_ ,. § . . 4 h... 11:73. in. r 3%, 34...“: . a :...Mv£....u.. u 3‘ 0 :33. fl .? 1! $2.: :7! If “:1? 3K litrdutwttii.‘ . .hfi: )1} :1 ,3... r. . b ‘L {e il‘ I. . , Elli»? 3:55.953» ESL? h: ~ :1. t. l. .L‘. 2. JEF- __3..13r . i3: llllljlllflflllfllllllfl'flllfllfllllIll 1 This is to certify that the dissertation entitled GRAVITY MODEL AND ECONOMIC INTEGRATION presented by Dongwook Han has been accepted towards fulfillment of the requirements for Ph.D. degree in Economics ; ,’/ "' / l ‘ I]?c”".“’."'llt”.“p-‘(7fi‘ ./ jzt 3H3 7.4.42.1, Major prol’ess‘or Date I??? 7; W? {)7 MSUisan‘” 1' . ' "1 '"r, ' ' 0-12771 V‘— —v~—fi'V—v—vV—fi——~._ fl“ A-,_‘,~- , _fi_-.~__v——. A ~ ‘r‘ ‘v—w—o—‘rj. f..._~_ WW4 \ ‘v—‘ —~,i LIBRARY Michigan State University ' PLACE IN RETURN Box to remove this checkout from your record. TO AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE "Y r _ nan MR .91' 1M WMJGS-p. 14 GRAVITY MODEL AND ECONOMIC INTEGRATION by Dongwook Han A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirement for the degree of DOCTOR OF PHILOSOPHY Department of Economics 1999 . Fr»- “.VQ4 o — I“Vr\y- ”‘b'vs ‘ 1 p v. n 0 Oh... V6 1‘ .v' ‘ (I) () D! Dr) (I) (I) .Lr ‘ i" (U ABSTRACT Gravity Model and Economic Integration By Dongwook Han Since the early 19805 there has been a proliferation of regional arrangements on a worldwide level. EC countries are moving to incorporate eastern Europe, while NAFTA.contemplates inclusion of South America. And APEC, the loose form of free trade area, is very active now. The objective of this dissertation is to identify a possible enlargement of NAFTA, to estimate trade creation and diversion of various trading blocs and to estimate the growth effect of exports expansion and regional integration. First, I examine the possible enlargement of NAFTA, and in which direction NAFTA should be enlarged so as to maximize welfare: south, incorporating Latin America (EAI, Enterprise for America Initiative); East, in a deal with Europe (TAFTA, Trans .Atlantic Free Trade Area); or West, (APEC). Second, by using a gravity model, trade creation and diversion of .ASEAN, EC, ANDEAN and MERCOSUR is estimated. Third, by using a growth accounting equation, I estimate the effect of exports expansion on growth. I find that a 1% growth in export causes a 0.22% increase in the real GDP. At a second stage, I estimate the growth effect of economic blocs by grafting the cross-sectional results of gravity model onto the growth accounting equation. In that fashion, important questions of international commercial policy and the dynamic effect of economic integration can be addressed. While most of the previous studies of regionalism focused on the static effects and excluded dynamic effects, this thesis identifies and quantifies the growth effects of regional integration. It also constitutes a first attempt to determine the “best” direction of possible enlargement in a way which would maximize welfare. ACKNOWLEDGEMENTS First of all, I would like to thank my dissertation chairman, Mordechai Kreinin for his excellent and dedicated supervision throughout the preparation of this dissertation. I am especially grateful for his kindness and intimate guidance. I wish to thank the other dissertation committee members Stephen Matusz, John D. Wilson for their helpful comments. My best thanks must go to my wife Kyunghwa and my beloved son Tae-Joon and daughter Ye-Jean for their encouragement and patience. I also grateful to my brother Myung-Wook and sister Eunjee for their help. Finally I give my thanks to my parents—in-law Jeoung- Yeon Kim and Soon-Mo Yoon for their material and moral help. iv TABLE OF CONTENTS LIST OF FIGURES . . . . . . . . . . . . . . . . . . Vii LIST OF TABLES . . . . . . . . . . . . . . . . . . viii LIST OF APPENDICES . . . . . . . . . . . . . . . . ix CHAPTER 1. INTRODUCTION . . . . . . . . . . . . . 1 CHAPTER 2. THE WELFARE ECONOMICS OF TRADE BLOC . . I. Revival of Regionalism, Why? . . . . . II. Static Effects of Trading Bloc . . . . 10 III. Welfare Implications of Natural Economic Bloc . . . . . . . . . . . . . . . . . . 20 IV. Dynamic Issues of Trade Bloc . . . . . . 26 CHAPTER 3. GRAVITY MODEL APPLICATION: REVIEW . . . . . 31 I. What is gravity model? . . . . . . . . . . 31 II. Theoretical foundations of Gravity Model . 34 III. Variables of the gravity model . . . . . 37 IV. Past Use of the Gravity Model: A Review . .41 CHAPTER 4. POSSIBLE ENLARGEMENT OF TRADING BLOC: NAFTA, WHAT IS THE NEXT STEP? . . . . . . . 59 I. Introduction . . . . . . . . . . . . . . 59 II. Model Specification . . . . . . . . . . 60 III. Data . . . . . . . . . . . . . . . . . . 62 IV. Empirical Results . . . . . . . . . . . 63 V. Summary and Conclusion . . . . . . . . . 68 CHAPTER 5. TRADE CREATION AND TRADE DIVERSION OF ASEAN, EC, ANDEAN AND MERCOSUR: BY GRAVITY MODEL 69 I. Introduction . . . . . . . . . . . . . . 69 II. Methodologies for Estimating Trade Creation and Trade Diversion . . . . . . . . . . 70 III. Estimates of Trade Creation and Trade Diversion . . . . . . . . . . . . . 73 IV. Summary and Conclusion .83 CHAPTER 6. GROWTH EFFECTS OF EXPORTS EXPANSION . . . 84 I. Introduction . . . . . . . 84 II. Model Specification . . . . . . . . 87 III. Data . . . . . . . . . . . 90 IV. Empirical Results . . . . . . . 91 V. Summary and Conclusion . . . 92 CHAPTER 7. GROWTH EFFECTS OF REGIONAL INTEGRATION . 93 I. Introduction . . . . . . . 93 II. Model Specification . . . . 95 III. Data . . . . . . . . . . 96 IV. Empirical Results . . . . . . . . . 98 V. Summary and Conclusion . . . 110 APPENDICES . . . . . . . . . . . . . . . . 112 BIBLIOGRAPHY vi 193 LIST OF FIGURES Static Effect of Trade Union (From Kreinin (1998) . . . . . . . . . . . l6 Welfare Effects of Customs Union . . . . . . . . 19 World Welfare and Number of Blocs . . . . . . 23 Natural Economic Bloc and Gravity Model . . . I 62 vii LIST OF TABLES Trading Neighbors: Ratio of Share of Trade to Partner's Share of World Output,1989 . . . . . . 25 Empirical Results of Previous Studies . . . . . 54 List of variables used in previous gravity model studies . . . . . . . . . . . . . . . . . . . . . 58 Empirical Results of Gravity Model . . . . . . . 66 Who is the best partner of NAFTA? . . . . . . . . 67 Cross-Sectional Results (11) form Model (5.1) . . 74 GTC, TD and TC of ASEAN, EC, ANDEAN, and MERCOSUR .76 Summary of Various Exports-Growth Model Formulations and Estimations . . . . . . . . . . . . . . . . . .90 Export-Growth Regression by Continents . . . . . 92 Cross-Sectional Results (35) form Gravity Model . .99 Bloc-Growth Regression Eq.(7.4) . . . . . . . . 100 Bloc-Growth Regression Eq.(7.5) . . . . . . . . .102 Bloc-Growth Regression Eq.(7.6) . . . . . . . . .103 Bloc-Growth Regression Eq.(7.7) . . . . . . . . .104 Cross-Sectional Results (B5) form Gravity Model . 106 Growth Effects of Regional Groupings . . . . . . 110 viii Appendix Appendix Appendix Appendix LIST OF APPENDICES Countries involved in Gravity Model . . .112 Trade Creation and Diversion of Trading Blocs . . . . . . . . . . . . . . . . . .114 Countries involved in Chapters 5 and 7. .187 D : Export-Growth Regression by Countries . .188 ix CHAPTER 1 INTRODUCTION Since the early 19805 there has been a proliferation of regional arrangements in the world. According to Bhagwati (1992) regionalism is not only back, but it is here to stay. Lester Thurow (1992) proclaimed that "The GATT is dead" and argued that the world will shift to a tripolar system with three blocs centered on Europe, the United States, and Japan which will each have free trade internally and managed trade externally. In any event the regional movement is here to stay and it is expanding. EU countries are moving to incorporate eastern Europe, while NAFTA contemplates inclusion of South America. There are two extreme views among economists concerning the regional movement. Those who favor the regional approach maintain that the world is likely to move toward global free trade far more rapidly if the number of negotiating countries is reduced via bloc formation, and that forming regional blocs is a quicker and more certain way of reaching multilateral agreements. They also maintain that regional blocs are more capable 0' ..o— . . . I _ _. . I: ... . . Vi. . ~ I s v‘ . . . . . t .: r». L» o . v c .3 vs .i« s ‘ Q~ v. .3 .o ”c. r. .3 .. . v. 2. Ct at 2. ex na. 2. . I =. .P- a a .u- .u. we. nu . a .6 ~ .1“ ‘5 Q.» a e Aves . § \N§ vTI K-& .7— a v v; Pd fie. ‘5 Q» A“ Ali of “deeper” integration than the multilateral system. But those who oppose regionalism argue that large blocks have greater market power and will impose higher tariffs on each other. So the blocks would tend to be more inward— looking and thus harmful for the multilateral trading system. The objective of this dissertation is to identify a possible enlargement of NAFTA, to estimate trade creation and diversion of various trading blocs, and to estimate the growth effect of regional integration. First, by using the natural economic bloc concept and gravity model with new dummy variables which represents various combination of blocs, the possible enlargement of NAFTA is identified. I will examine which of the following alternatives to NAFTA enlargement will maximize welfare: a southward expansion incorporating Latin America (EAI, Enterprise for America Initiative); eastward in a deal with Europe (TAFTA, Trans Atlantic Free Trade Area); or westward (APEC).1 Second, by using gravity model trade creation and diversion of ASEAN, EC, ANDEAN and MERCOSUR is estimated. Third, by using a gravity model with growth accounting equation the effect of economic bloc on growth is estimated. In that fashion, important questions of international commercial policy 1See the appendix A for the country grouping. up, v--~ Rfiv it... ~-, i.~ ,- L-u., .A 5‘ ‘ rev; "WA. on: _ a.“ \ Kr‘ 9‘ ‘l and the dynamic effect of economic integration can be addressed. While most of the previous studies of regionalism focused on the static effects and excluded dynamic effects, this thesis will identify and quantify the growth effects of regional integration. It also constitutes a first attempt to determine the “best” direction of possible enlargement in the way which would maximize welfare. Chapter 2 develops the analytical framework of economic integration. Theoretical considerations needed to evaluate the economic effect of trading bloc are reviewed including the static and dynamic effects of custom union and natural economic bloc model. The gravity model, which will be used in empirical analysis, will be reviewed in Chapter 3. Theoretical foundations and various applications of the model in international trade area are discussed. The empirical analysis which is the center-piece of this dissertation is presented in Chapters 4, 5, 6 and 7. Chapter 4 attempts to identify the possible enlargement scenarios of regional grouping. Using a gravity model and natural bloc argument the direction of NAFTA is identified with the “best" partners of NAFTA to maximize welfare. While chapter 4 focuses on NAFTA and the broad continent-sized groupings that are under discussion (the Americas, Europe and East Asia) Chapter 5 analyzes the effect of trade creation and trade diversion of ASEAN, EC, ANDEAN and MERCOSUR, trading blocs that are already in existence. While Chapters 4 and 5 is based on the static effects of regional integration, Chapter 6 and 7 focus on the dynamic effect of exports expansion and regional groupings. By using a gravity model with growth accounting equations, the effect of regional integration (ASEAN, ANDEAN, and EC) on the growth rate is estimated. In addition, the growth effect of the broad continent- sized groupings (the Americas, Europe, and East Asia) will be analyzed. CHAPTER 2 The‘Welfare Economics of Trading Blocs What are the motives for the current revival of regionalism? How is the regional movement likely to impact the welfare of the world, and individual participants? Section I reviews the motives for the formation of regional blocs. Section II analyzes the static effects of custom union while Section III focuses the welfare implications of natural economic blocs. The dynamic effect of economic integration is analyzed at section IV. I. Revival of Regionalismw‘Why? How do we explain the current revival of regionalism around the world? Jagdish Bhagwati (1993) argues that the single most important reason why regionalism is making a comeback is the conversion of the U.S. approach from multilateralism to regionalism or bilateralism. Disappointed by a lack of progress at the GATT(now the WTO)1:negotiations, the United States has \ 1 On January 1, 1995 GATT was expanded and made into a formal World Trade Organization (WTO). decided to switch course and to conclude first the Canada-US Free Trade Agreement (CUSTA) and then NAFTA. The United States has also announced its intention to negotiate free-trade agreements with groups of other Latin American countries under the EAI(Enterprise for the Americas Initiative)2. Concurrently, the European Community has continued to widen and deepen its integration process. These developments have, in turn, led other countries to consider the regional option. East Asia, in particular, is coming to believe that a regional bloc may be the only way to meet the challenge posed by developments in the Americas and Europe. A key reason for the United States’ conversion to regionalism in the early 1990's was the slow progress at the GATT. Krugman (1991b) offers several reasons for the erosion of the GATT process. First, the decline in US leadership has made it more difficult to run the system. With the US accounting for a progressively smaller share of gross world product, and with US dominance in productivity and technology progressively declining, it has been losing both the means and the desire to serve as the global trade hegemon. Second, the number of players participating in the process has grown large, making negotiations difficult and the free rider problem harder to handle. Third, the character of 2 See Appendix A for the countries involved in EAI. J‘- una ‘ ,- .0- r! I" ‘7‘?“ .‘a‘w ' ~“’, g ‘ .y‘.‘ .._ . 3" no“ a.’ ‘-.= v...“ ’I‘ . l kw: \ protectionism has changed. The presence of VERs, anti- dumping(AD) mechanisms, and other forms of administered protection made the negotiations vastly more complicated than it had been in the past. Factors beyond the Uruguay Round also appear to have played an important role in the trend toward regionalism. Regional trading arrangements are pursued for a variety of reasons that may differ across groupings and across participating countries within a given bloc. Kreinin (1998) outlines the motives for the formation of regional blocs. “These motives are many and varied and include the producers’ anticipation of trade diversion benefits (with little resistance from consumers); the expectation that a larger market, which facilitates scale economies, will contribute to greater productivity and thus enhance their competitive position; a frustration with the slow progress of global trade liberalization, and perception by some countries that they fail to benefit from the WTO process; the desire to enhance the nations' bargaining power in the WTO, and for LDCs to reduce dependence on the markets of industrial countries; and the possible hope of moving toward greater political cohesion.” The prospect of enhanced economic growth (stemming from the opportunity to exploit scale s A u —n~ A... i‘ llli C. .c n.“ ,s‘ . § r“ a- “V «an M‘ .3 1 on L. ... 3. v. Av .. VI. .3 . . an a. v. .. .. g“ T. .. A. C. v. a“ L. C. a. no .1 .2 a u Au 5 v \V. O o p. Q .Fu qt‘u a!“ A. i K a a ‘ § § \ ~\~ a: a a . economies and regional specialization, as well as attracting investment by expanding the regional market) is a motivation present in virtually every regional trading arrangement, in both industrial and developing countries. The realization of scale effects was a major consideration underlying the Unified Market Program in the EU. It is also an explicit goal of AFTA? and MERCOSUR“. They aim to exploit scale economies, deepen the division of labor within the region, and attract foreign direct investment (FDI) by presenting the region as a stable and prosperous single market. The dynamic growth effects expected by Mexico, especially the anticipated surge in FDI, were also a key motivation for Mexican interest in NAFTA. Second, regional initiatives may be viewed as a means to promote a broad range of noneconomic objectives, from enhancing regional political cohesion to various foreign policy considerations, such as managing immigration flows and promoting regional security. The formation of the EU had strong political 3 ASEAN (Association of Southeast Asian Nations) was formed in 1967 to promote economic, social, and cultural cooperation among Indonesia, Malaysia, Philippines, Singapore, and Thailand. A series of talks beginning in 1992 led to the decision to create the PETA“ The FTA is to be achieved in 2003. The goal is to reach a zero to 5 percent preferential tariff on manufactured goods by 2003. ‘ MERCOSUR (Southern Cone Common Market) is a common market involving Argentina, Brazil, Paraguay and Uruguay. It is the most significant regional trade bloc in Latin America. Its goals are common market, coordination of fiscal and exchange rate policy, and accelerating economic development. roots, as did the formation of the Association of South East Asian Nations (ASEAN). The desire of a number of EFTA countries to join the EU was also motivated, in part, by noneconomic objectives. MERCOSUR is perceived as a means of fostering cooperation between its member states. The promotion of political and economic stabilization and control of immigration flows were also important elements underlying both NAFTA and the association agreements of the EU with Eastern European countries. Lastly, there may exist some kind of domino effects. According to Baldwin(l993), as new regional trading arrangements form, or existing ones expand or deepen, the opportunity cost of remaining outside an arrangement rises. Nonmember exporters could experience costly reductions in market shares if trade is diverted to members. This may be sufficient in some countries to tip the political balance in favor of accession, as exporting interests begin to dominate import-competing interests. In turn, as new members join the arrangement, trade diversion from other outsiders may lead to a second round of accessions. The domino effect, or the anticipation of such, appears to have been prominent in the initiative of EFTA countries to apply for accession to the EU. East European countries were similarly .‘Y n can‘- in. ... .- int. I) .- .lv. .'. can. An, bdth "v... ‘ut ‘r. n. C. 10 interested in improving access to Western European markets and in not being left out of the emerging Unified Internal Market. The negotiations between Mexico and the United States to form a free trade area (FTA) may have started a comparable process in the Western Hemisphere. Canada's interest in NAFTA was strongly influenced by the potential erosion of the benefits expected from the CUSFTA, were it not to join the newly emerging NAFTA. In a similar vein, the large number of bilateral trade arrangements between Mexico and several Latin American countries is viewed by Mexico's partners as a first step toward joining NAFTA. Access to NAFTA has become an important objective of many Latin American countries as a way to correct the expected trade and investment diversion toward Mexico. II. Static Effects of Trading Blocs There are two basic forms of economic integration. First, a customs union (CU) involves two or more countries that eliminate tariffs among member countries and impose a Common External Tariff (CET) against outsiders. The EU (European Union) is a customs union. Second, free-trade area (FTA) eliminates tariffs on imports from member countries. But there is no CET against nonmember countries; each country is free to AV it- 07' b- U C: 1"" fiv“ a- s.