing countries. Contrary to the
Linder’s prediction, the developing countries have shown even
stronger income similarity effect than the developed countries
do. Due to their rapid economic growth in recent years, the
developing countries have shifted their major exports from
primary goods to manufactured goods. Not surprisingly, trade
among the developing countries may grow much faster than that
among the developed countries.
Moreover, given the inconclusive findings in this issue, Chow
et al. (1999) examined the income similarity effect on promoting
trade between the developed countries in East Asia and Europe.
 Economic Integration Among . . . 25

They concluded that the income similarity effect may only apply
to the developed countries with extremely small differences in
per capita incomes. The developed countries with similar per
capita incomes tend to have very similar demand patterns that
can facilitate high trade between them.
Most of the recent research has analyzed the effect of the
APEC FTA on member countries. Previous studies focused on
Downloaded by [University Of Pittsburgh] at 17:36 05 November 2014

the trade enhancing effect of the FTA implementation since the

mid-1990s. Using more recent data sets, recent studies (Endoh,
2000; Sharma et al., 2000; Soloaga et al., 2001) have applied
the gravity model to examine whether the FTA has led to
higher trade among the member countries. They concluded
that the partial implementation of the FTA has not caused
any significant trade increase between the member countries
during the 1990s. This finding is not entirely surprising given
the fact that a more substantial number of FTA provisions
have yet to be implemented. Nonetheless, the results in these
studies conclude that the high level of trade between the APEC
countries has appeared even before the APEC formation. This is
particularly true for the member countries with extensive trade
to spur their economic growth: United States, Canada, Japan,
Mexico, Australia, South Korea, Singapore, Hong Kong, Taiwan,
Malaysia, and Thailand.
Given its major impact on global trade, most empirical
studies have estimated the trade-related benefits caused by the
FTA implementation. Using the applied general equilibrium
analysis, Adams (1998) estimated the potential FTA effects on
the intra-industry trade among countries from fourteen regions.
The overall results suggest that the FTA implementation would
lead to high trade and income growth among the member
countries. In particular, he found that the developing member
countries may benefit even higher trade and income increases
compared to the developed member countries. The developing
 26 THE INTERNATIONAL TRADE JOURNAL

countries with more trade protection measures would derive

higher benefits as a result of the complete tariff reductions. A
recent study by Scollay et al. (2000) reached similar conclusion
on this issue. Their studies primarily focused on the effect of
APEC trade liberalization on welfare gains. As expected, the
implementation of the FTA would increase the welfare gains
for all member countries, with a larger effect on the developing
Downloaded by [University Of Pittsburgh] at 17:36 05 November 2014

member countries in particular. Additional results reveal that the

welfare gains on member countries would increase even further
as the FTA includes more new members in the future. In light
of its global trade approach, a majority of these studies confirm
that the full-scale FTA would benefit all member countries as
well as non-member countries in the long run.

IV. METHODOLOGY
The gravity model has been widely used to predict trade flows
between two countries. The amount of bilateral trade largely
depends on economic and geographical proximity variables. The
gravity model specifies that trade between two countries would
increase directly with their incomes but decrease with their
geographical distance. Later gravity models add the population
and regional dummy variables to better measure trade flows
among various trade blocs (Anderson, 1979; Bergstrand, 1985;
Linneman, 1966).
This study will examine whether the income similarity effect
exists among the APEC developed and developing countries
in particular. As suggested by previous studies, the income
similarity effect may be more applicable to developed countries
as their products are mostly tradable products. This study will
employ the modified gravity model specified by Arnon et al.
(1998) to test the income similarity effect on fourteen APEC
countries for the period 1985–1999. The specification of the
 Economic Integration Among . . . 27

modified gravity model includes the income similarity variable to

measure the differences in per capita incomes between member
countries. Under the Linder hypothesis, the member countries
with similar incomes would trade more with each other than
those with different incomes.
The use of the modified gravity model is highly appropriate
for several reasons. First, to control for the time effects on trade,
Downloaded by [University Of Pittsburgh] at 17:36 05 November 2014

