Regional arrangements among the industrial countries generally aim to improve the welfare of participating countries by reducing or eliminating restrictions impeding closer integration of members’ national markets. The arrangements under review have varied in terms of the model of integration adopted, their intermediate objectives, the integration instruments used, and the scope and depth of integration sought. (See Box 1 for information on the arrangements reviewed, and the Appendix for their membership.)

Regional arrangements among the industrial countries generally aim to improve the welfare of participating countries by reducing or eliminating restrictions impeding closer integration of members’ national markets. The arrangements under review have varied in terms of the model of integration adopted, their intermediate objectives, the integration instruments used, and the scope and depth of integration sought. (See Box 1 for information on the arrangements reviewed, and the Appendix for their membership.)

Differences among arrangements have existed at one point in time, and also within and among arrangements over time. Regional integration arrangements have been modified in response to experience, and to changes in the global economic environment and in the economic policy orientations of participant countries. These arrangements have covered trade in goods and, more recently, services. The treatment of “sensitive” sectors such as agriculture, where protection is relatively high, has differed between free trade areas and customs unions; in the former they are either not covered or subject to longer phase-in periods, while in the latter these sectors have been subject to common policies that have resulted in the harmonization of distortions. As tariffs and other border measures have been reduced—both multilaterally and within regional arrangements—increased attention has been given within regional arrangements among industrial countries to other policies that impede access to markets; these include subsidies, standards, customs procedures, government procurement, investment, and competition laws. Intra-area factor mobility has also been included, but commitments on capital mobility have generally exceeded those on labor mobility.

With regard to the models of economic integration, the EC from the outset emphasized the common market model, with the 1957 Treaty of Rome calling for the eventual achievement of unrestricted movement among members of goods, services, labor, and capital—the “four freedoms”—as well as the adoption of common commercial policies vis-à-vis the rest of the world. More recently, with the increased harmonization of policies, the EC has taken important steps in the direction of an economic union. Further policy harmonization is envisaged under the announced plans to establish a monetary union. By contrast, the other cases under review (Australia-New Zealand Closer Economic Relations Trade Agreement (ANZCERTA), United States-Canada, and the EFTA) are free trade areas (FTAs), involving liberalization within the region without requiring a common commercial policy vis-à-vis nonmembers.

The instruments and institutional arrangements for integration have varied according to the model of integration. Within the EC, there is significant sharing of sovereignty among members through the establishment of a body of community law as well as supranational institutions and mechanisms for common decision making and enforcement. The EC’s integration instruments have therefore emphasized common policies and coordination of national policies. Under the EC 1992 single market program, for example, the removal of remaining barriers to free circulation of goods, services, capital, and people procedurally involves the “transposition” into national legislation of “directives” proposed by the Commission—the EC’s administrative and technical body—and adopted by the EC Council. By contrast, the FTAs under review are not supported by significant supranational in stitutions or elaborate mechanisms for common decision making, although in all cases special institutional arrangements for dispute settlement and for ongoing consultations on matters of common interest have been established.

The coverage of regional arrangements among industrial countries has been fairly broadly based; in most cases trade in both goods and services and investment issues have been covered. In the goods markets, the arrangements have called for an across-the-board removal of conventional barriers to trade (tariffs and quantitative restrictions), within specified time periods and, by and large, in an automatic manner. “Sensitive” sectors—those enjoying significant protection and needing considerable structural adjustment—have been handled somewhat differently by FTAs and by the EC. Under FTAs, nontariff barriers on sensitive sectors (for example, agriculture, textiles and clothing, leather products, and steel) have often not been removed, while tariffs have been subject to reductions over relatively longer time periods. (In the case of the United States-Canada FTA, “cultural” industries have been excluded because they are considered sensitive by Canada.) In the EC, by contrast, sensitive sectors have tended to be subject to common policies, which have harmonized distortions, although efforts have recently been intensified to reform and liberalize these policies.

Selected Regional Trade Arrangements in Industrial Countries1

article image
Note: QRs = quantitative restrictions.

See Appendix for membership.

Data from World Bank, World Development Report, 1991: and IMF, Direction of Trade Statistics.

