Catalog

	Table of Contents
p. 73 of 77 BACK | NEXT Foreword, Acknowledgments, Contents, Acronyms and Abbreviations, Definitions I. Developing Asia and the World II. Economic trends and prospects in developing Asia III. Routes for Asia's Trade Introduction The drivers of trade and integration in Asia The rise of bilateralism Trade scenarios: Potential benefits and risks An agenda for trade and integration in Asia Conclusions Endnotes and references Statistical appendix

Asian Development Outlook 2006 : III. Routes for Asia's Trade

---

Trade scenarios: Potential benefits and risks

The global tide of bilateralism on which Asian countries are being carried could pose significant risks to the multilateral trading system. At one level, these risks stem from rules of origin that confer preferences and complicate investment and trade decisions. At another level, bilateralism could jeopardize multilateralism as countries refocus their energies on reaching reciprocal deals with their trading partners, to the exclusion of others. And where bilateralism coexists with high MFN tariffs, bilateralism may raise administered protection through, for example, the use of antidumping measures (James 2000). Whether bilateralism turns out to be a building block or a stumbling block to trade will ultimately depend on the underlying motivations and ambitions of the countries concerned, and how these shape individual agreements.

In this section, an attempt is made to compare multilateral liberalization scenarios with the status quo, and with the possible trajectory of trade and income in a world where the dynamics of bilateralism lead to the emergence of hub-and-spoke configurations. Of course, there is much that can be done to protect against the threats posed by bilateralism and, as is argued in the section An agenda for trade and integration in Asia, below, bilateral agreements have the potential to act as a catalyst for deeper integration.

Trade scenarios

The following subsections examine the potential income repercussions of several trade liberalization scenarios. The scenarios compare the impacts of possible reductions in tariff or tariff-equivalent barriers to merchandise trade at multilateral, regional, and bilateral levels. However, tariff barriers on merchandise trade are but one part of a broader integration agenda. Other ingredients could prove to be more significant, some of which may be more easily advanced on a regional and bilateral basis rather than in a multilateral setting. For example, there is reason to think that the liberalization of trade in services and cross-border investment, trade facilitation, and the movement of persons could all have a more pronounced impact on trade integration and income than a reduction in merchandise trade protection per se. The political economy of liberalization may also exercise an important influence on outcomes. If, for example, regional or bilateral agreements strengthen the bargaining position of exporters and "weed out" domestic protectionist influences, they may accelerate a more expansive liberalization agenda.

In theory, an upper bound on gains from liberalization of merchandise trade would occur in a world in which all tariff and tariff-equivalent barriers have been completely eliminated. Though probably naïve politically, a cooperative multilateral process that ends in global free trade is generally considered a "first best" solution and as such provides a useful benchmark against which other scenarios can be assessed. In the scenarios considered here (Box 3.4), it is assumed that all tariff barriers on merchandise trade are eliminated by 2025, with initial reductions (over baseline) starting in 2007.

This "perfect scenario" is contrasted with an "intermediate" scenario that postulates that "open regionalism" germinates in Asia. Open regionalism is represented by the Asian free trade scenario. As trade within Asia is liberalized, barriers to non-Asian partners are brought down in step, but barriers elsewhere in the global economy are unchanged. Again, it is assumed that moves toward the establishment of this hypothetical Asian Free Trade Area start in 2007, and are completed in 2025.

These multilateral and regional liberalization scenarios are compared with a scenario in which bilateralism proliferates. The basic dynamic is set out by Baldwin (2002). Essentially, it is assumed that the costs of not having a bilateral FTA with a trading partner increases as more competitor countries have them. Exporters that suffer losses tilt the political equilibrium in favor of a new FTA, triggering a "domino" or ripple effect in other countries. Since the political forces favoring liberalization are more likely to concentrate their energies on opening large markets ("hubs"), and protectionist forces are more likely to have success in resisting free trade with smaller, peripheral markets ("spokes"), the pattern that emerges is likely to have an overlapping "hub and spoke" configuration (Zhai 2005). Bilateralism leads to a point where large countries (hubs) have multiple FTAs with smaller trading partners (spokes), but the spokes tend not to liberalize among themselves. Such hub-and-spoke systems would contain multiple layers of discrimination, usually codified in complex rules of origin. Further, they may place the spokes at a disadvantage over time as free trade in the hub country favors the location of investment there but not in the spokes, as hubs gain more in terms of market access (Puga and Venables 1997, Baldwin 2002).