“ OH“ -NU . 3" bass ~v- -Q. i V Q is x. ‘ evil C. a .: a: .9. ; ‘ x t he. a . A c in F.“- ll impose its own tariffs. Examples of FTA are NAFTA, EFTA, or US-Israel. In any event, whatever the ultimate goal of a regional arrangement, increased intra-regional trade is given a high priority. A” Pre4Vinerain‘View Prior to the publication of the Customs Union Issue by J. Viner (1949) it was thought that since free trade is the optimal condition any movement towards it improves welfare. Hence preferential trading arrangements (PTA) necessarily constitute an improvement in global welfare. B. Viner : Trade Creation vs. Trade Diversion Jacob Viner identified two of the static welfare effects of economic integration, trade creation and trade diversion. Trade creation, the static welfare gain of economic integration, results from substitution of partner country imports for domestic output as the tariff declines to zero. It is favorable because it causes a more efficient allocation of resources. Trade diversion, the static welfare loss, is the substitution of non-partner country imports by imports from the partner country because of discrimination against the (I) -v- 9“ u, .. loav- -0. u... b..- p A. ‘n MP V “‘U u (I) 12 former. It is unfavorable because it causes a less efficient allocation of economic resources. In fact, this concept was debated over NAFTA. Suppose that starting with a non-discriminatory tariff on all trading partners, the United States forms an FTA with Mexico. Suppose further that shoes are produced under constant costs everywhere and that the FTA results in the United States importing shoes from Mexico. Is. this a change for the better or worse? The answer, reasoned Viner, depends on who is the pre-FTA supplier of shoes. If the United States produced its own shoes in the initial equilibrium, it must have produced at a higher cost than Mexico. In this case, the FTA shifts shoe production from a higher to a lower-cost source and is trade creating: welfare of the union and of the world rises. If, on the other hand, the United States initially imported shoes from a third country, say, the Republic of Korea, that country must be a lower cost producer of shoes than Mexico. In this case, the FTA causes shoe production to shift from a lower to a higher-cost source. There is trade diversion and the welfare of the union and the world declines. In this example, trade creation occurred by the replacement of some high-cost domestic production by imports from the partner country. It is favorable to 13 world welfare because it rationally reorganizes production within the FTA. By contrast, trade diversion occurred by increased trade within the FTA at the expense of trade with the rest of the world and it is unfavorable because it reorganizes world production less efficiently. Figure 2.1, reproduced from Mordechai E. Kreinin, International Economics: A Policy Approach, is a partial equilibrium model of trade creation and trade diversion that occur with economic integration. Before economic integration, domestic producers satisfy demand if the price of producing the commodity domestically is less than the price of importing it and paying the exporting country’s production cost plus the non- discriminatory tariff. This may result in a misallocation of resources if the exporting country is a lower resource cost producer. If the commodity is imported, the decision to purchase imports form Country B or country C is determined by relative prices. Before economic integration, when imports from countries B and C are subject to the same ad valorem tariff, the lowest resource cost producer supplies the domestic country A. After economic integration, the price of imports from the partner country falls relative to the price of domestic output and relative to the price of imports 14 from the non-partner. In the domestic country, there is substitution away from domestic output towards imports from the partner country (trade creation) or substitution away from imports from the non-partner country and towards imports from the partner country (trade diversion). To illustrate Viner’s concept of trade creation and trade diversion, let Q be a homogeneous commodity produced in countries A, B, C. Assume that Country A is the highest cost producer of Q, country B is the intermediate cost producer and country C is the lowest cost producer. Therefore, if the cost of producing Q domestically is greater than the price of importing Q from country C, and if countries B and C are subject to the same ad valorem tariff on Q imposed by country A, then country C will supply Q to the country.A. Economic integration between countries A and B reduces the price of Q imported from country B below the price of Q imported from country C ( the most efficient producer, who is still subject to the tariff ). Consumers substitute Q produced by the partner country B, for Q produced by country C. Resources are allocated less efficiently. Trade diversion occurs—the static cost of economic integration. Also eliminating the tariff between country A and B lowers the price of Q in country do '7‘ ‘O U. "r o- A. 15 A; consumers substitute away from Q produced domestically and towards the commodity Q produced by, and imported from, country B. Trade creation occurs as producers in B increase production of Q, and producers in A reduce its production. According to this analysis, economic integration is either trade-creating or trade diverting. If the value of trade created exceeds that of trade diverted, then economic integration is welfare improving. If the opposite is true, economic integration is welfare-decreasing. The welfare gain (loss) results form changes in economic efficiency as resources are reallocated away from less (more) efficient producer to more (less) efficient producers. A customs union causes losses when it leads to net "trade diversion" that is, instead of specializing more and increasing efficiency, countries that form a trading bloc may substitute each others' more expensive goods for goods from outside the bloc, leading to a loss of efficiency. The essential message of the Vinerian approach was that PTAfs, as distinct from nondiscriminatory trade liberalization, could harm both a member country and world welfare. Whether overall welfare rises or declines becomes empirical and depends on whether TC>TD or TD>TC. 16 Figure 2-1 Static Effect of Trade Union (From Kreinin (1998)) P Smmw Pl §\ 0 Q1 Q2 0 under free trade: QlQZ imports from most efficient country C 0 under uniform tariff (P1P3) : tT imports from most efficient C 0 ‘under a customs union: bB imports from country B 0 trade diversion(tT) : elimination of imports from C 0 trade creation (br): elimination of inefficient domestic production 0 favorable consumption effect (qB) It is a standard result of international trade theory that free trade in commodities maximizes world welfare in distortion-free world. However a world ridden by multiple distortions (e.g. tariffs, quotas and exchange control) will not necessarily be moved closer to Pareto Optimality by the removal of one distortion. Counter-intuitively, we may move away from Pareto '( (fl QA‘,‘ V». . . b a.. - vu‘ D fine A \~. V» ‘h ‘k. . v l7 Optimality. Regional integration which involves a partial movement toward free trade is not necessarily a Pareto improvement. This is the fundamental proposition of the General Theory of Second Best. C. Partial Equilibrium.Mbdel of Trade Creation and Trade Diversion Figure 2.2 Welfare Effects of Customs Union, reproduced from Mordechai E. Kreinin, International Economics: A Policy Approach, is a static, partial equilibrium model of trade creation and trade diversion that occur with economic integration. It is a one commodity, three country model. In figure 2.2 we assume that Sa and Da are the internal supply and demand curves in country A.for a given product. Sb and So are the export-supply curves of countries B and C to country A, with C being a more efficient producer than B. Sbt and Set are the same two supply curves subject to a 100 percent tariff imposed by country A. Curve St indicates total supply curve of the commodity in country A (Sa+Sbt+Sct). Prior to integration Price P1 is established. Country A produces Qa domestically and imports Qb and QC from countries B and C, respectively. When countries A and B form a customs union to the exclusion of C, the relevant supply curve in country B becomes Sb, while Sct remains in effect in country C. 18 Total supply in country A’s market becomes Scu, consisting of Sa+Sb+Sct. The price in country A.drops to P2; domestic supply declines to Qal; imports from country B rise to le; and impodrts from country C diminish to ch. These changes can be quantified in terms of their effect on producers' surpluses in all three countries, and on consumers' surpluses and government tariff revenue of country A. The following observations relate to each panel of the diagram: (a) Country.A enjoys an increase of consumers' surplus of BAP1P2 and suffers a reduction of producers' surpluses amounting to BAYZ. There is a gain of ZYP1P2. (b) Country B enjoys a gain in producers' surplus of HGCD. Country A.faces a loss in government tariff revenue of CDEF. Since area DCHI is common to both, we obtain in part (b) a loss of EHIF and again of ICG. (c) Tariff revenue of country .A decline from RPLM to NOKJ. Subtracting the area NOSM, common to both, we get a loss of RNOP+POSL and a gain of MUKS..At the same time, producers' surpluses of country C decline by.MUKL. This figure yield the following loss: RNOP+POSLfMUKS+MUKL=RNOP+POSL+LSK=RNOP+POKL. 19 .Area ZYP1P2 [ the gain in (a) ] is equal by construction to areas EFGH in (b) plus RNOP in (c). Subtracting from this net gain in (a) the losses EHIF in (b) and RNOP in (c), we are left with a net gain FIG in (b). Adding it to the earlier gain CIG, we obtain a net gain of CFG in part (b), to be weighed against the net loss of POKL in part (c). Figure 2-2 : Welfare Effects of Customs Union P ‘ P 1 : E Q - E : QalQaZ Qb le ch QC on a» (Q The net effect on world welfare depends on the relative size of the two areas. A.priori it cannot be determined whether the gain exceeds the loss or vice versa. Empirical estimates are necessary to determine the trade creation and trade diversion effects of economic integration and therefore, whether economic integration yields a net welfare gain. O. A. a r -e. [In (I) .h -. w.. y -A 5v A,“ 5" u‘~ "‘ v Nu 'v ‘-t‘~ . 20 III.‘Welfare Implications of Natural Economic Bloc The welfare effect of a trade bloc is assessed in terms of trade creation and trade diversion. Because measuring trade creation and diversion is an arduous task, economists developed general economic criteria for assessing whether a regional grouping is likely to result in net trade creation. For example, the larger the economic area encompassed by a bloc, and the greater the share of intra-bloc trade in their total trade prior to integration, the more likely it is to be trade creating — the extreme case being an FTA that encompasses the entire world. Partly because of its simplicity, the most widely used rule of thumb is the share of intra-regional trade in the bloc's total trade prior to integration. Krugman and Summers (Economist 1991) have argued that if the share of intra-regional trade exceeds roughly one half of the region's total trade it can be regarded as a natural bloc. Kreinin and Plummer(1994) developed an alternative, more sophisticated, formulation: “In contrast with a measure which focuses on the volumes or value of trade, the alternative approach focuses on the pattern of trade. In terms of economic efficiency it is beneficial for a country to join a regional bloc if such a step would not 21 greatly distort its comparative advantage.” They develop a method for assessing the extent to which comparative advantage would be preserved if and when a country integrates with others. If such preservation occurs in all member of a proposed FTA, then the grouping can be labeled a ‘natural’ bloc. Krugman(1991a) analyzes the welfare consequences of the number of trading blocs in the world economy, whereby each block imposes its optimal tariff and maintains free trade within the bloc. The interesting result is that minimum welfare obtains when the number of blocks equals three. In the symmetric Nash equilibrium of his policy game, the result is necessarily Pareto efficient when the world consists of a single trade block because this situation amounts to free trade. Consequently, starting with one block, initial increases in the number of blocks reduce welfare. On the other hand, if the world consists of many blocks, the monopoly power of each one becomes smaller and so further increase in the number of blocks will raise welfare. Krugman uses a highly stylized model of differentiated products in which each bloc imposes optimal tariff. He finds, by simulation, that world welfare declines with customs union formation until the 22 number of blocs reaches three. Beyond this point CU formation is found to be welfare improving as in figure 2-3. This figure also shows a startling result: for the full range of elasticities considered, world welfare is minimized when there are three blocs. Why does he get this result and is it really plausible? Krugman argues that at one extreme, with as many blocs as there are countries, each bloc is too small to have any market power. Therefore, competitive behavior maximizes world welfare. At the other extreme, with one trading bloc, at the free trade welfare is maximized. In between, welfare is lower. Starting with one bloc, if we divide the world into two blocs of equal size, each bloc exercises monopoly power over its products and imposes a optimal tariff on imports from its rival. There is trade diversion and each bloc suffers a loss of welfare. Next suppose we divide the world into three equal blocs. This leads to only one-third rather than half of the goods being subject to free trade and there is further trade diversion. But the reduced size of each bloc also reduces its market power and the optimal tariff declines. This generates trade-creation effect. With both trade diversion and trade creation taking place simultaneously, welfare may rise or fall. As the number of blocs rises, the optimal tariff continues to 23 Figure 2-3 World Welfare and Number of Blocs From Krugman (1991a). \Vdfiue Elasticity of Substitution =10 Elasticity of Substitution =4 Elasticity of Substitution =2 1 2 3 4 10 Number of Blocs decline and at some point must become sufficiently small to yield a larger trade-creation than the trade- diversion effect. The critical question is then the number of blocs at which this turning point obtains. Surprisingly, Krugman finds that for a variety of parameters, the number of blocs for which a declining welfare begins to rise again is three. He contends that the three-possible bloc world based on America, Europe and Asia will be harmful for the welfare of the world. r 1 11‘ 24 In reality, however, Krugman (1991b) claims that the sets of countries (Europe and CUSFTA among G7 countries) that are now engaging in free trade agreements are indeed natural trading partners, who would have done much of their trade with one another even in the absence of special arrangements. If trading arrangements follow the lines of natural trading regions, they will have a much better chance of improving welfare. Because intra-trade among countries of these regions is already substantial, the trade creation effects of regional integration are likely to dominate the trade diverting effect. Summers (1991) claims that while global liberalization may be best, regional liberalization is very likely to improve welfare because given the existing structure of trade, plausible regional arrangements are likely to have trade creating effects that exceed their trade diverting effects. The issue of natural trading blocs is crucial because to the extent that blocs are created between countries that already trade disproportionately with each other, the risk of large amounts of trade diversion is reduced. Table 2-1 by Summmers shows the importance of natural trading blocs. By dividing the ratio (the trade volume of two trading countries/world trade) with the ratio (Partner’s I (I) 25 GNP/world output) it compares the ratio of observed trade to the trade one would expect if it were equiproportional to GDP. For example, in the table 2-1, the number 6.06 between the US and Canada indicate that the US does 6 times more trade with Canada than is suggested by Canada’s share of world output. Summers claims that existing trade arrangements link the nations that are already natural trading partners. From the disproportionate share of U.S. trade with Canada, of trade within developing Asian , and of trade within the industrialized Europe, he concludes that the CUSFTA, EC, and ASEAN can be regarded as a natural bloc. Table 2-1: Trading Neighbors: Ratio of Share of Trade to Partner’s Share of World Output,1989 __I@er--W_.1V_itfir__-__ _- W W --_l_--_----__--_ W- -_ .. United Canada Other Japan Developing EC ERMBS Anmfiam Ash United States --- 6.06 2.38 0.87 2.34 0.61 Canada 2.63 0.66 0.47 0.97 0.39 Other 1.13 0.63 3.16 0.31 0.57 0.67 Amwfiam Japan 0.95 1.15 0.75 --- 4.33 0.53 Developing 0.73 0.62 0.43 1.26 4.83 0.54 Asia EC 0.22 0.30 0.42 0.17 0.63 1.75 Ikmm:&mmmms(fl¥fl) According to the above argument, a natural bloc is a welfare improving bloc. Using the natural bloc 26 concept, I will try to measure the degree or strength of natural trading blocs by the coefficient of dummies representing FTA or customs union in a gravity model instead of the intra-trade shares or the ratio of trade to GDP as done by Summers(199l). And by comparing these coefficients, I can identify what will be the desirable enlargement of NAFTA. This will be addressed in chapter 4. IV.Dynamic Issues of Trade Blocs ANDynamic or Growth Effects of Customs Union In considering the implications of a customs union, the static effects concern mainly changes in resource allocation and consist of trade creation, trade diversion, and the terms of trade effects. By contrast the dynamic effects are the long-run consequences of increased market size for the growth rate of the integrating region. Trade liberalization may also give rise to effects that produce a sustained increase in economic growth through information transfers, increased competition, accelerated technological change and the perception of improved investment opportunities. These effects are occasionally cited among the reasons for pursuing regional trading arrangements. To the extent that a grouping stimulates regional growth, 27 it may offset the static trade diversion effects on non- members and produce an expansion