The effect of APEC on the import trade of members v.s. non-members,

The effect of APEC on the import trade of members v.s. non-members, 1990-2007 BY Fung Ka Ying Sandi 06009972 Jiang Yang 06050395 Applied Economics Option An Honors Degree Project Submitted to the School of Business in Partial Fulfillment of the Graduation Requirement for the Degree of Bachelor of Business Administration (Honors) Hong Kong Baptist University Hong Kong April 2009

Abstract The proliferation of Regional Trade Agreements (RTAs) not only has been encouraging trade liberalized but also aims to reduce trade barriers. People continued to debate about whether such establishment of RTAs is an effective way of achieving free trade RTAs are trade creation (increase total trade) or trade diversion (the extra trade occurs at the expense of non-members). The paper examines the effects of one of the RTAs called Asia- Pacific Economic Cooperation (APEC) by using a gravity model to see if it has been trade creating or trade diverting. Annual data of members and non-members of APEC for the year 1990-2007 was used. The results show that the effects of Asia -Pacific Economic Cooperation (APEC) keep change in time. The RTA tended to be trade creating after 2001, that is, it fostered greater trade with trading partners and with the rest of the world. 2

Acknowledgement We would like to take the opportunity to thank our supervisor Dr. Yun-Kwong Kwok from Economics Department for his continuous guidance, support and contribution to this honor project. He kindly devoted his precious time in guiding us on the paper. We won t be able to complete the project without his help. Not only have we gained much knowledge, but also our independent mind and critical thinking are greater enhanced under his guidance. Fung Ka Ying Sandi Jiang Yang 3

Content 1. Introduction 2. Background of APEC 3. Literature Review 4. Methodology 5. Data 6. Estimation procedure and results 7. Limitation 8. Conclusion References Table Appendix 4

1. Introduction The enlargement of Regional Trade Agreements (RTAs) has been one of the major international developments in recent years. For example, European Union (EU), North American Free Trade Association (NAFTA), Association of South East Nation (ASEAN), etc, have shown successfully. Basically, these RTAs aim to reduce trade barriers in a preferential way, which is same as the World Trade Organization (WTO) has been encouraging trade liberalized. However, there is continued criticism as to whether such RTAs are ways to achieve free trade effectively, or the increased trade among members just caused less trade with non-members countries. Asia-Pacific Economic Cooperation (APEC), originally just a kind of forum for the dialogue between East and West coat of pacific ocean, now it has transformed and actively stepping toward the goal of free and open trade. Particularly, the APEC is including the two largest economies, China and the US, which should have a large impact of the world trade. This paper addresses the question of whether the APEC enhances welfare. This is examined by using a gravity model to analyze the effect of APEC on the import trade of both members and non members of APEC. To test whether an RTA enhances welfare or not, we will see if it is trade creation or trade diversion. Trade creation occurs when the introduction of an RTA allows an importing country to purchase products at lower cost than the country did in the past. No doubt, this benefits both the importing countries and the world as a whole. In contrast, trade diversion is the substitution of a more costly source of supply within an RTA for a less costly source outside, and this would negatively affect welfare. As the introduction of an RTA may have both trade creation and trade diversion effects, or either nor neither. Therefore, it is the net effect that needs to be assessed when deciding whether an RTA enhances or hinders welfare. This 5

paper empirically attempts to test for the existence of trade creation or trade diversion as a result of APEC by applying econometric gravity model and extends with tax, exchange rate, dummy variables in this paper. The rest of this paper is structured as follows. Section two introduces the background of APEC. Section three is the review of literature. Section four is the empirical methodology which gives a detailed demonstration of the gravity model that applied in the paper. Section five provides a description of the data used. Section six explains the estimation procedures and results. Section seven discusses the limitations of the model estimations. Section eight gives conclusion of the paper. 2. Background of APEC Asia-Pacific Economic Cooperation (APEC) is the earliest forum for facilitating economic growth, cooperation, trade and investment in the Asia-Pacific region. APEC operates on the basis of non-binding commitments, open dialogue and equal respect for the views of all participants. It is the only inter governmental grouping in the world. APEC has no treaty obligations required of its participants, which is different from the WTO or other multilateral trade bodies. APEC makes decision with consensus and commitment on a voluntary basis. APEC was launched in 1989 which aims to further enhance economic growth and prosperity for the region and to strengthen the Asia-Pacific community. APEC has 21 members - referred to as Member Economies - which account for approximately 40.5% of the world's population, approximately 54.2% of world GDP and about 43.7% of world trade 1. These 21 Member Economies are Australia, Brunei 1 Refers to: http://www.apec.org/apec/about_apec.html 6

Darussalam, Canada, Chile, China, Hong Kong, Indonesia, Japan, South Korea, Malaysia, Mexico, New Zealand, Papua New Guinea, Peru, The Philippines, Russian, Singapore, Taiwan, Thailand, The United States and Viet Nam. Since its establishment, APEC has worked to reduce tariffs and other trade barriers across the Asia-Pacific region, creating efficient domestic economies and dramatically increasing exports. The key to achieve APEC s vision is referred to the Bogor Goals of free and open trade and investment in the Asia-Pacific by 2010 for industrialized economies and 2020 for developing economies. These goals were adopted by the political leaders at their 1994 meeting in Bogor, Indonesia. Through policy alignment and economic and technical cooperation, APEC also works to create an environment for the safe and efficient movement of goods, services and people across border of the region. APEC member economies generated nearly 70% of global economic growth, in order to sustain this growth, APEC member economies work together and promise to follow the commitment to open trade, investment and economic reform. Because of the reduced trade barriers and a more economically competitive region, the price of daily base goods and services are lower, APEC member economies on average enjoy lower cost of living. APEC works in three broad areas to achieve the Bogor goals, these three key areas are known as Three Pillars. They are: Trade and Investment Liberalization, Business facilitation and Economic and Technical Cooperation. The outcomes of these three areas enable APEC member economies to strengthen their economies by pooling resources within the region and achieving efficiencies. Consumers in the APEC region also get tangible benefit through greater choices in the market place, 7

