Towards an Evaluation of Regional Integration in Latin America in the 1990s Robert Devlin and Ricardo Ffrench-Davis '»«»ŕ- ITD December 1998. Working Paper 2. The Institute for the Integration of Latin America and the Caribbean (INTAL) Integration, Trade and Hemispheric Issues Division (ITD) of the Integration and Programs Department of the IDB have organized a joint publication series: Working Papers Refereed technical studies providing a significant contributk to existing research in the area of trade and integration. ^sional Papers articles, speeches, authorized journal reprints and other document that should be of interest to a broader public. Inter-American Development Bank Integration and Regional Programs Department Institute for the Integration of Latin America and the Caribbean IDB - INTAL Esmeralda 130, 16th and 17thFloors (1035) Buenos Aires, Argentina Integration, Trade and Hemispheric Issues Division 1300 New York Avenue, NW Washington D.C. 20577 United States The opinions expressed herein are those of the authors and do not necessarily reflect the official position of the IDB and/or INTAL-ITD, or its member countries. Printed in Argentina INTAL - ITD Towards an Evaluation of Regional Integration in Latin America in the 1990s. Buenos Aires, 1998. 44 pages. Working Paper 2. I.S.B.N 950-738-077-9 The decade of the 1990s has witnessed a wave of regional integration initiatives in Latin America. The renaissance of regional integration has not been uncontroversial. Some, interpret regional integration as a costly policy option, which is harmful to the countries involved, and the multilateral systems more generally. Here we argue that regional integration is a potentially valuable tool for modernization and development in a second best world, arising from imperfect and incomplete markets at home and abroad. This article reviews the strategic dimensions and rationale of regional integration, places potential costs and benefits in their proper perspective and outlines directions for future research and economic policy. We conclude with policy suggestions contributing to amplifying the benefits and minimizing the costs of regional integration agreements. An earlier version of this paper was prepared for the Conference "Regional Economic Integration and Global Economic Cooperation: The Challenges for Industrial, Transitional and Developing Countries", organized by FONDAD, The Hague, Netherlands, on 18-19 November, 1997. This article is forthcoming in World Economy. The authors are, Robert Devlin, Division Chief of the Integration, Trade and Hemispheric Issues Division of the Inter-American Development Bank (IDB) and Ricardo Ffrench-Davis, Principal Regional Advisor of the United Nations Economic Commission for LatinAmerica and the Caribbean (ECLAC). The authors are grateful to F. Ballestero, A. Jessen, and A. Panagariya for their comments and Brenda Simonen and Francois Dionne for their statistical assistance. TABLE OF CONTENTS INTRODUCTION 1 SECTION! THE GROWTH IN TRADE 3 SECTION II. THE NEW FACE OF REGIONAL INTEGRATION IN LATIN AMERICA AND THE CARIBBEAN 9 SECTION III. WHY DO THE COUNTRIES PURSUE REGIONAL INTEGRATION? 13 Some common motivations for regional integration 13 Politics 13 Practicing the art of the possible in trade liberalization 14 Strategic Impulse to Development 15 A ttracting Foreign Direct Investment (FDI) 17 The costs of regional integration 17 SECTION IV PUTTING THE COSTS AND BENEFITS INTO PERSPECTIVE 19 SECTION V CONCLUSIONS 23 APENDIX The Composition and Technological Content of Intra-regional Trade Composition of intra-regional trade 27 Technological Intensity 27 BIBLIOGRAPHY TOWARDS AN EVALUATION OF REGIONAL INTEGRATION IN LATIN AMERICA IN THE 1990S Robert Devlin and Ricardo Ffrench-Davis INTRODUCTION The decade of the 1990s has witnessed a wave of regional integration initiatives in Latin America: more than 14 agreements -free trade areas or customs unions- since 1990 with a handful more in varying degrees of negotiation (see Table 1). However, this was not just a Latin American phenomenon, as regionalism has more than ever become a global trend (Mistry [1996]). Indeed, now Japan, South Korea and Hong Kong are the only World Trade Organization (WTO) members which are not signatories to at least one preferential trade agreement (WTO [1995]).1 Table 1 RECENT REGIONAL TRADE AGREEMENTS IN LATIN AMERICA3 Agreement Andean Pact (Bolivia, Colombia, Ecuador, Peru & Venezuela) Central American Common Market (Costa Rica, El Salvador, Guatemala, Honduras & Nicaragua) G-3 (Colombia, Mexico & Venezuela) MERCOSUR (Argentina, Brazil, Paraguay & Uruguay) Chile - Mexico Chile - Colombia Chile - Venezuela Chile - Ecuador Chile - MERCOSUR Chile - Canada Mexico - Bolivia Mexico - Costa Rica Mexico - Nicaragua NAFTA (Mexico, Canada & U.S.A.) Bolivia - MERCOSUR Some agreements in discussion c MERCOSUR - Andean Community MERCOSUR - European Union Chile - Central America Chile - European Union Mexico - El Salvador, Guatemala, Honduras Mexico - Belize Mexico - European Union Mexico - Ecuador Mexico - Peru Mexico - Panama Dominican Republic Dominican Republic Central America CARICOM CARICOM - Central America 1990 b 1991 ' 1994 1991 1991 1993 1993 1994 1996 1997 1994 1994 1998 1993 1997 Source: IDB, Department of Integration and Regional Programs, Integration, Trade and Hemispheric Issues Division. Notes:a Excludes partial agreements of the Latin American Integration Association (LAIA or ALADI in Spanish).b Date of reactivation of the free trade zone.c In different stages of development. Of course, Hong Kong has just been transferred to China, which yet has to become a member of the WTO. 1 Regional integration is not new to Latin America. Economic integration played an important role in the region's early Post-War economic history. The 1960s and 1970s saw a number of very ambitious initiatives inspired by the successful Western European experience (Ffrench-Davis, Munoz and Palma [1994]). Indeed, at its peak in the late 1960s and early 1970s, the topic of integration was hard to avoid in the discussion of Latin American development. However, disillusionment with integration processes had clearly set in by the late 1970s and the discussion of regional integration was all but silenced by the external crisis of the early 1980s. The renaissance of regional integration has not been uncontroversial. Some, including the authors, view regional integration as a potentially valuable tool for modernization and development in a second best world. Others, however, interpret regional integration as an inferior, costly policy option, which is harmful to the countries involved, and to the multilateral system more generally. The purpose of this article is to review the strategic dimensions and rationale of regional integration, place potential costs and benefits in their proper perspective and outline directions for future research and economic policy. Section I is a brief review of trends in intra- and extra-regional trade and some of the factors driving it. This is followed by a discussion, in Section II, which places regional integration in the broader Latin American policy context that is conditioning its effects. Section III outlines what countries expect to achieve from regional integration and the costs that can be confronted. Section IV attempts to put the costs and benefits into perspective and offers directions that might help us better evaluate the full effects of integration and thereby soften some of the rougher edges of the policy debate. Section V concludes with policy suggestions that should contribute to amplifying the benefits and minimizing the costs for the participants in regional integration agreements and the world community at large. 2 SECTION I. THE GROWTH IN TRADE The 1990s have witnessed a rebound in the region's trade after the crisis of the previous decade. Between 1990 and 1996, the region's exports expanded by 73 percent (95 percent for 1990-97); imports grew even faster, at 127 percent. Imports as a percentage of the region's GDP now equals more than 20 percent, up from 10 percent in 1990.2 Moreover, the region's growth of imports has consistently exceeded that recorded at the world level: according to WTO estimates, the value of world imports grew by an average 7 percent a year between 1990 and 1996, compared to 15 percent for Latin America. A closer look at the region's trade performance in the 1990s highlights the following trends: • Intra-regional trade has grown more rapidly than trade with countries outside the region. This trend is particularly pronounced in the case of exports (Table 2). Since 1990, the value of intra-regional exports has grown by 18 percent a year on average, compared to 9 percent for extra-regional exports. Intra-regional exports now account for 18 percent of total Latin American and Caribbean exports, up from 12 percent in 1990. Without Mexico, the figure reaches 27 percent, from 15 percent at the beginning of the decade. • It is interesting to compare real (constant prices) changes in exports to GDP growth. Overall, the GDP of the region grew 29 percent between 1990 and 1997. The quantum of total exports rose 72 percent, thus increasing by one-third the export ratio of Latin America. The leading destinations by far were intra-regional markets, which expanded by 215 percent. However, extra-regional exports also increased much faster than GDP, more than doubling the latter's rate of growth. • Growth rates for intra- and extra-regional imports have been more homogeneous (Appendix Table A.2). While intra-regional imports expanded by an average 18 percent a year between 1990 and 1996, extra-regional imports also grew very fast, by 14 percent a year, reflecting a generalized import boom in the region. This, coupled with the aforementioned important growth of extra-regional exports, confirms that regional integration has been consistent with open regionalism. • The marked difference in the growth rates of the region's overall exports and imports reflects a large imbalance in the growth of trade with extra-regional markets, as imports from these sources expanded nearly twice as fast as exports. When Mexico is excluded, the figures are 17 and 7 percent, respectively. 