the modified gravity model will be estimated for the entire period
(1985–1999) and the three sub-periods (1985–1989, 1990–1994,
and 1995-1999), respectively. The recession and financial crises
in the early and late 1990s may bias the results in this study. The
estimation for different time periods would avoid and reduce the
cyclical effects in the bilateral trade data. Moreover, this analysis
would allow us to better compare the trade enhancing effect of
the APEC on member countries over time. A comparison of the
results from pre- and post-1990 periods would indicate whether
the APEC FTA has caused any trade increase after its formation
in 1989.
Second, to identify the country effects on trade, the modified
gravity model will be estimated separately for three country
groups: high-income, low-income, and different-income groups.
As suggested by the Linder hypothesis, the results would indicate
whether a strong income similarity effect exists among the high-
income countries, whereas a weak income similarity effect among
the low-income countries. Finally, the results would suggest
whether the APEC FTA has promoted the income similarity
effect on trade between these countries. It is expected that the
implementation of the FTA provisions would contribute to the
income similarity effect by increasing trade between the member
countries with similar incomes.
The modified gravity model in this study includes the two ba-
sic gravity variables (income and distance variables) to explain
trade flows between the APEC countries. The model will also in-
 28 THE INTERNATIONAL TRADE JOURNAL

clude the income similarity variable to measure the income simi-

larity effect on member countries with different per capita incomes.
The modified gravity model used in this study is given as:

(1) log(Tradeij ) = B0 + B1 (log GDPi ) + B2 (log GDPj )

+ B3 (log Distij )
+ B4 (log GDPCi − log GDPCj )2 + eij
Downloaded by [University Of Pittsburgh] at 17:36 05 November 2014

where Tradeij is export values from country i to country j;

GDPi and GDPj are the incomes of exporting and importing
countries i and j; Distij is the physical distance between exporting
and importing countries i and j; (log GDPCi − log GDPCj )2
measures the income similarity effect between countries i and
j, where GDPC is the GDP per capita for countries i and j; and
finally eij measures the error term. All variables are estimated in
logarithms.
Eq. (1) is estimated by the ordinary least squares (OLS)
method for the entire period and the three sub-periods. The
sample countries in this study include the fourteen APEC
countries: the United States, Canada, Japan, Australia, New
Zealand, China, Hong Kong, Malaysia, Singapore, South Korea,
Mexico, Indonesia, the Philippines, and Thailand. A total of
seven other APEC countries are omitted from this analysis. In
particular, Russia, Peru, and Vietnam are excluded in this study
as they joined the APEC only recently. The true intraregional
trade effect among these countries will not be significant in such
a short period of time. Moreover, Taiwan, Brunei, Papua New
Guinea, and Chile are excluded due to data unavailability.
The dependent variable, Tradeij , is the bilateral trade (ex-
port) flows between the APEC countries.1 To adjust for inflation

1 Export figures are obtained from the Direction of Trade Statistics (Interna-

tional Monetary Fund, various issues).

 Economic Integration Among . . . 29

during the sample period, the export figures are measured in con-
stant U.S. dollars. The export figures are converted to constant
1990 prices by using the U.S. GDP deflators.2
The first two independent variables, GDPi and GDPj , are
the income variables for the exporting and importing countries
in GDP figures. The GDP figures are again measured in constant
U.S. dollars.3 The GDP figures are converted to constant 1990
Downloaded by [University Of Pittsburgh] at 17:36 05 November 2014

prices by using the U.S. GDP deflators. The GDP variables

measure the size of their economies. The GDPi reflects the export
supply of exporting country. The amount of trade should increase
with the size of economy. Therefore, the GDPi should have a
positive effect on the trade flow. In addition, the GDPj reflects
the import demand of the importing country. The amount of
trade should increase with the size of economy. Therefore, the
GDPj should also have a positive effect on the trade flow. The
GDP figures are measured in nominal U.S. dollars rather than
in purchasing power parity (PPP) units.4 Although the PPP-
based GDP figure is a useful indicator to measure the real wealth
of a country, it may not accurately reflect the export supply
and import demand situations of a country. The total amount of
goods and services that a country can trade with another country
depends largely on the GDP figures in nominal U.S. dollars (Gros
et al., 1996). Therefore, the GDP figures in nominal U.S. dollars
would better measure the trade potentials of a country.

2 U.S. GDP deflators are obtained from the World Development Indicators

CD-ROM (World Bank, 2001).

3 GDP figures are taken from the World Development Indicators CD-ROM

(World Bank, 2001).

4 The coefficients on incomes are also estimated using the PPP-based GDP

figures. The overall results are similar to those using the GDP figures in nominal
U.S. dollars. However, the explanatory power of the regression models with GDP
figures in nominal U.S. dollars is consistently higher than that with PPP-based
GDP figures (R2 = 0.60 compared to R2 = 0.40). The results are available upon
request.
 30 THE INTERNATIONAL TRADE JOURNAL

The distance variable, Distij , represents the level of geograph-

ical proximity between exporting and importing countries, as
measured by their physical distance in kilometers.5 The bilat-
eral trade between distant countries would decrease as the larger
distances between these countries would result in higher trans-
portation costs. Conversely, due to lower transportation costs,
the neighboring countries would trade more with each other.
Downloaded by [University Of Pittsburgh] at 17:36 05 November 2014

