Customs Unions: Trade Law’s Forgotten Child
By Michal Ovádek|michal.ovadek@kuleuven.be
Until customs unions (CUs) started to make headlines in British newspapers following the UK’s decision to depart from the EU, there has been little public or legal-academic interest in the concept, perhaps with the exception of Turkey whose estimation of its CU with the EU has decreased in proportion with the probability of its joining the Union. Has the disinterest been warranted? Not quite – more than 110 countries around the world are members of at least one CU and the concept dates back to at least 19th century Germany, where it played a part in German unification. In a recent research project, my colleague Ines Willemyns and I have looked at the state of international law governing CUs in a bid to contribute to the understanding of this particular form of regional economic integration, not least because it has been a source of considerable confusion in the Brexit discourse.
The whispers and silence of WTO law
The most important global legal regime regulating CUs is World Trade Organization (WTO) law, more specifically the General Agreement on Tariffs and Trade (GATT). Article XXIV GATT, which has remained practically unchanged since 1947, constitutes a ‘regional integration’ exception from the most-favoured nation obligation of Article I GATT available to states forming free trade areas (FTAs) or CUs. In other words, if an FTA or a CU fulfils the respective requirements, WTO Members are allowed to grant reciprocal preferential treatment to other Members under such an arrangement without violating WTO law.
Article XXIV:8 GATT defines a CU as ‘the substitution of a single customs territory for two or more customs territories’ and lays down two minimum conditions for a CU to be WTO law-compliant: (1) ‘duties and other restrictive regulations of commerce (…) are eliminated (…) with respect to substantially all the trade in products originating in such territories’ and (2) ‘substantially the same duties and regulations of commerce are applied by each of the members of the union to the trade of territories not included in the union’. In addition, as provided in paragraph 5 of Article XXIV GATT, the formation of the CU cannot raise overall barriers to the trade with third countries, who should not be adversely affected by the regional integration. Only measures that are necessary for the formation of the customs union can benefit from the exception in Article XXIV GATT, as confirmed by the Appellate Body (AB) in Turkey – Textiles and Argentina – Footwear (EC).
Reading the GATT requirements immediately brings up several questions: what are ‘other regulations of commerce’? How much trade is ‘substantially all trade’? What should the common external policy encompass? Which measures are ‘necessary’ for the formation of a CU? Few of these questions have been answered with any level of specificity over the years. In the leading case concerning CU requirements, Turkey – Textiles, which concerned Turkish alignment with the EU’s quantitative restrictions on the import of textiles from India, the AB found that the condition of ‘substantially all trade’ does not mean ‘all’ trade but it does require the elimination of trade barriers on ‘more than some’ trade. Similarly, even such a basic question as whether the term ‘other regulations of commerce’ includes quantitative restrictions has not been explicitly answered to date (even though the answer should be in the affirmative for reasons of fundamental viability of CUs). Neither has the AB addressed whether trade remedies can be maintained between members to a CU.
Making customs unions work
Looking at a potential CU between the UK (or at least Northern Ireland) and the EU following Brexit illuminates the lack of guidance offered by WTO law. If the UK and the EU were to pursue a CU which – unlike the EU itself – intended to comply with only the bare minimum of WTO obligations on account of the UK’s desire to preserve sovereignty (of the Westphalian type), which regulations of commerce could be divergent? Could the UK adopt different phytosanitary (SPS) and technical (TBT) regulations? Could the EU require compliance with rules of origin? Could either party impose trade remedies on the other?
The vagueness and the silence of WTO law ultimately result in states doing more or less what they want when designing CUs. However, what states want to do is on average a lot less prejudicial to state sovereignty (as traditionally understood) than the case of the EU could suggest. In the EU, SPS measures and technical regulations are harmonised to facilitate internal trade. Such harmonisation might, however, be characterised as an element of the single market rather than the CU, the former representing a higher stage of economic integration building on the latter. There are no rules of origin requirements for goods produced and traded within the EU, nor can any of the Member States impose trade remedies, which are entirely replaced by a supranational competition policy, including a robust regime regulating the provision of state aid.
However, upon the UK’s exit from the EU, a potential CU between the two parties could look entirely different and yet still be consistent with the GATT obligations. Many SPS and TBT measures are harmonised within the European Economic Area (EEA), an enhanced FTA comprising certain EFTA countries and the EU, as the participating countries wish to reduce trade frictions at borders. However, most CUs do not entail the harmonisation of SPS and TBT measures, nor do they typically eliminate rules of origin and the possibility to use trade remedies for the simple reason that all of this would constrain the sovereignty of the CU members. Such a lack of harmonisation implies that borders and customs checks are also part of a ‘median’ CU (i.e. not the EU); they are also part of the EEA – however ‘frictionless’ the traffic – as the harmonisation of SPS and TBT measures is not in itself sufficient to allow the removal of border checks (The EEA is not a CU and therefore does not harmonise the external trade regime).
For a CU to function properly, the members need to apply not only the same tariffs (the common external tariff) but also preferential treatment with respect to third countries. Failing that, goods could be introduced into the CU at differential rates to the detriment of some members. In such a case, members must choose whether to let the ‘unharmonised’ goods flow freely through the CU and cause trade deflection or whether to conduct customs checks and require compliance with rules of origin in order to ‘catch’ any ineligible goods and potentially charge an additional levy on them. The latter correctional option is frequently taken precisely because members are not able to enact a unified external trade regime, as the EU has accomplished.
One particular aspect of the external trade regime proving difficult for many CUs to bring under common governance is the negotiation of preferential trade agreements. Some CUs, such as the Caribbean Community (CARICOM), allow their members to conclude trade agreements individually, under certain conditions. This can create divergences in the external trade regime and consequently undermine the integrity or efficiency of the CU. On the contrary, trade policy is one of the exclusive competences of the EU, which cannot be exercised by the Member States separately. Similarly, in the Southern African Customs Union (SACU), members can only conclude trade agreements individually with the consent of the rest. However, the SACU has struggled to create an effective common negotiation mechanism, as South Africa has been unwilling to share competences with its much less powerful co-members of the CU. Power imbalances and regional hegemonies can also be observed in the Eurasian Economic Union (Russia) or the EU-Turkey CU (EU). In the latter CU, Turkey is under an obligation to align itself with EU trade policy without having any say in it. More understandably, the same applies to San Marino and Andorra.
No matter the Brexit outcome, the confusion which enveloped the issue of customs unions is at least partly down to the imprecise WTO rules governing the subject and the resulting diversity in what passes off as a customs union in name. The EU model might be the most effective one but looking at the fine print it is certainly not the most prevalent. Sovereign statehood still reigns supreme in international relations and the functions performed by a CU can easily get in the way of the established order of things. Countries must weigh the costs of pooling a portion of their sovereignty against the benefits such as stronger international leverage. The costs and benefits in that equation are, however, far from purely economic, as the UK’s example shows.
Michal Ovádek is a researcher at the Leuven Centre for Global Governance Studies and a PhD candidate at the Faculty of Law, KU Leuven. He works primarily on constitutional questions of EU law, in particular the rule of law, fundamental rights and the concept of legal basis. Tweets @michal_ovadek.