3 Table 2 WESTERN HEMISPHERE: TOTAL AND INTRA-REGIONAL EXPORTS (US$ millions and percentages) 1992 1993 1994 1995 1996 1997 1990-1997 average Western Hemisphere b Total Exports 658,234 684,995 727,241 765,511 859,185 996,045 1,052,001 1,158,617 % growth 7.9 4.1 6.2 5.3 12.2 15.9 5.6 10.1 8.4 Extra-hemispheric Exports 341,515 357,391 364,017 365,905 394,303 472,187 487,847 513,761 % growth 5.4 4.6 1.9 0.5 7.8 19.8 3.3 5.3 6.0 Intra-hemispheric Exports 316,719 327,605 363,224 399,606 464,881 523,858 564,154 644,856 % growth 10.7 3.4 10.9 10.0 16.3 12.7 7.7 14.3 10.7 Intra/Total 48.1 47.8 49.9 52.2 34.7 52.6 53.6 55.7 Latin America c ^^^^^B Total Exports 137,781 136,242 140,234 150,691 176,018 212,325 239,646 268,294 % growth 10.5 -1.1 2.9 7.5 16.8 20.6 12.9 12.0 10.0 Extra-Latin America Exports 121,412 116,249 115,291 121,058 141,020 170,431 195,867 216,766 % growth 10.9 -4.3 -0.8 5.0 16.5 20.9 14.9 10.7 8.6 Intra-Latin America Exports 16,369 19,993 24,943 29,633 34,998 41,894 43,779 51,528 % growth 7.3 22.1 24.3 19.3 18.1 19.4 4.5 17,7 17.8 Intra/Total 77.9 74.7 7 7.7 79.0 79.3 79.7 7S.3 79,2 Andean Community ^^_ Total Exports 31,605 28,630 28,390 29,654 34,256 38,843 45,479 49,568 % growth 26.1 -9.4 -0.9 4.5 13.5 13.4 17.1 9.0 6.6 Extra-Andean Exports 30,310 26,912 26,224 26,858 30,952 34,268 40,817 43,959 % growth 26.2 -11.2 -2.6 2.4 14.9 11.1 19.1 7.7 5.5 Intra-Andean Exports 1,295 1,719 2,156 2,796 3,404 4,575 4,662 5,609 % growth 23.5 32.7 25.4 29.7 21.7 34.4 1.9 20.3 23.3 Intra/Total 4.1 6.0 7.6 9.4 9.9 11.8 70.3 7 7.3 CARICOM ^H Total Exports 4,762 4,771 4,875 4,837 5,933 6,211 ... ... % growth 6.3 0.2 2.2 -0.8 22.7 4.7 ... ... Extra-CARICOM Exports 4,224 4,308 4,408 4,286 5,346 5,407 ... ... % growth 4.9 2.0 2.3 -2.8 24.7 1.1 ... ... Intra-CARICOM Exports 555 463 467 551 587 815 ... ... % growth 23.3 -13.9 0.8 19.1 6.5 38.9 ... ... Intra/Total 77.7 9.7 9.6 77.4 9.9 73.7 — — C AC M Total Exports 4,058 4,138 4,697 5,065 5,509 6,864 7,786 8,243 % growth 12.7 2.0 13.5 7.9 9.9 24.6 13.4 5.9 10.7 Extra-CACM Exports 3,402 3,356 3,697 3,961 4,290 5,408 6,200 6,417 % growth 12.4 -1.3 10.1 7.1 8.1 26.4 14.6 3.5 9.5 Intra-CACM Exports 656 782 1,000 1,105 1,229 1,456 1,586 1,826 % growth 14.6 19.1 27.9 10.4 11.3 18.4 8.9 15.1 15.7 Intra/Total 76.2 7S.9 21.3 21.8 22.3 21.2 20.4 22.2 MERCOSUR ^^B Total Exports 46,425 45,911 50,561 54,162 62,112 70,401 74,998 82,369 % growth -0.3 -1.1 10.1 7.1 14.7 13.3 6.5 9.8 8.5 Extra-MERCOSUR Exports 42,302 40,808 43,341 44,132 50,157 56,018 57,960 62,215 % growth -1.2 -3.5 6.2 1.9 13.7 11.7 3.5 7.3 5.7 Intra-MERCOSUR Exports 4,123 5,102 7,220 10,031 11,955 14,394 17,038 20,154 % growth 10.8 23.8 41.5 38.9 19.2 20.3 18.4 18.3 25.4 Intra/Total 8.9 7 7.7 74.3 18.5 79.2 20.4 22.7 24.5 Group of Three Total Exports 65,162 65,117 67,451 74,367 86,020 107,625 128,914 146,719 % growth 22.2 0.9 36.1 10.3 17.1 23.8 19.8 13.8 12.3 Extra-Group of Three Exports 64,127 63,937 65,675 72,023 83,456 104,319 125,749 142,553 % growth 15.5 -0.3 2.7 9.7 15.9 25.0 20.5 13.4 12.1 Intra-Group of Three Exports 1,035 1,180 1,776 2,344 2,565 3,306 3,165 4,166 % growth 47.0 14.0 50.4 32.0 9.4 29.9 -4.3 31.6 22.0 Intra/Total 7.6 1.8 2.6 3.2 3.0 3.7 2.5 2.8 NAFTA Total Exports 561,164 591,440 627,933 661,752 738,494 856,598 907,809 1,000,755 % growth 7.8 5.4 6.2 5.4 11.6 16.0 6.0 10.2 8.6 Extra-NAFTA Exports 320,667 341,997 354,468 360,444 396,434 461,079 480,978 514,926 % growth 5.2 6.7 3.6 1.7 7.2 19.3 4.3 7.1 7.0 Intra-NAFTA Exports 240,497 249,443 273,465 301,308 352,060 395,520 426,831 485,829 % growth 11.5 3.7 9.6 10.2 16.9 12.3 7.9 13.8 10.6 Intra/Total 42.9 42.2 43.6 45.5 47.7 46.2 47.0 48.5 Source: IDB, Statistics and Quantitative Analysis Unit of the Integration and Regional Programs Department based on DATAINTAL. Notes:" The exports of Mexico include maquila in all years.b The Western Hemisphere includes Latina America (see note c), the United States and Canada. : Latin America includes Argentina, Bolivia, Brazil, Chile, Colombia, Costa Rica, Ecuador, El Salvador, Guatemala, Honduras, Mexico, Nicaragua, Panama, Paraguay, Peru, Uruguay y Venezuela. 4 Some factors influencing current trends were: Geography. Areas dense in capital and population often tend to naturally interact and trade relatively more intensively with increasing specialization (Ballestero [1996]). Among the economic factors behind this are the positive externalities of location and agglomeration. The tendency can be further enhanced when income levels, cultures, tastes and languages are similar, as they are in Latin America, and when differentials exist in transport costs between contiguous and non-contiguous countries. On these criteria, large natural geographic areas of economic integration would appear to exist in Latin America in its Southern Cone, Venezuela-Colombia-Ecuador, Central America and North America for Mexico. Indeed, the gravity of geography seems to matter. If intra-regional trade is disaggregated into border and non-border trade, it is seen that border trade (i.e., that with neighboring countries) represents the bulk of intra-regional commerce (see Figure 1). Relaxation of the external restriction. The decline of world interest rates, debt relief and a return of external capital flows in the 1990s (Devlin, Ffrench-Davis and Griffith-Jones [1995]) dramatically increased import capacity in the region with consequent reactivation of economic activity. Since intra-regional imports equal intra-regional exports, the generalized import boom was reflected in the marked growth of intra-regional exports during the 1990s. Real exchange rate appreciation. The region's external trade performance was influenced by the exchange-rate behavior of Latin American and Caribbean countries. The simultaneous liberalization of the capital account in many countries, coupled with a surge in supply of foreign capital and the use of exchange-rate anchors in support of stabilization programs, contributed to real currency appreciations in an important number of countries (just when the opposite, a real depreciation, was needed to facilitate export-led growth). An exchange rate index, weighted by GDP, gives an average revaluation of 25 percent between late 1980s and 1994. This situation tended to encourage imports while, at the same time, discouraging the region's exports.3 Moreover, since real appreciation of exchange rates with respect to the rest of the world has been simultaneous among a significant number of neighboring countries in Latin America in the 1990s, the dampening effects on exports have been relatively stronger in the extra-regional market. Economic reforms. The structural reforms undertaken in the late 1980s and 1990s have energized private market activity, facilitated the emergence of new investors and trade. Unilateral trade liberalization in particular has been a key in exposing natural market opportunities for exports to neighboring countries that heretofore were hidden behind the wall of national protection (IDB [1996a]). 3 Some countries, like Colombia and Chile, have actively tried to slow down pressures for real appreciation, by avoiding anchors and resorting to, among other things, foreign exchange regulations and imaginative financial engineering (Devlin, Ffrench-Davis and Griffith-Jones [1995]). 5 Figure 1 INTRA-REGIONAL AND BORDER TRADE, 1994 Exports (percentage of total) Urate Uigiy BSdiBlr Paegtff Itou Paren» Méd htnlES Gianda Riirir CtstaRba Ctfcrtia (He Bdíte 2D Border 30 | Regional 43 33 60 Imports (percentage of total) lUlmam 1 1 r.™.taRir^ ^^H 1 20 Border 30 I Regional 40 SO 60 Source: IDB, Department of Integration and Regional Programs, Integration, Trade and Hemispheric Issues Division. Note: The countries included are those that share borders and have available information. Mexican border trade does not include trade with the United States. 6 Subregional Trade Agreements. The aforementioned explosion of subregional and bilateral trade agreements in the 1990s has stimulated intra-regional trade through many mutually reinforcing effects; for example: an increased flow of information and public attention on opportunities in an adjacent market ("agreement-led" growth in trade). preferences are an integral part of the regional integration agreements and provide incentives for intra-regional trade. The absolute level of the preference over time will depend on the evolution of external tariff rates. Nevertheless, it is important to point out that many of the preferences of the older trade agreements in the region have been progressively eroded by the unilateral liberalization of trade in the late 1980s and early 1990s. in contrast to unilateral opening, the free trade arrangements have given the private sector reciprocal and legally binding market access which has reduced the risks of trade and investment barriers emerging in the affected market. This in turn increases private sector confidence. An example of the especially strong nature of a reciprocal commitment is Mexico, which during the peso crisis exempted NAFTA partners from a temporary increase of tariffs on 500 items (5 percent of total tariff lines). a preferential agreement can signal the continuing commitment of public authorities to trade liberalization, during a conjuncture in which unilateral or multilateral liberalization is not possible or desirable. In other agreements such as MERCOSUR, subregional trade liberalization is accompanied by an additional commitment involving a much broader political message, pursued at the highest official levels, to promote deep economic integration and political cooperation among member countries. 