TY - JOUR
AU - Tarr, David, G.
AB - Abstract The effects of the Uruguay Round are quantified using a numerical general equilibrium model which incorporates increasing returns to scale, 24 regions, 22 commodities, and steady state growth effects. We conclude that the aggregate welfare gains from the Round are in the order of $96 billion per year in the short run, but could be as high as $171 billion per year in the long run after capital stocks have optimally adjusted. Despite these global gains, we identify some developing countries that lose from the Round in the short run. In the long run, almost all gain, and the Round will allow developing countries to gain further through their own unilateral liberalisation. The Uruguay Round is a complex agreement comprising many elements, and with diverse impacts across countries. We employ a 24 region, 22 commodity applied general equilibrium model with two principal objectives: (1) to quantitatively assess the most important elements of the Round; and (2) to facilitate critical understanding of the results in our model, as well as why other estimates differ from ours. We find that the world as a whole gains substantially from the reforms of the Uruguay Round (UR): about $96 billion annually in our increasing returns to scale (IRTS) static model and $171 billion in our IRTS steady‐state model. However, the dollar gains are concentrated in developed countries, especially the United States, the European Union and Japan. In the static model the United States gains $13 billion (principally from removal of the Multi‐Fibre Arrangement), the European Union gains $39 billion (principally from the reduction of agricultural subsidies), and Japan gains $17 billion (principally from the reduction of import protection in agriculture). The overall explanation for this pattern is intuitive. The industrialised countries, especially the United States, Japan and the European Union, ‘gave up’ the most in the UR in the (misleading) mercantilist sense. In other words, these countries are reducing policies that are very costly in terms of foregone welfare to themselves, most notably Multi‐Fibre Arrangement (MFA) protection and agricultural distortions. On the other hand, developing countries reduce agricultural distortions relatively less and do not restrict imports under the MFA. The only general exception is that developing countries reduce protection in manufactures by more than OECD countries, since OECD countries have relatively lower benchmark protection on average in this area. Our decomposition of the effects shows that reduction of protection in manufactures is the reason that developing countries fare about as well as industrialised countries in terms of the percentage of their own low GDP levels. We evaluate the elimination of the MFA and find that it results in losses for developing countries as a whole, except when we assume implausibly high elasticities of demand for textile and apparel products in the markets of countries restraining imports under the MFA. The reason is that MFA liberalisation results in the transfer of MFA quota rents from developing to industrialised countries. For developing countries as a whole, their own efficiency gains, combined with terms‐of‐trade gains in unprotected third markets, are typically insufficient to overcome the quota rent losses. The adverse impact of the MFA removal is especially severe for the marginally inefficient exporters who also lose market share under free trade. However, the highly efficient developing country suppliers of textiles and clothing, led by China, are estimated to gain under the most plausible elasticity values. Unlike some previous studies, we find that some developing countries will be net losers from the UR in the short‐run. The comparative static losses for developing countries derive primarily from two effects: (1) the reduction of agriculture subsidies in the United States, European Union and EFTA, which results in terms‐of‐trade losses for some net food importing countries; and (2) the MFA liberalisation induces losses for some developing countries. As the poorest region in the world, we focus on Sub‐Saharan Africa. Notably, it is among the regions that lose in the short run. African trade negotiators were ‘successful’ in playing the mercantilist game of extracting concessions from the rest of the world, while not liberalising their own trade at all. Losses to Africa then follow from terms‐of‐trade losses from agricultural liberalisation and the end of the MFA. Among our most important results for Africa is that we show that high and non‐uniform domestic taxation (both explicit and implicit) in selected agricultural sectors is the most important distortion. Africa can make enormous gains from reduction of these distortions. To facilitate understanding, we begin, in Section I, with a description of our constant returns to scale (CRTS), perfect competition, comparative static model and database. Results using this ‘base’ model are presented in Section II; this includes a detailed evaluation of the MFA, and an analysis of the impact on Sub‐Saharan Africa. In Section III we present results from our preferred models that incorporate imperfect competition, increasing returns to scale (IRTS) and the steady state. We explain why we find that the impact of IRTS is considerably less important in the context of the UR than previously estimated. In Section IV, we compare and reconcile our results with other models of the UR, and show that the reasons for the differences in the estimates of different models are relatively easy to understand and evaluate critically. Through this sequence we offer a coherent statement of the mapping from alternative modeling assumptions to quantitative conclusions. I. A Multi‐Regional Trade Model We begin with a presentation of our ‘base’ model, which is static CRTS with 24 regions and 22 production sectors in each region, listed in Table 1. In order to assess the consequences of regional aggregation, for some simulations we also employ a 12‐region aggregation of the base model. Table 1 Sectors and Regions 22 Sectors (All Models) PDR Pdddy rice WHT Wheat GRO Grains (other than rice and wheat) NGC Non‐grain crops FOR Forestry, fishing, lumber, wood, paper and wool PCR Processed rice MIL Milk products TEX Textiles WAP Wearing apparel CRP Chemicals, rubber and plastics I_S Primary iron and steel NFM Non‐ferrous metals FMP Fabricated metal TRN Transport industry T_T Trade and transport CGD Investment good MEA Meat products and livestock ENR Energy and energy products MIN Minerals and mineral products FOO Food, beverages and tobacco MAC Machinery, equipment and other manufacturing SER Services and utilities 24 Region Model AUS Australia NZL New Zealand CAN Canada USA United States JPN Japan KOR South Korea E_U E.E.C. −12 IDN Indonesia MYS Malaysia PHL Philippines SGP Singapore THA Thailand CHN China HKG Hong Kong TWN Taiwan ARG Argentina BRA Brazil MEX Mexico LAM Rest of Latin America SSA Sub‐Saharan Africa MNA Middle East and North Africa EIT Eastern Europe and Former Soviet Union SAS South Asia EFTA Other European (EFTA, Switzerland, Turkey, South Africa) 12 Regional Model USA United States JPN Japan E_U Europe CHN China SSA Sub‐Saharan Africa MNA Middle‐East and North Africa EIT Eastern Europe and Former Soviet Union SAS South Asia OOE Other OECD 22 Sectors (All Models) PDR Pdddy rice WHT Wheat GRO Grains (other than rice and wheat) NGC Non‐grain crops FOR Forestry, fishing, lumber, wood, paper and wool PCR Processed rice MIL Milk products TEX Textiles WAP Wearing apparel CRP Chemicals, rubber and plastics I_S Primary iron and steel NFM Non‐ferrous metals FMP Fabricated metal TRN Transport industry T_T Trade and transport CGD Investment good MEA Meat products and livestock ENR Energy and energy products MIN Minerals and mineral products FOO Food, beverages and tobacco MAC Machinery, equipment and other manufacturing SER Services and utilities 24 Region Model AUS Australia NZL New Zealand CAN Canada USA United States JPN Japan KOR South Korea E_U E.E.C. −12 IDN Indonesia MYS Malaysia PHL Philippines SGP Singapore THA Thailand CHN China HKG Hong Kong TWN Taiwan ARG Argentina BRA Brazil MEX Mexico LAM Rest of Latin America SSA Sub‐Saharan Africa MNA Middle East and North Africa EIT Eastern Europe and Former Soviet Union SAS South Asia EFTA Other European (EFTA, Switzerland, Turkey, South Africa) 12 Regional Model USA United States JPN Japan E_U Europe CHN China SSA Sub‐Saharan Africa MNA Middle‐East and North Africa EIT Eastern Europe and Former Soviet Union SAS South Asia OOE Other OECD Open in new tab Table 1 Sectors and Regions 22 Sectors (All Models) PDR Pdddy rice WHT Wheat GRO Grains (other than rice and wheat) NGC Non‐grain crops FOR Forestry, fishing, lumber, wood, paper and wool PCR Processed rice MIL Milk products TEX Textiles WAP Wearing apparel CRP Chemicals, rubber and plastics I_S Primary iron and steel NFM Non‐ferrous metals FMP Fabricated metal TRN Transport industry T_T Trade and transport CGD Investment good MEA Meat products and livestock ENR Energy and energy products MIN Minerals and mineral products FOO Food, beverages and tobacco MAC Machinery, equipment and other manufacturing SER Services and utilities 24 Region Model AUS Australia NZL New Zealand CAN Canada USA United States JPN Japan KOR South Korea E_U E.E.C. −12 IDN Indonesia MYS Malaysia PHL Philippines SGP Singapore THA Thailand CHN China HKG Hong Kong TWN Taiwan ARG Argentina BRA Brazil MEX Mexico LAM Rest of Latin America SSA Sub‐Saharan Africa MNA Middle East and North Africa EIT Eastern Europe and Former Soviet Union SAS South Asia EFTA Other European (EFTA, Switzerland, Turkey, South Africa) 12 Regional Model USA United States JPN Japan E_U Europe CHN China SSA Sub‐Saharan Africa MNA Middle‐East and North Africa EIT Eastern Europe and Former Soviet Union SAS South Asia OOE Other OECD 22 Sectors (All Models) PDR Pdddy rice WHT Wheat GRO Grains (other than rice and wheat) NGC Non‐grain crops FOR Forestry, fishing, lumber, wood, paper and wool PCR Processed rice MIL Milk products TEX Textiles WAP Wearing apparel CRP Chemicals, rubber and plastics I_S Primary iron and steel NFM Non‐ferrous metals FMP Fabricated metal TRN Transport industry T_T Trade and transport CGD Investment good MEA Meat products and livestock ENR Energy and energy products MIN Minerals and mineral products FOO Food, beverages and tobacco MAC Machinery, equipment and other manufacturing SER Services and utilities 24 Region Model AUS Australia NZL New Zealand CAN Canada USA United States JPN Japan KOR South Korea E_U E.E.C. −12 IDN Indonesia MYS Malaysia PHL Philippines SGP Singapore THA Thailand CHN China HKG Hong Kong TWN Taiwan ARG Argentina BRA Brazil MEX Mexico LAM Rest of Latin America SSA Sub‐Saharan Africa MNA Middle East and North Africa EIT Eastern Europe and Former Soviet Union SAS South Asia EFTA Other European (EFTA, Switzerland, Turkey, South Africa) 12 Regional Model USA United States JPN Japan E_U Europe CHN China SSA Sub‐Saharan Africa MNA Middle‐East and North Africa EIT Eastern Europe and Former Soviet Union SAS South Asia OOE Other OECD Open in new tab Except for tariff data discussed below, the data employed to calibrate the model come primarily from the GTAP database for 1992 (Version 2) documented in Gehlhar et al. (1997). The 24 region version of the model retains all the regions of the GTAP database. The sectors were selected such that those in which significant reduction in distortions occurred are retained as individual sectors. This should minimise aggregation bias. Production entails the use of intermediate inputs and primary factors (Labour, Capital and Land). Primary factors are mobile across sectors within a region, but are internationally immobile. Each region has a single representative consumer, as well as a single government agent. We assume Constant Elasticity of Substitution (CES) production functions for value added,1 and Leontief production functions for intermediates and the value added composite. As depicted in Fig. 1, demand is characterised by nested CES utility functions for each agent, which allow multi‐stage budgeting. Demand at the top level, for the composite ‘Armington’ aggregate of each of the 22 goods in Table 1, is Cobb–Douglas. Important demands are modelled with a traditional Armington