Agriculture is a prominent example of a sensitive sector that illustrates the differences in treatment between FTAs and the EC. The EFTA has not included agricultural products in its intra-area liberalization program, as this sector is heavily protected and subsidized at the national level by many EFTA countries. Although the United States-Canada FTA calls for the eventual reduction to zero of all tariffs on agricultural goods and for the elimination of a few nontariff barriers (for example, voluntary export restraint arrangements (VERs) on meat products), it does not cover the major protective farm policies in these countries. In the case of the ANZCERTA, agriculture is already substantially liberalized inside the FTA and also vis-à-vis the rest of the world, although a few areas of agriculture remain protected in Australia. By contrast, the EC has from the outset maintained a highly controlled common market for agricultural products under its Common Agricultural Policy (CAP), which includes a system of support prices, variable levies on imports, and subsidies for exports to compensate for differences between support prices and world market prices. (Other sensitive sectors subject to common policies in the EC include coal, steel, and shipbuilding.)

With most tariffs reduced to relatively low levels in industrial countries and conventional quantitative restrictions on manufactures eliminated in most industrial countries through successive rounds of GATT negotiations, other government-induced impediments to the integration of national markets—including those associated with different levels of subsidies, cumbersome customs formalities, technical and regulatory standards, and national preferences in government procurement practices—have become binding. Regional integration arrangements among industrial countries have increasingly incorporated provisions to reduce all or some of these barriers. In the area of subsidies, the ANZCERTA has gone the farthest among FTAs; effective July 1990 all subsidies and bounties on intra-area exports were eliminated. The EC, under its single market program, has stepped up efforts to reduce distortions in trade and investment arising from different subsidy practices across members by tightening the enforcement of its competition policy (Xafa and Kronenberg, forthcoming). The United States-Canada FTA includes a commitment to phase out subsidies affecting bilateral trade in the automotive sector; however, this FTA could not resolve any other key subsidy issue “because the political costs of removing the [subsidy] programs far outweighed the concessions that could be offered by the other country” (Schott, 1988, p. 31).

The areas of standards, customs formalities, and government procurement have received increasing attention in regional arrangements among industrial countries. Over half of all directives proposed in the context of EC 1992 concern technical and regulatory standards (OECD, 1990). The EC’s general approach in this area has been to set “minimum” community-wide standards where necessary and to apply the principle of mutual recognition of national standards in the rest of the cases. As part of the program to create a European Economic Area, the EFTA and EC aim to harmonize technical standards through the European Committee for Standardization (CEN), the Committee for Electro-Technical Standardization (CENELEC), the European Telecommunications and Standardization Institute (ETSI), and through an organization for testing and certification (The European Organization for Testing and Certification, EOTC). The ANZCERTA includes explicit commitments to harmonize standards and other regulations that may impede trade. The United States-Canada FTA has no provisions in this area, although future negotiations are envisaged on technical standards, including testing and certification requirements. With regard to customs formalities, the EC as part of the single market program aims to abolish all such formalities, while the ANZCERTA includes commitments to harmonize customs processing procedures. All regional arrangements under review contain provisions to reduce discrimination in government procurement for member countries.

Recent regional arrangements among industrial countries have covered services and investment, areas hitherto not covered by GATT disciplines. Among FTAs, the United States-Canada agreement broke new ground by establishing firm contractual bilateral obligations to “level the playing field” in services, although the FTA was “better on rule making than on liberalization; almost all existing restrictions [were] grandfathered” (Schott, 1988, p. 31). The United States-Canada FTA includes a positive list of services to be liberalized, notably financial services (where the right of establishment for members” financial institutions and the elimination of certain Canadian restrictions to access are considered particularly important), but also in computer-based services, tourism, architecture, and professional labor services. In the investment area, the FTA commits Canada and the United States to extend national treatment to each other in most cases. The ANZCERTA broke new ground by establishing a negative list of services specifically excluded by each member country;20 capital movements between the two member countries are already relatively free. Both free trade in services and free movement of capital are key features of the single market program; directives have already been implemented covering air and road transport, telecommunications, audiovisual works, and financial services (OECD, 1990).

The EC has from the beginning been covered by a community-wide pro-competition policy, which is embedded in Articles 85–94 of the Treaty of Rome. Its enforcement has been significantly tightened under EC 1992 and a new Merger Control Regulation gives the Community authority to scrutinize all mergers and acquisitions that meet certain threshold requirements. Among FTAs, the ANZCERTA is unique in having incorporated (during the 1988 review of the agreement) a commitment to make business laws and regulatory practices of the two members compatible. This was facilitated by strong similarities in competition policies and legal systems, and by the high degree of pre-existing harmonization in areas such as restrictive business practices, intellectual property laws, and consumer affairs (OECD, 1990). Consistent with this agreement, the competition policies of each country apply to trans-Tasman trade rather than antidumping laws (but not countervailing laws). By contrast, in the absence of a commitment to harmonize competition policies, Canada and the United States continue to apply existing antidumping and countervailing laws to trade within the FTA. The FTA, however, established special procedures to deal with bilateral disputes arising in connection with antidumping and countervailing duty cases.

