You’re familiar with the story about the horse designed by a committee? Well horse races designed by committee, too, are prone to be clumsy and bizarre. After three months of consultations on the next Director-General of WTO, during which a tribunal ran a straw poll disguised as a consensus procedure, we now face a consensus decision (finally) on whether to accept their recommendation. There is a non-trivial risk that there’ll be no consensus on this recommendation. Let’s hope this doesn’t happen, however, because the fall back procedure designed by the race committee can’t work. The contest for the position of the Director-General of WTO is not over: there’s a decision to be made at the end of May in the General Council in which it is still possible that no consensus will emerge around the recommendation of the Council Chair that members appoint Mr Pascal Lamy. I hope not. Mr Lamy is an excellent candidate as was Ambassador Carlos Perez del Castillo. If the result of the process—the recommendation of Mr Lamy—is accepted by WTO Members, then that’s all that matters and we should get on with the job. But the “process(download a copy of the procedures)”:http://docsonline.wto.org/imrd/directdoc.asp?DDFDocuments/t/WT/L/509.doc does not give me confidence that there will be consensus around Mr Lamy, although that is what it was designed to do. It was created by the WTO members to avoid a repeat of the extended public wrangle that occurred four years ago over the (eventual compromise) appointment of Mike Moore and Supachai Panichpakdi. The idea of the “race committee” was that private consultations by three senior Ambassadors with the representatives of all WTO members would allow them to eliminate on a progressive basis those candidates “least likely to attract consensus”. The process would end when the tribunal was left with one candidate to be recommended to the General Council for a—presumably foregone consensus—decision to appoint. In fact, a developing country champion in this race, Mauritian Trade Minister Cuttaree—who withdrew after the penultimate consultation round last month—has already hinted at the potential for failure of this consensus in the final step: bq. “The selection process leaves a bitter taste in the mouth,” Cuttaree told the AP by telephone from Port Louis, Mauritius. “I think we have to look again at this process. “It is very unfortunate that the developing world will not be able to control this organization, although maybe they are to blame because they make up two-thirds of its membership.” (“Forbes.com”:http://www.forbes.com/home/feeds/ap/2005/05/13/ap2026586.html) Cuttaree is referring here to the outstanding failure of the developing country majority (some 120 of the 148 members) to secure the Director-General’s position: only once in sixty years of the GATT/WTO has a developing country candidate held the job (the current DG, Dr Supachai). In the past ten years the industrial countries have filled the job three out of four times # Peter Sutherland: UK, the last DG of GATT who filled the job on a temporary basis for three months in 1995 pending the resolution of a struggle over the appointment of
# Renato Ruggiero: Italy, who held the job until April 1999 when there was a five month hiatus while members wrangled behind the scenes and finally produced a ‘double-header’
# Mike Moore: NZ, who presided for the first half of the four-year term
# Supachai Panichpakdi: Thailand, who is now about to take up the Secretary General’s position in UNCTAD You can understand how frustrated the developing country leadership in WTO must feel. Each of them, naturally, considers him/herself a candidate for the top job. Since the formation of WTO when it was obvious that the trans-atlantic deal over sharing the leadership of GATT and the OECD could no longer hold, the industrialized countries have continued to rule the roost with one of their own as the rooster-in-chief. How could this happen? Cuttaree implies that the numbers actually work against the developing world: how can this be? Cuttaree’s final remark may seem obscure but it points to the heart of the problem in the appointment of WTO Directors General: the consensus factor. About consensus A consensus agreement is one which is adopted without objection. That’s not the same as a decision that has unanimous support: there’s no way to be sure, but probably few consensus decisions would be unanimous. You can say only that none of the opponents of a consensus decision were so strongly opposed to it that they wanted to cause a ‘spill’. The WTO attempts to work always by consensus rather than by taking a vote. This is a unique way of working: in other key multilateral agencies, such as the U.N. General Assembly or the Security Council or even in the World Bank and IMF (where the votes are not equally weighted), the Members vote. You’ve probably seen the photos of the Ambassadors around the horse-shoe table of the Security Council, with ear-cups on listening to the interpretation, raising their country name-plate high with their right hand. Members avoid voting in WTO, as if religiously, probably because they believe that consensus keeps the peace. That’s an objective that has been high on the agenda of the multilateral trading system since it started in 1947: to avoid ‘trade wars’ becoming real wars. Critics are quick to say: consensus means only that votes are replaced with bullying and influence-peddling behind the scenes to ensure that dissenters ‘toe the line’ or to blackmail the majority by holding agreement hostage. But experience of the system over sixty years shows that neither of these is characteristic. To see the characteristic of consensus you must look at its impact on the content of the decision, not at the ‘voting’ dynamic. There is no objective data about the latter, but the different drafts of a decision to be made by consensus are often available and they provide a characteristic track through the consensus decision-making process. In summary, they show that consensus decision-making is a process of agenda-building, coalition-building, co-option, defection, negotiation, compromise and (since sovereigns are involved) ‘comity’. Nothing is settled, in a typical consensus decision, until everything is settled; there is no decision until everyone is on-board and no-one is ready to ‘jump ship’. The text of a consensus decision is a palimpsest of drafts, square-brackets and (less visible) special meanings. Can a race be won by consensus? No: for the very reason that consensus is about the nature of the decision, not about the number of ‘votes’ (there aren’t any). You can’t compromise and contort, revise and re-submit