impact analysis of normalized revealed comparative...

Working Paper in Economics and BusinessVolume V No. 3/2016

Impact Analysis of Normalized Revealed Comparative Advantageon ASEAN’sNon-Oil and Gas Export Pattern Using Gravity-Model Approach

Umar FakhrudinFithra Faisal Hastiadi

June 2016

University of Indonesia

Working Paper in Economics and BusinessChief Editor: Hera SusantiEditors: Muhammad Halley Yudhistira, Rus’an NasrudinSetting: Rini Budiastuti, Moslem Afrizal

Copyright c©2016, Department of EconomicsDepartment of Economics Building 2nd FloorDepokWest Java, Indonesia 16424Telp. 021-78886252Email: [email protected]: http://econ.feb.ui.ac.id/katagori/working-paper/

http://econ.feb.ui.ac.id/katagori/working-paper/

Contents

Contents 3

List of Tables 4

List of Figures 4

1 Introduction 1

2 Conceptual Framework 52.1 Comparative Advantage, Pattern Export and Normalized Revealed Comparative

Advantages (NRCA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52.2 Gravity Model in International Trade . . . . . . . . . . . . . . . . . . . . . . . . . 7

3 Empirical Research Review 8

4 Hypothesis 10

5 Methodology 10

6 Result and Analysis 126.1 Comparative Advantage of ASEAN Trade Commodities in ASEAN Market Based

on NRCA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126.2 Gravity Model Estimate with Additional Dummy and ∆NRCA . . . . . . . . . . . 15

7 Conlusion 22

8 References 23

List of Tables

1 Descriptive analysis of variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122 Grouping of SITC3 Commodities at Level 3 Digit Based on Production Factors and

Level of Technology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133 NRCA Results and Ranking of Entire ASEAN Commodities in ASEAN Market . . 164 FEM Estimates in Different Scenarios . . . . . . . . . . . . . . . . . . . . . . . . . 20

List of Figures

1 Intra-ASEAN Trade, 1962-2012 Source: UNCOMTRADE (2014, processed) . . . . 22 Non-oil and gas Intra-ASEAN Trade, 1962-2012 Source: UNCOMTRADE (2014,

processed) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 Average Changes of NRCA ASEAN inASEAN Market, 1989-2012 Source: Source:

Own Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

Impact Analysis of Normalized Revealed ComparativeAdvantageon ASEAN’s Non-Oil and Gas Export Pattern Using

Gravity-Model Approach

Umar Fakhrudin1, Fithra Faisal Hastiadi21 Center for Foreign Trade Policy, Ministry of Trade, Republic of Indonesia

2 School of Economics, University of IndonesiaEmail: [email protected]; [email protected]

Abstract

The development of global economic challenges hasforced ASEAN countries to further deepen itseconomic integration within the ASEAN Economic Cooperation (AEC) and to incorporate sev-eral ASEAN Plus agreements into Regional Comprehensive Economic Partnership (RCEP). Underthiscircumstance, the ASEAN members need to distinguish how the difference in comparative ad-vantage of each export commodity affects and influences the pattern of ASEAN’s non-oil exports.This study attempts to identify the impact of comparative advantage, represented by NormalizedRevealed Comparative Advantage (NRCA) index, on the non-oil export pattern of the ASEANcountries using the augmented gravity model as its research method. The results indicate thatcomparative advantage has a positive influence on ASEAN’s non-oil exports and that the compar-ative advantages in agricultural commodities have the biggest influence.

Key words: Comparative Advantage, Export Pattern, NRCA, ASEAN, Gravity ModelJEL Code : F11, F12,F13, F14, F15

1. Introduction

International trade occurs when two or morecountries that trade with each other enjoy ad-vantages. The expected profit is through com-parative advantage which comes in form of in-creased production efficiency where each coun-try is able to purchase products at a lowerprice. Conversely, a country is able to sell itsproducts abroad at a relatively higher price(Sarwedi, 2010).

Market openness through the establishmentof free trade agreements, in theory, is ableto provide either advantages or possible lossesto the countries involved in the cooperation.The advantage is a more efficient allocation of

natural resources in productionspecialization,which means increasing comparative advantageof countries (trade creation). On the contrary,this will further lower comparative advantagewith the presence of trade diversion (Widyas-anti, 2010).

The early establishment of the Associationof Southeast Asian Nations (ASEAN) in 1967had a major agenda to continue to develop eco-nomic cooperations, one of which was trade.ASEAN is one of the regional market exportdestinations that continues to evolve into an in-creasingly open market, especially for its mem-ber countries. ASEAN realizes that the bestway to cooperate is by opening up each econ-omy in order to create regional economic in-

1

[email protected]

[email protected]

Fakhrudin, U., Hastiadi, F.F. /Impact Analysis of Normalized ... 2

Figure 1: Intra-ASEAN Trade, 1962-2012Source: UNCOMTRADE (2014, processed)

tegration, followed up by, among others, theestablishment of the ASEAN Free Trade Area(AFTA) in 1992.

There are at least four periods in ASEANmilestones: pre-AFTA (pre-1993), post-AFTApre-Asian Crisis (1993-1998), post-AFTA post-Asian Crisis (1999 to present), and AFTAopen trade cooperation with partners outsideASEAN (ASEAN Plus). ASEAN Plus imple-mentation phase began in mid-2005 with theimplementation of ACFTA (ASEAN-People’sRepublic of China Comprehensive EconomicCooperation Agreement). Furthermore, thephase continued with the establishment of theASEAN Economic Community (AEC) in 2015and the planned merger of several ASEANPlus cooperations into Regional Comprehen-sive Economic Partnership (RCEP) (ASEAN,2011).

According to UNCOMTRADE data (2014),the SITC trade data nomenclature version 2, aspresented in Figure 1, recorded a total intra-ASEAN trade in 1967 (when ASEAN onlycomprised five founding states) of US$1.63 bil-lion or 16.66% of total ASEAN trade to theworld which later increased to US$ 26.99 bil-lion or 18.71% of the total ASEAN trade to theworld in 1983-one year before Brunei Darus-salam joined ASEAN. This means that for aperiod of 16 years an increase in the averageintra-ASEAN trade reaches 20.18% per year

with the highest increase in 1979 (45.98%) anda slight decrease (-1.84%) in 1975.

In 1993 when ASEAN Free Trade Area(AFTA) began to take into effect, the mar-ket share of intra-ASEAN trade was still un-able to hit 25% until 2012, although the valueof intra-ASEAN trade had reached US$584.67billion. In addition, the increase in trade sincethe establishment of ASEAN (1967) until theyear 2012 was 14.35% per year. This indi-cates 12.36% increase per year since AFTA wasimplemented (1993-2012) or only 11.91% peryear since the establishment of ASEAN Plusfor the first time (2005-2012). Likewise, theshare of intra-ASEAN exports and imports inthe period of 1967-2012 jumped from 13.66%to 25.85% (for exports) and from 15.01% to22.48% (for imports).

In addition, total intra-ASEAN non-oil tradein 1967 as shown in Figure 2 was recordedat US$1.48 billion. This value is equivalent to16.18% of the total value of non-oil trade ofASEAN to the world. Then, the value climbedto US$14.47 billion, but its share dropped to14.58% in 1983. In general, ASEAN trade overa period of 16 years since the formation ofASEAN, posts an increase average of 15.37%per year. During the period, the highest in-crease occurred in 1973 (53.56%), while thelowest decline (-7.60%) took place in 1975.

Meanwhile, intra-ASEAN non-oil trade also


Figure 2: Non-oil and gas Intra-ASEAN Trade, 1962-2012Source: UNCOMTRADE (2014, processed)

has not been able to touch 25% until 2012 eventhough the trade value has reached US$428.75billion in 2012. This means that the increase innon-oil trade since the establishment of intra-ASEAN ASEAN (1967) until the year 20121

was 14.09% per year. In other words, there is anincrease in intra-ASEAN trade by 11.56% peryear since AFTA is implemented (1993-2012)or only 9.57% per year since the formation ofASEAN Plus for the first time (2005-2012).

Similarly, the share of intra-ASEAN exportsand imports in the period 1967-2012 soaredfrom 20.78% to 23.15% (for exports) and from12.59% to 20.51% (for imports). Neverthe-less, the market share increase remained atthe range of 25%. The share of intra-ASEANtrade is still relatively small compared tothatof intra-European Union trade. The EuropeanUnion (EU) is a form of economic integra-tion which is considered the most advanced inthe world today. The share of intra-EU tradehas reached an average in the range of 66%,both for the whole commodity and for non-oil commodities, during the period 2003-2012(UNComtrade, 2014). Lowering intra regionalASEAN trade is in line with what was men-tioned by Elliott and Ikemoto (2004) who ar-

1The intra-trade assumption here only reflects tradeamong ASEAN member states without including tradewith members of ASEAN Plus and non-ASEAN (China,Japan, Korea, India, Australia and New Zealand).

gued that the orientation of foreign trade ofmember countries of ASEAN (both exportsand imports) generally still tended to be ori-ented from and to non-ASEAN nations, suchas China, Japan, United States and the coun-tries of Western Europe.

As global challenges upsurge, which resultfrom the impact of the Asian crisis in the late90s and the global financial crisis in 2008, aswell as from the rapidly growing China andIndia economies, ASEAN is spurred to fur-ther deepen itseconomic integration. In 2003,the idea of the ASEAN Economic Commu-nity (AEC) emerged. AEC has a goal to makeASEAN as a single market and productionbase, competitive economic region, a regionwith equitable economic development and in-tegration with the global economy. AEC cov-ers liberalization and facilitation of trade ingoods, trade in services and investments, in-cluding protection and promotion of invest-ment; narrowing differences in development;and openness of trained manpower and capi-tal flows (Chia, 2013). In addition to the es-tablishment of the AEC, ASEAN has also de-veloped FTA cooperation with several trad-ing partners within ASEAN Plus2 and plansto merge numerous ASEAN Plus cooperationsinto Regional Comprehensive Economic Part-nership (RCEP) (ASEAN, 2011).