Therefore, distance should have a negative effect on the trade

flow.
The most important variable in this study is the income
similarity variable, (log GDPCi − log GDPCj )2 . The level of
income similarity between exporting and importing countries can
be computed by taking the differences in their GDP per capita
incomes. The GDP per capital figures are obtained from dividing
the GDP figures by the population figures.6 In particular,
countries with similar per capita incomes should have a small
difference in GDP per capita incomes, whereas countries with
different per capita incomes should have a large difference in
GDP per capita incomes. As stated by the Linder hypothesis,
countries with similar incomes would have higher trade flows
between them as they are more likely to share similar demand
patterns (Arnon et al., 1998). Therefore, the income similarity
variable should be negatively related to the trade flow.
The Linder hypothesis suggests that the income similarity
effect may be more applicable to the high income countries as
their products are mainly tradeable goods. To test the validity of
this hypothesis, the APEC countries are classified into either high
income or low income countries based on their Gross National

5 The data on physical distance are obtained from the Surface Distance between

Points of Latitude and Longitude Database (Swedish University of Agricultural

Science, 2001).
6 The data on the U.S. GDP, U.S. GDP deflator, and population figures are

obtained from the World Development Indicators CD-ROM (World Bank, 2001).
 Economic Integration Among . . . 31

Product (GNP) per capita figures.7 The APEC countries are

considered as high income countries if their U.S. GNP per
capita is more than $20,000. Otherwise, the APEC countries are
considered as low income countries. The high income countries in
this study include the United States, Canada, Japan, Australia,
Singapore, and Hong Kong, while the low income countries
include New Zealand, Malaysia, South Korea, China, Indonesia,
Downloaded by [University Of Pittsburgh] at 17:36 05 November 2014

Mexico, the Philippines, and Thailand.

As mentioned earlier, this article will compare the income
similarity coefficients between the high-income and low-income
countries for the three sub-periods. A direct comparison of these
regression coefficients may be inappropriate as these coefficients
are estimated from different equations. To properly compare
the magnitude of the coefficients, the regression coefficients are
converted to the standardized regression coefficients (Tamhane
et al., 2000). The standardized regression coefficients (B∗i ) are
computed by scaling the coefficients of independent variables (Bi )
using:

(2) B∗i = Bi (Sxi /Sy )

where Sy and Sxi are the sample standard deviations of the de-
pendent variable and independent variable, respectively. Hence,
the results of the OLS coefficients in Eq. (1) are transformed to
standardized regression coefficients by using Eq. (2).

V. RESULTS
OLS Regression Results
Table I reports the results for the entire period (1985–1999),
whereas Tables II–IV report the results for the three sub-periods

7 The data on the U.S. GNP per capita figures are obtained from the World

Development Indicators (World Bank, 1998).

 Downloaded by [University Of Pittsburgh] at 17:36 05 November 2014

Table I
Regression Coefficients for Income Similarity Effects: 1985–1999
Trade between Similar Countries

Trade between All High-Income Low-Income Trade between

Variables All Countries Countries Countries Countries Different Countries

Intercept −1.18 −3.52 2.72 3.73 1.71

(−3.92) (−6.68) (4.67) (4.64) (5.29)
Log GDPi 0.73∗ 0.75∗ 0.52∗ 0.43∗ 0.63∗
(48.46) (29.24) (24.41) (9.95) (37.59)
Log GDPj 0.77∗ 0.83∗ 0.53∗ 0.69∗ 0.65∗
(50.87) (32.28) (24.84) (15.67) (37.70)

32
Log Distij −1.19∗ −1.06∗ −0.89∗ −1.36∗ −1.19∗
(−45.74) (−24.23) (−19.96) (28.07) (−41.81)
Log (GDPCi − GDPCj )2 −0.0044 −0.09∗ −0.13† 0.03† 0.01
(−0.872) (−6.69) (−2.50) (2.34) (0.99)
Adjusted R2 0.67 0.68 0.76 0.55 0.70
F-Statistics 1362.05 665.36 352.74 249.89 824.72
Observations 2711 1271 450 821 1440

Note: t-statistics are in parentheses.

