Abstract

An assessment of the economic costs and benefits of regional integration depends on the yardstick against which these are judged.3 Theoretical arguments and historical experience suggest that nondiscriminatory free trade is the “first best” policy option, both from global and individual country perspectives. This chapter identifies the economic benefits and costs of regional trading arrangements relative to the status quo, that is, relative to a policy of maintaining unchanged levels of protection.4 There are two types of gains and losses: static—those stemming from a once-and-forall reallocation of an existing stock of capital, labor, and other resources; and dynamic—those associated with the effects of regional economic integration on productive capacity and potential output (dynamic perspective). For simplicity, the discussion in this chapter focuses, unless otherwise specified, on the benefits and costs of customs unions—which entail liberalization of intraregional trade coupled with a common external trade policy vis-à-vis the outside world.

An assessment of the economic costs and benefits of regional integration depends on the yardstick against which these are judged.3 Theoretical arguments and historical experience suggest that nondiscriminatory free trade is the “first best” policy option, both from global and individual country perspectives. This chapter identifies the economic benefits and costs of regional trading arrangements relative to the status quo, that is, relative to a policy of maintaining unchanged levels of protection.4 There are two types of gains and losses: static—those stemming from a once-and-forall reallocation of an existing stock of capital, labor, and other resources; and dynamic—those associated with the effects of regional economic integration on productive capacity and potential output (dynamic perspective). For simplicity, the discussion in this chapter focuses, unless otherwise specified, on the benefits and costs of customs unions—which entail liberalization of intraregional trade coupled with a common external trade policy vis-à-vis the outside world.

Static Effects on Members

The static effects of a customs union depend on the relative size of gains owing to trade creation and losses from trade diversion (and suppression).5 Trade creation occurs when a member replaces goods produced domestically at a relatively high cost before the union with goods imported from another member at a relatively lower cost. Trade diversion—or suppression—arises when low-cost goods previously imported from the outside world are replaced by higher-cost imports from other members, or where external protection after the union is raised by higher-cost domestic production (trade suppression). Whether one outweighs the other is essentially an empirical question.

On the supply side, gains from trade creation most frequently occur as resources previously engaged in costly production for the home market are reallocated to producing exports that enjoy preferential access to other members” markets (intra-union or internal trade creation). Gains may also arise by stimulating the production of exports for the outside world (extra-union or external trade creation). On the demand side, consumers gain from trade creation because of lower prices and a greater variety of goods and services. Losses from trade diversion accrue to consumers as imports are now purchased from suppliers within the union that are less efficient than suppliers outside the union. The higher cost of these goods offsets the lower internal market prices attributable to lower intra-union trade barriers. From the point of view of the member country, this trade diversion means that tariff revenues for the government (or quota rents to importers) are now paid implicitly as a subsidy production in other member countries. (The losses associated with diverting trade from more efficient suppliers outside the union may, however, be offset to some extent by improved efficiency in production in a member country, resulting from the availability of duty-free imports of intermediate inputs from other member countries.)

Trade diversion may also affect the terms of trade in favor of members of unions whose trade constitutes a large share of world trade. By buying fewer imports from—and selling fewer exports to—nonmembers, a large union would, other things being equal, lead to lower world prices for its imports and higher world prices for its exports. The terms of trade will not necessarily improve for each member; whether they do or not depends on the structure of trade and demand and supply elasticities.

Some gains from a customs union would also arise from cost savings from lower trade costs. In this case, members would capture the natural rents from being able to export to other members at lower transport costs than for exports to the rest of the world.

An additional source of static gains for exporters in the union are the rents from preferential market access (Wonnacott and Wonnacott, 1981). While exporters from the outside world sell their products to the union at the international market price, exporters from within the union can charge the price in effect in the market of the member country, that is, the world price plus the external tariff or the markup made possible by nontariff barriers. Of course, the counterpart to this rent is, for the importing member country, trade diversion and a loss of tariff revenue (or of importers” quota rents).6 But the exporting country would realize a net gain, provided that the exporter’s rent is greater than the dead-weight loss associated with expanding production of exports to the union at a marginal cost higher than world market prices.

In contrast with a customs union, a free trade area does not commit members to a common external trade policy and thus allows them to minimize trade-diversion effects. By lowering tariffs unilaterally, a member of a free trade area can switch imports back to more efficient suppliers; this action would, however, simultaneously undercut the value of its trade concessions (and thus of available rents) to other members. Rules of origin (typically based on national value-added requirements) are generally established in free trade areas to prevent “trade deflection,” whereby goods originating in nonmember countries are transshipped from the member with lowest external protection to members with higher levels of external protection. In contrast with free trade areas, customs unions appear to be more conducive to higher degrees of economic integration among members, especially when endowed with supranational structures for common decision making.