As previously discussed, amid deepening eco-


nomic integration, the share of intra-ASEANtrade stays at the range of 25% although itonly covers non-oil exports. The concern is thatthe ASEAN member countries need to knowthe extent of non-oil exports pattern in theASEAN market. In this case, they need to findout which commodities that have comparativeadvantage, especially those having the largestand most significant influence on the patternof ASEAN’s non-oil exports.

The knowledge clearly becomes necessary toavoid ASEAN member countries only beingtarget market when ASEAN actually opens upAEC market within RCEP concept. Accordingto Chia (2013), the spirit of establishment ofAEC itself is to make ASEAN a single marketand production base, a competitive economicregion, and a region with equitable economicdevelopment and global economy integration.Coverage of AEC includes liberalization andfacilitation of trade in goods, services and in-vestments, together with protection and pro-motion of investment; narrowing developmentdifferences; and openness of trained manpowerand capital flows.

With regards to trade patterns3 (in thiscase is a pattern of non-oil exports), studiesthat employ gravity models are already quiteabundant, particularly those related to tradeflows (Kepaptsoglou et.al., 2010). Neverthe-less, studies applying approach of commoditiestrade pattern remain limited.

2ASEAN Plus is ASEAN-People’s Republic ofChina Comprehensive Economic Cooperation Agree-ment (ACFTA), implemented in mid-2005; ASEAN-Korea Comprehensive Economic Cooperation Agree-ment (AKFTA), implemented in mid-2007; ASEAN-Japan Comprehensive Economic Partnership (AJCEP),implemented in late 2008; ASEAN-India Comprehen-sive Economic Cooperation Agreement (ASEAN-IndiaCECA) and ASEAN-Australia and New Zealand FreeTrade Agreement (ASEAN-ANZ FTA) which was im-plemented in early 2010.

3Trade pattern is anything that is traded by a coun-try (be it goods or services), with whom, and to whichdirection (export or import) the trade is conducted.Trade pattern itself is one of the main goals of trade

Studies related to ASEAN trade flows are,among others, the ones conducted by Elliottand Ikemoto (2004) and Hapsari and Mangun-song (2006). Both studies utilize export pat-tern approach in viewing trade flows. Both re-searches have taken into account a variablewhich is a derivative product of comparativeadvantage, namely complementarity and simi-larity indices. Both of these variables are con-sidered able to capture a comparative advan-tage in terms of differences in endowment fac-tor and to explain product differentiation andinequality in product demand. Yet, the com-parative advantage is in aggregate, whilst thedifference in either each commodity or eachgroup of commodities cannot be shown bythese two variables.

Yue and Hua (2002) in their study whichaims to identify the effect of comparative ad-vantage on export patterns do not use the grav-ity model, but the model of export supply in-stead. Such research is able to explain that thecomparative advantages affect export perfor-mance with index variable of Revealed Com-parative Advantage (RCA) as an approach.Considering also that RCA has several draw-backs, the authors also refers to Yu et al(2009) who modifies RCA index equation intoNormalized Revealed Comparative Advantages(NRCA) index which can cover the shortcom-ings of RCA index.

Therefore, to answer the two questions re-lated to the pattern of non-oil export ofASEAN member countries in the ASEAN mar-ket, this study will use gravity model by addingindex NRCA as one of the variables. It is thenexpected that NRCA has impact on the patternof ASEAN non-oil exports. Moreover, this canbe one of the considerations for ASEAN mem-ber countries in determining trade policy, par-ticularly in terms of which commodities thatneed to be focused on to increase non-oil ex-

theory, particularly in terms of which goods or servicesthat will be exported or imported by a country (Dear-dorff, 2010).


ports.In general, this study aims to identify the

determinants of ASEAN’s non-oil exports inthe ASEAN market. Furthermore, this studyseeks to determine the effect of comparativeadvantage, that is, NRCA, on the pattern ofASEAN’s non-oil exports during the period1989-2012.

Utilizing panel data analysis, this researchfinding suggests that comparative advantagehas positive effect on non-oil exports and com-parative advantage of natural resources-basedcommodities has the greatest impact.

This paper consists of: Part 2 describing theconceptual framework of comparative advan-tage and trade patterns through gravity modelapproach; Section 3 outlining a number of rel-evant empirical researches; Section 4 introduc-ing the model, variables, data, and estimationmethods in this study; Section 5 discussing theresults of the estimates; and Section 6 Conclu-sion.

2. Conceptual Framework

2.1. Comparative Advantage, Pattern Exportand Normalized Revealed ComparativeAdvantages (NRCA)

Comparative advantage or Ricardian Modelis a classical economic theory which arguesthat a country to another are interdependentand can mutually benefit each other, and oneof which is economic benefit. The two coun-tries can conduct exchange transactions in ac-cordance with their comparative advantage, inthis case is the comparison of number of laborsused to produce one unit of product (Ricardo,1817; Edward and Schoer, 2002). Ricardianmodels shows that countries with a higher com-parative advantage in a product will tend tofocus its production factors on producing andincreasing the amount of production and subse-quently exporting to countries that have lowercomparative advantage for that product. In op-position, the country will tend to reduce oreven not to produce products that have lower

comparative advantage and subsequently willimport such product from countries that havehigher comparative advantage (Appleyard etal, 2006).

In further development, the difference in pro-duction endowment factor is considered havingan impact on international trade (Heckscher,1919; Ohlin, 1933). Based on Heckscher - Ohlintheory, Edward and Schoer (2002) suggest thatexports are made by countries with abun-dant production endowment factor as theyhave lower opportunity cost compared to othercountries. Therefore, differences in productionendowment and opportunity cost are the coreof comparative advantages, in addition to dif-ferences in technological development (Salva-tore, 2002 and Costinot, 2009).

In line with the above explanation, it can besaid that comparative advantage is consideredbeing able to show export performancepattern.This is consistent with the definition of tradepattern, i.e.anything traded by a country (ei-ther goods or services), with whom trade trans-actions are carried out, and to which direc-tion (export or import). Trade pattern itselfis one of the main objectives of trade theory,especially in terms of which goods or servicesthat will be exported or imported by a country(Deardorff, 2010).

Index of Revealed Comparative Advantage(RCA) or Balassa Revealed Comparative Ad-vantage (BRCA) is one tool to measure thelevel of comparative advantage in empiricalstudies (Balassa, 1965). Although useful in ex-amining whether a country has a comparativeadvantage in specific products, it has some lim-itations (Hillman, 1980; Bowen, 1983, 1985,1986; Ballance et al., 1985, 1986; Deardorff,1994; and Hoen and Oosterhaven 2006). Mostresearches apply BRCA index only to identifythe relative ranking of the comparative advan-tages of a country for different commoditieseven though generally it remains problematicin its relative order (Yeats, 1985).

Some RCA alternatives have been developed


to overcome the weaknesses of BRCA, amongothers, BRCA log (Vollrath, 1991), Symmet-rical Revealed Comparative Advantage (srca)(Laursen, 1998), Weighted Revealed Compara-tive Advantage (WRA) (Proudman and Red-ding, 1998), Additive Revealed ComparativeAdvantage (ARCA) (Hoen and Oosterhaven2006). Although the indices develop some as-pects of BRCA, but none of those indices couldbe the one that can be generally applied tocomparison between spaces (either commodi-ties, state, or region) and time.

To answer the problem of BRCA limitationsand some of the alternative RCA indices, Yuet al (2009) has developed BRCA index into aNormalized Revealed Comparative Advantage(NRCA). NRCA posseses attributes that canindicate the rank and comparable in compar-ative advantage across commodities, countries,and time spans and it is expected to show acountry’s trade pattern, thereby enabling iden-tification of which types of commodities thathave good potential in a market and at a spe-cific time. NRCA index value for each commod-ity from each country as a whole is zero or neu-tral. This is in line with the assumption thatno country has a comparative advantage for allcommodities.

The NRCA equation is as follows:

NRCAik =

∆Xik

X=Xik

X− XkXi

XX(1)

where,NRCAi

k is the difference in the comparativeadvantages of country i for product k in aspecific market;Xik is commodity export k of country i to aspecific market;Xi is the total exports of country i to a specificmarket;Xk is the world’s commodity exports k tocertain markets; andX is the world export to a specific market.

NRCA value span ranges from neutral value(0) is -0.25<NRCA<0 and 0<NRCA<0.25.

This signifies that a commodity has actual ex-port value lower than the comparative advan-tage in its neutral value if the NRCA value issmaller than 0. On the contrary, a commodityhas actual export value greater than the com-parative advantage in its neutral value if theNRCA value is smaller than 0.

The symmetrical range of values causes thetotal number of NRCA for all commodities ofa country or trading market to become zero orneutral. Consequently, in a country or market,if one of the commodities from a country expe-riences increase in comparative advantage, thesame commodity from other countries will ex-perience a decline in comparative advantage.NCRA can be used to determine the level ofspecialization of a country. In this condition,NRCA can be used to look for the comparativeadvantage among commodities in a country us-ing the following equation:

NCRA can be used to determine the level ofspecialization of a country. In this condition,NRCA can be used to look for the comparativeadvantage among commodities in a country us-ing the following equation:

∆NRCAikl =

Xik

X− ∆Xil

X

=

[(Xik

X− XkXi

X*X

)−(Xil

X− XlXi

X*X

)](2)

where,∆NRCAi

k is the difference in the comparativeadvantages of commodity k with commodity lin Country i;Xik is commodity export k of country i to aspecific market;Xil is commodity export l of country i to aspecific market;Xi is the total exports of country i;Xk is commodity export k of country i;Xl is commodity export l of country i; andX is the world export to a specific market.


When comparing comparative advantage ofcommodity with partner countries within aparticular market, be it country, regional orglobal, the NRCA equation becomes:

∆NRCAi−jk =

∆Xik

X−

∆Xjk

X

=

[(Xik

X− XkXi

X*X

)−(Xjk

X− XkXj

X*X

)](3)

where,

∆NRCA(i−j)k is the difference in the compar-

ative advantages of commodity in country iwith commodity j for produk k in a specificmarket;Xik is a commodity export k of country i to aspecific market;Xjk is a commodity export k of country j to aspecific market;Xi is total exports of country i to a specificmarket;Xj is total exports of country j to a specificmarket;Xk is a world’s commodity export k to aspecific market; andX is world’s exports to a specific market.