cheaper goods and services, increased training and employment opportunities and improved to access to international markets. The objective of Borgor has not been exactly practice until Jun 2001, when APEC Trade Ministers endorsed a set of APEC Trade Facilitation Principles which was developed by an ad hoc task force led by Hong Kong. The APEC Trade Facilitation Principle reaffirmed the importance in achieving the Bogor goal of free and open trade and investment in the Asia-Pacific, in providing significant benefits for both governments and business, and in generating welfare gains for the economy as a whole. 3. Literature Review The Regional Trade Agreements (RTAs) is considered to be a step toward free trade and welfare enhancing. The RTAs aim not only to reduce trade barriers to members but also did not increase any trade barriers to non-members. However, Viner (1995) brought an idea that the RTAs lead to both trade creation and trade diversion. Trade creation happens when the formation of an RTA allows an importing country to purchase products at lower cost than was previously the case. This benefits both the importing country and the whole world even though the countries are non-members of the RTA. But trade diversion is the substitution of a more costly source of supply within the RTA for a less costly source outside, and this would result negatively affect welfare. As an RTA will lead to both trade creation and trade diversion effects, we have to assess the net effect in order to decide whether it enhances welfare. Later, Johnson (1960) developed an equilibrium diagram to explain the economic effects of trade creation and trade diversion. The partial equilibrium 8

diagram shows that the affect of an RTA is a sum of several effects, and that countries may be better or worse off in markets where trade is diverted. Johnson s conclusions are confirmed when using recent techniques to compute the general equilibrium models. Pomfret (1997) and Schiff and Winters (2003) both have good expositions of Johnson s partial equilibrium diagram. Low (2003) and Lloyd and MacLaren (2004) have an excellent survey article of analyzing the RTAs. 2 Comparing with other RTAs in the world, such as the Australian and New Zealand Closer Economic Relations (CER), the Southern Cone Common Market (MERCOSUR) and the North American Free Trade Association (NAFTA) which have been established to reduce trade protection among member states, APEC is a little different of structure. APEC was built on the concept of open regionalism, which means that any members reducing its trade barriers to other members should offer the same reductions to non-members. 4. Methodology 4.1 Gravity models Gravity model is the most common empirical tool to use to estimate the effect of RTAs. They are very popular in analyzing economic phenomena related to the flow of goods and/or services. A gravity model involves regressing trade on a number of explanatory variables, then using dummy variables to ascertain whether this relationship is affected by the existence of RTAs. Beckerman (1956), Anderson (1979), Bergstrand (1989), Oguledo and Macphee (1994), Frankel and Wei (1995), Kreinin and Plummer (1998), Krueger (1999), Cernat (2001), Havemann, Nair- Reichert, and Thursby (2003), Adams, Dee, Gali, and McGuire (2003), Filippini and 2 The literature review cited from: Bhavish Jugurnath, Mark Stewart and Robert Brooks. (2007). Asia/Pacific Regional Trade Agreements: An empirical study. ScienceDirct: Journal of Asian Economics 18 947-987. 9

Molini (2003) and Tang (2005) are just some of the studies which have used gravity models to estimate the trade creation and trade diversion effects of various RTAs, including the North American Free Trade Association (NAFTA), the Southern Cone Common Market (MERCOSUR), the Australian and New Zealand Closer Economic Relations (CER) and the Association of South East Asian Nations (ASEAN). 3 Trade volume is directly related to GDP, population and land area of the countries involved, as well as transaction costs which are usually proxied by distance and cultural similarities (a common language is often used). This paper tests these factors and adds the cost variables of the exchange rate and taxes to the list to decide the effect of APEC on its members and non-members. 4.2 Dependent variables 4.2.1. Imports Typical gravity model uses total trade (imports plus exports) as the dependent variable, but in this paper, we focus on imports since they are more closely substitute the effects of domestic trade barriers. The real import is obtained and deflated by the GDP deflator to account for inflation. 4.3 Independent variables 4.3.1. Gross Domestic Product, GDP In this model, we have two GDP variables, GDP i for the importing country 4 and GDP j for the exporting country. GDP deflator is used to account for inflation then obtains the real GDP. It is anticipated that richer countries trade more in term of 3 The literature review cited from: Bhavish Jugurnath, Mark Stewart and Robert Brooks. (2007). Asia/Pacific Regional Trade Agreements: An empirical study. ScienceDirect: Journal of Asian Economics 18 974-987. 4 Although we know that some members of APEC are not officially be treated as independent countries but economies, in our regression, we simply name them as country for generally speaking. 10

goods and services. Therefore, we expect a positive relationship between these variables and imports to country i. 5 4.3.2. Population, POP There are also two variables here, POP i for the importing country and POP j for the exporting country. It is expected that countries with larger populations will both import and export more. But Aitken (1973) suggested that the larger a country s population, the larger the domestic market compares to the foreign market, so potential export supply must be smaller. However, Bergstand (1989) pointed out that larger population would allow for economies of scale, which may increase the price competitiveness of the exporter s production, causing higher exports. Therefore, the sign on the coefficient of population of the exporter may be indeterminate, while expecting the sign for the importer is positive. 4.3.3. Distance, DIST DIST ij is the geographical distance between the capital cities of the importing country i and the exporting country j. The physical distance between the trading countries represent the transport cost, therefore, it is expected DIST would be negatively correlated with trade. 4.3.4. Surface area, AREA This is the total land area of a country, including areas under inland bodies of water and coastal waterways. AREA i is the surface area of the importing country, and AREA j is for the exporting country. We anticipate that larger countries will both 5 Note that through out the analysis, i refers to the importing country, while j refers to the exporting country. 11