As noted above, FTAs among industrial countries have increasingly included undertakings to maintain and improve intra-area capital mobility, particularly through mutual concessions on financial services and investment. In this regard, FTAs have approached the philosophy of common markets. FTAs have not yet included, however, significant provisions to allow broad based intra-area labor mobility—a key feature of the EC common market. Nonetheless, in the case of the United States-Canada FTA, the provisions to liberalize trade in professional labor services implied a relaxation of temporary entry restrictions for professionals. Moreover, the ANZCERTA was signed against the background of relatively free labor mobility between the two partner countries (OECD, 1990). Provisions allowing free labor mobility within EFTA countries only exist within the Nordic countries (Abrams, and others, 1990).

Implementation Record

Regional integration agreements among industrial countries have been implemented on schedule and often ahead of schedule. The EC, for example, formally removed intra-union tariff and quota restrictions and established the common external tariff by July 1968, one and a half years ahead of the original deadline. Members that joined the EC subsequently (Denmark, Ireland, and the United Kingdom in 1973; Greece in 1981; and Portugal and Spain in 1986) have also adopted the common external tariff and removed tariffs and quota restrictions on intra-union trade in line with mutually agreed schedules. The EFTA also removed barriers to intra-area trade in manufactures in keeping with the original six and a half year timetable. A series of bilateral trade agreements between EFTA countries and the EC—most completed by 1973—have converted nearly all of Western Europe into an effective free trade area for most manufactured products.

The record of successful implementation by the EC of the Treaty of Rome during the 1960s and 1970s must be qualified. As was increasingly realized in the 1970s and 1980s, the formal removal of tariffs and conventional quota restrictions on intra-EC trade did not lead in all cases to the intended “free circulation” of goods. With intra-EC free trade established as a legal requirement in the Treaty of Rome, barriers to the free circulation of goods tended to involve less visible restrictions, including those associated with different subsidy practices among members and discriminatory public procurement practices, as well as those resting in the “nooks and crannies of administrative arrangements” (Holmes and Shepherd, 1983). Similarly, the formal adoption by the late 1960s of a common external tariff vis-à-vis the rest of the world did not effectively harmonize external protection for EC member countries. The degree of trade openness has continued to vary substantially among members owing to a significant number of sector-specific bilateral restraint arrangements maintained at the national level, including those maintained under Article 115 of the Treaty of Rome.21 These restrictions—which typically have taken the form of “voluntary” export restraint arrangements (VERs), and “voluntary” undertakings by foreign suppliers to maintain a minimum price—have mainly affected such sensitive areas as agriculture, textiles and clothing, automobiles, and consumer electronics. In this regard, the EC has operated during most of its existence as an imperfect free trade area more than as a common market (El-Agraa, 1988). The single market program aims to remove the remaining obstacles to a truly unified EC market.

The implementation of more recent regional initiatives (ANZCERTA, EC 1992, and the United States-Canada FTA) has been broadly on schedule. In the case of the ANZCERTA, all tariffs, quantitative restrictions, and export subsidies on trans-Tasman trade in goods were fully phased out as of July 1990, five years ahead of the original deadline. In the case of the United States-Canada FTA, there have been two accelerated reductions in tariffs. In the case of EC 1992, about 70 percent of the 282 proposals submitted by the EC Commission had been approved by the EC Council by mid-1991. The transposition of EC directives approved by the EC Council into the national legislation of member countries has been slower than initially expected. There are, however, indications that the pace of transposition has been accelerating; as of mid-1991, about 70 percent of all approved directives had been transposed into national legislation. A few major difficulties remain in completing the single market program, including in the area of free movement of people (particularly nonprofessional workers and refugees), the harmonization of indirect taxation, and the treatment after 1992 of imports in some sensitive sectors currently subject to national restrictions.22 In the case of Japanese vehicle imports, national restrictions previously applied by four member countries will be converted into an EC-wide restriction.23

Effects of Arrangements

Regional trading arrangements among industrial countries have undoubtedly been more successful than arrangements among developing countries in liberalizing intra-area trade, particularly in manufactures (see next section and Charts 1 and 2).

Chart 1.
Chart 1.