of trade both inside and outside the grouping. There is, however, the additional risk for outsiders that improved investment opportunities, combined with restrictive or nontransparent rules of origin, or both, may divert direct investment flows from non-members. This was the main concern in Asia over NAFTA. This effect is likely to be less significant from a worldwide perspective if a regional grouping maintains relatively low MFN tariffs or the grouping is economically small. Further, the stronger the conviction that multilateral trade liberalization will proceed apace, the less the incentive to alter longer-term investment plans in response to current regional trading arrangements. The dynamic effects of economic integration are reflected in long run changes in the level of gross national income and output. However there are immense difficulties in assessing the impact of integration on growth, arising from the fact that a multitude of factors influence the growth rate, and it is not easy to isolate the effect of integration. They will be addressed in chapter 6. «i~ y . ‘u. .F 11' ‘v. Q ‘v ‘0 v v ‘\ 28 B. Dynamic Time-Path Issue Equally important to the dynamic effects of customs union is the dynamic time-path question: whether the effect of customs union is to accelerate or decelerate the continued reduction of trade barriers toward the goal of reducing them worldwide. Bhagwati(1992,1993) contends that even if a particular regional scheme moves the world towards freer trade, over time it may result in a more protectionist world by imposing higher tariffs with greater market power. From the viewpoint of the world trading system, more critical than static effects are the dynamic time-path implications of the regional approach. Here the disagreements between economists are even deeper. Summers (1991) argues that the world is likely to move toward global free trade far more rapidly if the number of negotiating parties is reduced to three via bloc formation. Three parties with a lot to gain from a successful negotiation are more likely to complete it than are many parties, each with only a small amount to gain. It may well be that a smaller number of trade blocs are more likely to reach agreement than a larger number of separate countries. He doubts that the existence of the EC has complicated the process of reaching multilateral trade agreements. Instead, the 29 ability of Europe to speak with a more common voice would have helped, not hurt, over time. The flip side of the argument, however, is that large blocs have greater market power and, in the absence of cooperation, may impose higher tariffs on each other. Taking this latter view, Bhagwati(1993) notes that larger countries often tend to be more inward-looking than smaller countries. Once a bloc is large enough, the need to be open to extra-bloc countries is reduced. Bhagwati is also skeptical of the argument, made by regionalists, that the regional approach is quicker and more certain. As for speed, even the best example of regionalism, the European Community, started four decades ago (1957) and is still incomplete. The transition has not been instantaneous any more than the negotiated reductions of trade barriers under GATT; and this despite the enormous political support for a united Europe. Observe agriculture. The record of regional trade blocs dealing with agricultural trade liberalization is dismal; the CAP is not exactly the European community’s crowning achievement. In fact, if it were not for multilateralism (i.e., the Uruguay Round), it is difficult to imagine that the process of unraveling the CAP could even have begun. 30 Depending on the relative power of different interest groups, trading blocs may turn inward over time. Interest groups within the bloc may take the View that the bloc’s markets belong to them and resist extra- bloc liberalization. CHAPTER 3 Gravity Mbdel Application: Review I. ‘What is the gravity model? A gravity model is rooted in Newtonian physics that investigates the universal force of attraction which affects all matter. Issac Newton held that every particle of matter in the universe attracts every other particle with a force that is proportional to the product of their masses and inversely proportional to the square of the distance between them. Mathematically this is expressed by the classical formula Gmm (3.1) F=_—dlz—2—, where F is the force of attraction, m are the masses, d is the distance, and G is a universal gravitational constant. It is interesting that this gravity model has been applied to a wide variety of goods and factors of production moving across regional and national boundaries 3] 32 under different circumstances. The model has been successfully applied to flows of varying types, such as migration, flows of buyers to shopping centers, recreational traffic, commuting, patient flows to hospitals, telephone calls and interregional as well as international trade.1 For the international trade flow, the gravity model states that the size of trade flows between two countries is determined by supply conditions at the origin, by demand conditions at the destination, and by stimulating or restraining forces relating to the specific flows between the two countries. So the size of trade flows between countries is positively related to GDP and negatively to the distance. The basic form of the gravity model in international trade flows is: (GDP. )" (01313)” T = c (3.2) g 1%, a where Tij = value of trade (exports+imports) between country i and country j c = constant GDPi, GDPj = income in the exporting and importing 1 Gravity models are much in use in geographical analysis and regional science. For the various applications of gravity model, see Hua and Porell (1979) and citations thereof. 33 countries Dij = distance between countries 1 and j finflmf=parameters of gravity equation to be estimated. In View of the similarity between this equation and the law of gravity in physics, models of this sort have come to be called "gravity models.” Taking the log form to estimate and adding dummies to estimate trade bloc effect, we have log(7;.) = a + ,81 log(GDR) + ,62 log(GDPj) + ,6310g(N,-) ( 3 . 3) + A log(Nj) + Alog(DISTANCE.y.) + mAwACENCig.) + y,(NAFTAy.) + y2(EC.g.) + 73(EAy.) + at]. where Ni, Nj = population of exporting and importing countries ADJACENCY = adjacent dummy variable which takes the value of 1 if both countries 1 and j are adjacent and 0 otherwise NAFTA,EC,EA(East Asia)= Dummies for country grouping 34 II. Theoretical Foundation of Gravity Medal The gravity model has a long history of empirical success and has been justified theoretically by Linnemann(1964), Leamer and Stern (1970), Anderson (1979), and Bergastrand(1985). Early studies based on general equilibrium approach (Tinbergen, 1962) concluded that incomes of the trading partners and the distance between them are statistically significant and of the expected positive and negative signs, respectively. This gravity model was further developed by Linnemann (1966), who proposed it as a pragmatic way of combining three sets of determinants of the size of a bilateral international trade flow: the importer's demand, the exporter's supply and the costs of doing business. Linnemann(1966) asserts that the gravity model is reduced form from a four—equation quasi-Walraisan equilibrium model of export supply and import demand. Starting with a simple three country model, his four equations are ( 3 . 4 ) Xi’. = Dn(Y...pt.pz,p3.tzttn) (3 . 5) X102 = Diz(Yz.apt,p2,p3.112’t32) ( 3 . 6) X8 = Drg(}é.,ptpz,p.-t..,t23) (3.7) Xf=S,(K,,,p,) where -X5 = demand for the product of country i in country j 35 X? = supply of the product of country i X = national product or, income, of country i K} = production capacity of country i [a = price of a product unit of country i in country j (y = transport costs between countries 1 and j for a product unit of country i Equality of supply and demand is given by (3.8) Xf=Xfi+X3+XP3 and prices are always excluded since they merely adjust to equate supply and demand. Aggregate income proxied the level of demand in the importing country and the level of supply in the exporting country. Distance proxies transport costs which drive a wedge between demand and supply. The gravity model is viewed as a reduced-form equation for trade volume (proxied by value) in which prices do not appear because they are endogenous. Its theoretical foundations have never been made entirely secure and yet it has great intuitive appeal. Despite its widespread empirical use, the gravity equation has been a model in search of a theory. Several 36 different theories have been developed in support of the model, and the differences in these theories help explain the many different forms of the gravity equations and differences in their results. Works by Anderson(1979) and Bergstrand (1985,1987) have produced increasingly complete derivations of gravity type equations from traditional neoclassical theory. Using the pure expenditure system model, Anderson(1979) derives the simplest gravity model as follows; the imports of goods from country i by country j can be written as H.9) AQ=QK where Q = the share of importables in country j's total expenditure K = country j’s total income If it is assumed that income must equal sales, the trade balance equation for country i can be written as (3.10) Y,=bJ-ZYJ- This trade balance equation state that the income of county i must sum to the total imports of country j 37 and it is assumed in equation (3.9) that non-traded goods have zero value. From equation (3.10) we have (3.11) b-: If equation (3.11) is substituted into equation (3.9), the result becomes 11”.- (3.12) M”'=ZY J Equation (3.12) gives the simplest form of gravity equation.2 II. ‘Variables of the gravity model A. GNP and Population The income variables Yi and Yj determine the potential export and import. Since greater productive capacity and incomes promote trade, the coefficients of Y1 and Yj are expected to be positive. Populations are generally used to proxy country size. The more populous countries are assumed to be endowed with a greater quantity and variety of natural resources. This greater self-sufficiency leads to less 2 For more rigorous derivation of gravity equation, see Bergstrand (1985, 1987) and Deardorff (1998). a: V a A: .—.~ T!- 38 reliance on international trade with the expectation of negative values for the coefficient for Ni and Nj. On the other hand a large domestic market promotes the division of labor and thus creates opportunities for trade in a variety goods. Moreover a large market better compensates foreign suppliers for the fixed cost of entry. Thus the coefficient of Ni and Nj cannot be signed a-priori and there is some disagreement regarding the effect of Ni and Nj on trade.3 B. Distance ‘ The distance variable represents resistance to trade. Dij is a proxy variable for natural trade resistance which is a composite of transportation cost or transport time. It is commonly held that people are better informed about conditions prevailing in near-by countries: propinquity leads to better business 3 While most authors find that both countries' population has a negative effect on trade flows ( Linnerman (1966), Aitken(l973), Hewett (1976),Bikker(1987)), Brada and Mendez (1983) found population sizes to have a positive impact on trade flows. Also Brada and Mendez (1985) found that the effect of Ni and Nj is negative and positive respectively. 4Most gravity model does not include the trade barriers (tariff and non-tariff barriers) in the model because of the difficulties of data measurement, such as what relative weights should be given to the import duties levied on the different commodities. Also most gravity model assumes that the trade barriers have equal trade-resisting impact on all the trade flows; possibly differences in impact on individual flows are supposed to be due to the random factors only such as political factors (all dealings with the Communist countries). However Oguledo and Macphee(1994) used import tariffs in the model and find negative effect on the trade volume. 39 information, greater familiarity with laws, institutions, habits, and language of the partner country, and greater similarity in the way of life and in preference patterns. Thus the coefficient of Dij should be negative. C. Adjacency Distance is supplemented by an adjacency dummy which is 1 if i and j share a common land border and 0 otherwise. This variable reflects reductions in both cultural and transportation friction between adjacent countries over and above the effect of distance. Neighboring countries (Aij) can be expected to have an additional stimulus to trade because of similarity of tastes and an awareness of common interests. So the coefficient of Aij should be positive. D. Dummy CU or FTA The use of this model permits us to analyze the preference area effects through the use of dummy variables which are the 1 if both countries i and j belong to country group (FTA) and 0 otherwise. For example, if the estimate of the dummy coefficient is 1.5, it means that the countries belonging to the same country group trade with each other four and half times more [exp (1.5)=4.5 when we transform the log form to F. ‘1‘ fl .“ y a (n 4O exponential] than normal trade which is explained by the gravity variables (GNP, population and distance). So I use this estimate as the measurement of the degree or strength of natural bloc. The dummy variables which I use in this study are the formal trading arrangements that are already in effect such as NAFTA, EC, ASEAN and MERCOSUR and the broader continent-sized groupings that are under discussion such as EA(East Asia)5, EAI(Enterprise for America Initiatives)6,.APEC and TAFTA (proposed Trans-Atlantic Free Trade Area=EC+NAFTA). E. Price and Exchange Rate Price and exchange rate variables were not included in the analysis. The gravity model analyzes imports or exports for many countries at a single point in time, and being based on cross-section data excludes price variables. This exclusion stems from the general equilibrium nature of the analysis, in which prices are endogenous and merely adjust to equate supply and demand. As Leamer and Stern (1970) observe, this does not imply that prices are not effective in allocating resources. 5 See Appendix A for country grouping. East Asia countries are the countries of APEC members which exclude USA, Canada, Mexico, Chile, Australia and New Zealand. Even though East Asia and.APEC are not the trade bloc, gravity model make it possible to estimate the effect of these grouping. 6 See Appendix A for country grouping. EAI includes most of the western hemisphere countries in both North and South America. 41 On the contrary, pries are assumed to adjust quickly, and demand and supply are assumed to be responsive enough to prices to bring about equilibrium rapidly. Bergstrand (1985) notes that the absence of these data is the only reason for their exclusion. A large percentage of the sample used in this study were developing countries: yearly price index figures for many of these countries are unavailable or, at best, unreliable. Official exchange rates figures are generally not market rates: thus they are of limited usefulness. Both price and exchange rate variables affect competitiveness and trade. This can be the main limitation of the model in spite of empirical success in its applications. III. Past Use of the Gravity Model: A.Review The gravity model has been applied to a wide variety of goods and factors of production moving across regional and national boundaries under different circumstances. The model has been successfully applied to flows of varying types, such as migration, flows of buyers to shopping centers, recreational traffic, commuting, patient flows to hospitals, telephone calls and interregional as well as international trade. In this 42 section the past use of the gravity model in the international trade area is reviewed and the model estimates are reported in Table 3-1 with the list of variables used (Table 3-2). A. Bloc effect The gravity model has been used to quantify the effects of economic integration on trade flows. Tinbergen (1962) estimated the preferential effect of the British Commonwealth and Benelux CU. The purpose of his study was to determine the pattern of international trade that would prevail in the absence of discriminating trade impediments. Using a simplest form of gravity model (including only GNP of exporting and importing country, distance and bloc dummies as explanatory variables) he finds that both preferences have significant and positive effect on the size of international trade flows. Further he finds that there are deviations in actual trade from the normalized trade pattern of the gravity model. The negative deviations or negative error terms of the model are interpreted as the evidence of the existence of special barriers and obstacles to the optimum flow of .7 international trade. Linnemann(1966) estimated the 7 See Tinbergen (1962) Appendix VI for the results of deviations from the model. 43 preference effect for the British Commonwealth of Nations, French Community group and Portuguese and Belgian colonies.8 Aitken(l973), estimated a gravity model that included dummy variables for common membership in a free trade area, and thus was able to estimate the trade creation and diversion effects of the EEC and EFTA. By regressing the gravity model without the bloc dummy variable he got the estimated trade values without preference and then this estimated values were subtracted from actual trade to estimate the trade creation and diversion effects of the EEC and EFTA. Hewett (1976) used a gravity model to explore the East-West trade. He finds that typical western trade volume lies above that of typical communist trade volume when he compares the western and eastern trade-volume to the normal trade volume from gravity model. The ratio of intra- CMEA(Council of Mutual Economic Assistance) trade to typical eastern trade turns out to be much higher than the ratio of intra-EEC or intra-EFTA trade to typical western trade. Pelzman(l977) also studied trade creation and diversion in the CMEA during 1954-70 using gravity model. The difference between the actual CMEA trade flows and the hypothetical CMEA trade flows from gravity a See Linneman (1966) Table 4.5 for country grouing. 44 model is taken to be indicative of gross trade creation.9 The difference between actual trade flows with nonmembers and the projected trade flows with nonmembers indicates the trade diversion effects. The resulting difference between the GTC and TD effects measure the TC effects. Instead of employing OLS he uses maximum likelihood technique and GLS. He finds that the CMEA countries have experienced the cumulative growth in trade creation over the integration period 1965—70. The estimates of gross trade creation ranged from $9.2 billion in 1965 to $13.2 billion in 1970. The estimates of trade creation ranged from $9.9 billion in 1965 to $13.1 billion in 1970. The estimates of trade diversion ranged from -$0.7 billion10 in 1965 to 0.1 billion in 1970. Brada and Mendez (1983) compared the economic integration of developed and developing countries by employing a gravity model. They find that economic integration among developing countries (Andean Pact, Central American Common Market, and Latin American Free Trade Area) can have the same positive effect on intra- 9 The TD and TC effects combined result in GTC, which signifies a growth in trade among the member countries, regardless of replacing domestic production (TC) or the replacement of nonpartner imports by partner country imports (TD). See Balassa (1967) for GTC. m Negative trade diversion represents external trade creation which is possible in the case of a customs union when members reduce their tariff to the level of the common external tariff. 