import and export more. But relative size may also be important for comparative advantage reasons. Therefore, the sign for the AREA coefficients may also be indeterminate. 4.3.5. Exchange rate, EXR EXR i is the exchange rate of the importer, and EXR j is for the exporter. EXR is defined as the local currency value per $US divided by the US GDP deflator. For country i, an increase in EXR i represents a depreciation of the local currency or a fall of that country s relative price, so the coefficient of it is expected to be negative. And we expect a positive sign for country j. 4.3.6. Taxes on goods and services, TAX TAX i is the simple average final bound for the importer, and TAX j is for the exporter. Simple average final bound is the simple average of final bound duties excluding unbound tariff lines. 6 Since all taxes (except lump sum taxes) are distorting, it is expected that both of these variables have a negative relationship with the imports to country i. 4.3.7. Language, LANG Linnermann (1966) stated that a common language between two countries would influence the volume of trade. 7 LANG ij is a dummy variable which is assigned a value of one if the two trading countries have the same official language and zero otherwise. We anticipate countries which have a common language will trade more. Therefore, a positive relationship is expected. 6 Refers to: http://www.wto.org 7 Linnermann, H. (1966). An economic study in international trade flows. Amsterdam: North Holland. 12

4.4 Model specification The basic gravity model uses the above variables and is shown as Eq. (1) Log IMPORT ijt = α 0 + α 1 log GDP it + α 2 log GDP jt + α 3 log POP it + α 4 log POP jt + α 5 log DIST ij + α 6 log AREA i + α 7 log AREA j + α 8 log EXR it + α 9 log EXR jt + α 10 TAX it + α 11 TAX jt + α 12 LANG ij + µ ijt (1) We use a log specification because it gives the best results typically. And i is the importing country, while j is the exporting country. To examine the effects of APEC within the framework of this equation, dummy variables are then added. When this technique has been used by the others, a consensus that RTAs are generally trade creating has emerged. For instance, Frankel and Wei (1995) and Frankel (1997) showed trade creation in Asian and in North American trading blocs. And a recent paper by Rose (2000) also found that RTAs, were trade creating in general. However, Hassan (2001) found evidence of trade diversion for the South Asian Association for Regional Cooperation (SAARC). 8 To look at the effects of APEC, the basic gravity model outlined in Eq. (1) is extended using three dummy variables. The dummy variables take value of one if the county is a member of APEC and zero otherwise, regardless of the membership status of the trading partners. Eq. (2) is shown as below. 8 The literature review cited from: Bhavish Jugurnath, Mark Stewart and Robert Brooks. (2007). Asia/Pacific Regional Trade Agreements: An empirical study. ScienceDirect: Journal of Asian Economics 18 974-987. 13

Log IMPORT ij = the above +α 13 APEC i APEC j + α 14 APEC i +α 15 APEC j (2) APEC i takes the value of one if the importing country i is a member of APEC and zero otherwise. If we get a positive coefficient for this variable (α 14 >0), it means countries that are members of APEC will import more than an equivalent country that is not a member, and this would imply that APEC is trade creating. The second dummy variable, APEC j takes the value of of one if the exporting country j is a member of APEC and zero otherwise. If α 15 > 0, countries are importing more from other countries which are members of APEC, and again this indicates that APEC is trade creating. When positive coefficient on α 14 and α 15 imply trade creation, negative coefficient means that trade is lower, and the formation of APEC may be trade diverting. If the dummy variable, APEC i APEC j values one, it means both the importing country and exporting country j are members of APEC. Therefore, α 13 > 0 means that APEC countries import more from countries that are also members of APEC. An indication would be that, if α 14 and α 15 are significant and negative, while α 13 is significant and positive, APEC is trade diverting. 4.5 Random-effects model (REM) or Error-components model In this paper, the gravity model is examined with many different countries in different years. Therefore, we might wish to choose a pooled cross-section and timeseries model in which error terms may be correlated across time and individual units. So, we select the random-effects model instead of the ordinary least square (OLS) pooled data model to run regression for the gravity model. The random-effects model is as follows: 14

Y it = α i + βx it + e it for i =1, 2,, N (number of cross-section units) and t =1, 2,, T (number of time periods) and e it = u i + v t + w it 2 Where u i ~ N(0, σ u ) = cross-section error component 2 v t ~ N(0, σ v ) = time-series error component 2 w it ~ N(0, σ w ) = combined error component Here the individual error components are assumed to be uncorrelated with each other and not auto-correlated across both cross-section units and time periods. Now the intercept term in the model is treated and decomposed into two random variables, one is a time series variable and the other is a cross-section variable. 5. Data The Data used in this paper covers a period of 18 years from 1990 to 2007 and includes 10 APEC member countries and 6 non-member countries. The 9 member countries that we choose are Australia, Canada, China, Japan, Republic of Korea, Malaysia, New Zealand, The Philippines, Singapore and United States. The 6 non-member countries that we choose are Brazil, Egypt, France, Germany, India and United Kingdom. It can be seen from the Appendix that all these 16 countries selected are mostly ranked as the top 50 GDP countries. They are very outstanding countries of 15