formulation; imports by source substitute with each other in a CES utility function, and the important composite and domestic output substitute in a higher‐level CES utility function. In our model the important composite in each region is a composite of the import good from the other 23 regions. Firm varieties enter in the imperfect competition model described in Section III below, where competition is at the level of the firm. Fig. 1. Open in new tabDownload slide Firm level product differentiation within an Armington structur. Fig. 1. Open in new tabDownload slide Firm level product differentiation within an Armington structur. Relying on our a priori beliefs as to plausible values for these elasticities,2 we assume that the lower‐level elasticity of substitution σMM is 8 and the higher‐level elasticity σDM is 4. These values are systematically perturbed in the sensitivity analysis presented in Harrison, Rutherford and Tarr (1995), but unless otherwise discussed (as in the MFA analysis) are held at these values throughout. Exports are not differentiated by country of destination in the base model. All distortions are represented as ad valorem price‐wedges. These include factor taxes in production, value‐added taxes, import tariffs, export subsidies, voluntary export restraints (represented as ad valorem export tax equivalents) and non‐tariff barriers (represented as ad valorem import tariff equivalents). Lump‐sum replacement taxes or subsidies ensure that government revenue in each region stays constant at real benchmark levels.3 The primary source of the data on pre and post‐UR import tariffs in our model is the database assembled by the International Economics Division of World Bank based on the Integrated Data Base (IDB) of the GATT Secretariat. The IDB is based, in turn, on submissions from the contracting parties (the GATT IDB is documented in GATT (1994)). We refer to this data set as the World Bank (WB) tariff data. For manufacturing sectors this database is a trade weighted aggregation, to the regions and sectors of the GTAP database, of the integrated database of the GATT Secretariat. For the agricultural sectors it was necessary to estimate the ad valorem tariff equivalents of the specific tariff offers incorporated in the commitments of the GATT contracting parties. These adjustments are documented in Ingco (1996) and Hathaway and Ingco (1995), and are incorporated in our basic data set on tariffs.4 These data consist of a matrix of import tariffs before and after the UR, showing the ad valorem tariff applying on imports of each good from each other region.5 Further details are provided in Harrison, Rutherford and Tarr (1995). Our study is distinguished from earlier estimates of the impact of the UR since our estimates are based on the actual agreed offers, rather than projections of what might be agreed. II. Results from the Base Model II.1. What Do We Model? The UR is a complex agreement which covers agreements in many areas: (i) tariff reductions in manufactured products; (ii) tariffication of non‐tariff barriers in agriculture and binding commitments to reduce the level of agricultural protection; (iii) the reduction of export and production subsidies in agriculture; (iv) the elimination of Voluntary Export Restraints (VERs) and the MFA; (v) institutional and rule changes, such as the creation of the World Trade Organisation (WTO) and safeguards, antidumping and countervailing duty measures; (vi) new areas such as Trade Related Investment Measures (TRIMs), Trade Related Aspects of Intellectual Property Rights (TRIPs), General Agreement on Trade in Services (GATS); and (vii) areas receiving greater coverage, such as government procurement. There are a number of studies that have qualitatively characterised the UR (see, for example, Francois et al. (1994), GATT (1994), United States Government Accounting Office (1994), Schott and Buurman (1994) and Baldwin (1994); we shall not repeat those descriptions. Rather we note that we quantify the impact of the UR in the first four of the above areas. To the extent that there are additional benefits or losses from UR changes in the other areas, our estimates are an underestimate or overestimate of the gains from the Round.6 Our quantitative assessment of pre‐UR and post‐UR distortions in each of the areas we model is provided in Harrison, Rutherford and Tarr (1995; Appendix A). In manufactures, tariff cuts are based on the submissions of the Contracting Parties to the GATT Secretariat, as discussed above. In the area of agricultures we assess the impact of the tariffication of non‐tariff barriers and the reduction of tariffs. In addition to tariff reduction the UR calls for a reduction of the budgetary outlay on export subsidies in agriculture by 36% for developed countries and 24% for developing countries at the tariff‐line level. We model export subsidies as ad valorem export subsidies and model their reduction as a percentage reduction in the ad valorem rate of export subsidies.7 The Aggregate Measure of Support (AMS) to agriculture, which includes subsidies to domestic production, should be reduced by 20% for developed countries (16.8% for the EU), and by 13% for developing countries. Some commentators have expressed skepticism if the AMS will actually bind the contracting parties. In our base simulations we reduce domestic input subsidies by these percentages. However, we decompose the impact of the reforms in agriculture into their various components so that we may evaluate the impact of the quantitative impact of the agriculture reforms with and without the reduction in domestic input subsidies. II.2. Our Base Estimates: 24 Regions, 22 Commodities, CRTS We begin with our CRTS static model. The aggregate welfare effects of the individual components of the UR, as well as the complete reform package (FULL), are reported in billions of 1992 US dollars in Table 2 for all 24 regions of the GTAP database with 22 commodities. We also report the welfare effects of the FULL scenario as a percentage of base GDP in each region, to help scale the results for economies of different size. Table 2 Welfare Effects in ‘Base’ CRTS Static Model (1992 $bn) Regions . AGR . MFA . MFRS . Full . Full (%) . Australia 0.7 0.0 0.4 1.1 0.4 New Zealand 0.3 0.0 0.1 0.4 1.0 Canada 0.2 0.9 −0.0 1.2 0.2 USA 1.7 10.1 0.8 12.8 0.2 Japan 15.2 −0.5 2.0 16.6 0.5 Korea 4.6 −0.5 0.5 4.6 1.5 European Union (12) 28.5 7.6 2.3 38.8 0.6 Indonesia 0.2 0.6 0.6 1.3 1.1 Malaysia 1.2 0.1 0.7 1.9 3.3 Philippines 0.6 −0.0 0.4 0.9 1.6 Singapore 0.6 −0.1 0.5 0.9 2.1 Thailand 0.7 0.1 1.7 2.4 2.1 China −0.6 0.9 0.9 1.2 0.3 Hong Kong 0.6 −1.7 −0.2 −1.3 −1.4 Taiwan 0.0 −0.5 0.8 0.4 0.2 Argentina 0.4 0.0 0.2 0.6 0.3 Brazil 0.3 −0.0 1.1 1.3 0.3 Mexico −0.0 −0.1 0.3 0.1 0.0 Latin America 1.4 −0.5 0.3 1.2 0.4 Sub‐Saharan Africa −0.3 −0.1 −0.0 −0.4 −0.2 Middle East and North Africa −0.4 −0.5 0.6 −0.4 −0.1 Eastern Europe & FSU −0.2 −0.6 0.5 −0.4 −0.1 South Asia 0.1 0.6 2.7 3.3 1.0 Other European 2.4 0.0 1.7 4.1 0.3 Developing Countries (Total) 9.2 −2.3 11.6 17.7 0.4 Industrialised Countries (Total) 49.1 18.3 7.2 75.2 0.4 World 58.3 16.0 18.8 92.9 0.4 Regions . AGR . MFA . MFRS . Full . Full (%) . Australia 0.7 0.0 0.4 1.1 0.4 New Zealand 0.3 0.0 0.1 0.4 1.0 Canada 0.2 0.9 −0.0 1.2 0.2 USA 1.7 10.1 0.8 12.8 0.2 Japan 15.2 −0.5 2.0 16.6 0.5 Korea 4.6 −0.5 0.5 4.6 1.5 European Union (12) 28.5 7.6 2.3 38.8 0.6 Indonesia 0.2 0.6 0.6 1.3 1.1 Malaysia 1.2 0.1 0.7 1.9 3.3 Philippines 0.6 −0.0 0.4 0.9 1.6 Singapore 0.6 −0.1 0.5 0.9 2.1 Thailand 0.7 0.1 1.7 2.4 2.1 China −0.6 0.9 0.9 1.2 0.3 Hong Kong 0.6 −1.7 −0.2 −1.3 −1.4 Taiwan 0.0 −0.5 0.8 0.4 0.2 Argentina 0.4 0.0 0.2 0.6 0.3 Brazil 0.3 −0.0 1.1 1.3 0.3 Mexico −0.0 −0.1 0.3 0.1 0.0 Latin America 1.4 −0.5 0.3 1.2 0.4 Sub‐Saharan Africa −0.3 −0.1 −0.0 −0.4 −0.2 Middle East and North Africa −0.4 −0.5 0.6 −0.4 −0.1 Eastern Europe & FSU −0.2 −0.6 0.5 −0.4 −0.1 South Asia 0.1 0.6 2.7 3.3 1.0 Other European 2.4 0.0 1.7 4.1 0.3 Developing Countries (Total) 9.2 −2.3 11.6 17.7 0.4 Industrialised Countries (Total) 49.1 18.3 7.2 75.2 0.4 World 58.3 16.0 18.8 92.9 0.4 AGR, Agricultural reform; MFA, MFA reform; MFRS, Manufacturing sector reforms; Full, Complete UR; Full (%), Complete UR as a percentage of base GDP. Open in new tab Table 2 Welfare Effects in ‘Base’ CRTS Static Model (1992 $bn) Regions . AGR . MFA . MFRS . Full . Full (%) . Australia 0.7 0.0 0.4 1.1 0.4 New Zealand 0.3 0.0 0.1 0.4 1.0 Canada 0.2 0.9 −0.0 1.2 0.2 USA 1.7 10.1 0.8 12.8 0.2 Japan 15.2 −0.5 2.0 16.6 0.5 Korea 4.6 −0.5 0.5 4.6 1.5 European Union (12) 28.5 7.6 2.3 38.8 0.6 Indonesia 0.2 0.6 0.6 1.3 1.1 Malaysia 1.2 0.1 0.7 1.9 3.3 Philippines 0.6 −0.0 0.4 0.9 1.6 Singapore 0.6 −0.1 0.5 0.9 2.1 Thailand 0.7 0.1 1.7 2.4 2.1 China −0.6 0.9 0.9 1.2 0.3 Hong Kong 0.6 −1.7 −0.2 −1.3 −1.4 Taiwan 0.0 −0.5 0.8 0.4 0.2 Argentina 0.4 0.0 0.2 0.6 0.3 Brazil 0.3 −0.0 1.1 1.3 0.3 Mexico −0.0 −0.1 0.3 0.1 0.0 Latin America 1.4 −0.5 0.3 1.2 0.4 Sub‐Saharan Africa −0.3 −0.1 −0.0 −0.4 −0.2 Middle East and North Africa −0.4 −0.5 0.6 −0.4 −0.1 Eastern Europe & FSU −0.2 −0.6 0.5 −0.4 −0.1 South Asia 0.1 0.6 2.7 3.3 1.0 Other European 2.4 0.0 1.7 4.1 0.3 Developing Countries (Total) 9.2 −2.3 11.6 17.7 0.4 Industrialised Countries (Total) 49.1 18.3 7.2 75.2 0.4 World 58.3 16.0 18.8 92.9 0.4 Regions . AGR . MFA . MFRS . Full . Full (%) . Australia 0.7 0.0 0.4 1.1 0.4 New Zealand 0.3 0.0 0.1 0.4 1.0 Canada 0.2 0.9 −0.0 1.2 0.2 USA 1.7 10.1 0.8 12.8 0.2 Japan 15.2 −0.5 2.0 16.6 0.5 Korea 4.6 −0.5 0.5 4.6 1.5 European Union (12) 28.5 7.6 2.3 38.8 0.6 Indonesia 0.2 0.6 0.6 1.3 1.1 Malaysia 1.2 0.1 0.7 1.9 3.3 Philippines 0.6 −0.0 0.4 0.9 1.6 Singapore 0.6 −0.1 0.5 0.9 2.1 Thailand 0.7 0.1 1.7 2.4 2.1 China −0.6 0.9 0.9 1.2 0.3 Hong Kong 0.6 −1.7 −0.2 −1.3 −1.4 Taiwan 0.0 −0.5 0.8 0.4 0.2 Argentina 0.4 0.0 0.2 0.6 0.3 Brazil 0.3 −0.0 1.1 1.3 0.3 Mexico −0.0 −0.1 0.3 0.1 0.0 Latin America 1.4 −0.5 0.3 1.2 0.4 Sub‐Saharan Africa −0.3 −0.1 −0.0 −0.4 −0.2 Middle East and North Africa −0.4 −0.5 0.6 −0.4 −0.1 Eastern Europe & FSU −0.2 −0.6 0.5 −0.4 −0.1 South Asia 0.1 0.6 2.7 3.3 1.0 Other European 2.4 0.0 1.7 4.1 0.3 Developing Countries (Total) 9.2 −2.3 11.6 17.7 0.4 Industrialised Countries (Total) 49.1 18.3 7.2 75.2 0.4 World 58.3 16.0 18.8 92.9 0.4 AGR, Agricultural reform; MFA, MFA reform; MFRS, Manufacturing sector reforms; Full, Complete UR; Full (%), Complete UR as a percentage of base GDP. Open in new tab We find that the world as a whole gains about $93 billion annually, and that the gains in dollar terms are concentrated in the developed countries, especially the United States, Japan and the EU, who gain $13, $17 and $39 billion, respectively. Some smaller countries gain significantly in percentage terms: Malaysia gains 3.3% of GDP, Singapore and Thailand about 2.1% each, and Korea and Philippines about 1.6% each.8 Although developing countries as a whole gain from the Round, a few LDC regions are estimated to lose on balance. We decompose and explain these results across countries in detail below, and focus on sub‐Saharan Africa in Section II.4, but the overall explanation for this pattern is intuitive: it is the industrialised countries, especially the United States and the EU, that ‘gave up’ the most in the UR in the mercantilist sense. That is, these countries are reducing policies that are very costly to themselves. The exception to this is the reduction of protection in manufacturing, since the OECD countries have relatively lower protection on average in this area. Thus, as shown in the scenario for manufacturing reform (MFRS), developing countries tend to gain more from the reduction of protection in manufacturing. This is the primary explanation why developing countries gain as much as the industrialised countries as a percentage of their own low GDP. On the other hand, the developing countries reduce agricultural distortions relatively less (although the reduction of production subsidies is important in some cases). Moreover, they do not restrict imports under the MFA. Indeed, the MFA is crucial to understanding the consequences of the UR on developing countries, so we provide below a detailed evaluation of the MFA results. In column AGR in Table 2 the combined effects of all the components of agricultural reform are assessed jointly. In Harrison, Rutherford and Tarr (1995) we decompose the impact of