乌拉圭回合谈判的关税协定及欧盟反倾销调查作业
07-15, 2014
在大多数情况下征收税率反映倾销幅度，除非这个重大的损害也可以以一个较低的税率被挽回（'更低的税收法则”）。该反倾销进口税可能是一个从价税，一个特定的税收或一个可变的税收（即最低进口价格）。此外，在大多数情况下本着WTO与反倾销措施期限为5年的规定。针对当事人可以请求一个可能会引起较低税率的中期审评。
没有出口到欧盟的新成员在调查的时候，也受到后者的反倾销税的影响（主要是为了防止规避）。鉴于这一过程评估和税收的制定，和一个反倾销调查指数微小的变动都可能影响企业的行为，该委员会已消除外国竞争相当大的自由裁量权，保护国内产业与国外生产商。
The imposed duty rate in most cases reflects the dumping-margin unless the material injury could also be withdrawn with a smaller duty rate (‘lesser duty rule’). The anti-dumping import tax may either be an advalorem duty, a specific duty or a variable duty (i.e. a minimum import price). Moreover, in line with WTO regulations antidumping measures are in most cases imposed for a period of 5 years. Targeted parties may however ask for an interim review which may result in lower duty rates.
Same exporting country (Macrory et al., 1991). Newcomers, which did not export to the EU at the time of the investigation, are also subjected to the latter antidumping duties (mostly in order to prevent circumvention). Given the nature of this process of evaluation and duty setting, and that the mere initiation of an AD investigation may affect firms’ behaviour, the Commission has considerable discretion to eliminate foreign competition and to protect the domestic industry against foreign producers.
Uruguay Round Tariff Commitments and EU Antidumping Investigations During the Uruguay Round the European Union agreed to reduce its bound tariffs by almost a third, with considerable variation across industries and individual product lines. Table 1 (below) provides an overview of the EU’s bound Uruguay Round MFN tariff cuts, per industry. The sector with the largest average decline in tariff protection was the tobacco industry, with a cut of around 24 percentage points.15 Containing a much larger number of individual HS 8-digit product lines, the iron and steel sector comes second showing an average reduction in tariff protection of approximately 5.1 percentage points, followed by the processed food, furniture, paper, beverages and chemicals industries.16 In addition, coefficients of variation displayed in Column (4) also reveal that the tariff cuts within individual industries were not conducted uniforMLy and were subject to considerable intraindustry.
The most successful issuer of dumping complaints was the iron and steel industry with 491 investigated product-country pairs, followed by the textiles (232), industrial chemicals (114), fabricated metals (112), footwear (99) and electrical machineries (69) industries.The potential of antidumping constraints to provide import protection to the domestic industry has also been highlighted by Messerlin and Reed (1995), who find that 90% of all AD measures are implemented on the basis of rather loose injury criteria - such as simple differences in prices rather than actual predatory pricing behaviour.#p#分页标题#e#
The relatively large average tariff reduction in the tobacco sector has to be interpreted with some caution as the tobacco sector only counts 6 HS 8-digit product lines, whereas the iron and steel industry includes 573 HS 8-digit products.The latter industries show average tariff cuts of 4.3, 4.2, twice 3.9 and 3.4 percentage points, respectively.Despite the much smaller average reductions in the latter sectors (relative to the tobacco sector), the former arestill considerably above the average MFN tariff reduction which amounts to 2.7 percentage points for allmanufacturing industries.
A list of the countries targeted by an EU antidumping investigation over the considered time horizon is provided in Annex table . A similar ordering emerges when analysing the final imposed antidumping duties, with the iron and steel sector being the prime user of antidumping measures counting 218 product-country pairs subject to an 7number of AD-targeted product lines instead of product-country pairs (Table 1, Column 6) results in a very similar ordering with the iron and steel, chemicals and textile industries representing the sectors with the highest number of AD targeted product lines.
Notes: The above statistics are based on the author's own calculation using product-country level import data from Comext, bound Uruguay Round tariff changes from the WTO's schedule of concessions and antidumping data from the World Bank’s global antidumping database. The statistics displayed in Table 1 are based on 2,680,408 observations. It is worthwhile noting that while the above table includes all country-specific HS 8-digit EU import flows between 1996 and 2008 our estimations only include countries and 4-digit ISIC industries where at least one antidumping investigation had been initiated over the considered time horizon. Introducing
lagged regressors and growth variables further reduces the estimating sample to 701,272 observations including 1061 antidumping antidumping duty. Further sectors with a rather high incidence of imposed duties are the fabricated metal , industrial chemicals , footwear and electrical machineries sectors.