45 member trade as it does among developed countries (EEC and EFTA). They also find that the those bloc effects are influenced by the level of development of the integrating countries and by the distance between them. Brada and Mendez (1985) also examined economic integration among developed, developing and centrally planned economies. Using a gravity model they find that even though effective integration is possible for both developed and developing countries, such as those in Latin America, distances between members may severely limit the benefits of integration. Thoumi(l989) used a gravity model to study economic integration among the LDCs of the Caribbean Basin. He finds that the integration systems of the region have had varying degrees of success in generating trade. Those that lowered trade barriers against outsiders appear to be relatively successful. The CACM (Central American Common Market) and CARICOM (Caribbean Community) appear to have had a substantial impact on intra-Basin exports, while LAFTA (Latin American Free Trade Association) has failed to promote trade significantly in this subregion. Wang and Winters(1992) studied the trading potential of Eastern Europe using a gravity model. They find that actual trade between 46 Eastern-bloc countries and market economies is just one- quarter of its potential as estimated by a gravity model. Frankel(l993), Frankel and Wei(1993) and Frankel, Stein and Wei(1995) use gravity models to find the effect of various blocs on international trade flows. With recent interest on regionalism, they revitalized the gravity model to study bloc effects. Using the gravity model to examine bilateral trade patterns throughout the world, they find that the European Community, the Pacific, and the Western Hemisphere have trade bloc effects. Intra-regional trade turns out to be greater than could be explained by natural determinants: the proximity of a pair of countries, their sizes and GNP/capita ratios, and whether they share a common border or a common language. Frankel(l993) focuses on East Asia and the Pacific; He reaches several conclusions regarding the Yen Bloc that Japan is allegedly forming in East Asia and the Pacific. First, gravity-model estimates of bilateral trade show that the level of trade in East Asia is biased intra—regionally, to a greater extent than can be explained by distance. Second, there is no evidence of a special Japan effect because the estimates of bloc dummies between Japan and other East Asian countries are not significant. Third, 47 once one properly accounts for rapid growth in Asia, the statistics do not bear out a trend toward intra-regional bias of trade flows. So he concludes that beyond the- evident fact that Japan and other Asian countries were growing rapidly, there is no evidence that Japan is concentrating its trade with other Asian countries in any special way. Frankel and Wei(1993) focus on the EC and EFTA- They find that in Europe, it is the EC that operates as a bloc, not EFTA. EC members trade an extra 55 percent more with each other, beyond what can be explained by proximity, size, and GNP/capita. They also find evidence of trade-diversion in 1990 by the negative sign in the bloc dummy. Second Enlargement of EC in 1986 caused much trade diversion. Recently, McCallum (1995) used a gravity model and detailed Canadian data on interprovincial and international trade to demonstrate the effect of the US- Canada border in diminishing trade in goods. He finds that trade among 10 Canadian provinces is on average 22 times larger than trade between 10 Canadian provinces and 50 U.S. states. So he concludes that national borders in general continue to matter. McCallum(1996) also used a gravity model of the 1988-90 merchandise trade flows among Canadian provinces and between Canadian provinces and U.S. states. He shows that Quebec trades twenty ,- Ar-.." - b ‘ b; 5.. - ‘ C F ”nu- P vvv.“'"‘ . \ .F“ ...d- an Oh“ Dete: B. ..u CC 81‘ an ‘- n u» -e. ”L. I Q s. x... ‘y‘. Ne~~‘e AA‘.,.,' ‘~. 48 times more with other provinces than it does with U.S. states of similar size and distance. The results imply that the fabric of national economies is far tighter than that of the global trading system, even for countries operating without substantial trade barriers such as U.S. and Canada. B. Determinants of Trade Flows Sattinger(1978) uses a gravity model to study whether trade between countries is motivated by differences or by similarities in agricultural land per capita, per capita income, the urban proportion of the population, and temperature. He finds that trade between countries is led predominantly by differences rather than similarities and that greater differences in agricultural land per capita, per capita income, the urban proportion of the population, and temperature, results in greater trade between countries. Using the basic variable of a gravity model (GDP, Distance and Adjacency) along with several other variables, Srivastava and Green (1986) studied the determinants of bilateral trade flows. They find that distance, product category, political stability, cultural similarity, colonial past, membership in an economic union, and standard demographic variables '1 .- . .7 .5 . . E . . E .3 C Cw .flv C #9. u .- .~ ‘ S. .A‘ .1; V; .C a X a (C .2 r I e v. E r. I a a t. e a . .hi“ 1' w or» .. . ...& . . “w. an by .«e a o v. r. A: a: H a 3 . o a . Co . .6 Lu pro vi . c C C um c. "1.. “v a . . . ~u. 49 such as GDP and population are the important determinants of trade flows. C. Exchange Rate Regime (Fixed or Float) Gravity models were also used to study trade flows under flexible and fixed exchange rates. Abrams(l980) estimates the trade losses from exchange rate variability. Adding an exchange rate uncertainty variable to the basic gravity model, he estimates that the trade losses which may have occurred during the 1973- 76 period as a result of exchange rate uncertainty were greater than in the pre-l973 fixed-rate period. The exchange rate uncertainty variable is defined by any bilateral exchange rate variability (VEX) and the variation of bilateral exchange rates from trend(VTREX).11 Even though the study cannot be generalized to show the superiority of one exchange rate regime over another he maintains that other thing being equal, increased exchange volatility is detrimental to trade. Thursby and Thursby(1987) also used a gravity model to study the effect of exchange risk on bilateral If in year t, j’s exchange in terms of i’s currency is Exi,j,t, then V1i')0,j,t=EKEXI’J‘J:-E—Xi,j,t—l)—l]2 where k represents the months of t-l. If monthl-y changes in bilateral exchange rates are AEXij, then WREXI’JJ=:Z:[(Am,j,k—AEXi,j,t-l)-l]’ where k represents the months of t-l. RA‘S'R-‘ VV‘...‘ (""8” u...“ i :M .Ffl ‘ "J: 84 €.__ 1 ‘ “‘-u ’1 "A-‘ . l . IT‘S». “AVQ‘* x,- . 7:. A» .‘ ‘:~&.c- NA " Vs Hash] ~ V‘Au v 1.71"” Q Nye». ~ (‘1‘ s... c » ‘1 s. p .3 y‘ “EV ‘3 ‘ ‘4 ‘L \ 50 trade flows and to examine the Linder hypothesis that countries with similar and sufficiently high incomes will engage in much trade . Using a sample of seventeen countries for the period 1974-82, the authors find strong support for the hypothesis that increased exchange-rate variability affects bilateral trade flows negatively; and also overwhelming support for the Linder hypothesis.12 Brada and Mendez(1988) used a gravity model to study the effect of the exchange rate regime on the volume of international trade. By adding the dummies for the exchange rate regime to the basic gravity model they find that bilateral trade flows among countries with floating exchange rate are higher than those among countries with fixed rates. Also they find that while exchange rate uncertainty does lower the volume of trade among countries, regardless of the nature of their exchange rate regime, its effects are less than the trade-reducing effects of restrictive commercial policies imposed by fixed rate countries. D. Political Economy Summary(1989) attempts to identify and quantify the factors affecting bilateral trade flows between the ‘zThey use a variable of the absolute difference in per capita income in the two countries and find the negative effect of this variable on the trade volume. 80 this result support the Linder Hypothesis that the similar countries trade more. ict P L. - C1»- ‘, -c C a“- R- -- C 4V“V-. ' -A " Q" P? -eAb"-.‘ n ' I PM“. ,5. A? vb--'.'v~ h ’ j. .p‘ ‘5'“ W We ‘91.- .3... qu HU..-‘ “9"" Tu ‘- H. .a Trad. 51 United States and other countries by developing a gravity model. She adds semi economic variables and international political factors such as arms transfer, political rights, civilian employees and foreign agents to the basic gravity model and finds that these variables are important determinants of U.S. bilateral trade. Pollins (1989) studies the effect of international interactions on bilateral trade flows by adding international corporation and conflict variable to the basic gravity model. He finds that the international conflict has a significant negative effect on the bilateral trade flows. Recently Hufbauer, Elliot, Cyrus and Winston (1996) used a gravity model to study the effect of Us economic sanctions (against countries such as Cuba, Iran and Libya) on trade, jobs and wages. By adding economic sanction dummies on the basic gravity variables, they find that US exports were $15 billion and $19 billion lower than they would have been if not for the effects of sanctions put in place in 1995. E. Trade Disaggregated By Commodities Some authors tried to estimate trade in disaggregated commodity group such as manufactured goods, agricultural products, fuels, and other raw materials I id. Co. . . . . a . v . CL A: a. «G .3 . . C a C a t l k O. L“ e . u .1... .G a . .1 .. . a e a... t t .. . D. .C m 5 z . r : . do v. S .. . C e 7: t o .. . A: z . CC u a . r. .3 +1. to no. a .. Q. t. . . .2 at . . a.“ r. C. .3 v. at am .3 .9... a-“ .3 .3 a o .3 .3 A O a o. A.» .3 kl. D. ih... .2 at n. F. T“ _. ma 2. a. at . . at a» at Tho ; .U‘ v‘ TH ‘ts. L).- 4“ 52 instead of aggregated total bilateral trade data. The estimates of the gravity model changed little in most cases. Frankel(l992) finds that the Asian grouping coefficient has the highest estimates in the raw materials and has the highest significance in the manufactured goods if judged by t-statistics. Christerson(l994) uses a gravity model to study world trade in apparel. He finds that for low value apparel products, which tend to compete in price, labor costs were a significant determinant of trade flows, causing production to concentrate in low-wage areas. On the other hand for high value products, which tend to compete in quality, fashion, and quick response to changing demand conditions, production for export tend to take place near fabric suppliers and final markets, which tend to be in higher-wage areas. He concludes that proximity to markets and suppliers often outweighs the importance of labor costs, particularly for high-end apparel production. F. Direct Foreign Investment Hufbauer, Lakdawalla and Malani(l994) use a gravity model to study determinants of direct foreign investment and its connection to trade. To analyze the determinants of direct foreign investment they use data . ;].-’-“‘,’.‘.V‘-r~v " V‘~c§. Ivu . H (r ‘ 7‘ ‘ ““‘ r . w I. 5‘... -, OC‘S CL‘ e.“ H .; Iv" ‘\ fl ‘1‘ I 53 on direct investment from Germany, Japan and the United States. Using investment stock and investment flows as dependent variables instead of bilateral trade flows, they find first that regionalism (regional dummies: EU, Asia Pacific rim and western Hemisphere) plays a significant and consistent role only in the investment stock placements of United states and Japan. Second, the size and openness of partner economies are important determinants of the distribution of investment stocks. Third, they find that Japanese firms are more conservative in the sense that new Japanese investment tends to follow established locations. To analyze trade and investment links they use basic gravity variables adding investment stock as independent variable. The empirical results show that DFI of Japan tends to promote imports more than exports of Japan while DFI of the United States seems to increase exports more than imports of the United States. The estimates of the gravity model reviewed in this chapter are reported at the Table 3-1 with the list of variables used (Table 3-2). Table 3-1 54 : Empirical Results of Previous Studies List of Tinberg Linneman Aitken Hewett Pelzman Brada Brada Gravity on n (1973) (1976) (1977) and and Variable (1962) (1966) Mendez Mendez (1983) (1985) GDPi 0.84 0.86 0.911 0.97 0.954 0.357 1.092 GDPi 0.62 0.98 1.052 0.75 0.788 0.131 0.157 Ni -0.14 -0.369 -0.11 -0.283 0.899 -0.291 Nj -0.21 -0.331 -0.03 -0.177 0.680 0.574 Distance -0. 56 -0.77 -0.349 -0. 78 -1.229 -0. 760- -0. 543 Adjacency 0.02 0.892 EC 0.887 0.51 2.307 3.11 EFTA 0.572 1.23 2.095 2.46 NAFTA CMEA 4.30 2.788 ANDP 0.346 1.51 CACM 1.916 2.50 LAFI‘ A -1.467 1.15 C 0.05 1.27 PB 0.04 NOB 306 3532 132 322 17921 789 Data Year 1958 1958-60 1967 1970 1954-70 1954-77 1970 R-Squared 0.84 0.63 0.87 0.86 0.58 0.56 0.651 Indepenc Variable 55 Independe Thoumi Wang Frankel Frankel, Frankel, McCallu . (1989) and (1993) aand Stein m Variable Winters Wei and (1983) (1992) (1993) Wei (1995) GDPi 1.009 1.02 0.787 0.75 0.75 1.21 GDPi 0.241 1.17 0.787 0.75 0.75 1.06 Ni -0.22 N] -0.38 Distance -0.898 -0.75 -0.589 -0.55 -0.56 -1.42 Adjacency 0.249 0.78 0.732 0.79 PERGDPi 0.468 0.078 0.09 0.09 PERGDPj 0.218 0.078 0.09 0.09 EC 0.341 0.52 0.49 EFTA 0.04 -0.05 NAFTA 0.05 EA 0.66 WH 0.934 0.93 APEC 1.597 1.32 TAFTA ANDP 0.90 CACM 3.805 LAFT A 1.044 CARICO 4.261 M Inter-prov. 3.09 MERCO. 2.09 NOB 5700 1953 1647 1573 683 Data Year 1971 1990 1990 1990 1988 R-Squared 0.618 0.75 0.77 0.77 0.81 P———' ' lndepen; . Variable 77 ~ (‘7‘- slliik s CD '7 56 Independe Sattinger Srinivasta Abrams Thursby Brada and . (1978) and (1980) and Mendez Variable Green Thursby (1 98 8) (1986) (1987) GDPi 0.91 0.218 0.76 2.03 0.479 GDlj 0.79 0.012 0.65 0.55 0.393 N1 -0.004 0.291 Nj 0.089 0.277 Distance -0.97 -0.449 -0.25 -2.389 -0. 775 Adjacency 1 .461 PERGDPi 0.25 PERGDPj 0.08 EC 0.81 0.3 13 EFTA 0.97 0.24 NAFTA PRF 0.784 FIXi -1 .919 M -0. 839 FLOATii -0.851 VEX -0.05 -0.95 EXR -4. 126 XUV -3.891 Aij 0.25 Bij -0.59 Uij 0.42 Iij 0.78 Tij 0.09 NOB 380 3690 76 144 Data Year 1972 l 977 l 973 -76 1 974-82 1977 R-Squared 0.80 0.3095 0. 80 0.64 0.6845 Independ thfidfle Sununa ry (1989) Pollins (1989) CHJPi 1.136 (EEflfi (142 1.386 bfi 4113 IDknance 4143 -O.752 Aufiacenc IHERIHD PEIUSI) BIZ (1719 EIWHK bLAFqflk (122 14120 Pofifiafl (fivfl (104 iFonfign (185 VVfiiSorp (1036 (hATWT '1030 AUHECI 'TAHVTAL ADVIMP (CAHCLJ InAFTYX 21409 (:hflEAt 11216 IQCHB 552 Ihna lfimn' 1982 1973 Squanxl (162 (1567 57 ' GDPi GDP] 1 Ni ...\'i h Dmance Adéacenc; we PERGDP EC K l A\1)P ~52 21/57“ 77 7 #777- 58 Table 3-2: List of variables used in previous gravity model studies GDPi Exporter GDP GDPj Importer GDP N1 Exporter Population Nj Importer Population Distance Distance Between Eaggorter and Importer Adj acency Adjacent Country Dummy PERGDPi Per Capita GDP of Exporter PERGDPj Per Capita GDP of Importer EC Eurgrean Community EFTA European Free Trade Area NAFTA NAFTA preference SO Socialist Exporter LOME Lome preferences EA East Asia_preference WH Western Hemisphere APEC Asia-Pacific Economic Corporation TAFT A Trans-Atlantic Free Trade Area ANDP Andean Pact CACM Central American Common Market LAFT A Latin American Free Trade Area C CommonwealtMreferences AA Assoc. African EC preference TM Tunisia-Morocco-French preferences G P GSP Portugues preference FA Other French Africa preferences F French preferences B Belgian preferences PB Benelux Preferences 1.1 FA"; . t v. are C. l u: w“ v~ A...» rs ~ AV E» 9“ ?s «an ~ .L ....n 7.» nu. .. CHPTER 4 POSSIBLE ENLARGEMENT OF NAFTA; ‘What is the next step? I. Introduction A tripolar trade system centered on Europe, United States, and Japan has become a popular topic among media and popular writers. What’s more, EU countries are moving to incorporate eastern Europe, while NAFTA contemplates inclusion of south America. Some authors worry that three is an unstable number and that parties of three tend to split into a two and a one. The proposal for a Trans-Atlantic Free Trade Area (TAFTA), a tariff-free common market uniting North America and Europe1 may be motivated by these concerns. And APEC (Asian-Pacific Economic Corporation), the loose form of free trade area, is very active now. This chapter attempts to identify a possible enlargement of NAFTA by using the natural economic bloc concept and gravity model with new dummy variables which represents various combination of regional grouping. I examined which of the following alternatives to NAFTA 1 See Wall Street Journal (May, 2,1995; Sec C, p 20) and Chicago Tribune (May 21, 1995; Sec 7, p 1) 59 F‘ '5‘ n o— .‘D--- u - 7 '2 .. ‘-‘.‘~ ‘ II. MO< \" “H- ‘ in - ‘ In. . ‘- .n ‘1‘. . ‘N ..‘) W~hgy .L I (We a 4*." ‘1 .4 t A ‘2‘ . “Pi ~£ f‘ I O, enlargement will maximize welfare: A southward expansion incorporating Latin America (EAI, Enterprise for America Initiative); westward in a deal with Europe (TAFTA, Trans Atlantic Free Trade Area); or eastward (APEC, Asian-Pacific Economic Corporation). II. Model Specification The model to identify a possible enlargement of NAFTA is a gravity model with various combination of country group dummies. To see where the NAFTA.“should” expand, the TAFTA, EAI and APEC dummy variables are added to gravity model specification. log(7;) = a + fl. log(GDP.) + A log(GDP,) + A log(N.) (4 . 1) + A log(Nj) + ,6510g(DISTANCEy) + ,66(ADJACENCY;) + 7,(TAFTAg.) + 72(EAIy.) + 73(APECQ.) + uij. where Tij = value of trade (exports+imports) between country i and country j2 GDPi, GDPj = income in the exporting and importing countries 2 See Appendix A for the countries involved in this study. -vn'f“ \ ,‘ \ M d‘gAen- 1'; 1'- .‘o’ “ ‘H-“.‘ .- yh, .ouhraot'v “25¢:- . . ..~.. - ‘1 I 61 DISTANCEij = distance between countries 1 and j Ni, Nj = population of exporting and importing countries ADJANCENCYij = adjacent dummy variable which takes the value of if both countries 1 and j are adjacent and 0 otherwise TAFTA, EAI, APEC = dummies for country grouping The coefficients of the dummy variables are compared to each other to see which grouping will maximize welfare. Also by combining NAFTA with individual countries, the “best” partners of NAFTA are identified. How these dummy variables are related to natural grouping is illustrated in figure 4-1. The gravity model offers a systematic framework for measuring what patterns of bilateral trade are “normal” around the world. In addition the coefficients of regional grouping dummies offers the measurement of the degree of strength of natural grouping. A positive coefficient of regional dummy (Case 1 and 2) shows that the regional grouping has a positive effect on the regional trade flow. The higher the coefficient of bloc dummy (Case 2) the closer is the grouping to the natural economic bloc which means a more welfare-improving group. The negative ... .C e i . .. v. p. . C. u. D. C. 3. AU AU .