the agricultural reform package into reductions in export subsidies, production subsidies and import protection. We show that the gains to the EU of over $28 billion come from the reduction of its agricultural subsidies. On the other hand, Japan and Korea gain from the reduction of their high import protection. Although China and some of the aggregate developing country regions lose somewhat small amounts, there are surprisingly few loser, given the concern about losses of the ‘net food importing’ countries. This is explained by the fact that most regions have something to gain by reducing their own production subsidies and most also export some food, even if they are net food importers overall. II.3. The Impact of Removal of the Multifibre Arrangement The results of the MFA scenario show that although there are substantial benefits to the world from elimination of the MFA ($15.9 billion), in aggregate developing countries lose. The large gains to the United States, the EU and other quota‐constrained countries from MFA reform are simple enough to explain: they capture quota rents; and there are gains in efficiency from reducing excessive domestic production and increased consumption at lower prices. For households in the United States, for example, the aggregate price of domestic and imported wearing apparel products declines by 9% implying substantial reductions in the cost of living.9 Importers of textiles and apparel that were previously unconstrained, of which Japan is the largest example, lose from the removal of the MFA. Exporters divert sales to the previously constrained markets and this results in a terms‐of‐trade deterioration in previously unconstrained markets such as Japan. For quota‐constrained exporting countries, the welfare effects are mixed. There is: (i) the loss of quota rents in export markets that were constrained; (ii) the potential gain in efficiency to the extent that they shift resources into textiles and clothing, assuming they have an ex post comparative advantage in these industries; and (iii) the potential gain in terms‐of‐trade on sales of textile and apparel products to previously unconstrained markets such as Japan. The theoretical literature has tended to emphasise that since lost quota rents are a ‘rectangle’, and efficiency gains are only a ‘triangle’, exporting countries will likely lose unless there are significant terms‐of‐trade gains on sales in unconstrained markets. The share of sales to previously constrained markets will strongly influence this latter element.10 Our numerical results shows the importance of these elements, but also shows that elasticities are important for this latter element as well. In Table 3 we provide some information that facilitates understanding the results. The total value of export quota rents by country (to all their export destinations) is listed for both textiles and for clothing.11 The quota premia rates shown are the specific rates applying on exports to the EU or United States, which are the two most important constrained markets. Table 3 also displays a share parameter for the share of total exports (of the quota‐constrained countries) in quota‐constrained markets. The smaller this share the more likely an increase in rents in unconstrained markets will compensate for the loss of rents in constrained markets. Table 3 Quota Rents for Textiles and Clothing . Textiles (TEX) . Clothing (WAP) . . . . Quota premium (%) . . . Quota premium (%) . . Value of quota rent ($mn) . % exports constrained . EU . USA . Value of quota rent ($mn) . % exports constrained . EU . USA . Korea 119 16 10 10 555 55 19 23 Indonesia 97 24 17 12 512 52 48 47 Malaysia 65 100 12 10 330 100 32 37 Philippines 7 50 10 9 363 81 28 34 Singapore 7 11 10 8 365 100 28 31 Thailand 53 40 13 9 396 42 36 35 China 378 19 27 18 2,223 31 36 40 Hong Kong 48 13 8 8 1,249 100 16 18 Taiwan 95 13 12 8 515 81 22 19 Brazil 65 100 14 9 43 77 18 20 Mexico 41 60 14 9 181 99 18 20 Latin America 46 45 14 9 619 86 18 20 Middle East and North Africa 84 78 7 5 390 97 9 10 Eastern Europe and FSU 87 78 9 6 430 97 12 13 South Asia 566 46 27 18 1,375 85 36 40 . Textiles (TEX) . Clothing (WAP) . . . . Quota premium (%) . . . Quota premium (%) . . Value of quota rent ($mn) . % exports constrained . EU . USA . Value of quota rent ($mn) . % exports constrained . EU . USA . Korea 119 16 10 10 555 55 19 23 Indonesia 97 24 17 12 512 52 48 47 Malaysia 65 100 12 10 330 100 32 37 Philippines 7 50 10 9 363 81 28 34 Singapore 7 11 10 8 365 100 28 31 Thailand 53 40 13 9 396 42 36 35 China 378 19 27 18 2,223 31 36 40 Hong Kong 48 13 8 8 1,249 100 16 18 Taiwan 95 13 12 8 515 81 22 19 Brazil 65 100 14 9 43 77 18 20 Mexico 41 60 14 9 181 99 18 20 Latin America 46 45 14 9 619 86 18 20 Middle East and North Africa 84 78 7 5 390 97 9 10 Eastern Europe and FSU 87 78 9 6 430 97 12 13 South Asia 566 46 27 18 1,375 85 36 40 Open in new tab Table 3 Quota Rents for Textiles and Clothing . Textiles (TEX) . Clothing (WAP) . . . . Quota premium (%) . . . Quota premium (%) . . Value of quota rent ($mn) . % exports constrained . EU . USA . Value of quota rent ($mn) . % exports constrained . EU . USA . Korea 119 16 10 10 555 55 19 23 Indonesia 97 24 17 12 512 52 48 47 Malaysia 65 100 12 10 330 100 32 37 Philippines 7 50 10 9 363 81 28 34 Singapore 7 11 10 8 365 100 28 31 Thailand 53 40 13 9 396 42 36 35 China 378 19 27 18 2,223 31 36 40 Hong Kong 48 13 8 8 1,249 100 16 18 Taiwan 95 13 12 8 515 81 22 19 Brazil 65 100 14 9 43 77 18 20 Mexico 41 60 14 9 181 99 18 20 Latin America 46 45 14 9 619 86 18 20 Middle East and North Africa 84 78 7 5 390 97 9 10 Eastern Europe and FSU 87 78 9 6 430 97 12 13 South Asia 566 46 27 18 1,375 85 36 40 . Textiles (TEX) . Clothing (WAP) . . . . Quota premium (%) . . . Quota premium (%) . . Value of quota rent ($mn) . % exports constrained . EU . USA . Value of quota rent ($mn) . % exports constrained . EU . USA . Korea 119 16 10 10 555 55 19 23 Indonesia 97 24 17 12 512 52 48 47 Malaysia 65 100 12 10 330 100 32 37 Philippines 7 50 10 9 363 81 28 34 Singapore 7 11 10 8 365 100 28 31 Thailand 53 40 13 9 396 42 36 35 China 378 19 27 18 2,223 31 36 40 Hong Kong 48 13 8 8 1,249 100 16 18 Taiwan 95 13 12 8 515 81 22 19 Brazil 65 100 14 9 43 77 18 20 Mexico 41 60 14 9 181 99 18 20 Latin America 46 45 14 9 619 86 18 20 Middle East and North Africa 84 78 7 5 390 97 9 10 Eastern Europe and FSU 87 78 9 6 430 97 12 13 South Asia 566 46 27 18 1,375 85 36 40 Open in new tab In order to illustrate the channels through which the MFA works, in Table 4 we present results with the base model in which we vary key elasticity parameters The most important elasticities are the trade elasticities, σDM and σMM, which we discussed in Section I. In column MFA we reproduce our base model specifications, where we choose σDM = 4 and σMM = 8. Table 4 Decomposing the Welfare Effects of MFA Reform ($bn) Region . MFA . MFA1 . MFA2 . MFA3 . MFA4 . MFA5 . MFA6 . Australia 0.1 0.0 0.0 0.0 0.0 0.0 0.1 New Zealand 0.0 −0.0 0.0 −0.0 0.0 −0.0 −0.0 Canada 0.9 0.7 1.0 1.3 1.0 0.7 1.3 USA 10.1 8.3 10.3 14.7 10.5 8.5 15.4 Japan −0.5 −0.2 −0.6 −0.2 −0.6 −0.2 −0.2 Korea −0.5 −0.6 −0.5 0.6 −0.4 −0.5 0.7 European Union 7.6 6.7 8.0 7.1 7.9 6.9 7.4 Indonesia 0.6 −0.2 1.0 0.0 0.7 −0.2 0.1 Malaysia 0.1 −0.3 0.3 −0.3 0.1 −0.2 −0.3 Philippines −0.0 −0.2 0.0 0.1 0.1 −0.2 0.2 Singapore −0.1 −0.3 −0.1 −0.3 −0.1 −0.3 −0.3 Thailand 0.1 −0.3 0.1 0.0 0.1 −0.2 0.1 China 0.9 −1.2 1.4 0.5 1.3 −1.0 1.0 Hong Kong −1.7 −1.4 −1.8 −0.7 −1.7 −1.4 −0.7 Taiwan −0.5 −0.6 −0.5 0.2 −0.4 −0.5 0.3 Argentina 0.0 0.0 0.0 0.1 0.0 0.0 0.1 Brazil −0.0 0.1 −0.0 0.1 −0.0 −0.1 0.1 Mexico −0.1 −0.2 −0.1 0.4 −0.0 −0.1 0.5 Latin America −0.5 −0.6 −0.6 0.5 −0.4 −0.5 0.6 Sub‐Saharan Africa −0.1 −0.1 −0.1 0.3 −0.1 −0.1 0.2 Middle East and North Africa −0.5 −0.5 −0.6 1.4 −0.4 −0.4 1.5 Eastern Europe & FSU −0.6 −0.6 −0.7 0.1 −0.6 −0.5 0.2 South Asia 0.6 −0.9 1.0 0.1 0.9 −0.7 0.4 Other European 0.1 0.3 −0.1 2.5 0.1 0.3 2.6 Developing Countries (Total) −2.3 −7.8 −1.3 2.9 −0.9 −7.0 4.8 Industrialised Countries (Total) 18.3 16.0 18.7 25.5 18.9 16.2 26.7 World 16.0 8.0 17.4 28.4 18.0 9.2 31.5 Region . MFA . MFA1 . MFA2 . MFA3 . MFA4 . MFA5 . MFA6 . Australia 0.1 0.0 0.0 0.0 0.0 0.0 0.1 New Zealand 0.0 −0.0 0.0 −0.0 0.0 −0.0 −0.0 Canada 0.9 0.7 1.0 1.3 1.0 0.7 1.3 USA 10.1 8.3 10.3 14.7 10.5 8.5 15.4 Japan −0.5 −0.2 −0.6 −0.2 −0.6 −0.2 −0.2 Korea −0.5 −0.6 −0.5 0.6 −0.4 −0.5 0.7 European Union 7.6 6.7 8.0 7.1 7.9 6.9 7.4 Indonesia 0.6 −0.2 1.0 0.0 0.7 −0.2 0.1 Malaysia 0.1 −0.3 0.3 −0.3 0.1 −0.2 −0.3 Philippines −0.0 −0.2 0.0 0.1 0.1 −0.2 0.2 Singapore −0.1 −0.3 −0.1 −0.3 −0.1 −0.3 −0.3 Thailand 0.1 −0.3 0.1 0.0 0.1 −0.2 0.1 China 0.9 −1.2 1.4 0.5 1.3 −1.0 1.0 Hong Kong −1.7 −1.4 −1.8 −0.7 −1.7 −1.4 −0.7 Taiwan −0.5 −0.6 −0.5 0.2 −0.4 −0.5 0.3 Argentina 0.0 0.0 0.0 0.1 0.0 0.0 0.1 Brazil −0.0 0.1 −0.0 0.1 −0.0 −0.1 0.1 Mexico −0.1 −0.2 −0.1 0.4 −0.0 −0.1 0.5 Latin America −0.5 −0.6 −0.6 0.5 −0.4 −0.5 0.6 Sub‐Saharan Africa −0.1 −0.1 −0.1 0.3 −0.1 −0.1 0.2 Middle East and North Africa −0.5 −0.5 −0.6 1.4 −0.4 −0.4 1.5 Eastern Europe & FSU −0.6 −0.6 −0.7 0.1 −0.6 −0.5 0.2 South Asia 0.6 −0.9 1.0 0.1 0.9 −0.7 0.4 Other European 0.1 0.3 −0.1 2.5 0.1 0.3 2.6 Developing Countries (Total) −2.3 −7.8 −1.3 2.9 −0.9 −7.0 4.8 Industrialised Countries (Total) 18.3 16.0 18.7 25.5 18.9 16.2 26.7 World 16.0 8.0 17.4 28.4 18.0 9.2 31.5 Removal of the MFA in the CRTS Base Model with elasticities as follows: MFA, σMM = 8 and σDM = 4 for MFA sectors; MFA1, σMM = 4 and σDM = 2 for MFA sectors; MFA2, σMM = 10 and σDM = 4 for MFA sectors; MFA3, σMM = 0 and σDM = 15 for MFA sectors; MFA4, σMM = 8 and σDM = 4 and σDC = 2 and σDG = 2 for MFA sectors; MFA5, σMM = 4 and σDM = 2 and σDC = 2 and σDG = 2 for MFA sectors; MFA6, σMM = 0 and σDM = 15 and σDC = 2 and σDG = 2 for MFA sectors. Open in new tab Table 4 Decomposing the Welfare Effects of MFA Reform ($bn) Region . MFA . MFA1 . MFA2 . MFA3 . MFA4 . MFA5 . MFA6 . Australia 0.1 0.0 0.0 0.0 0.0 0.0 0.1 New Zealand 0.0 −0.0 0.0 −0.0 0.0 −0.0 −0.0 Canada 0.9 0.7 1.0 1.3 1.0 0.7 1.3 USA 10.1 8.3 10.3 14.7 10.5 8.5 15.4 Japan −0.5 −0.2 −0.6 −0.2 −0.6 −0.2 −0.2 Korea −0.5 −0.6 −0.5 0.6 −0.4 −0.5 0.7 European Union 7.6 6.7 8.0 7.1 7.9 6.9 7.4 Indonesia 0.6 −0.2 1.0 0.0 0.7 −0.2 0.1 Malaysia 0.1 −0.3 0.3 −0.3 0.1 −0.2 −0.3 Philippines −0.0 −0.2 0.0 0.1 0.1 −0.2 0.2 Singapore −0.1 −0.3 −0.1 −0.3 −0.1 −0.3 −0.3 Thailand 0.1 −0.3 0.1 0.0 0.1 −0.2 0.1 China 0.9 −1.2 1.4 0.5 1.3 −1.0 1.0 Hong Kong −1.7 −1.4 −1.8 −0.7 −1.7 −1.4 −0.7 Taiwan −0.5 −0.6 −0.5 0.2 −0.4 −0.5 0.3 Argentina 0.0 0.0 0.0 0.1 0.0 0.0 0.1 Brazil −0.0 0.1 −0.0 0.1 −0.0 −0.1 0.1 Mexico −0.1 −0.2 −0.1 0.4 −0.0 −0.1 0.5 Latin America −0.5 −0.6 −0.6 0.5 −0.4 −0.5 0.6 Sub‐Saharan Africa −0.1 −0.1 −0.1 0.3 −0.1 −0.1 0.2 Middle East and North Africa −0.5 −0.5 −0.6 1.4 −0.4 −0.4 1.5 Eastern Europe & FSU −0.6 −0.6 −0.7 0.1 −0.6 −0.5 0.2 South Asia 0.6 −0.9 1.0 0.1 0.9 −0.7 0.4 Other European 0.1 0.3 −0.1 2.5 0.1 0.3 2.6 Developing Countries (Total) −2.3 −7.8 −1.3 2.9 −0.9 −7.0 4.8 Industrialised Countries (Total) 18.3 16.0 18.7 25.5 18.9 16.2 26.7 World 16.0 8.0 17.4 28.4 18.0 9.2 31.5 Region . MFA . MFA1 . MFA2 . MFA3 . MFA4 . MFA5 . MFA6 . Australia 0.1 0.0 0.0 0.0 0.0 0.0 0.1 New Zealand 0.0 −0.0 0.0 −0.0 0.0 −0.0 −0.0 Canada 0.9 0.7 1.0 1.3 1.0 0.7 1.3 USA 10.1 8.3 10.3 14.7 10.5 8.5 15.4 Japan −0.5 −0.2 −0.6 −0.2 −0.6 −0.2 −0.2 Korea −0.5 −0.6 −0.5 0.6 −0.4 −0.5 0.7 European Union 7.6 6.7 8.0 7.1 7.9 6.9 7.4 Indonesia 0.6 −0.2 1.0 0.0 0.7 −0.2 0.1 Malaysia 0.1 −0.3 0.3 −0.3 0.1 −0.2 −0.3 Philippines −0.0 −0.2 0.0 0.1 0.1 −0.2 0.2 Singapore −0.1 −0.3 −0.1 −0.3 −0.1 −0.3 −0.3 Thailand 0.1 −0.3 0.1 0.0 0.1 −0.2 0.1 China 0.9 −1.2 1.4 0.5 1.3 −1.0 1.0 Hong Kong −1.7 −1.4 −1.8 −0.7 −1.7 −1.4 −0.7 Taiwan −0.5 −0.6 −0.5 0.2 −0.4 −0.5 0.3 Argentina 0.0 0.0 0.0 0.1 0.0 0.0 0.1 Brazil −0.0 0.1 −0.0 0.1 −0.0 −0.1 0.1 Mexico −0.1 −0.2 −0.1 0.4 −0.0 −0.1 0.5 Latin America −0.5 −0.6 −0.6 0.5 −0.4 −0.5 0.6 Sub‐Saharan Africa −0.1 −0.1 −0.1 0.3 −0.1 −0.1 0.2 Middle East and North Africa −0.5 −0.5 −0.6 1.4 −0.4 −0.4 1.5 Eastern Europe & FSU −0.6 −0.6 −0.7 0.1 −0.6 −0.5 0.2 South Asia 0.6 −0.9 1.0 0.1 0.9 −0.7 0.4 Other European 0.1 0.3 −0.1 2.5 0.1 0.3 2.6 Developing Countries (Total) −2.3 −7.8 −1.3 2.9 −0.9 −7.0 4.8 Industrialised Countries (Total) 18.3 16.0 18.7 25.5 18.9 16.2 26.7 World 16.0 8.0 17.4 28.4 18.0 9.2 31.5 Removal