∗ Significant at the 1 percent level.
† Significant at the 5 percent level.
‡ Significant at the 10 percent level.
 Downloaded by [University Of Pittsburgh] at 17:36 05 November 2014

Table II
Regression Coefficients for Income Similarity Effects: 1995–1999
Trade between Similar Countries

Trade between All High-Income Low-Income Trade between

Variables All Countries Countries Countries Countries Different Countries

Intercept −0.78 −2.83 2.32 7.22 1.72

(−2.44) (−4.49) (2.07) (5.24) (2.96)
Log GDPi 0.72∗ 0.73∗ 0.58∗ 0.35∗ 0.60∗
(43.84) (21.47) (13.07) (5.01) (20.54)
Log GDPj 0.77∗ 0.89∗ 0.60∗ 0.56∗ 0.63∗
(46.04) (25.95) (13.51) (8.05) (21.05)

33
Log Distij −1.26∗ −1.24∗ −1.01∗ −1.42∗ −1.12∗
(−46.14) (−25.31) (−12.54) (18.97) (−23.57)
Log (GDPCi − GDPCj )2 −0.01† −0.03† −1.64† 0.01 0.02†
(2.52) (−2.20) (−2.42) (0.38) (2.34)
Adjusted R2 0.67 0.64 0.75 0.60 0.69
F-Statistics 1212.72 440.33 112.20 106.69 269.86
Observations 908 428 150 278 480

Note: t-statistics are in parentheses.

∗ Significant at the 1 percent level.
† Significant at the 5 percent level.
‡ Significant at the 10 percent level.
 Downloaded by [University Of Pittsburgh] at 17:36 05 November 2014

Table III
Regression Coefficients for Income Similarity Effects: 1990–1994
Trade between Similar Countries

Trade between All High-Income Low-Income Trade between

Variables All Countries Countries Countries Countries Different Countries

Intercept −1.29 −4.80 2.96 4.96 1.99

(−3.68) (−7.70) (2.86) (3.29) (3.54)
Log GDPi 0.73∗ 0.80∗ 0.52∗ 0.25∗ 0.61∗
(41.08) (25.06) (13.48) (3.16) (20.91)
Log GDPj 0.78∗ 0.88∗ 0.52∗ 0.66∗ 0.61∗
(43.52) (27.67) (13.32) (8.17) (20.44)

34
Log Distij −1.23∗ −1.07∗ −0.89∗ −1.22∗ −1.16∗
(−40.40) (−19.41) (−11.40) (−14.78) (−23.50)
Log (GDPCi − GDPCj )2 −0.01 −0.08∗ −0.18∗ 0.02 0.01
(−0.68) (−4.76) (−3.18) (0.85) (0.99)
Adjusted R2 0.66 0.70 0.74 0.49 0.67
F-Statistics 1016.35 509.39 107.76 66.39 251.35
Observations 908 428 150 278 480

Note: t-statistics are in parentheses.

∗ Significant at the 1 percent level.
† Significant at the 5 percent level.
‡ Significant at the 10 percent level.
 Downloaded by [University Of Pittsburgh] at 17:36 05 November 2014

Table IV
Regression Coefficients for Income Similarity Effects: 1985–1989
Trade between Similar Countries

Trade between All High-Income Low-Income Trade between

Variables All Countries Countries Countries Countries Different Countries

Intercept −2.09 −4.07 1.89 5.76 2.10

(−5.15) (−6.36) (2.12) (4.07) (3.81)
Log GDPi 0.74∗ 0.76∗ 0.54∗ 0.40∗ 0.62∗
(36.90) (26.67) (16.56) (4.83) (21.13)
Log GDPj 0.80∗ 0.78∗ 0.55∗ 0.53∗ 0.65∗
(40.03) (27.32) (16.97) (6.29) (21.20)

35
Log Distij −1.16∗ −0.90∗ −0.88∗ −1.39∗ −1.25∗
(−33.03) (−15.55) (−12.02) (−15.18) (−24.46)
Log (GDPCi − GDPCj )2 −0.03∗ −0.15∗ −0.73∗ 0.07∗ −0.004
(−3.96) (−6.79) (−3.53) (3.01) (−0.49)
Adjusted R2 0.66 0.74 0.79 0.47 0.68
F-Statistics 839.59 512.63 144.91 59.67 258.03
Observations 895 415 150 265 480

Note: t-statistics are in parentheses.

∗ Significant at the 1 percent level.
† Significant at the 5 percent level.
‡ Significant at the 10 percent level.
 36 THE INTERNATIONAL TRADE JOURNAL

(1985–1989, 1990–1994, and 1995–1999). All the regression mod-

els fit the data very well as the independent variables explain
more than 65 percent of the variation in the trade flows. As
predicted by the Linder hypothesis, the results would indicate
whether the APEC countries with similar per capita incomes
would trade more with each other. In addition, the results for
the period 1990–1999 would reveal whether the APEC FTA has
Downloaded by [University Of Pittsburgh] at 17:36 05 November 2014

boosted the income similarity effect on the intra-APEC trade.