trade in the whole world. The countries have very close export and import relationships with each other. Some countries even become main export and import partners. Advanced & high income OECD countries 9 Australia, Canada, France, Germany, Japan, Korea, New Zealand, Singapore, United Kingdom and United States Developing countries Brazil, China, Egypt, India, Malaysia and Philippines Not only did these selected countries join the APEC, they joined other RTAs as well. These countries remain active trading countries over the world. The following table shows the list of Regional Trade Agreements (RTAs). PTA 10 ASEAN or the AFTA (Association of South East Asian Nations or the Asian Free Trade Association) APEC (Asia-Pacific Economic Cooperation) CER (Australia-New Zealand, Closer Economic Relations Trade Agreement) MERCOSUR (Southern Cone Common Market) NAFTA (North American Free Trade Agreement) Date of Selected Countries Entry 1967 Malaysia, Philippines and Singapore 1989 Australia, Canada, China (1991), Japan, Korea, Malaysia, New Zealand, Philippines, Singapore and United States 1983 Australia and New Zealand 1991 Brazil 1994 Canada and United States are chosen. Considering that the paper has limited data, countries from all over the world 9 OECD: Organization for Economic Co-operation and Development 10 PTA: Preferential trading area, a trading bloc which gives preferential access to certain products from certain countries 16

Continents Asia Africa North America South America Europe Oceania Selected Countries China, India, Japan, Korea, Malaysia, Philippines and Singapore Egypt United States Brazil France, Germany and United Kingdom Australia and New Zealand Hence, these total 16 countries are taken into the consideration in the study. The data on GDP, populations and exchange rate are obtained from the International Financial Statistics of International Monetary Fund. And the import data between countries are from the Direction of Trade Statistics of International Monetary Fund. The other variables, definitions and sources are listed in Table 1. Insert Table 1 6. Estimation procedure and results The data used in this paper covers a period of 18 years from 1990 to 2007 and includes 16 countries from both members and non-members of APEC. This model is estimated for three sub-periods: 1990-1995, 1996-2001 and 2002-2007. We use the pool data to estimate a single regression equation, which allow the coefficients to be different in the three group periods. The model is also estimated for the entire period using the full panel. The results of these equations are presented in Tables 2 and 3. 17

Insert Table 2 Insert Table 3 6.1 The basic gravity model results Table 2 reports the basic gravity model results when using pooled data for the three groups of observations as well as for the entire sample of 1990-2007. In general, these equations fit the data well, most of the individual coefficients are statistically significant, and most of the signs satisfy the theoretical estimations, indicating that the proposed explanatory variables are significantly related to bilateral trade. The coefficients of determination (R 2 ) range between 31% and 90%.The F test (p-value) results show that collectively the models are highly significant. These results are in line with the usual gravity model findings of other papers. Table 2 shows that the coefficients on GDP i and GDP j are all significant and positively signed. And the population coefficients are all positively signed and are always significant. It indicates that, richer and more populace countries tend to trade more. Also, as we expect for the coefficients of the distance variables, DIST ij are all significant and negatively signed. This tells us that transportation costs play an important role when we determine the volume of trade between countries. The coefficients on AREA i and AREA j are all negatively signed and are almost always significant. Frankel (1997) also found a similar result to this, and he suggested it was the reason that large countries tend to have more natural resources and they often trade less with others. The coefficients on LANG ij are all significant and positive. This implies that countries with common language have cultural similarities, so their trade contract will be easier to make. 18

When we come to the exchange rate, let s recall that an increase in EXR i and EXR j infers a depreciation of the real exchange rate. In line with our expectations the coefficients associated with the importing country, EXR i are all negative and almost always significant, while EXR j are mostly positive and significant for the earlier years 1990-1995, 1996-2001 as well as for the whole period 1990-2007. The coefficients on taxes are always negative and significant. This implies that higher taxes reduce trade between countries, which is in line with our expectations. 6.2 The gravity model including APEC The estimates for APEC show that this RTA had a mostly positive effect for the whole world during the period 1990-2007. But the main purpose of this paper is to look into the details of different sub-periods. For the sub-period 1990-1995 and 1996-2001, the coefficients on APEC i, APEC j and APEC i APEC j are positive, while the probability for APEC j is not significant. This indicates that the effect of APEC to the selected countries was not that obvious when the RTA has just formed. And we can not conclude whether APEC was trade creating or trade diverting for the following two reasons. First, APEC began as an informal dialogue group or a forum between countries in 1989. Then it took APEC a few starting years to set its vision which is the Bogor Goals of promoting for free and open trade between members and Individual Action Plans. Second, though APEC followed the aim of WTO to encourage trade liberalization, it was formed on a voluntarily base. Members are not forced to follow all the actions. But in the paper of Bhavish Jugurnath, Mark Steward, Robert Brooks (2007) shows that APEC was trade diversion in these two sub-periods. 19

For the sub-period 2002-2007, the result changed a lot. The coefficients on APEC i, APEC j and APEC i APEC j are positive and significant. This represents that APEC became trade creating. This may result from the objective of Borgor in Jun 2001 when APEC Trade Ministers endorsed a set of APEC Trade Facilitation Principles. The APEC Trade Facilitation aims to promote the application of the APEC Business Travel Card (ABTC). The APEC travel card was originally trialed in 1997 there were only a few members joined the scheme. But in the year 2001, the scheme was expanded to most of the members. And now there are 19 economies currently participating in the ABTC Scheme: Australia Brunei, Darussalam, Chile, China, Hong Kong (China), Indonesia, Japan, Korea, Malaysia, Mexico, New Zealand, PNG, Peru, Philippines, Singapore, Chinese Taipei, Thailand and Vietnam. The ABTC aims to provide fast and efficient travel for business people within the APEC region. Card holders enjoy fast entry and exit through special APEC lanes at major airports, not needing to individually apply for visas for entry permits each time when business person travels and card is valid for three years from first issue. Also the e-apec Strategy was adopted in 2001, which sets out the agenda to strengthen the structures of market and institutions, facilitate infrastructure investment and technology for on-line transactions and promote entrepreneurship. The effect of APEC may change over time. It is believed that the trade creation effect of APEC will become stronger in the future. For example, the advanced APEC members have completed their tariff reductions until 2010, while the developing members have completed until 2018. 20