In the interim, the equation of comparativeadvantage comparison of commodity in timerange changes becomes:

∆NRCAik,t+1 =

∆Xik,t+1

Xt+1−

∆Xik,t

Xt

=

(Xik,t+1

Xt+1−Xk,t+1Xi,t+1

Xt+1*Xt+1

)−(Xik,t

Xt−Xk,tXi,t

Xt*Xt

)(4)

where,∆NRCAi

(k,t+1) is the difference in the com-parative advantages of commodity in countryi for commodity k between a time range oft+ 1 and t;X(ik,t+1) is a commodity export k of country i

at time t+ 1;X(ik,t) is a commodity export k of country i attime t;X(i,t+1) is total exports of country i at timet+ 1;X(i,t) is total exports of country i at time t;X(k,t+1) isworld’s commodity export k at timet+ 1;X(k,t) isworld’s commodity export k at time t;X(t+1) is world’s exports at time t+ 1; andXt is world’s exports at time t.

According to Yu et al (2009), NRCA is con-sidered consistent in measuring the compara-tive advantage that is symmetrical, additive forthe range of countries and commodities, andcomparable across countries, commodities andtime. Therefore, NRCA can be used in timeseries analysis as well as in comparative analy-sis of comparative advantage among countrieswith panel data analysis.

2.2. Gravity Model in International Trade

The use of the gravity model approach in in-ternational trade flows is developed separatelyby Tinbergen (1962) and later Pyhnen (1963).Linneman (1966) adds the variable and movesfurther by establishing a theoretical justifica-tion in the form of Walrasian general equilib-rium system.

Although initially there was no theoreticalsupport for gravity model, since the late 1970s,there were numerous developments that hadfilled this gap. Anderson (1979) conducted thefirst formal attempt to derive the equation ofgravity based on Cobb-Douglas expendituresystem, assuming identical homothetic prefer-ences between regions and products were differ-entiated by region of origin. Bergstrand (1985,1989) also explored the theoretical support ofbilateral trade in some papers, where the gravi-tational equationwas associated with the modelof monopolistic competition and product dif-ferentiation (not intercountry-based). Helpman(1987) used the framework of product differen-


tiation with increasing returns to scale (IRS)to justify the gravity model.

Deardorff (1995) derived the gravitationalequation from two extreme cases based onHeckscher-Ohlin (H-O) model. The first caseis in the form of free trade with homogeneousproducts where both producers and consumersdo not differ in choosing one of the manytrading partners. The second case is wherethe country produces different goods and eachhas different Cobb-Douglasor constant elastic-ity of substitution (CES) preference. Deardorffdrew the conclusion that the gravity modelwas consistent with the existing trading model.Eaton and Kortum (1997) as cited by Rahman(2003) derived the gravity equation in a Ri-cardian framework. Evenett and Keller (1998)argued that the gravitational equations couldbe derived from the H-O model with both per-fect and imperfect product specialization con-ditions.

In the past ten years, development of grav-ity model indicates that out of 75 studies usingthis model, most are related to the impact ofFTA trade policy, particularly regional FTA,the rest is related to the flow of trade in gen-eral (Kepaptsoglou, Karlaftis, and Tsamboulas2010). Similarly, in the new determinant in in-ternational trade, there are several variablesthat have good potential to explain the grav-ity model of trade, namely the level of devel-opment, trade policy, affinity of language andcolony, geography, relative population density,common currency, and membership in a re-gional trade agreement (Yamarik and Ghosh,2006).

3. Empirical Research Review

Empirical studies using gravity model interms of economic integration (FTA) againstflows and trade patterns have been commonlyconducted until the last decade. The use ofgravity model is empirical model other thanthe Computional General Equilibrium (CGE)model. With regards to the use ofgravity model

in examining the pattern of trade in ASEAN,Elliott and Ikemoto (2000) as well as Hapsariand Mangunsong (2006) both apply the aug-mented gravity model to investigate the de-terminants of trade flows among AFTA mem-ber states. Both studies identify the influ-ence of AFTA formation on intra-regional andextra-regional trades by comparing the tradepatterns among countries involved in AFTAscheme and non-AFTA countries.

Gravity model on both studies utilize somebasic variables by adding some other controlvariables. These basic variables have some con-nections with trade, consistent with the re-sults of empirical studies with other gravitymodel. In their studies this time, Elliott andIkemoto (2000) as well as Hapsari and Mangun-song (2006) add two variables, namely Com-plementarity Index (COM) and Similarity In-dex (SIM). COM and SIM are added into themodel because thoughdifferences in economiclevel is able to describe the differences in en-dowmentfactor, the variable is explicitly unableto explain product differentiation and demandinequality (Deardorff, 1984 in Elliot and Ikke-moto, 2004).

Ng and Yeats (2003) suggest that COM canseparate the effect of the commodity compo-sition from other factors that encourage tradeflows. Besides, it represents the alignment be-tween export and import structures in a bi-lateral trade with the assumption that tradedcommodity reflects innate ability factor. It isalso said that Complemetarity is one of theproducts of comparative advantage if assum-ing that the pattern of export and import de-scribes resource endowment and demonstratesthe existence of economic resources and whichproduction structure that complements (Drys-dale, 1967). At the interim, SIM, according toNg and Yeats (2003) provides information onwhether the export structure of trading be-tween two countries has a common key exportproduct or not. One of the shortcomings ofCOM and SIM is its more aggregate nature.


In this regard, both indices cannot capture thecomparative advantage which later can portraythe difference inendowment factor from differ-ent commodities, both within a country or inits comparison with other countries.

In their study, Elliott and Ikemoto (2004)has a specific purpose, in addition to thosedescribed above, to determine whether an in-crease occurs in trade between AFTA members(intratrade) or in trade with partners outsideAFTA. Thus, the effect of increased trade willbe utilized as much by members of AFTA. Fur-thermore, it will become part of trade policy-making and negotiations. Specification of esti-mates made in this study is by making a com-parison between the time before AFTA and thetime afterwards. The division of the period forcomparison purposes comprises time periods of1982 - 1987, 1983-1987, 1988-1992, 1993-1997and 1998-1999 as well as a summary of two pe-riods, 1983-1997 and 1993-1999. Moreover, theestimations are conducted three times, namely(i) using a single intra-regional bias dummyRTA to determine the pattern of bilateral tradewith comparable RTA other than ASEAN; (ii)using two intra and extra regional bias to iden-tify trade creation and trade divertion; and(iii) is similar to stage (i), but is only intendedforASEAN.

In the meantime, Hapsari and Mangunsong(2006) employ data of 1988-2003 and in themodel (which is almost similar to that of Elliottand Ikemoto) they utilize price as an additionalcontrol variable. Next, estimation is conductedin three stages using Ordinary Least Square(pooled data), namely (i) estimating standardgravity model equation (economic size and dis-tance variables) which is added with tariff andgeographical dummy variables; (ii) estimatingthe same model as step (i) and adding tariff,Dummy RTA (ASEAN), trade creation, andtrade diversion variables; and (iii) estimatingthe same model as step (ii) by adding a vari-able index of complementarity and similarity.

The results obtained from these two studies

are slightly different, yet they give complemen-tary explanation. In general, AFTA affects in-crease in trade among ASEAN member coun-tries and it also causes slight reduction of wel-fare for non-ASEAN countries, due to the di-version of trade to ASEAN countries (Hapsariand Mangunsong, 2006).

According to Elliott and Ikemoto (2004),based on the range of AFTA implementa-tion period in the first five-year period, thereis no significant improvement in trade flowsin ASEAN, due to the limitations of institu-tional progress of each ASEAN government.Moreover, this is attributable to the effectsof enlargement of share of non-ASEAN ex-porters, such as China and Latin Americancountries. The Asian crisis in the late 90s ac-tuallty spurred the increase in trade amongASEAN member countries and after the cri-sis passed, trade situation returned to becomemore outward looking.

The research related to the use of Com-parative Advantage, in this case the RCA in-dex, as a determinant of exports or exportperformance pattern was conducted by Yueand Hua (2002). This study aims to determinewhether comparative advantage, which is iden-tified through the RCA index, affects the de-velopment of China’s exports. Estimations aremade by two stage least square (2SLS) regres-sion to find out the export determinant fromboth the demand and supply sides as well aspanel data regression to examine the relation-ship of comparative advantage with export per-formance.

The assumption used in this study is Chinaas a price taker. The comparative advan-tage index used is RCA for chemical prod-ucts (RCA5), raw materials RCA, and finishedgoods manufacturing industry (RCA 68), andmachinery and transportation equipment RCA(RCA7). The data used cover the period 1980 -2000 for export data, in accordance with SITC1-digit level, that is SITC 0-8 and SITC level 3.The results and conclusions of the study signify


that RCA index is able to explain the patternof China’s export appterwhich grows accordingto their comparative advantage.

4. Hypothesis

As presented in the previous section, thisstudy aims to find out the determinant of non-oil exports of ASEAN countries, especially toidentify the effect of comparative advantage onthe pattern of non-oil exports of the ASEANmember states.

The hypothesis to be tested in this studyis that the greater the difference between thecomparative advantage of one of the membercountries of ASEAN with its trading partnercountry in the ASEAN market (∆NRCA), thebigger the non-oil exports. Comparative advan-tage will have a significant positive relation-ship with non-oil exports. As for this study, thecomparative advantage used is based on com-modity groups.It is then expected that the sizeof influence of comparative advantage of eachcommodity on non-oil exports from ASEANmember countries is obtained.

5. Methodology

This study refers to and modifies the modelused by Elliott and Ikemoto (2004) and Hapsariand Mangunsong (2006) which modifies aug-mented gravity model from basic model varia-tions of Tinbergen (1962) and Pyhnen (1963).Additionally, this research model also refersto Yue and Hua (2002) and Yu et al (2009).Therefore, the modification in this study to themodels used in the two previous studies men-tioned above is replacing the complementarityindex variable with Normalized Revealed Com-parative Advantages (NRCA) variable index,which can present the export pattern basedon of comparative advantage comparison fromcommodity groups. The equation model in this

study is

log(Xijt) = α0 + α1logPGDPit + α2logPGDPjt

+ α3log(POPit) + α4log(POPjt)

+ α5log(TCijt) + α6AFTAjt

+ α7ASEANPLUSit + α8CRISISjt

+ α9k∆NRCAi−jkt + µijt + εijt

(5)

where each variable can be described as follows:

1. Xijt is a non-oil exports from country i tocountry j at time t. Exports aredependentvariable as an approach to trade amongASEAN member countries as well as be-tween an ASEAN member country and anon-ASEAN country. Numerous indepen-dent variables are used as determinant ofthose exports. Use of exports as depen-dent variable is initial variable from grav-ity models (Tinbergen, 1962 and Pyhnen,1963).