- 0 an». an. at e 9. Ce it C» . A‘ .a An: A . v A, b A u ade 1'. III. D. y. ‘1 SSte 62 coefficient (Case 3) shows that the regional bloc is possibly a trade diverting grouping. Figure 4-1: Natural Economic Bloc and Gravity Model Tmfle NmnnfihmkPmmnuw tmmmmflbmefiw vmmmksfififlfiDBme and Adjacency) (HHPIfimmmwlhfimmeI Case 1: Bloc Dummy Effect: Positive Case 2: Bloc Dummy Effect: Positive : Effect bigger than Case 1 Case 3: Bloc Dummy Effect: Negative III. Data Included in this chapter are 122 countries: Western Hemisphere (28 countries), EU(12), EFTA(6), East Asia (10) Other Asia (10), Middle East(13), Africa(40), Oceanic(2) and USSR. Bilateral trade flows among those countries yield 2557 data points. These trade data in \(3 \L) . dyn- 5-..! y..- .o—o- “v- Ir, ‘W. no. _ ‘ 3‘7 .4". I ll.‘ A. 63 1993, measured in millions of U.S. dollars, were obtained from the International Monetary Fund, Directions of Trade(1993). GDP and population figures are derived from International Financial Statistics by International Monetary Fund. GDPs are measured in millions of US dollars and populations are measures in millions of persons. Distances (in miles) are the distances between the countries' major harbors and were obtained from the internet by Jon D. Haveman3. IV. Empirical Result I've tried several different forms of the gravity model by using OLS method. Results are reported in Table 4-1. Except the model (4.3) which was regressed on bilateral exports, most of the model was regressed on bilateral trade (exports+imports). 1080;) = a + [3.108(GDP.) + 19401311) + 53108017,) + 13. 10g(N,) (4 . 2) + ,6,log(DISTANCEg) + ,66(ADJACENCY.U.) + y‘(NAFTA,.j) + 72(ECD.) + 73(EA9.) + u”. Iog(X.) = a + A 10g(GDR) + fl.(GDP,) + .6. Iog(N.) + 16.10am.) (4 . 3) + ,Qlog(DISTANCEu) + ,B‘(ADJACENCY;) + 7'(NAFTAU) + y2(ECu) + y3(EAu) + u”. where Xij : Country i’s export to country j 3 The internet address is http://intrepid.mgmg.purdue.edu/pub/Trade.Data/distance.txt. From this source I also obtained the adjacency of countries. y: as“. W- 'Yl'!‘ «- 9 - J3 Jr. .fi‘s a, To compare the effect of TAFTA, APEC, EAI (Enterprise for Americas Initiative: Western Hemispheres) and to find the significant effect, several forms of gravity model are regressed by combining TAFTA, APEC and EAI with existing bloc. Eq(4.4) combines TAFTA with EA while Eq.(4.5) APEC with EC. Eq(4.6) combines TAFTA with NAFTA, EC, and EA, while Eq(4.7) EAI with EC and EA. All those models are cross-section for 1993. log(T.) = at +19. 103(GDP.) + ,3. log(GDP,) +19. log(N.) + AW.) (4 . 4) + ,6, log(DISTANCEu) + A(AD1ACENCI;) + yl(TAFTAu) + 72(EAU) + u”. log(1;)= a + 19. 1040012) + )6. 10461912) + fl. log(N.> + fl.(N,) (4 . 5) + ,8, log(DISTANCEv) + fl‘(ADJACENCYu) + yl(APECu) + 72(ECU) + u”. 1030;) = 0 +19. 10g(GDP.) +fl. 10g(GDP,) + Alog(N.) + AW.) (4 . 6) + ,8, log(DISTANCEu) + ,BAADIACENCYL) + 7|(NAFTA0) + y,(ECu) + 7,(EA) + 74(TAFTA) + u”. low.) = a + fl. log(GDR) + mama) + 13. IogI' v-‘I :‘Aiyca‘l‘b‘ N‘.’Y 'qO-‘Q Aouyu' CW9: wg§$b 67 square of the regression among dummy variables (0.617 the highest between EC and TAFTA dummies, 0.01 between NAFTA and TAFTA, and 0.00 between NAFTA and EC). So we don't have to worry about the multicollinearity. Should the US should follow the TAFTA (Trans- Atlantic Free Trade Area) or EAI (Enterprise for America Initiative) or APEC? It turns out that the US should give priority for focusing on APEC if we follow the natural trade bloc argument. The regression the coefficient for APEC is the highest (1.89) and TAFTA(O.52) was a distant second while that of Western Hemisphere(EAI) was the lowest(0.03). As we’ve seen from Figure 4.1, APEC gives the highest jump in the level of intra-regional trade. Hence the APEC is the most significant natural grouping. The next question is which countries of APEC will be the “best” partners of NAFTA in terms of welfare maximization. By adding each individual member of APEC to NAFTA dummy, the best candidates for natural bloc formation turns out to be Hong Kong (1.99) and Singapore (1.96). The rank of the coefficients are reported in the Table 4-2. 68 Table 4—2 : Who is the best partner of NAFTA? Country Name Coefiicient Ikmglkmg 199 smgmmxe L96 nnmwsa 179 Komm 166 thmi 157 Chma 144 Thdhmd 104 Inmnwda 09 IWaNZbdmml 08 (fifle 078 Auflnma 074 lflmkmmes 072 PqnmJWmN -LOS (Ema: V. Summary and Conclusion In this chapter, I examined in which direction NAFTA.should expand, to the south (EAI, Enterprise for America Initiative), east (TAFTA, Trans-Atlantic FTA) or west (APEC, Asian-Pacific Economic Corporation). To do this I have used the natural bloc concept combined with gravity model. It turns out that the most “natural” extension of NAFTA is towards APEC and within APEC the best candidates are Hong Kong, Singapore Malaysia and Korea. The US should focus more on Asia, than on TAFTA or EAI. CHAPTER 5 Trade Creation and Trade Diversion of ASEAN, ANDEAN, EC and.MERCOSUR: By Gravity Mbdel I. Introduction While chapter 4 analyzes the regional groupings which do not have formal agreements, this chapter focuses the trade creation and diversion effects of ASEAN, EC, ANDEAN and MERCOSUR, trading blocs that are already in existence. While ASEAN (Association of Southeast Asian Nations) was formed in 1967 to promote economic, social, and cultural cooperation among Indonesia, Malaysia, Philippines, Singapore, and Thailand, ANDEAN (Andean Common Market) which involves Bolivia, Colombia, Ecuador, Peru and Venezuela was formed in 1969. MERCOSUR(Southern Cone Common Market) which includes Argentina, Brazil, Paraguay and Uruguay is relatively new, having been created in the spring of 1991. I included the EC to compare the size of trade creation and diversion with ASEAN, ANDEAN, and MERCOSUR. Also these blocs will be analyzed for the growth effects in Chapter 7. The methodology for estimating trade creation and trade diversion in this chapter is ex-post gravitational 69 Y ‘ weL‘a .- co"; 3 sexy--9 II. Met DIVI EVA. ‘ ~ .oler “A“ A ‘ § .13 _r1 C; sly Vs 70 approach. In this chapter I am estimating the values of trade creation and trade diversion, not measuring the welfare effects. In section II, the model specifications and methodologies will be discussed and the estimation of trade creation and trade diversion will be reported in section III. II. Methodologies for Estimating Trade Creation and Trade Diversion .A modeler who intends to employ an ex-post estimation procedure possesses all relevant post- integration data, but needs to know what those data would have been in the absence of integration. Thus, the problem that manifests itself in ex-post estimation is in constructing the "antimonde." In ex-post models, trade creation estimates are obtained by first estimating what expected total imports would have been in the absence of integration. Once this is obtained , it is necessary to subtract the expected total imports from the actual value of total imports, in order to deduce the change in total imports due exclusively to integration. Trade diversion is estimated by subtracting the actual value of non- partner imports from the estimated value of external imports in the absence of integration. 71 The construction of the antimonde and its comparison to actual data are essential for the estimation of trade creation and trade diversion. As long as a realistic, unbiased, and consistent technique is used to formulate the antimonde, the possession of actual data will enhance the plausibility of the estimates. Thus, the construction of a realistic antimonde is of the essence. The gravity model offers alternative means for doing this. First, estimate a gravity model that includes dummy variables for common membership in a free trade area and find the effect of the bloc by the significance of bloc dummy. log(My) = a + ,61 log(GDR) + ,1?2 log(GDPj) + ,6, log(N,.) < 5 . 1 ) + A log(Nj) + fl,log(DISTANCEy. ) + [36(ADIACENC13.) + yl(BLOC) + uij. where Mij = value of imports in country i from country j GDPi, GDPj = income in the importing and exporting countries DISTANCEij = distance between countries 1 and j Ni, Nj = population of importing and exporting countries 72 ADJANCENCYij = adjacency dummy variable which takes the value of 1 if both countries i and j are adjacent and 0 otherwise BLOC = dummy variable for ASEAN,EC, ANDEAN and MERCOSUR Second, the gravity model without the bloc dummy variable will be used to get the projected imports values in the absence of integration. Lastly, these projected values will be subtracted from actual imports to estimate the trade creation and diversion effects. The trade creation (TC) effect refers to increase of imports from the partner countries replacing domestic production. The trade diversion(TD) refers to the replacement of nonpartner imports by partner country imports. As defined by Balassa (1967) gross trade creation (GTC) refers to the increase in intra-member trade, regardless of replacing domestic production (TC) or the replacement of nonpartner imports by partner country imports (TD). External trade creation (ETC) will refer to increase in imports from nonpartner countries. It is possible in the case of a customs union when members reduce their tariff to the level of the common external tariff. 73 III. Estimates of Trade Creation and Trade Diversion .A. TC and TD Estimates The gravity equation(5.l) results are reported in Table 5-1. While the ANDEAN shows a negative bloc effect, ASEAN and EC show the positive effect on intra-bloc imports. The effect of ASEAN, EC, ANDEAN and MERCOSUR was variant depending on the year of beginning of the trade agreement. So the focus of trade creation and diversion is limited to 1970, 1975, 1980, 1985, 1990 and 1995. To obtain the projected import values without integration, the gravity model without the bloc dummy variable is used. Thus projected imports are obtained by plugging the data of GDPi, GDPj, Ni, Nj, DISTANCEij, and ADJANCEij into the gravity model results without the bloc dummy. For the case of ASEAN 1995, the following gravity result without the bloc dummy log(Mv.) = 1122183 + 1.474437log(GDR) + 1275097log(GDPj) 1 5 . 2 ) - O.728792210g(Ni) - 0.5358907log(Nj) - 0.9471 35610g(DISTANCEfi) + 05008433(ADJACENCI;) is used to get the projected imports. The results of projected imports are reported in the Appendix B. Then these projected values were subtracted from actual imports to estimate the trade creation and diversion effects. 74 Table 5-1: Cross—Sectional Results (71) form Model (5.1) Year ASEAN EC ANDEAN MERCOSUR 1961 . . . 1962 0.767669 . . 1963 . 0.773136 . . 1964 0.717059 . . 1965 . 0.730029 . . 1966 0.696454 . . 1967 . 0.643965 . . 1968 1.314757 0.752288 1969 2.014907 . . 1970 1.2672 . . . 1971 2.168632 . -0.98439 . 1972 2.167735 . -0.74398 . 1973 1.683634 —0.73991 . 195T 1746115 . -1.49944 1975 1.528872 0.428787 -0.91525 . 1976 1.655361 0.488971 -0.94935 . 1977 1.614157 0.510943 -0.88977 . 1978 1.618101 0.535997 -0.97717 . 1979 1.722675 0.606116 -0.95377 . 1980 1.777118 0.534293 -0.98084 . 1981 1.418053 0.423394 -1.03618 . 1982 1.398816 0.371004 -O.82411 . 1983 1.361215 0.372246 -1.25841 . 1984 1.301678 0.35264 -1.5406 1985 1.448796 0.457558 -1.44854 . 1986 1.487884 0.687565 -1.45581 . 1987 1.557849 0.674339 -1.18801 . 1988 1.490453 0.614078 -1.46756 . 1989 1.719471 -O.51665 . 1990 1.667796 . . . 1991 1.472287 . . . 1992 1.409535 . . . 1993 1.461568 . . 1994 1.480809 . . 1995 1.237432 . . 1.070952 75 The difference between the actual intra-bloc imports and the projected intra-bloc imports is taken to be the indicative of the gross trade creation as defined by Balassa(1967). It is the increase in intra-trade regardless of whether domestic production (TC) or the imports from nonpartner countries are replaced by partner countries(TD). By adding those differences in intra-bloc imports, GTC of bloc is estimated. The GTC of ASEAN, EC, ANDEAN and MERCOSUR is reported in table 5-2. In 1985, the GTC of EC (increase in intra-EC imports ) was $119 billion while the GTC of ASEAN (increase in intra-ASEAN imports) was $11 billion. While the GTC of ASEAN and EC show positive , ANDEAN shows negative GTC ($0.6 billion in 1985). Negative GTC means that the ANDEAN bloc causes the decrease in intra-ANDEAN imports. From this negative GTC of ANDEAN we can speculate external trade creation (increase in imports from nonpartner countries)of ANDEAN which is possible when members reduce their tariff to the level of the common external tariff. In 1995 the MERCOSUR'S GTC was $9 million. The negative difference between the actual imports from nonmember and the projected nonmember imports (Appendix B) indicates the trade diversion effects. By adding those differences the trade diversion of blocs is 76 estimated. The trade diversion of EC shows the highest trade diversion among the blocs ($12 billion) while ANDEAN shows the next highest ($4 billion) in 1985. The trade diversion of ASEAN was $1 billion in 1985, $1 billion in 1990 and $2 billion in 1995. In the case of MERCOSUR, the effect of bloc shows up only in 1995 and trade diversion was $13 billion in 1995. Table 5-2: GTC, TD and TC of ASEAN, EC, ANDEAN, and MERCOSUR($ Million) Year ASEAN EC ANDEAN MERCOSUR 1970 GTC 205 . TD 156 . . . TC 49 . 1975 GTC 3,410 53,131 -165 TD 595 15,166 1872 . TC 2,815 37,965 -2073 . 1980 GTC 11,899 119,388 -531 . TD 1,025 32,072 1,193 . TC 10,874 87,316 -1,724 . 1985 GTC 9,886 87,478 -599 . TD 1,009 11,680 4,126 TC 8,877 75,798 -4,725 . 1990 . GTC 25,556 . TD 1,136 . . TC 24,420 . . 1995 GTC 55,768 8,537 TD 2,127 13,202 TC 53,641 -4,665 77 The resulting difference between the GTC and TD effects will be indicative of the trade creation effects of blocs. Table 5-2 shows that the ASEAN and EC are trade creating blocs while ANDEAN is trade diverting bloc. In the case of MERCOSUR, and trade diversion ($13 billion) was bigger than the GTC ($8 billion). So MERCOSUR has negative trade creation which indicates trade diverting bloc and there was no trade creation. The gravity model was useful for identifying the countries from which the trade diversion effect originates. From Appendix B, for ASEAN, the trade diversion effects originate from Europe (France, Italy, Spain, Switzerland and Denmark) and Latin America (Mexico, Brazil, and Venezuela). For ANDEAN, the trade diversion originates from.Asia (China and Japan) and from North and South America (Brazil, Mexico, USA and Canada). In case the of EC, diversion comes from Latin America (Mexico and Brazil) and Asia (China, Japan, Korea, and Taiwan). For MERCOSUR, the trade diversion mostly affects North America (USA, Canada, Mexico), Asia ( Japan, Korea, Indonesia, and Philippine) and Europe ( United Kingdom, France). 78 B. Credibility of The Results To check the credibility of the results against other relevant estimates, first of all, I did independent estimation of trade creation and diversion of ASEAN and EC by using import growth approach suggested by Kreinin (1981). Second, I compared the results of EC with the previous estimates of trade creation and diversion of EC. The import growth approach formulate an antimonde based on what import growth rates would have been in the absence of integration. Once such growth rates in the antimonde are estimated, it is possible to perform pairwise comparison in order to derive trade creation and trade diversion. The standard normalized approach estimates an antimonde import growth rate by using a control country as a normalizer. Pioneered by Kreinin (1972), this approach posits a control country’s growth rate in the antimonde. In other words, the growth rates in the integrating area are normalized by the growth rates of similar ratios in different countries over the same period. While Korea is used as normalizer To estimate the TC and Td of ASEAN, Unites States is used as normaizer for the estimation TC and TD of EC. 79 Table 5—3 presents the relevant import data for the ASEAN and Korea to estimate TC and TD of ASEAN in 1995. Table 5-3 Imports of ASEAN and Korea (1) 1989/90 (2) 1995 (3)Ratio average Col.2/Col.1 Total Imports 123 293 2.17 of ASEAN External 97 221 2.27 Imports of ASEAN Imports of 55 106 1.92 Korea Had total imports of ASEAN grown at the same rates as that of Korea(1.92 times) it would have been $236 billion in 1995. That figure represents hypothetical imports in the absence of integration. The difference between actual ($293 billion) and hypothetical imports - $57 billion - is the estimated annual trade creation of ASEAN. To estimate annual trade diversion, in the absence of integration the ratio of external to total imports would have remained at the base period level of 78%(=97/123). That yields hypothetical external imports of (293*78%) $228 billion. Trade diversion is the difference between this figure 80 and actual external imports in 1995 or (228-221) $7 billion. Compared to the gravitational results in Table 5-2, $53 billion TC and $2 billion TD in ASEAN 1995, these($57 billion TC and $7 billion TD) are very close figures. So the results of gravitational estimates are credible. Table 5-4 presents the relevant import data for the EC and USA to estimate TC and TD of ASEAN in 1995. Table 5-4 Imports of EC and USA (1) 1970/71 (2) 1985 (3)Ratio average Col.2/Col.1 Total Imports 101.5 527 5.27 of EC External 38.5 199 5.16 Imports of EC Imports of USA 39.5 186 4.70 Had total imports of EC grown at the same rates as that of USA(4.7O times) it would have been $474 billion in 1985. That figure represents hypothetical imports in the absence of integration. The difference between actual ($527 billion) and hypothetical imports - 81 $53 billion — is the estimated annual trade creation of ASEAN. To estimate annual trade diversion, in the absence of integration the ratio of external to total imports would have remained at the base period level of 38%(=38.5/101.5). That yields hypothetical external imports of (527*38%) $205 billion. Trade diversion is the difference between this figure and actual external imports in 1995 or (205-199) $6 billion. Compared to the gravitational results in Table 5—2, $75 billion TC and $11 billion TD in EC 1985, these ($53 billion TC and $6 billion TD) are close figures. The trade creation estimates of EC has consistency with other studies. Kreinin(1981) finds TC in EC $28 billion (in 1977/1978) and McConnel (1981) $27 to $37 billion (in 1977/1978) and Owen(1983) $40 to $96 billion (in 1980). My estimates of TC in EC (in 1980) is $87 billion which is in the range of Owen (1983). Previous estimates of trade creation and trade diversion of EC is summarized in the Table 5-5. 