7. Limitations of the model estimations In this paper, all GDP are obtained from the International Financial Statistics, since they are in national currency, for the ease of comparisons, we have to convert the data into US dollars by using the exchange rate between two countries, and the exchange rates are always volatile and frequently fluctuated. Another limitation is that we only include 16 countries in our estimations for this paper, so the results for the gravity model may not perfectly reflect our expectations. Also, policies of the countries which may be related to trade are not being considered in our model, in fact, these policies may affect the volume of the trade flows. The last limitation is that the data of taxes (simple average of final bound duties) we obtained from the WTO only provided the data of 2006 and 2008. So we use the data of 2006 for all the years 1990-2006 and the data of 2008 for 2007. And this limits our observation for the effect of taxes cause to trade flows. 8. Conclusion This paper examines bilateral trade of APEC based on a gravity model, with data of 16 countries from 1990 to 2007 which involves 10 members and 6 nonmembers. And analyze the effect of APEC by three sub-periods: 1990-1995, 1996-2001 and 2002-2007. The estimated coefficients from the basic gravity model show that GDP, population, distance between capital cities of trading partners, cultural similarity (a common official language) and physical area explain much cross country trade. Some price variables, such as real exchange rate and taxes are also used in this study. And the study finds that the final results line up with prior expectations. That is, taxes have a negative effect towards free trade. The movements of real exchange rate 21

have mostly the expected effect. The depreciation of an economy s currency encourages exports while discourages imports. From the result of the regression estimated in this paper that the APEC did not have a significant effect to its member and non-member for the period of 1990-1995 and 1996-2001. It can not be concluding that whether APEC was trade creating or trade diverting. This is because that APEC was acted as a forum in the early years and it was built on a base of voluntary. We can see significant trade creation among members and non-members of APEC during the period 2002-2007 when APEC Trade Facilitation Principles and e- APEC Strategy was adopted in 2001. As well as the expanded application of APEC Business Travel Card brings fast and efficient travel for business people within the APEC region. The study expects that the trade creating effect of APEC will become stronger in the future as the advanced APEC members have completed their tariff reductions until 2010, and the developing members have completed their tariff reductions until 2018. 22

Reference: APEC Business Travel Card (ABTC), http://www.businessmobility.org/key/abtc.html APEC Official Website, http://www.apec.org Area and Population of Countries, http://www.infoplease.com/ipa/a0004379.html Bhavish Jugurnath, Mark Stewart and Robert Brooks. (2007). Asia/ Pacific Regional Trade Agreements: An empirical study. ScienceDirect: Journal of Asian Economics 18 974-987. Database and Browser: Direction of Trade Statistics. International Monetary Fund. Database and Browse: International Financial Statistics. International Monetary Fund. Giorgio Fazio, Ronald MacDonald and Jacques Melitz. (2005).Trade Costs, Trade Balances and Current Accounts: An Application of Gravity to Multilateral Trade. Presented at CESifo Area conference on Macro, Money & International Finance, February. Havemann, J. (2002). International Trade Data 20.08.2004, http://www.macalester.edu/research/economics/page/haveman/trade.resources/ TradeData.html International Monetary Fund Home Page, http://www.imf.org/external/index.htm Jeffrey A. Frankel, Shang-Jin Wei and Ernesto Stein. (1994). APEC and Regional Trading Arrangements in the Pacific. Institute for Inernational Economics. Linnermann, H. (1966). An economic study in international trade flows. Amsterdam: North Holland. László Mátyás. (1998). The Gravity Model: Some Econometric Considerations. Budapest University of Economics, and Erudite, Université de Paris XII. Matthieu Bussière, Jarko Fidrmuc and Bernd Schnatz. (2008). EU Enlargement and Trade Integration: Lessons from a Gravity Model. Rebiew of Development Economics 12(3) 562-576. Measuring Worth, http://www.measuringwoth.com/index.html Richard E. Feinberg. (2003). Political Economy of United States Free Trade Arrangements. Blackwell Publishing Ltd. Tingsong Jiang, Warwick J. Mckibbin. (2008). What Does a Free Trade Area of the Asia- Pacific Mean to China. Brookings Global Economy and Development. 23

Vinod K. Aggarwal. (2007). Chapter 3: The Political Economy of a Free Trade Area of the Asia Pacific: A U.S. Perspective. An APEC Trade Agenda? The Political Economy of a Free Trade Area of the Asia-Pacific. The World Bank, http://www.worldbank.org The World Factbook of Central Intelligence Agency, https://www.cia.gov/library/publications/the-world-factbook WTO website, http://www.wto.org 24

Table 1: Variables, definitions and sources Variables Expected signs Definitions Measures Sources IMPORT It is counted as the total of bilateral trade that is total trade, consisting of merchandise trade and services trade divided by the GDP deflator. GDP + Sum of gross value added by all resident producers in the economy plus any product taxes and minus any subsidies not included in the value of the products. It is counted without making deductions for depreciation of fabricated assets or for depletion and degradation of natural resources. Then deflated by the GDP deflator to obtain the real GDP. POP +/- Total population is based on the de facto definition of population, which calculates all residents regardless of legal status or citizenship, except for refugees not permanently settled in the country of asylum, whom are generally considered part of the population of their country of origin. DIST - Distance between capital cities of pairs of countries. Million US$ Million US$ Millions International Monetary Fund, Direction of Trade Statistics(2008) International Monetary Fund, International Financial Statistics(2008) International Monetary Fund, International Financial Statistics(2008) km Havemann (2002) International trade website and Fitzpatrick and Modlin (1986) book of direct line distances AREA -/? A country s total area. Km 2 Infoplease EXR +/- It refers to the rate determined in the legally sanctioned exchange market. It is calculated as an annual average (local currency units relative to the US dollar). The US GDP deflator is used to account for inflation. Average units International Monetary Fund, International Financial Statistics(2008) 25