2. PGDPit and PGDPjt is the Gross Domes-tic Product (GDP) per capita of exporter(i) and importer (j). This variable is usedas an indication of purchasing power ofboth exporter and importer. Besides, GDPper capita is a proxy of capital endowmentratio (Bersgstrand, 1985; Shon, 2005). TheGDP per capita is also considered a proxyto determine the influence endowment fac-tor on fragmentation (Kimura et. al, 2007)as well as the proxy of infrastructure en-dowment and the skills of the workforce inthe trading country (Trkcan, 2011). GDPper capita is also an approach to economicdevelopment level that has positive im-pact on international trade (Frankel, et.al,1995; Elliott and Ikemoto, 2004).

3. POPi dan POPjt adalah is population ex-porter country i and population of thetrading partners country (importer) j, re-spectively. The population itself can be


used as a proxy for the magnitude of de-mand or market. Population can also sig-nify presence of import substitution ef-fect where domestic production receivesincentives on the expanding market. It alsoshows the effect of absorption where, de-spite rising population and increasing pro-duction, the produced goods are more ab-sorbed in the domestic market than for ex-port (Razzaghi et al., 2012). On the otherhand, population can depict the economicdirection of a country. If the direction isoutward-oriented, the import demand willincrease with the increase in population.Conversely, if the orientation direction isinward-oriented, the demand for importsdeclinesalong with the increase of popula-tion (Tayyebi, 2005).

4. TCijt is the index of the cost of transporta-tion from country i to country j, wherethere is the assumption that the amountof bilateral trade increases with the sizeof their economies and decreases with in-creasing transportation costs due to dif-ferences in distance (Tinbergen, 1962 andPylnen, 1963). Index Trade Cost is thetrading cost which is calculated using In-verse Gravity Framework based on a re-search by Novy (2009). TC is an estimateof the cost of trading based on the cost ofbilateral trade and gross national output.

5. The dummy variables in this researchmodel are AFTA, ASEANPLUS and CRI-SIS. AFTA is worth 1 if the importer is acountry belonging to AFTA, while 0 if theimporter is a non-AFTA. ASEANPLUSis worth 1 if the importer is a countrythat joins in the scheme of ASEAN Plus,while 0 if the importer is a non-ASEANPluscountry. In the interim, CRISIS is in-tended to show the time of the economiccrisis. Dummy CRISIS’ value is 1 if theeconomy is in crisis, that is the Asian cri-sis in 1997-1999 and the global financial

crisis in 2007-2009, and is valued 0 if thereis no crisis.

6. ∆NRCAi−jkt is a variable used to reflect

the different comparative advantages ofa country commodity i with its trad-ing partner (j) in the ASEAN market.∆NRCAi−j

kt is calculated based on equa-tion 3. Given the relatively small ∆NRCAvalue, ∆NRCA is multiplied by 100. Inline with the findings by Yue and Hua(2002) that comparative advantage is con-sidered capable of showing patterns of ex-ports, it is expected that the higher the∆NRCAi−j

kt , the greater the impact on theamount of exports from i to j.

Sources of data used in the study are in ac-cordance with the variables in the empiricalmodel described in Table 1.

The trade data used in the calculation ofNRCA is based on the data of three-digit SITCversion 3 (SITC3) which are then grouped ac-cording to UNCTAD (2013). This is a group-ing of commodities based on the origin of theraw material and the level of technology andskills of the workforce of an industry. The useof SITC data at three-digit level is becauseat this level the characters of commodities bytechnology similarity and production factorscan be seen (Greenway and Milner, 1986; andMenon, 1996). The commodity grouping in thisresearch can be seen in Table 2.

As for the purpose of analysis, the re-searchers create three alternative combinationsof commodities grouping by Commodity Groupcodes, namely:

• Combination I: 1A, 2, 3, 4, and 5

• Combination II: 2, 3, 41A, and 5; and

• Combination III: 2, 3A, 3B, 3C1, 3C2,3C3, 3D1, 3D2, 3D3, 41A, and 5.

Combination I and II are similar to oneanother. Combination II integrates Coal andDerivatives (Code 1A) with Mining Goods


Table 1: Descriptive analysis of variables

Variables Descriptions Relationship Sources

Log(Xijt)Log of bilateral trade between countries i UNComtrade,and country j in year t, in million US$ accessed throughaccording to SITC classification version 3 WITS

Log(PGDPit)Log of constant GDP per capita in 2005 of +

WDI, World Bankcountry i in year t, GDP is in US$

Log(PGDPjt)Log of constant GDP per capita in 2005 of +

WDI, World Bankcountry j in year t, GDP is in US$

Log(POPit) Log of POP of country i in year t - WDI, World Bank

Log(POPjt) Log of POP of country j in year t + WDI, World Bank

Log(TCijt)Trade cost index of country i and -

ESCAPcountry j in year t

AFTAijt Cooperation dummy of AFTA in year t +/-

(Code 4) into Mining Goods including Coal andDerivatives (Code 41A). Combination III aresimilar to combination II but the commoditiesin combination 3 are disaggregated into numer-ous groups of industry or manufacturing com-modities based on the different skill levels ofthe workforce and technology.

Descriptive analysis of the calculation resultsof NRCA will be addressed using combinationIII coupled with other energy commodities (1B,1C, and 1D). The descriptive analysis focusesnot only on non-oil commodities alone, but willbe more thorough. NRCA calculation is basedon equation 1. In addition, to identify changesin comparative advantage over time, the fol-lowing periods are used: 1989 (the period priorto AFTA), 1996 (the period after AFTA andbefore the 1998 crisis), 2004 (after the Asiancrisis, before the establishment of the ASEANPlus, and before the global financial crisis) and2012 (after the establishment of the ASEANPlus and post-global financial crisis).

For commodity based on Combinatiob I;(iv) spesifikasi In the meantime, the gravitymodel estimation is carried out in five sce-narios, namely: (i) the estimates specification[1] of basicgravity model; (ii) model specifi-cation [2] of basic gravity model estimationis added with dummy AFTA, ASEANPLUS

and CRISIS; (iii) model specification [3] showsthe gravity model [2] which is added with

∆NRCA(i−j)k variables for commodities that

are based on combination I; (iv) specification[4] is the gravity model with specification sim-ilar to specification [3], but using ∆NRCAvariables of commodities based on combina-tion I; and (v) model specification [5] is gravitymodel specification that is similar to specifica-tion [4], but with ∆NRCA that is based oncombination III.

6. Result and Analysis

6.1. Comparative Advantage of ASEAN TradeCommodities in ASEAN Market Based onNRCA

General overview of the average change ofNRCA ASEAN as a single state entity in theASEAN market during the period 1989-2012is illustrated in Figure 3. The average NRCAvalue is the average of the NRCA commodi-ties that have NRCA value above zero. Fig-ure 3 illustrates that the trend of comparativeadvantageaverage change of ASEAN tends tofall by 3.73% during 1989-2012. NRCA aver-age value of all commodities, namely energycommodities other than coal (consisting of oil


Table 2: Grouping of SITC3 Commodities at Level 3 Digit Based on Production Factors and Levelof Technology

Commodity Group Code Description

1. Energy Commodity

1A Coal and Coal-Based Products (except gas)

1BPetroleum and petroleum-based products (except gas) includinglubricants and asphalt (Tar)

1C Gas, including Liquefied Natural Gas, Gas from Oil and Coal

1D Electric Power

2. Agricultural Commodities

2Agriculture, Animal Husbandry, Plantation and OtherAgricultural Products

3. Industrial/Manufacturing Products Commodities

3A Labour-intensive and resource-intensive manufacture produscts

3B Low-skill and technology-intensive manufacture products

3C1Electronics including medium-skill and technology-intensivemanufacture Products

3C2Parts of Electronic and Electrical Products that include medium-skilland technology-intensive manufacture Products

3C3Non-electronic and non-electrical products that include medium-skilland technology-intensive manufacture Products

3D1Electronics that include high-skill and technology-intensivemanufacture Products

3D2Parts of Electronic and Electrical Products that include high-skilland technology-intensive manufacture Products

3D3Non-electronic and non-electrical products that include high-skilland technology-intensive manufacture Products

4. Mining Commodities

4 Minerals (other than coal), metals, and other minerals

5. Other Commodities

5 Other commodities(Unspecified)


and gas and electricity) and non-oil commodi-ties was 0.775 in 1989 and then climbed to0.788 in 1996. However, the NRCA average in-crease contracted to 0.410 in 2004 then slightlyjumped to 0.467 in 2012. Of these changes, itappears that in the period before the Asian eco-nomic crisis in 1997, the comparative advan-tage of ASEAN in ASEAN market was twiceas much as that of post-crisis.

For the meantime, the trend ofASEAN’saverage non-oil commodity com-parative advantage indicates a sharper decline(-4.69%) when compared to that of the entirecommodities (-3.73%) during 1989-2012.The decline in ASEAN’snon-oil commoditycomparative advantage is the result fromchanges in the pattern that occurred beforethe global financial crisis. The changes aremainly in the increase in comparative advan-tage of energy commodities which indicatedan annual increase of 20.03% over the period2007-2012. The rise of energy commodities’advantage is able to slightly drive increase incomparative advantage of all commodities,but also compensates in a form of pressure tothe decline in non-oil commodities’ advantage.This also indicates that intra-ASEAN trade inover the last decade is more focused on energycommodities, especially oil and gas.

