1. Introduction
The multilateral trading institution that governs contemporary inter-state trade relations, the World Trade Organization, emerged in 1995 out of the 1947 General Agreement on Tariffs and Trade (GATT). The GATT itself was signed as a ‘temporary agreement’ during the negotiations for the International Trade Organization (ITO) and was intended as a part of the ITO (Jackson, Reference Jackson2000). The ITO was planned as the third institution, alongside the International Monetary Fund and the World Bank, of the post-war multilateral economic system. Although 54 states signed the ITO Charter in Havana in March 1948, the US Congress never ratified the Charter, with the Truman Administration eventually abandoning the quest for its approval in 1950. The loss of its primary architect ended the ITO. Because the ITO is a failed organization, the scholarly literature has paid relatively little attention to it, and when it has done so, it has been with the intention of explaining that very failure (Bidwell and Diebold, Reference Bidwell and Diebold1949; Diebold, Reference Diebold1952; Goldstein, Reference Goldstein and Ruggie1993).
But, there is much to be gained from analyzing the ITO Charter itself. Given the ITO was intended as the central multilateral trade organization of the postwar era, the study of the ITO advances our understanding of the origins of the current multilateral trading system. The ITO's provisions capture the intentions of the designers of the institution, the compromises they struck, and the negotiations they underwent with key domestic audiences. In this respect, analysis of the ITO also contributes to the study of multilateral institutional design, which is a prevalent concern in the study of international relations (e.g., Koremenos et al., Reference Koremenos, Lipson and Snidal2001a and Reference Koremenos, Lipson and Snidalb). Further, it helps advance analysis on US behavior in the construction of the post-war order, as the US actors' decision to constrain their power within extensive institutional settings has been deemed a unique historical moment, with repercussions lasting into the current day (e.g., Ikenberry, Reference Ikenberry2000).
Drawing on archival material on inter-state negotiations on the ITO, internal US Department of State documentation, as well as Congressional testimonies alongside scholarly works, this paper focuses on one of the most crucial, but yet under-analyzed dimensions of the ITO – the nature of member state representation in the institution, particularly members' voting power.Footnote 1
The ITO endorsed one-member-one-vote, which is puzzling for a number of different reasons. To begin with, the ‘ITO constitution was far more detailed and wide ranging than was the GATT’ (Goldstein and Gowa, Reference Goldstein and Gowa2002: 157). The institution's relatively generous scope and formalization raises questions about why the USA would agree to voting equality, especially because the distribution of voting rights matters for ‘control’ over the institution (Koremenos et al., Reference Koremenos, Lipson and Snidal2001a). Importantly, the ITO agreement explicitly foresaw the extensive usage of voting, including on: the decisions of the Conference and the Executive Board; the Conference's delegation of ‘any power or duty of the Organization’ to the Executive Board; waivers on an ‘obligation imposed upon a Member by the Charter’; sponsoring of ‘agreements with respect to any matter within the scope of [the] Charter’; and amendments to the Charter as well as possible membership suspensions that could result from lack of adherence to such amendments (Articles 75, 77, 79, 100). Further, the Conference of the ITO, where all members would be present, would make decisions generally by majority voting, creating a danger that the USA ‘could be outvoted on almost any issue’ (Goldstein and Gowa, Reference Goldstein and Gowa2002: 158).Footnote 2 Simply, voting mattered and the consensus decision-making that has governed GATT/WTO decisions was not originally planned. Hence, as the rest of the article will show, the designers of the institution labored over voting, and the importance of voting only increased as the negotiations evolved.
Also making the outcome of voting equality curious, the US Congress – the domestic ratification body for the USA's postwar multilateral agreements – was sensitive to US formal power within the nascent system. Congressional members found the US voting power in the IMF and the World Bank, which preceded the ITO by several years, to be inadequate even though the USA would be the top shareholder with veto power in both institutions (Gardner, Reference Gardner1980). Given this domestic policy framework, how the ITO ended up with symmetric voting deserves analysis.
Further complicating any foregone conclusion about voting equality at the ITO, the USA's main partner in the construction of the post-war system, the United Kingdom (UK), preferred weighted voting. The post-war system rested on the close cooperation of the UK and the USA, with the UK managing to pass a number of its key preferences onto the new order (e.g., Gardner, Reference Gardner1980; Eichengreen, Reference Eichengreen1987; Ikenberry, Reference Ikenberry1992). How the divergent views of their primary partner affected the US officials' thinking is worthy of scrutiny.