Table 2 Basic gravity model estimates - pooled regression analysis Variable Expected signs 1990-1995 1996-2001 2002-2007 1990-2007 C 3.999 0.0000*** 4.917 0.0000*** 3.897 0.0000*** 2.799 0.0000*** GDPi + 0.323 0.0000*** 0.194 0.0000*** 0.281 0.0000*** 0.405 0.0000*** GDPj + 0.076 0.0000*** 0.067 0.0003*** 0.212 0.0000*** 0.155 0.0000*** POPi + 0.040 0.2307 0.079 0.0164** 0.031 0.3345 0.039 0.1838 POPj? 0.142 0.0000*** 0.189 0.0000*** 0.122 0.0001*** 0.224 0.0000*** DISTij - -0.228 0.0000*** -0.272 0.0000*** -0.301 0.0000*** -0.190 0.0000*** AREAi - /? -0.101 0.0000*** -0.060 0.0005*** -0.058 0.0005*** -0.113 0.0000*** AREAj - /? -0.021 0.2474-0.046 0.0074** -0.064 0.0001*** -0.077 0.0000*** LANGij + 0.125 0.0691* 0.226 0.0053** -0.019 0.0314** 0.311 0.0000*** EXRi - -0.091 0.0000*** -0.043 0.0054** -0.001 0.2028-0.023 0.0624* EXRj + 0.064 0.0000*** 0.028 0.0607* -0.001 0.9213 0.130 0.0000*** TAXi - 0.000 0.9603-0.012 0.0000*** -0.011 0.5858 0.003 0.1203 TAXj - -0.035 0.0000*** -0.033 0.0000*** 0.167 0.0000*** -0.028 0.0000*** Weighted Statistics R-squared 0.895 0.318 0.543 0.812 Adjusted R-squared 0.894 0.312 0.539 0.812 Prob(F-statistic) 0.000 0.000 0.000 0.000 DW stat 1.270 1.299 0.828 0.550 Cross-sections 240 240 240 240 Total pool observations 1440 1440 1440 4320 Dependent variable: IMPORT. Method: pooled random effects. *p<0.10, **p<0.05, ***p<0.001, significant at the 10%, 5% and 1% level, respectively. * Base on the personal judgment, import data has been calculated by two ways, one is to deflate import by the GDP deflator of the country in question, and the other one is to deflate import by the US GDP deflator. Since we have done a lot of trial and error, the results of using the GDP deflator of the country in question shows the best result, and we believe that this way of calculating the import data fit the model the most, so we choose this to present in the paper. The results of using the US GDP deflator are shown in Appendix 2 for comparison. 26

Table 3 Extended gravity model estimates on APEC - pooled regression analysis Variable Expected signs 1990-1995 1996-2001 2002-2007 1990-2007 C 3.796 0.0000*** 3.924 0.0000*** 3.159 0.0000*** 2.764 0.0000*** GDPi + 0.313 0.0000*** 0.221 0.0000*** 0.250 0.0000*** 0.366 0.0000*** GDPj + 0.067 0.0000*** 0.089 0.0000*** 0.192 0.0000*** 0.120 0.0000*** POPi + 0.067 0.0234** 0.127 0.0000*** 0.119 0.0000*** 0.092 0.0004*** POPj? 0.148 0.0000*** 0.217 0.0000*** 0.200 0.0000*** 0.238 0.0000*** DISTij - -0.228 0.0000*** -0.221 0.0000*** -0.265 0.0000*** -0.209 0.0000*** AREAi - /? -0.105 0.0000*** -0.085 0.0000*** -0.077 0.0000*** -0.118 0.0000*** AREAj - /? -0.020 0.1889-0.064 0.0000*** -0.083 0.0000*** -0.069 0.0000*** LANGij + 0.107 0.0782* 0.161 0.0190** 0.020 0.0000*** 0.226 0.0001*** EXRi - -0.102 0.0000*** -0.071 0.0000*** -0.062 0.7522-0.058 0.0000*** EXRj + 0.054 0.0000*** 0.004 0.7652-0.044 0.0011** 0.098 0.0000*** TAXi - 0.002 0.2292-0.004 0.1576 0.002 0.3283 0.006 0.0009** TAXj - -0.034 0.0000*** -0.027 0.0000*** -0.006 0.0032** -0.027 0.0000*** APECi + 0.211 0.0000*** 0.204 0.0481** 0.317 0.0011** 0.368 0.0000*** APECj + 0.052 0.2330 0.016 0.8693 0.317 0.0010*** 0.228 0.0000*** APECij + 0.104 0.0308** 0.467 0.0001*** 0.427 0.0002*** 0.096 0.0828* Weighted Statistics R-squared 0.896 0.402 0.611 0.816 Adjusted R-squared 0.895 0.395 0.607 0.815 Prob(F-statistic) 0.000 0.000 0.000 0.000 DW stat 1.281 1.292 0.860 0.565 Cross-sections 240 240 240 240 Total pool observations 1440 1440 1440 4320 Dependent variable: IMPORT. Method: pooled random effects. *p<0.10, **p<0.05, ***p<0.001, significant at the 10%, 5% and 1% level, respectively. 27