Precisely where bilateralism will lead is difficult to predict (as discussed in the previous and following sections). Two illustrative scenarios are considered. These are not intended as a prediction of likely future trading agreements within Asia, but rather to illustrate what the implications of hub-and-spoke systems might be relative to alternative multilateral and regional liberalization scenarios. Two-hub, multiple-hub, and overlapping-hub-and-spoke configurations could emerge, but here attention is confined to two stylized scenarios in which ASEAN and the PRC emerge as single Asian hubs. In each scenario, the hub gradually eliminates tariff and tariff-equivalent barriers with all other Asian countries over the period 2007 to 2025, but the spokes do not liberalize with each other.

Embedded in all scenarios are a number of common assumptions. First, it is assumed that conditions in the global economy remain generally benign, with industrial economies growing broadly in line with their potential rates. Although negative "shocks" are probable, their impacts are assumed to be quickly reversed. Also, any turbulence originating within the region is assumed to be short-lived. On a longer view, the structural transformations—for example, the massive population shift from countryside to towns that is likely to occur in Asia—that both cause and are a consequence of growth may themselves pose risks, and may eventually constrain or even derail progress. Each scenario assumes that institutions and policies adapt to new circumstances and that underlying transitions are successfully navigated. The associated challenges are undoubtedly serious and complex, but they are not of primary interest here.

A global "computable general equilibrium" model (GEMAT, Box 3.5)⁹ is used to trace the numerical impact of each scenario. Although computable general equilibrium approaches have their limitations, they are generally considered to be the best tool available for analyzing trade policy issues. To some extent, limitations are mitigated as the major interest here is the relative, rather than absolute, magnitude of impacts. The particular model used here embodies the most recent description of the structure of the global economy, including a comprehensive description of trade protection and preferences, i.e., the Global Trade Analysis Project (GTAP) release 6 dataset (Dimaranan and McDougall 2006). One important feature of this database is that its import protection data incorporate tariff preferences under PTAs such as the Everything But Arms initiative, under which Bangladesh's textile and clothing exports enter the EU market.

Baseline

The baseline scenario represents the status quo in terms of the global trade regime, but illustrates some of the more pronounced transformations that could occur over the next two decades in developing Asia. Although uncertain and contingent, these transformations echo earlier experiences of robust growth and economic catch-up in East Asia. Similar transitions are implied by the other scenarios.

In the baseline, per capita incomes in developing Asia continue to close the gap with developed-country averages. But given the wide gaps to start with, per capita incomes for most countries continue to lag far behind developed-country averages by 2025 (Figure 3.3). By 2025, baseline per capita income in the PRC is just 25% of that in developed countries, but this compares favorably to a starting ratio of 15% in 2005 (measured in constant 2001 purchasing power parity [PPP] terms). In terms of the PRC's aggregate income (also in PPP terms), the baseline suggests that it could become the largest national economy in the world in around 2015.¹⁰ If incomes are instead measured at market exchange rates, the PRC retains its momentum but its economic mass is downsized. At market exchange rates, the PRC could rank as the third-largest national economy in the world by 2025, after the US and Japan. With rapid economic growth, developing Asia's share in world GDP would increase steadily over the period from 28% to over 40% in constant 2001 PPP terms (Figure 3.4).

As productivity growth lifts income, the baseline suggests that there will be an exodus of workers out of agriculture. Growth of services absorbs almost all "surplus" agricultural labor, and services' share in output grows by more than manufacturing's. In India, services play a particularly important role in absorbing resources released from agriculture. In some countries, the model suggests that the movement of labor from low- to higher-productivity occupations could add as much as 0.5 percentage points to annual GDP growth.

The model suggests that the sources of growth may vary somewhat across Asia. In the PRC and East Asia, the contribution of labor force expansion to income growth is likely to taper off, and by 2015 could be negative. But in India, provided that jobs can be generated for young workers, growth could be buoyed by a "demographic dividend." In Southeast Asia, growth should accelerate in the next decade as the effects of the crisis years recede, and as investment rates and productivity growth revert to their longer-run trend. Capital accumulation continues to be an important growth driver across the region, but especially in the lower-income countries, including Indonesia, Philippines, and Viet Nam.

Global free trade

The welfare gains from trade liberalization would be maximized if all countries eliminated protection. Box 3.6 illustrates the various mechanisms through which trade liberalization can influence income.

The results of the global free trade scenario are presented in Table 3.4. By 2025, it is estimated that free trade in merchandise goods would raise world income by $341.4 billion dollars a year (in 2001 prices).¹¹ This translates to about 0.5% of baseline GDP, an estimate not dissimilar to that of Anderson et al. (2006).¹² But estimated gains for developing Asian countries are substantially larger than this average as a fraction of their baseline incomes.