Finally, the outcome of voting equality at the ITO becomes even more intriguing when one considers that the USA floated its own proposal for weighted voting during the negotiations only to abandon it later. What role did the weighted voting proposal serve?
The paper finds that the outcome of voting equality can primarily be explained with reference to the US negotiators' belief in such voting serving the intended functions and purposes of the institution better than its alternatives. Here, the negotiators' implicit or explicit references to the Bretton Woods institutions appear as a significant factor in defining their inclination toward voting equality. Unlike these two institutions, the ITO would have been largely charged with making recommendations, which the member states themselves would have to implement. This conceptualization of the ITO limited the size of its bureaucracy, its budget, and its instruments of influence. Thus, many of the functions of the Bretton Woods institutions and the institutional logic those functions produced were absent in the case of the ITO, weakening any rationale for weighted voting. For example, the ITO would have no loans to make.
Despite this preference for voting equality, when the inter-state negotiations threatened US preferences on one of the most important issues for the designers – the resolution of balance of payment (BOP) questions – the negotiators presented a proposal for mirroring the IMF's weighted voting at the ITO on this specific issue. Even in the midst of this proposal, however, the US drafters internally remained wary of weighted voting and reasoned that establishing the IMF's decisions as dominant over the ITO's on this particular matter was a viable alternative to asymmetric voting. And, when the ultimate draft of the ITO established the ITO‒IMF linkage they desired, the US drafters reverted back to their original equal voting proposal.Footnote 3
At the same time, although non-British states did not categorically rule out weighted voting, their general tendency for voting equality appears to have reinforced the US drafters' own inclinations. For example, other states' views were one of the reasons for the US negotiators to find acceptable an IMF‒ITO connection on BOP in place of asymmetric voting. In contrast, both the UK's preference for weighted voting as well as the key domestic audiences', including Congressional members, objections to voting equality appear to be insignificant for the outcome of one-member-one-vote in the ITO. This last point is of particular importance, as voting equality contributed to the domestic unpopularity of the ITO in the USA.
Beyond illustrating institutional design through closely tracing the evolution of negotiations, the paper's findings are relevant for a number of debates. They advance the understanding of the kind of ‘restraint’ US officials exercised in binding themselves to institutional frameworks at the peak of US economic power at the end of World War II (e.g., Goldstein and Gowa, Reference Goldstein and Gowa2002; Ikenberry, Reference Ikenberry2000). While the USA could have plausibly pushed others to either pursue weighted voting or face a US exit during the negotiations, the US designers did not follow such a strategy. The paper's discussions also illuminate the role institutional frameworks play in adjusting the importance of the underlying economic power asymmetries among states. Although the USA was the preeminent economic power in the construction of the entire post-war order, that economic weight manifested itself differently across the three integral institutions, the IMF, the World Bank, and the ITO.Footnote 4 The ITO's unique intended purposes and functions undermined a rationale for the underlying economic asymmetries to be translated onto the institutional setting as asymmetries in formal political power.
The next section provides an overview of the inter-state negotiations on the ITO (Section 2), followed by a section that explains the evolution of those negotiations (Section 3). The conclusion highlights the paper's broad implications for academic and policy analyses (Section 4).
2. Background on the ITO and inter-state negotiations on voting
The ITO Charter was signed in Havana, Cuba, in March 1948, following two years of negotiations among more than 50 states.Footnote 5 The US designers hoped the trade institution would engender economic cooperation and freer trade among states, which would in turn facilitate peace (US Department of State, 1945; Truman, Reference Truman1949). Also, given that the US negotiators feared the significant upsurge US exports experienced during the Second World War was an aberration, they envisioned the ITO would continue to facilitate the openness of foreign markets to US exports (Wilcox, Reference Wilcox1949a). While the origins of the ITO date back to discussions between the US officials and their British counterparts in the early 1940s, the skeleton of the institution emerged with the Proposals for the Expansion of World Trade and Employment at the end of 1945 (herein, ‘the Proposals’; US Department of State, 1945). The Proposals then evolved into the ‘Suggested Charter for an International Trade Organization’ (US Department of State, 1946a). This Suggested Charter formed the basis of inter-state negotiations on the ITO starting in London in 1946 and ending in Havana in 1948. In Havana, two State Department officials –William Clayton, the Under Secretary for Economic Affairs, and Clair Wilcox, who headed the US delegation at the Charter negotiations and was the Director of the Office of International Trade Policy – signed the ITO Charter. An inter-departmental/agency committee, known as the Executive Committee on Economic Foreign Policy (ECEFP) formulated the Charter, with the Department of State officials playing a central role (ECEFP, 1948).