of the MFA in the CRTS Base Model with elasticities as follows: MFA, σMM = 8 and σDM = 4 for MFA sectors; MFA1, σMM = 4 and σDM = 2 for MFA sectors; MFA2, σMM = 10 and σDM = 4 for MFA sectors; MFA3, σMM = 0 and σDM = 15 for MFA sectors; MFA4, σMM = 8 and σDM = 4 and σDC = 2 and σDG = 2 for MFA sectors; MFA5, σMM = 4 and σDM = 2 and σDC = 2 and σDG = 2 for MFA sectors; MFA6, σMM = 0 and σDM = 15 and σDC = 2 and σDG = 2 for MFA sectors. Open in new tab In column MFA1 of Table 4 we lower the values of these trade elasticities to 2 and 4, respectively, and characterise these results as short run. Losses to developing countries as a whole increase substantially in this short run scenario, from a loss of $2.3 billion to a loss of $7.8 billion. This is because the developed countries import considerably less with high elasticities of demand, and thereby allow less resource reallocation in developing countries and less terms‐of‐trade gains for developing countries in unconstrained markets. In column MFA2 of Table 4 we show the impact of greater substitutability among imports of textiles and clothing from different regions by increasing σMM to 10 while keeping σDM at our base value of 4. Aggregate gains to the world increase since demand elasticities increase, but China, South Asia, Indonesia, Thailand, and Malaysia become the dominant gainers among developing countries. This pattern arises because our data indicates that these countries or regions are the most efficient suppliers (they have the highest quota premia). They expand exports, at the expense of marginally inefficient developing countries, when importing nations more readily substitute among alternative foreign suppliers. This is the scenario feared by the marginally inefficient suppliers who have been able to maintain sales and obtain rents only due to the quota. Hence we also find some relatively inefficient MFA exporters losing even more in MFA2 in comparison to MFA: for example, rest of Latin America, Middle East and North Africa, and Eastern Europe and the former Soviet Union. What assumptions regarding trade elasticities would be most likely to lead to the marginally inefficient exporters gaining from reform of the MFA? The logic of our analysis and base model results would suggest that one would need to have a large demand increase in developed countries due to lower prices (hence set σDM equal to 15). One would also need to ensure that substitution away from marginally inefficient suppliers is restricted (hence set σMM equal to 0). Column MFA3 in Table 4 reports the results of such a specification. As expected from a high value of σDM, there are considerably more gains to the world, and developing countries as a group now gain. Moreover, as expected, the two scenarios with high σDM and low σMM (MFA3 and MFA6) are the only scenarios where most of the marginally inefficient suppliers gain, i.e., Sub‐Saharan Africa, Eastern Europe and the former Soviet Union, Latin America, Middle East and North Africa, Taiwan, Mexico, Brazil and Korea all gain. An interesting result is that, despite these possibly ‘optimistic’ trade elasticity assumptions, some suppliers (Malaysia, Singapore and Hong Kong) still register losses in welfare. The reason is seen from Table 3: these are the suppliers who have all of their benchmark clothing sales in quota‐constrained markets. Thus, in scenario MFA3, with σMM = 0, they cannot share in the terms‐of‐trade improvement realised by exporters to previously unconstrained markets. Thus Malaysia, which is one of the most efficient suppliers based on the quota premium on its wearing apparel sales, gains in our scenarios if and only if σMM is high. Singapore and Hong Kong, however, are marginally inefficient suppliers, with no sales in unconstrained markets, and thus lose in all scenarios. In our model the representative consumer agent and the government agent in each region consume 22 goods (at the top level of Fig. 1). Unless otherwise stated, the elasticities of substitution between these 22 aggregate goods are unity, i.e. σDC = σDG = 1 in each region. For higher elasticity values, consumers will shift purchases toward goods whose relative prices decrease. In particular, consumers in import quota‐constrained countries will shift income toward textiles and apparel purchases in the MFA scenarios. In scenarios MFA4 and MFA5 we increase the top‐level elasticity of substitution to 2, otherwise holding elasticities at the values assumed in scenarios MFA and MFA1, respectively. Since the relative prices of textiles and clothing are decreasing in the constrained markets, the gains to both developing countries and the industrialised countries increase, but the magnitude is not overwhelming. The reason is that households consume a composite of imported and domestic clothing, and in the case of the United States for example, composite clothing prices decrease only by 7.6%. Of course, higher values of σDC and σDG will magnify the impact observed in MFA4 and MFA5, but there is a question of plausibility of values much higher than 3. Scenario MFA6 combines the logic of our analysis of the effects of trade elasticities (in MFA1 to MFA3) and the effects of final consumption elasticities (in MFA4 and MFA5). It represents the most optimistic configuration of elasticities for developing countries with respect to MFA reforms generating beneficial welfare reforms. Even so, Malaysia, Singapore and Hong Kong still lose due to the low value of σMM and the absence of initial sales to unconstrained markets as discussed above. The results in Table 4 demonstrate that predicting which exporting countries will gain from MFA removal is complex. The mid‐level elasticity σDM is central to the overall welfare gains possible from MFA reform, while the bottom‐level elasticity σMM is central to how costs of MFA reform are distributed among developing countries. The results for Malaysia show that an efficient supplier can overcome the handicap of a small initial share in previously unconstrained markets and gain from the removal of the MFA, provided the substitution possibilities are sufficiently great.12 This possibility had been overlooked in previous theoretical analyses. These analyses suggested that it would only be countries such as Indonesia, which has substantial initial sales of wearing apparel in unconstrained markets that would gain (due to rectangles of terms‐of‐trade gains). The initial share in constrained markets is important, but when substitution possibilities among exporters are high, the most efficient suppliers capture these markets. In sum, import quota‐constrained countries gain from removal of the MFA due to quota rent gains as well as efficiency gains, and unconstrained importers lose due to a terms‐of‐trade loss. The most efficient suppliers among the developing countries also typically gain, but the gains vary substantially, increasing in the long run when elasticities are expected to be higher. Many of the marginally inefficient suppliers among the developing countries lose, because they lose quota rents and in the long run lose market share to more efficient suppliers among developing countries. II.4. Regional Aggregation We can simulate our CRTS base model after aggregating to 12 regions (see Table 1 for the aggregation). The principal result is that the gains from the UR are reduced by about 5–10%, because aggregation reduces the dispersion of distortions. Since resource misallocation costs increase roughly with the square of the distortion, a model with decrease dispersion will produce estimates of the welfare impacts from the protection that are based downward. More importantly, at the individual country level the results are sometimes magnified. For example, China gains much more from elimination of the MFA in the 12 sector model. The reason is that Malaysia and Indonesia, which are the most effective competitors in the region, are aggregated as part of ‘East Asia’, thereby producing a region of only average efficiency. Thus, we regard the greater level of geographic detail in our model to be an important extension relative to the other aggregative models of the UR.13 11.5. The Impact on Sub‐Saharan Africa As the poorest region in the world, the impact on Sub‐Saharan Africa is especially important. The trade negotiators for Africa were quite successful in the trade negotiation game, in the sense that they did not incur any obligations under the Round to reduce their own considerable distortions. Thus, the principal impact on Sub‐Saharan Africa from the UR is the adverse terms‐of‐trade impact from the increase in its import bill in agriculture.14 There is also a slight adverse terms‐of‐trade impact from elimination of the MFA, as prices in export markets decline under increased competition. Africa is not assumed to be collecting quota rents from the MFA, so it does not suffer greatly from MFA elimination. Moreover, we do not find that Africa gains in the long run either from the UR.15 These results show the costs to a small country when its trade negotiators are successful at playing their game of obtaining maximum access and giving up the least. What can Africa do to improve its welfare in the post‐UR environment? Our data show that the most important distortions in Africa are the high explicit and implicit domestic taxes of selected agricultural sectors. United States Department of Agriculture estimates indicate that agricultural sectors are typically taxed at very high rates in Africa. In our data set, rice is taxed at a rate of 71%, whereas the maximum for any other sector is 11%. Thus, due to the large dispersion in the domestic taxes, too little resources are devoted to rice production. When we reduce taxes of rice, so that the maximum is 11% (no higher than any other sector in the economy) we find a very large increase in African welfare (9.7% of GDP). We caution, however, that these results are dependent on estimates of domestic distortions in agriculture, which vary considerably over time and across countries and products. Given our estimates of domestic distortions in agriculture, we find that simple across the board cuts in border taxes actually result in very small welfare losses (0.2% of GDP for a 25% cut in tariffs). The reason is that the tariff on rice is also high (70%); the tariff cuts reduce consumption deadweight losses but induce an exit of resources from rice production. This increases the production distortions which are slightly dominant. We find that the optimal strategy for Africa is to both reduce its domestic taxes in sectors where they are disproportionately high (no higher than 11% in any sector) and also reduce its border taxes by 50%. Given a simultaneous reduction in domestic distortions (which generates large gains as mentioned above), the 50% reduction in its border taxes will result in additional welfare gains that are slightly larger (0.23%) than the loss Africa suffers under the UR. III. Increasing Returns to Scale and Steady‐State Effects III.1. Imperfect Competition and Increasing Returns Due to evidence that many sectors experience IRTS at the plant or firm level, we modify our base model to incorporate IRTS. Our imperfectly competitive model is applied only to those sectors which are subject to IRTS, as explained below. Thus, we continue to model some sectors using CRTS and competitive markets. For sectors which are subject to IRTS and imperfect competition, on the demand side of the model we assume that consumers have preferences for the products of firms (i.e. firm‐level product differentiation) and we allow for the possibility that consumers have preferences for the products of their home firms. This allows us to accommodate situations where national producers produce goods that conform to tastes of the citizens of their countries. For example, Japanese consumers would have a preference for autos with the steering wheel on the right hand side and hence, other things equal, prefer Japanese made vehicles to American vehicles which do not possess this property. More importantly, Japanese vehicles would be better substitutes for each other than American vehicles would be for Japanese vehicles. We implement this preference structure beginning with our nested CES (Armington) structure discussed above, but add an additional nest for the firms in all countries at the bottom level. Our structure