Trade Developments in Industrial Country Regional Arrangements

(As a percent of regional GDP)

Sources: World Bank, World Development Report, 1991; and IMF, Direction of Trade Statistics.1 The figures for 1960 are for the original six members, with a GDP of $190 billion. Figures for 1990 are for the EC 12, with a GDP of $6 trillion.
Chart 2.
Chart 2.

Trade Developments in Developing Country Regional Arrangements

(As a percent of regional GDP)

Sources: World Bank, World Development Report, 1991; and IMF, Direction of Trade Statistics.

In sharp contrast to developing country cases (see next section), regional trading arrangements in the industrial world have typically involved countries that were previously highly integrated through trade. With the exception of the ANZCERTA, the regional arrangements under review have been established among countries that have previously been each other’s major trading partners (Table 5).

Table 5.

Selected Industrial Country Regional Arrangements: Intraregional Exports

(In percent of region’s total exports)

article image
Sources: IMF, Direction of Trade; International Financial Statistics; World Economic Outlook; and IMF staff estimates.

The more recent regional initiatives among industrial countries—namely, EC 1992 and the United States-Canada FTA—have not existed long enough to determine their relative effects on intraregional and extraregional trade. In the cases of the EC and EFTA, which have been in effect for more than 30 years, the importance of intraregional trade increased substantially in the first 10–15 years following the establishment of these arrangements but tended to level off or even decrease in subsequent years. In the case of the ANZCERTA, which has been in effect for less than ten years, the relative importance of intra-area trade has not increased significantly (Tables 5 and 6).

Table 6.

Selected Industrial Country Regional Arrangements: Changes in Trade to GDP Ratios

article image
Sources: IMF, Direction of Trade; International Financial Statistics; World Economic Outlook; and IMF staff estimates.

Chart 1 shows the rapid growth, particularly in intraregional trade, within the EC during the first two phases. Trade flows after the creation of the EFTA show relatively fast growth during the first decade, with stabilization at a lower rate after the United Kingdom joined the EC. Furthermore, EFTA exports to both the EFTA and EC have not increased relative to total exports. The trade openness ratios show, however, that international trade activities have gained importance in EFTA economies. For the ANZCERTA, while it is difficult to separate the impact of the trade agreements from those of privatization, deregulation, and unilateral trade liberalization, trade statistics show an acceleration of intra-area trade after the ANZCERTA signing, following the leveling off of intratrade in the latter years of the previous trade agreement. Between 1980 and 1990, intraregional trade as a percent of total trade increased to 7.4 percent from 6.4 percent. At the same time, Australia’s traditional trade surplus with New Zealand remained broadly unchanged in nominal terms. Although the trade openness of both countries decreased, this was due primarily to a substantial unit price decrease in both imports and exports. In fact, import volumes in New Zealand expanded considerably in the wake of the trade liberalization measures.

A number of elements explain relatively successful implementation and the greater impact of arrangements on trade within the industrial group. First, trade liberalization within industrial country regional arrangements has involved mainly intra-industry trade specialization (that is, trade in differentiated manufactures); such trade can be achieved without major shifts in factor proportions and entails relatively lower transitional adjustment costs, particularly in terms of labor dislocation.24 Second, while structural adjustment costs attributable to resource shifts in line with inter-industry specialization have been incurred in certain areas (particularly with intra-area liberalization among the more developed and less developed regions of the EC), these have been easier to sustain in the presence of fairly steady growth—partly reflecting the dynamic effects of regional integration. In the case of the EC, localized structural adjustment costs have been partly alleviated by a system of compensatory transfers—the so-called “structural funds.”25 Third, as noted earlier, structural adjustment in key sensitive sectors has been postponed because these sectors have not been covered by FTAs or because distortions have been harmonized under common policies within the EC (for example, the CAP) and in some cases within the context of broader multilateral agreements (for example, the Multifibre Arrangement or MFA).

By definition, regional trading arrangements provide preferential treatment for members and inevitably entail some discrimination against non-members. Discrimination in industrial-country arrangements has been contained in the past in several ways, including the fact that they have mainly affected trade in manufactures on which members generally maintained low levels of protection vis-à-vis nonmembers. In some cases, restrictions against nonmembers were also reduced as regional arrangements were implemented (the EC, for example, reduced restrictions on manufactured products in the context of multilateral trade negotiations, and Australia and New Zealand have reduced restrictions on agricultural and manufactured products, as well as on services on a unilateral basis). In other cases, restrictions against third countries were relatively low at the time that the arrangements were formed (as were those on manufactured products covered by the United States-Canada FTA). Trade diversion in some cases was also limited by the similar economic structures of the countries forming the agreement (United States-Canada FTA). In addition, with the exception of the ANZCERTA, regional arrangements among industrial countries have involved countries that were important trading partners prior to forming a preferential trading arrangement.