Transitional adjustment costs would be incurred as the economies in the union move from one equilibrium position to another. These costs may be substantial in the short run, with transitional welfare losses reflecting temporary labor unemployment and idle capacity during a period when firms—faced with competition from other regional suppliers—go out of business, while resources are reallocated to better uses only sluggishly. Clearly, such transitional losses would be less the greater the degree of labor and capital mobility within and among the members of the regional group.

The net static effects of a customs union on members” welfare depend on the relative importance of the elements discussed. If the focus is only on trade-creation and diversion effects, some key parameters that would maximize net gains may be identified a priori. Trade diversion would be minimized where initial protection is either very high or very low. Initially very high or prohibitive import barriers ensure diversion of few or no imports. Initially very low protection levels ensure that imports from within the union would only be marginally more costly than those from outside the region they displace. (Initially very low protection may also imply small gains from trade creation.) Diversion is also minimized if regional partners are important trading partners with each other prior to forming a preferential trading arrangement. Trade creation is greater the more responsive supply and demand conditions are in member countries to changes in prices induced by intra-union liberalization. This responsiveness would be positively related to the degree of flexibility in labor and capital markets, with the stage of development of securities markets of particular importance in allowing structural adjustment through, among other things, mergers and acquisitions.

Net gains will also be affected if members have a societal preference for industry; Cooper and Massell (1965) suggest that certain countries are willing to accept a degree of inefficiency associated with industrial protection. Under these circumstances, a customs union could serve to raise members” welfare by reducing excessive costs of inefficient industrial production while at the same time respecting the constraint of maintaining a desired level of industrial activity. This assumes, however, considerable room for industrial complementarity and for economies of scale within the union; it also assumes significant flexibility among member countries to reconfigure their industrial structure by “swapping” industries and specializing in lines that minimize costs. One conclusion from the Cooper Massell model is that welfare might be raised even if trade diversion and suppression outweigh internal and external trade creation. Trade diversion and trade suppression would be a signal of success if a regional trading arrangement is established mainly to promote import-substituting industrialization through wider—but still protected—markets. These welfare gains, however, would evaporate over the medium to long run to the extent that accumulated inefficiencies render import substitution nonviable as a development model.

The concepts of trade creation and diversion continue to underpin the theory on customs unions. However, as these concepts emphasize the welfare of—and trade between—geographically defined nation states, their relevance is increasingly undermined by the globalization of investment and production. For example, the gains from trade creation would also accrue to firms of nonmember countries with a physical presence (branches or subsidiaries) in the region or with other forms of linkage to firms in the region (licensing agreements, cross-shareholding arrangements, strategic alliances, and so on). By the same token, the losses from trade diversion may also affect firms owned by the residents of the region located in nonmember countries, or firms located in the region but with strong linkages to firms in nonmember countries.

In principle, there are two different—but not incompatible—scenarios under which trade creation might be maximized. First, trade—predominantly inter-industry trade—may expand on the basis of differences in resource endowments (and thus in productive structures) across members, that is, through specialization guided by comparative advantage. Alternatively, where production structures and factor endowments are similar, trade creation based on an expansion of intra-industry trade, product differentiation, and economies of scale is most likely among countries with: (1) similar factor endowments and production structures; (2) overlapping demand structures and relatively high per capita incomes; and (3) a relatively large combined market size. In either case, the potential for trade diversion is less where countries are important trading partners prior to implementing regional integration schemes.

Dynamic Effects on Members

Regional integration arrangements may lead not only to a one-time increase in income owing to static efficiency gains, but also to a sustained increase in the rate of growth of income. These dynamic effects may arise through a number of channels, including economies of scale; spillover effects (such as transfers of know-how from producer to user industries); increased competition; an improved investment climate; and a stepped-up pace of technological change. (Another aspect of dynamic gains from trade integration is related to consumption smoothing. Unrestricted access to world markets enhances the ability of agents to smooth consumption during business cycles.) Clearly, the potential dynamic gains for members from a regional arrangement would be greater to the extent that the arrangement goes beyond the reduction in tariff barriers toward higher degrees of integration in other areas, including the removal of remaining obstacles to the free circulation of goods and services, free factor mobility, and the harmonization of macroeconomic and other relevant policies.

While economies of scale may also be considered a static gain—by assuming unchanged stocks of capital, labor, and technology—their potential for generating dynamic gains bears emphasizing. Economies of scale are made possible by larger regional markets. Gains from scale economies internal to the firm would come through lower costs and increased productivity, as companies previously operating at below minimum efficiency can now expand output and move down their cost curves (in static models) and their learning curves (in dynamic models). Larger regional markets may also widen opportunities to achieve economy-wide or industry-wide economies of scale, which are based on such externalities as spillover effects (for example, transfers of know-how from producer to user industries) and may be strongest in high-tech sectors. Regional cooperation and market broadening through regional integration could also help secure scale economies in infrastructure—such as transportation and communications networks—and in research and development.