A general feature of computable general equilibrium models is that they tend to suggest that changes in levels of protection (in either direction) have seemingly modest effects on income. But such models are likely to understate the gains from liberalization, possibly substantially. Most immediately, the estimates here allow for only limited impacts on factor productivity. If greater openness provides a direct boost to productivity or has other beneficial externalities, as many assume (e.g., Anderson et al. 2006), then estimated gains would measurably increase. Another point worth noting is that, at an aggregate level, the "level of ambition" embodied in the liberalization of merchandise trade is quite modest. Tariff rates are already low across much of the developed world, so that their elimination is unlikely to have large impacts on a global scale. However, for individual countries, impacts may be large. Also, there is assumed to be no change for trade in services, which is hobbled by countless frictions, and where perhaps the greatest potential gains from multilateral liberalization would lie (Part 1 of the Asian Development Outlook 2006).

It is also possible that the model may overstate some gains. In particular, no allowance has been made for any adjustment costs associated with liberalization. Where countries' capacity to adjust is seriously constrained, international development assistance—particularly technical assistance—may help countries design programs that allow them to capture better opportunities and influence positively their willingness to liberalize. Of course, if the political economy is such that those who stand to lose from liberalization can impede it, the scenario design is questionable. But from a narrow economic accounting perspective, adjustment costs, which are fixed and may be spread out over time, are likely to be dominated by the recurrent benefits of liberalization.

With global free trade, Southeast Asian economies such as Malaysia, Singapore, Thailand, and Viet Nam are major winners. By 2025, their real incomes rise by 3.5—5.5% of their GDP. Percentage income gains in South Asia are smaller. The income gain is 1.8% of baseline GDP in Sri Lanka, and 1.0% in India, but Bangladesh benefits only marginally. In part, the muted impact for Bangladesh reflects an assumption that it bears losses when its preferences in EU and other markets are ended. If, in reality, developed-country importers capture a large portion of the rents conferred by preferences, or complicated rules of origin deter exports (see The rise of bilateralism, above),¹³ the losses borne by countries like Bangladesh on the erosion of preferences may be more limited and their gains from liberalization commensurately larger. More generally, however, the limited scale of benefits in some South Asian countries illustrates some fundamental weakness of exporting sectors in these countries, suggesting the importance of structural and institutional reforms­ (as well as international development cooperation­) in gearing them to trade and increasing their openness. The benefits from trade liberalization would be magnified by such reforms. Likewise, steps to improve infrastructure and to lower trade costs that lift trade volumes would also intensify the benefits of liberalization (see the section An agenda for trade and integration in Asia, below).

Trade volumes and patterns undergo some significant shifts in the global liberalization scenario. By 2025, global merchandise trade expands by 10% over the baseline. Yet in developing Asia, the expansion is more pronounced due to its relatively high initial level of protection: merchandise trade expands by 20.1%, raising the weight of the region in world trade by 2.4 percentage points. Agriculture, food, and textiles and clothing emerge as major sources of export growth in developing Asia, reflecting the existence of large initial trade policy distortions in these sectors and the geographic locus of comparative advantage.

Notably, however, liberalization does not benefit electronics in Asia, perhaps the most dynamic sector of the last decade. Electronics exports from the region decline relative to the baseline. This is because the sector is already highly globalized, with complex international production-sharing agreements and geographically fragmented supply chains. Even today, electronics trade is dominated by the intra-industry shipments of intermediate goods, including parts, components, and semifinished products. Trade costs associated with logistics, customs clearance, and time delays are the binding constraints to growth for the electronics sector, not tariff distortions.

3.7  Sources of gains

The payoff from trade liberalization includes both static gains, following a reallocation of resources ("trade creation"), and "dynamic" gains associated with additional capital accumulation. In GEMAT, there are also gains that accrue from scale and variety effects that are important when markets are imperfectly competitive. The box figure provides a breakdown of estimated gains. In a straightforward model of perfect competition (with no scale or variety effects), estimated global income gains are just 60% of those estimated in the presence of such effects. Irrespective of the nature of competition, just over one third of global gains are attributable to accumulation effects on future consumption percolating from higher investment rates, which are lifted by higher income and saving. But for developing Asia, the dynamic gains account for about half the estimated gains. These larger gains from accumulation reflect higher propensities to save in Asia.

Box 3.7 explains the composition of gains at an aggregate regional level in terms of their static and dynamic components.

Asian free trade

Table 3.5 summarizes results for the Asian free trade scenario. The estimated benefits from free trade in Asia (with the extension of MFN to the rest of the world) are nearly as large as those from global free trade. It is also of note that no country or global region loses under the assumptions of Asian free trade. This result reflects the nondiscriminatory nature of the liberalization that is assumed in this experiment. Tariffs are lowered for exports from all countries to Asia in step with the reduction of tariffs within Asia.