is depicted in Fig. 1, where each of the import sources represents the composite of varieties from the firms in that country. There has been much confusion in the CGE literature concerning the implications of this type of specification. One point that has occasionally been overlooked is that when elasticities at all levels of a nested CES structure are equal, the structure collapses to a single level CES. This property of CES functions implies that a special case of our model is firm‐level product differentiation with elasticities of substitution equal for all firms and products in the model. The advantage of our general structure, which builds on the CRTS structure, is that it allows a pure comparison of the results of a CRTS model to the IRTS model. That is, we do not need to change the assumed market demand curves of consumers in order to implement the IRTS model. Moreover, if there are different elasticities of substitution between home and imported varieties of firm products, we can incorporate that in our structure, whereas a single level CES function cannot. In other words, we can accommodate a more general Slutsky substitution matrix with preferences defined over varieties. On the supply side we assume the standard IRTS implementation that firms produce with constant marginal costs and a given fixed cost. If the same industry output is produced with fewer firms, there is a rationalisation gain as firms slide down their average cost curve, producing more output with the same fixed costs. This approach requires that we obtain estimates of the extent of unrealised economies of scale. These are derived from econometric estimates of the cost disadvantage ratio (CDR), primarily from Pratten (1987) and Neven (1990).16 We assume that firms are symmetric and determine price and quantities in an oligopoly model in which there is entry and exit, and hence zero profit in equilibrium. Firms compete in a quantity adjusting oligopoly framework where the quantity conjectures are calibrated to be identical for all firms in each sector and each country, and do not change in the counterfactual. In each sector, define mrr′ as the Lerner markup (price minus marginal costs divided by price) of firms in region r selling in region r′ Harrison, Rutherford and Tarr (1995, Appendix C) show that our formulation yields: (2) where rr′ is the market share of firms from region r in country r′ is the market share of imports in region r, and Ωrr′ is the conjectural variation for firms from region r selling into region r′. When tariffs are lowered, domestic firms lose market share but gain market share on export markets. The top part of the above equation refers to the home market. Since σDM exceeds unity, with a lower domestic share rr′ firms lower their markup on domestic sales. But since firms are subject to a zero profit constraint, lower markups induce losses and firm exit. Firm exit allows remaining firms to increase their output and achieve rationalisation gains by sliding down their average cost curves. Exit continues until rationalisation gains are sufficient to restore zero profit with the lower markups. Elasticities affect markups both directly and indirectly. There is a direct effect on price. With a larger direct effect on price from higher elasticities, the market share effect will be larger. Then the indirect effect through the markup equation will be larger. From equation (1) a key point is evident: the induced reduction in firms and rationalisation benefits is derived from changes in the shares, which typically are not large. Our static IRTS results are shown in the columns labelled ‘static’ in Table 5. The most striking feature of these results is how similar they are to the CRTS results. Global welfare increases from $93 billion in the CRTS version to just $96 billion in the IRTS version, contrary to the folklore in the CGE literature17 that IRTS always generates much larger welfare gains due to rationalisation gains. Table 5 Welfare Effects of the Uruguay Round under Preferred IRTS Assumptions: Static and Steady State Results ($bn 1992) . Agricultural reform . MFA reform . Manufacturing sector reform . Complete Uruguay Round . Complete reform as a percentage of GDP . . Static . Steady state . Static . Steady state . Static . Steady state . Static . Steady state . Static . Steady state . Australia 0.7 0.9 0.0 0.1 0.5 2.3 1.2 3.3 0.5 1.3 New Zealand 0.3 0.5 0.0 0.0 0.1 0.9 0.4 1.4 1.2 4.4 Canada 0.3 0.2 0.9 1.0 0.1 1.3 1.3 2.6 0.3 0.6 United States 1.8 3.2 10.0 9.2 1.2 13.7 13.3 26.7 0.3 0.6 Japan 15.1 16.8 −0.6 −0.5 2.2 6.2 16.9 22.7 0.5 0.7 Korea 4.6 5.2 −0.5 −0.4 0.7 2.7 4.8 0.5 1.8 2.8 European Union (12) 28.3 26.4 7.6 7.8 3.0 14.9 39.3 0.9 0.7 0.9 Indonesia 0.2 0.3 0.6 0.9 0.6 1.4 1.3 2.6 1.2 2.4 Malaysia 1.2 2.2 0.1 0.3 0.7 2.6 1.8 5.0 3.7 10.2 Philippines 0.7 1.1 0.0 0.2 0.4 1.1 0.9 2.4 1.9 4.8 Singapore 0.6 0.5 −0.2 −0.2 0.5 0.4 0.9 0.7 2.3 1.9 Thailand 0.8 1.4 0.1 0.8 1.8 10.3 2.5 12.6 2.4 12.1 China −0.5 −0.8 1.0 1.7 0.9 1.2 1.3 2.0 0.3 0.5 Hong Kong 0.6 0.6 −1.7 −1.5 −0.1 −0.2 −1.2 −1.1 −1.4 −1.3 Taiwan Province of China 0.0 0.0 −0.4 −0.3 0.8 1.3 0.4 1.1 0.3 0.7 Argentina 0.4 0.7 0.0 0.1 0.3 1.6 0.7 2.3 0.3 1.1 Brazil 0.3 0.1 −0.0 0.1 1.2 4.0 1.4 4.3 0.4 1.3 Mexico −0.0 07 −0.1 0.2 0.3 1.4 0.2 2.3 0.0 0.7 Latin America 1.5 2.0 −0.5 0.3 0.3 3.2 1.3 4.7 0.5 1.9 Sub‐Saharan Africa −0.2 −0.5 −0.0 −0.1 0.1 0.2 −0.3 −0.7 −0.2 −0.5 Middle East and North Africa −0.3 0.1 −0.4 0.2 0.8 1.9 −0.3 1.5 −0.1 0.3 Eastern Europe and FSU −0.1 −0.0 −0.5 −0.3 0.8 2.3 −0.2 1.2 −0.0 0.2 South Asia 0.3 0.2 0.9 1.9 3.1 5.3 3.7 6.7 1.3 2.3 Other European countries 2.2 1.6 −0.2 −0.8 1.7 7.0 4.2 8.8 0.4 0.9 Developing countries (total) 9.9 13.9 −1.5 3.4 12.9 40.5 19.4 55.2 0.5 1.4 Industrial countries (total) 48.7 49.8 17.9 16.9 8.7 46.3 76.7 115.4 0.5 0.8 . Agricultural reform . MFA reform . Manufacturing sector reform . Complete Uruguay Round . Complete reform as a percentage of GDP . . Static . Steady state . Static . Steady state . Static . Steady state . Static . Steady state . Static . Steady state . Australia 0.7 0.9 0.0 0.1 0.5 2.3 1.2 3.3 0.5 1.3 New Zealand 0.3 0.5 0.0 0.0 0.1 0.9 0.4 1.4 1.2 4.4 Canada 0.3 0.2 0.9 1.0 0.1 1.3 1.3 2.6 0.3 0.6 United States 1.8 3.2 10.0 9.2 1.2 13.7 13.3 26.7 0.3 0.6 Japan 15.1 16.8 −0.6 −0.5 2.2 6.2 16.9 22.7 0.5 0.7 Korea 4.6 5.2 −0.5 −0.4 0.7 2.7 4.8 0.5 1.8 2.8 European Union (12) 28.3 26.4 7.6 7.8 3.0 14.9 39.3 0.9 0.7 0.9 Indonesia 0.2 0.3 0.6 0.9 0.6 1.4 1.3 2.6 1.2 2.4 Malaysia 1.2 2.2 0.1 0.3 0.7 2.6 1.8 5.0 3.7 10.2 Philippines 0.7 1.1 0.0 0.2 0.4 1.1 0.9 2.4 1.9 4.8 Singapore 0.6 0.5 −0.2 −0.2 0.5 0.4 0.9 0.7 2.3 1.9 Thailand 0.8 1.4 0.1 0.8 1.8 10.3 2.5 12.6 2.4 12.1 China −0.5 −0.8 1.0 1.7 0.9 1.2 1.3 2.0 0.3 0.5 Hong Kong 0.6 0.6 −1.7 −1.5 −0.1 −0.2 −1.2 −1.1 −1.4 −1.3 Taiwan Province of China 0.0 0.0 −0.4 −0.3 0.8 1.3 0.4 1.1 0.3 0.7 Argentina 0.4 0.7 0.0 0.1 0.3 1.6 0.7 2.3 0.3 1.1 Brazil 0.3 0.1 −0.0 0.1 1.2 4.0 1.4 4.3 0.4 1.3 Mexico −0.0 07 −0.1 0.2 0.3 1.4 0.2 2.3 0.0 0.7 Latin America 1.5 2.0 −0.5 0.3 0.3 3.2 1.3 4.7 0.5 1.9 Sub‐Saharan Africa −0.2 −0.5 −0.0 −0.1 0.1 0.2 −0.3 −0.7 −0.2 −0.5 Middle East and North Africa −0.3 0.1 −0.4 0.2 0.8 1.9 −0.3 1.5 −0.1 0.3 Eastern Europe and FSU −0.1 −0.0 −0.5 −0.3 0.8 2.3 −0.2 1.2 −0.0 0.2 South Asia 0.3 0.2 0.9 1.9 3.1 5.3 3.7 6.7 1.3 2.3 Other European countries 2.2 1.6 −0.2 −0.8 1.7 7.0 4.2 8.8 0.4 0.9 Developing countries (total) 9.9 13.9 −1.5 3.4 12.9 40.5 19.4 55.2 0.5 1.4 Industrial countries (total) 48.7 49.8 17.9 16.9 8.7 46.3 76.7 115.4 0.5 0.8 Open in new tab Table 5 Welfare Effects of the Uruguay Round under Preferred IRTS Assumptions: Static and Steady State Results ($bn 1992) . Agricultural reform . MFA reform . Manufacturing sector reform . Complete Uruguay Round . Complete reform as a percentage of GDP . . Static . Steady state . Static . Steady state . Static . Steady state . Static . Steady state . Static . Steady state . Australia 0.7 0.9 0.0 0.1 0.5 2.3 1.2 3.3 0.5 1.3 New Zealand 0.3 0.5 0.0 0.0 0.1 0.9 0.4 1.4 1.2 4.4 Canada 0.3 0.2 0.9 1.0 0.1 1.3 1.3 2.6 0.3 0.6 United States 1.8 3.2 10.0 9.2 1.2 13.7 13.3 26.7 0.3 0.6 Japan 15.1 16.8 −0.6 −0.5 2.2 6.2 16.9 22.7 0.5 0.7 Korea 4.6 5.2 −0.5 −0.4 0.7 2.7 4.8 0.5 1.8 2.8 European Union (12) 28.3 26.4 7.6 7.8 3.0 14.9 39.3 0.9 0.7 0.9 Indonesia 0.2 0.3 0.6 0.9 0.6 1.4 1.3 2.6 1.2 2.4 Malaysia 1.2 2.2 0.1 0.3 0.7 2.6 1.8 5.0 3.7 10.2 Philippines 0.7 1.1 0.0 0.2 0.4 1.1 0.9 2.4 1.9 4.8 Singapore 0.6 0.5 −0.2 −0.2 0.5 0.4 0.9 0.7 2.3 1.9 Thailand 0.8 1.4 0.1 0.8 1.8 10.3 2.5 12.6 2.4 12.1 China −0.5 −0.8 1.0 1.7 0.9 1.2 1.3 2.0 0.3 0.5 Hong Kong 0.6 0.6 −1.7 −1.5 −0.1 −0.2 −1.2 −1.1 −1.4 −1.3 Taiwan Province of China 0.0 0.0 −0.4 −0.3 0.8 1.3 0.4 1.1 0.3 0.7 Argentina 0.4 0.7 0.0 0.1 0.3 1.6 0.7 2.3 0.3 1.1 Brazil 0.3 0.1 −0.0 0.1 1.2 4.0 1.4 4.3 0.4 1.3 Mexico −0.0 07 −0.1 0.2 0.3 1.4 0.2 2.3 0.0 0.7 Latin America 1.5 2.0 −0.5 0.3 0.3 3.2 1.3 4.7 0.5 1.9 Sub‐Saharan Africa −0.2 −0.5 −0.0 −0.1 0.1 0.2 −0.3 −0.7 −0.2 −0.5 Middle East and North Africa −0.3 0.1 −0.4 0.2 0.8 1.9 −0.3 1.5 −0.1 0.3 Eastern Europe and FSU −0.1 −0.0 −0.5 −0.3 0.8 2.3 −0.2 1.2 −0.0 0.2 South Asia 0.3 0.2 0.9 1.9 3.1 5.3 3.7 6.7 1.3 2.3 Other European countries 2.2 1.6 −0.2 −0.8 1.7 7.0 4.2 8.8 0.4 0.9 Developing countries (total) 9.9 13.9 −1.5 3.4 12.9 40.5 19.4 55.2 0.5 1.4 Industrial countries (total) 48.7 49.8 17.9 16.9 8.7 46.3 76.7 115.4 0.5 0.8 . Agricultural reform . MFA reform . Manufacturing sector reform . Complete Uruguay Round . Complete reform as a percentage of GDP . . Static . Steady state . Static . Steady state . Static . Steady state . Static . Steady state . Static . Steady state . Australia 0.7 0.9 0.0 0.1 0.5 2.3 1.2 3.3 0.5 1.3 New Zealand 0.3 0.5 0.0 0.0 0.1 0.9 0.4 1.4 1.2 4.4 Canada 0.3 0.2 0.9 1.0 0.1 1.3 1.3 2.6 0.3 0.6 United States 1.8 3.2 10.0 9.2 1.2 13.7 13.3 26.7 0.3 0.6 Japan 15.1 16.8 −0.6 −0.5 2.2 6.2 16.9 22.7 0.5 0.7 Korea 4.6 5.2 −0.5 −0.4 0.7 2.7 4.8 0.5 1.8 2.8 European Union (12) 28.3 26.4 7.6 7.8 3.0 14.9 39.3 0.9 0.7 0.9 Indonesia 0.2 0.3 0.6 0.9 0.6 1.4 1.3 2.6 1.2 2.4 Malaysia 1.2 2.2 0.1 0.3 0.7 2.6 1.8 5.0 3.7 10.2 Philippines 0.7 1.1 0.0 0.2 0.4 1.1 0.9 2.4 1.9 4.8 Singapore 0.6 0.5 −0.2 −0.2 0.5 0.4 0.9 0.7 2.3 1.9 Thailand 0.8 1.4 0.1 0.8 1.8 10.3 2.5 12.6 2.4 12.1 China −0.5 −0.8 1.0 1.7 0.9 1.2 1.3 2.0 0.3 0.5 Hong Kong 0.6 0.6 −1.7 −1.5 −0.1 −0.2 −1.2 −1.1 −1.4 −1.3 Taiwan Province of China 0.0 0.0 −0.4 −0.3 0.8 1.3 0.4 1.1 0.3 0.7 Argentina 0.4 0.7 0.0 0.1 0.3 1.6 0.7 2.3 0.3 1.1 Brazil 0.3 0.1 −0.0 0.1 1.2 4.0 1.4 4.3 0.4 1.3 Mexico −0.0 07 −0.1 0.2 0.3 1.4 0.2 2.3 0.0 0.7 Latin America 1.5 2.0 −0.5 0.3 0.3 3.2 1.3 4.7 0.5 1.9 Sub‐Saharan Africa −0.2 −0.5 −0.0 −0.1 0.1 0.2 −0.3 −0.7 −0.2 −0.5 Middle East and North Africa −0.3 0.1 −0.4 0.2 0.8 1.9 −0.3 1.5 −0.1 0.3 Eastern Europe and FSU −0.1 −0.0 −0.5 −0.3 0.8 2.3 −0.2 1.2 −0.0 0.2 South Asia 0.3 0.2 0.9 1.9 3.1 5.3 3.7 6.7 1.3 2.3 Other European countries 2.2 1.6 −0.2 −0.8 1.7 7.0 4.2 8.8 0.4 0.9 Developing countries (total) 9.9 13.9 −1.5 3.4 12.9 40.5 19.4 55.2 0.5 1.4 Industrial countries (total) 48.7 49.8 17.9 16.9 8.7 46.3 76.7 115.4 0.5 0.8 Open in new tab There are two reasons for these small enhancements in the benefits. First, we have assumed smaller values for the CDRs than most studies, thus there are smaller rationalisation gains in the benchmark equilibrium to be realised from liberalisation. More importantly, however, we have studiously avoided incorporating a regime switch other than CRTS to IRTS. Many CGE implementations of IRTS change something other than the pure IRTS impact at the same time, such as elasticities, and it is the regime switch that is driving the larger numbers.18 III.2. Dynamic Effects While the dynamic benefits of trade liberalisation and the UR are often described, these benefits are rarely estimated. We employ a steady state approach to evaluation of trade policy change in a multi‐region model which, based on the original work of Hansen and Koopmans (1972) and Dantzig and Manne (1974), we first implemented in Harrison, Rutherford and Tarr (1994; 1996); our approach has also been used by Francois, McDonald and Nordstrom (1994; 1995). Briefly stated, we assume that the capital stock in each country is optimal given the rate of return on capital in the initial equilibrium. That is, increases in the rate of return on capital would induce an