Appendix 1 Information about Countries 11 Country Rank by GDP (IMF2007) Australia 15 Coal, gold, meat, wool, alumina, iron ore, wheat, machinery and transport equipment Canada 9 Motor vehicles and parts, industrial machinery, aircraft, telecommunications equipment, electronics China 4 Office machines, data Export Main Export Partners Import Main Import Partners processing equipment, Telecommunications equipment Japan 2 Cars, electronic devices and computers Korea 14 Electronic products, machinery and transport equipment Japan 20.3%, China 11.5%, Korea 7.9%, US 6.7%, New Zealand 6.5%, India 5% US 78.9%, UK 2.8%, China 2.1% US 19.1%, Hong Kong, 15.1%, Japan 8.4%, Korea 4.6%, Germany 4% US 20.4%, China 15.3%, Korea 7.6%, Taiwan 6.3%, Hong Kong 5.4% China 22%, US 12.5%, Japan 7.1% Machinery and transport equipment, computers and office machines, telecommunication lasers Machinery and equipment, motor vehicles and parts, electronics, crude oil, chemicals, electricity Electrical machinery, petroleum and related products, professional and scientific instruments Oil, foodstuffs, and wood Machinery, electronics equipment US 13.9%, China 13.7%, Japan 11%, Singapore 5.6%, Germany 5.6% US 54.1%, China 9.4%, Mexico 4.2% Japan 14%, Korea 10.9%, Taiwan 10.5%, US 7.3%, Germany 4.7% China 20.5%, US 11.6%, Saudi Arabia 5.7%, UAE 5.2%, Australia 5%, Korea 4.4% China 17.7, Japan 16%, US 10.7%, Saudi Arabia 5.9% 11 Refers to: http://www.cia.gov/library/publications/the-world-factbook 28

Country Rank by GDP (IMF2007) Malaysia 39 Electronic products, petroleum products, coconut oil, fruits New Zealand 51 Tourism destination, dairy products, meat, wood, fish and machinery Philippines 46 Semiconductors and electronic products, transport equipment, garments, copper products, petroleum products, coconut oil, fruits Singapore 45 Machinery and Export Main Export Partners Import Main Import Partners equipment (including electronics), consumer goods, chemicals, mineral fuels United States 1 Industrial supplies, production machinery and equipment US 15.6%, Singapore 14.6%, Japan 9.1%, China 8.8%, Thailand 5%, Hong Kong 4.6% Australia 19.6%, US 14.3%, Japan 11.4%, China 6. 3%, UK 5.1% US 17%, Japan 14.5%, Hong Kong 11.5%, China 11.4%, Singapore 6.2%, Malaysia 5%, Germany 4.3% Malaysia 12.9%, Hong Kong 10.5%, Indonesia 9.8%, China 9.7%, US 8.9%, Japan 4.8% Canada 20.1%, Mexico 11.7%, China 5.5%, Japan 5.1%, Germany 4.2%, UK 4.1% Machinery, electronics and electronic equipment, chemicals Machinery and equipment, vehicles and aircraft, petroleum, electronics Electronic products, mineral fuels, machinery and transport equipment, iron and steel, textile fabrics, grains, Machinery and equipment, mineral fuels, chemicals, foodstuffs Non-auto consumer goods, fuels, production machinery and equipment Japan 13%, China 12.9%, Singapore 11.5%, US 10.8%, Taiwan 5.7%, Thailand 5.3%, Korea 4.9% Germany 4.6% Australia 28.6%, Japan 10.7%, US 10%, China 6.6%, Germany 4.2%, Singapore 4.1% Japan 12.3%, US 14.1%, Singapore 11.2%, Taiwan 7.2%, Saudi Arabia 6.4%, Korea 5.9%, Malaysia 4.1% Malaysia 13.1%, US 12.5%, China 12.1%, Japan 8.2%, Taiwan 5.9%, Indonesia 5.6%, Korea 4.9% China 16.1%, Canada 16.0%, Mexico 10.3%, Japan 6.6%, Germany 4.7% 29

Country Rank by GDP (IMF2007) Brazil 10 Transport equipment, iron ore, soybeans, footwear, coffee, autos, automotive parts, machinery Egypt 52 Crude oil and petroleum products, cotton, textiles, metal products, chemicals France 6 Machinery and transportation equipment, aircraft, plastics, chemicals, iron and steel, beverages Germany 3 Machinery, vehicles, Export Main Export Partners Import Main Import Partners chemicals, metals and manufactures US 15.8%, Argentina 9.9%, China 7.9%, Netherlands 5.4%, Germany 4.7% US 9.7%, Italy 9,5%, Spain 7,6%, Syria 5.5%, Saudi Arabia 4.9%, UK 4.2% Germany 14.7%, Spain 9.6%, Italy 8.7%, UK 8.3%, US 7.2%, Belgium 7.1% France 9.7%, US 7.5%, UK 7.3%, Italy 6.7%, Netherlands 6.4%, Austria 5.4%, Belgium 5.3%, Spain 5% Machinery, electrical and transport equipment, chemical products, oil, automotive parts, electronics Machinery and equipment, foodstuffs, chemicals, wood products, fuels Machinery and equipment vehicles, crude oil, aircraft, plastics, chemicals Machinery, vehicles, chemicals, foodstuffs, textiles, metals US 11.9%, China 10.6%, Argentina 9.0%, Germany 7.5%, Nigeria 4.5%, Japan, 4.0% US 11.7%, China 9,7%, Italy 6.4%, Germany 6.3%, Saudi Arabia 4.7%, Russia 4.3% Germany 18.9%, Belgium, 10.7%, Italy 8,2%, Spain 7%, Netherlands 6.5%, UK 5.9%, US 5.1% Netherlands 12%, France 8,6%, Belgium 7.8%, China 6.2%, Italy 5.8%, UK 5.6%, US 4.5%, Austria 4.4% 30

