Reports Featured

Brexit, Tariffs, and GATT’s original intent: Why a forgotten MFN exemption merits a closer look

Brexit Tariffs
Written by Dr Thomas D. Grant

Exit from the EU on GATT/WTO terms is assumed to mean that the UK will find itself between a Scylla and Charybdis of either eliminating all tariffs on all countries or snapping into place tariffs on the remaining EU Member States at the same levels that the EU currently applies to the rest of the world.

The most-favoured nation rule—MFN—is indeed an important tool of the world trade system, and if the UK and EU were complete strangers (and all else being equal) there is no doubt MFN would apply in that way between them, and between both of them and third parties. However, there is a largely overlooked history of MFN exemptions for countries departing complex legal-political unions. Viewing the history together with the GATT/WTO principles that countries are not to use trade as a political lever and trade rules ought to promote more trade liberalisation (not more tariffs), it is arguably more consistent with GATT’s original intent to keep, at least on an interim basis, the existing frictionless trade between Britain and the EU after Brexit.


If the UK does not enter into a free trade agreement with the EU, interim or otherwise, then what happens to tariffs post-Brexit? Dominic Grieve says tariffs are inevitable in that scenario: ‘with no deal, we must put up barriers to trade with all EU member states’. Others, like Tesco CEO Dave Lewis, seem to suggest that Britain could avoid tariffs: Lewis contemplates the possibility that ‘as part of no deal there is no tariff’ on trade with the remaining EU States (EU27). Who is right?

There is some gravity in the concern that departure without a new agreement means tariffs, because Article I, paragraph 1, of the General Agreement on Tariffs and Trade (GATT), which is the treaty at the heart of the World Trade Organisation (WTO), stipulates that every GATT party accord ‘most favoured nation’ treatment to all others. Most-favoured nation treatment—MFN—requires States to which it is applicable to treat like goods alike, regardless of their origin. Under MFN, a State which, for example, lowers its tariffs to zero on cotton t-shirts from another State, must also lower its tariffs to zero on cotton t-shirts from all other States. The same goes the other way: if, instead, a State has tariffs of, say, 2 per cent on cotton t-shirts from one State, then it must also have tariffs of 2 per cent on cotton t-shirts from all other States. GATT provides a well-known exemption to MFN: if two or more States form a free-trade area or a customs union, under which they give one another preferences with respect to ‘substantially all’ their trade—like the zero-tariff on t-shirts but applied to substantially all goods—then they may treat States outside their area or union differently; they may have tariffs on t-shirts from the outsiders.

Those who predict customs delays at the UK’s borders and shortages in industrial supply chains, as well as on consumer-facing store shelves, make reference to MFN. It is MFN, they say, that will compel a change to tariffs. Leaving the customs union which forms one part of the EU’s overall system, the UK from the moment it leaves will be obliged, they say, to trade under MFN and, thus, the UK will be obliged to give no preference to EU goods that it does not also give to goods from elsewhere. Absent a new free trade agreement between the UK and the EU, so the thinking goes, goods traveling into the UK from the remaining EU Member States (the EU27) will have tariffs imposed on them; and the enforcement of the new tariffs will require customs checks—and delays—at UK ports—and shortages throughout the economy. This, at any rate, is what some commentators, such as Dominic Grieve, say must happen when a State leaves a free-trade area or customs union.

The UK’s departure from the EU, however, is not a departure from a mere free-trade area or customs union. The European Community, as it was when the UK joined, though most UK politicians in the 1970s didn’t understand it, was a treaty organisation with ambitions to be a State – or at least with ambitions to tie its members much more deeply than typical free trade arrangements do. The EU is not the first formation on the international stage to have embodied aspects of a State without being quite as much as that – but it is the first since GATT was originally adopted to develop so strikingly along such lines.

Earlier examples, by mid-century when GATT was drafted, had recently dissolved or where in their final phases. These included various empires, defunct or soon to be, and an assortment of other territories with more or less improvised legal-political ties to a State. The drafters of GATT recognised that such phenomena—and they were diverse; no single label described them—posed a question for the new trade rules: namely, how to treat the independent States emerging from them? The approach GATT’s drafters took was sensible: exempt those States from MFN, so that their emergence from earlier relationships was not unnecessarily complicated by the imposition of new tariffs. Thus, where between certain territories free or preferential trade had prevailed, GATT did not oblige its interruption, even though the relationships between them in other respects had come to a close. The UK’s exit from the EU today is much like those situations that the GATT in 1947 recognised merited an exemption from MFN. A consistent approach which gives effect to the original intent behind GATT to foster security and predictability in the face of such legal-political change would treat the UK’s exit and those situations alike.