7 SECTION II. THE NEW FACE OF REGIONAL INTEGRATION IN LATIN AMERICA AND THE CARIBBEAN Regional integration must be evaluated in the broader context of the overall contemporary economic policy which is conditioning its effects. The regional integration initiatives immediately following WWII inserted themselves into the prevailing development strategy of import substitution. Indeed, the integration schemes ofthat period were designed in part to enhance the efficiency of the import substitution model through a strategic expansion of the highly protected national market.4 While the integration initiatives achieved some important results -for example, the significant liberalization of reciprocal trade in Central America- outcomes fell far short of objectives. On the one hand, the strong national political commitments to domestic protection made opening up even among associate countries an extremely laborious negotiating process that rarely achieved more than very partial results. On the other, the costs of trade diversion were amplified due to the general presence of high average tariffs on third parties and extensive use of non-tariff restrictions. The traditional model of development in Latin America changed in the 1980s and this, in turn, has dramatically changed the face of regional integration itself. In effect, the regional integration of today has inserted itself into the broader overall strategy of opening up to the world economy. Countries have entered into multiple arrangements that are eliminating tariffs among partners across substantially all trade within a relatively short period of time, and which often involve other commitments that even go beyond the WTO's trade-related disciplines. In any event, Figure 2 and Table 3 respectively show, for a sample of regional agreements in Latin America, that the reciprocal liberalization process is typically completed for the bulk of trade within a period of 10 years and negotiated exceptions have rarely exceeded 6 percent of total tariff lines. In the initial stages of their development, regional integration arrangements link up with the overall economic reform process most obviously through its trade liberalization component. In effect, regional integration is a third tier of a three-tier liberalization process. Unilateral liberalization. The first and most dramatic level of liberalization has been through unilateral measures to open up economies. Profound trade reforms have been undertaken in Latin America as part of a broad-ranging process of change, in which international competitiveness and exports play a leading role. Most countries are in search of export-led development. In contrast with the experience of East-Asian nations, the main instrument of trade reform has been a relatively indiscriminate and rapid liberalization of imports (Agosin and Ffrench-Davis [1995] pp. 3, 9-58; ECLAC [1995] ch. V). The aim is to expose producers of importables, which had often been receiving a high level of protection, to outside competition, while also encouraging the output of exportables. It is expected that this will result in higher productivity, with the absorption of new technologies and increased specialization. 4 In some cases, as in the Andean Pact, there was the deliberate effect to reduce drastically the level and dispersion of effective tariffs. 9 120 100 80 60- 40- 20 Figure 2 SELECTED FTAS IN LATIN AMERICA. YEARS TO LIBERALIZATION (percentage of tariff items liberalized) .........................................................y^žĹizĽ................ľ................".................................................................................... .' Z* -■* ........................■ / */-':' i------------------------- I i / ' 1.............r / - -v, i hi i li i : ! 1 i í 1 10 11 12 13 14 15 16 17 1£ ----------Bolivia to Mexico (1995) ----------Colombia to Mexico (1995) Source: Estevadeordal (forthcoming). ■ Canada to Mexico (1994) Costa Rica to Mexico (1995) ■Colombia to Chile (1994) USA to Mexico (1994) Table 3 SELECTED LATIN AMERICA FREE TRADE AGREEMENTS: EXCEPTED PRODUCTS (Number of tariff lines) Sectors (SITC Rev. 2) Canada to Chile to Colombia Chile to Chile to Colombia Mexico to Costa Rica Mexico to Mexico Chile to Mexico to Mexico Colombia to Chile Ecuador Venezuela to Mexico Colombia to Mexico Costa Rica to Chile Mexico Venezuela Food and live animals chiefly for food 72 128 128 114 46 203 115 85 51 35 38 114 Beverages and tobacco 6 7 6 14 3 15 3 6 6 3 Crude materials, inedible, except fuels 15 14 17 24 39 136 4 7 1 13 177 Minerals fuels, lubricants and related materials 15 24 16 15 17 18 Animal and vegetable oils, fats and waxes 1 38 46 56 25 46 7 1 1 30 25 7 Chemicals and related products, N.E.S. 4 43 52 17 32 56 83 7 3 51 Manufactured goods classified chiefly by material 215 161 4 43 31 103 605 Machinery and transport equipment 172 9 3 153 26 65 15 Miscellaneous manufactured articles 40 37 2 6 180 146 145 Commodities and transactions not 5 4 4 classified elsewhere in the Site Other (n.e.) 1 12 10 7 7 Total 77 673 490 234 350 605 669 112 65 89 100 1128 Source: Estevadeordal (forthcoming). Most of the countries' trade reforms could be described as sudden and drastic. The experience is reflected in the fact that the average tariff in Latin America and the Caribbean has declined from 45 percent in the second half of the 1980s to 13 percent in 1995, accompanied by a sharp reduction of tariff dispersion as well (Figure 3). Furthermore, over the same period the share of the region's imports subject to non-tariff barriers declined from 31 percent to 11 percent. Specific tariffs have virtually disappeared even while they are still common in the industrialized economies (ECLAC [1995]; IDB [1996]). Figure 3 TARIFF LIBERALIZATION IN LATIN AMERICA (weighted on imports) 50% 0% I--------------1--------------1--------------1--------------1--------------1--------------1--------------1--------------1--------------1--------------1— 1985 1990 1995 Source: IDB, Department of Integration and Regional Programs, Division of Trade, Integration and Hemispheric Issues. Multilateral liberalization. The second level of external opening is multilateral. The region has assumed the new disciplines that emerged from the Uruguay Round. Indeed, it was the only developing region to bind 100 percent of its tariffs (although the binding was at a level more than double that of average applied tariffs). With Panama joining the WTO in 1995, all of the region is now subject to the rights and obligations of this world organization. Regional Integration. The third tier of opening has been through regional integration. It is often overlooked that in the new context of policy change in Latin America, regional integration is an additional instrument to open economies to competition and complements levels one and two of the trade liberalization process. Indeed, the insertion of regional integration initiatives into the overall liberalization strategy of the 1990s gives the process the character of "open regionalism". Moreover, the fact that tariffs are different from zero, but generally at relatively moderate levels, leaves space for reciprocal tariffs preferences with more limited trade diversion than in earlier trade agreements. SECTION III. WHY DO THE COUNTRIES PURSUE REGIONAL INTEGRATION? Regional integration has various dimensions. The simplest form of integration is a free trade area in goods only, in which there is a progressive elimination of tariffs on most trade among partner countries. The next level of commitment would be a more comprehensive free trade area that includes services and even WTO "plus" disciplines in other trade-related areas. Even deeper commitments would involve ceding sovereignty over commercial policy, by creating a customs union with a common external tariff protecting the liberalized subregional market. Deeper still is a common market with free movement of factors of production. All these schemes are currently reflected in the objectives of Latin America integration. While the benefits and costs on integration depend on which type of scheme one is considering and the overall policy context, limits on space permit only some general appreciations in this regard. Some common motivations for regional integration Without being exhaustive and recognizing that many goals can be interrelated, frequent objectives are outlined below. Politics Many initiatives emerge under a political umbrella designed to bring nations closer in more than market economics. The degree of commitment varies but the motive is often present. Latin Americans share a common heritage, language and culture. There is a side of Latin Americans that makes them like to be more together; these sentiments indeed extend back to Bolivar and independence. The phenomenon is sometimes hard for non-Latinos to understand and could seem contradictory given the long history of serious political disputes among Latin American nations. But the fact remains that the centrifugal forces of disagreement have co-existed with, and often have been overcome by, the centripetal forces of a common heritage and culture. These opposing forces of course can be especially intense in the various geographic sub-regions of Latin America. Today's integration in Latin America is often driven by powerful political objectives. This is most clearly manifest in MERCOSUR (including associates Chile and Bolivia), where countries with a history of conflictive relations are using economic integration to draw themselves more closely together into a common purpose of peace and prosperity. A similar phenomenon can be found with the reactivation of integration in other sub-regions such as the Andean Community and Central America. 