The US drafters proposed voting equality (one-member-one-vote) in the ITO's Conference, which would be composed of all members, in their original drafts for the institution, both in the Proposals and the Suggested Charter. Internal documents preceding the Proposals show that the US officials preferred voting equality in the Conference with special majorities on issues of particular concern to the USA (e.g., ECEFP, 1945a, 1945c). The Proposals indicated, ‘Each member of the Conference should have one vote’, but left it to the Executive Board to decide its own voting procedures (US Department of State, 1945: Sections C, D). In the Suggested Charter, the US officials proposed one-member-one-vote at the ITO's would-be 15 member Executive Board (EB) as well as at the Conference, where all members would be represented (Articles 52, 53).Footnote 6 The Suggested Charter also stated that the Conference would choose the members of the EB, though it did not specify how this selection would occur (Article 55.2). Despite this vagueness, the US drafters did not include a clause for permanent membership on the EB in the Suggested Charter, as they had done in the Proposals (Section D). The Suggested Charter required special majorities (two-thirds) on specific issues, particularly the Conference's decision to waive members' obligations under the Charter (Article 55; more on this below). In this way, the Suggested Charter specified the Proposals' general reference to ‘special voting arrangements’ (US Department of State, 1945: 25).Footnote 7
States' voting power and representation in the ITO were subject to debate in various key inter-state negotiations. The first of these was in November 1946 in London, which was followed by a meeting in Lake Success, New York, in January‒February 1947, and then a meeting in Geneva in May‒June 1947. It was following these meetings that 23 states signed the GATT in October 1947.Footnote 8 The GATT was signed as a component of the ITO, specifically as Chapter IV of the ITO Charter (Jackson, Reference Jackson2000). ‘Much of the GATT was taken verbatim from the draft of the ITO Charter as it stood at the end of the Geneva Conference deliberations on ITO’ (ibid.: 22). Moreover, the ‘government officials who drafted the GATT were, with few exceptions, the very same men who were drafting the text of the ITO Charter’ (Hudec, Reference Hudec and Howse1998: 21). The question of member state representation and voting power in the ITO was, however, not resolved in Geneva and was back on the agenda at the last set of meetings in Havana in November 1947‒March 1948 (see e.g., Bronz, Reference Bronz1949; Irwin et al., Reference Irwin, Mavroidis and Sykes2008; Wilcox, Reference Wilcox1949a).
Three key developments transpired during these various negotiations on member state representation and voting at the ITO. First, at the meeting in London, the USA and the UK tabled different plans for states' voting rights and representation in the new trading institution. While the USA went with its original plan for voting equality, the UK tabled a contrasting plan for voting. According to the UK plan, both the Conference and the EB would have weighted voting (UN, 1946a). Votes, the UK delegation explained, could be weighted by the following factors: external trade, national income, and population (UN, 1947a). The selection of countries to the Board, under this proposal, would also be based on weighted voting. The UK proposal, generally, underlined the importance of reflecting countries' asymmetric economic importance in the institution. One of the ways to achieve differential importance was by giving ‘a number of the more important trading countries … permanent seats on the proposed Executive Board of the Organization’, the proposal noted (UN, 1946a). Interestingly, though, the UK proposal also contained a number of votes distributed equally among members (akin to ‘basic votes’ in the IMF/World Bank).
