 |
Table of Contents |
 |
|
|
Asian Development Outlook 2006 :
I. Developing Asia and the World
Gauging the gains of Doha trade liberalization for the region
Viewed from a historical perspective, global free trade would appear an idealistic goal. As noted above, negotiations in the Doha Development Agenda have focused on a more modest agenda. To conjecture about the outcome in Doha, this section examines the economic impact in the context of a model that brings Asia into sharper relief. This exercise is conducted using an Asia- and trade-focused general equilibrium model (GEMAT), described more fully in Part 3.
1.5.4 Various Doha scenarios |
|
Deep Doha. This is essentially scenario 5 in Anderson et al. (2006).1 It assumes significant agreement in Doha by 2006 in a number of areas: nonagricultural tariff bindings are cut by 50%, and agricultural tariffs are cut using a tiered formula. The marginal cut is set at 45% for agricultural tariffs below 15%, 70% for tariffs within the 15—90% bracket, and 75% for tariffs above 90%.2 Agricultural export subsidies are eliminated for all countries. Domestic support for agriculture is also cut for the United States (by 28%), the European Union (by 16%), and Australia and New Zealand (by 10%). Special and differential treatment (SDT) is not applied and developing countries are assumed to liberalize to the same degree as developed countries. All these trade reforms are phased in over the 5-year period of 2007—2011. Possible reforms in services sector and trade facilitation are not incorporated in this scenario.
Doha-SDT. This scenario examines the consequences of including SDT. Specifically, there are four agricultural tariff brackets for developing countries, with inflexion points placed at tariff levels of 20%, 60%, and 120%. Their marginal rates of reduction are 35%, 40%, 50%, and 60% within each of the four bands. On NAMA, the cut in nonagricultural tariff bindings for developing countries is 33%. Least-developed countries are not required to undertake any reduction commitment. On the other hand, the required reductions in both agriculture and nonagriculture sectors for developed countries are the same as that in the Deep Doha scenario. The assumption regarding cuts in domestic support and export subsidies are also the same as those in Deep Doha.
Doha-SDT excluding sensitive agricultural products. Building on Doha-SDT, this "Doha Light" scenario further assumes a less ambitious agricultural agenda that excludes sensitive farm products. The developed countries are assumed to treat 2% of their HS6 agricultural tariff lines as sensitive and subject to just a 15% tariff cut. For developing countries, the corresponding figure is 4%.
1 See Anderson et al. (2006b) for the details of the design of this Doha scenario. The Doha scenario used here corresponds to their "Doha-All" scenario.
2 Since the cut is applied on bound tariffs, the cuts in applied rate may be small for some developing countries due to binding overhang.
|
Since the parameters (and, indeed, likelihood) of an agreement at Doha remains uncertain, the simulations adopt the ambitious assumptions about Doha outcomes that have been made by Anderson et al. (2006b) (Box 1.5.4). While a successful Doha package could spur further multilateral liberalization initiatives, these are not considered here.
Under the "Deep Doha" scenario, world income rises by $155.2 billion in 2025, measured in 2001 prices (Table 1.5.7). However, liberalization benefits developing Asia disproportionately. Around 70% of the estimated gains from Deep Doha accrue to Asia (including Japan). Box 1.5.5 compares ADB current estimates with those of earlier World Bank studies. Caveats about these exercises and their estimates are set out in Part 3. Importantly, as is further discussed in Part 3, the absolute values of these estimates tend to have a strong
downward bias, and, of course, are a function of the assumptions and parameters underlying the model. Hence, the absolute values are arguably less important than the rank ordering of the results and the relative magnitudes.
As Table 1.5.7 shows, the estimated gains from Doha are not evenly distributed either globally or within Asia. The disproportionately large gains accruing to Asia follow from Asia's openness and larger initial trade shares, as well as the assumption of a significant reduction in Asian import protection, especially in agriculture. Japan and Korea, which are large economies in absolute size, capture the bulk of the gains. Relative to income, Korea and Thailand gain most. Korea benefits significantly from a sharp reduction in its own agricultural tariffs as well as from growing export opportunities in the PRC. Thailand benefits as its agricultural exports are boosted by reduced distortions in the global agricultural trading system.
Although Deep Doha is a positive-sum game, there are prospective losers as well as winners. In Asia, Bangladesh and Viet Nam would appear to face the prospect of income losses. And Deep Doha would appear to offer no benefits to the Philippines and few to the PRC. As the possibility of prospective losses may impede or slow multilateral liberalization initiatives, it is important to look behind these results.
An important aspect of the Doha negotiations, reflected in this scenario, is that negotiations are about reductions in members' legally bound tariffs, not the actual tariffs that they apply. Therefore, where there is water in the tariffs, nominal reductions in bound tariffs may mean little or no reduction in actual levels. In the Deep Doha scenario assumptions presented here, this is the case for both Bangladesh and the Philippines. An important result from standard economic theory is that liberalization gains depend largely on the boosts to efficiency and productivity that occur when countries liberalize their own trade (Figure 1.5.1), not from tariff reductions by their trade partners. By volunteering cuts in bound tariff rates that would lower applied tariffs, countries such as Bangladesh and the Philippines would also benefit. The negotiating approaches of WTO obfuscate this fundamental point.