13 Practicing the art of the possible in trade liberalization Unilateral and multilateral liberalizations are always staged. Unilateral liberalization has the benefits of creating trade and raising competition through importables and lowering input costs; but it also has immediate fiscal costs, real resource costs by creating obsolete capital and redundant labor, and political costs because of the real and imagined threats of globalization. Thus, the process of adjustment must move in tandem with the political and economic capacity to digest those costs. The large and rapid liberalization of recent years encountered initially considerable "water" in national tariff schedules and the needed fiscal adjustments were quite straightforward. But now tariffs are presumably closer to actual differential margins of competitiveness between home and abroad, while fiscal options are narrower. Moreover, the liberalization process is made more complicated by exchange rate appreciations that have taken place in the 1990s in a number of Latin American countries, which has further cut margins of protection, and the simultaneous action of stabilization programs which usually raise the cost and reduce the supply of domestic credit (ECLAC [1995]). Meanwhile, multilateral rounds come in spurts, the schedule of which is largely out of the control of developing countries. In the case of the Uruguay Round the region still is in the process of digesting existing commitments. In this context, authorities can use regional integration as a window of opportunity for continued liberalization, but in the more predictable and controlled environment of a reciprocal agreement of circumscribed scope. In fact, regional integration adds a compensatory ingredient to import liberalization, by fostering reciprocal exports in tandem with reciprocal imports. Hence, the doses of positive and negative impulses to economic activity and investment are more balanced with regional integration, than is the case in pure unilateral import liberalization. Moreover, the regional agreement does lower the average level of protection vis-á-vis the status quo, creating trade, raising competition and promoting specialization in the subregional market. The arrangement for liberalization can meet less political resistance (and indeed even be quite popular as in the case of MERCOSUR) because of a number of associated factors such as public sentiments about "getting together" with a known neighbor, compensation through reciprocity with guaranteed market access, and more limited impacts on fiscal income (the starting point for most integration agreements has involved modest levels of trade and rather symmetric tariff structures). A regional arrangement can additionally serve to signal authorities' commitments to investors and lock-in policy reforms that otherwise might be more easily reversible. North-South agreements, anchored by a credible developed country market, in particular are often cited for these "non-traditional" confidence-building effects (Fernandez [1997]; Ethier [1998] pp. 449, 1149-61). Agood example is the incorporation of Southern Europe into the EU, which was instrumental in the former's economic transformation and consolidation of democracy. 14 Strategie Impulse to Development Regional integration also builds on strategic considerations arising from imperfect and incomplete markets at home and abroad, which handicap the spread of efficiency gains in certain sectors and the development of new productive patterns with progressively higher degrees of value added. The conventional literature on the benefits and costs of economic integration focuses on tariff preferences in a framework of optimal competitive equilibrium. This equilibrium is assumed to be disturbed only by the existence of import restrictions. In this framework, integration is beneficial only if it implies a move toward free trade; that is, if the effects of trade creation (shift toward cheaper sources of supply) are larger than those of trade diversion (shift toward more costly sources of supply).5 The crucial issue, however, is how costs are measured; in the standard approach it is at actual market prices net of tariffs, assuming away transitional costs and incomplete markets, as well as acquirable competitiveness. The assumptions lead to the obvious conclusion that overall unilateral liberalization is always the optimal national policy and better than regional integration. But the real world is more complicated. For many non-traditional products, access to markets is more limited and unstable, making economies of scale, the emergence of externalities of location and agglomeration and specialization more difficult to achieve. It is for these types of products that regional integration becomes a potential platform for diversifying growth of exports, and to improve trade's contribution to development. In the face of distortions in world markets, guaranteed access to regional foreign markets can be a catalyst for exploiting potential externalities; indeed, this is a leading objective of policy-makers and a major force encouraging regional integration. Moreover, in face of economies of scale, what otherwise would be a costly trade diversion can eventually become a cost-reducing and welfare-enhancing effect (Corden [1972] pp. 3, 465-75; Ffrench-Davis [1980] pp.44, 41-8). Meanwhile, local factor markets are incomplete or distorted. Labor training, technology and long-term capital are scarce, with inexistent or infant markets and foreign direct investment (FDI) -a potential bearer of some of these scarce factors- is frequently coquettish, playing one national suitor off against the other in a world of imperfect information.6 These market failures are more significant for non-traditional exports of differentiated products, whether of natural resources, manufactures or exportable services. If access to external markets is improved for these exportables, it can be a catalyst for completing markets and diluting segmentation. There is also the effect, ä la Lipsey, of an eventual reduction of price distortions on consumption. Additionally, it must be recalled that FDI represents only 6 to 10 percent of capital formation in the world. 15 Infrastructure, trade financing and knowledge of markets (marketing channels, organized transportation, standards, etc.) have often been biased against intra-regional trade in LDCs. All these "factors" of trade have been traditionally more developed for transactions with the great metropolitan centers (often linked to the cumulative effects of a colonial past) while they are inexistent or rudimentary for trade among LDCs neighbors. This is a significant variable partly explaining why intra-regional trade has been lower among Latin American countries than what the gravity of geography would often suggest. These are serious restrictions on the expansion of production and trade in goods and services with relatively more knowledge content and longer learning curves, elements which are now recognized as key components of the growth process. Regional integration can be a strategic tool to partially overcome these obstacles by: • expanding market size to facilitate greater specialization and industrialization through economies of scale and possibilities to exploit economies associated with the agglomeration of production activity. • enhancing the forces of competition, enlarging a market with guaranteed reciprocal access, and intensifying the specificity of information flows, all of which in turn should induce new domestic investment and permit better conditions to attract efficient foreign investment. • creating the security of subregional market access, preferences, and exploiting the familiarity of neighborhoods, which combine to accelerate the emergence of new exporters of manufactured goods. In effect, the learning curve associated with subregional export experience can serve as a platform for new international exports. This is an important consideration since history has shown that developing countries can achieve new dynamic comparative advantage on the road of their long term convergence with industrialized countries.7 The expected enhanced international competitiveness brought about by regional integration should build confidence and prepare countries for globalization and further advances in multilateral liberalization. And as mentioned earlier, regional integration can also be viewed as a way to move ahead with liberalization while the region awaits consensus on development of a new round of reciprocal multilateral disciplines. To appreciate the strategic dimension of integration it is useful to examine the profile of intra-regional exports. Intra- and extra-regional exports from Latin America display marked differences For an example of intra-industry trade patterns set off by an important subregional agreement, see Echavarria [1997]. 16 in terms of their product structure and technological content, with manufactures accounting for a much larger share of intra-regional commerce. This pattern is evident even if Mexico -whosemaquila trade with the United States accounts for a large share of Latin America's overall exports- is discounted from the regional average. Excluding Mexico, manufactures account for approximately 50 percent of intra-regional exchanges, compared to around 23 percent for extra-regional exports. The appendix to this article discusses the composition of intra-regional exports and their technological content in more detail (see Appendix Table A. 1). Attracting Foreign Direct Investment (FDI) There is recognition that foreign direct investment can be a useful tool for preparing for globalization as foreign firms bring technology, modern corporate practice, export networks (fully one-third of world trade is intra-firm), and long term external finance. Ethier [1998] points out that developing country competition for FDI is sufficiently intense that significant distinguishing features in a country can be decisive in attracting investors, which tend to cluster and locate together. The literature recognizes that regional integration schemes can create an impact which is attractive for FDI (Bloomstrom and Kokko [1997]). The bigger the sub-regional market, the greater the change in the economic environment on account of the agreement, and the greater the locational advantages, the more likely the initiative will stimulate foreign investment, including that which is intra-regional in character. MERCOSUR -the sub-regional agreement in Latin America with the highest international profile- has certainly attracted the attention of foreign investors. All these aspects of regional integration are of course potential developments rather than guaranteed outcomes; what happens in practice depends on the nature of policy implementation, a point we will return to in the concluding section. The costs of regional integration The potentially positive aspects of regional integration are accompanied by costs too. Some of these have received much attention in the literature of late. Abrief summary of the most frequently cited costs would be: i) Preferences in regional trading arrangements can divert trade away from possibly more efficient firms which are located in non-member countries (Yeats [1996]; Bhagwati and Panagariya [1996]). This has costs for domestic consumers and for the non-member countries that lose market share. The trade diversion risks locking the partner economies into patterns of inefficient production. 