82 Table 5-5: Previous Estimates of Trade Creation and Diversion of EC ($ Bil.) Author Area Year TC TD Balassa EEC 1965 1.9 0.1 (1967) Kreinin EEC 1963 0.08 0.05 (1969) 1964 0.07 0.02 1965 0.04 0.09 Truman EEC 1968 9.2 -0.1 (1969) Willamson & EEC 1969 8.3 3.5 Botrill (1971) EFTA EEC 1965 1.7 0.6 Secretariat 1966 2.2 0.7 (1972) 1967 2.3 0.9 Kreinin EEC 1969/19 7.2 to -4.2 to (1972) 70 20.5 2.4 Verdoorn & EEC 1968 10.1 1.1 Schwartz (1972) Aitken EEC 1967 9.2 0.6 (1973) Kreinin EEC 1970 5.3 3.9 (1973) 1.7 Sellekaerts EEC 1972 -24.6 (1973) Prewo EEC 1970 19.8 -2.5 (1974) Balassa EEC 1970 11.3 0.3 (1975) Rensick & EEC 1968 1.8 3.0 Truman (1975) Truman EEC 1968 3.0 2.0 (1975) Kreinin EEC 1977/19 28.0 5.0 (1981) 78 McConnel EC 1977/19 27.0 to 7.9 (1981) 78 36.8 Owen EC 1980 40 to 96 (1983) 83 V. Summary and Conclusion This chapter uses the gravity model to estimate trade creation and trade diversion of ASEAN, ANDEAN, EC and MERCSUR. By using the gravity model as antimonde the expected bilateral imports are estimated and these estimates are subtracted from actual imports to get trade creation and trade diversion. This gravity model shows that the ASEAN and EC are trade creating blocs while ANDEAN and MERCOSUR are trade diverting blocs. From these results we can conclude that while the export-promotion policy with low rates protection by ASEAN and EC has fostered trade creation the import-substitution policy with high rates of protection of ANDEAN and MERCOSUR has affected trade diverting negatively. CHAPTER 6 GROWTH EFFECT OF EXPORTS EXPANSION I. Introduction There are strong logical and empirical grounds supporting the hypothesis that exports are a key factor in the growth process. The logical grounds can be documented in terms of both direct and secondary effects of exports on the economy. There are many direct benefits from a high export growth rate that help in promoting general economic growth. Export development tends to concentrate investment in the most efficient sectors of the economy—those in which the country enjoys a comparative advantage. Specialization in the products in which the country has a comparative advantage increases productivity. These benefits follow the traditional line of emphasizing specialization and reallocation of existing resources. In addition to these static effects of reallocating an unchanged quantity of resources, there are dynamic effects which are the increases in economic well-being that accrue to an economy because trade expands the resources of a country and induces an increase in the 85 productivity of existing resources. An increasing level of exports generally means that the country has the wherewithal to step up its level of imports. These imports include capital goods which are especially important in contributing to economic growth. The country is enabled to take greater advantage of the international division of labor, procuring desired goods from abroad at considerable savings in terms of productive factors. This helps increase the efficiency of industry, which is a major factor in economic growth. The country also gains from economies of scale, since the international market added to the domestic market obviously permits larger- scale operations than does the domestic market alone. The necessity of remaining competitive in international markets tends to maintain pressure on the export industries to keep costs low and to constantly strive for more efficient operations. The competitive pressures also tend to lead to improvements in the quality of the export product, and in general to inhibit the establishment of the inefficient export industries. In addition to direct benefits of providing part of wherewithal for economic development, and stimulating more efficient use of resources, a dynamic export sector also produces substantial secondary benefits. These include increased investments and technological 86 advancement elsewhere. Profitable export industries tend to stimulate additional investment, both domestic and foreign. Where exports of a primary product are profitable and expanding, there is a stimulus to domestic investment in both the existing industries and in the various processing industries associated with the product in its various stages of production. Expanding exports also encourage investment in ancillary industries set up to supply and service the operations of the main export industries. A rapid growth in exports also serves as an inducement to foreign investment in the country, particularly where the investment climate is propitious from the viewpoint of foreigners. In addition to stimulating domestic and foreign investment, a growing export sectors also encourages an increased flow of technological and market innovations, as well as managerial skills. Under the pressure of competition and the desire to continue expanding foreign sales, foreign techniques and methods are imported to further improve productivity and quality. This is beneficial for both the domestic exporter and the foreign importer, the latter often pressing for the new techniques in order to improve his own sales and profit position. 87 All of these factors tend to reinforce each other, stimulating further expansion of exports and investment. The results is a substantial growth in real GDP. Such export-led growth is important in many countries. This chapter attempts to measure the effect of export expansion on the economic growth. II. Mbdel Specification A model of the relation between export and growth will be based on the growth accounting equation. Assume the following Cobb-Douglas production function incorporating three factors: (6.1) Yi=A KFLFEg’ where )2 = country 1’s real GDP A = a technological constant K3= country i’s capital stock ln= country i’s labor force inputs E} = country i’s exports The third factor, exports, has been included to estimate the effect of export on the growth rate. As we mention in the introduction exports has been included on the grounds that there are scale effects and externalities associated with export production and 88 sales. Also following the international comparative advantage, ceteris paribus, exports can have independent effect on the output growth through the reallocation of existing resources. By differentiating equation (6.1) with respect to time and dividing through by (6.1) we obtain the following linearly estimable equation: 1,2 g g E, i 6.2 -—-—=—-+a-——+ + —— ‘ ’ Y,- A K) ’94- ’15,- where superscript 0 represents the change in the variable with respect time. This model was used in the development literature [ Emery(1967), Syron and Walsh(1968), Balassa (1978,1985), Tyler(1981), Feder(1982), Kavoussi (1984), Chu(1988), Fosu(1990) ] to study the relation between export and growth. Various exports-growth model formulations and estimations are summarized in table 6-1. The basic idea of this model is that the growth in export is an important factor for economic growth. It is also clear that capital formation, labor and technological growth contribute significantly to GDP growth. 89 Table 6-1: Summary of Various Exports-Growth Model Formulations and Estimations Authors IModel Formulation Estimation of Export Coefficient 1Emery(1967)1 y = a + bx 0.3295 (1c and DC) 'Syron and y = a + bx 0.3327 (Low-income iDC) 0.3871 (Middle- FWalsh(l968) income DC) a1assa(1978)2 AY = ak, +ka +c(AL)+d(AX') 0.04 (Semi- industrialized DC) lBalassa(1985)3 AY = ak, + bk , + c(AL) + d(AX) + eY+ flux 0.182(DC) yler(1981? y1 = a + bk + C! + ab: 0.57 (Middle-income) y2 =a+bk+cl+dmx lFeder(1982)5 y = a + bl + c(I / Y) + dx(X / Y) 0.422 (Semi- industrialized DC) [Kavoussi (1984)‘ y, = a+bk +cl+dx 0.105(all DC) y2 =a+bk+c1+d,mx+d2mx osu(l990) y=a+bk+cl+dr 0.123(Afn'can DC) Notes: [Cs and DCs are industrialized and developing countries, respectively ‘ y = GNP per capita growth rate, x = export growth rate 2 AY,AL,AX = changes in GNP, labor force, merchandise export , k, =domestic capital growth, It, = foreign capital growth 3 AY,AL,AX = changes in GNP, labor force, merchandise export , It, =sum of gross domestic investments less current account balances fi'om initial year terminal year, k, = sum of current account balances from initial year terminal year. ‘ y,lr,l,x= GNP, capital formation, labor, and export growth mx=manufactured export growth 5 y = GDP growth, I = labor forth growth, I /Y = investment output ratio 5 Similar to Tyler’s except mx =product of the share of manufactured goods in total merchant exports III. Data Empirical work which test the impact of exports on the GDP growth uses mostly 19605 and 1970s cross country data sets with relatively limited number of countries. Recently compiled international data set, Penn World Table (PWT Mark 5.6) is more accurate, more comparable between countries, and covers more countries than used in previous studies. The countries (total 152 countries) included in this data covers Africa (50) Central and North America (22), South America(12), Asia (32), Europe (28), and Oceania (8). Data Years are from 1960 to 1992. The data used in this chapter are: X = country i’s GDP growth rate Penn World Table 5.6 by Summers and Heston (1994) RGDPL(Real GDP per capita, Column 3) growth rate K} = country 1’s capital stock growth rate Penn World Table 5.6 by Summers and Heston (1994) KAPW(Non-residential capital stock per worker, column 20) growth rate L, = country i’s labor force inputs growth rate Penn World Table 5.6 by Summers and Heston (1994) Labor Force Participation growth rate £3 = counttry i’s export growth rate 91 IV; Empirical Results First, I report the results of the export-growth formulation using Penn World Table data. By pooling the cross section and time series data, I have the following result: Y.- A K, 1;, E, (6.4) —=-—+a—+fl—+y— Y A K L. E t i r 1 Estimates:4.519 0.313 0.579 0.2194 R-square 0.3277 t-values :(0.03)(3.21)23.27)(21.82) # of obs. 1546 As expected exports have a positive effect on the growth rate as in the previous studies: a 1% growth in export causes a 0.2194% increase in the real GDP. This estimate is in the middle of previous results which ranged from 0.105 to 0.57. Second, if we look at the results by continents, the growth of export in African countries has the highest contribution to the GDP growth (0.32), followed by the Asian countries(0.22), Oceania (0.17), Europe(0.16) and America (0.13). The export- growth results for each countries will be given at appendix D. 92 Table 6—2: Export-Growth Regression by Continents Contine Const . Capital Labor Export R- # of -nts Growth Growth Growth square obs . .Africa. 3.39 0.22 -O.lO 0.32 0.50 291 (7.33) (3.80) (-0.33) (14.93) America 4.03 0.50 1.07 0.13 0.23 400 (9.67) (7.08) (2.40) (7.94) .Asia 5.94 0.10 2.00 0.22 0.33 295 (10.31) (1.35) (3.00 ) (11.00) Europe .5.75 0.21 0.77 0.16 0.19 510 (14.94) (3.57) (2.37) (9.99) Oceania. 4.59 -0.51 4.06 0.17 0.32 50 (2.37) (-1.70) (1.58) (3.14) Parentheses are the t-statistic ‘V. Summary and Conclusion The main conclusion to be drawn from this study is that export expansion is crucial factor to economic growth. It would appear that countries ought to aim at 2.5 percent expansion of exports to obtain a 1 percent expansion of per capita real GDP. of the above conclusion is that countries eager to The policy implication increase their growth rates should adopt the type of policies that will stimulate exports. This suggests that countries which neglect export sectors and adopt the policy of imports substitution are likely to have lower economic growth. CHAPTER 7 GROWTH EFFECT OF ECONOMIC INTEGRATION I. Introduction It is customary to divide the effect of a customs union into static and dynamic. The static effect is concerned with allocative efficiency while dynamic or growth effect is concerned with the long-run growth consequences of increased market size. This makes possible production on a larger scale and infuses competition into markets. The fact that a multitude of factors influence the growth rate makes it difficult to assess the impact of integration on growth. .Also most of the previous studies of regionalism focused on the static effects and ignored the dynamic effects. The objective of the research reported in this chapter is to identify and measure these growth effects. Research on the growth effect of economic integration is rather recent and the literature regarding econometric evaluations on the growth effect are sparce. Coe and.Moghadam (1993) analyze growth effect of EC integration on France. Using the ratio of intra-EC trade to total EC output as proxy for integration, hours worked 93 94 in the nornfarm business sector, the stock of capital and the stock of R&D capital are used to explain the growth of French GDP. They finds that 0.3 percentage points of the French annual growth rate can be attributed to EC integration. Italianer (1994) analyzes the growth effect of the EC, using growth of capital stock, labor force participation, and an EC proxy (defined as intra-EC trade as a share of total EC trade). He finds that the EC- proxy is positively and significantly related to the growth rate of EC. In both studies, however, growth in intra-EC trade does not necessarily mean the effect of the EC bloc. It could be the result of natural factors, i.e., rapid growth in per capita GDPs or of the increase in economic size of the member countries. To investigate the extent to which regional policy initiatives influence trade flows and growth rate it is necessary to hold constant natural economic determinants. The gravity model offers a systematic framework for measuring the effect of bloc formation on trade flows. Hence, using the results from chapter 4 with a growth accounting equation, the growth effect of various economic blocs can be estimated. I will apply the cross-section results of the gravity model to a time-series model of a growth equation using the yearly results from 1960 to 1992. 95 This is a new and more correct approach to measuring the growth effect of economic integration. II. Model Specification A model of the relation between export and growth which was used in chapter 6 will be used basic model to measure the dynamic effects of economic integration. E . 1,1 12 _L,'_ i ‘ l (7.1) —=—+a—+ +7— Yi A Ki flLz' Ei where )2 = country i’s real GDP A = a technological constant K3= country i’s capital stock In: country i’s labor force inputs E3 = country i’s exports and superscript 0 represents the change in the variable with respect time. By replacing the export growth rate in (7.1) with the cross-section results of the gravity model, we can identify the effect of economic integration on the growth rate. The following equations will be estimated: 96 (7 2) —.L=fi+a£+flfi+y— ’ Yi A Kl. Li B where.B= regional effect from the gravity model The basic idea of this model is that the growth in export is a important factor for economic growth and the trade integration has effect on the export growth which results in the growth of GDP. Therefore the regional effect derived from the gravity model can be used to estimate the dynamic effect of block formation on the growth rate of the national economy. III. Data The countries included in the analysis of the growth effect are North and South America (Western Hemisphere), East Asia and Europe(total of 60 countries)1, the broader continent-sized groupings that are under discussion. Even though the years covered by the gravity model were 1961- 1995, only data from 1961 to 1992 were used for analysis because data from the Penn 1 See Appendix C for the countries involved in the analysis of chapter 5. 97 World Table (Mark 5.6) were only available up to 1992. The data used in this chapter are: K = country 1’s GDP growth rate Penn World Table 5.6 by Summers and Heston (1994) RGDPL(Real GDP per capita, Column 3) growth rate K} = country i’s capital stock growth rate Penn World Table 5.6 by Summers and Heston (1994) KAPW(Non—residential capital stock per worker, column 20) growth rate Z, = country i’s labor force inputs growth rate Penn World Table 5.6 by Summers and Heston (1994) Labor Force Participation growth rate .3 ==regional effect growth rate from the gravity model Gravity Model Estimation in chapter 4. Data Years are from 1960 tol992. Bilateral export data were collected from the International Monetary Fund (IMF) Directions of Trade. IMF International Financial Statistics provided the GDP and Populations of the world. Most of the growth equation data were collected from Penn World Table 5.6 by Summers and Heston (1994).2 2 . I 901‘— this Penn World Table 5.6 by Summers and Heston (1994) from internet. 98 IV} Empirical Results .A. Growth Effects of ASEAN, ANDEAN, and EC As we discussed in chapter 4, the gravity model offers a systematic framework for measuring what patterns of bilateral trade are “normal” around the world. In addition the coefficients of regional dummies offer the effect of the bloc formation on trade flows. The results of these estimates will be grafted on to the growth accounting model to see how these blocs will affect of the growth of the region. The gravity model used in this chapter is log(X,.) = a + ,6, log(GDR ) + ,6,1og(GDPj) + ,B,log(DISTANCE :1) 7.3 ( ) + ,8,log(AwACENT,) + ,6,log(Bloc) where Bloc: dummy variable for ASEAN, ANDEAN, EC To get enough data points for the growth equation, I chose ASEAN, ANDEAN and EC as regional group and the estimates of these bloc are reported in the Table 7-1. MERCOSUR was excluded because it was significant only in one data year 1995. Because I report only the bloc effect which is significant at least at the 90% level there are some missing years. These results will be grafted onto the time-series data for growth analysis. 99 Table 7-1: Cross-Sectional Results (05) form Gravity Model Year ASEAN .ANDEAN EC 1961 0.710227 1962 0.767669 1963 . . 0.773136 1964 . 0.71706 1965 . 0.730029 1966 0.696455 1967 . 0.643965 1968 1.278732 0.752288 1969 1.805311 1970 0.903383 . . 1971 2.215933 -0.9844 . 1972 2.154016 -0.74398 . 1973 1.866969 -0.73991 . 1974 1.945677 -1.49944 . 1975 1.607181 -0.91525 0.428787 1976 1.662562 -0.94935 0.488971 1977 1.62151 -0.88977 0.510943 1978 1.558911 -0.97717 0.535997 1979 1.741579 -0.95377 0.606116 1980 1.785182 -0.98084 0.534293 1981 1.371986 -1.03618 0.423394 1982 1.300219 -0.82411 0.371004 1983 1.310376 -1.25841 0.372246 1984 1.409654 -1.5406 0.35264 1985 1.56851 -1.44854 0.457558 1986 1.584085 -1.45581 0.687565 1987 1.636484 -1.18801 0.674339 1988 1.623224 -1.46756 0.614078 1989 1.825043 -0.51665 . 1990 1.783225 . 1991 1.581453 . . 1992 1.556086 . . 100 As seen in the table 7—1, while the EC and ASEAN have positive effect on the export, the ANDEAN has negative impact on trade flows. This negative regional effects from the ANDEAN represents significant trade diversion. Next, I run the growth (time-series) regression using the regional integration results (cross-section) from the gravity model. Pooling the cross-section and time-series data yields Table 7-2 results. Kl. Ll. =—+a—+ —+ — fiLi 7 (7.4) —1' Y. .A KC .8 z I Table 7-2 Bloc-Growth Regression Eq.(7.4) Variable ASEAN ANDEAN EC \Bloc A 7.8163 4.7258 5.9067 (6.86) (4.91) (18.09) K 0.4431 0.8621 0.5333 (2.08) (4.28) (10.59) L 0.0619 1.93 1.75 (0.02) (1.26) (5.05) B -0.0237 -0.0660 0.0315 (—1.33) (-3.27) (2.89) R-Square 0.11 0.27 0.16 # of Obs. 44 90 146 Parentheses are the t-statistic 101 As shown in Table 7—2, while the EC shows positive effect on the growth rate of their member countries, ANDEAN shows the negative effect. The ASEAN does not show any significant effect on the growth by the bloc formation. The positive estimate of EC suggest that 1% growth of intra-EC exports form bloc formation causes 0.0315% increase in real GDP. The negative estimate of ANDEAN indicates that 1% increase in the growth of intra- ANDEAN exports cause 0.066% decrease in the real GDP growth. From this result we can speculate that EC is trade-creating bloc which is favorable to growth and that ANDEAN is trade-diverting bloc which is unfavorable to growth. Because R-square was low compared to the export- growth regression of Chapter 6, I tried several other specification. While ASEAN does not show any significant effect, EC has significant positive effect and ANDEAN has significant negative effect on the member countries. First, I added the growth rate in export of the individual countries. (7.5) NH Table 7-3 Bloc-Growth Regression Eq.