GATT’s forgotten MFN exemption

A curious thing about GATT is that it does not say much, if anything, about what happens when a typical Free Trade Area or Customs Union dissolves or a Member exits — but it does say something about the dissolution of and exit from a variety of other relationships. The relationships it addresses in this way are unions formed under constitutions or under treaties or created by empires; protectorates; and a range of other dependencies and international territorial constructs. GATT reflects that such relationships existed and that those which had survived as late as 1947, were in the process of breaking up or undergoing major change. GATT permitted States which had left or were leaving a relationship like these to continue an existing tariff régime or single market with regard to one another—without having to implement the same preference to the rest of GATT’s parties. Thus, though the relationship was otherwise over or ending, its trade aspect was preserved without attracting the MFN obligation. Noteworthy for present purposes, while GATT Article XXIV, paragraph 5, exempts States entering free-trade areas or customs unions from MFN, other parts of the treaty exempt States exiting, where what they were exiting was a legal-political relationship that included a single market or other preferential treatment but also had gone deeper than that. The first MFN exemption is well-known; the second, all but forgotten.

What situations precisely did the GATT drafters have in mind, when they adopted this other exemption to cover States leaving relationships that had entailed aspects of a free-trade area or customs union but that had had other aspects as well, in particular of historical and political nature? The Ottoman Empire is one example named in terms (Art. I para. 3). Then-current or recently former territories of the British Empire are also covered (Art. I para. 2(a) Annex A). So are the French protectorate relations (e.g., with Morocco), the French Union (which included Viet Nam) (Art. I para. 2(b) Annex B), the Philippines (then recently independent from the USA) (Art. I para. 2(b) Annex D), and Cuba (which had been under a sort of protectorate with the USA but by 1947 had full political independence) (Art. I para. 2(c)). GATT accorded former constituents of the Raj — i.e., India and Pakistan — MFN exemption as well (Art. XXIV para. 11). Trieste, which the Peace Treaty separated from Italy and declared a ‘Free Territory’ under Security Council responsibility, was permitted to have trade advantages vis Italy and Yugoslavia without having to give the same advantages to others (Art. XXIV para. 3(b)).

The express indication of these cases as exceptions to MFN would seem — at first blush — to mean that no other cases are to be treated the same way (expressio unius). However, on closer inspection, another curious thing comes to view: the GATT drafters named practically every situation that they knew of, where an historical or political union had given way, or was giving way, to separate parts as at 1947. They left no stone unturned, and for each post-union situation, they disapplied MFN. The disapplication of MFN, thus, was practically complete for such acts of separation as at the time of the drafting of GATT. The drafters did not provide in explicit terms for the possibility that similar cases of separation might arise in the future. In view of the path that international society was on in their day, it is understandable that they did not.

Why GATT addresses the post-union MFN exemption the way it does

International society, in the day of the GATT drafters, was undergoing rationalization. Before, there had been a bewildering array of diversity in the forms that international legal actors took. Consider the protectorates. Each of these was the result of a highly idiosyncratic treaty relation; no two were alike; nearly all entailed aspects of political union, even as each also entailed the preservation of existing rights in the protected territory; the precise balance among these divergent aspects was set on special terms, differing from treaty to treaty. Then there were the special arrangements of sui generis character, like the U.S. semi-protectorate over Cuba; or Britain’s government of India and the variety of ‘indirect rule’ arrangements throughout the (then-dissolving) Empire. It was arrangements like these that were disappearing as the century progressed. Capitulatory régimes under protectorate treaties were being denounced. Colonies were being decolonized. Old unions were breaking up. How statesmen of the time built their international organisations reflected the trend. For example, where the Covenant of 1919 had allowed Dominions and even Colonies, in addition to States, to be admitted as Members of the League of Nations (Covenant Art. 1), the Charter of 1945 opened the new United Nations only to States (Charter Art. 4). The process of rationalization was steady and, it seemed, relentless. The world of the future would be a world of States. This is the world in which the drafters of GATT did their drafting.

The point is this: the GATT drafters addressed the cases of disappearing unions, protectorates, and other international law exotica which they knew, and they exempted the former constituents thereof from MFN. They were drafting at a time when these cases were a residual category; nobody thought there would be anything quite like them in future, because the trend was toward rationalization, not variegation, of international legal forms. The MFN exemption was a housecleaning exercise – and nobody thought anybody would untidy the house again.

It is true that the GATT drafters envisaged future FTAs and customs unions. They addressed these, however, begrudgingly — as an exception to the intended application of GATT, which otherwise was to be universal and uniform. Moreover, they understood that they would be precisely what they called them: free-trade areas and customs unions. They did not see that they would be, nor did they expect them to become, new political organisms acting on the international plane like States but exhibiting unique characteristics making them hard to categorise. The French protectorate treaties, the British Empire, the various other colonial or quasi-colonial administrative arrangements – each of these involved preferential trade for some or all purposes, but that was not the only, or even the most, salient thing about them. Most of them embodied military, strategic, political, and other characteristics well beyond a mere free-trade area or customs union.