First, the most significant finding of this study is related
to the income similarity effect among the high income and low
income countries. The results support the Linder hypothesis
that the income similarity effect is only confirmed for the high-
income countries. As shown in Table I, the coefficient on the
income similarity (−0.13) is negative and statistically significant
for the entire period 1985–1999, indicating that the high-income
countries with similar per capita incomes would trade more with
each other as expected. In contrast, the low income and different
income countries have shown no evidence of income similarity
effect during the study periods. In fact, the coefficients for the
low income and different countries are positive although they are
only statistically significant in the first sub-period 1985–1989.
Contrary to the Linder’s prediction, this suggests that these
countries with large difference in per capita incomes may trade
more with each other. In general, the overall results are consistent
with the Linder hypothesis that income similarity effect is only
applicable to the high income countries, as their main outputs
are mostly tradable products (Arnon et al., 1998).

Standardized Regression Results

As mentioned earlier, a direct comparison of the income sim-
ilarity coefficients across different country groups may be inap-
propriate as these coefficients are estimated from different equa-
tions. To properly compare the magnitude of the coefficients,
 Economic Integration Among . . . 37

the regression coefficients are normalized to become the stan-

dardized regression coefficients. The OLS coefficients in Tables
I–IV are converted to standardized coefficients by using Eq. (2).
The standardized regression results are presented in Tables V–
VIII (Table V for 1985–1999 and Tables VI–VIII for 1985–1989,
1990–1994, and 1995–1999, respectively). All the standardized
coefficients have the expected signs similar to those in Tables I–
Downloaded by [University Of Pittsburgh] at 17:36 05 November 2014

IV. Among the three country groups, the income similarity effect
is only confirmed for the high income countries. As shown in Ta-
ble V, the coefficient on income similarity is negative for the
high-income countries, whereas it is positive for the low-income
and different-income countries for the entire period. This indi-
cates that high-income countries with similar per capita incomes
would trade more with each other. Moreover, the income simi-
larity effect seems to appear for the high-income countries after
1990, as the coefficient changes from positive in 1985–1989 to neg-
ative value in 1990–1999. The APEC FTA has caused higher in-
come similarity effect on the high-income countries. In particular,
the implementation of the FTA since 1994 has probably boosted
the income similarity effect by increasing trade among the high-
income countries. In sum, the result is consistent with the fact
that the high-income countries have established closer trade and
economic ties among themselves even before the APEC forma-
tion. The export-oriented member countries (the United States,
Japan, Australia, Hong Kong, and Singapore) have traded ex-
tensively with each other as early as in the mid-1980s (Soloaga
et al., 2001). The income similarity effect along with the FTA im-
plementation has probably contributed to the consistently high
trade flows among the high-income countries.
In contrast to the high-income countries, the low-income
countries have experienced no income similarity effect over the
entire period. The coefficient on income similarity is positive
for 1985–1999, indicating that the low-income countries with
 Downloaded by [University Of Pittsburgh] at 17:36 05 November 2014

Table V
Standardized Regression Coefficients for Income Similarity Effects: 1985–1999
Trade between Similar Countries

Trade between All High-Income Low-Income Trade between

38
Variables All Countries Countries Countries Countries Different Countries

Log GDPi 0.55 0.48 0.57 0.23 0.60

Log GDPj 0.58 0.53 0.58 0.37 0.61
Log Distij −0.53 −0.39 −0.46 −0.67 −0.67
Log (GDPCi − GDPCj )2 −0.01 −0.11 −0.06 0.05 0.01
 Downloaded by [University Of Pittsburgh] at 17:36 05 November 2014

Table VI
Standardized Regression Coefficients for Income Similarity Effects: 1995–1999
Trade between Similar Countries

Trade between All High-Income Low-Income Trade between

39
Variables All Countries Countries Countries Countries Different Countries

Log GDPi 0.53 0.42 0.59 0.19 0.57

Log GDPj 0.57 0.51 0.61 0.31 0.59
Log Distij −0.57 −0.49 −0.52 −0.72 −0.66
Log (GDPCi − GDPCj )2 0.03 −0.04 −0.11 0.01 0.06
 Downloaded by [University Of Pittsburgh] at 17:36 05 November 2014

Table VII
Standardized Regression Coefficients for Income Similarity Effects: 1990–1994
Trade between Similar Countries

Trade between All High-Income Low-Income Trade between

40
Variables All Countries Countries Countries Countries Different Countries

Log GDPi 0.54 0.49 0.57 0.14 0.60

Log GDPj 0.57 0.54 0.57 0.36 0.60
Log Distij −0.54 −0.37 −0.48 −0.64 −0.68
Log (GDPCi − GDPCj )2 −0.01 −0.09 −0.13 0.04 0.03
 Downloaded by [University Of Pittsburgh] at 17:36 05 November 2014