This said, however, some arrangements that liberalized intraregional trade maintained high protection—and thereby, significant discrimination—vis-à-vis nonmembers. The extent and nature of discrimination against nonmembers has varied among regional arrangements. Here, the stance of policies in the EC and in the United States-Canada FTA are globally significant given their importance in world trade. Even if these arrangements increase global welfare, they involve large negative effects for countries whose production and exports are concentrated in sectors where discrimination has increased or remains high.

Bilateral liberalization under the ANZCERTA has not led to notable discrimination against non-members, even though the previously relatively minor importance of intra-area trade in the ANZCERTA’s total trade (see below) might, in principle, have accentuated this risk. The absence of sizable discrimination reflects the fact that both Australia and New Zealand have been aggressively pursuing unilateral, nondiscriminatory trade liberalization—on goods and services, particularly financial services—simultaneously with the implementation of the ANZCERTA (OECD, 1990). Areas where discrimination is a concern include textiles and apparel, footwear, motor vehicles, and—in Australia—a few agricultural subsectors. Some quota restrictions and relatively high tariffs remain in these areas against third parties, even though tariffs and quota restrictions have already been fully phased out on virtually all trans-Tasman trade. The degree of discrimination in these cases, however, is being progressively reduced as Australia and New Zealand continue to implement their unilateral trade liberalization programs, which seek to eliminate all remaining quota restrictions and reduce all tariffs to low and fairly uniform levels. Other discriminatory elements in the ANZCERTA include mutual preferences in public procurement practices not extended to nonmembers, and in rules of origin. The latter are biased in favor of sourcing within the area, as products eligible for free trade within the region must have at least 50 percent of their value attributable to raw materials or value added from within the ANZCERTA. The discriminatory elements in the ANZCERTA are globally insignificant since this FTA accounts for only a minor share of world trade (about 1.5 percent in 1990).

The extent of discrimination against nonmembers of the United States-Canada FTA has been mitigated to a degree by the substantial integration of trade between the two countries prior to entering into the FTA (Chart 1); by relatively low average tariff rates on manufactures vis-à-vis nonmembers—although rates are somewhat higher in Canada than in the United States; and by the high proportion of bilateral trade already subject to zero tariffs prior to the agreement. The similar economic structures of the two countries also limit the scope for trade diversion while expanding scope for trade creation via intra-industry competition. (In the area of financial services, the FTA allows U.S. financial institutions a degree of preferential access to Canada. The ongoing liberalization and the opening up of the Canadian financial sector suggests, however, that this preferential access is unlikely to endure.) For example, in 1988, about 70 percent of U.S. exports to Canada, and nearly 80 percent of Canadian exports to the United States, entered the other country duty free; at the same time, Canada purchased nearly one fourth of U.S. exports and the United States was the destination of over three fifths of Canadian exports. However, some discrimination will persist as high tariffs or nontariff barriers remain for nonmembers for some products on which tariffs and nontariff measures on intra-area trade will be removed within ten years. These include textiles and clothing, footwear, some chemicals, soft plywood and furniture, steel, tires, and a few subsectors in agriculture and energy (OECD, 1990). Developing countries with a comparative advantage in these products would be most affected by this trade diversion. A multilateral agreement in the Uruguay Round to phase out the MFA and reduce tariff peaks on textiles and footwear would reduce the scope for trade diversion in these sectors. Discrimination against nonmembers is also likely to result from rules of origin, which are biased in favor of intra-area sourcing, and public procurement practices—which include bilateral concessions that go beyond the GATT Procurement Code—to which both countries are signatories.

In contrast to FTAs, the EC common market has by definition sought to maintain common trade policies vis-à-vis nonmembers. This approach has been a two-edged sword with regard to discrimination, depending on whether the creation of common external barriers—or their extension to new members—has resulted in higher or lower external protection. On balance, discrimination against third parties has been tempered by the reduction of the common external tariff to low levels on most manufactures through successive rounds of GATT negotiations. A low common external tariff has influenced members that joined the EC after its foundation—particularly Greece, Spain, and Portugal—to liberalize in areas where these new members” prior protection exceeded the common external tariff. (For instance, Spain’s average tariff on most-favored-nation (MFN) trade fell to 5 percent from about 13 percent as a result of joining the EC and adopting the EC’s common external tariff.)