To find out how much NRCA value and rank-ing of each commodity traded by ASEAN coun-tries as a single entity in the regional mar-ket in a given year, refer to Table 3. This ta-ble describes that, in general, in the ASEANmarket, petroleum commodities and their pro-cessed products (1B) have the highest compar-ative advantage compared to that of other en-ergy commodities (gas (1C) and electric power(1D)) and of non-oil commodities. The calcu-lations indicate that petroleum and its prod-ucts always ranks first or second. When ex-plored further, it is found that ASEAN mem-ber that has the highest comparative advantagefor petroleum and its products in the ASEANmarket is Singapore. The comparative advan-

tages of Singapore petroleum also becomes amajor contributor to the high NRCA value forASEAN petroleum. This finding becomes evi-dence that a country which possesses naturalresources does not always have the advantagein the natural resources products, whileanon-producing country can actually become a cen-tral and control the commodity trade. Addi-tionally, logistics readiness and adequate tradeservices are the driving factors that boost thecomparative advantage. Singapore in this casetook fifth position in global Logistics Perfor-mance Index (LPI 2014) and ranked first inease of doing business (Doing Business 2015,2014).

ASEAN commodities including non-oil showdiverse changes in their comparative advan-tage. Commodities that belong to the low-skill and technology-intensive manufactureproducts (3B), Non-Electronic and Electri-cal Products which belong to medium-skilland technology-intensive manufacture Prod-ucts (3C3), Non-Electronic and ElectricalProducts which are classified as high-skilland technology-intensive manufacture prod-ucts (3D3), minerals (including coal), metals,other minerals (41A), as well as other com-modities (unspecified) (5) are one of five groupsof commodities that experience discomparativeadvantage, where NRCA value during the pe-riod 1989-2012 is relatively always below zero.

On the side of the competitor, 3C3 com-modities from Japan have the highest compar-ative advantage and are always ranked first inthe ASEAN market during the period 1989-2012, although the level of excellence in 2012dropped to half compared to that of 1989. Thefall of Japan’s 3C3 commodity advantage is at-tributable to, among others, the increasing ad-vantage of European Union’s 3C3. Meanwhile,China has begun to emerge as a new competi-tor of 3C3 commodities with comparative ad-vantage in the ASEAN market since 2011. Theemergence of China as an exporter which has arelatively high comparative advantage for elec-


Figure 3: Average Changes of NRCA ASEAN inASEAN Market, 1989-2012Source: Source: Own Calculation

tronic products with medium-level technologyis one of the factors that causes the increasinglyfragmented electronics industry in the country.This is evident from the increasing number ofsuch products as laptops and mobile phoneswhich are manufactured and marketed fromChina to ASEAN. Meanwhile, the EuropeanUnion and the United States remain 3D3 com-modity exporters with the highest comparativeadvantage over the period 1989-2012.

The ASEAN commodities which are labor-intensive and resource-intensive manufactureproducts (3A) showed superiority in theASEAN market in the late 1980s to the early1990s, but the trend continued to decline sothat in the period 1995-1996 it turned into dis-comparative advantage products and remainedthat way until 2012. The condition was madepossible because of the growth of the advan-tages of the electronic commodities that belongto medium-skill and technology-intensive man-ufacture products (3C1) as well as spare partsand parts of electronic and electrical goodswhich belong to medium-skill and technology-intensive manufacture products (3C2).

Similarly, agricultural, livestock, plantationsand other agricultural commodities(2), eventhough they are in the advanced category(NRCA>0) during the years 1989-2012, theyindicate a decreasing pattern of comparativeadvantage and lean towards discomparative ad-vantage. This is probably due to the lack of

intra-ASEAN trade for this commodity. Thisis evidenced by the increasing NRCA value ofcommodities 2 from Australia, India and theUnited States.

For the meantime, the electronic commodi-ties belonging to high-skill and technology-intensive manufacture products (3D1) aswell as spare parts and parts of electronicand electrical goods belonging to high-skilland technology-intensive manufacture prod-ucts (3D2) apparently need to become prod-uct focus of ASEAN considering that two com-modity groups demonstrate a sufficiently goodcomparative advantage. The main competitorof commodities 3D1 in the ASEAN market isChina. The potential to take over the Chi-nese market gap is also quite large providedthat ASEAN market’s exporters for 3D1 whichhas the relative advantage are only China andASEAN countries. As for commodities 3D2 inthe ASEAN market, the comparative advan-tages of ASEAN outpace other competitorssuch as Korea and China.

6.2. Gravity Model Estimate with AdditionalDummy and ∆NRCA

In this study, the estimated gravity modelemploys Fixed Effect Model (FEM) to estimatepanel data. The method is chosen because it isconsidered being able to overcome the Multilat-eral Trade Resistance (MTR) with proxy. It isacceptable in theory through country-specific


Tab

le3:

NR

CA

Resu

ltsan

dR

ankin

gof

Entire

ASE

AN

Com

mod

itiesin

AS

EA

NM

arket

Com

modity

Gro

up

Code

1989

1996

2004

2012

NR

CA

PA

KD

SN

PA

KA

N5

NR

CA

5P

AK

DSN

PA

KA

NN

RC

AP

AK

DSN

PA

KA

NN

RC

AP

AK

DSN

PA

KA

N

1B

2.6

96

12

1.1

72

25

0.8

03

28

2.4

24

12

1C

0.0

49

638

0.1

05

633

0.0

98

632

0.4

37

319

1D

0.0

00

847

0.0

00

749

0.0

08

847

0.0

01

843

21.8

91

24

0.4

01

413

0.1

52

528

0.0

26

740

3A

0.6

13

311

-0.1

23

10

106

-0.4

91

11

128

-0.7

84

12

133

3B

-1.0

00

12

134

-0.7

54

12

134

-0.7

17

13

134

-1.0

83

14

136

3C

1-0

.032

10

84

-0.0

21

879

0.0

25

744

0.0

31

638

3C

2-0

.029

983

0.2

02

520

0.2

89

318

0.1

01

530

3C

3-2

.853

14

140

-2.2

00

14

140

-0.8

92

14

137

-0.8

80

13

135

3D

10.0

45

739

0.9

46

36

0.2

64

420

0.2

07

424

3D

20.0

78

431

1.9

00

12

1.6

44

13

0.5

09

217

3D

3-1

.342

13

138

-1.1

64

13

137

-0.5

56

12

132

-0.4

75

11

131

41A

0.0

52

537

-0.3

89

11

125

-0.3

40

10

122

-0.1

59

9101

5-0

.167

11

109

-0.0

74

9101

-0.2

87

9120

-0.3

56

10

119

Avera

ge

NR

CA

valu

esof

0.7

75

0.7

88

0.4

10

0.4

67

com

modities

with

NR

CA>

0

Avera

ge

NR

CA

valu

esof

non-o

iland

0.5

36

0.8

62

0.4

75

0.1

75

gas

com

modities

with

NR

CA>

0

Avera

ge

NR

CA

valu

esof

oil

1.3

73

0.6

38

0.3

03

0.9

54

and

gas

com

modities

inclu

din

gelectrica

lp

ower

with

NR

CA>

05P

AK

DSN

:In

ter-com

modity

Rankin

gin

aC

ountry

6PA

KA

N:

Inter-co

mm

odity

Rankin

gA

mong

Countries


fixed effect MTR, which is the concept thatthe bilateral trade between the partner coun-tries is not only influenced by partner countriesbut also by their interaction with other coun-tries in the global region (Adam and Cobhan2007; Feenstra, 2004; Melitz, 2007; Rose andvan Wincoop, 2001).

The FEM estimation results with numerous∆NRCA indices are based on three combina-tions previously described in the methodologysection in Table 4. There are five scenarios ofestimated gravity model. Scenario [1] exhibitsthe estimation results from basic gravity modeland signifies that all the basic variables of thegravity model has high significance and showssign of an appropriate relationship with the ex-isting theory. The specification estimation re-sults of scenario [2] is to add the basic grav-ity model with trading partner membership inAFTA and ASEAN Plus dummy as well asCRISIS dummy.

The estimation results of model scenario [3]reveals the results of the gravity model esti-

mation by incorporating ∆NRCA(i−j)k , vari-

ables, where there is ∆NRCA between the ex-porter’s NRCA and importer’s NRCA for Com-bination A. Section [4] unveils the results ofgravity model estimation with similar specifi-cations to scenario [3] using ∆NRCA variablecommodity based on combination B. Next, sce-nario [5] is based on combination C. Therefore,the main purpose of Table 4 is to demonstratethe robustness test of variable group test in themodel and the basis for selecting which spec-ification model that will become the researchfocus.

Based on Table 4, it can also be concludedthat the AFTA adan ASEANPLUS variablesare dummy that need to be included in themodel. This is reflected in the level of sig-nificance of both variables (see scenario [2],[3], [4], and [5]). In addition, if the dummyis not included as a variable in the model, itwould reduce the significance of the other vari-ables, especially ∆NRCA, the main variable

in this study. Similarly with CRISIS variable,although it is not a significant variable, it af-fects the significance of the other variables ifthey are omitted from the model. Related tothe significance of the variables in the model,CRISIS variable will not be discussed further.

The model that becomes the focus of re-search is the model [4] while model [5] is afurther modification to model [4] as an addi-tional discussion that tries to look deeper athow big the influence of ∆NRCA of manu-facturing commodity groups that are disaggre-gated according to differences in levels oflaborskills and technology.

Based on scenario [4], several findings can bedescribed in more details as follows. First, GDPper capita of ASEAN as exporter and GDPper capita of its trading partners display thesame results and are consistent with researchthat utilizes gravity model, that is GDP signif-icantly and positively affects exports (Frankel,et.al, 1995; Elliott and Ikemoto, 2004). Theestimation results indicate that the level ofASEAN economy as exporter has a greater in-fluence on the increase in ASEAN non-oil ex-ports, when compared to the economic level ofits trading partners (both ASEAN and non-ASEAN) as importers. This also means thatfrom the results of model estimation, the elas-ticity of GDP per capita increase of ASEANis greater than GDP per capita of its trad-ing partners. Based on the rule of ceterisparibus, every 1% increase of GDP of ASEANwill increase ASEAN non-oil exports by nearly2.00%. Meanwhile, the rise of GDP of ASEANtrading partner by 1% will increase ASEAN’snon-oil exports by 1.21%. It means that theexportpattern of ASEAN follows the conceptof growth leads to export, where the internalfactors of economic growth become the greaterbenchmark when compared to its export mar-ket conditions.

