What is the Meaning of “Consensus” in International Decision Making?

On 28 March, the negotiators at the Final UN Diplomatic Conference on the Arms Trade Treaty failed to adopt the Arms Trade Treaty (see BBC report here) by consensus.  A few days later the Arms Trade Treaty was adopted by the United Nations General Assembly (by a vote of 154 in favour, three against and 23 abstentions – for a really useful account of the negotiations see the Arms Trade Treaty Legal Blog). At the Diplomatic Conference, the rules of procedure required that the treaty be adopted by “consensus”. However, at the end of that conference, Syria, Iran and North Korea objected to the text. According to the bloggers at the ATT Legal Blog there then ensued a discussion of whether the objection by these three States could stand in the way of the adoption of the treaty by consensus, with some States taking the view that acceptance of the text by the overwhelming majority of States was sufficient to establish consensus despite the expressed opposition of three States. However, the President of the Conference ruled that there was no consensus and that the treaty could not be adopted.

The wranglings about the meaning of “consensus” have a broad importance for decision-making with regard to treaties and in other international conferences. Since the 1970s it has become standard practice in many important diplomatic conferences that decisions are taken, where possible, by consensus. Although this procedural device is ubiquitous, as well as being important for the way in which international law is made, the arguments at the ATT conference lay bare the ambiguity that lies at the heart of this concept. There is no consensus on the meaning of “consensus”. The consensus procedure is an important device for achieving broad based agreement on international treaties. It is therefore important to have some clarity on what it means. Lack of clarity on the procedure might well have an adverse impact on the process by which treaties and other international decisions are reached with the result that the substantive outcomes might be less desirable.

The ATT Legal Blog reports that:

“Mexico said that the overwhelming majority of States were in a position to adopt the treaty text. Mexico suggested to proceed to the adoption of the text as there is no established definition of the term “consensus” in the United Nations. Nigeria supported Mexico. Japan also supported Mexico. Costa Rica then supported Mexico’s statement. Chile then supported Mexico’s statement. Colombia “resolutely” supported Mexico.

But Russia then took the floor to say that not enough had been done to achieve consensus. A “strange thing has happened”, that “we should simply ignore the rules of procedure”. This, it said, is “quite unacceptable” and “a manipulation of consensus”. Russian delegation categorically opposes this. The President then gave his understanding that one State could block consensus and he took it that Russia was blocking consensus. Iran then asked for the floor and said that everyone has to play by the rules. President concluded that there was no consensus in the room.”

It is clear that consensus does not mean unanimous agreement. Not all States have to express affirmative agreement for a text for it to be adopted by consensus on it. Neutrality on a text or even dislike for it will not be regarded as blocking consensus. To block consensus there will be need to be expression of objection to the text. Or is that expressions of objection? The heart of the ambiguity is that it is not clear whether any expression of objection to the adoption of a text is sufficient to block consensus and prevent adoption of the treaty or relevant decision. In particular, will a single objection block consensus? To allow a single objection to block consensus is to allow one State to veto the treaty or decision. At the Cancun Climate Change Conference in 2010, the Mexican Foreign Minister, Patricia Espinosa, President of the Conference (the 16^th Conference of the Parties of the UN Framework Convention on Climate Change) overruled Bolivia’s express objection to the texts of the final decisions and declared consensus “to thunderous applause.” [Rajamani, “The Cancun Climate Agreements: Reading the Text, Subtext and Tea Leaves”, (2011) 60 ICLQ 499, 515]. So at least Mexico is being consistent. One wonders about Japan though. At the Kampala Review Conference of the parties to the Statute of the International Criminal Court, it was Japan that spoke up, in dramatic fashion, just before the amendments on aggression were adopted. Many feared that Japan was going to block consensus but though the Japanese delegate outlined the problems they had with the text he said Japan would not block consensus. Had Japan really been opposed would it have taken the view that one State could not block consensus?

If a single objection will not suffice, then how many objections are required to block consensus? Even then, does the consensus procedure give a veto to only a few States and allow a tiny minority to hold the majority to ransom? To prevent this can the President of a conference simply declare consensus, while denying delegations that wish to object the right to take the floor to express that objection. Lavanya Rajamani notes that:

“There are examples of presiding officers presuming consensus in the face of imminent opposition. It took such a well-judged presumption to reach agreement on the FCCC. The Chair of the Intergovernmental Negotiating Committee, Jean Ripert, gaveled the FCCC through with Member States of the Organization of Petroleum Exporting Countries (OPEC) and Malaysia requesting the floor. The President of the first Conference of Parties, Angela Merkel, gaveled the Berlin Mandate through with member states of OPEC waving their flags. These are instances of imminent but not express opposition.” [Rajamani, “The Cancun Climate Agreements: Reading the Text, Subtext and Tea Leaves”, (2011) 60 ICLQ 499]

To deny States the right to take the floor, in order to prevent them from expressig their opposition to a text, and to create an appearance of consensus seems to be an abuse of office by the presiding officers. The use of this tactic suggest that the officers are themselves aware of a lack of consensus.