Of course, even if countries do reduce their own tariffs, multilateral liberalization may not lead to gains in all cases. If liberalization means that a country loses preferential access to important markets, or export subsidies are removed on goods that have a large weight in the import basket, or subsidized entry into other markets is halted (as, for example, in the case of the Fiji Islands' sugar exports to the EU), countries may face the prospect of significant losses. The model suggests that the erosion of preferential access for clothing and textiles to the EU, with the EU lowering its most-favored-nation tariffs, could dampen Bangladesh's exports, resulting in losses in income and terms of trade. Bangladesh also loses from terms-of-trade effects driven by higher agricultural prices (including cotton). Also, given that the PRC already has substantially reduced its tariffs on merchandise trade, a deterioration of its terms of trade may be needed to sustain fast export growth. The case of Viet Nam, which is not yet a member of WTO, illustrates another point. In the Doha scenario, levels of protection in nonmember countries are assumed unchanged. Consequently, as WTO member countries liberalize, nonmember countries such as Viet Nam suffer from a diversion of their trade to other locations where costs are now lower. If Viet Nam successfully concludes negotiations to enter WTO, it too would benefit from Doha, raising estimated impacts.
Under the scenario of Doha-SDT (Table 1.5.8), the global gain in 2025 shrinks by around 30% in comparison with the Deep Doha scenario, to $110.6 billion. But for developing countries,15 the gain would be only $56.4 billion, only 63% of that in the Deep Doha scenario. If higher-income Asian countries (Hong Kong, China; Korea; Singapore; and Taipei,China) are excluded, the gain of developing countries is only $19.3 billion, or less than half of that in the Deep Doha scenario (Figure 1.5.2). This exercise suggests that no regions (as defined) would be better off from the introduction of SDT, even for LDCs like Bangladesh. This result squares with the theoretical proposition that developing countries need to cut their own trade protection to reap the benefits of multilateral trade liberalization. SDT does not serve the interests of developing countries.
If the Doha-SDT scenario is further weakened by assuming that the sensitive agricultural products are subject to lower tariff cuts, global gains in 2025 would be reduced to $68.4 billion, reflecting the importance of agricultural liberalization in the Doha trade liberalization agenda. In Asia, the exception of sensitive agricultural products are important for Japan, Korea, Thailand, and, to a lesser extent, Viet Nam, given their high interests in agricultural trade liberalization. However, for other Asian economies, the exception of sensitive agricultural products only have marginal impacts on their welfare gains from Doha.
1.5.5 Gains from Doha: A Comparison with other quantitative estimates |
|
Care needs to be taken in comparing different estimates of the potential benefits of the Doha Development Agenda. Over the years, the World Bank has revised its estimates of Doha benefits significantly downward. The estimate here is also lower than earlier World Bank estimates (2001 and 2004). An important reason for this is that use of the GTAP V6 (Global Trade Analysis Project) database implies a reduction in baseline trade protection compared to the earlier GTAP datasets used in the initial World Bank studies. The benefits of the cessation of quotas on textiles and clothing and the entry of the People's Republic of China into the World Trade Organization are now embedded in the baseline, so Doha assumptions now remove fewer distortions (van der Mensbrugghe 2006).
But there are also technical differences between the estimate of the General Equilibrium Model of Asian Trade (GEMAT) and more recent World Bank estimates. These differences illustrate the sensitivity of model estimates to differences in technological and behavioral specifications, and parametric assumptions. In particular, estimated benefits are sensitive to assumptions about trade price elasticities (Armington elasticities), returns to scale and product variety, and to sector and geographic aggregation.
The most recent World Bank estimates are based on trade price elasticities that are about one third higher than those used here. The elasticities in GEMAT are closer to the traditional GTAP values, which have been estimated econometrically (Hertel et al. 2003). This difference in parameter assumptions tends to boost the World Bank's estimates of trade benefits.
On the other side of the coin, GEMAT's assumption of increasing returns and imperfect competition generates larger benefits than those derived in a perfect competition model with constant returns technology, such as that used by the World Bank. Aggregation also matters. Higher commodity and geographic aggregation in GEMAT cuts estimated benefits. For essentially an identical scenario, GEMAT's construction trims about 15% of the global benefits reported in recent World Bank studies.1
A recent study by the International Food Policy Research Institute shows that ambitious Doha round trade liberalization can induce a global real income gain of $104 billion in 2019, around 0.19% of world GDP in 2019. This study was conducted using MIRAGE, a global computable general equilibrium (CGE) model which also incorporates scale economy and firm-level productivity variety (Bchir et al. 2002). The lower estimates, in comparison with the GEMAT results, may be due to its assumption about the imperfect mobility of unskilled labor between agricultural and nonagricultural activities, which limits the gains from more efficient resource reallocation.
Another recent study by Polaski (2006) (the "Carnegie model"), using a static global CGE model, estimates global gains of $58.6 billion, or 0.19% of 2001 world GDP, according to its Central Doha scenario. This gain is larger than the equivalent static impact in GEMAT. The difference may be found in scenario design and model specification. The Central Doha scenario in Polaski (2006) assumes relatively modest agricultural trade liberalization and more ambitious manufacturing trade liberalization, which would induce larger gains from manufacturing liberalization and smaller gains from agricultural liberalization. Moreover, Polaski (2006) assumes unemployment in the urban unskilled labor market and a rural-urban wage differential in developing countries. This further magnifies the gains of some developing countries from manufacturing trade liberalization, as their manufacturing sector expands following the increased export opportunities induced by trade liberalization, but may induce losses for developing countries from Doha agricultural liberalization, which diverts unskilled labor away from manufacturing and into agriculture.
Simulation approaches clearly have limitations and until there is better information, estimates of the benefits of trade liberalization will necessarily be subject to important qualifications. However, the relative magnitude of the welfare impacts and the sources of gains and losses obtained from these modeling exercises do provide useful insights for policy analysis.
1 For Asian countries, the estimated income gains from GEMAT are generally larger than the World Bank estimates, reflecting that the impacts of low trade elasticities are more than offset by the introduction of scale and variety effects.
|
 Back
Key issues in the Doha Development Agenda | Next
Beyond Doha |
Publications 
Top of page