(7.5) Variable ASEAN ANDEAN EC \Bloc A 5.5753 3.6224 5.09 (4.77) (4.06) (2.25) K 0.3773 0.7682 0.2628 (2.23) (4.23) (1.90) L 0.0083 1.5347 0.5185 (0.01) (1.18) (0.83) E 0.1788 0.1528 0.2001 (3.07) (4.73) (5.86) B -0.0253 -0.034 0.0292 (-1.12) (-1.77) (2.25) R-Square 0.31 0.42 0.22 # of Obs. 46 90 146 Parentheses are the t-statistic As expected, the export growth have a positive effect on the GDP growth rate. Both Band E can go into equation (7.2) because 8 is intra-regional export growth effect from the gravity model while E is individual country's export growth. For the problem of multicollinearity, as we discussed in Chapter 4, we don’t have to worry because R—square from the regression (0.22 to 0.42) exceeds the R-square of these two variable regressed (0.0016 to 0.06). While the EC shows positive effect on the growth rate of their member countries, ANDEAN shows the negative effect. The ASEAN does not show any significant effect on the growth by the bloc formation. Again, from this result we can speculate that EC is trade-creating bloc which is NB favorable to growth and that ANDEAN is trade-diverting bloc which is unfavorable to growth. Secondly, I added the growth rate of total trade (exports plus imports) of individual countries. Again the bloc effect of ANDEAN is negative while the effect of EC is positive and ASEAN is not significant. It might be concluded that regional bloc of ANDEAN is harmful to the GDP growth of their members. Y. .4 K. 1;. r', B (7.6) —=—-+a—-‘-+fl—L+6—+y— Y, A K L T B i i I where T is trade (export +import) Table 7-4 Bloc-Growth Regression Eq.(7.6) Variable ASEAN ANDEAN EC \Bloc A 4.3748 2.8452 4.58 (3.75) (3.12) (7.56) K 0.3824 0.7778 0.2458 (2.45) (4.48) (1.92) )L 0.2232 1.5607 0.6793 (4.30) (5.62) (1.18) T 0.2232 0.1872 0.2452 (4.30) (5.62) (7.97) B -0.0252 -0.034 0.03437 (-1.25) {-1.65) (2.86) R-Square 0.42 0.46 0.33 # of Obs. 46 90 146 Parentheses are the t-statistic 104 Lastly, I tried to use several other variables, like population growth rate and investment growth rate. While the R-square increases, the bloc effect of ANDEAN remains as negative and EC as positive. Still ASEAN does not show any significant effect. Further the investment growth has a positive effect. .4 '. 7,. B Y. K. .1. I. P 7.7 ——=—+a—-’-+a—-’—+ ——’—+ —’+6—+ — ( ) Y; A K 1. 'BL 'BP T 713 i t i i i where I is investment and P is population Table 7-5 Bloc-Growth Regression Eq.(7.7) Variable ASEAN ANDEAN EC \Bloc A, 5.6271 -0.8543 4.9146 (1.33) (-0.22) (8.51) K 0.4772 0.7663 0.3052 (3.58) (4.42) (2.44) I 0.2139 0.1158 0.1171 (4.55) (3.51) (4.76) L -.8583 0.7726 0.6501 (-0.35) (0.58) (1.19) P -.4011 1.7348 -0.5102 (—0.24) (1.13) (-0.88) T 0.1632 0.1519 0.2213 (3.67) (4.63) (7.51) B -0.0252 -0.034 0.0221 (-1.67) (-2.08) (1.92) R-Square 0.62 0.54 0.42 # of Obs. 46 90 146 Parentheses are the t—statistic 105 B. Growth Effects of 3 continent-Sized Grouping While section A focuses on the regional blocs that are already in existence, this section analyzes the growth effect of the broad continent-sized groupings that are under discussion (the Americas, Europe and East Asia). In this section I use the broader continent-sized grouping which divides the 60 countries into 3 continent- sized grouping, the Americas, East Asia, and Europe. The cross-sectional results of the gravity model will be grafted on the growth accounting model as in section B to see how these 3 continent-sized grouping will affect on the growth of the 60 countries overall. Because I report only the bloc effect which is significant at least at the 90% level there are some missing years. As seen in the table 7-5, while the European and East Asian have positive effect on the export, the Americas (Western Hemisphere) has negative impact on trade flows. This negative regional effects from the Western Hemisphere represents significant trade diversion. Sixty countries are used to estimate for this gravity model involves in the Americas, East Asia, and Europe. So for each year, the results are based on 3422 bilateral trade flows between those 60 countries. HM Tablg)7—6: Cross-Sectional Results (85) form Gravity Eq.(7.2) Year WHg East Asia Europe“ 1961 -0.55894 0.599458 1962 0.4243 0332419 1963 -0.36684 0.263489 1964 024903 1965 -0.32861 . . 1966 072137 0.331027 0452826 1967 -0.67419 0.519714 0.316495 1968 -0.57849 0.373262 0.24313 1969 4.0537 0.552019 0.634534 1970 -0.97916 0.561083 0.647863 1971 105572 0.555645 0.489642 1972 -1.1299 0.468725 0560499 1973 102209 0.637404 0.581412 1974 -1.0556 0.592496 0.597018 1975 4.05253 0.58355 0.704656 1976 -0.98131 0.601357 0.698752 1977 -090957 0.577845 0.650634 1978 095115 0.529319 0.647793 1979 -1.08554 0.473935 0.857138 1980 -1.06426 0.383767 0.828112 1981 -100513 0.494312 0.705946 1982 094917 0.463394 0.691483 1983 -0.87907 0.471411 0.596273 1984 -0.84129 0.450836 0597431 1985 099354 0.424192 0.710772 1986 -1.17258 0.318773 1.044629 1987 -1.20856 0.382472 1.100312 1988 -1.20645 0.388737 1.023031 1989 0.2029 0.877039 -0.19738 1990 -0.13831 0.883637 -0.26512 1991 0.828377 -0.43284 1992 0.884387 0.44559 3 “H1 stands for Western Hemisphere including the North and South Amernuh ‘ Europe consists of EC(12 countries) + EFTA(G countries). See Appendix C for the countries involved in the gravity model. 107 Next, I run the growth (time series)regression using the regional grouping results(cross-section) from the gravity model. Pooling the cross section and time series data yield following result: K. L. 1', .i -A l l —=—+a—+ —+ — Y1. A Ki flLi 73 (7.8) Estimates: 5.732 0.541 1.876 ~0.0069 R-square 0.1660 t—values : (18.81)(10.91)(5.52)(-2.66) # of observation :706 The most startling result is that the 3 continent- sized grouping has a significant negative effect on the growth rate in three continents (the countries analyzed here includes the North and South America, East Asia, Europe) over all. The estimate suggests that the 1% increase in the intra-exports among three continents- sized groupings has the negative effect on the growth rate of the real GDP by 0.0069%. I tried several other specification by adding several other variables. But all resulted in the same negative effect of regional integration on real GDP growth rate. This negative growth effect of three continent-sized groupings is consistent with to Krugman(1991a) where he finds that world welfare is minimized when there are three blocs. 108 First I added the growth rate in export of the individual countries. As expected, the export growth has a positive effect on the GDP growth rate but regional bloc still have a negative impact. K‘. i. E. +a—l—+,B—'—+6—L+y— (7.9) “i*=_‘ Y. A K. L. E. B 1 1 1 1 Estimates: 3.927 0.473 1.434 0.144 -0.0064 R-square 0.2884 t-values : (15.17)(lO.19)(4.54)(10.82)(-2.67) # of observation :706 Secondly, I added the growth rate of total trade (exports plus imports) of individual countries. Again the bloc effect is negative while the trade growth has a positive effect on the Real GDP growth. It might be concluded that the regional groupings is harmful to the GDP growth and free trade is superior policy for growth. K B Y. A 1; 1 =—+a——‘—+ —‘+6—'+ — A ’91. 7 (7.10) —— Y. K. . T. B where T is trade (export +import) Estimates: 3.927 0.451 1.402 0.183 -0.0058 R-square 0.3562 t-values: (13.03)(10.08)(4.55)(14.08)(-2.54) # of observation :690 109 Thirdly, I tried to use several other variables, like population growth rate and investment growth rate. While the R-square increases, the bloc effect remains negative. Further the investment growth has a positive effect and population growth has negative effect on real GDP growth rate. (7.11) €—=:4—+a£+a'1i+fl£’¥+fl—i+§—+y— where I is investment and P is population Estimates:4.589 0.457 0.028 1.274 -0.036 0.180 -0.0055 R-square 0.3679 t-values: (10.91)(10.28)(2.86)(4.07)(-2.15)(13.60)(-2.41) # of observation :690 Lastly, I ran growth regression by regional grouping. As shown in table 7-6, the European grouping has a positive effect on the growth of their member countries. The regional grouping of America (Western Hemisphere) has negative impact on the GDP growth within the region. East.Asian has not significant effect on the growth of their region. Table 7-7: Growth Effects of Regional Groupings 110 Model/ (7.7) (7.8) (7.9) Continent s America -0.0061 -0.0053 -0.0050 (-2.14) (-1.96) (-1.86) East Asia 0.0155 0.0117 -0.0031 (1.01) (0.77) (-0.23) Europe 0.0026 0.0026 0.0031 (2.16) (2.21) ( 2.63) Parentheses are the t-statistic V; Summary and Conclusion The fact that a multitude of factors influence the growth rate makes it difficult to assess the impact of regional integration on growth. However gravity model offers a systematic framework for measuring the effect of regional bloc on trade flows. So the results of these estimates were grafted on to the growth accounting model. The results of bloc-growth regression shows that effect of ANDEAN is negative. These negative results show that the ANDEAN bloc is harmful to the member country's growth. However, EC has positive effect on the growth of their members and EC integration has contributed to the EC's growth positively. The most startling result is that the 3 coninent- sized grouping has a significant negative effect on the world growth rate(60 countries on America, Europe and 111 Asia) even though the effects of regional groupings on the intra-trade flows are positive. This suggests that the regional integration is not the best policy for world growth. This empirical result supports the fundamental proposition of the General Theory of Second Best. Free trade in commodities maximizes world welfare in distortion-free world. However a world ridden by multiple distortions (e.g. tariffs, quotas and exchange control) will not necessarily be moved closer to Pareto Optimality by the removal of one distortion. Counter- intuitively, we may move away from Pareto Optimality. The regional integration which involves a partial movement toward free trade is not necessarily Pareto improvement. Regarding the dynamic issue whether the regional movement is the building or stumbling bloc, this results shows that the world that divided into three blocs can be the stumbling bloc to world growth. Further, the result that the total trade growth has a positive effect on the ‘world growth rate suggests that we have to focus more on free trade. A ndixA° n 'esinvlvedinG vi NAFTA 3 USA Canada Mexico Western Hemisphere (WH 28) : EAI (Enterprise for American Initiative) USA Bahamas Bolivia Colombia Ecuador Guyana Jamaica Panama Suriname Venezuela EC(12) Belgium-Luxembourg Germany Netherlands Sweden BM 15) = EC(12) + Austria, Finland and Greece TAFTA=NAFTA+ EC EFTA Austria Iceland East.Asia Japan Indonesia Myanmar Thailand Canada Barbados Brazil Costa.Rica El Salvador Haiti Mexico Paraguay Trinidad.Tbg. Denmark Ireland Portugal United Kingdom Finland Norway China Korea.RP Philippines 112 Argentina Belize Chile Dominican Rp. Guatemala Honduras Nicaragua Peru Umstmy France Italy Greece Switzerland Hong.Kong Malaysia Singapore Other Pacific Australia APEC = NAFTA + East Asia + Other Pacific + Chile Other Asia Bangladesh Laos Pakistan S1i.Lanka Middle East Cyprus Bahrain Israel United.Arab.Emir Algeria Burundi Chad Cote.D'ivoire Gabon Guinea Madagascar Mauritnia Mozambique Rwanda Sierra Leone Tanzania Uganda Zimbabwe 113 New.Zealand Fiji Mongolia Papua.N.Guinea Malta Egypt Jordan Saudi.Arabia Benin Cameroon Comoros Djibouti Gambia Guinea—Bissau Malawi Mauritius Niger Senegal South Afiica Togo Zaire India Nepal Solomon Islands Turkey Kuwait Syrn.Arab.RP BurkinaFaso CentralAfizRP Congo Ethiopia Ghana Kenya Mali Morocco Nigrria Seychelles Sudan Tunisia Zambia 114 Apmndix B: Trade Creation and Diversion of Trading Blocs ASEAN :1970 ( 8 Million) Importer Exporter Actual Imports Projected Imports GTC TD lndonee Argent 0.01 1 .94 -1 .93 Indonee Anabel 38 7.38 3167 Indonee Auntie 1 .8 1.3 0.41 Indonee Balm 0 . . Indonee Berbedo 0 0.04 0.04 Indonee Beuum 22 23 19.64 Irrdorm Bebe 0 . . Iridonee Bolivia 0 0m -0.m Indonee Bra! 0 1.82 -1 .82 Indonee Canada 0.5 394 3.44 lndonee ChIIe 0 0.54 -0.54 Indonee Chhe 0 3.6 -3.6 Indonee Colcmbl 0 0:5 «0.5 Irtdortee Cceta.R 0.01 0.128 -0.07 Indonee DOMINIC 0 0.07 «0.07 Indcnee Datum 13.6 1.79 11.70 Indenee ELSALV 0 cm are Indonee Ecuador 0 0.1 1 0.1 1 Indonee FInImd 1 .13 1.3 0.19 Indonee France 22 9.23 1277 Indonee Germy Q 1123 87.77 Indenee Greece 0.31 0.88 0.5 Indonee Guetema 0 0.13 -0.13 Indonee Guyem 0 cm «0.02 Indonee Hell 0 OTB -0.(B Indcnee Honda: 0 0.04 -0.04 Indonee Hono.Ko 28 1 21 24.79 Indonee 1001.10 0.113 are 0.5 Indenee Irelmd 0.71 0.41 0.3) lndonee ItaIy 19 7.67 1 1.3 Indonee Janice 0 013 «am Indonee Jepm a 21.18 614.82 Indonee Kore-R 19.84 1.04 18.!) Indcnee Novel 0 2.18 -2.18 Indcnee Medco 0 2CD am Indcnee Myerrmer 0 0.20 0.20 lndonee Nether! 6 3.31 6.9 Irrdortee NewZee 1.04 1.42 0.3 Indonee NIcerw 0 0.07 -0.07 Indonee Norway 1.6 115 om Indonee PAPUAN 0 0.21 0.21 Indonee Patents 0 013 one Indonee Perm 0 0.04 -0.04 Indonee Peru 0.8 0.3 0.41 Indonee PI“ 31 1.6) 29.6 Indcnee Portage 1.24 0.51 0.73 Indonee Slngepo 0 1.6 -1 .6 Induces Speh am 3.31 1.78 Indonee Saturn 0 0.04 0.04 Indonee Sweden 282 310 0.2 Indcnee am 10.72 288 7.84 Indenee TRINIDA 0 018 -018 Indonee Taiwan 11% 1.12 10.74 Indonee Thelen 19.25 2% 17.19 Irtdenee USA 182 33.13 1691 Indorm UIM. 1 7 8.51 8.0 fifiéifififififiififiiEEEHEEHEEEEHQEEHHEHHEEQ$133353ii §§¥§EEE§§§EEE§§ CoeuR DOMINIC S ELSALV E‘fiig’é’ifi iii firms: NewZee Wining: TRINIDA gasizgiéfi 115 0.13 6.5 7.16 418 1 18 1 .4 5.% 0.1 4.56 1 .5 0.12 4.87 §§§§§§§§§§i3§i§i§§iiiiiiiiiiiiiiiiiiiiiiifiiifiiiiiiiiiiéiiiiiiiéi §§¥§E§E§§ CoateR DOMINIC 3 ELSALV iiéiifiiiiiéiifiigfii SEE PAPUAN 11313313 TRINIDA TaIwm 5? Urflted. UNIS-Y Venezue §3¥5§E§Eii 116 3.11 10:49 11% 2.37 21% 3.& 18 0.22 8.18 19.4 0.13 3E) 0.11 2.04 8.47 8.13 0.56 0.13 0.11 2.77 0.13 0.17 1 .24 4.87 0.24 4.3 3.5 1 2.9 0.31 1 .(B 0.33 3.24 0.74 0.02 . 1.28 0.04 2C5 12:76 0.02 3.34 -0.04 1.10 17.5 11111111133iiiiiiiiiiiiiiiiiiiiiiiifiiiiiiiiiiiiiiiiiiiiiiiiiiiii gr: 3 117 1227 8.14 13 0.3 0.01 0.41 ~99 p a or Opp 8220882=8o880888 800000 N dooouoo mi 2.5 7.13 0.78 8.07 0.81 0.14 0.1 2 3.6 -0.1 2 4.5 -8.13 0.78 -8.07 «0.81 -0.1 4 -0.1 2 -1 .20 iiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiii?ii 3 e 118 0.5 0.27 12 0.5 0.04 0.01 0.13 0.27 0.1 1 0.“ 3m 1 52 8.13 0.07 8.04 0'27 29.45 18.11 0.37 1.19 3.11 0.37 -1.17 119 ASEAN :1975 ( 3 Million) GTC TD AcmaIImpom Projectedlmpom m .9 1 1 -8 0 .1 .1 -12 -1 1 -2 -19 -2 -1 -1 1 -1 -1 -7 -x 91 .4 -1 0 0 0 18 -1 -1 .7 -1 $6 94 so -12 5 1w .5 0 .4 -2 0 0 -2 22 3 Q1 -1 1 0 -12 -1 1 fi .3 -8 1 754 -2 -1 -5 9&8 on crumplwalal18117$“5100081.21armounalfinlosn‘ooalmaunmounava 1 3 cl 112 33 1 5 1 One/1.000"0009000009ooowmauooooanlaMOMMMOGMalozoooomom7oz1aoammoo 1- Emma:m...mmmmaaWm.“mmmmmwmmwmmmmmm mmmmmEmmmmmmmmmmmmm-mmmmm 120 1&7 0% 06$QM0040045QW14..0006JBOEOMMJsmm024204$3uQ0W4dwwmmose13 0.... 11 7&7 om omW2W2017016$$410008152won7w2w505100283360um18%”815938 0“ 8440.0 0.04 0 2007001 50000 0 200 37 5030201 6 0 60 925105. 50.02 Emma. m Wmmmmmmm m“mammmmemmmmmummmmmm mmwmmmmmmmmmmmmmmmmmmmmmmmmmmmmmm 1H .1 .10 426344344386140009443$4WM9M2$5434004338054JflJM®434BQ 0w 0M30$ 3119118QQ51000w1W3$1$MB417716110346m0wafi¢w1nnmm166n5 .07 .0731” 1 0.01m350006003 $m20000704000 mano3nqfil~o30m0mwmnonnoawmmogom3ao70.02 2&1 6 1 APUA. N WWW SW SW mmewmwwa TRINIDA mmxmmmmwmam m: Emmmfl £me M mm:EmMmmmm:mmmmmmmgamma“:maximum 122 9004003 4000 0 440 7 70 04 30035... 1 «mm m m mom mes yea a 1 1 mma 121 .9 -5 -2 0 15 -1 -12 -19 -2 -8 3 1 -1 .3 -1 -1 10050155m310006044l2m0$5$7140404100162un09816mmfl139m40l .05 .13523311w11 1 woow01«(ammluoo0030”1Romfiw“flamom$&05$5”onmoxmwmnloogBoomooW6Oflmm0“00 m w :wmm MMMWMMWWWW WWW TRINIDA mm mamm E.“ .mm m: Ham m m m 0 WW m wmmwwwwwmmwmmwwmmmwwwmwwmmwwMWWWWWWWWMMWWWWWWWWmMMMMMMMWMMMMMMMM 123 eems4ooom4m mm -2 -3 -1 812 -14 5&541aoosmemmosoammn4s gese1ooou1nas1 nauams1711ozn4gaommzuma1e 1 1 1 mmmmmmwwmmWmmemmmmmwmm mEmmmmmmmEmmmmmmmmmmm GTC TD Projected Imports 124 MMWmm ASEAN 21980 (S Million) -23 57 -19 32 -1 3 0 -1 101 32 a 58 a -1 -1 22 -1 3 -1 7 15 282 -8 -2 0 0 -1 1 21 -2 4 1 m -1 270 8 -m -2 377 88 -1 -1 2 3 0 -1 -5 157 0 246 -7 0 -22 -§ 731 -8 1 36 so 3 -1 4 -7 1517 8&W11301flm8$811w13flmmu2001m25m1M%&$33M1W3115%83M03N43u5mamm oaoaosoomaooooooooomw4ooOOQOaIEO mam Q15 “@05010018mu082m05mm003 1 MMMMMMMMMMMMMMMMMMM MMMMMMMMMM MMMMMMMM MMMMMMMMMMMMMMM MMMMMMM MMu mun mmu Indonee MMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMM 125 154 47 .1 -1 94 o -1 .17 13 -2 179 <1 -1 .1 3 -1 -1 .15 364 can .4 -2 o o .1 177 -2 -27 5 we -1 32m 40 ~23 16 m1 40 -1 o o 2 -1 a 173 4 31$ 103 41 .4 .31 70 2344 -14 .14 13 o 2 19 o sw11mo1assme11M1anmmnzoo1szasm1mms1au1n2111zamnomn1mmmmaumaw1omo 9 00 00713 300 022 3000020 0 0 2 072301 2 0 009 103 52 0 mm m 11 a a mu m 313 wm1mmm 1 m1mm nm 5mm 1mmo a MMmmwMMMMMMMMWMWMM :MMmMM :MMM MmMmmMMMMMMmMMMMMMMM MM MMMMM :MMM EmmmmmmmmmmmMmmmmmmmmmmMmmmmmmmmm 126 1 4ou¢m544n4aammqqooom4uen4mmsaemao41w435emos4eaammqmumzaoso4amnmw 1zs1ns11w12uamm2001a134n1manmzmwow2111enxomaasnmnzuncn1ono1mmaaa 1. 11 1 o 04 4 44 4 1o 3 11- 300 00 4 9 ommummoumooummauoaamamwmownma ma 1umoamnaamwa mama «M1 n mm1n1 ”Mmmwwwmm mmm meMmmmumewmmmw PAPUA. N Ea Wmm ”ZEN “:3 Emma. mmmmmmmmmmmmmmmmmmmwmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmwwwwmwmmmmwm IN 21 -1 2 3 265 531 113 -1 o o o 1169 -1 337 7 as o 1347 274 2756 21 5 181 343 o 62 134 216 -1 7 257 15 a) o m 43 -2 -19 815 1773 479 5 -8 -2o 10 -2 -1 -1 76 o -1 -1 1 .4 -2 53 .5 -1 -1 1 -1 <3 -14 11H12m$fl8100021Q3$0 G5 ma 2380H1113wSEOWWZQEM$28mmm11&O1$$5m611fl13w O1&047mflm0 :Oom ammo mflmemmfl w:m§09$00w%fi7ommwmomoomaummoomoos 4| 1 4| PAPUA. N mmmwamwmmm EL. SALV Rum WWW Wmm TRINIDA Tmm mg m .E ammm 2%..me mmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmm m USA mm; mm.R DOMINIC iii???2311131117???§§§§§§E§E§E¥EE§EEE§§E TRINIDA Wmn USA United. WWW Wmm 1% 610 3.9 0.1 31 7 0.2 0.9 151 115 121 14 $38882a§m8~8ao -L d Q838805808AA-MBa d d 011.18 111 noi$§338§9§Lfim§b§$oohé§ @555: L m 33828583o 129 ASEAN :1985 ( $ Million) M__ __ ___ M - ,2 ._ __ 69399! Immijmm_ _ -_ , GISTQ_ Indona Argent! O 23 -23 lndonos Austral 179 1:3 51 Indones Austria 16 26 -1O lndonas Bahama O 2 -2 Indore: am 0 1 -1 lndones Edgium 54 38 16 Indones Belize 0 0 O lndones 601M: 0 1 -1 lndones Brazil 8 46 (5 Indona Canada so 66 -26 lndona Chic 0 6 -6 Indones China 3:!) 1 19 21 1 Indomc Colombl O 6 <8 lndones Costa R 0 2 -2 lndona DOMINIC O 2 -2 lndonos Dentnark 3 28 -25 lndomo ELSALV O 1 -1 lndonu Ecuador 0 3 -3 Indona Flnhnd 1 26 -5 lndones France 176 144 32 lndoms Gummy 33 1% 217 lndoms Gmm 3 18 ~15 lndoms Guam: 0 2 -2 lndanu 0 O 0 lndons Hui O 1 -1 lndoms Honda: 0 1 -1 lndona Hong.Ko 151 64 87 Indons Iceland 0 2 -2 lndoms Ireland 2 9 -7 Indone- lUy 247 133 114 lndoneo W O 1 -1 lndona W 10192 465 9737 lndonu KamaR as 54 615 lndones W 13 w 43 lndoneo Medea 1O 6 S lndonu Myanmu 4 6 -2 Undone; Nathan 20 5) 190 lndones M29: 75 23 52 lndones Nissan 0 1 -1 lndaneo Nanny 2 27 -6 Indiana: PAPUAN 0 4 -4 m Pm 0 1 -1 lndoms Fungus O 1 -1 lndonos Pom O 5 -5 lndona Philip 1% 3) 1w lndonu P011119: 2 1 1 -9 Indones Shown O 84 -84 lndones Spain 76 5 23 Indonu Saturn 0 O O Indonu Sweden 24 44 .z) lndonos 81141121! 