The second primary development on the voting issue surfaced during Spring 1947, when the US delegates presented a plan for limited weighted voting. The US drafters proposed weighted voting in the ITO on two different issues: (a) where the ITO and the IMF had a ‘common interest’, namely discussions on countries' balance of payments; and (b) when the ITO Conference, comprising all member states, selected the members to the ITO's EB. Weighted voting in the ITO on these two issues would mirror the weighted voting in the IMF (UN, 1947b, c, d). Under this scenario, because of the distribution of voting in the IMF, the USA, the UK, the USSR, China, France, and India, in that order, would have had the highest number of votes and therefore would be Board members. Other states would have to build coalitions to be elected to the Board. Other issues, according to the US plan, would be governed by one-member-one-vote.
The third major development regarding voting happened during the negotiations in Havana, when the US negotiators abandoned their aforementioned plan for weighted voting in December 1947 (New York Times, 1947). The Havana negotiations resulted in the ITO consisting of, as the US drafters had all along planned, a Conference and an EB. Both the Conference and the Board would make decisions through equal voting. The Conference would guide the agenda of the Organization and steer its work, including the selection of the Board and the delegation of certain tasks to it. In this regard, the Conference held the ‘final authority’. The Board would consist of 18 member states, eight of which would be countries of chief economic importance, and the Conference would select the rest of the Board with due attention to regional representation. The ITO Charter indicated the members would consider ‘shares in international trade’ in assessing countries' economic importance (UN, 1948).
3. Explanations
What explains the evolution and the outcome of states' negotiations on members' voting rights at the ITO? Specifically, why did the US designers prefer voting equality initially, but later tabled a proposal for weighted voting, only to return to voting equality eventually?
The USA's original preference for voting equality
The evidence suggests that the US designers originally preferred formal voting equality at the ITO, especially at its Conference, because they thought such formal equality would best serve the intended functions and purposes of the institution, and these functions were not merely geared toward the actualization of collective goals, but also toward the fulfillment of key US interests.
The US drafters conceptualized the ITO as a platform for inter-state bargaining on members' trade policies, given the institution would be a ‘centralized agency for the coordination of the work of Members regarding the reduction of trade barriers’ (US Department of State, 1946a: Article 1; also US Department of State, 1945). Toward this end, they reasoned that the institution would facilitate ‘consultation and collaboration among Members' (US Department of State, 1946a). Further, the ITO was envisioned as the centralized information agency on members' trade policies (ibid.: Article 5).
This conceptualization of the institution as a bargaining and consultation platform, in turn, necessitated members' widespread participation, which militated against weighted voting. Particularly, given the nature of the ITO as a forum for ‘continuous consultation’, each member would be expected to reveal information about its trade policies and would have to be engaged in an open discussion with others regarding those policies (Wilcox, Reference Wilcox1949a, Reference Wilcoxb). And, had a member state's willing acceptance of a specific trade policy been lacking during the inter-state negotiations, then that state's enforcement of the policy could not be trusted. As a US negotiator explained:
since most of the acts of the Conference [of the ITO], and for that matter, the executive board, are in the nature of recommendations to governments for action, which has to be implemented by governments, the importance of the recommendation is not merely determined by the weights of the votes that you can collect on an issue, but on the number of countries that are willing to accept the obligations. (US Senate, 1947: 533)Footnote 9
US drafters expressed this opinion also during internal discussions, noting that in an ‘essentially recommendary body’ like the ITO there would not be much to ‘gain by allowing the more heavily weighted Members to vote down the smaller Members’ (ECEFP, 1946c: 4). It was better to grant the members the voice to be heard during the negotiations, so that their inclinations toward compliance with the outcome of the negotiations could be ascertained.Footnote 10 The key rules of international trade embodied in the institution also emphasized the importance of inter-state bargaining. Consider, for instance, the most-favored-nation (MFN) rule, which has survived to date. While MFN generally precluded a member from treating different members of the ITO unequally, the substance of this rule would gain meaning only through inter-state negotiations.