The time of GATT, however, did not seem to be the beginning for the likes of these diverse international legal forms. It seemed to be their end. As at 1947, many of the relationships noted in the Treaty were already defunct, and the passing of the rest was just a matter of time. The GATT drafters, to accommodate their break-up and disappearance, acknowledged them and gave the exemption from MFN to their parts. The exemption was extended on specific terms; it was not phrased as if to benefit a generally-defined category. This is not surprising. The sheer diversity of those legal arrangements defied generalization. And why even try? They were on the wane. They were artefacts from a past age of international society, not models for a brave new world. As far as the GATT drafters knew, we would, with apologies to Horatio and Hamlet, ‘goodly or otherwise… never see the likes of [them] again’.

A surprising entry—and exiting without surprises

But, lo and behold: enter the ‘ever closer Union’, as the Preamble to the Treaty on the Functioning of the European Union and Article 1 of the Treaty on European Union call the body they govern. The European Community, which did not exist in 1947, gradually emerged on the scene. What later it came to be was something very much unexpected — not a mere tariff régime, but, instead, a sui generis international law actor displaying much deeper integration, a body functioning for some purposes as a State, and yet not quite fitting that standard form. European federalists even in the 1940s dreamed of European Union; what they yearned for was a State. What in fact emerged was something rather less than that—but much more than a system of trade preferences. What emerged defies easy categorization. It interacts with the WTO under GATT as a customs union; but nobody thinks that the EU is only that. It is certainly progressive in many of its social and political aims – but, in its legal morphology, the EU is a throwback. It recalls the earlier era when the world was populated with considerably more diverse legal-political forms.

So, ask oneself this: If the EU had existed in 1930; and then a Member separated from it in 1946, how would the GATT drafters have treated the situation? It is at least open to consideration that they would have treated it as exempt from MFN, just as they treated practically every former and waning union, protectorate, and empire at the time. The counter-factual history here is to a legal purpose: it is to suggest an interpretation of GATT today that accords with the original intent of the drafters.

MFN is not an end unto itself. It is a tool within the wider system started by GATT.

For one thing, MFN is a tool to promote trade liberalization, one of the central purposes of GATT. It is hard to reconcile that purpose to an interpretation of MFN that compels the snapping into place of tariffs between two of the world’s largest economies. The GATT drafters, considering the Philippines and the USA, Morocco and France, Pakistan and India, and various others, exempted them from MFN — for the reason that the raising of tariff barriers between them, quite apart from frustrating new-found independence and perhaps in some cases even driving countries back into political relationships they chose to leave, would have been a reversal of the trade liberalization GATT was seeking to promote. True, these were relatively modest economies—but, if size matters here at all, the rationale for the MFN exemption applies even more, not less, the larger the separating economies concerned.

MFN serves another purpose, related to but distinct from liberalisation. MFN serves to curtail the use of tariffs as a political lever. If GATT parties use tariffs as a political lever, then the trade régime will be subject to the surprises that politics brings. MFN was intended to avoid such surprises, not to multiply them. Opponents of Brexit invoke the threat of a tariff wall between Britain and Europe for the purpose of exercising political leverage—and, seemingly oblivious to the paradox, they say that MFN makes such a wall inevitable.

The interpretation of GATT that refrains from introducing an MFN lever to Brexit politics is preferable on general principles. Also, it is an interpretation supported by how the GATT drafters applied an MFN exemption to analogous situations of political change.

It is to be submitted that the GATT drafters’ approach is better than that urged by people who invoke MFN to deter Brexit. Imposing tariffs between the exited State and the remaining Union is not how like cases were addressed in 1947. It is not how the separation of the UK from the EU should be addressed in 2019.

Concluding points

On the basis of the considerations set out above, these concluding points may be made:

(i) States exiting complex legal-political relationships in the past have been exempt from MFN under GATT (post-union exemption);

(ii) GATT has the object and purpose of promoting global trade liberalization and preventing States from using tariffs as a political lever;

(iii) MFN should not be applied in a manner that requires tariffs between the UK and the remaining EU Member States after the UK exits, because the more consistent approach with past practice and with the object and purpose of GATT is to apply the post-union exemption;

(iv) a new trade agreement between the UK and EU, post-Brexit, is desirable if reached on appropriate terms – but it is not necessary for avoiding MFN; MFN, even absent a new trade agreement, need not immediately apply;

(v) in the event that the EU insisted that no MFN exemption is available post-Brexit, then this would be tantamount to saying that GATT, though it has rules for entry, does not have rules for exit from that Union; such a lacuna in the global trade régime would be damaging and destabilizing. The better view is that the post-union exemption applies.

About the author

Dr Thomas D. Grant