17 ii) Regional integration agreements can improve the terms of trade of member countries at the expense of non-member countries and give rise to incentives for maintaining or increasing preferences and protection.8 iii) When there are serious asymmetries in average tariff levels among prospective partners of an integration agreement, the loss of tariff revenue in the preferential liberalization process can have serious redistribute effects among the countries (Panagariya [1996] pp. 5, 485-515). In effect, part of what would have had realized as tariff revenue on imports from the partner country prior to the agreement is transferred to the partner's producers as tariffs are preferentially eliminated. iv) While regional integration can clearly induce foreign direct investment in the expanded subregional market, it can locate unevenly and -in the absence of harmonized incentives- be a source of competition among partners and a fiscal drain. Integration induced investment can also involve diversion of FDI from more efficient non-members (Winters [1997]). v) In regional integration benefits are often asymmetrically distributed and initially concentrated in some members while others are dependent on uncertain spillover effects (Puga and Venables [1997]). vi) An explosion of free trade areas creates a spaghetti pattern of agreements with multiple hubs and spokes that give rise to distortions in trade, administrative costs, rent seeking and a difficult to predict distribution of gains among countries (Wonnacott and Wonnacott [1995] pp. 2, 107-19). vii) The emergence of regional agreements can create defensive reactions, in which a country joins an agreement not because it is the best option, but because of the real or potential costs of being left out of an integration process. viii) Regional integration distracts attention from multilateral rounds of liberalization and delays further unilateral opening. 8 Looked at from another angle, rather than improving the terms of trade, regional integration may be able to soften a worsening or highly volatile terms of trade that could occur if Latin American countries keep on producing more of the same basket of traditional exports. 18 SECTION IV. PUTTING THE COSTS AND BENEFITS INTO PERSPECTIVE Evaluating regional integration processes and their costs and benefits is no easy task. Part of the problem is the nature of the subject matter. First, regional integration is a complex general equilibrium phenomenon with dynamic processes, making it difficult to dissect for purposes of causal explanation. The process involves issues that link growth to technology, learning, externalities, political economy and politics, all of which economists have trouble grappling with at a national level not to mention among several countries simultaneously. A further complication which one finds in Latin America is that the integration processes are an integral part of the profound structural reforms that have touched all levels of the economy and create big changes. Moreover, initial conditions, and the phases and sequencing of these reforms, are usually quite different among the partner countries. Second, regional integration is a medium/long term process. When successful, one expects to see initial costs compensated by benefits that play out over the medium and long term. Third, regional integration is very much a second best world where generic prescriptions can be especially dangerous. Fourth, regional integration is typically evaluated in light of what would have happened in its absence. Moreover economists are interested in measuring changes in welfare; given the complications of defining this for a particular subregion they often use a proxy expressed in a summary statistic reflecting growth or trade (Winters [1997]). These characteristics place great burdens on analysts. It is well known that counterfactual analysis faces a daunting epistemological problem: contrary to fact conditionals can never be verified by realizing their antecedent (the "if clause); thus the resulting explanation is never correct or incorrect but rather only persuasive or not persuasive. We also know that counterfactual s are more likely to be persuasive the: (i) the more simple the causal process studied; (ii) the shorter the time period in question; (iii) the smaller the changes considered; and (iv) the less analysis turns on exact magnitudes. Reflecting back on the characteristics of integration processes it can be seen that our counterfactual analysis is challenged on all these counts. Conclusions about regional integration rarely are based on the entire story. Much of the debate centers on static trade creation and trade diversion effects. This is partly because many economists consider these effects to be the fundamental dimension for evaluating regional integration. One problem, however, is that the static analysis frequently uses a partial competitive equilibrium framework to jump to general conclusions about a process that is a general equilibrium phenomenon. 19 Worse, the existence of trade diversion alone (never mind the net effects with trade creation) in new integration agreements has been sufficient for some to categorically condemn them and regional integration more generally (Yeats [1996]).9 But more importantly, trade creation vs. diversion is clearly only part of the story and many other economists (e.g., Mistry [1996]; Fernandez [1997] and the authors) would argue that is not the major part. This is because the net benefits of the dynamics of integration, coupled with the so-called non-traditional effects such as signaling and lock-in, can be several times larger than their static reallocation effects. Problems exist here too because our models of dynamics and empirical foundations for testing them are very deficient, so much so that some have even characterized analysis in this area as "mystical" (Winters [1997]). It is true that the empirical foundation of dynamic analysis is still weak. Nevertheless, the models of dynamics are sufficiently specified to suggest that the benefits behind the dynamics of integration are potentially large. It therefore is worth the effort to go beyond static trade creation-diversion analysis (which has its ambiguous dimensions as well) to begin to better understand, even if only very imperfectly, the longer term dynamics. The empirical bottlenecks to understanding Latin American integration should not be underestimated. Even basic data such as the evolution of preferences, rules of origin, non-tariff measures, intraregional investment flows, firms cost structures, etc., are unavailable or incomplete. The many gaps sometimes induce questionable ad hoc compromises in our analytical techniques or cause us to ignore important phenomena altogether through the convenient use of the ceteris paribus. Better data development and more field research will not eliminate the debate over regionalism but it would certainly help to ground the debate more in reality and probably help narrow our differences. The starting point is to better complement our powers of scientific deduction with much more empirical field work and case studies of the disaggregated dimensions of the dynamics of regional integration. In other words, instead of examining what would have happened in the absence of integration we might want to spend more time discovering what is actually happening and how it is happening in Latin American integration. In effect, one would examine the different objectives of a specific integration arrangement, see whether these different objectives are being realized, and begin to catalogue the causal factors contributing to developments without necessarily being overly concerned about precise weights. For example, one frequently stated obj ective of regional integration is to enhance competition; hence we can examine how sectoral markets are changing their competitive structure and the forces behind that. Is intra-industry specialization increasing in the subregional 9 This latter study set off a major controversy by concluding that MERCOSUR was harmful to itself and the rest of the world due to trade diversion. It concluded this by discovering intense intra-regional trade in some sectors with high preferences. The study, however, did not control for the fact that protection might be independent of MERCOSUR nor for other potentially important explanatory factors, and overlooked indications of much trade creation in the 1990s (Devlin [1997] pp. 7, 171-5 and IDB [1996b]). 20 market? Are the different parameters of the integration agreement stimulating firms to invest? Are firms' technology and cost structures improving in the direction of greater international competitiveness and is there room to reduce preferences? Are new international exports and comparative advantages emerging out of experiences in the subregional market? This type of research is at "ground zero" and examines the integration agreement from the bottom up. Field research does not generate elegant analytical structures. It is time consuming and expensive, often requiring the building of primary data bases. It also will not generate summary statistics of welfare or permit categorical evaluations of integration processes. But it has four potential benefits: i) It will allow for better observation of what is actually happening in the different dimensions of integration. The analyst gets "inside" the process where the action is and examines the dynamics of sectoral markets and firms which actually move the process forward. ii) By working at relatively low levels of aggregation one might be able to begin to identify causal factors that are not easily captured in existing theory or more aggregated analysis. iii) While such analysis will not permit the adding up of effects into a summary statistic of welfare, the examination of multiple disaggregated dimensions of an integration process will permit a series of analytical "vignettes" which taken together can build a tentative story of whether the integration process is achieving expected goals in strategic areas. iv) The empirical work will feed our economic modeling of integration with better informed assumptions and better data for testing. In sum, the suggested approach of more intensive interaction between deductive and inductive research methods should enhance our powers of discovery and evaluation of a process that is ever more present in the world economy. Any major transformation has costs, usually concentrated up-front. Therefore it is no surprise that regional integration has costs. For instance, since regional integration is a strategic compromise among economies with different economic and political characteristics, a degree of unwanted trade diversion is inevitable.10 However, countries justify these costs by the greater benefits that are expected, which are derived from a combination of political returns, lock-in effects, trade creation and the aforementioned dynamic forces of transformation which are spread out over a longer period of time. 10 Trade diversion growing out of a regional negotiation can reflect trade-offs from a process of endogenous protection which is essential for the political sustainability of commitments. Also, as mentioned earlier, in a "dynamic" setting some trade diversion could be a benefit to the extent it ultimately would contribute to lower costs, increase competitiveness and growth. 