Some areas (for example, agriculture, coal and steel, shipbuilding, and textiles and apparel) have enjoyed high levels of protection for decades, and maintenance of a common EC commercial policy vis-à-vis nonmembers appears to have raised average effective protection in those sectors for some EC members above levels in the absence of the common market.26 This problem is most pronounced in the case of the CAP, which has introduced substantial discrimination against nonmembers in virtually the entire range of agricultural tradable products. In this regard, the EC differs from FT As (other than the ANZCERTA), as the latter cover few agricultural subsectors. Protection and subsidization of farm products is high in many industrial countries participating in FTAs. By excluding agricultural products from intra-area liberalization, the FTAs neither help reduce this protection nor add to discrimination against non-members. The CAP sets intra-EC agricultural prices far above world prices, and generates structural oversupply—reflected in high stockbuilding and heavy subsidization of exports—and thus tends to depress world prices and to drive efficient producers out of business.27 In the absence of a substantial reform of the CAP, discrimination against third countries in agriculture may widen to the extent that new members with relatively lower protection in this area are admitted to the EC. Spain and Portugal had to raise their external protection on agricultural products significantly when they joined the EC and embraced the CAP in 1986.28 A similar situation may arise in the case of Sweden, which has recently applied for EC membership and currently plans to reduce protection substantially for agriculture over a three-year period that began at the end of 1991. The CAP is costly not only for foreign suppliers, but also for EC consumers, and it results in losses in productive efficiency within the EC owing to the misallocation of resources.

In addition to the areas mentioned above, there are concerns that other elements of the single market program may discriminate against nonmembers. These concerns are related to public procurement policies, the possibility that “reciprocal treatment” may be demanded from third countries whose firms seek access to the EC internal market, and the possible conversion of some national trade restrictions into EC-wide restrictions.

The extent of future discrimination would be affected by a number of factors. First, discrimination in regional arrangements in industrial countries tends to be concentrated in such sensitive areas as agriculture and textiles, currently subject to multi-lateral negotiations (see Chapter VI). A successful completion of the Uruguay Round could reduce discrimination against third countries associated with these arrangements and help ensure that they serve as building—rather than stumbling—blocks to global trade liberalization.

Second, the adverse static effects on nonmember countries attributable to trade diversion may be offset by dynamic gains generated by regional arrangements. As discussed in Chapter II, to the extent that these arrangements maintain an open external trade regime, nonmember countries may benefit from increased demand by member countries for outside imports; lower-cost exports originating in the regional grouping (achieved, for instance, through the exploitation of economies of scale); and access to a large regional market no longer segmented by administrative, technical, or regulatory barriers to trade. These benefits to non-members may, however, be dampened to the extent that regional groups achieve higher growth on the basis of “defensive” foreign direct investment—that is, investment diverted to member countries to enable foreign firms to “jump” existing trade barriers and hedge against the perceived risk that a regional grouping may adopt a “fortress mentality.” (Such investment might be considered a “second best’ solution, in that it results in a different pattern of investment than would occur in the absence of trade barriers.)

Third, the globalization of investment and production partially undermines the validity of basing the above analysis of discrimination on nation states as the main trading units. Foreign firms that have internationalized their operations and are already located within the geographic boundaries of the regional group may gain from discrimination. Profit repatriation and other less tangible returns (such as learning by doing in managerial and technical areas, market knowledge, and so on) could reduce, and perhaps reverse, the adverse effects of discrimination against third countries to which these firms belong. Lipsey (1990) and Peck (1989) have argued that non-EC multinational firms long since established in the EC have greater experience than many EC firms in treating the EC as a single market and are, therefore, relatively well positioned to take advantage of the unified market. A corollary is that third countries in the developing world that are not well integrated into the globalization process—largely as a result of their own domestic policies—stand to gain significantly less than other nonmembers from regional arrangements among industrial countries.

Quantifying Net Gains

Numerous studies have attempted to quantify the net gains and losses of regional integration arrangements among industrial countries. The impact of regional arrangements on third countries depends on whether trade diversion is outweighed by trade creation owing to both static and dynamic effects. This section summarizes the results of some such studies. It covers studies of arrangements where sufficient historical evidence provides the basis for an ex post analysis (the EC, EFTA, and ANZCERTA), and a number of studies on the prospective impact of the EC 1992 program and the United States-Canada FTA.