The GDP per capita which is a proxy of thecapital-endowment ratio), in addition to indi-cation of purchasing power (Shon, 2005), shows


increased domestic capability as an incentivefor domestic producers to at least improve pro-duction quality or quantity. This will createlarger economies of scale able to produce ex-port goods, which in turn can boost exports.On the other hand, the economic capacity ofpartner countries (importers) leads to an in-crease in demand for goods that have an im-pact on an increase in imports of goods fromoutside.

Second, the POP or population variablewhich is indicated by the population of export-ing countries (ASEAN) that demonstrates asignificant effect and is contrary to the non-oil export of ASEAN. The population of im-porters (ASEAN and non-ASEAN) gives con-trary effect, that is positive effect on the in-creasing exports of ASEAN. An increase in thepopulation of ASEAN member states by 1%causes a decline in the value of non-oil exportsof ASEAN by 0.48% (ceteris paribus) and anincrease in the population of importers by 1%causes an increase in ASEAN’s non-oil exportsby 0.42%.

The influence of population on exports is inline with the studies by Filippini and Molini,(2003) as well as Razzaghi, et al. (2012). Bothstudies suggest that the negative effects of thepopulation, particularly in exporting countries,on exports indicate an incentive for domesticproducts as a result of an increase in the num-ber of markets in the country, which can bereferred to as import substitution effect. Thereis also the effect of absorption where domes-tic production that increases with the numberof population will be absorbed in the domes-tic market first before being exported abroad.Meanwhile, the increase in the population ofthe importing country will increase the mar-ket size. Market growth in trading partners willeventually become one of the factors for the in-crease in demand in the export market.

Third, differences in the value of non-oil ex-ports that is the result of Free Trade Agree-ment (FTA) in ASEAN also show significant

gains. If the ASEAN trading partners are coun-tries belonging to the AFTA scheme, then thereare differences in the value of non-oil exportsamounting to 18.23%, lower than that of thetrading partners not included in the AFTAscheme. This condition matches with the find-ings of Elliott and Ikemoto (2004) who arguethat the member countries of ASEAN are morelikely to be outward looking, so that more trade(in this case export) are conducted with trad-ing partners outside ASEAN. It is also rein-forced in the findings in this study, that thepresence of ASEAN Plus, which was marked bythe commencement of ACFTA in 2005, showsdifferences in non-oil exports by 8.85%, greaterwhen exports are made to the partner withouta trade agreement.

AFTA and ASEANPLUS also indicate thatASEAN intra-regional market is consideredless beneficial for ASEAN member countriesthemselves. Besides, the estimation of dummyAFTA and ASEAN Plus demonstrate that inorder to increase intra-regional trade, seen fromthe side of exports, ASEAN countries need toadopt policies to better utilize the ASEAN Plusscheme which will then be merged into RCEP.

Fourth, the trade cost index (Trade Cost, TCindex), as a proxy for the cost of trade, showsa negative effect on exports. A 1% increase incosts affect the decline in the value of non-oilexports by 1.28% (ceteris paribus). This is con-sistent with the statements of Tinbergen (1962)and Pylnen (1963) that exports will decline asthe cost of trade increases. Trade costs are notonly material, but also include the quality oftrade facilitation itself.

Lastly is the influence of the main vari-ables, ∆NRCA, against non-oil exports. Re-sults from this study indicate that the im-pact of ∆NRCA is positive and significanton the increase in non-oil exports to ASEANtrading partners that can be described as fol-lows: (i) increase in 1 unit of agricultural com-modities ∆NRCA will increase non-oil exportsby 10.92% (ceteris paribus); (ii) increase in 1


unit of manufacturing commodities ∆NRCAwill lead to 5.84% rise in non-oil exports;(iii) increase of 1 unit of mining commodities∆NRCA (including coal, mineral and gem-stone) will translate into a 13.55% increasein non-oil exports; and (iv) increase of 1 unitof other commodities ∆NRCA will lead toan increase in non-oil exports amountiing to8.55%. The above findings are in accordancewith the Theory of Comparative Advantage orRichardian Model stating that comparative ad-vantage will increase export. Since ∆NRCAisthe comparative advantage NRCA index of ex-porting countries which is susbtracted withNRCA index of importing countries, the in-crease in ∆NRCA index can be interpreted asan increase in the comparative advantage of theexporter or the comparative decline of impoter.Thus, exporting countries will tend to focus itsfactors of production to produce and increasethe amount of production and subsequent ex-port to countries that have lower comparativeadvantage for these products (Appleyard et al,2006). Additionally, it appears that non-oil ex-ports in ASEAN are more reliant on exportsof agricultural commodities and mining. Thismeans that natural resources products remaintop ASEAN’s non-oil exports as the impact ofchanges in comparative advantage for miningand agricultural commodities is significant.

In overall, the results of the estimation[5] above point out that the determinants ofASEAN’s non-oil exports are; (a) the level ofeconomy proxied by GDP per capita of ASEANmember countries and their trading partnerswith each impact is positive; (b) the size of themarket with a population approach in ASEANmember countries and their trading partners,where the population of ASEAN has positiveinfluence and the population of ASEAN trad-ing partners has a negative impact; (c) thecost of trade from exporters (ASEAN) to im-porting countries (ASEAN’s trading partners)which has a negative impact; (d) there is adifference in value of non-oil exports with the

presence of AFTA cooperation. ASEAN’s non-oil exports value is lower when trade is per-formed with trading partners under AFTA co-operation; (e) non-oil exports of ASEAN be-come larger when trade is done with membersof the ASEAN Plus compared to when it isconducted with non-ASEAN Plus trading part-ners; and (f) if ∆NRCA between ASEAN andits trading partners are greater, the non-oil ex-ports from ASEAN to its trading partners willincrease.

Further, the value of Adjusted R-squarespecifications [4], which amounts to 0.9823, in-dicates that the model is capable to identify98, 23% of the variation of ASEAN’s non-oilexports through the free variables inside themodel. The remaining 1.77% are a variable in-fluence outside the model. This means that thegoodness of fit of the model [4] is 98.23%. Inthe specification of this model, all variableshave a high significance, which is significant atα = 1%. All coefficients in the model are alsovisible in accordance with the model, charac-terized by a high F-statistic with probabilityof 0.000000.

What is becoming more interesting to findout further is how big the effect of changes incomparative advantage relative differences inmanufacturing commodities to non-oil exportswhen manufacturing commodities are disaggre-gated according to the skill level of its work-force and the level of technology. For this pur-pose, the results of the estimate are shown insection [5].

The estimation results of scenario [5] showsthat the biggest influence and significantchanges in ∆NRCA, especially ∆NRCA formanufacturing commodities, to the non-oil ex-ports is the ∆NRCA change for Electronic andElectrical Products belonging to medium-skilland technology-intensive manufacture prod-ucts (Commodity Code 3C3); some examplesof 3C3 are various kinds of tires, engine blocks,textile machinery, and other mid-sized indus-trial machinery parts. Hence, an increase of 1


Table 4: FEM Estimates in Different Scenarios

Regressand Model Specifications

Log(Xij) [1] [2] [3] [4] [5]

C -7.3532*** -9.101739*** -10.81938*** -6.210795*** -13.39251***(2.350912) (2.398969) (2.489238) (2.455375) (2.714252)

Log(PGDPit) 1.914438*** 1.891433*** 1.750732*** 1.993184*** 1.856974***(0.06286) (0.064094) (0.082768) (0.07588) (0.092573)

Log(PGDPjt) 1.273707*** 1.224691*** 1.255565*** 1.214314*** 1.146371***(0.042106) (0.042942) (0.048675) (0.048635) (0.056535)

Log(POPit) -0.443149*** -0.502671*** -0.354787*** -0.477653*** -0.394052***(0.113862) (0.114341) (0.114015) (0.114419) (0.128541)

Log(PGDPjt) 0.438797*** 0.641821*** 0.641756*** 0.419105*** 0.819746***(0.141069) (0.141866) (0.146203) (0.146196) (0.162279)

Log(TCijt) -1.223267*** -1.285705*** -1.268065*** -1.278837*** -1.244777***(0.059217) (0.060284) (0.059653) (0.060258) (0.062055)

AFTAJt -0.148625*** -0.159802*** -0.182333*** -0.168761***(0.03541) (0.034756) (0.03547) (0.036219)

ASEANPLUSJt 0.072423*** 0.09481*** 0.08853*** 0.093463***(0.019036) (0.020133) (0.020555) (0.022418)

CRISISJt 0,02056 0,007544 0,012648 -0,005609(0.012783) (0.012523) (0.012815) (0.013405)

∆NRCAi−j1At 0.892188***

(0.12822)

∆NRCAi−j2t 0,054658 0.109195*** 0.186367***

(0.038332) (0.038584) (0.045076)

∆NRCAi−j3t 0.066943*** 0.058419***

(0.011501) (0.011592)

∆NRCAi−j3At -0,031382

(0.02876)

∆NRCAi−j3Bt 0.080431**

(0.034523)

∆NRCAi−j3C1t 0,422641

(0.464218)

∆NRCAi−j3C2t 0,157684

(0.106042)

∆NRCAi−j3C3t 0.105891***

(0.019969)

∆NRCAi−j3D1t 0.072898*

(0.040335)

∆NRCAi−j3D2t 0.03534**

(0.01593)

∆NRCAi−j3D3t -0,00816

(0.025617)

∆NRCAi−j4t 0.066397**

(0.033602)

∆NRCAi−j41At 0.135475*** 0.176966***

(0.031982) (0.033784)

∆NRCAi−j5t 0.140072*** 0.085459*** 0.129712***

(0.034849) (0.033767) (0.036458)

R-squared 0,981038 0,981642 0,983054 0,982303 0,983236Adjusted R-sq 0,980135 0,98071 0,982104 0,98133 0,98219F-statistic 1085.423*** 1052.674*** 1034.854*** 1009.223*** 939.3961***Num of Obs 1056 1056 1056 1056 1056

Standard error in parentheses; ∗p < 0.1, ∗∗p < 0.05, ∗∗∗p < 0.01


unit of ∆NRCA3C3 leads to an increase in non-oil exports by 10.59% (ceteris paribus). Thesecond largest impact occurs at 3B commodi-ties (commodities that are classified as low-skilland technology-intensive manufacturing prod-ucts). Examples of commodities and derivativeproducts are steel pipes, steel plates, house-hold appliances, and so forth. Every increaseof 1 unit ∆NRCA3B translates into a 8.04%increase against non-oil exports of ASEAN.