The exact industry ordering is iron and steel , industrial chemicals , textiles ,footwear , electrical machinery , fabricated metals. Moreover, analysing the distribution of imposed preliminary
and final duties per industry (Annex table 4) delivers further interesting insights. The highest preliminary duties were, on average, imposed in the non-metallic minerals, the non-electrical machinery and the leather products industries, while the sectors with the highest average of imposed final duties were the electrical and non-electrical machinery, as well as the footwear and wearing apparel industries .
Targeted product-country pairs. Annex table 4 displays the distribution of AD investigations per industry for the estimating sample and shows an almost identical frequency distribution across different industries.#p#分页标题#e#
Conceptual Framework
Our empirical analysis is motivated by Anderson and Schmitt’s (2003) theoretical framework which provides a rationale for investigating the impact of tariff liberalization on
the use of quotas and antidumping measures. Building on Brander and Krugman’s (1983) model of reciprocal dumping, they show that countries tend to resort to antidumping measures when tariffs and quotas are credibly restricted by coordinated (e.g. multilateral) commitments. Focusing on a two-country, two-firm, Cournot framework in which each firm sells the same good in both countries, Anderson and Schmitt (2003) analyse the choice between different trade policy instruments by means of a government objective function in the presence and absence of a multilateral liberalization commitment.
where ti represents an ad-valorem tariff rate for product i and qri and adi denote advalorem tariff equivalents of a binding quota or an antidumping restriction, respectively. 20 For simplicity the authors exclude the potential entry and exit of firms. Since some form of market imperfection is need in order to explain the use of quotas and antidumping constraints, the authors assume an oligopolistic market structure. Adopting a strategic Cournot interaction implies a tariff quota equivalence if quotas are auctioned off (see for instance Hwang and Mai, 1988). Anderson and Schmitt (2003) include transportation costs as an additional trade barrier in their model, but we simplify here by omitting these barriers.
Assuming complete discretion for the government to set trade policy tools as freely as it chooses,23 Anderson and Schmitt establish that tariffs, when set unilaterally, are the most efficient protectionist trade policy tool. The intuition is that, while all three alternatives are likely to exert a similar impact on domestic prices, and by consequence on domestic producers’ profits, non-tariff trade barriers are assumed to be more costly for governments since the latter won’t generate any revenue gains.24 Anderson and Schmitt also show that, given any tariff rate t, unilaterally imposed quantitative constraints are preferred to antidumping constraints, since the latter may exert a negative impact on the domestic industry’s export profits whereas the former leave export profits unaffected.25 As a result, the following preference ranking of trade policy tools emerges:
Allowing both countries to commit to (reciprocal) trade liberalization via an internationally binding agreement, it is assumed that the tariff commitment reduces and restricts the use of external tariff protection relative to the unconstrained Nash equilibrium. Enhanced trade flows and declining local market power lead to Pareto improvements for both signatory countries [i.e. Uc(.) > Un(.)]. Given a government’s (still present) incentive to change the terms-of-trade to its own advantage the former may, however, decide to explore alternative (and potentially more subtle) ways of import protection following the binding tariff agreement.27 In this context, the government is assumed to first negotiate binding multilateral tariff cuts in order to internalize terms-of-trade effects, and then to look for 23 Several authors note that the use of antidumping measures is influenced by political-economy forces and thus may lead to a less strict interpretation of the (anti-)dumping legislation (Moore, 1992; Baldwin and Steagall, 1994; Zanardi, 2004).#p#分页标题#e#
Anderson and Schmitt (2003) assume that quotas cannot be licensed off and are thus lost to foreign producers. The authors therefore find that any quota-cum-tariff or antidumping-cum-tariff protection can be achieved by using a (higher) optimal tariff. Since the latter additionally generates tariff revenue the use of tariffs maximizes .
In this context, it is interesting to note that the Uruguay Round substantially reduced bound MFN tariffs for most signatory countries thereby representing a credible internationally-binding commitment on the use of tariffs. Moreover, the trade round also required its signatory countries to ‘tariffy’ quantitative restrictions and to limit their use in the future.29 The use of NTBs, and in particular of antidumping measures, is however much less regulated. In light of very limited WTO-restrictions on the use of antidumping actions, we hence argue that the Uruguay Round trade agreements may represent a suitable policy setting for a potential substitution effect of declining tariff protection for an enhanced use of product country level antidumping measures. We do not consider trade barriers stemming from technical and safety regulations given their rather less precise nature and the prevalent difficulties in finding adequate product level measures.30 Moreover, by focusing on the Uruguay Round we based our empirical examination on an institutional framework in which policy-makers were enabled to credibly commit to binding tariff and quota restrictions. As a result, the policy context we study investigates the relationship between different forms of trade policy, and guided by the theoretical framework illustrated in this section, serves as a vehicle to address the question of trade policy substitution following major, coordinated trade reforms.