Ex Post Studies

Ex post studies have tended to focus on the static trade-diversion and trade-creation effects of regional trading arrangements. They have concluded that trade diversion exceeds trade creation for the EC and EFTA, with the reverse true for the ANZCERTA.

With regard to the EC, El-Agraa (1985a, 1985b, and 1988) argues that trade diversion exceeded trade creation in the first 10–15 years after the EC’s establishment and that trade diversion was largely concentrated in agriculture, where—as noted earlier—highly distortive protection was harmonized under the CAP. Petith (1977) concludes that the formation of the EC led to an improvement in its terms of trade vis-à-vis the outside world—partly because of the lack of retaliation by the latter. Furthermore, until the mid-1980s, researchers found no significant evidence that the formation of the EC had led to higher-than-otherwise foreign direct investment, except perhaps from Japan (see Mayes, 1985; and El-Agraa, 1988).

A number of studies have attempted to measure the impact of the EFTA agreement on its member countries. Studies undertaken in the early 1970s focused on trade-creation and trade-diversion effects during the EFTA’s first decade (EFTA, 1969, p. 24; Robertson, 1970, p. 91) and estimated trade creation to be higher than trade diversion (about $0.5-$2.0 billion, compared with $0.4-$0.9 billion in 1967). Later studies focusing on the period after the tariff cuts and the removal of quantitative restrictions did not find significant trade creation or diversion in EFTA countries. They showed, however, some trade diversion from trading regions outside the EC and EFTA (Utne, 1984; and EFTA, 1980).

Studies of the ANZCERTA show ambiguous results. Banks (1990), in a study of the effects on the Australian manufacturing sector, found the 1983 agreement provided more trade creation than trade diversion in Australia. The study also attempted to quantify the gains from increased specialization and economies of scale in both countries; it found that these sources of trade gains, as well as those provided by trade creation, were substantially higher for New Zealand than for Australia. Crocombe (1991), by contrast, focusing on the competitiveness of the New Zealand economy and based on a microeconomic approach, found few advantages of intra-area trade liberalization for the New Zealand economy.

Prospective Studies

In the case of EC 1992, the “Cecchini Report” (Cecchini, 1988) concludes that the single market program would lead to a once-off increase in the EC’s real income of between 2.5 and 6.5 percent by improving resource allocation. Peck (1989) argues that the Cecchini Report overestimates the onceoff gain from improved microeconomic efficiency by a factor of 2 to 3, partly on the grounds that it is doubtful that the single market program will be implemented in full. Baldwin (1989), by contrast, suggests that the Cecchini Report significantly underestimates the economic benefits because it ignores the indirect dynamic effects of EC 1992 over the medium and long run. He estimates that the direct efficiency gain of the single market adds 0.9–2.8 percentage points to Cecchini’s growth range in the medium term. The 1991 United Nations Conference on Trade and Development (UNCTAD) Trade and Development Report considers some of the spillover effects of EC 1992. Using a figure of 5 percent (in the middle of the Cecchini range) for EC income growth, UNCTAD estimates that this enhanced income could lead to $10 billion (7 percent of base-year exports) in increased exports from the developing countries to the EC.

Empirical estimates of the real income gains from trade within the United States-Canada FTA are highly sensitive to the underlying model used, whether macroeconomic or computable general equilibrium (CGE). Macroeconomic models typically allow for unemployment in factor markets and focus on the short- and medium-term impact of the FTA on such variables as output, investment, unemployment, and the trade balance. By contrast, CGE models usually assume full employment and allow for a detailed specification of complex micro-economic behavioral relationships and intersectoral and inter-country interactions. While macro-econometric models estimate the parameters, CGE models borrow the parameters from other studies or simply chose them on the basis of educated guesses. The numerical results of CGE models should be seen as simulations—rather than as forecasts—consistent with a given set of parametric and theoretical specifications.

Studies of the United States-Canada FTA based on macroeconomic models predict real income gains for Canada.29 Studies based on computable general equilibrium models provide conflicting results and are highly sensitive to theoretical specifications (particularly market structure and pricing behavior) and elasticity and scale parameters. Some models that assume constant returns to scale and other studies reach the opposite conclusion, although the size of the gains and losses are relatively small. Studies incorporating increasing returns to scale tend to suggest larger gains for Canada, but within a wide range (1–9 percent of real income). The inconclusive results regarding the impact of the United States-Canada FTA on economic growth in the two countries implies that its likely impact on third countries is also inconclusive. To the extent that real income gains in both countries result from free trade, an increase in the demand for imports from the rest of the world can be expected, which would offset at least part of any losses incurred by third countries owing to trade diversion resulting from relative price changes.

