摘要

在世界貿易體系目前形式旨在減少貿易壁壘。事實上，就大部分發達經濟體而言，過去成功地達成多邊貿易談判導致大量關稅消減。然而，有其他形式的進口保護也發生了嗎？隨后是否有大量的關稅減讓被使用其他形式的進口保護所取代？在本文中，我們實證研究談判達成的外部關稅的削減和隨后歐盟使用反傾銷訴訟之間的關系。證據發現烏拉圭回合關稅的更大幅度的削減增加了隨后的反傾銷調查的概率。

歐洲的反傾銷政策和烏拉圭回合談判關稅的消減

傳統的進口保護措施的執行通過關稅與貿易總協(GATT)定支持下的多邊貿易談判和世界貿易組織（WTO），被越來越多的限制。鑒于全球關稅降低的措施，在非關稅壁壘和或有保護措施的使用顯著增加，促使一些評論家表示關注這些潛在的保護政策的替代品。
 

Abstract
The world trading system in its current form aims at reducing multilateral trade barriers across
the board. Indeed, the last successfully concluded multilateral trade negotiations led to substantial tariff concessions on the part of most developed economies. What, however, happened to other forms of import protection? Have substantial tariff concessions subsequently been replaced by the use of alternative forms of import protection? In this paper we empirically investigate the relationship between negotiated external tariff cuts and the subsequent use of antidumping actions by the EU. Evidence is found for larger Uruguay Round tariff cuts increasing the probability of subsequent antidumping investigations.

European Antidumping Policy and Uruguay Round Tariff Concessions

The use of traditional measures of import protection has been increasingly restricted by multilateral trade negotiations under the auspices of the General Agreement on Tariffs and Trade (GATT), and the World Trade Organisation (WTO). In light of globally declining tariffs and quantitative measures, the remarkable increase in the use of non-tariff barriers and contingent protection measures has prompted several commentators to express concerns over their use as potential protection substitutes.

 Antidumping measures feature prominently amongst the forms of alternative import protection and their use tends to be much less constrained by coordinated, multilateral trade agreements. Given the way in which antidumping policies and procedures are designed and implemented, an alternative channel for protection lobbying emerges providing ample room for political-economy influences and hence a rationale for the substitution of trade policy instruments.

This paper investigates the hypothesis of trade policy substitution, following a major, coordinated trade reform, by examining the relationship between multilateral tariff concessions and the subsequent use of antidumping actions by the European Union (EU). Motivated by Anderson and Schmitt’s (2003) theoretical contribution, which provides a theoretical framework for an increasing use of antidumping measures following binding tariff and quota restrictions, we address the substitution hypothesis by focusing on the impact of the Uruguay Round tariff cuts on subsequent EU antidumping investigations at a detailed HS 8-digit product-country level. The Uruguay Round substantially reduced bound MFN tariffs for most signatory countries and also required its signatory countries to ‘tariffy’ and reduce quantitative restrictions. Given that antidumping actions largely remain WTO-unconstrained, we 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. Our results provide support for Anderson and Schmitt’s (2003) theoretical contribution. A statistically highly significant, however in absolute size limited, positive impact of bound MFN tariff cuts on the probability of subsequent antidumping investigations is found; having controlled for other influences. Employing a variety of different econometric techniques, including random-effects and a Chamberlain- Mundlak approach to control for unobserved heterogeneity, this finding is shown to be robust to a series of sensitivity tests.

Our results point to a substitution of different forms of trade policy instruments following the Uruguay Round, with a persistent and statistically highly significant impact of bound MFN tariff concessions on the probability of future antidumping investigations being identified. The average partial effects vary between 0.010 and 0.002, which indicates a positive, but quantitatively limited, probability increase of up to 0.00010 percentage points per one percentage point tariff reduction agreed upon during the UR. Our findings stand in contrast to previous studies in the literature.

These studies either find no statistically viable relationship between tariff liberalization and antidumping measures for developed economies, or even a weak negative link between the latter two forms of import protection as shown by Moore and Zanardi (2011) and Feinberg and Reynolds (2007), respectively.50 Common to both of the latter two studies is that they focus on rather broad industry-level tariff and antidumping actions. However, antidumping measures are, in general, imposed on a very disaggregated product-level, giving rise to a potential bias. Interestingly, our results tend to corroborate those of Bown and Tovar (2011) who analyse HS 6-digit product-level tariff cuts in India and find evidence for tariffs being substituted by an enhanced use of antidumping and safeguard measures. 50 Note that Moore and Zanardi (2011) focus on sectoral applied and not on bound MFN tariff changes between 1991 and 2001. Feinberg and Reynolds (2007) find a significant positive link between bound MFN tariff cuts and subsequent antidumping actions mainly for developing countries and a rather weak negative relationship for traditional AD using countries (i.e. Australia, Canada, EU, New Zealand, and the USA).