Manufacturing commodities ∆NRCA thathas an influence on other non-oil exports is∆NRCA3D1 which is ∆NRCA of commoditiesclassified as Electronic commodities belongingto high-skill and technology-intensive manu-facture products. The estimation results findthat an increase of 1 unit of ∆NRCA3D1 willhave an impact on increasing non-oil exportsby 7.29%. Commodities that are grouped under3D1 are, among others; digital computers, dig-ital processing units, color television receiversand digital radios. Additionally, an increase of∆NRCA3D2 by 1 unit will impact to the in-crease in non-oil exports of ASEAN by 3.53%.3D2 commodities are spareparts and parts ofelectronic and electrical goods which belongto high-skill and technology-intensive manufac-turing products. Examples of commodities 3D2are, among others, non-cellular phone telecom-munications equipment, spareparts of radio,and television tubes spare parts.

In the meantime, ∆NRCA for manufac-ture commodities in group 3A (labor-intensiveand resource-intensive manufacture products),3C1 (electronic commodities which belong tomedium-skill and technology-intensive man-ufacturing products), 3C2 (spareparts com-modities and parts of electronic and elec-trical goods classified as medium-skill andtechnology-intensive manufacture products),and 3D3 (commodities from non-electronic andelectrical products included in high-skill andtechnology-intensive manufacture products) donot show a significant effect.

Based on the findings in model [5], several

elements need to be studied further. In thisstudy, commodities that are used in the cal-culation of ∆NRCA are still in a relativelyaggregated form. There is a possibility of ag-gregate bias from the ∆NRCA index resultswhich will then affect the results of model es-timation. It becomes necessary to disaggregatevariable component which are still aggregatein terms of commodity side. The possibility ofaggregate bias seen in the model of this studyis the large variation of significance and influ-ence of ∆NRCA against non-oil exports whenindustry/manufacture commodities (commod-ity 3) are disaggregated into numerous groupsbased on the level of workforce skills and thelevel of technology.

Moreover, it is interesting to further researchon the interaction of changes in comparativeadvantage of a commodity to changes in com-parative advantage of other commodities. Suchassumption is necessary to find empirical evi-dence of trade specialization from the conceptof comparative advantage of a commodity in acountry.

In the research model, evidence of such in-teractions has not been found, yet an indica-tion towards further proving of the theory ofcomparative advantage is made possible by thediscovery of negative marks on a few commodi-ties in the model specification [5] although theestimation results are not statistically signifi-cant. For instance, if in case there was a jus-tification variable in the model that demon-strated the interaction between commodity 3Awith 3C3, it could be said that when ∆NRCAof commodities categorized as labour-intensiveand resource-intensive manufacture productswas down by one unit, it would have an effecton increasing non-oil exports by 3.13%. The in-crease is the compensation of 1 unit increase ofcommodities ∆NRCA that include Electronicand Electrical products classified as high-skilland technology-intensive manufacture prod-ucts that impact the increase in non-oil ex-ports by 10.59%. It can therefore be said that


the ASEAN countries have more specializa-tion in Electronic and Electrical products com-modities classified as high-skill and technology-intensive manufacture products rather than inthose of labour-intensive and resource-intensivemanufacture products.

7. Conlusion

In general, the study finds that non-oil ex-ports of ASEAN are influenced by the level ofASEAN economies and that of their tradingpartners, the population of ASEAN and trad-ing partners reflecting the market size of ex-porters and importers, trading costs, the sta-tus of AFTA and ASEAN Pluscooperations,as well as changes in the difference betweenASEAN’s comparative advantage as exporterand trading partner’s comparative advantagefor certain commodities in the ASEAN regionalmarket.

Specifically, the study proves that compara-tive advantage has an influence on the patternof non-oil exports. It explains that the increas-ing changes of ASEAN’s NRCA against itstrading partner’s NRCA in the ASEAN mar-ket will increase the non-oil exports of ASEANto its trading partners. In other words, theincrease in the comparative advantage in ex-porting countries or the decline in compara-tive advantage in importing countries increasesthe volume of exports from the exporter to theimporter. This is in line with the Theory ofComparative Advantage or Richardian model,in which a country will have specialization inexporting goods that have higher comparativeadvantage.

The study also reveals that the effects of∆NRCA changes on ASEAN non-oil exportsis ∆NRCA changes for mining commodi-ties (including coal), agricultural commodities,and other commodities. Meanwhile, ∆NRCAchanges for industrial products/manufacturecommodities indicate the smallest effect onnon-oil exports. This signifies that non-oil ex-

ports of ASEAN tend to depend on commodi-ties that derive from natural resources.

If the industrial/manufacture commoditiesare disaggregated, the estimation results sug-gest that ASEAN has a tendency to special-ize in manufacture products other than elec-tronic and electrical products that belong tomedium-skill and technology-intensive manu-facture products as well as products derivedfrom the low-skill and technology-intensive in-dustries. Furthermore, it also indicates thatASEAN has enough advantage in spare partsfor high-tech electronic products. Meanwhile,ASEAN actually has a very small compara-tive advantage in, electric and electronic partsproducts for medium technology as well ashigh-tech electrical products and electronics.

This paper also recommends that, first, fur-ther research is required by disaggregating∆NRCA of commodities that are still aggre-gate because of possible bias aggregate. This isindicated from the presence of significant vari-ations and major variations in the effect ofthe ∆NRCA change against non-oil exportsif industrial/manufacture commodities (com-modity 3) are disaggregated into groups basedon the level of several manufacture commoditygroups, based on the work force skills and thelevel of technology.

Second, it is interesting to study further theinteraction of changes in one comparative ad-vantage of a commodity with changes in com-parative advantage of other commodities. As-suming the existence of this interaction be-comes important, it can be used as one empiri-cal method to find evidence of trade specializa-tion of the concept of comparative advantage.Viewed from the concept of trade specializa-tion according to comparative advantage, wewill show more about the pattern of export andtrade of a country in a given market.

Third, ASEAN member states need to adoptpolicies to better utilize the ASEAN Plusscheme which will then be merged into RCEP.The consideration is the empirical results of


this study that signify that it is actually ex-ports to trading partners under ASEAN Plusthat give higher exports difference, comparedto those of non-ASEAN Plus countries. Mean-while, when ASEAN’s trading partners are in-cluded in AFTA cooperation, it indicates thatthe value of exports is lower.

8. References

[1] Adam, C. & Cobham, D. (2007). Exchange RateRegimes and Trade. Manchester School 75 (s1),44-63.

[2] Anderson, J.E. (1979). A theoretical foundation forthe gravity equation. American Economic Review,69 (1), 106-116.

[3] Appleyard, D.R., Field, A.J.Jr., dan Cobb, S.L.(2006). International Economics, fifth edition,Now York: McGraw-Hill.

[4] Assosiation of South East Asian Nations (ASEAN)Secretariat. (2013). ASEAN Statistical Yearbook2012. Jakarta: ASEAN Secretariat.

[5] Assosiation of South East Asian Nations(ASEAN). (2011). ASEAN and FTA Part-ners Launch The World’s Biggest Regional FreeTrade Deal. April 25, 2014. http://www.asean.

org/news/asean-secretariat-news/item/

asean-and-ftapartners-launch-the-world’

s-biggest-regional-free-trade-deal-html.[6] Balassa, B. (1965). Trade Liberalization and

Revealed Comparative Advantage, ManchesterSchool of Economic and Social Studies, 33, 99-123.

[7] Ballance R, Forstner H, Murray T. (1985). Onmeasuring revealed comparative advantage: a noteon Bowen’s indices. Weltwirtschaftliches Arch121:346-350

[8] ergstrand, J.H. (1985). The Gravity Equation inInternational Trade: Some Microeconomic Founda-tions and Empirical Evidence. The Review of Eco-nomics and Statistics, 67 (3), 474-481.

[9] Bersgstrand, J.H. (1989). The Generalized Grav-ity Equation, Monopolistic Competition, and theFactor-Proportions Theory In International Trade.Review of Economics and Statistics, 71(1), 143-53.

[10] Bowen, H.P. (1983). On The Theoretical Interpre-tation Of Indices Of Trade Intensity And RevealedComparative Advantage. WeltwirtschaftlichesArchiv, 119, 464-472.

[11] Bowen, H.P. (1985). On Measuring Compar-ative Advantage: A Reply And Extension.Weltwirtschaftliches Archiv, 121, 351-354.

[12] Bowen, H.P. (1986). On Measuring Com-parative Advantage: Further Comments.Weltwirtschaftliches Archiv, 122, 379-381.

[13] Chia, S.Y. 2013. The ASEAN Economic Commu-nity: Progress, Challenges, and Prospects.ADBIWorking Paper 440. Tokyo: Asian Devel-opment Bank Institute. http://www.adbi.

org/workingpaper/2013/10/25/5916.asean.

economic.community.progress.challenges/

[14] Costinot, A. (2009). On The Origins Of Com-parative Advantage. Journal of International Eco-nomics 77, 255-264.

[15] Deardorff, A.V. (1995). Determinants of Bilat-eral Trade: Does Gravity Work in A NeoclassicalWorld? NBER Working Paper, 5377, CambridgeMass.: National Bureau of Economic Research.

[16] Deardorff, A.V. (1994). Exploring The LimitsOf Comparative Advantage. WeltwirtschaftlichesArchiv, 130,1-19.

[17] Deardorff, A.V. (2010). Deardorffs’ Glossaryof International Economics. May, 25, 2014.http://www-personal.umich.edu/~alandear/

glossary/t.html#TradePattern.[18] Deardorff, Alan V., 1984, ”Testing Trade Theo-

ries and Predicting Trade Flows”, Chapter 10 padaRonald W. Jones and Peter B. Kenen, eds., Hand-book of International Economics, Volume 1, Ams-terdam: North Holland.

[19] Doing Business.(2014). World Bank Group. Febru-ary 25, 2014. http://www.doingbusiness.org.

[20] Drysdale, P. D. (1967), Japan, Australia and NewZealand: the prospects for Western Pacific integra-tion, Economic Record, Vol. 45(111), September,pp.321-342.