increase in investment until the marginal productivity of capital is driven down to the initial rate of return. The UR will produce a new equilibrium, where for almost all countries the rate of return on capital increases (relative to a price index of consumption) due to a more efficient allocation of resources. This implies that in a dynamic sense the new capital stock can no longer be optimal: investment will be forthcoming until the marginal productivity of capital is reduced to the long run equilibrium rate of return on capital. In the static calculation we allow the rental rate of capital to vary within each country, while holding constant the aggregate stock of capital in each country. The steady state calculation essentially reverses this: we allow the capital stock in each country to be endogenously determined while holding constant the rental rate of capital in each country. This expansion of the capital stock then works through our model like an ‘endowment effect’, generating larger welfare gains since there are more resources to be employed. In addition, as the income of the world increases, demand for goods and the derived demand for factors increases. In this model the quantity of capital will increase in response, producing a further endowment effect. Since our steady state calculation ignores the foregone consumption necessary to obtain the larger capital stock, we stress that this calculation measures an upper bound on potential welfare gains in a long run, classical Solow‐type growth model. It could, however, be an underestimate of the long run gains, since it fails to capture endogenous growth effects such as those arising from induced improvements in productivity or innovation (so‐called ‘learning by doing’). The results of applying this steady state variant to our 24 region IRTS model are reported in the columns labelled ‘steady‐state’ in Table 5. The obvious difference is the magnification of global welfare gains from the UR, from $96 billion in the comparable 24 region static case to $171 billion. IV. Comparisons to Other Estimates A number of other studies of the quantitative impact of the UR and its components have been conducted, and can be usefully compared to the results we offer.19 Where significant differences in the estimates exist among these studies, we emphasise that these differences typically may be intuitively explained. Indeed, subject to a statistical margin of error (which we quantify), what is remarkable is our ability to interpret the reasons for the differences in these studies, and their broad consistency in terms of the overall benefits and the relative importance of various components. This does not imply that all estimates are equally valid, but that the reader may transparently determine the reasons for the differences and assess which are the most appropriate. We begin with a discussion of the overall estimates and then examine other estimates of MFA and agricultural reform. IV.1. Overall Evaluations In addition to our model, there are two other general models that employ the actual changes agreed in the UR by the Contracting Parties: the GATT/WTO team of Francois et al. (1995), and Hertel et al. (1995) who used the GTAP model. The principal (updated) estimates of these models are summarised in Table 6. Table 6 Summary of Estimates of the Welfare Effects of the Uruguay Round by Team and Modelling Assumption . Harrison, Rutherford, Tarr . Francois, McDonald Nordstrom . Hertel et al. . . Billion dollars . GDP (%) . Billion dollars . GDP (%) . Billion dollars . GDP (%) . CRTS* – Static 93 0.4 40 0.2 — — IRTS† – Static 96 0.4 99 0.4 — — IRTS – Steady state 171 0.7 193 0.8 — — CRTS – Static 2005 projection 162 0.7 — — 258§ 1.0 IRTS – Steady state 2005 projection — — 510‡ 2.0 — — . Harrison, Rutherford, Tarr . Francois, McDonald Nordstrom . Hertel et al. . . Billion dollars . GDP (%) . Billion dollars . GDP (%) . Billion dollars . GDP (%) . CRTS* – Static 93 0.4 40 0.2 — — IRTS† – Static 96 0.4 99 0.4 — — IRTS – Steady state 171 0.7 193 0.8 — — CRTS – Static 2005 projection 162 0.7 — — 258§ 1.0 IRTS – Steady state 2005 projection — — 510‡ 2.0 — — * Constant returns to scale, perfect competition. † Increasing returns to scale and imperfect competition in selected sectors. ‡ Based on Francois et al. (1994). § Based on 2005 projected GDP. Source: Francois, McDonald and Nordstrom (forthcoming); Hertel et al. (1995); calculations of the present authors. Open in new tab Table 6 Summary of Estimates of the Welfare Effects of the Uruguay Round by Team and Modelling Assumption . Harrison, Rutherford, Tarr . Francois, McDonald Nordstrom . Hertel et al. . . Billion dollars . GDP (%) . Billion dollars . GDP (%) . Billion dollars . GDP (%) . CRTS* – Static 93 0.4 40 0.2 — — IRTS† – Static 96 0.4 99 0.4 — — IRTS – Steady state 171 0.7 193 0.8 — — CRTS – Static 2005 projection 162 0.7 — — 258§ 1.0 IRTS – Steady state 2005 projection — — 510‡ 2.0 — — . Harrison, Rutherford, Tarr . Francois, McDonald Nordstrom . Hertel et al. . . Billion dollars . GDP (%) . Billion dollars . GDP (%) . Billion dollars . GDP (%) . CRTS* – Static 93 0.4 40 0.2 — — IRTS† – Static 96 0.4 99 0.4 — — IRTS – Steady state 171 0.7 193 0.8 — — CRTS – Static 2005 projection 162 0.7 — — 258§ 1.0 IRTS – Steady state 2005 projection — — 510‡ 2.0 — — * Constant returns to scale, perfect competition. † Increasing returns to scale and imperfect competition in selected sectors. ‡ Based on Francois et al. (1994). § Based on 2005 projected GDP. Source: Francois, McDonald and Nordstrom (forthcoming); Hertel et al. (1995); calculations of the present authors. Open in new tab We begin, however, with a discussion of the much publicised earlier estimate of the team of the GATT Secretariat. Francois et al. (1994; table 11 b) find that the steady state impact of the UR on global welfare will be $510 billion per annum in 1990 dollars. This relatively high estimate was obtained by expanding the welfare estimates from their IRTS‐steady state model forward to the year 2005 (when all the UR changes would be implemented) and evaluated the impact of the UR on the larger world economy of 2005.20 Given a much larger world economy, the same percentage cuts in tariffs and export subsidies and the same percentage gains in percentage of GDP yield larger estimated absolute dollar gains. That is, even though the gains from the UR estimated in the GATT study are comparable to our long run estimates in percentage terms, the dollar value is much higher due to the size of the economy on which the tariff and export subsidy cuts are applied. Francois (1995) later estimate the impact of the UR changes based on the economy of 1992, as we do. They then obtain estimated gains of $193 billion in their steady state IRTS model, rather than $510 billion. There is nothing inherently correct about using either 2005 or 1992 as the base year of the model, but when the estimates are in terms of dollars rather than percentage of GDP it is clearly important to keep the year of the estimate in mind. The estimate by the WTO team of $193 billion is comparable to our estimate of $171 billion, since both estimates were obtained in a model that evaluated the UR in a long run steady state model with IRTS. In the static or short run version of their IRTS model, the WTO team obtain $99 billion per year of estimated gains, compared to our estimate of $96 billion. That is, taking into account dynamic or steady state effects roughly doubles the estimated gains from the UR in both models. Based on the systematic sensitivity analysis, reported in Harrison, Rutherford and Tarr (1995), the standard deviation of our estimate is about $3 billion. Thus the estimate of Francois et al. (1995) in the IRTS static case is within one standard deviation of our estimate. Model variants with lower elasticities produce lower estimated gains in the models of the WTO team and in ours. However, given that the UR will be implemented over 10 years, we do not report estimates from low elasticity versions. Francois et al. (1995), however, employ a lower set of demand elasticities in their CRTS static model where they obtain $40 billion. It is primarily the smaller elasticities that produce their smaller estimate of $40 billion. When we employ the Francois et al. (1995) elasticities in our CRTS static model, our CRTS static estimates are also significantly reduced (albeit not as dramatically) to $US 67 billion; moreover, when in our CRTS model we employ the elasticities Francois et al. (1995) used in their IRTS model, the welfare estimates increase to $98 billion. Hertel et al. (1995) estimate the impact of the UR in a CRTS model, where the economy is projected forward to the year 2005 based on projections of the amount of capital, labour and human capital that will be available in the year 2005. Their estimate for the gain in world welfare is $258 billion. Again, the same percentage changes in protection produced a larger dollar gain in welfare based on the year 2005 rather than our estimate of $US 93 billion with our static CRTS model. For the purpose of more precise comparison, we have scaled up the factors of production in our model in same proportions as Hertel et al. (1995), and then implemented the same reduction in distortions that we implement in our other model variants. We obtain estimated welfare gains in this model of $162 billion, which is an increase of 74% over the comparable CRTS static model with a 1992 base. Our estimate is slightly lower than that of Hertel et al. (1995) for two apparent reasons: most importantly, they employ somewhat higher demand elasticities, roughly choosing σDM = 5 and σMM = 10, where we employ values of 4 and 8, respectively; and they project that the MFA would be more binding in 2005 than in 1992, so they employ higher export tax equivalents. IV.2. MFA Reform Other CGE assessments of the consequences of the elimination of the MFA have been undertaken by Yongzheng et al. (1996), Francois et al. (1994; 1995), Hertel et al. (1995) and Trela and Whalley (1990a). Our results are closest to those of Hertel et al. (1995). Yongzheng et al. (1996) and Francois et al. (1994; 1995) report the MFA as a much larger share of the UR. For example, Francois et al. (1994; 1995) report that MFA liberalisation is responsible for about 50% of the gains from the UR in many of their scenarios, whereas we obtain about 17%. The reason is straightforward: our study, as well as Hertel et al. (1995), employed updated estimates of the tariff‐equivalents of MFA quotas from the GTAP data base. The updated values, which are significantly scaled down from the original values, correct an earlier error so that the GTAP database is consistent with its background documentation. The model of Hertel et al. (1995) is a 13 sector aggregation of the GTAP data and produces estimates that the MFA is responsible for about 20% of the gains from the UR, not far from the 17% of our estimates. Their estimates are based on removing quotas in a projected economy of the year 2005. The world economy is assumed to grow faster than the quotas under the MFA until that time, so the quotas become more binding and the MFA increases in relative importance in their model. But the pattern of results across countries, in terms of which developing countries gain or lose, is very similar in their model and ours. The most important finding from the Trela and Whalley (1990a) analysis is that the vast majority of exporting countries under the MFA actually gain from MFA reform, notwithstanding losses in quota rents. In their base model (their Table 3), the only countries to lose in welfare terms are the Dominican Republic, Haiti, Hong Kong, Macau and Singapore; the aggregate welfare gain to exporters is $8.078 billion in 1986, with aggregate losses for the above five countries totalling only $0.569 billion. The key difference is that they treat all imports and domestic products of any particular textile and apparel product as homogeneous. Two way trade is netted out in the initial equilibrium of their model, so only net trade is reported. Then the elimination of the MFA leads to a large decrease in production in industrialised countries. This would be equivalent in our model to taking σDM = σMM to be infinite. In columns MFA3 and MFA6 of Table 4 we have shown that large values for the elasticity of substitution between