Mexico’s view at the ATT conference that there is no established definition of consensus is reminiscent of the view expressed before the UN Conference on the Law of the Sea (UNCLOS) by the President of that Conference: Mr Amerasinghe of Sri Lanka. Amerisinghe argued that UNCLOS should operate by consensus but argued that the consensus procedure should only be contained in a gentleman’s agreement and not in the rules of procedure of the conference. He doubted that one could “possibly write into the rules of procedure a rule to the effect that there should be a consensus, because a consensus is beyond definition; it is impossible of definition.” [cited by Louis Sohn, in “Voting Procedures In United Nations Conferences for  the Codification of International Law” (1975) 69 Am. J. Int’l L. 310 at 333]. Amerasinghe’s views prevailed and UNCLOS became the first major UN conference to seek decision making by consensus. It is perhaps hard to recall how innovative this was at the time. However, the number of writings on the consensus procedure with regard to UNCLOS around the time of the conference is perhaps an indication of how much of a departure from previous practice this was. There were two articles on this in the 1975 American Journal of International Law alone. [the Sohn piece above and Vignes, “Will The Third Conference On The Law Of The Sea Work According To The Consensus Rule? ” (1975) 69 Am. J. Int’l L. 310. Sohn’s AJIL piece was preceded by L. B. Sohn, “United Nations Decision-making, Confrontation of Consensus? (1974) 15 Harv. Int. L. J. 438. AJIL later published Buzan, “Negotiating By Consensus: Developments In Technique At The United Nations Conference On The Law Of The Sea”, (1981) 75 Am. J. Int’l L. 324]

However right from the beginning there has been the fear that decision making by consensus would give vetos to a small minority. For this reason, some States initially objected to attempts to use consensus in UNCLOS, and insisted that in the event of a failure to achieve consensus there should be possibility of decision making by voting. [see discussion in Sohn’s AJIL piece]

The discussion regarding the introduction of the consensus procedure and the insistence that there be a backup procedure of decision making by consensus is instructive in two ways. First of all, that history and the initial opposition to consensus indicates that, at least many, States, were of the view that consensus means that there should be no expressed opposition to the final text. This is the only way in which one can explain the statements that using the consensus procedure should not be regarded as a right to veto the treaty and the insistence that there should be a back-up voting procedure (which ended up being a requirement of a 2/3rds majority). Secondly, UNCLOS demonstrates that if there is a fear the minority being held hostage by the majority, a way out can be provided in advance by having a back up procedure. Indeed, this is what ended up happening with the ATT, as the draft treaty referred to the General Assembly for adoption in that forum.

Although there is a clear downside to adopting the view that the consensus can be blocked by any State expressing objection to a text, there are disadvantages if it were to be established that decisions can be taken by consensus even where objections are expressed. In the first place, it would have to be determined how many objections can be overridden. Just one? Or more? Once a number is established then in reality one is in the terrain of a voting procedure. If one or some objections would not prevent consensus but no particular number of objections not blocking consensus was fixed, a second possible disadvantage would emerge : the possibility that States would not be treated equally. On this scenario, one wonders whether an objection by the United States, Russia or China would be treated the same as that from a smaller country. It is interesting that at the ATT Conference it is only after Russia spoke that the President stated that he understood that one State could block consensus and he took it that Russia was blocking consensus (not Iran, or Syria or North Korea!). Indeed, it should be remembered that it was the larger, more influential States that had originally favoured the consensus procedure at UNCLOS as a means of counteracting the collecting voting power of developing countries.

A third and more important downside to the view that some objections would not block consensus is the effect that this would have on the process of treaty drafting. The very purpose of consensus procedure was to make sure that negotiators carried all States along for as long as possible. The procedure was initiated to make it less likely that the views of particular States would be ignored simply because those States were in a minority or lacked voting power. Consensus means that negotiators need to do their best to ensure that treaty texts reflect the views of the broadest possible group of States. One can only guess that some treaties would have been different if the consensus procedure had not been adopted. If it were clear that the objections of particular States can be overridden, negotiators might have been more tempted not to make as much effort to deal with those objections.

However, the concern that one or very few States might frustrate the will of the international community remains real. As stated above, and as was done at UNCLOS, and at the International Criminal Court’s Kampala Review Conference, perhaps the way forward is to have a back up voting procedure. This would keep consensus as meaning lack of stated objection and presumably reaching agreement by this means would be prized by States as it would signal that the treaty in question was truly global. When that fails though the treaty could still be adopted. It might also be important to identify those treaties where consensus is prized and those were it is might not be.