21 Thus when examining up-front static costs, analysts should be careful to interpret them as only a piece of a story which plays out over a longer term, and thereby refrain from categorical overall assessments, except in the most extreme cases. Meanwhile, since regional integration is a strategic decision, participants should have their objectives clearly articulated. One objective is to minimize costs; thus any constructive analysis that sheds light on them also should be welcome. Moreover, vigilance about costs is extremely important. On the one hand, while fashionable, not all integration arrangements make economic sense. On the other, even those that do can potentially go awry. Finally, there are systemic costs arising from the sum effect of many regional initiatives which individually may make sense.11 11 The clearest example of this is the spaghetti effect of many integration arrangements in the hemisphere which reduce transparency and raise transaction costs. However, the problem may be less severe than appears at first blush because the majority of arrangements follow the umbrella concepts laid out by ALADI (e.g., Chilean bilaterals) or NAFTA (e.g., the Mexican bilaterals). Moreover, the complex network of arrangements provides incentives for consolidation, as witnessed in MERCOSUR'S emerging free trade association in South America, the Free Trade of the Americas process and perhaps someday soon at the world level in a new multilateral round aiming at a target of zero tariffs (which would eliminate the simplest free trade areas). 22 SECTION V. CONCLUSIONS Regional integration is a fact of life in Latin America and indeed in most of the world. Regional integration is being pursued in Latin America for political reasons as well as for its value as a strategic tool of development in a second best world. No one denies that it can have important costs. But much of the attention on costs has been focused on the short term up-front costs that are part of any major transformation. These costs must be measured against the benefits, and the bulk of these are expected in the medium to long term through the dynamics of economic transformation. These dynamic processes could be better understood and measured if more attention were given to micro and sectoral field research where much of the process of regional integration takes place. This ground zero research would complement our more aggregate analysis and model building.12 Since regional integration is here to stay it also is constructive to promote policies that are likely to minimize the risks of unacceptable costs and to amplify potential benefits. Some of the major challenges in this regard are: a) Progressive elimination of imperfections in subregional integration schemes. There is a need for full implementation of agreements and effective enforcement. Politically feasible formulas are needed to gradually eliminate existing exceptions to agreed trade liberalization because the opening up of sensitive sectors is usually very rich in trade creation effects. Integration of services is largely a frontier that still must be crossed, but special caution is needed regarding integration of financial services due to potential negative macroeconomic side-effects (Zahler and Budmevich [1997]). Remaining trade distorting non-tariff measures (NTMs) must be eliminated or harmonized to the extent possible, and very importantly, rules of origin in free trade areas should be gradually replaced for common external tariffs or, alternatively, simplified and relaxed sufficiently (within the confines of rigor) to respect the status quo in trade patterns (Garay and Estevadeordal [1995]; Simpson [1997] pp. 6, 9-10; Serra, et. al. [1996]). Latin America must also replace its tradition of settling disputes through diplomatic channels (perhaps effective when the economies were state-dominated) for modern transparent dispute settlement mechanisms (Devlin [1995]). Now it is necessary that integration arrangements be transparently rule-based; only in this way will the full potential for productive private investment -that is so important for the efficient specialization which is at the heart of successful integration agreements- be realized. Rationalization of regional institutions is necessary. In the case of some traditional integration schemes which modeled themselves after Europe, the task is to downsize an overdimensioned and 12 The Institute for the Integration of Latin America and the Caribbean (INTAL) in Buenos Aires is sponsoring a series of sectoral studies for three subregional schemes (MERCOSUR, Andean Community and the Central American Common Market). ECLAC is also researching the sectoral effects of integration. 23 underfinanced institutional structure. For the new schemes of the 1990s the task isjust the opposite: fortify incipient institutional arrangements so that instruments are compatible with objectives. Of particular concern for ambitious integration schemes is the need to have mechanisms in place to ward off dangerous imbalances in the distribution of costs and benefits of the process. Another area of concern is infrastructure (Vera [1997]). There is a need to create effective institutional mechanisms for the coordination of regional infrastructure networks and their financing, in order to better exploit the advantages of location. Finally, integration schemes, especially deep ones, must improve official mechanisms for the interchange of information and analysis on macroeconomic developments in the sub-regions and monitor the processes of convergence which should emerge (Ben-David [1996] pp. 3-4, 279-98). b) Macroeconomics and structural economic reforms. These reforms, including those leading to macroeconomic stability, have been underlying the recent success of intra-regional trade. However, there is concern that exchange rate appreciation linked to capital surges and indiscriminately open capital accounts, coupled with the use of the exchange rates primarily as anchors of domestic prices, is distorting resource allocation and trade, and may reduce the sustainability of macroeconomic balances and their contribution to growth (IDB [1997]; Ffrench-Davis [1996] pp. 60, 21-37). Sometimes short-run targets of stabilization also have tended to contribute to a weakening of mezzoeconomic policies (such as education, labor training, support to technological improvement of medium and small firms, infrastructure). All this tends to differ the reaping of profitable opportunities in processes of economic integration and opening, and worsens the balance of benefits and costs. In extreme instances of instability in the face of volatile capital flows, it would actually threaten the viability of the integration project itself. There clearly is a need to foster development of more direct instruments for stabilization policy. c) Operational WTO review of regional integration processes. Article XXIV of the GATT and Article V of the GATS are meant to ensure consistent, fair and transparent multilateral monitoring of integration arrangements. However, the articles and their implementation, even with the important clarifications of the Understanding attached to Article XXIV in the Uruguay Round, still suffers from a degree of imprecision. The doubts that sometimes are raised about regional integration could be more constructively dealt with in the context of operational Article XXIV reviews with multilaterally agreed criteria and strong empirical foundations (Serra, et al. [1996]). Moreover, improved multilateral guidelines would help to broaden the common base among agreements and mitigate the potential costs of the spaghetti bowl of arrangements in the hemisphere and the world. d) Reducing the volatility of International Capital Flows. Volatility of capital flows and contagion undermine any major economic initiative, whether at the national or sub-regional levels. While countries' macroeconomic stance and financial regulatory settings are a first line of defense, improved international cooperation to deal with this growing problem is long overdue. 24 As long as countries are clear in their strategic objectives for an integration agreement, are vigilant of costs, continue to perfect shortcomings and ensure the process remains an integral part of an overall policy framework of structural economic reform, there is good reason to be cautiously optimistic about the ability of regional initiatives to serve as an effective instrument of growth and development. The other essential element for Latin America and the rest of the world is further development of the multilateral system. Multilateral initiatives to assuage the volatility of capital flows are essential for the sustained world growth and the open markets which are the foundation of the new regionalism. Latin America also has a vested interest in ensuring that regionalism is consistent with a healthy and progressively more liberalized rules-based world trading system, if for no other reason that 80 percent of its trade is extra-regional. Fortunately, there is growing consensus among economists and policy makers about the potentially positive contribution of the new open regional integration to the world trading system. The trend is well captured in a recent WTO Secretariat study which states "...to a much greater extent than is often acknowledged, regional and multilateral integration initiatives are complements rather than alternatives in the pursuit of more liberal and open trade" (WTO [1995] p. 56). 