On this point, the ITO's aim was to create a multilateral mechanism for reducing trade barriers. The multilateral nature of the institution would work toward assuaging states' hesitations about lowering trade barriers, since state A's reservations about lowering barriers to imports from State B could, theoretically, be offset by the gains that state A's exporters would make in other members' (States C, D, etc.) economies (Hoekman and Kostecki, Reference Hoekman and Kostecki2001). US drafters of the ITO emphasized this point: ‘If each country could be attracted by the chance of securing new opportunity from all the rest, it might be bolder and more assured in granting opportunity to others’ (Feis, Reference Feis1948). Without each state's participation, the multilateral nature of negotiations on reductions to trade barriers would be undermined, so again the ITO would need to rely on as widespread member participation in the reductions to trade barriers as possible (US Senate, 1947: 543).Footnote 11
Additionally, the US negotiators reasoned that the ITO would witness fluid coalitions and shifting interests, which insured against stable coalitions in opposition to the USA. In Wilcox's words, in the ITO, ‘[t]here [would be] no sharp division of interest that [would] appear in every case. The line-up of votes [would] be constantly shifting, from issue to issue and from time to time’ (Wilcox, Reference Wilcox1949a: 147). Such fluidity stemmed directly from the ITO's standing as a bargaining platform, and shifting coalitions meant that the effects of one-member-one-vote on the USA would be mitigated because there would likely be no permanent coalitions against the USA. In contrast, divisions between creditor and debtor countries were expected to mark the IFIs, with the USA on the creditor end (Gold, Reference Gold1972, Reference Gold1981). In this light, the US drafters did not see a danger in voting equality.
Similarly, according to the designers, the manner in which commodity councils, which would be formed to oversee trade in commodities, would be governed did not necessitate weighted voting. While each member of a commodity council would have one representative, voice across countries importing and exporting the commodity would be evenly distributed (Article 63). Here, weighted voting did not make sense to US designers, where they noted ‘the proportionate share of a member in world trade as a whole [which would be the basis for weighted voting] would not be an appropriate criterion in these cases’ (ECEFP, 1947b: 2). Instead, equal voice across importers and exporters ensured the protection of US interests at these councils.
Even more, on certain issues, where the USA predicted stable groupings against its views, particularly on ‘restrictive business practices’, the US drafters thought weighted voting would be to the detriment of the US interests. In this case, others would be able to obtain ‘a large number of votes … and be in a better position to outvote the US’, so equal voting was relatively more desirable (US Department of State, 1947c: 9; also United States Delegation, 1947a).Footnote 12 In short, the designers reasoned that unfavorable coalitions against US interests could not be properly dealt with through weighted voting, either because weighting the votes would hurt the USA or because there was a more appropriate solution.
The nature of member states' financial contributions, which can be seen as a by-product of the institution's intended functions and purposes, also played a role in the US designers' thinking in affirming the choice of voting equality.Footnote 13 The US financial commitment to the ITO would have been higher than those of other members given contributions would be based on a pre-negotiated ability to pay. Yet, one of the leading State Department officials emphasized, the USA was not placing ‘any capital’ in the organization, which in turn undermined a rationale for weighted voting. In this respect, the ITO's reliance on member finances for its operations would have been much lower than the Bretton Woods (BW)’ institutions. The official explained, ‘in [the IMF] we have a weighted voting position because we put in capital there’ (US Senate, 1947: 37). The US contributors were concerned that if there were to be weighted voting at the ITO, an institutional logic was needed for it. For instance, William Clayton remarked:
As other countries get into production, our percentage [of international trade] will, of course, go down. It is a thing that would not be static like, for example, our contributions to the International Bank and Fund. They are fixed there where we have a right to demand, and we did demand and we got a weighted voting provision. (US Senate, 1947: 37)
A contemporary view of the USA's contributions to the BW institutions as ‘static’ is not meaningful. But, at the time of the creation of these institutions, many were concerned about increasing original levels of commitment, preferring to see them as more or less fixed (US House of Representatives, 1945). Clayton primarily suggests the capital contributions in the financial institutions created an institutional logic for weighted voting, and this logic, not the underlying power differentials, constituted the reason for asymmetric voting. Yet, such logic would have been lacking in the ITO.
Other state representatives' views
While other states' preferences and attitudes did not categorically rule out weighted voting in the ITO, they likely reinforced the US inclination toward voting equality. Similar to US drafters, other countries' officials reasoned that the contribution of weighted voting to inter-state negotiations on the reduction of trade barriers remained unclear. The Benelux countries emphasized ‘it is desirable to find a solution which will give satisfaction to all countries represented, for the success of the ITO depends largely on the extent to which various member states will feel that their opinions have been taken into consideration and their interests safeguarded’ (UN, 1946b). Thus, the delegation for the trio backed one-member-one-vote in the Conference. Similarly, the Chinese delegation noted: ‘The general application of weighted voting to all provisions of the Charter would … not only be unfair to Members in the early stages of industrialization but would also be unrealistic’ (UN, 1947e). And, some of those against weighted voting argued the ‘successful functioning of the Conference would depend in large measure upon a feeling of equality’ (UN, 1946c).
