An overview of the WTO principle of MFN and bilateral/regional agreements: How does it work?

I’ve been getting some questions lately both at work and through mail from people curious about just what the relationship is in the WTO between the multilateral rule of Most Favoured Nation and the “further deepening” of preferences that is evident in the plethora of preferential trade agreements, whether bilateral or regional. “How can this go together with MFN in the WTO?” I am asked. “Shouldn’t MFN mean that if they liberalize something in a regional or bilateral deal that everyone else in the WTO should get the same since they are members too? What am I missing here?” Is another commonly heard phrasing.

I’ll do an overview of the main issues here, because these are truly important things to understand and examine, not just in Trade & Environment, but for geopolitics at large.

Note before I get into it 1: Please understand that there is enough material here to write a book or three and that by necessity I’ll have to keep it breezy. Hopefully this small overview will iron out most of the general confusion.

Note before I get into it 2: This is highly relevant to Trade & Environment issues. I have previously noted that we seem to be moving towards a “messier” landscape of global governance where alternate pathways forward will be explored while multilateralism remains stalled. Understanding the general gist of how bilateral and regional deals interact with current multilateral WTO rules is obviously very important in such a landscape (hence the recent questions from people not usually interested in the WTO I suppose).

Note before I get into it 3: Loose use of terminology confuses a lot of people in this area. In WTO Article XXIV, which regulates exceptions to MFN through preferential agreements, there are only two things mentioned: Free Trade Agreements (FTA:s) and Customs Unions (CU:s). Thus, the key thing to get in terminology is that the vast majority of agreements that people talk about when they go on about “bilaterals” or “regional agreements” or “preferential trade agreements” etc etc are strictly speaking to sort under FTA:s in the WTO terminology. Customs Unions are relatively rare by comparison.

First things first – Yes, it is allowed and possible to veer from the principle of MFN in the WTO through FTA:s (whether they be bilateral or regional agreements) due to Article XXIV in the GATT. There is also a special “enabling clause” adopted by the WTO in 1979 that particularly and even more favourably allows such exceptions between developing and developed countries and between two or more developing countries.

Great. So WTO law does technically leave room for exceptions to MFN, and countries lately use these exceptions a lot to create preferential trade situations and “play favourites” through these agreements. So what is all the fuss about? Let me do my best at explaining.

The role of preferential trade agreements have been debated since the very formation of the GATT itself and the history of this is fascinating. None other than John Maynard Keynes himself was at first against the rule of MFN, arguing with passion for keeping the freedom to impose discrimination as a country saw fit i.e. “preferentialism”. Of course, he disliked the word discrimination and phrased it rather as freedom of choice for countries to conduct their own trade policy rather than being universally bound by MFN (and arguably thinking in no small part about the welfare of the fraying British Empire).

Displaying some impressive intellectual flexibility however, he later on came to embrace the MFN criteria fully and instead became a vehement opponent to the preferential “freedom” that he had previously espoused, fearing instead that it would lead to just the type of rampant protectionism and dangerous regional blocs mentality that had been at cause for the second world war.

Thus, Keynes ultimately came to hold the view that lately has become pretty much unanimous among trade economists: MFN good, preferential possibilities bad because they ultimately lead to beggar-thy-neighbor policies between nations by opening up the possibility of “playing favourites”. As we all know, when someone is the favourite, there are others who aren’t, and this leads to bad political relations. Add a couple of other negative factors that hollows out trust between nations and before you know it, you have world war two.

Historically, it’s really interesting to note however that this unanimous embrace of MFN didn’t become prevalent until pretty late in the game. When Free Trade Agreements (FTA:s) were still relatively rare in the world trading system, a lot of people did not see much harm in them. Tipping my hat to Jagdish Bhagwati of Columbia, it could perhaps be fair to say that he saw where this “death by a thousand cuts” was taking us, giving some early warnings about the dangers of setting upon the path of undermining the principle of MFN through exceptions.

Wait, you may say, if this possibility for “preferentialism” was recognized as bad by both Keynes and as of late by everybody else, then why was Article XXIV instituted in the first place? Surely there must be lots of explanation in its origin containing rigorous logic and very carefully laid out reasons?


The origins of this main MFN exception provision of GATT Article XXIV is shrouded in mystery. Bhagwati posits that it was probably envisioned as being used extremely rarely. If that is the case, paying mind to the hundreds of FTA:s we now have and counting, that rationale was mistaken.