Table VIII
Standardized Regression Coefficients for Income Similarity Effects: 1985–1989
Trade between Similar Countries

Trade between All High-Income Low-Income Trade between

41
Variables All Countries Countries Countries Countries Different Countries

Log GDPi 0.52 0.53 0.64 0.23 0.62

Log GDPj 0.57 0.54 0.65 0.30 0.65
Log Distij −0.48 −0.30 −0.46 −0.76 −0.71
Log (GDPCi − GDPCj )2 −0.06 −0.14 0.14 0.14 −0.01
 42 THE INTERNATIONAL TRADE JOURNAL

similar per capita incomes do not trade more with each other. As
suggested by the Linder hypothesis, the income similarity effect
may not apply to the low-income countries as their main outputs
are mostly non-tradable raw materials products. Moreover, their
incentive to trade with each other has been rather weak given
the fact that most of them are resource-rich countries.
On the other hand, the result which shows low trade among
Downloaded by [University Of Pittsburgh] at 17:36 05 November 2014

the low-income countries may be somewhat surprising. The rapid

economic growth among these countries has drastically changed
their export structures in recent years. As these countries become
more industrialized, they would shift their primary exports from
raw materials to manufactured products (Choudhry et al., 2000).
In particular, substantial trade increase has occurred among
the industrializing low-income countries with similar per capita
incomes: Malaysia, South Korea, Indonesia, and Thailand.
Second, the positive coefficient on income similarity also
suggests that the low-income countries with large difference in
per capita incomes may trade more with each other. In fact,
Malaysia and South Korea trade substantially more with China
and Philippines than with Indonesia and Thailand. This trade
enhancing effect appears to be strong in the late 1980s but
declines considerably in the 1990s. It is worth noting the huge
decrease in the coefficient from 0.14 in 1985–1989 to 0.01 in 1995–
1999, indicating that the low-income countries with different
per capita incomes decrease trade with each other. While these
countries gradually expand trade with each other, they still
maintain a significant portion of trade with the high-income
countries (Sharma et al., 2000). This is most evident among the
high-growth low-income countries. As more of these countries
become industrialized, the bilateral trade between them would
further increase.
Finally, the results show no evidence of income similarity
effect on the different income countries. The coefficient on income
 Economic Integration Among . . . 43

similarity is again positive for 1985–1999, indicating that trade

would increase among countries with large differences in per
capita incomes. This suggests that the high-income countries
would trade more with the low-income countries. In fact, there
is a small increase in trade between them after 1994, as the
coefficient increases from 0.03 in 1990–1994 to 0.06 in 1995–
1999. The closer trade linkages between these countries may
Downloaded by [University Of Pittsburgh] at 17:36 05 November 2014

be explained by their high dependence on trade to industrialize

their economies (Greenaway et al., 1998; Krueger, 1998). This
is especially true for the low-income countries such as Malaysia,
South Korea, Thailand, Indonesia, and China. To promote high-
growth performance, these low-income countries have maintained
a considerable amount of trade with the high-income countries,
United States, Japan, Australia, and Singapore in particular.
There is some evidence to suggest that their overall trade
will increase in the next two decades. First, trade among these
countries has been growing steadily in the past decade. In fact,
the annual rate of trade increase among the APEC countries has
exceeded that among the EU countries since 1990 (Frankel et al.,
1997). The substantial trade increase has been largely sustained
by their high growth performances during these periods. More-
over, trade between the APEC countries would likely increase as
their exports to other trade blocs gradually decrease. The issue
of common currency within the EU will help to promote high
trade among its member countries. The EU monetary integra-
tion along with its expansion toward East Europe may result
in lower trade with the APEC countries. To maintain current
level of trade increase, the APEC countries would expand trade
not only with each other, but also with the potential member
countries in other regions.
Second, the implementation of the full-scale APEC FTA will
likely increase trade flows and welfare gains among the member
countries (Adams, 1998; Scollay et al., 2000). In particular, the
 44 THE INTERNATIONAL TRADE JOURNAL

low-income countries with high trade protection measures may

experience a larger trade increase compared to the high-income
countries. More open trade would allow them to have access to
larger markets for their exports. Furthermore, when the low-
income countries become more industrialized, they would need to
import more intermediate goods from the high-income countries
to produce manufactured goods for exports. Hence, the low-
Downloaded by [University Of Pittsburgh] at 17:36 05 November 2014