没有出口到欧盟的新成员在调查的时候，也受到后者的反倾销税的影响（主要是为了防止规避）。鉴于这一过程评估和税收的制定，和一个反倾销调查指数微小的变动都可能影响企业的行为，该委员会已消除外国竞争相当大的自由裁量权，保护国内产业与国外生产商。
The imposed duty rate in most cases reflects the dumping-margin unless the material injury could also be withdrawn with a smaller duty rate (‘lesser duty rule’). The anti-dumping import tax may either be an advalorem duty, a specific duty or a variable duty (i.e. a minimum import price). Moreover, in line with WTO regulations antidumping measures are in most cases imposed for a period of 5 years. Targeted parties may however ask for an interim review which may result in lower duty rates.
Same exporting country (Macrory et al., 1991). Newcomers, which did not export to the EU at the time of the investigation, are also subjected to the latter antidumping duties (mostly in order to prevent circumvention). Given the nature of this process of evaluation and duty setting, and that the mere initiation of an AD investigation may affect firms’ behaviour, the Commission has considerable discretion to eliminate foreign competition and to protect the domestic industry against foreign producers.
Uruguay Round Tariff Commitments and EU Antidumping Investigations During the Uruguay Round the European Union agreed to reduce its bound tariffs by almost a third, with considerable variation across industries and individual product lines. Table 1 (below) provides an overview of the EU’s bound Uruguay Round MFN tariff cuts, per industry. The sector with the largest average decline in tariff protection was the tobacco industry, with a cut of around 24 percentage points.15 Containing a much larger number of individual HS 8-digit product lines, the iron and steel sector comes second showing an average reduction in tariff protection of approximately 5.1 percentage points, followed by the processed food, furniture, paper, beverages and chemicals industries.16 In addition, coefficients of variation displayed in Column (4) also reveal that the tariff cuts within individual industries were not conducted uniforMLy and were subject to considerable intraindustry.
The most successful issuer of dumping complaints was the iron and steel industry with 491 investigated product-country pairs, followed by the textiles (232), industrial chemicals (114), fabricated metals (112), footwear (99) and electrical machineries (69) industries.The potential of antidumping constraints to provide import protection to the domestic industry has also been highlighted by Messerlin and Reed (1995), who find that 90% of all AD measures are implemented on the basis of rather loose injury criteria - such as simple differences in prices rather than actual predatory pricing behaviour.#p#分页标题#e#
The relatively large average tariff reduction in the tobacco sector has to be interpreted with some caution as the tobacco sector only counts 6 HS 8-digit product lines, whereas the iron and steel industry includes 573 HS 8-digit products.The latter industries show average tariff cuts of 4.3, 4.2, twice 3.9 and 3.4 percentage points, respectively.Despite the much smaller average reductions in the latter sectors (relative to the tobacco sector), the former arestill considerably above the average MFN tariff reduction which amounts to 2.7 percentage points for allmanufacturing industries.
A list of the countries targeted by an EU antidumping investigation over the considered time horizon is provided in Annex table . A similar ordering emerges when analysing the final imposed antidumping duties, with the iron and steel sector being the prime user of antidumping measures counting 218 product-country pairs subject to an 7number of AD-targeted product lines instead of product-country pairs (Table 1, Column 6) results in a very similar ordering with the iron and steel, chemicals and textile industries representing the sectors with the highest number of AD targeted product lines.
Notes: The above statistics are based on the author's own calculation using product-country level import data from Comext, bound Uruguay Round tariff changes from the WTO's schedule of concessions and antidumping data from the World Bank’s global antidumping database. The statistics displayed in Table 1 are based on 2,680,408 observations. It is worthwhile noting that while the above table includes all country-specific HS 8-digit EU import flows between 1996 and 2008 our estimations only include countries and 4-digit ISIC industries where at least one antidumping investigation had been initiated over the considered time horizon. Introducing
lagged regressors and growth variables further reduces the estimating sample to 701,272 observations including 1061 antidumping antidumping duty. Further sectors with a rather high incidence of imposed duties are the fabricated metal , industrial chemicals , footwear and electrical machineries sectors.
The exact industry ordering is iron and steel , industrial chemicals , textiles ,footwear , electrical machinery , fabricated metals. Moreover, analysing the distribution of imposed preliminary
and final duties per industry (Annex table 4) delivers further interesting insights. The highest preliminary duties were, on average, imposed in the non-metallic minerals, the non-electrical machinery and the leather products industries, while the sectors with the highest average of imposed final duties were the electrical and non-electrical machinery, as well as the footwear and wearing apparel industries .