25 APPENDIX The Composition and Technological Content of Intra-regional Trade Composition of intra-regional trade The Latin American economies provide very important, and dynamic markets for the sales of manufactures for many countries of the region (ECLAC [1994]). For instance, for Argentina, Chile, Colombia, Uruguay and Venezuela, this is by far the biggest market, be it for traditional industries, basic inputs or new industries. Latin America continues to be the almost exclusive destination for exports from the new industries of several countries; these exports have been markedly dynamic in the 1990s. The same is true of their respective sub-regional market as regards the new industries of Costa Rica and Guatemala. Brazil has channeled its export manufactures to different markets; the United States continues to be the main buyer of traditional industries, followed by Europe; as for basic inputs, other developing regions have displaced Latin America as the main destination, but in the case of new industries, the region is the most important market for Brazil. An exception is the case of Mexico, where the regional market has less relative importance than the United States.13 Technological Intensity Development based on a growing and sustained international competitiveness is boosted by the dynamic effects derived from technological apprenticeship. The strategies to improve international linkages, based on productive development, emphasize the role played by trade in the process of stimulating the development of activities which make intensive use of knowledge and technology. In this sense, it is interesting that trade among developing countries is characterized by concentration in goods that are more technology-intensive than exports from developing to industrial countries. This is the broad conclusion of a study by Buitelaar [1993] which combines data on foreign trade and on production. Three main conclusions emerge from the research: i) The production of goods which depend to a greater extent on intra-regional trade has more sophisticated technological features. Such goods are to be found mainly in the chemical sector, non-electrical machinery and transport equipment. They are also sectors in which international demand tends to be more dynamic. Their price trends are more stable and evolve more positively over the long term than prices of traditional exports. 13 Regressions carried out by ECLAC for the period 1970-91 show that there was a strong positive relationship between the importance of Latin America as a destination and the share of new-industrial products in total exports of Argentina, a relationship that is positive but less intense for Brazil. In Chile, the exercise revealed a strong positive correlation for all manufactures. For Mexico, however, no significant relationship was obtained. 27 ii) The sectors which exhibit a strong export drive toward the region also tend to show (sometimes with a lag) a drive towards extra-regional markets, which suggests that the promotion of intra-regional trade complements the promotion of extra-regional exports. iii) These same sectors are those in which the region has a high dependency as regards extra-regional intermediate imports; therefore, intra-regional trade benefits from having access to inputs and equipment which may be imported from third countries. Thus, relaxation of excessive import restrictions has contributed to foster and upgrade exports. To sum up, intra-regional trade, because of its characteristics, associated with location and the diverse channels which facilitate such trade, complement the Latin American countries linkages with the global economy and provide a dynamic context of technological apprenticeship, leading to greater international competitiveness and a more diversified, balanced pattern of specialization. Table A. 1 LATIN AMERICA (14 COUNTRIES):3 COMPOSITION OF EXPORTS BY DESTINATION, 1970-74 AND 1995 (in percentages) Latin America and the ,. . „ u u Total Caribbean 1970-74 1995 53.6 32.2 29.9 16.0 6.2 4.0 17.6 12.2 46.0 65.6 33.6 30.5 9.5 11.9 6.0 5.2 9.2 9.1 8.9 4.4 12.4 35.1 4.3 5.8 1.9 7.2 3.6 10.0 0.6 1.7 1.6 4.5 1.4 3.8 2.6 12.1 0.3 0.9 1.7 10.1 0.6 1.2 0.4 2.2 100 100 Source: ECLAC, on the basis of official figures. Note: " Argentina, Bolivia, Brazil, Chile, Colombia, Costa Rica, Ecuador, Guatemala, Honduras, Mexico, Paraguay, Peru, Uruguay and Venezuela. For 1995, Mexico's figures were adjusted excluding maquila, in order to homogenize with the 1970-74 period figures. 1970-74 1995 A. Commodities 1. Agricultural products 2. Mining products 3. Energy products B. Industrial products 1. Semi-manufactures 1.1. Agriculturally-based, labor-intensive 1.2. Agriculturally-based, capital-intensive 1.3. Mineral-based 1.4. Energy-based 2. Manufactures 2.1. Traditional industries 2.2. Basic inputs 2.3. New labor-intensive industries a) Low technological content b) Mid-range technological content c) High technological content 2.4. New capital-intensive industries a) Low technological content b) Mid-range technological content c) High technological content C. Other 0.2 0.5 TOTAL 100 100 51.0 19.7 11.7 10.3 1.0 2.4 38.3 7.0 48.8 79.7 23.3 29.9 7.5 10.3 3.1 4.8 6.4 8.4 6.2 6.5 25.5 49.8 4.8 8.9 4.8 10.2 8.7 13.0 1.7 2.3 4.1 6.3 3.0 4.4 7.2 17.8 0.7 1.2 5.0 14.7 1.5 2.0 28 Table A.2 WESTERN HEMISPHERE: TOTAL AND INTRA-REGIONAL IMPORTS (US$ millions and percentages) 1992 1993 1994 1990-1996 average b Western Hemisphere c Global Imports 747,493 758,937 837,042 913,399 1,046,065 1,161,200 1,238,749 % growth 5.4 1.5 10.3 9.1 14.5 11.0 6.7 8.8 Extra-Hemispheric Imports 431,751 423,603 459,534 512,927 585,874 642,112 672,737 % growth 4.5 -1.9 8.5 11.6 14.2 9.6 4.8 7.7 Intra-Hemispheric Imports 315,743 335,334 377,507 400,473 460,191 519,088 566,012 % growth 6.6 6.2 12.6 6.1 14.9 12.8 9.0 10.2 Intra/Total 42.2 44.2 45.7 43.8 44.0 44.7 45.7 Latin America and the Caribbean ^^M Global Imports 110,235 128,880 157,007 174,272 205,546 226,317 250,306 % growth 11.9 16.9 21.8 11.0 17.9 10.1 10.6 14.6 Extra-LAC Imports 93,221 108,082 131,190 144,680 171,237 183,965 204,990 % growth 12.1 15.9 21.4 10.3 18.4 7.4 11.4 14.0 Intra-LAC Imports 17,014 20,798 25,817 29,592 34,310 42,352 45,317 % growth 11.0 22.2 24.1 14.6 15.9 23.4 7.0 17.7 Intra/Total 75.4 16.1 76.4 77.0 76.7 18.7 18.1 Latin America and the Caribbean (excluding Mexico) ^^H Global Imports 68,643 78,914 94,877 109,163 125,367 153,876 162,140 % growth 7.7 15.0 20.2 15.1 14.8 22.7 5.4 15.4 Extra-LAC Imports 52,685 59,703 71,021 81,874 93,983 113,531 118,979 % growth 6.9 13.3 19.0 15.3 14.8 20.8 4.8 14.5 Intra-LAC Imports 15,958 19,210 23,855 27,289 31,384 40,345 43,161 % growth 10.1 20.4 24.2 14.4 15.0 28.6 7.0 18.0 Intra/Total 23.2 24.3 25.7 25.0 25.0 26.2 26.6 CA CM ^^^m Global Imports 6,535 6,868 8,874 9,456 10,224 12,087 12,304 % growth 6.0 5.1 29.2 6.6 8.1 18.2 1.8 11.1 Extra-CACM Imports 5,895 6,058 7,805 8,326 8,950 10,580 10,743 % growth 6.2 2.8 28.8 6.7 7.5 18.2 1.5 10.5 Intra-CACM Imports 640 810 1,069 1,131 1,274 1,507 1,561 % growth 3.9 26.5 32.0 5.8 12.7 18.3 3.6 16.0 Intra/Total 9.8 11.8 72.0 72.0 72.5 72.5 72.7 Andean Communíty ^^^m Global Imports 17,425 22,311 27,220 29,398 30,617 38,300 36,814 % growth 3.0 28.0 22.0 8.0 4.1 25.1 -3.9 13.3 Extra-Andean Imports 16,243 20,665 25,129 26,753 27,345 33,423 31,954 % growth 1.5 27.2 21.6 6.5 2.2 22.2 -4.4 11.9 Intra-Andean Imports 1,182 1,646 2,091 2,645 3,272 4,877 4,860 % growth 30.4 39.3 27.0 26.5 23.7 49.1 -0.4 26.6 Intra/Total 6.8 7.4 7.7 9.0 70.7 72.7 73.2 MERCOSUR ^^M Global Imports 27,326 32,140 38,536 47,823 61,851 75,311 83,217 % growth 12.3 17.6 19.9 24.1 29.3 21.8 10.5 20.4 Extra-MERCOSUR Imports 23,204 27,016 31,266 38,427 49,989 61,218 66,125 % growth 13.1 16.4 15.7 22.9 30.1 22.5 8.0 19.1 Intra-MERCOSUR Imports 4,122 5,125 7,270 9,396 11,862 14,094 17,092 % growth 8.0 24.3 41.9 29.2 26.2 18.8 21.3 26.8 Intra/Total 15.1 75.9 18.9 79.6 79.2 18.7 20.5 NAFTA ^^M Global Imports 678,850 680,023 742,165 804,493 919,866 1,007,336 1,076,608 % growth 5.2 0.2 9.1 8.4 14.3 9.5 6.9 8.0 Extra-NAFTA Imports 443,190 430,927 462,031 510,164 578,310 627,931 655,530 % growth 4.8 -2.8 7.2 10.4 13.4 8.6 4.4 6.7 Intra-NAFTA Imports 235,660 249,096 280,134 294,329 341,556 379,405 421,078 % growth 5.9 5.7 12.5 5.1 16.0 11.1 11.0 10.2 Intra/Total 34.7 36.6 37.7 36.6 37.7 37.7 39.7 GROUP OF THREE ^^M Global Imports 54,168 66,073 82,264 86,588 100,433 97,549 111,838 % growth 14.4 22.0 24.5 5.3 16.0 -2.9 14.6 12.8 Extra-G3 Imports 53,450 65,110 80,795 84,626 98,242 94,379 108,600 % growth 14.2 21.8 24.1 4.7 16.1 -3.9 15.1 12.5 lntra-G3 Imports 719 963 1,470 1,963 2,191 3,170 3,238 % growth 31.7 34.0 52.6 33.6 11.7 44.7 2.1 28.5 Intra/Total 1.3 7.5 1.8 2.3 2.2 3.3 2.9 Source: IDB, Department of Integration and Regional Programs, Integration, Trade and Hemispheric Issues Division, based on DATAINTAL. Caribbean, US and Canadian data, as the source country, are from IMF Direction of Trade. Notes:" Mexico's exports include maquila. In principle, intra-regional imports should equal intra-regional exports. Slight variations between values of intra-regional exports from Table I and intra-regional imports in this table are due to reporting differences between the countries.b Average for Latin America and the Caribbean is for 1990-1996.c Western Hemisphere includes Latin America and the Caribbean (see following definition), the United States and Canada.d Latin America and the Caribbean here is Argentina, Bolivia, Brazil, Chile, Colombia, Costa Rica, Ecuador, El Salvador, Guatemala, Honduras, Mexico, Nicaragua, Panama, Paraguay, Peru, Uruguay, Venezuela, Bahamas, Barbados, Belize, Guyana, Jamaica, Trinidad and Tobago, Haiti and the Dominican Republic, e = estimate. 