Although there is yet no general theory on imperfect competition and the dynamic effects of liberalization are still poorly understood, scholars increasingly agree that economies of scale and investment effects are key to understanding and assessing the impact of the United States-Canada FTA and the EC 1992 single market program. This view appears to be corroborated by available data on investment and corporate restructuring. For instance, according to Howell and Cozzini (1990), international merger and acquisition activity targeting Western Europe rose sharply from $9 billion in 1986, to $31 billion in 1988, and to $52 billion in 1989. Most of this activity came from Western European countries themselves, as intra-European cross-border merger and acquisition deals increased from $5 billion in 1986 to $22 billion in 1988, and to nearly $28 billion in 1989. In the case of North America, international merger and acquisition deals targeting this area also increased markedly, from $29 billion in 1986, to $68 billion in 1988, before declining to $60 billion in 1989. The intra-North America component, although relatively less important than in the European case, has been significant: cross-border merger and acquisition activity within North America doubled from $5 billion in 1986 to about $10 billion annually in 1988–89.


Upon adopting the 1988 Services Protocol, New Zealand’s exemptions included international aviation, telecommunications, shipping, stevedoring, and postal services. Australia’s exemptions included certain banking areas, airport services, construction, domestic and international aviation, telecommunications, coastal shipping, broadcasting and television, engineering and general consultancy, health insurance, and postal services (OECD, 1990). The review of the Services Protocol at the end of 1990 resulted in Australia eliminating the exemptions on construction, engineering, and general consultancy. New Zealand eliminated its inscriptions on stevedoring and telecommunications.


Article 115 empowers the Commission to authorize member countries to apply protective measures against imports from third countries if such imports threaten domestic production of the product concerned and cause injury. These authorizations temporarily restrict free circulation of goods within the Community.


National restrictions must be eliminated or replaced by EC-wide restrictions because the Commission signaled its intention not to give approval to member states” requests for Article 115 exemptions after 1992.


IMF (forthcoming) discusses the difficulties raised by the agreement on Japanese vehicles from the point of view of trade and competition policies.


Krugman’s (1981) model demonstrates that gains from trade are spread more evenly between capital and labor where similarity in factor proportions, and thus intra-industry trade, dominates. Ex post econometric evidence of links between intra-industry trade and regional integration in the EC and/or EFTA is found, among other things, in Havrylyshyn and Civan (1983), Balassa (1986), and Balassa and Bauwens (1987 and 1988). Greenaway (1989) discusses the general issue of the between regional trading arrangements and intra-industry trade.


Gordon (1991, p. 22) argues that, for all their problems, EC transfers via structural funds appear to have been “carefully targeted toward meeting their objectives [of] . . . assisting less developed and declining regions and removing structural rigidities from labor markets.”


Strictly speaking, textiles and apparel are not subject to an EC common policy to the same extent as agriculture, coal and steel, and shipbuilding. However, many import quotas on textiles and apparel under the MFA are negotiated for the EC as a whole and these seem appreciably higher than intra-EC barriers affecting this sector. Intra-EC trade in textiles, however, continues to be hampered by technical regulations and a complex scheme to distribute EC-wide quotas among members.


The conditions of access to EC agricultural and other controlled markets have varied significantly for nonmembers because of a complex hierarchy of preferential arrangements maintained by the EC. These arrangements include FTAs with individual EFTA countries; trade and cooperation agreements with certain Mediterranean, African, Caribbean, and Pacific States (ACP), and Eastern European countries; and the EC’s generalized system of preferences, which is granted unilaterally to developing countries. The reduction of discrimination against non-EC members participating in these preferential arrangements is, of course, at the expense of other non-EC countries.


For instance, tariffs on corn and sorghum went from about 20 percent to a tariff-equivalent of over 100 percent when Spain joined the EC (Callahan, 1989). As discussed in Callahan (1989), however, the lowering of Spain’s tariffs on manufactures to match the common external tariff of the EC was more important to nonmembers as a whole, and more than offset the adverse discriminatory effect of higher protection on agriculture.

  • View in gallery

    Trade Developments in Industrial Country Regional Arrangements

    (As a percent of regional GDP)

  • View in gallery

    Trade Developments in Developing Country Regional Arrangements

    (As a percent of regional GDP)