income countries would probably increase trade with the high-

income countries.
Tables V–VIII also report the standardized coefficients on the
GDP and distance variables. All of these coefficients have similar
results for the three country groups. First, the GDP coefficients
on the exporting and importing countries are indeed positive and
remain about the same magnitude for all sub-periods (around
0.55 and 0.58, respectively). This indicates that trade among
the APEC countries would increase directly with their incomes.
As these countries become more developed, they would possibly
increase the investment in infrastructure improvements, which
can further enhance trade among these countries (Choudhry
et al., 2000). Second, the trade enhancing effects of incomes are
higher for the high-income and different-income countries, as the
magnitudes of their income coefficients are larger than those for
the low-income countries during 1985–1999. High trade among
the high-income countries is hardly surprising given the fact that
they produce mainly tradable goods for exports. Moreover, high
trade between the different-income countries (i.e., low-income
and high-income countries) can be attributed to the use of
export-oriented growth policies among the low-income countries.
As mentioned earlier, these countries have traded extensively
with the high-income countries in order to industrialize their
economies. This has helped these countries to maintain high
growth performances for the past two decades.
Finally, the distance coefficients are negative for all three
country groups over the entire period. This confirms that the
 Economic Integration Among . . . 45

neighboring APEC countries would trade more with each other as

shorter distance would reduce the transportation costs between
them (Bayoumi et al., 1997). Interestingly, the magnitudes of
the coefficients are higher for both low-income and different-
income countries compared to the high-income countries (−0.66
and −0.68, respectively, compared to −0.37 for high-income
countries).
Downloaded by [University Of Pittsburgh] at 17:36 05 November 2014

Physical distance seems to be a major trade resistance factor

for the low-income countries. As most of the low-income countries
are divided by water, the lack of modern port facilities would nat-
urally incur higher transportation costs among these countries
(Sharma et al., 2000). Nevertheless, the physical distance will not
become a trade resistance factor in the long-run for two reasons.
First, to facilitate high trade, the high-income countries have
started to use more advanced transport technology in order to re-
duce transportation costs significantly (Waters et al., 1990). The
rapid trade increase among these countries certainly justifies for
the use of more expensive transportation modes. Second, despite
the high transportation costs among them, the high-income coun-
tries have maintained steady trade with the low-income coun-
tries given their strong demand for various products from these
countries (Polak, 1996). This is particularly true for the United
States, Australia, Japan, Hong Kong, and Singapore. The rev-
enue generated from exports among the high-income countries
would allow them to pay for the high import costs. By and large,
it is expected that distance would not become a crucial factor in
predicting trade flows between these countries.

VI. CONCLUSIONS
Using the modified gravity model, this article examines
whether the income similarity effect may apply to the APEC
member countries for the period 1985–1999. This paper also
analyzes whether the implementation of APEC FTA in the 1990s
 46 THE INTERNATIONAL TRADE JOURNAL

has helped to promote trade between these countries. Overall,

the results are very consistent with the Linder’s prediction that
the income similarity effect is only confirmed for the high-
income countries. This suggests that the similarity of per capita
incomes has increased trade between the high-income countries
but not between the low-income countries. Given the trade-
oriented growth policy among the high-income countries, this
Downloaded by [University Of Pittsburgh] at 17:36 05 November 2014

result may not be surprising at all. Moreover, the implementation

of the FTA has facilitated the income similarity effect on the
high-income countries, as their bilateral trade has increased
considerably since 1990. As more of the FTA provisions will
become effective in this decade, trade between the high-income
countries will likely increase even further.
In light of the rapid trade integration, the study of the income
similarity effect on APEC would have significant implications
for both member and non-member countries. There are several
important issues in this area that future research can explore
further. First, this research suggests that the impact of the
FTA on income similarity effect will likely increase, as the full-
scale FTA will be implemented by 2010. In particular, the low-
income countries may experience substantial trade increase due
to the FTA effect. As a result, they may show stronger income
similarity effect comparable to those among the high-income
countries. A re-examination of this issue in the future will better
estimate the changes of income similarity effect among these
countries. Second, consistent with open regionalism principle,
the APEC member countries will expand trade with non-member
countries especially from other trade blocs. Given their consistent
high trade over the years, the APEC countries have tried to
establish more formal trade liberalization measures with the EU
countries. Future research can analyze the trade enhancing effect
of the possible APEC-EU free trade area. Thus, the estimation
technique used in this study can be used to test the income
 Economic Integration Among . . . 47

similarity effects on both trade blocs. These studies with a larger

country sample would provide more accurate assessment of the
income similarity effect on trade.

ACKNOWLEDGMENT
The author would like to thank two anonymous referees for
their valuable comments and suggestions. The usual disclaimer
Downloaded by [University Of Pittsburgh] at 17:36 05 November 2014

applies.