Targeted product-country pairs. Annex table 4 displays the distribution of AD investigations per industry for the estimating sample and shows an almost identical frequency distribution across different industries.#p#分页标题#e#
Conceptual Framework
Our empirical analysis is motivated by Anderson and Schmitt’s (2003) theoretical framework which provides a rationale for investigating the impact of tariff liberalization on
the use of quotas and antidumping measures. Building on Brander and Krugman’s (1983) model of reciprocal dumping, they show that countries tend to resort to antidumping measures when tariffs and quotas are credibly restricted by coordinated (e.g. multilateral) commitments. Focusing on a two-country, two-firm, Cournot framework in which each firm sells the same good in both countries, Anderson and Schmitt (2003) analyse the choice between different trade policy instruments by means of a government objective function in the presence and absence of a multilateral liberalization commitment.
where ti represents an ad-valorem tariff rate for product i and qri and adi denote advalorem tariff equivalents of a binding quota or an antidumping restriction, respectively. 20 For simplicity the authors exclude the potential entry and exit of firms. Since some form of market imperfection is need in order to explain the use of quotas and antidumping constraints, the authors assume an oligopolistic market structure. Adopting a strategic Cournot interaction implies a tariff quota equivalence if quotas are auctioned off (see for instance Hwang and Mai, 1988). Anderson and Schmitt (2003) include transportation costs as an additional trade barrier in their model, but we simplify here by omitting these barriers.
Assuming complete discretion for the government to set trade policy tools as freely as it chooses,23 Anderson and Schmitt establish that tariffs, when set unilaterally, are the most efficient protectionist trade policy tool. The intuition is that, while all three alternatives are likely to exert a similar impact on domestic prices, and by consequence on domestic producers’ profits, non-tariff trade barriers are assumed to be more costly for governments since the latter won’t generate any revenue gains.24 Anderson and Schmitt also show that, given any tariff rate t, unilaterally imposed quantitative constraints are preferred to antidumping constraints, since the latter may exert a negative impact on the domestic industry’s export profits whereas the former leave export profits unaffected.25 As a result, the following preference ranking of trade policy tools emerges:
Allowing both countries to commit to (reciprocal) trade liberalization via an internationally binding agreement, it is assumed that the tariff commitment reduces and restricts the use of external tariff protection relative to the unconstrained Nash equilibrium. Enhanced trade flows and declining local market power lead to Pareto improvements for both signatory countries [i.e. Uc(.) > Un(.)]. Given a government’s (still present) incentive to change the terms-of-trade to its own advantage the former may, however, decide to explore alternative (and potentially more subtle) ways of import protection following the binding tariff agreement.27 In this context, the government is assumed to first negotiate binding multilateral tariff cuts in order to internalize terms-of-trade effects, and then to look for 23 Several authors note that the use of antidumping measures is influenced by political-economy forces and thus may lead to a less strict interpretation of the (anti-)dumping legislation (Moore, 1992; Baldwin and Steagall, 1994; Zanardi, 2004).#p#分页标题#e#
Anderson and Schmitt (2003) assume that quotas cannot be licensed off and are thus lost to foreign producers. The authors therefore find that any quota-cum-tariff or antidumping-cum-tariff protection can be achieved by using a (higher) optimal tariff. Since the latter additionally generates tariff revenue the use of tariffs maximizes .
In this context, it is interesting to note that the Uruguay Round substantially reduced bound MFN tariffs for most signatory countries thereby representing a credible internationally-binding commitment on the use of tariffs. Moreover, the trade round also required its signatory countries to ‘tariffy’ quantitative restrictions and to limit their use in the future.29 The use of NTBs, and in particular of antidumping measures, is however much less regulated. In light of very limited WTO-restrictions on the use of antidumping actions, we hence argue that the Uruguay Round trade agreements may represent a suitable policy setting for a potential substitution effect of declining tariff protection for an enhanced use of product country level antidumping measures. We do not consider trade barriers stemming from technical and safety regulations given their rather less precise nature and the prevalent difficulties in finding adequate product level measures.30 Moreover, by focusing on the Uruguay Round we based our empirical examination on an institutional framework in which policy-makers were enabled to credibly commit to binding tariff and quota restrictions. As a result, the policy context we study investigates the relationship between different forms of trade policy, and guided by the theoretical framework illustrated in this section, serves as a vehicle to address the question of trade policy substitution following major, coordinated trade reforms.