26 3 .1 1 Indonu TRINIDA 312 4 38 lndona Taiwan 54 so 2% lndonu 171911311 618 64 3 Indonan USA an 06 4447 Indonu Unhd 2G5 130 73 lndonas Uruguay 0 3 3 lndona Venezue O 15 -15 130 -7 161 23 -1 o 63 o -1 -1 43 .2 1m .2 -1 -1 5 -1 -1 -15 211 531 .5 -1 o o -1 117 -2 .44 2 92 o as 1 1m 1111111111w1m11111mam1 111111 mmm1oxo1au4ws1«In12mmmm1001$2m7m1wca7mmdump/.114n9mc03330mam2nafin1ow gmeoomoommzm4007015ma80000m0m9m1 mmsnwmoz450132mwoamowwmm11mw SMMW in? fig :31” ”:1 ammfimmmmmm 11me11me wmmwwwWWWwwmwmwmwmWWWmwmwmwmwwwmwwmwwwmwwmwmwmwmwwwwwmwwmwmmmmmm 131 o -1 -15 :9 1 14 .4 -1 -1 6 o -1 -1 ‘ flme1.1000WindsMowmnmam70317046m1h0504wmmm1u£8m640&04$m14m o1n12m411n12n$n81000m1m4aomfimm3290n11025MaomnzmmmgnlwlumwalomOalnA‘an 1121mmm1mmmm"mw11111311111m1m11m 115mmmmmamgnm 4| mmmmmmmmm00$oowm7w 1 6.2 PUA N mmwwmmmm :mmmm E15mmMW“firmmmmmmmmmmm A 1mm1111mm31111131131111mmmmEmmmmmmwmmmmmm 132 124 -7 -1 -1 6 -1 .3 m44fid?Amamdooomomflmommmaammoommd1m1commammmm14%€340$04433 111m1zmamn1oooa1msnommmu1muow111amomosnzsmmu11nns11xo1sn1m71zm13 Wmmw mawm m mmmmm 11111131.. Emflmmammmmw EmEmmmmEmma:awning?MEMEEMMEM 133 anfi4004measmommmmfiwadn44 4 Azamaoemammean 25 134 156 17 O O 1 88 2 E 8 1 433 ammxau1mz111mmemoaxasmm2 1 111111111 1111111111111111111111111 11.111 1111111111.11.111.1111111111111111111:1111 ASEAN : 1990 (3 Million) 134 Impottet Exporter Actual Impom Projected Imports GTC TD lndonee Amend 1 S -34 lndonee Austral 40 175 274 Indonee Auntie 47 62 -15 lndoneo Baum O 2 -2 Indonee Berbedo O 1 -1 lndonee Belgian 199 70 129 lndonee Beta: 0 O O Indonee Bolivia 0 2 -2 Indonee Brazil 18 75 57 Indonee cm 191 1 13 78 lndonee Chile 5 10 5 lndonee China 80 130 719 Indonee Colombi 4 9 .5 . lndonee 00:13.1? 0 3 3 lndonee DOMINIC O 3 -3 lndonee Denmuk 54 55 -1 lndonee ELSALV 0 2 -2 lndonee Eouedo: 1 4 -3 lndonee Fitiend 15 61 -46 lndonee France 53 25) 239 lndonee Gennany Q1 :1)? 614 lndonee Gleeoe 9 31 -22 Indonee Gmm 1 3 -2 Indoneo Guyana 0 O O lndonee Hell 0 1 -1 lndonee Hondure O 3 -3 Indonee Hong.Ko 579 97 62 lndonee loelmd O 6 -6 lndonee Ireland 5 21 14 lndonee ltfly 437 246 161 lndonee Janice 0 2 -2 lndonee Jun 12744 85 118$ lndoneo KoreaR 1d!) 134 1466 lndonee Meleyei 316 126 190 Indonee Modoo Z) 6 -§ lndonee Myanmar 4 32 .m Indonee Netheu 518 90 a Indonee NewZee 7B 31 47 lndonee Nioeng O 1 -1 lndoneo Nam 13 47 -34 Indonee PAPUAN 8 8 0 Indonee Pm 19 2 17 lndonee Pm 0 3 -3 lndonee Pew 0 1O -1O lndonee Phalip 161 5‘2 1CD Indonee Poduga 17 5 -8 lndome Shana 19m 182 1720 lndonee Spin 192 1 18 74 lndonee Surinem O 1 -1 lndonee Sweden $ 86 33 Indonee Swine! 4a 63 <5 lndonee TRINIDA O 3 <3 lndonee Tim 8‘ 13 711 lndonee Thflen 193 113 & lndonee USA $1 550 3131 lndonee Unwed. 563 am 375 lndonee Union-y O 4 -4 lndonee Venezue 1 12 -1 1 Inlays! Amend 12 21 -9 135 wwoowoosagma10m21namn50w1xOWflaflwmnmmaomamalOyflmmommowm . 1 1111111111111 Am Austria Bil-nee Bub-do Belgium Belize Bolivia Brazil Cenede Chile Chile Colombl Coot-R DOMINIC Denmak ELSALV Eon-dot thnd Frmoe Gummy Gteeoe Gtmeme Guyan- Ham Ham HomKo Ioelmd lndonee Ireland New Jemdoe anen KateeR Mexico Mv-nmr Nethed Newzee Nloueg Mom-y PAPUAN Penum- Pm- Fem Philip Pm 3W Speh Sminem Sweden Swim TRINIDA mmmmmmmmmmmmwmmmmmmmmmmmmmmwmmmmmm 136 -1 -28 1 19 0 .25 .5 -2 -2 4 -1 -2 -14 71 m -13 -2 O -1 -2 194 3 3 4 34 -1 1334 137 1 Z) 27 -18 1 3 -7 -1 -1 6 O 37 O 5 O 272 O -1 -Z) -21 -2 O O 3250 271 -2 .5 14 1 231 57 30 O 1 72 O -2 19 4% 3 744 9 15n511221123mm n2112mamum1wmammam1ma11s4«m1aaznmmmzyumaz1aozau7m7 1 O7msm000xoonmm4oooon05wmo “mafia m 023310441maummonwmmo13mma1moonmoww APUA. N 1111111111111111111 TRINIDA 11111111111111111111111111111111111111111111 1111111111111111111111111111111111111111111111111111111111111111 é?i?E‘EE‘E‘EE‘E‘?E3’E‘é‘EiE§§§§§§§§§§§§§§§1Wfifififfififfifififfiffififii" commie S ELSALV Ecuador §W33???ééiésiéaimggzgggggg TRINIDA China Costa}? DOMINIC ELSALV Ecuador Finlmd 137 fiqfiqoofiaa .5 Booawoa 51a3-8aoq1118-88938§§1§331811°1§§§311e11 d d Anawu68m8§noganu q A -2 -2 .2 14 1618 111 151 1679 iLbBoab§1§$1o§o38381u iiiiiiifiimiiifififiiE'E'E'E‘Eiifiiéfiifi 138 §AEOOONB§S 3 A gwafi ficx3§ 030808 110 1%2 178 183 a13§§12§-§3831111a-m§§s 171 163 «1 ~11 ASEAN 21995 ( 3 Million) 139 Impede: Exponer Indonee Argent! Indonee Austral Indonee Austria Indonee Belarus Indonee Babedo Indonee Belgian lndonee Belize lndonee 80M lndonee Brazil Indonee Canada Indonee Chile lndonee Chine lndonee Colombl lndonee CostaR Indonee DOMINIC lndonee Denmak Indonee ELSALV lndonee Ecuador lndonee Finland lndonee France lndonee Gummy Indonee Suede indonee Gudeme lndmee Guys: lndonee Ham lndonee Hondun Indonee HongKo Indonee blend lndonee lmlmd Indonee m Indonee Jam Indonee ann Indonee KoreeR lndonee Malaya! lndmee Memo lndonee Myanmu Indonee Nether! lndonee NewZee lndonee Niacin Indonee Nam-y ltudome PAPUAN lndonee Pam Indonee Pengu- Indonee Peru Indonee Pimp lndonee Ponuoe Induces Shame lndonee Spain lndonee Sub-m lndonee Sweden lndonee Swine: Indonee TRINIDA Indonee Taiwan lndonee Mn lndonee USA Indonee Unwed. lndonee Uruguay Indonee Venezue Mel-ye! Argent! _‘____A#ctual Impacts Projected Impacts GTC TD 4 88 -84 1&6 5Q 57 93 126 -33 0 6 -6 1 2 -1 369 151 238 0 O O 1 6 -5 197 183 14 479 300 179 75 44 31 2:53 914 113 8 43 -$ 0 8 -8 O 8 -8 141 1% S O 3 <3 2 12 -1O 6 as -44 RD 50 $7 1939 610 1379 77 7O 7 2 11 «9 1 1 0 O 1 -1 1 3 -2 1633 442 1191 3 9 «6 52 46 6 977 529 448 O 3 3 141m 1934 12215 3322 56 2777 1221 6Q 32 167 161 6 47 3 8 1-1% 68 1% 84 G 0 2 -2 47 107 m 28 21 7 43 6 37 3 7 -4 12 21 -9 say 173 $7 50 67 ~17 as 771 -73 702 51 61 O ‘l -1 93 1$ 53 Q 15 <37 2 11 -9 157 4Q 153 672 am 12 7% 1774 6181 1426 514 914 3 14 -11 13 56 6 61 6 -4 immiifififiHES!HES!Him!!!EEEEHHHEEHE!HEEEEEEEHEEE g‘s’zgaésgfifi mun DOMINIC S ELSALV 5‘33??? 11% iigfiiiiigi PAPUAN 140 1279 1244 31$ ofiixaonieko §§ Biuooggggo .5 .5 OVOO 248 d omenséfia Koobggibofiba§ 19.1, 45 -1 16533 313 -1 1914 1- 1- -11 -18 -14 -2 -1 141 -2 -13 a) 20 -174 -10 3 -2 <5 -1 .4 23 166 673 -1o -3 o o -1 523 .3 -25 33 a) -1 2466 316 36 17 -1 1 as 2 -1 -14 -1 44 o -2 -a 1016 36 o -16 10 a 192 as 6784 401 -1 -16 94 ms 66 64 -2 367 o .4 322 641 w 2729 -12 zamnamaaa14mmmmaoo1mamum1mamanmasa42zemunoeoammmmamamm42mo4mmnmm ommmmeaO18001.ammoooomomnmommmwomnMOWQu‘ZA.ommommommmazzmmmmomooammmw mmw mmwmwmmm mmmmmmmm M:MMWMWWMWWWwwwmmmmwwmmmmmmw Mm mmmwmmmwmm MW Chilo MMMMMWMWMMWMMWWMmmmmmmmwmmmmmmmmmmmmmmmmmwmmmmmmmmmwmwwmwwwwwmmw 1111111113111EE’E‘iiif§§§§i1€111mi§§§§§§§§§§§§§§§§§§§§§§fiffiffif DOW-MIC 5 ELSALV 142 151 833383 0 a dam—n... ngflgaSMd-q8gggmndmm 48656.633a6§8§§§§ iii???Eié’iiiii’EifififiiiiiifiiiiéifiEii E . 143 117 21 41 4 6418 717 815 187 312 146 ~22 -1 -1 -2 186 1 7 ~24 3m§§§3§34§§b 144 EC :1975 ( $ Million) 1 H “ Actual Impom -_ MMMMMWWWMMWWWWWWWMM mmmmmmmm WWMW wflwWwmemmwm Wmmwwmwmm 41 24 31 4% “$55 4 5ma3m623m42flwwm1102n8flflm304H53mm0m44251 W212mn2$ ” 4 . 1813 4 28 _$%m.148m9m935m47amm$6132943mwm flosmaanzazflu “$411am5 Sm " FM. WW “mmmxoogmmum in amze mmmm :7234m ansemamaameaa174aexmmamm§ mm Um Venezue TRINIDA Tim M USA United. _ “mmwmmmmwwmmmmmwwwwwwwwwwwwmmmmwwmmwmmwwwwmmmwmwwwmmmwmwmwww gngmgm MR DOMINIC 5 ELSALV ifififiifiigfifiiifiigfi PAPUAN 1151113111 TRINIDA Tmm 1116.136: 141 .5 g @ 3338383368§u§88§4§5838.a.69§§333669388359§ug§§ gngfigaao§seaa§a. 15 112 1Q 110 811 442 .2 3374 1(1) -173 -19 -1 3 ~71 -1O miii3mi35533553333333333%§§§§§§§§§§§§335§§§§§§§§§5§§§§§§ 33 $3117; ELSALV E‘fifiii HangKo 1% lm 1:11:11! ii? PAPUAN 115113151 TRINIDA waen 13 U1“. Um Venezue fiegamgfisg 146 47 1 207 114 1% 1881 335:9 ggxfifiiéfiz§~§§8§39§§§69§86§a-.6 353: 4116 ooSm-Aogaaol‘uosw 31 371 423 15 113 § I 1710 8» 8111.11» 147 444n444m$61044445a1660M4714904444n2m04m1afimmoasmm am 1mmJ@WGQMA7 1 mm M MMMMMMWMMMMMWMMMMMMMMMMMMMMMMWWWmmmmmmmmmmmmmmmmmmmmmmm mmmmmmrmm ammm mmmmwm mmmmmmmwmfimm WWW If“!!! M ltdl'ld Irdlnd llfllnd Mid 148 110 mmmmas 97m9m§9mas9mmm9me9vaams ma mmmmnmmwmamm.2m.smmnmz11ms9mmmm92 mmmm9224aennexmmmzmmumz usmmmmma nm199ummmmmmmzma mammm17annmmmma9 Unned. Um Venezue W Mmmmwmewwmmmwmm wwmmmmm TRINIDA Tau-n mmmwmmemwmmm ”WWWWWW mmm ”mmwwmmmm m: 9999999999999999999999mmmmmmmmmmMMMMMMMMWwwmwwmwwwwwwwwwwwwwwwww EEEEEEEEEEEEEEEEififififififiififiEEEEEEEEEEEEE ii 149 @3283§68-a~$uaaaa:3m§§ams.. £3 @38885 a - ”‘3 3:33: Bqfififlio 11% 8338258 §§m$omafifiggg8 a-I ”(AU aa§.§sam d a883wun§aom8 L‘.’ fiSaBaBBaggafi an. $8§§3_§m.&8 aaafiaa ‘ an. BBhAEBSBwa- S kioghga 8m~3308.8§88asam d 38 §m é fififiugkobgbgmk u figfiahwozg ‘ United Unfied‘ Unled. Unned. United. Unwed Unfied. Untied Unled. Untad. Urfiad. Untad. Unwed. Umted. United. Unified. Unfied. United. United. United. United. Untad. Unfied. United. 150 1248 13 810 118 14 2a 12 13 27 15 113 121 70 13 1&4 1753 116 1m 1% 21 188 1 13 mgafiamfi 11 277 N 10 51 218 15 11% 514 18 2212 107 123 215 27 218 §-S§bn 58838§§0g§§$3q3m a a EC1198 Immu 3919mm 151 mmm_ §§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§i i g JMWMWfiflmEWEU 15 1% 2G55Ma8§§3§3a§8m388§§aau~§tm ‘ A d O Rw§.§ dd 08 d d 383mmw8&o§ mqéfiaaafifiugu 1W 152 «mgomzomemms1a1ommaseoommmmam4ma -10 am 12 1 86 -2 -2 52 -63 1 1678 1S 2 -22 20 m 1641 9 18 15 121 -212 2 -3 4831 3%W11m03mm9mm3538mmmwfi122mw43MQZM$MQBQ52W223mn65m1mm52mmm4mmmmau7 anam1m23nmw&M4248wmmfi412amman1mmomwmn3m1c1mn$wm2mu7ommwmammmn w m 1 :15; 3° Hfflfif §§§§§§§§§§§§ 153 1172 27 18 10 275 18 1310 fiamficaégmfifimu @383 a... N” 19323 a NI an. ‘ d fiqgasaaaazfiaeaafisfifi éfiné M -12 4% -1& 115 4&3 -14 -1 1a -1a -14 -11 1455 -17 443 fiififiaaaso§§833§89 m355355555§§§§§§§§§§5§§§§§§5§535333353353335? 1m 2 § PAPUAN §i§§§§35§§ TRINIDA i 154 .d§g§§m§a8§ 1” d dawnawuhaano§~a8$=m3 aamfiafiaagafi dd a§§a38§ d Antudagmgfinofioeb 112 ~14 2147 § ¢Aa$uhaiaaoo§~on£¢g8§g EEEEEEEEEEEEEEEEEEEEEEEE 155 ou8§o§3gaznuoba5§ogfiaaafiwnadoonfiggg 5.7 1188 N aaaifiwmaaouagfim d zmuemo§~momi--8-~3~ea auifi 3 174 1m 1287 889?» 137 13 19 msiggfiaugm LLohohfifiga gbmauhsongfiambbofis om§§§a5§$aaoaao8o £§§i88 406 mmmmmmmmmmmmmmmmmmmmmmmmmmmMmmMMWWWmmMWwWWWWWWWWWMWWWMWWWMWMWWWW §§§§§§§§§§§§§§§§§§§§§§g;gg§§ggzsssz333:33:3333;333:3233333333333 E; 156 115 b§8§§33§3333 aaaafiaag 141 mgflagafi 134 182 m-H~w§mg§38§ M Q “858%- 13 “Q 19 179 078 5% 14 9838mm» -13 -11 m.mL§§§§~m§aa8§Bbkq1gao§ Smré €113;ESEEEEEEEEEEEEEEESEEEE 157 415 17 38% d... a? Qa§§8$§8~8m8 flu 3§88aau flfind .5 U qo§§§aag§95 gfiaxfiamé§a§38~§ 1$7D 18 18 81 20 74') 18 133 171 1781 1518 UMed. Untied Untied, Unted. Untied. United; Urmd, Unied UM. Um‘led. United. Unfid. UMed. Untied, um. UMed. United. United. Untied. United, Myanmar Named NewZaa Nicarag Nanny PAPUAN 3%? Phillip 9%? dmm 51 TRINIDA Taiwan USA UMUIV Venozuo 158 5218 3721 1175 [1'1 0 I—I \O m m A a O a V 1% §§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§ “WWWQWaFQMNWME §r figamonos 9 u §m§§86§;uamg§§fiabgaom 8% 104 47 41 287 2 1 O 5 1GB 190 a A fiqaaaam§gg§§m§om8§ S N .5 #1510806 #3458 (.16 fisa smgfiasflfigdgo GTC 1:D_ ~13 2141 E13§E§E§§§§§i max DOMINIC ELSALV E‘fiigfifi 11% fififififigfiéfiéiiii‘i 1m BaBBUdON ‘ afinggwafiggaa d gagaafiinwa§~m§4358§~§85maunamflggéqumuagmiguo§aa 9a.5 23mg? § 11“ -11 215 ~194 1157 L;ad HLbbL§030855983853§$8§A 6§q§goa 1- iffflfifififii fifififififi 522mg CodaR DOMINIC 3 ELSALV Egfiifii HomKo 18 If“!!! s§§i§§fizzgiiggzszizs TRINIDA T . 3' USA Ulited. Una-y fiiiéfimsgig 161 12841 1877 210 afiflgn fig S§3§3313Q538388a ~2 ~1 2 $1 210 -918 -5 ~7 ~§ ~15 ~15 ~$ ~1814 314 ~277 110 125 -23 112 .2 103 13a 15 618 -1 14) 26 13533353551315???i§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§335 E mmR DOMINIC i ELSALV éggszigsssiggiggmg mum NW zfiifiiings TRINIDA Tum ii 162 1 874 1G mfifiafiafi N a O fig 00800080800080a3885$8§§ §§5m353.53§§wa§c85 $33: §zc§3§ A ‘ ans.~a§a88u0§--Esufimgfififimgiagafifisau 27 245 11 1247 1513 iiiffié’fiffififififififiE?1Wfifiimii1 1% ~5 u 575 4 O ~2 ~2 -8 ~2 ~6 ~5 «Q ~2 ~141 1 75 ~15 -$ a Q 2 -2 ~28 ~1 O ~1 ~10 ~18 s32 <3 ~13 ~1 -:9 ~51 4mfimmam£8272moe$wemsaqa1dmmwx1o flamooooqloommoum130n7omo21ogfia 07$025mm015m082a001$27541m23$mm$30 4| 1 UAN mmmm W:Mm£mmmmmmmam w Ema mmmmmmmwwmmflm MEEEEmmm:Emmmmmmmmmmmmmmm Mmmmmmm ggggggggg g g; q q — +-_-— — - a g? 333 §§§E§§¥§§§§§¥§§§§§§§§§§§5§§§§§§§ 1m 333§§§§§3§§§3E33333333333E3333Egggg E? 164 §B~§S§8m 3.1311§§ flomréafi 83 411 1mm 12731 10 14 13 1% 1m -. a 5... ~18- mum..- fianggofinamm-AAadnfiwga -I § 126 gigaaud m figaszx12§§rfiaa§33 gnouflgomokbfohuofinfigbuu .. § 0100 abaafiga-Las:a§szaso§11§§m tog-L ~185 é: iiEEEEEEiiEiiififiifififiififififii133mm33333333333333333353 m: was Nam éigififing: “5? 1% B - - ozfiagfifisaosufi 8 83 mm §a§0a§00§§833§ ngaofifi J " 38" 823$.afi : 1m ”1 66 M 1% <5 m 46 §L§g uua§§§§6a§fiubgm§8~mgamggg nah.§§2m§: 4m EEEEEE§E§§§§§§§§§§§E§§§§§§§§ $1 166 17 fitmusbfim: 83833 “figaia gasfiamf .3.5 (9.. 517 1472 18 :8 18 Qfifiss d 3.. m @96888385 $28§§.&ha-§aa~ 515 122 10 261 31 (B ~197 1814 ~13 ~10 1374 ~1118 “Eafigfififiw832§$a§1 Unted. Unted. United. Umad. United. Wed. Unied. um. United. Unled. United. United. Panama Patagua Pam Phillip Portuga SW Spin Surham Sweden Swflzar TRINIDA Tainan Thalan USA Um Venezue 167 74 21 121 12 17m 122 212 1573 217 19 18 12) 61 37 1028 158 1% 191 8| Jae. AN’DEJ 85558855555 168 i t §§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§E§§§§§§§§§§§§§§ ---.A¢2§' 1mm -, JMW ‘13u7 -‘ (a) (”NO-.0000 as b. an. oooooooL-o OOOO-‘dO-‘OO-IH‘NOOONOOOOOB c an babooooonoooh ‘ .0 N 194 4A 24 0.3 3.5 6.2 1 1.0 8.2 9.9 8.2 1 .0 1 .3 2.0 1 .1 3.8 1 .8 13.8 15.1 1 .7 1 .8 0.5 0.8 0.8 0.8 0.1 2.4 0.9 1 1 .4 0.8 13.7 1 .8 0.8 14.8 0.4 4.8 1 .3 09 1 .5 0.2 0.9 1 .8 __,__GT9 TD 1 1 1 .3 ~41 ~2.4 0.3 27.5 ~%.2 -8.0 4.7 ~99 ~7.1 ~1 .0 ~1 .3 ~20 ~1 .1 ~38 -1 .8 ~38 4.9 ~1 .7 ~1 .8 ~05 ~0.8 ~0.8 1 .4 ~01 ~2.4 ~09 -9.4 ~0.7 7.3 ~1 .8 ~1 .4 ~07 1 1 .9 ~22 1w 27.2 o 804008b34004 N coho - o o 803 omfiflubasouua oouo' 9.1 6.8 1 7.8 12.8 -9.1 ~19.8 ~1 7.8 ~2.8 ~1 .8 8.4 ~10.9 -8.0 ~7.3 6.5 ~15.1 5.4 -8.2 ~7.1 2.4 ~0.8 Siii’ifififigg 170 9 N .A P . oomnoaaa 208‘ _.O fiU‘lOOOOOO cannot»:“cartoons'-~ m Mahmaooowoonooo O a O uBmupm 8 m u3fimoaooma M4 1WD 64 $8 13 34 an 63 $3 491 Q” 38 48 38 28 -m8 40 n3 EEEEEEEEHESSSS’SHHHH§5§§§§§§§§§§§§§§§§§§§5SHEEN E 7? USA Emma: Chile é? mmR DOMINIC i ELSALV 171 18 0.7 0.2 1.2 10.1 0 3 woowa gm . 801000 U-‘OOOOOOO .0 .09 GOO-BA... pw dob .3 h ‘ a-§emos g80933§oo$3§3§ou33uo 1&9 1 13.3 18.5 73.4 1 1 1 .7 10.9 15.7 10.8 4314 79.7 ~14.7 52.2 ‘E; Wmm \Mum Wmm Wmm Wm» Wmm Wan Wmm Wan Wmm Wmm Wmm Wum Wum Wmm Wum Wmm Wan Wan Wan Wmm Wan Wan “um Wmm Wan hum Wmm Wmm Wmm Wan Wan Wum Wum Wmm Wum Wan Wuw Mum Wan Mum 172 d dd Bogooogofio080808003gaoooagoowofiaam ch d A aggoofiu 173 mmw I §§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§i on 121. $350$00aOA Qoooobo Bomooooooooosto 000080000 0 -o 2; W8 0 100 no 0 doonaono 21% A a; €§£§E§§§§§i 3 § 174 1 1'9 143 39 0.6 0.3 10 184 0.9 175 17 26 0.2 10.1 0.2 13.8 13.4 1 as 0.4 151 14.5 0.1 1327 1.3 183 61.1 19 12 1 4.5 ~50 18.4 ~4.7 ~3.8 19.8 ~1 .3 ~16.1 {Q1 .2 ~102.4 6.1 ~132.2 ~21 .2 ~58 31 .7 ~15.3 1&0 ~5.1 531 .1 ~18.8 ~27.9 ~4.3 ~14.4 ~7.8 ~27.4 sissgzssiazzgggggg §§¥§E§E §§§§§§ Eifiitiifiigitfiiififi PAPUAN §§§§§§3§1§ TRINIDA Tmm 1’51 Um Wmm §2§§§t§§§§§§ 175 o d USahaob' _. .0 .noo-‘OOOO'lb 30604503 N Odo“ E Sgfiaooooaoad eaaéfiozSoang ES§§§§§§§§§33§§§i333H3133HHHEHS’S’HSHS’EHS (E; < oomimc E E? 176 808§0£S§00§00;0838§ dd «8 11.8 133.4 Wmm Wmm Wmm Wmm Wmm Wmm Wan Mum Wmm Wmm Wan Wan Wmm Wmm Mum Mum me me Wan Wmm Wmm Wmm Wan Wmm Wan Wan Wan Wan Wmm Wmm Wan Wan Wan Wmm Wan Wan me Fma i§§§§ HMM mmm Eggsifiim mun PAPUAN 3%? V Mp §§§§§§ mwm Tmm gm 177 agooéoggo c. .5 8000 8§a0000§0g0003000680§0§0030 E E. g {It a 178 i §§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§ MWWE? i b ! . . 9 a0a00~00050000¢0 0| 0 800000a050000000000000a980 .0 d OOO-‘OOUIOOOO megmmu u. 31 .0 179 21 .9 8.2 19.8 10.4 116 0.5 152 13 1 34 0.1 3.6 12.6 24.5 5.6 4.1 5.4 1 .2 1 6.9 31 7.0 1%.5 33.5 1 6.3 33.1 21.6 1 7.1 1 19.9 1 7.4 1&0 1 74.7 1 7.5 8.1 1 4.0 1 3.7 10.2 1 .7 &.6 10.5 13.3 13.6 167.5 10.0 -1 7.6 -1 2.3 -210.9 -1 4.6 «8.9 -1 64.1 -Z$.0 -1 1 .1 -3.3 .1 .7 -1 .1 -3.0 -0.4 -6.9 HHHHES’S’S’H TRINIDA Taiwan 3‘ U$A Unwed. Um Venom: §i§§§§§§§i Ch“ 3g? .R 180 81 .4 141 13.1 41.1 103.4 15.5 ~21 .6 HHHHS’SHES3333'Hfié’iii’ififié’fiifiifiiH3 Wmm Wan Wan Wmm Wan Wmm Wan Wan Wmm 181 05 M “9 1“ M4 §Awaa§§§§§§300§9o9‘- $0 \Mum Wmm Wmm Wmm Wmm Wmm Wmm Wmm Wan Wmm Wmm Wmm Wan Wmm Wan Wan Wan Wan Wan Wan Wan Wan Wmm Wan Wan Wmm Wan Wan Mum Wan Wan Wan Wan Wmm Wan Wmm Mum W33 l& 1187 g - d 1. 0N03000Um080 93 $3 ‘ d3 baflqdu 5% -mo MA 183 MERCOSUR: 1995 (3 Million) GTC ETC TD m m Projected lmputs Actual Imports 0 _ _ mm anommmmas.....awm % .3.mas.aamm.mm.mmammwamm45mmmflmmmm nmsamoammwwmuammeaamem1asm4mmnsmmmmama4muamnsamm1wmmmnawmm $9.0mommmmmmmmas. mmm 00-nom8m «mammn0m10mmnmmmommmm-awmm TRINIDA UM. mummmmmmmmfimmwmm Emmm mmfimmmm mmmmmmmmmmmmmmmwmmmmmmmmmmmmmmmw 184 Importer Exporter Actual Imports Projected Imports GTC ETC TD Brazl Amati 041 211» 1932 Bruil Austral $3 523 43) Brazl Austria 201 272 -71 Brazi Bahama 10 17 -7 Brazl albedo 6 11 -5 Brazil Belgium 1771 335 136 Brazl Belize 1 1 0 Brazil Bolivia 5!) 14) 390 Brazil Canada 625 14¢ -623 Bruil Chile 1210 $6 622 Brazl China 1226 101 463 Brm Colombi 67 933 -476 Brazl CoetaR 5 44 41 Brad Denmuk 217 213 4 Brazil DOMINIC e3 57 9 Brazil Ecuador 23 113 99 Brazil ELSALV 34 1B 16 Brazil Finland 107 156 49 Brazl France 1% 151 6 Brazl Germany 367 21!) 1287 81:2! Greece 217 151 as Brazil Gwema 61 64 «3 But! Guyma 12 1O 2 Bruil Hall 7 1O <3 Brazl Hondua 32 1B 14 Blazil Hong.Ko 413 130 m Brazl Iceland 3 12 -9 Brazil Indcnee m 4:0 4!) Brazl Ireland as 107 -41 Brazil Italy 2039 156 261 Brazil Jamaica 47 1B 29 Brazil Japan M 2441 1&7 Brazl KoraaR 136 523 86 Brazil Mdayei 303 179 129 Brazil Mexico 1117 45 Bruil Myarimer 1 24 -23 Brazl w 150 53 1014 Brml New.Zea 6 93 -51 Brazil Nicarag 5 12 -7 Brazil Norway 113 205 93 Brazl Panama 92 3 57 Brazl PAPUAN 3 ' 7 4 Brazl Puagua 1331 179 1122 81:2! Peru a 444 -6 Brazl Philip 275 93 177 Blazll Pomona 413 225 186 Brul Slngapo am as 274 81:! Spain 115 $3 152 Brazl Sam 16 1O 6 Brazil Sweden 175 an .125 Btazl Swlzer 24$ 316 a Brull Tm ()7 264 143 Brazil Thalan 501 324 177 Brull TRINIDA a 7D -1 Brazl Unwed. 155 1864 4329 Brazil Uruguay 612 226 566 Brazl USA can 14124 «ace Brazil Venezue Q1 1126 66 185 we4ooaoamwaaaao41oaeseaoooaoqamooae - 44 O 38 -2 O -Q 4 O .4 1 12 0 19 040504mflmm GTC TD m9400503mmm~fl713120233210003071floy83n07203104131304445153m8 Projected Impom ”0000201“3%”10303O1m§000oosooomomoqufluoenunvalooozw1a004xsosumm 5 Actual Imports Emmwmmmwmmmmmmmmmwmmmmmmmmmmmmwwm mEmmmmmmmmfimmmmEmmEmmmmmmmmmfi 186 Importer Exporter Actual Imports Projected Imports GTC TD Uruguay Argenti 262 fl -1 77 Uruguay Austral 4 20 -16 Uruguay Amtria 3 B 5 Uruguay Bahamas 0 O 0 Uruguay Barbedo O O 0 Uruguay Belgium 8 1 1 <3 Uruguay Belize 0 O 0 Uruguay Bolivia 2 2 0 Uruguay Brazil 737 206 531 Uruguay Canada Z) 40 J) Uruguay Chle 4') 24 16 Uruguay China 5 so 5 Uruguay Colombi 5 13 7'2 Urmuay CoataR 1 1 0 Uruguay Denmark 4 6 -2 Uruguay DOMINIC 1 1 O Umway Ecuador 7 3 4 Uruguay ELSALV 1 1 O Urmuay Finland 9 4 5 Urugmy France 37 53 -16 Uruguay Germany 1 1B 61 57 Uruguay Greece 2 4 -2 Uruguay Guatema 4 2 2 Uruguay Guyana 0 O 0 Uruguay Haiti 0 O 0 Uruguay Hondua O 1 -1 Uruguay Hong.l