Economies of scale and spillover effects may provide a rationale for regional trading arrangements based on temporarily high external barriers. By allowing firms to move down their cost curves, temporary protection would serve as a springboard for subsequent export expansion.7 The argument hinges on regional markets being fairly large. But even then it is doubtful that a policy of “import protection for export promotion” would lead to sustainable gains. For one thing, trade policy instruments could not be shown to be optimal intervention tools in this context. For another, the policy carries serious risks. Under strong lobbying from vested interests, this policy could easily be turned into a defense of “sunset industries,” thus postponing gains from needed structural adjustment. Also, monopolistic firms nurtured by the initial protection may subsequently offer formidable resistance to the lowering of regional external barriers. Moreover, protection in a large regional bloc, even if temporary, would be likely to elicit retaliatory protection elsewhere. On balance, it would seem that outward-oriented regional trading groups would be much better suited than inward-oriented ones to generate sustained gains from economies of scale and spillover effects.

Significant dynamic gains may also stem from increased competition spurred by regional liberalization. Competition strengthens the reliability of relative prices as indicators of relative scarcities, which leads to more efficient and transparent markets and forces improvements in resource allocation. Even in the context of the oligopolistic and monopolistic market structures that tend to be associated with product differentiation and economies of scale, intensified competition within larger markets would limit social costs associated with collusion and other abuses of market power. Competition would also spur corporate restructuring and industry rationalization, as well as modernization through the adoption of superior technologies.

By providing larger markets through “binding” plurinational commitments, regional integration arrangements are likely to improve the investment climate. The attractiveness of investment in the region would increase for investors in member as well as nonmember countries, the latter also possibly reflecting defensive moves to avoid the adverse effects of trade diversion. Larger regional markets would increase investment opportunities in various ways, including by raising the profitability of innovation—as the fixed costs of research and development would be spread over a larger market—and facilitating the exploitation of economies of scale and scope. Also, the processes of corporate restructuring, rationalization, modernization, and technological change spurred by competition would further raise the level and efficiency of investment. Moreover, binding commitments supported by increasingly harmonized policies and regulations within the region would lead to reduced uncertainty; this would reduce the risk premia demanded by investors, lowering the value of postponing investment decisions, enlarging planning horizons, and so on.

Effects on Nonmembers

Regional groups, especially those that account for a sizable share of world trade, can have significant effects on nonmembers. The net impact is an empirical question and depends on net static and dynamic effects of trade creation and diversion. On the one hand, a large regional group may divert substantial amounts of trade from suppliers outside the region, shifting terms of trade against nonmembers, and possibly driving producers in nonmember countries out of markets where they may have a clear comparative advantage. Furthermore, if the union’s external barriers are high, or if there is a perceived high risk that a “fortress mentality” may develop within the union, foreign direct investment that otherwise would have gone to nonmembers may now flow to the union. This latter adverse effect on nonmembers would be accentuated, with an erosion of confidence in the future viability of multilateralism.

On the other hand, nonmembers may reap gains to the extent that the discrimination inherent in the regional group is low (that is, there are minimal external barriers to trade and investment) and progressively reduced by liberalization. Through economies of scale and efficiency gains induced by intensified competition, a large union may be able to export at lower cost; this would imply a favorable move in the terms of trade for nonmembers, offsetting to an extent the deterioration in terms of trade that nonmembers may experience as a result of trade diversion. Other channels for these gains to nonmembers include, among other things, the spillover effects of increased demand for outside imports by the regional grouping and the reduced cost of access to a large market no longer segmented by differential regulations, technical standards, and customs formalities.

3

Xafa (forthcoming) provides a comprehensive survey of the theoretical literature on economic integration, including the possible implications of the “new trade theory.” Gunter (1989) and El-Agraa (1988) provide other surveys.

4

Other policy counterfactuals against which the benefits and costs of regional trading arrangements may be measured include unilateral liberalization—opening up of one’s market to international trade without discriminating between trading partners and irrespective of their actions—and multilateral liberalization—reciprocal trade concessions among contracting parties to the GATT. Theoretical arguments and historical experience suggest that nondiscriminatory free trade is the “first best” policy, both from a global and individual country perspective. The greater benefits of nondiscriminatory free trade, relative to regionalism, would not have to be foregone, however, if regionalism is conducive to, and supportive of, multilateral liberalization.

5

The economic analysis of customs unions in terms of their trade-creation and trade-diversion effects dates back to the pioneering work of Viner (1950), Meade (1955), and Lipsey (1957).

6

From the point of view of the union as a whole, there are no net gains from rents associated with preferential access, except if such access allows union firms to realize efficiency gains from economies of scale that would not be captured otherwise (see next section on dynamic effects of customs unions).

7

This is Krugman’s (1984) “import protection as export promotion” argument applied to regional integration. Studies by Baldwin and Krugman (1986) and Venables and Smith (1986) suggest that this type of effect has been important for certain industries in industrial countries with large domestic markets.