REFERENCES
Adams, P. 1998. Long-Run Effects of APEC Trade Liberaliza-
tion: An Applied General Equilibrium Analysis. World Econ-
omy 21(7):931–952.
Anderson, J. E. 1979. A Theoretical Foundation for the Gravity
Equation. American Economic Review 69:106–116.
Arnon, A. and Weinblatt, J. 1998. Linder’s Hypothesis Revisited:
Income Similarity Effects for Low Income Countries. Applied
Economics Letters 5:607–611.
Bayoumi, T. and Eichengreen, B. 1997. Is Regionalism Sim-
ply a Diversion? Evidence from the Evolution of the EC
and EFTA? In T. Ito and A. Krueger, eds. Regionalism ver-
sus Multilateral Trade Arrangements. University of Chicago
Press, Chicago.
Bergstrand, J. H. 1985. The Gravity Equation in International
Trade: Some Microeconomic Foundations and Empirical Ev-
idence. Review of Economics and Statistics 67:474–481.
Cheong, I. 1998. The Potential Obstacles to Economic Coopera-
tion in the Asia-Pacific Region. International Trade Journal
12(4):445–459.
Choudhry, S., Abu-Bakar, A., and Wylie, P. 2000. A Yen Bloc
in Pacific Asia: Natural Economic Symbiosis or Overblown
Rhetorics? Applied Economics Letters 7(4):215–218.
 48 THE INTERNATIONAL TRADE JOURNAL

Chow, P., Kellman, M., and Shachmurove, Y. 1999. A Test of the

Linder Hypothesis in Pacific NIC Trade 1965–1990. Applied
Economics Letters 31(2):175–182.
Endoh, M. 2000. The Transition of Postwar Asia-Pacific Trade
Relations. Journal of Asian Economics 10:571–589.
Ethier, W. 1998. The New Regionalism. Economic Journal 108:
1149–1161.
Downloaded by [University Of Pittsburgh] at 17:36 05 November 2014

Frankel, J. and Wei, S. J. 1997. Open Versus Closed Trade Blocs.

In T. Ito and A. Krueger, eds., Regionalism Versus Mul-
tilateral Trade Arrangements. University of Chicago Press,
Chicago.
Greenaway, D., Morgan, W., and Wright, P. 1998. Trade Reform,
Adjustment and Growth: What Does the Evidence Tell Us?
Economic Journal 108:1547–1561.
Gros, D. and Gonciarz, A. 1996. A Note on the Trade Potential
of Central and Eastern Europe. European Journal of Political
Economy 12:709–721.
International Monetary Fund. Direction of Trade Statistics.
International Monetary Fund (various issues), Washington,
D.C.
Krueger, A. O. 1998. Why Trade Liberalization Is Good for
Growth. Economic Journal 108:1513–1522.
Krueger, A. O. 1999. Are Preferential Trading Arrangements
Trade-Liberalizing or Protectionist? Journal of Economic
Perspectives 13(4):105–124.
Langhammer, R. 1999. Regional Integration APEC Style: Are
There Lessons to Learn from Regional Integration EU Style?
ASEAN Economic Bulletin 16(10):1–17.
Linder, S. B. 1961. An Essay on Trade and Transformation. John
Wiley, New York.
Linneman, H. 1966. An Econometric Study of International Trade
Flows. North Holland, Amsterdam.
 Economic Integration Among . . . 49

Polak, J. 1996. Is APEC a Natural Regional Trading Bloc? A

Critique of the Gravity Model of International Trade. World
Economy 19(5):533–543.
Scollay, R. and Gilbert, J. 2000. Measuring the Gains from APEC
Trade Liberalization: An Overview of CGE Assessments.
World Economy 23(2):175–197.
Sharma, S. C. and Chua, S. Y. 2000. ASEAN: Economic Inte-
Downloaded by [University Of Pittsburgh] at 17:36 05 November 2014

gration and Intra-Regional Trade. Applied Economics Letters

7(3):165–169.
Soloaga, I. and Winters, L. A. 2001. Regionalism in the Nineties:
What Effect on Trade? North American Journal of Economics
and Finance 12:1–29.
Swedish University of Agricultural Science. 2001. Surface Dis-
tance Between Points of Latitude and Longitude Database.
Department of Plant Protection Sciences, Uppsala, Sweden.
Tamhane, A. and Dunlop, D. 2000. Statistics and Data Analy-
sis: From Elementary to Intermediate. Prentice Hall, Upper
Saddle River, NJ.
Waters, C. and Soman, S. 1990. Containerized Freight Transport
in Developing Countries. Canadian Journal of Development
Studies xi(2):297–310.
World Bank. 1998. World Development Indicators. World Bank,
Washington, D.C.
World Bank. 2001. World Development Indicators CD-ROM.
World Bank, Washington, D.C.