imports and domestic products (as well as between textiles and apparel and other aggregate products) produce results almost as positive for developing countries as those of Trela and Whalley (1990a). In simulations in our model not reported in Table 4, we have employed still higher values for these elasticities which closely approximates their assumption of homogeneous products, i.e. σDM between 15 and 30 and σMM between 8 and 15. With these high elasticities we find that developing countries gain between $13.4 billion (lower elasticities) and $25.5 billion (higher elasticities) from removal of the MFA, or between 0.34% and 0.64% of their GDP. This demonstrates that these elasticities are key to the large gains reported by Trela and Whalley (1990a). But these elasticities are not supported by econometric evidence or our a priori beliefs. IV.3. Agricultural Reform Although the estimates in Harrison, Rutherford and Tarr (1995) are the first to decompose the consequences of agricultural liberalisation into its three major components, there are three CGE assessments of the total benefits from the reduction of agricultural distortions. All use the same underlying ‘RUNS’ model: Goldin et al. (1993), Brandão and Martin (1993) and Goldin and van der Mensbrugghe (1994). Goldin et al. (1993) and Goldin and van der Mensbrugghe (1994) projected the world economy forward to the year 2002, which as we have seen increases the estimated dollar benefits. Equally importantly, as with all studies that estimated the impact of the UR before its completion, overly optimistic assessments of what the Round would accomplish were employed. In particular, projected 36% cuts in all distortions were employed. To investigate the impact of these optimistic projections, we simulated 36% cuts in import tariffs and export subsidies in our CRTS static model, leaving other changes in distortions the same as in our ‘FULL’ scenario. We obtain gains of $128 billion (up from $93 billion). Thus, we can easily explain the larger aggregate gains in their studies.21 Why does their estimate that the agricultural liberalisation component of the UR constitutes 89% of the total gains, whereas our static CRTS model attributes only 63% of the share of the gains to agricultural reform? To reconcile these differences we implemented a RUNS‐like elasticity structure in the 12 region version of our model. The principal RUNS‐like differences we implemented are: (i) any agricultural good in RUNS is homogeneous with the same good in other countries, i.e. no Armington or CET assumption for these goods; and (ii) the use of a CET in non‐agricultural commodities, between domestic output destined for domestic markets and export markets. After altering elasticities in our 12 region CRTS model and steady state CRTS model to approximate the RUNS formulation, we found that agriculture reform was responsible for 87% of the gains in the static version (close to the RUNS results), although only 60% in the long run. The use of the Armington and CET formulation for the non‐agriculture sectors but the homogeneous assumption for agricultural sectors, results in substantially increasing the relative importance of agriculture.22 The key issue then is the plausibility of the parameter values chosen. V. Conclusion Our evaluation of the UR leads us to estimate the static welfare gains at $96 billion annually in 1992 dollars, with an upper bound estimate in the steady state of $171 billion annually. Although our estimates and others vary with the assumptions that are made with respect to trade elasticities, distortion data assumed, and modelling choices, we have emphasised an approach that should lead to transparency of results. The reader is then able to draw informed judgements as to their credibility for different time horizons. We estimate that the European Union, United States and Japan gain the most from the Round, as these are the regions that reduce distortions the most. Although there are likely to be some losers from the UR, especially in the short run, we are optimistic that unilateral liberalisation of tariffs and production distortions can be implemented to ensure that all regions can gain. In the case of Sub‐Saharan Africa, we have verified this and found that the region could gain enormously from these policies, especially from the reduction of the variance in its domestic distortions in agriculture. Footnotes 1 The elasticities of substitution for these value added production functions are taken from Harrison, Rutherford and Wooton (1991; table 1, p. 101), which are unpublished estimates by Harrison, Jones, Kimbell and Wigle (1993) from time‐series data for the United Stated between 1947 and 1982. Contrary to many of the estimates employed in the CGE literature, the econometric specification used in this case corresponds to the functional form assumed in the model. 2 The available econometric evidence suggests values which are much lower than these (see Reinert and Roland‐Holst (1992) and Shiells and Reinert (1993)). But elasticities would be expected to increase over time, and this model assesses an adjustment of about 10 years, a rather long period in the context of these econometric estimates. 3 Hence we do not capture the marginal efficiency cost of governments having to raise extra revenues through a distortionary domestic tax system. For LDCs these costs could be quite significant, since the revenue losses from trade reform could be sizeable. 4 The GTAP data set had to be rebalanced with the new data on tariffs and subsidies. We impose consistency in the benchmark data through adjustments in the vector of stock change. There have been some revisions to the data of Ingco since the publication of her study, all of which we incorporate. 5 The GATT IDB reflects only MFN tariffs. Although the WB database shows some geographically discriminatory tariffs, the discrimination is only due to different aggregations based on different trade weighting of MFN tariff lines. In particular, GSP preferences are ignored. Thus, the possible losses to LDCs from the erosion of GSP preferences in the UR are ignored in the WB and GATT databases. 6 Notably, although tariffication of non‐tariff barriers may have been protectionist in the short run, as countries choose to bind at a higher level of protection than applied on average (see Ingco (1996)), the longer‐run gains may be substantial as GATT‐style tariff‐cutting formulae can be more easily applied to tariff barriers. Moreover, Martin and Francois (1994) show that even if the bound rate exceeds the average ad valorem‐equivalent protection level, it will probably truncate extremely high equivalent protection levels from occurring. Since those higher rates are likely to generate disproportionately higher welfare losses, the net gains may be substantial even in the short run. In addition, without a successful completion of the UR, it is possible that higher protection would have developed. Perroni and Whalley (1994) evaluate the impact of a trade war if the UR had failed. 7 The agreement on export subsidy reduction also includes a reduction in the volume of exports subject to export subsidies of 20%. It is possible that this will impose an additional constraint on export subsidies so that the impact would be greater than we estimate. 8 We report welfare as a percentage of full GDP, where GDP is derived from the GTAP database. In Harrison, Rutherford and Tarr (1995) welfare changes were reported as a percentage of private consumption, and hence were therefore somewhat higher. 9 The other sector to MFA reform is the TEX sector. The United States experiences a small reduction of 0.8% in TEX prices; this also serves to reduce the cost of living in the United States, but the welfare impact is mitigated by the loss to US producers on exports. The reduction in prices in the EU is 1.9% for TEX and 7.2% for WAP. 10 See Yongzheng et al. (1997) and Tarr (1987) for formal derivations of the welfare analysis. 11 This is likely an overestimate of LDC quota rent losses from MFA elimination to the extent that it ignores rent dissipation costs in LDCs associated with these quota rents. See Hamilton (1986), Trela and Whalley (1990b) and Tarr (1994) for applied discussions of rent dissipation. 12 And, if substitution possibilities are high, a marginally inefficient supplier such as Korea will typically lose even though its share in previously constrained markets is relatively high. 13 Detailed results for the 12 sector version may be found in Harrison, Rutherford and Tarr (1995). 14 Since Africa taxes its agricultural sectors, part of the loss of consumers’ surplus from the increase in the world price is offset by an increase in tax revenue as well as domestic producers’ surplus. Unlike the results of Tyers and Falvey (1989), who showed that an increase in the world price is immiserising for an exporting country which is exporting due to an export subsidy, the loss in consumers’ surplus is dominant. 15 Given the adverse terms of trade effects, the rate of return for capital declines for Africa in the static model; this induces a diminution of investment and the capital stock in Africa results from the UR changes in the steady‐state results, which induces further losses. See Section III for an explanation of the steady state results. 16 Full details of our estimates are provided in Harrison, Rutherford and Tarr (1994; Appendix B). The CDR values we use are: PCR 13%, MEA 10%, FOR 5%, ENR 3%, MIN 8%, FOO 3%, TEX 6%, I_S 5%, TRN 11%, FMP 5%, CRP 4%, NFM 5%, and MAC 6%. We apply these CDR estimates for each region in the specified sector, for lack of any data that convinces us that they vary across regions in any known manner. Other sectors are modelled as CRTS. 17 Exceptions to that folklore include Nguyen and Wigle (1992) and de Melo and Tarr (1992; ch. 7). 18 For example Francois et al. (1994; 1995) increase their elasticities of demand when they introduce IRTS, and it is the impact of the elasticities that is primarily responsible for the larger estimated welfare gains in their IRTS scenarios. We quantify the impact below. Harris (1986) employed a pricing rule for IRTS sectors which was highly sensitive to the tariff‐inclusive price of imports. Harrison, Jones, Kimbell and Wigle (1993) demonstrate, using his model, that the pricing rule drives the big welfare gains. 19 We only discuss in detail models that evaluated the UR after its completion. In general, earlier models which estimated the impact of the UR produced larger welfare gains because the early projections of what would be achieved by the UR were overly optimistic. For example, Nguyen et al. (1993) assumed that: agricultural subsidies would be cut by 30% in all countries; agricultural import protection would be cut by 20%; tariffs and non‐tariff barriers in manufactures would be cut by between 30 to 50%, depending on country and product; and non‐tariff barriers in services would be cut by 40%. As a result they obtained a relatively large estimate, for a comparative static CRTS model, that the welfare of the world would increase by 1.1%. 20 The 2005 estimates for each region in their table 11 b are obtained by multiplying the corresponding 1990 estimates for that region in their table 11 a by a scalar which is based on OECD and World Bank growth projections for that region. Specifically, they estimate annual growth rates for OECD countries over the 1990–2005 period of 3.3%, for China of 8.5%, for Taiwan of 5.7%, and for Developing and Transitional countries of 4.8%, resulting in scalar adjustments of 63%, 340%, 230% and 202%. 21 For example, agricultural export taxes, as exist in Sub‐Saharan Africa, were cut in these studies, while there is no requirement in the UR that export taxes be cut. Moreover, a 36% reduction in agriculture import taxes is considerably larger than the cuts estimated by Ingco (1994). 22 Another reason that agriculture is responsible for such a high proportion of the benefits in the RUNS studies is that, given the aggregate characterisation of the manufacturing sector, these studies did not model the impact of elimination of the MFA. See Harrison, Rutherford and Tarr (1995) for details, as well as a discussion of some other assumptions we modified to approximate the RUNS model. References Baldwin , Robert ( 1994 ). ‘ An economic evaluation of the Uruguay Round agreements .’ Mimeo Department of Economics, University of Michigan . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Brandão , Antônio Salazar P. and Martin , Will J. ( 1993 ). ‘ Implications of agricultural trade liberalization for the developing countries . Agricultural Economics , vol. 8 , pp. 313 – 43 . Google Scholar Crossref Search ADS WorldCat Brooks , Michael A. and Heijdra , Ben J. ( 1988 ). ‘In search of rent‐seeking.’ In The Political Economy of Rent‐Seeking (ed. C. K. Rowley, R. D. Tollison and G. Tullock). Boston: Kluwer Academic Publishing . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Dantzig , George and Manne , Alan S. ( 1974 ). ‘ A complementarity algorithm for an optimal capital path with invariant proportions . Journal of Economic Theory , vol. 9 , pp. 312 – 23 . Google Scholar Crossref Search ADS WorldCat Francois , Joseph F. , McDonald , Bradley and Nordström , Håkan ( 1994 ). ‘ The Uruguay Round: a global general equilibrium assessment .’ Mimeo GATT Secretariat , Geneva, September. Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Francois , Joseph F. , McDonald , Bradley and Nordström , Håkan ( 1995 ). ‘Assessing the Uruguay Round.’ In The Uruguay Round and the Developing Countries (ed. Will Martin and L. Alan Winters). Discussion Paper No. 307, Washington, DC: The World Bank . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Gehlhar , Mark , Gray , Denice, Hertel , Thomas W., Huff , Karen, Ianchovichina , Elena, McDonald , Bradley J., McDougall , Robert, Tsigas , Marinos E. and Wigle , Randall. ( 1997 ). ‘Overview of the GTAP data base.’ In Global Trade Analysis: Modeling and Applications (ed. T. W. Hertel). New York: Cambridge University Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC General Agreement on Tariffs and Trade (GATT) ( 1994 ). The Results of the Uruguay Round of Multilateral Trade Negotiations . Geneva: GATT Secretariat, November . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Goldin , Ian , Knudsen , Odin, and van der Mensbrugghe , Dominique ( 1993 ). Trade Liberalization: Global Economic Implications . Washington, DC: OECD and World Bank . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Goldin , Ian and van der Mensbrugghe , Dominique ( 1994 ). ‘ The Uruguay Round: an initial assessment ‘, mimeo, World Bank, December 9. Hamilton Carl ( 1986 ). ‘ASEAN systems for allocation of export licences under VERs.’ In The Political Economy of Manufacturing Protection: Experiences of ASEAN and Australia (ed. C. Findlay and R. Garnaut). Sydney: George Allen & Unwin . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Hansen , Terje and Koopmans , Tjalling, ( 1972 ). ‘ On the definition and computation of a capital stock invariant under optimization . Journal of Economic Theory , vol. 5 , pp. 487 – 523 . Google Scholar Crossref Search ADS WorldCat Harris , Richard G. ( 1986 ). ‘Market structure and trade liberalization: a general equilibrium assessment.’ In General Equilibrium Trade Policy Modeling , (ed. T. N. Srinivasan and J. Whalley). Cambridge, MA: MIT Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Harrison , Glenn, W. , Jones , Richard, Kimbell , Larry J. and Wigle , Randall ( 1993 ). ‘ How robust is applied general equilibrium analysis Journal of Policy Modeling , vol. 15 , no. 1 , pp. 99 – 115 . Google Scholar Crossref Search ADS WorldCat Harrison , Glenn W. , Rutherford , Thomas F. and Tarr , David G. ( 1993 ). ‘ Trade reform in the partially liberalized economy of Turkey . The World Bank Economic Review , vol. 7 , no. 2 (May) pp. 191 – 217 . Google Scholar Crossref Search ADS WorldCat Harrison , Glenn W. , Rutherford , Thomas F. and Tarr , David G. ( 1995 ). ‘Quantifying the Uruguay Round.’ In The Uruguay Round and the Developing Countries , Discussion Paper No. 307 (ed. Will Martin and L. Alan Winters). Washington, DC: The World Bank . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Harrison , Glenn W. , Rutherford , Thomas F. and Tarr , David G. ( 1996 ). ‘ Increased competition and completion of the market in the European union: static and steady state effects . Journal of Economic Integration , vol. 11 , no. 3 (September) pp. 332 – 65 , earlier version with appendices appeared as ‘Product standards, imperfect competition and the completion of the market in the European Community.’ Policy and Research Working Paper, no. 1293, The World Bank, April, 1994. Google Scholar Crossref Search ADS WorldCat Harrison , Glenn W. , Rutherford , Thomas F. and Wooton , Ian ( 1989 ). ‘ The economic impact of the European Community . American Economic Review (Papers & Proceedings) , vol. 79 , (May) pp. 288 – 94 . OpenURL Placeholder Text WorldCat Harrison , Glenn and Rutherford , Thomas F. and Wooton , Ian ( 1991 ). ‘ An empirical database for a general equilibrium model of the European Communities . Empirical Economics , vol. 16 , pp. 95 – 120 . Google Scholar Crossref Search ADS WorldCat Harrison , Glenn W. , Rutherford , Thomas F. and Wooton , Ian ( 1995 ). ‘ Liberalizing agriculture in the European Community . Journal of Policy Modeling , vol. 17 , pp. 223 – 55 . Google Scholar Crossref Search ADS WorldCat Harrison , Glenn W. and Vinod , H. D. ( 1992 ). ‘ The sensitivity analysis of applied general equilibrium models: completely randomized factorial sampling designs . Review of Economics & Statistics , vol. 74 , (May) pp. 357 – 62 . Google Scholar Crossref Search ADS WorldCat Hathaway and Ingco , M. ( 1995 ). ‘Agricultural liberalization and the Uruguay Round.’ In The Uruguay Round and the Developing Countries , Discussion Paper no. 307 (ed. Will Martin and L. Alan Winters). Washington, DC: The World Bank . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Hernandez , E. ( 1985 ). La Productividad y el Desarrollo Industrial en México . Mexico City: Fondo de Cultura Económica . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Hertel , Thomas, W. , Martin , Will, Yanagishima , Koji and Dimaranan , Betina ( 1995 ). ‘Liberalizing manufactures trade in a changing world economy.’ In The Uruguay Round and the Developing Countries , Discussion Paper no. 307 (ed. Will Martin and L. Alan Winters). Washington, DC: The World Bank . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Hertel , Thomas W. and Tsigas , Marinos E. ( 1996 ). ‘Structure of GTAP.’ In Global Trade Analysis: Modeling and Applications (ed. T. W. Hertel). New York: Cambridge University Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Ingco , Merlinda ( 1996 ). ‘ Tariffication in the Uruguay Round: how much liberalization . World Economy (July) pp. 425 – 46 . Martin , Will and Francois , Joseph ( 1994 ). ‘ Bindings and rules as trade liberalization .’ mimeo International Trade Division, International Economics Department, World Bank (November). Melo , Jaime de and Tarr , David ( 1992 ). General Equilibrium Analysis of U.S. Foreign Trade Policy . Cambridge, MA: MIT Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Neven , Damien J. ( 1990 ). ‘ Gains and Losses from 1992 . Economic Policy (April), pp. 13 – 61 . Nguyen , Trien , Perroni , Carlo and Wigle , Randall ( 1991 ). ‘ The value of a Uruguay Round success . World Economy , vol. 14 , pp. 359 – 74 . Google Scholar Crossref Search ADS WorldCat Nguyen , Trien , Perroni , Carlo and Wigle , Randall ( 1993 ). ‘ An evaluation of the draft Final Act of the Uruguay Round . Economic Journal , vol. 103 (November) pp. 1540 – 9 . Google Scholar Crossref Search ADS WorldCat Nguyen , Trien and Wigle , Randall ( 1992 ). ‘ Trade liberalisation with imperfect competition: the large and the small of it . European Economic Review , vol. 36 , pp. 17 – 35 . Google Scholar Crossref Search ADS WorldCat Perroni , C. and Whalley , J. ( 1994 ). ‘ The new regionalism: trade liberalization or insurance ?’ NBER Working Paper No. 4626. Cambridge, MA: National Bureau of Economic Research . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Pratten , C. ( 1987 ). ‘ A survey of economies of scale .’ Report prepared for the EC Commission, Brussels. Reinert , Kenneth A. and Roland‐Holst , David W. ( 1992 ). ‘ Armington elasticities for United States manufacturing sectors . Journal of Policy Modeling , vol. 14 , no. 5 , pp. 631 – 9 . Google Scholar Crossref Search ADS WorldCat Roland‐Holst , David W. , Reinert , Kenneth A. and Shiells , Clinton R. ( 1992 ). ‘ North American trade liberalization and the role of nontariff barriers .’ In Economy‐Wide Modeling of the Economic Implications of a FTA with Mexico and a NAFTA with Canada and Mexico. (United States International Trade Commission) . Washington, DC: USITC Publication 2508. Rutherford , Thomas ( 1992 ). ‘ Extensions of GAMS for complementarity and variational problems arising in applied economics ,’ Economics Working Paper 92–9, Department of Economics, University of Colorado at Boulder . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Schott , Jeffrey and Buurman , Johanna ( 1994 ). The Uruguay Round: An Assessment . Washington, DC: Institute for International Economics . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Shiells , C. R. and Reinert , K. A. ( 1993 ). ‘ Armington models and terms‐of‐trade effects: some econometric evidence for North America . Canadian Journal of Economics , vol. 26 , no. 2 , pp. 299 – 316 . Google Scholar Crossref Search ADS WorldCat Tarr , David G. ( 1987 ). ‘ Effects of restraining steel exports from the Republic of Korea and other countries to the United States and the European Economic Community . World Bank Economic Review , vol. 1 , no. 3 , (September), pp. 397 – 418 . Google Scholar Crossref Search ADS WorldCat Tarr , David G. ( 1994 ). ‘ The welfare costs of price controls for cars and color televisions in Poland: contrasting estimates of rent‐seeking from recent experience . World Bank Economic Review , vol. 8 , no. 3 , (September), pp. 415 – 43 . Google Scholar Crossref Search ADS WorldCat Tarr , David, G. (forthcoming) ‘ Effects of restraints on steel exports of Korea .’ In Export Restraints and Emergency Protection (ed. Hugh Patrick). Trela , Irene and Whalley , John ( 1990a ). ‘ Global effects of developed country trade restrictions on trade and apparel . Economic Journal vol. 100 (December), pp. 1190 – 205 . Google Scholar Crossref Search ADS WorldCat Trela , Irene and Whalley , John ( 1990b ). ‘ Internal quota allocation schemes and the costs of the MFA .’ Mimeo Department of Economics, University of Western Ontario (September). Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Tullock , Gordon ( 1988 ). ‘Rents and rent‐seeking.’ In The Political Economy of Rent‐Seeking (ed. C. K. Rowley, R. D. Tollison and G. Tullock). Boston: Kluwer Academic Publishing . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Tyers , Rod and Falvey , Rod ( 1989 ). ‘ Border price changes and domestic welfare in the presence of subsidized exports . Oxford Economic Papers , vol. 41 , pp. 434 – 51 . Google Scholar Crossref Search ADS WorldCat United States Government Accounting Office ( 1994 ). The General Agreement on Tariffs and Trade: Uruguay Round Final Act Should Produce Overall U.S. Economic Gains . Washington, DC: General Accounting Office (GAO/GGD–94–83b). Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Winters , L. Alan ( 1984 ). ‘ Separability and the specification of foreign trade functions . Journal of International Economics , vol. 17 , pp. 239 – 63 . Google Scholar Crossref Search ADS WorldCat Yongzheng , Yang ( 1994 ). ‘ The impact of MFA phasing out on world clothing and textiles markets . Journal of Development Studies , vol. 30 , no. 3 (April) pp. 892 – 915 . OpenURL Placeholder Text WorldCat Yongzheng , Yang , Martin , Will and Yanagishima , Koji ( 1997 ). ‘Evaluating the benefits of abolishing the MFA in the Uruguay Round package.’ In Global Trade Analysis: Modeling and Applications (ed. T. W. Hertel). New York: Cambridge University Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Author notes The views expressed are those of the authors alone and should not be interpreted as the opinion of the World Bank. We are grateful to Will Martin for comments. © Royal Economic Society 1997
TI - Quantifying the Uruguay Round
JF - The Economic Journal
DO - 10.1111/j.1468-0297.1997.tb00055.x
DA - 1997-09-01
UR - https://www.deepdyve.com/lp/oxford-university-press/quantifying-the-uruguay-round-poyPYHognv
SP - 1405
EP - 1430
VL - 107
IS - 444
DP - DeepDyve
ER -