If one looks at the text of Article XXIV, it actually does seem to put up quite harsh requirements. The definition in Article XXIV says that duties and regulations on substantially all the trade between the two trading partners should be removed for an agreement to qualify as a Free Trade Area. Also, these agreements cannot mean that the “duties and other regulations of commerce” are higher to third countries than they were before the FTA was made (i.e. leaving the door arguably quite open for huge trade diversion while focusing more on the technical similarity of treatment as before the FTA).

But wait, you say, surely a lot of the regional and bilateral trade agreements that we are seeing can’t qualify for such a high treshhold? What exactly does “substantially all the trade” mean? From a textual interpretation, such a critical view is perfectly understandable, as it does seem like a very high treshhold to fulfill to qualify for MFN exception. In 1994 there was an interpretative aid to article XXIV adopted that tried to give some clarity on this mysterious Article. Maddeningly, the interpretation of Article XXIV (8) where the “substantially all the trade” requirement is located, is left out however. Why? Perhaps due to a mix of disagreement, confusion and an unwillingness to rock the boat when everybody is “doing the preferential”.

The WTO Appellate Body gave a bit of guidance in the case Turkey – Textiles where they essentially said, speaking of customs unions, that the meaning of “substantially all” was not agreed upon by WTO members, but that it “accords some flexibility” (to customs union members). They also said that it did not mean “all the trade” and it definitely meant more than “some of the trade”. And that was about as much clarity we received from the Appellate Body in this case regarding the meaning.

With such confusion mixed with widespread use, it’s no wonder that the late Gardner Patterson, a former Deputy Director General of the WTO remarked:

Of all the GATT articles, this is [Article XXIV] one of the most abused, and those abuses are among the least noted. Unfortunately, therefore, those framing any new [free trade area] need have little fear that they will be embarrassed by some GATT body finding them in violation of their international obligations and commitments and recommending that they abandon or alter what they are about to do.”

This statement succinctly summarizes the situation, although the “least noted” part is probably not true today, now that people are getting worried about just where this will all lead.

There is no consensus rule in the WTO with regards to making Free Trade Agreements for WTO Members. The only requirement is that they (when using Article XXIV to avoid MFN) have to be notified to the other WTO members. Now, technically, there is a requirement that these notifications are to be examined in the Committee on Regional Trade Agreements (CRTA) with regards to their WTO consistency. However, due to never being able to reach consensus, which is required, these reports are basically never adopted, leaving this to be of marginal relevance in practice.

On a final technical note: Article XXIV is thus widely used to avoid MFN, which is Article I of the GATT. Can Article XXIV be used as a defence in other WTO Agreements than the GATT? The jury is still out on this one with regards to general applicability, but trade professor Michael Trebilcock of University of Toronto posits that Turkey-Textiles gives certain indication that this may be possible if another WTO Agreement “incorporates” Article XXIV of the GATT by referring to it somehow (though not necessarily directly).

So what does this all mean?

It means, in my opinion, that by leaving the door open for “preferentialism” through Article XXIV we did not inoculate the trading system to the “beggar-thy-neighbor” policies that brought the world to the brink of destruction once before.

This was all well and good when there were still relatively few FTA:s in existence, but with the “explosion” of agreements that utilize the exceptions to MFN, we are starting to see some eerie similarities to the “Untrusting Trading Blocs” situation that ultimately ended so poorly in the 1930’s. Particularly as I believe there is a high degree of risk that such trading blocs will tend to be dominated by a single economic hegemon. Proponents for this approach tend to say that its main benefit lie in being able to achieve a deeper degree of economic integration due to having less players at the table. This of course can be true. But it’s important to not stop there when looking at this. Assumptions tend to be made that once these huge regional blocs have been built up through preferential agreements and “playing favourites”, that they will then just naturally sit down and start negotiating and integrating with each other. I believe this to be a Pollyanna version of reality, and am quite unconvinced that it will happen. More generally, I believe it was simply a big mistake to put in exceptions to MFN, because it has made it much easier to avoid the hard questions on global economic governance cooperation.

Yet here we are.

For our purposes here, I hope this basic primer on a deceptively deep area of international trade has been useful.

Print pagePDF page
This entry was posted in Posts. Bookmark the permalink.

2 Responses to An overview of the WTO principle of MFN and bilateral/regional agreements: How does it work?

  1. Srikar says:

    Very well summarized!