[21] Eaton, J. dan Kortum, S. (1997), Technology andBilateral Trade, in NBER Working Paper, 6253,Cambridge, MA: National Bureau of Economic Re-search.

[22] Edward, L. dan Schoer,V. (2002). Measures ofCompetitiveness : A Dynamic Approach to SouthAfrica’s Trade Performance in the 1990s. TheSouth African Journal of Economics, 70 (6).

[23] Elliott, Robert J.R dan Ikemoto, K. (2004).”AFTA and the Asian Crisis: Help or Hindrance toASEAN Intra-Regional Trade?” Asian EconomicJournal, 18 (1), 1-23.

[24] ESCAP-World Bank. (2014). Interna-tional Trade Cost. April 20, 2014. http:

//databank.worldbank.org/data/views/

variableselection/selectvariables.

aspx?source=ESCAP-World-Bank:

-International-Trade-Costs.[25] Evenett, S.J. and Keller, W. (1998), On the The-

ories Explaining the Success of the Gravity Equa-tion, in NBER Working Paper, 6529, Cambridge,MA: National Bureau of Economic Research.

[26] Feenstra, R. (2004). Advanced International Trade:Theory and Evidence. Princeton: Princeton Uni-versity Press

http://www.asean.org/news/asean-secretariat-news/item/asean-and-fta partners-launch-the-world's-biggest-regional-free-trade-deal-html




http://www.adbi.org/workingpaper/2013/10/25/5916.asean. economic. community.progress. challenges/



http://www-personal.umich.edu/~alandear/glossary/t.html#TradePattern

http://www-personal.umich.edu/~alandear/glossary/t.html#TradePattern

http://www.doingbusiness.org

http://databank.worldbank.org/data/views/variableselection/selectvariables.aspx?source=ESCAP-World-Bank:-International-Trade-Costs






[27] Filippini, C. dan Molini, V. (2003). The Determi-nants of East Asian Trade Fows: A Gravity Equa-tion Approach. Journal of Asian Economics 14,695-711.

[28] Frankel, J. A. (1997), Regional Trading Blocs inthe World Economic System. Washington, DC: In-stitute for International Economics.

[29] Frankel, Jeffrey A. and Rose, Andrew K. 1995. ”Es-timating theEffects of Currency Unions on Tradeand Output. ”Working Paper No. 7857, NationalBureau of Economic Research, August.

[30] Greenaway, D. and Milner, C. (1986), The Eco-nomics of Intra-Industry Trade, Oxford: Blackwell.

[31] Hapsari, I. M. dan Mangunsong, C. (Novem-ber, 2006).Determinants of AFTA Members’ TradeFlows and Potential for Trade Diversion. Asia-Pacific Research and Training Network on TradeWorking Paper Series, No. 21, November 2006.

[32] Hecksher, E.F. (1919). The Effect Of Foreign TradeOn The Distribution Of Income, In Ellis, H. S.And Metzler, L.A., eds., Readings in the Theoryof International Trade, American Economic Asso-ciation, Philadelphia: Blakiston.

[33] Helpman, E. (1987). Imperfect Competition andInternational Trade: Evidence From Fourteen In-dustrial Countries. Journal of the Japa-nese andInternational Economies, 1 (1).

[34] Hillman, A.L. (1980). Observation on the Rela-tion between Revealed Comparative Advantageand Comparative Advantage as Indicated by Pre-Trade Relatives Price. Weltwirtschaflitches Archi,116, 315-321.

[35] Hoen, A.R. dan Oosterhaven, J. (2006). On TheMeasurement Of Comparative Advantage. TheAnnals of Regional Science, 40, 677-691.

[36] Kepaptsoglou, K., Karlaftis, M.G., Tsamboulas,D. (2010). The Gravity Model Specification ForModeling International Trade Flows And FreeTrade Agreement Effects: A 10-Year Review OfEmpirical Studies, The Open Economics Journal,3, 1-13.

[37] Kimura F, Takahashi Y, Hayakawa K. 2007. Frag-mentation and parts and components trade: com-parison between East Asia and Europe. N Am JEcon Finance 18:23-40

[38] Laursen, K. (1998). Revealed Comparative Advan-tage And The Alternatives As Measures Of In-ternational Specialisation. DRUID Working Paper,98-30.

[39] Linnemann, H. (1966). An Econometric Study ofInternational Trade Flows. Amsterdam: North-Holland.

[40] Logistics Performance Index (LPI). (2014). WorldBank. February 25, 2014. http://www.worldbank.org.

[41] Mlitz, J. (2007). North, South and Distance in the

Gravity Model. European Economic Review, 91,971-91.

[42] Menon, J. (1996), Adjusting towards AFTA: TheDynamics of Trade in ASEAN, Singapore: Insti-tute of Southeast Asian Studies.

[43] Ng F. and A. Yeats. (June, 2003). Major TradeTrends in East Asia: What are their Implicationsfor Regional Cooperation and Growth., WorldBank Policy Research Working Paper, 3084.

[44] Novy, Dennis (2009). Gravity redux: measuring in-ternational trade costs with panel data. WorkingPaper. Coventry: University of Warwick, Depart-ment of Economics.

[45] Ohlin, B. (1933). Interregional and InternationalTrade, Cambridge, MA: Harvard University Press.

[46] Pyhnen, P. (1963). A Tentative Model forthe Volume of Trade between Countries:Weltwirtschaftliches Archiv, 90, 24-40.

[47] Proudman, J. dan Redding, S. (1998). OpennessAnd Growth. The Bank of England, United King-dom.

[48] Rahman, M. (2003). A Panel Data Analysisof Bangladesh’s Trade: The Gravity Model Ap-proach. April 26, 2014. https://eprints.usq.

edu.au/5794/1/Rahman_ACE09_AV.pdf

[49] Razzaghi, S., Ali, M., Azad, M., dan Sofi, Y.(2012). The Determinants of Trade Flows betweenD-8 Group Members through Gravity Model. J.Basic. Appl. Sci. Res., 2(6), 5726-5731.

[50] Ricardo, D., (1817). On the Principles of PoliticalEconomy and Taxation, London: John Murray.

[51] Rose, A.K. dan van Wincoop, E. (2001). NationalMoney as a Barrier to International Trade: TheReal Case for Currency Union. The American Eco-nomic Review, 91 (2), 386-390.

[52] Salvatore, D. (2002). International Economics(3rd edition). New York: Macmillan

[53] Sarwedi.(2010). Analisis Determinan PerubahanPenawaran Barang Ekspor Indonesia. BuletinEkonomi Moneter dan Perbankan, 12 (3), 355-376.

[54] Sohn C.H. (2005). Does The Gravity Model Ex-plain South Korea’s Trade Flows? Japanese EconRev, 56 (4), 417-30.

[55] Tayyebi, S. K. (2005), ”Trade integration inASEAN: An application of the panel gravitymodel,” in Charles H. (Ed.), New East Asian Re-gionalism, Causes, Progress and Country Perspec-tives, Edward Elgar, pp. 109- 127.

[56] Tinbergen, J. (1962). An Analysis of World TradeFlows, J. Tinbergen (ed.): in Shaping the WorldEconomy. New York: The Twentieth CenturyFund.

[57] Trkcan K. 2011. Vertical Intra-Industry Trade andProduct Fragmentation in the Auto-Parts Indus-try. Journal of Industry, Competition and Trade,Springer, vol. 11(2), pages 149-186, June

http://www.worldbank.org

http://www.worldbank.org

https://eprints.usq.edu.au/5794/1/Rahman_ ACE09_AV.pdf

https://eprints.usq.edu.au/5794/1/Rahman_ ACE09_AV.pdf


[58] UNComtrade. (2014). February, 20, 2014.http://www.wits.worldbank.org

[59] United Nations Conference on Trade andDevelopment (UNCTAD). (2013, June12). UNCTAD Product groupings. Febru-ary 2, 2014. http://unctadstat.unctad.

org/UnctadStatMetadata/Classifications/

UnctadStat.SitcRev3Products.

UnctadProductGroupingslist.Classification_

En.xls

[60] United Nations Conference on Trade and Devel-opment (UNCTAD). (2013, October 2). UNCTADManufactured goods by degree of manufacturing.February, 20, 2014. http://unctadstat.unctad.

org/UnctadStatMetadata/Classifications/

UnctadStat.SitcRev3Products.

DegreeOfManufacturing.Classification_En.

xls

[61] Vollrath, T.L. (1991). A Theoretical Evalu-ation Of Alternative Trade Intensity Mea-sures Of Revealed Comparative Advantage.Weltwirtschaftliches Archiv, 127, 265-280.

[62] Widyasanti, Amalia Adninggar. (2010). Do Re-gional Trade Areas Improve Export Competitive-ness? A Case of Indonesia.Bulletin of Monetary,Economics and Banking, 13 (1), 5-22.

[63] World Development Indicator (WDI). (2014).WorldBank. February 25, 2014. http://lpi.

worldbank.org.[64] Yamarik, S. dan Ghosh, S. (2006). A Sensitivity

Analysis Of The Gravity Model. The InternationalTrade Journal, 19 (1), 83-126.

[65] Yeats, A.J. (1985). On The Appropriate Inter-pretation Of The Revealed Comparative Advan-tage Index: Implications Of A Methodology BasedOn Industry Sector Analysis. WeltwirtschaftlichesArchiv, 121, 61-73.

[66] Yu, R., Cai, J., dan Leung, P. (2009). The Normal-ized Revealed Comparative Advantage Index. TheAnnalysis of Regional Science, 43(1), 267-282.

[67] Yue, C. dan Hua, P. (2002). Does Comparative Ad-vantage Explains Export Patterns In China? ChinaEconomic Review, 13, 276-296.

http://unctadstat.unctad.org/UnctadStatMetadata/Classifications/UnctadStat.SitcRev3Products.UnctadProductGroupingslist.Classification_En.xls





http://unctadstat.unctad.org/UnctadStatMetadata/Classifications/UnctadStat.SitcRev3Products.DegreeOfManufacturing.Classification_En.xls





http://lpi.worldbank.org

http://lpi.worldbank.org

impact analysis of normalized revealed comparative...

Documents