29 BIBLIOGRAPHY Agosin, M. and R. F French -Davis. "Trade Liberalization and Growth: Recent Experiences in Latin America", Journal of Inter-American Studies and World Affairs, 37. 1995. Ballestero, F. La Macroeconomia Regional y el Area Nacionál de Integración del MERCOSUR. Washington, D.C. : Integration, Trade and Hemispheric Issues Division, Inter-American Development Bank. 1996. Ben David, D. "Trade and Convergence Among Countries". Journal of International Economics, 40. 1996. Bhagwati, J. and A. Panagariya. Preferential Trading Areas and Multilateralism: Strangers, Friends and Foes? Washington, D.C. : Economic Development Institute, World Bank. 1996. Bloomstrom, M. and A. Kokko. Regional Integration and Foreign Direct Investment. Washington, D.C. : International Trade Division, World Bank. 1997. Bouzas, R. "Preferential Trade Liberalization in the Western Hemisphere: NAFTA and Beyond". In: J. J. Teunissen (ed.), Regionalism and the Global Economy. The Hague : FOND AD. 1995. Buitelaar, R. Dynamic Gains of Intra-regional Trade. ECLAC, Working Paper N° 18. Santiago : United Nations. 1993. Corden, M. "Economies of Scale and the Theory of Custom Unions". Journal of Political Economy, 80. 1972. Devlin, R. Comments on "Conflictos Comerciales y Mecanismos de Resolución de Controversias Bajo elAcuerdo de Libre Comercio entre Estados Unidos y Canada", by R. Dearden. In: Patricio Meiler and Richard Lipsey (eds.), NAFTA y MERCOSUR. Santiago, Chile : CIEPLAN and DOLMEN Editorial. 1995. ___________. "En Defensa del MERCOSUR". Archivos del Presente, 2. 1997. ___________, R. Ffrench-Davis and S. Griffith-Jones. "Surges in Capital Flows and Development: An Overview of Policy Issues". In: R. Ffrench-Davis and S. Griffith-Jones (eds.), Coping -with Capital Surges. The Return of Finance to Latin America. Boulder, Colorado: Lynne Rienner Publishers. 1995. 31 EcHAVARRiA, J. Trade Flows in the Andean Countries: Unilateral Liberalization or Regional Preferences. Washington, D.C. : Trade Unit, Organization of American States. 1997. ECLAC. Open Regionalism in LatinAmerica and the Caribbean. Santiago : United Nations. 1994. __________. Latin America and the Caribbean: Policies to Lmprove Linkages with the Global Economy. Santiago : United Nations, 1995. Revised and updated second edition published by Fondo de Cultura Económica, 1998. __________. Panoramade la LnserciónLnternacionalde America Latina y elCaribe, 1996. Santiago : United Nations. 1997. EsTEVADEORDAL, A. (forthcoming). "Negotiating Free Trade Areas in the Americas" Washington, D.C. : Integration, Trade and Hemispheric Issues Division, Inter-American Development Bank. Ethier, W. "The New Regionalism". The Economic Journal, 108. 1998. Fernandez, R. Returns to Regionalism: An Evaluation of Non-Traditional Gains from RTAs. New York : International Trade Division, World Bank and New York University. 1997. Ffrench-Davis, R. "Distorsiones del Mercado y Teória de las Uniones Aduaneras". Integración Latinoameričana, 5. 1980. __________. "Macroeconomic Policies for Growth". CEPAL Review, 0. 1996. __________, O. Muňoz and G. Palma. "The LatinAmerican Economies: 1950-1990". In: L. Bethell (ed.), Cambridge History of LatinAmerica, Vol. 6. Cambridge : Cambridge University Press, 1994. Second edition in: L. Bethell (ed.), Latin America: economy and society since 1930. Cambridge University Press. 1998. Garay, L. J. and A. Estevadeordal. Protection, Preferential Tariff Elimination andRules of Origin in the Americas. Washington, D.C. : Integration, Trade and Hemispheric Issues Division, Inter-American Development Bank. 1995. Inter- American Development Bank - IDB. "Trade Liberalization". 1996 Report in Economic and Social Progress, Part II, Chapter 2. Washington, D.C. : IDB. 1996a. _________. Integration and Trade in the Americas, Periodic Note, December. Washington, D.C. : Integration, Trade and Hemispheric Issues Division, IDB. 1996b. __________. Integration and Trade in the Americas, Periodic Note, July. Washington, D.C. : Integration, Trade and Hemispheric Issues Division, IDB. 1997. __________. Integration and Trade in the Americas, Periodic Note, August. Washington, D.C. : Integration, Trade and Hemispheric Issues Division, IDB. 1998. Mistry, P. Regional Integration Arrangements in Economic Development. The Hague : FONDAD. 1996. Panagariya, A. "The Free Trade Area of the Americas: Good for Latin America?". The World Economy, 19. 1996. Puga, D. and A. Venables. Trading Arrangements andIndustrial Development. Washington, D.C. : International Trade Division, World Bank. 1997. Serra, J., et al. Reflections on Regionalism. Washington, D.C. : Carnegie Endowment for International Peace. 1996. Simpson, J. "Statement". Partially reproduced in "US Treasury Official: NAFTA Origin Rules Shouldn't Be Model for FTAA". Inside NAFTA, 4. 1997. Vera, A. Transport Infrastructure in Datin America. Working Paper N° 221. Washington, D.C. : Integration, Trade and Hemispheric Issues Division, Inter-American Development Bank. 1997. Winters, A. Assessing Regional IntegrationArrangements. Washington, D.C. : International Trade Division, World Bank. 1997. Wonnacott, P. and R. WoNNACOTT. "Liberalization in the Western Hemisphere: New Challenges in the Design of a FxeeTxade Agreement". North American Journal ofEconomics and Finance, 6. 1995. World Trade Organization - WTO. Regionalism and the World Trade System. Geneva. 1995. Yeats, A. Does MERCOSUR s Trade Performance Justify Concerns About the Effects of Regional Trade Arrangements: YES! Washington, D.C. : International Trade Division, World Bank. 1996. Zahler, R. and C. Budnevich. Integration of Financial Services in Datin America. Santiago, Chile : Zahler and Co. 1997. INTAL PUBLICATIONS REGULAR PUBLICATIONS Integration & Trade. Three journal issues (in English and Spanish) by subscription including twelve issues of INTAL Monthly Newsletter (in English, Portuguese and Spanish). MERCOSUR Report. Annual publication (in English, Portuguese and Spanish). Informe Andino. Annual publication as from 1999. SPECIAL REPORTS Integración energética en el Cono Sur (Spanish). Mario A. Wiegers. 1996. Integracwn en el Sector Transporte en el Cono Sur (Spanish): Transporte Terrestre. Jose Sant'Anna. 1997. Puertos y vías navegables. Martin Sgut. 1997. Los ferrocarrilesy su contribución al comercio internacionál. Ian Thomson. 1997. WORKING PAPERS The Integration Movement in the Caribbean at Crossroads: Towards a New Approach of Integration. (English). Uziel Nogueira. 1997. MERCOSUL e Comércw Agropecuario (Portuguese). Ives Chaloult and Guillermo Hillcoat. 1997. Las relaciones de comercio e inversion entre Colombia y Venezuela (Spanish). Eglé Iturbe de Blanco. 1997. DISSEMINATION PAPERS Integración y Democracia en America Latina y el Canbe (Spanish). Alvaro Tirádo Mejía. 1997. Estado de evolución en la elaboración e implementación de las Normas ISO 14.000 y CODEX Alimentarius. (Spanish). Laura Berón. 1997. Evolución institucionaly juridica del MERCOSUR (Spanish). Vicente Gamelo. 1998. DATABASES DATALNTAL (CD-ROM) Sistema de estadisticas de comercio de America Base INTAL MERCOSUR (BIM) Base de datos bibliográficos (INTEG) Directono de las Relaciones Económicas de America Latinay el Canbe con Asia-Pacifico (CD-ROM) Instrumentos básicos de integración económica en America Latinay el Canbe-II INTAL/ITD PUBLICATIONS Una evaluación de la homogeneidad macroeconómicay del desarrollo de la región centroamencana. Florencio Ballestero. Documento de Trabajo INTAL-ITD 1. 1998. Towards an evaluation of regional integration in Latin America in the 1990s. Robert Devlin and Ricardo Ffŕench-Davis. Working Paper INTAL-ITD 2. 1998. ALCA: Unproceso en marcha. Nohra Rey de Marulanda. Documento de Divulgación INTAL-ITD 1. 1998. INT/ITD PUBLICATIONS WORKING PAPERS Common Market of the Southern Cone: MERCOSUR. Martin Arocena. Working Paper # 204. September 1995 (also available in Spanish). From Miami to Cartagena: Nine Lessons and Nine Challenges of the FTAA. Robert Devlin and Luis Jorge Garay. Working Paper # 211. July 1996 (also available in Spanish). Facts, Fallacies and Free Trade: A Note on Linking Trade Integration to Labor Standards. Donald J. Robbins. Working Paper # 214. May 1997. What can European Experience Teach Latin America About Integration. L. Alan Winters. Working Paper # 215. May 1997. Economic Integration and Equal Distribution. Willem Molle. Working Paper # 216. May 1997. Towards Free Trade in the Western Hemisphere: The FTAA Process and the Technical Support of the Inter-American Development Bank. Enrique V. Iglesias. Working Paper # 217. July 1997 (also available in Spanish). Convergence and Divergence Between NAFTA, Chile, and MERCOSUR: Overcoming Dilemmas of North and South American Economic Integration. Raul A. Hinojosa-Ojeda, Jeffrey D. Lewis and Sherman Robinson. Working Paper # 219. May 1997. Transport Infrastructure in Latin America. Arturo VeraAguirre. Working Paper # 221. July 1997 (also available in Spanish). MERCOSUR: Achievements and Challenges. Carlos Sepúlveda and Arturo VeraAguirre. Working Paper # 222. September 1997 (also available in Spanish). SPECIAL PUBLICATIONS Integration and Trade in the Americas (Periodic Note). July 1995, February 1996, August 1996, December 1996, July 1997, December 1997 and August 1998 (also available in Spanish). Extract from the 1996 Report on Economic and Social Progress in Latin America, Part II, Chapter 2: Trade Liberalization. 1996 (also available in Spanish). European Economic and Monetary Union: Recent Progress and Possible Implications for Latin America and the Caribbean. March 1997 (also available in Spanish). Globalization and Regional Integration: Consequences for Latin America. Enrique V. Iglesias. November 1997 (also available in Spanish). Protection, Preferential Tariff Elimination and Rules of Origin in the Americas -An Overview. Luis Jorge Garay and Antoni Estevadeordal. June 1995 (also available in Spanish). The New Face of Regional Integration in Latin America and the Caribbean. Enrique V. Iglesias. July 1997 (also available in Spanish). Free Trade Area of the Americas: From Miami to Belo Horizonte. Enrique V. Iglesias. May 1997 (in English, Portuguese and Spanish). Transpacific Partnership: Latin America's Role. Enrique V. Iglesias. September 1997 (also available in Spanish).