a competitor in a race. If you like some apect of Lamy and you like parts of Cuttaree you might want to propose a Cuttamy or a Lamaree who would beat the pants off a Perez del Castillo in a consensus decision. But it doesnt’ work like that. So what’s going on? The Chair of the General Council, and of the tribunal conducting the pre-consensus consultations on the D-G selection process, Ambassador Amina Mohamed of Kenya, has given us her understanding of what the tribunal has been directed to do. In her “remarks(download a copy)”:http://www.wto.org/english/thewto_e/dg_e/gc_chair_remarks_13may05_e.doc to a Heads of Delegation meeting yesterday (13 May) she said
… We have been guided strictly by the elements set out in paragraph 17 of the 2002 Procedures, which require us to assess Members’ preferences and the breadth of support for each candidate. Paragraph 17 goes on to say, and I quote, that:“the ultimate aim of the consultation process shall be to identify the candidate around whom consensus can be built. In order to do this, it may be necessary to conduct successive consultations to *identify the candidate or candidates least likely to attract such a consensus*”.Let me also recall paragraph 19 of the 2002 Procedures, which stipulates, and I quote, that:“At the end of the final stage of the consultative process, the Chair, with the support of the facilitators, shall submit the name of the candidate *most likely to attract consensus* and recommend his or her appointment by the General Council.”_Emphasis added_
The Ambassador does not say how the Tribunal made the highlighted decisions (a point Cuttaree makes). But I believe that they can have been nothing more than a straw poll based on a negative voting procedure. Although Ambassador Mohamed says that the question from the Tribunal to each Member was “What are your preferences?”, I believe that it would have made more sense for them to ask (and that therefore they did ask) “What are your negative preferences?” Let’s imagine we’re in the penultimate stage of the consultations when all but the final three candidates have withdrawn from the race as required by the rules, which say that candidates judged ‘least likely to attract consensus’ in each round should withdraw. This isn’t a consensus procedure in which you can change the terms of the question so you must pick between A, B and C. If the Tribunal asks for your positive order of preferences they will get an ambiguous answer. Here’s why. “A and C” you say, meaning that C is your top preference but if A or even B were the choice of the majority you’d go along without objection. Now the Tribunal asks for my preferences. “A and C” I say. I, too prefer C and would go along with A. But, unlike you, I mean that if B got up I’d object (there’s no way on earth I’d tolerate that B character!). These two, apparently “identical” preferences lead to very different consensus outcomes. Suppose, following the same procedure in the next round of consultations (in which you and I express a preference for A), the Tribunal concludes that candidate B is the candidate most likely to attract consensus and puts him up for appointment. Has the elimination process found the candidate most likely to attract consensus? Possibly, but it depends not on the degree of support for B but on the degree of opposition to B. If I’m really opposed to B then there will be no consensus reached at the final decision stage because it takes only one overt ‘no’ to overturn the consensus. The Tribunal members, as experienced WTO Committee chairs, know this risk. So my guess is they asked each Member to nominate whom they’d least favor as a candidate and added up the negative ‘votes’ in this straw-poll to determine which candidate(s) to eliminate in each round. Now you can see where this leads: to the success of the lowest-common-denominator candidate: the candidate with the fewest objections. Unfortunately, that’s the biggest risk of any negative-polling procedure. True consensus decisions can avoid the LCD outcome because the capacity to compromise can avoid a ‘race for the bottom’. A skillful Chair can sometimes extract a quantum-leap from a compromise text that was in none of the contending drafts. It happens. Even so, the residual candidate from a negative straw-poll is not guaranteed of attracting consensus as the “Financial Times”:http://news.ft.com/cms/s/31867894-c329-11d9-abf1-00000e2511c8.html has noticed. And worse still, there’s apparently no room for the ‘fall-back’ vote that the decision procedures say should take place if there is no consensus, because all the other candidacies have been withdrawn in accordance with the rules. Is this a problem in practice? Maybe: the problem lies in Cuttaree’s numbers. Why would the majority developing country group be at a disadvantage under these procedures when, it seems, they should make the book in any open voting procedure? Developing countries are not an homogenous group. They are more diverse than the industrialized countries just because they are a much larger group: more geographically dispersed, bigger income variations etc. This means that there are frequently more developing candidates for any job: in this appointment round they had three (Uruguay, Brazil and Mauritius). But the more candidates the developing country group puts up, the more likely it is to ‘split’ the vote of the group as a whole. In this round, as “Ben Muse reports”:http://benmuse.typepad.com/ben_muse/2005/04/a_brazilian_dip.html, the Brazilian candidate was put forward, according to those near to the Brazilian campaign, in an attempt to block the Perez del Castillo bid in a spirit of what can only be called vengance. It’s hardly guaranteed that an open vote would avoid the risk of this sort of ‘strategic’ behavior among developing country members (or among the industrialized, for that matter). But a “consensus” procedure that cannot be a consensus at all certainly increases the risks of a split or skewed vote. Although there is no evidence that bullying, bribery or blackmail characterize true consensus decision-making, there is an obviously greater potential for that sort of behavior in any iterative voting procedure that is conducted behind closed doors. In contrast, the greatest risk you run in a single vote conducted by open declaration is lobbying. Habemus ducem? The rules adopted for the recent papal election forbade lobbying—presumably on the theory that inspiration is more religious than influence—but even the Pope is elected by a ballot. It’s time for the WTO to accept that what is a good procedure for decisions on trade policy and the trading system is not necessarily good for everything.