Country Rank by GDP (IMF2007) India 12 Petroleum products, textile goods, gems and jewelry Export Main Export Partners US 15%, China 8.7%, UAE 8.7%, UK 4.4% United Kingdom 5 Machinery and transportation equipment, US 15%, Germany 11%, France 10%, Ireland 7%, Netherlands 6%, Belgium 6%, Italy 4% Import Main Import Partners Crude oil, China 10.6%, US machinery, gems, 7.8%, Germany fertilizer, chemicals 4.4%, Singapore 4.4% Machinery, Germany 14.2%, vehicles, chemicals US 8.6%, China 7.3%, Netherlands 7.3%, France 6.9%, Belgium 4.7%, Norway 4.7% 31

Appendix 2 (Import Data= Import (Measured by US dollars) / US GDP Deflator) Basic gravity model estimates - pooled regression analysis Variable Expected signs 1990-1995 1996-2001 2002-2007 1990-2007 C 4.157 0.0000*** 5.019 0.0000*** 3.652 0.0000*** 2.764 0.0000*** GDPi + 0.076 0.0000*** 0.149 0.0000*** 0.294 0.0000*** 0.220 0.0000*** GDPj + 0.084 0.0000*** 0.067 0.0003*** 0.208 0.0000*** 0.155 0.0000*** POPi + 0.158 0.0000*** 0.121 0.0002*** 0.033 0.2981-0.016 0.5974 POPj? 0.182 0.0000*** 0.196 0.0000*** 0.135 0.0000*** 0.173 0.0000*** DISTij - -0.187 0.0000*** -0.264 0.0000*** -0.289 0.0000*** -0.205 0.0000*** AREAi - /? -0.044 0.0117** -0.059 0.0008*** -0.053 0.0015** 0.004 0.7952 AREAj - /? -0.038 0.0283** -0.049 0.0050** -0.068 0.0000*** -0.056 0.0010*** LANGij + 0.176 0.0080** 0.234 0.0039** 0.180 0.0203** 0.302 0.0000*** EXRi - 0.111 0.0000*** -0.028 0.0671* -0.016 0.2902 0.233 0.0000*** EXRj + 0.078 0.0000*** 0.029 0.0541* -0.005 0.7185 0.129 0.0000*** TAXi - -0.032 0.0000*** -0.018 0.0000*** -0.000 0.9680-0.019 0.0000*** TAXj - -0.036 0.0000*** -0.033 0.0000*** -0.012 0.0000*** -0.027 0.0000*** Weighted Statistics R-squared 0.342 0.314 0.548 0.224 Adjusted R-squared 0.336 0.308 0.545 0.222 Prob(F-statistic) 0.000 0.000 0.000 0.000 DW stat 1.128 1.310 0.821 0.482 Cross-sections 240 240 240 240 Total pool observations 1440 1440 1440 4320 Dependent variable: IMPORT. Method: pooled random effects. *p<0.10, **p<0.05, ***p<0.001, significant at the 10%, 5% and 1% level, respectively. 32

Extended gravity model estimates on APEC - pooled regression analysis Variable Expected signs 1990-1995 1996-2001 2002-2007 1990-2007 C 4.110 0,0000*** 4.119 0.0000*** 2.930 0.0000*** 3.106 0.0000*** GDPi + 0.058 0.0001*** 0.173 0.0000*** 0.259 0.0000*** 0.160 0.0000*** GDPj + 0.072 0.0000*** 0.089 0.0000*** 0.190 0.0000*** 0.119 0.0000*** POPi + 0.181 0.0000*** 0.157 0.0000*** 0.120 0.0000*** 0.053 0.0574* POPj? 0.198 0.0000*** 0.225 0.0000*** 0.213 0.0000*** 0.206 0.0000*** DISTij - -0.182 0.0000** -0.211 0.0000*** -0.253 0.0000*** -0.216 0.0000*** AREAi - /? -0.048 0.0034** -0.079 0.0000*** -0.071 0.0000*** -0.007 0.6309 AREAj - /? -0.040 0.0144** -0.067 0.0000*** -0.087 0.0000*** -0.056 0.0001*** LANGij + 0.153 0.0152** 0.179 0.0089** 0.029 0.6493 0.235 0.0001*** EXRi - 0.093 0.0000*** -0.050 0.0005*** -0.061 0.0000*** 0.177 0.0000*** EXRj + 0.065 0.0000*** 0.004 0.7593-0.049 0.0003*** 0.096 0.0000*** TAXi - -0.031 0.0000*** -0.011 0.0000*** 0.003 0.1240-0.021 0.0000*** TAXj - -0.035 0.0000*** -0.027 0.0000*** -0.007 0.0015** -0.026 0.0000*** APECi + 0.060 0.1659 0.090 0.3806 0.302 0.0019** 0.116 0.0266** APECj + 0.063 0.1405 0.006 0.9518 0.314 0.0012** 0.225 0.0000*** APECij + 0.136 0.0044** 0.477 0.0001*** 0.444 0.0001*** 0.097 0.0906* Weighted Statistics R-squared 0.372 0.398 0.613 0.253 Adjusted R-squared 0.365 0.392 0.609 0.250 Prob(F-statistic) 0.000 0.000 0.000 0.000 DW stat 1.141 1.286 0.852 0.473 Cross-sections 240 240 240 240 Total pool observations 1440 1440 1440 4320 Dependent variable: IMPORT. Method: pooled random effects. *p<0.10, **p<0.05, ***p<0.001, significant at the 10%, 5% and 1% level, respectively. 33