A comparison of Asian free trade gains with those of the global free trade scenario shows that, for most developing countries in Asia, the bulk of gains from global liberalization originates from free trade within Asia. The point has already been made that for many countries, especially for South Asian countries and Viet Nam, the major benefit from liberalization comes from lowering their own tariff barriers, so this result should come as no surprise. For those countries that tend to have large intraregional trade shares, such as most ASEAN countries, the gains are larger still. But for Indonesia, its gains from Asian free trade are relatively small in comparison with those from global free trade, even though it has a high intraregional export share. This reflects the difference in commodity structure of exports between Indonesia and other major ASEAN countries, which results in higher tariff barriers faced by Indonesia in nonregional markets and, consequently, larger gains from global free trade for that country. For the PRC and Sri Lanka, the gains from Asian free trade are less than half those from global free trade. This reflects the PRC's strong trading links with markets outside the region, as well as both countries' small gains from unilateral liberalization due to their already low import protection.

In the case of Bangladesh and the Philippines, the Asian free trade assumptions generate estimated gains that exceed those from global free trade (Figure 3.5). Under the Asian free trade assumptions, Bangladesh retains preferential access for its garment exports to EU markets, which it loses in the global simulation. A similar story applies to the Philippines, which enjoys reduced or zero preferential tariff rates in the EU market in some sectors such as electronics, machinery, and other manufactured goods.

In spite of overall gains, the simulation suggests that Asian free trade could lead to a significant deterioration in the terms of trade for Asia's two mammoth economies, the PRC and India. Both these countries' major trading partners are outside Asia. The elimination of import tariffs on an MFN basis raises the PRC's and India's demand for goods produced in other parts of the world, but now there is no compensating increase in the demand for their exports outside Asia. This asymmetry lifts import prices relative to export prices, creating terms-of-trade losses.

In terms of the direction of trade, Asian free trade would increase regional integration on some traditional measures. The share of intra-Asian trade in global trade in 2025 rises to 19.3% compared to 15.3% in the baseline. But the "double-relative" measure of trade intensity presented in Table 3.6 suggests that the intensity of trade within Asia would decline slightly. However, this result has no welfare implications.

Asian hub and spokes (bilateralism)

Two hub-and-spoke configurations in Asia are simulated. The "PRC hub" configuration assumes that the PRC is a hub of Asian bilateral FTAs, and that all other Asian countries (including Japan) are the spokes of these FTAs. The "ASEAN hub" configuration assumes that ASEAN countries act as a hub collectively; also that ASEAN establishes its own free trade area and negotiates bilateral FTAs with other Asian countries as a whole. Implicitly, if ASEAN became a platform for bilateral agreements, it would connect some of the spokes that remain separated in the other bilateral hub configuration.

In both Asian hub-and-spoke configurations, all bilateral distortions—including tariff, tariff-equivalent nontariff barriers, and export subsidies—are phased out between hub and spokes over the period 2007—2025. Trade barriers are retained between spokes. As GEMAT does not include the impacts of rules of origin that can have high compliance costs and result in reduction of preference margin, other things equal, the simulation results may overestimate the welfare gains of these overlapping bilateral FTAs.

These two simulations forcefully illustrate the potential downside of bilateralism. Compared to the Asian free trade scenario that extends MFN status to countries outside Asia, the Asian hub-and-spoke scenario generates somewhere between one fifth and one quarter of the global gains. Within developing Asia, too, hub-and-spoke systems are inferior to regional free trade and to multilateral liberalization initiatives (Table 3.7). A PRC hub generates just over half the benefits of Asian free trade for Asian developing countries, and an ASEAN hub, which implicitly entails the removal of more trade distortions, about 70% of the benefits.

The inferiority of hub-and-spoke systems for Asian developing countries is most evident from the welfare impacts for South Asian countries, which gain very little or nothing from the hypothetical bilateral FTAs with hubs in East Asia and Southeast Asia. This result reflects a high level of own import protection in the countries of South Asia and their limited trade linkages with East Asia and Southeast Asia.

Only Korea appears to capture most of the benefits that might otherwise accrue from Asian free trade or global free trade through a bilateral agreement with the PRC. This is both because the PRC is one of Korea's most important export markets, and because their trade structures are largely complementary.

Although the two hypothetical hub-and-spoke configurations represent an inferior option toward trade liberalization in terms of global or regional gains, they generally induce larger benefits for the hub countries. The PRC's welfare gain from a regional FTA hub position is almost four times that from Asian free trade. ASEAN countries as well are better off as a hub than under the Asian free trade scenario. Therefore, potential hubs may have little incentive to pursue broader regional trade liberalization.

Back The rise of bilateralism	Next An agenda for trade and integration in Asia