There is, however, not enough evidence to suggest that these objections formed the basis of the US tendency for voting equality. Particularly, other states' protests of weighted voting followed the UK's proposal for extensive weighted voting during the first meeting in London, and the USA had already expressed a preference for voting equality prior to that meeting. And, as the below discussions will reveal, the US tabled its own weighted voting proposal after having heard others' views on asymmetric voting. Moreover, other states objected clearly to the indiscriminate application of weighted voting in the institution. The aforementioned representative quotations from the Benelux and Chinese representatives challenged the wholesale usage of weighted voting throughout the institution (in the spirit of what the UK had proposed). As the London draft on the negotiations noted, ‘several [who supported voting equality] expressed willingness to consider alternatives' (UN, 1946c).Footnote 14 Indeed, the Chinese and the Canadians at some point presented their own proposals for weighted voting, and the Brazilian delegates specifically asked their US counterparts to pursue weighted voting (Clayton and Wilcox, Reference Clayton and Wilcox1947; US Department of State, 1947b). The US drafters could have, thus, plausibly pushed for a combination of basic votes (equal votes) and some weighted voting, had they considered the latter important at the final stage. To sum, while other states' objections against weighted voting could have reinforced the US designers' tendency for voting equality, they do less well in explaining the origins of that inclination.
Informal power?
Could the US designers' confidence in the USA's projected informal power in the new era have affected their decision on voting equality at the ITO? Although the US negotiators did count on some level of informal power in the institution, it is difficult to suggest that this foreseen informal power explains the outcome. As one US negotiator put it:
I think it is fair to anticipate on the basis of experience in other international organizations, that even if the formula is one country or one member, one vote, there is no such thing as equality among votes; and the announcement that the United States, for example, intends to vote a certain way on a motion is far more important than the announcement that some small country intends to vote that way. (US Senate, 1947: 548; see also ECEFP, 1947c)
Here the US negotiator makes the reasonable point that voting equality does not mean equality of states in practice. Yet, not much more can be read into this quote, which merely suggests that informal power would assure US interests would be protected under voting equality instead of explaining the US preference for such equality. The US negotiators were not confident of a high level of US informal power well into the post-war era, assuming reasonably that they saw informal power as being underwritten by relative economic importance (e.g., Stone, Reference Stone2011). For instance, William Clayton reasoned the USA would face decreasing relative economic importance as the post-war era matured: ‘While perhaps we have the large proportion of international trade today, that may not continue, of course. As other countries get into production, our percentage will, of course, go down’ (US Senate, 1947: 37). And, as noted earlier, any confidence in expected informal power cannot explain the choice of institutional design, since that confidence would be consonant both with an institution that included weighted voting (as in the IMF and the World Bank) and one that did not (as in the ITO). Originally only US dollars were used in IMF loans, and it was expected that the US credit and the backing of the US financial markets would bankroll the Bretton Woods institutions in their initial years, which indeed materialized (Gold, Reference Gold1981). Given the US dominance in the post-war financial system (let alone other factors, such as these institutions' location), the US negotiators would have been assured of US informal influence in the two multilateral financial institutions, but they nonetheless pursued weighted voting.
Implications so far
These discussions suggest that US designers sought institutional logic for their proposals on member state representation at the ITO. They argued that because the member states would have to self-enforce the recommendations of the ITO, as much open participation as possible was needed. And, the lack of capital contributions to it did not justify weighted voting. It appears that comparisons to the already created Bretton Woods institutions seem to have helped the designers clarify the functions and purposes of the ITO. The US designers made explicit references to the two multilateral financial institutions, and not to the United Nations, given they regarded the ITO as the indispensable third institution in the multilateral economic system. From this angle, one could also suggest that the US designers were chiefly motivated by functional considerations and not normative concerns for sovereign equality in choosing voting equality. In addition to the absence of such normative discussions, lack of comparisons to the UN is telling, as the UN General Assembly symbolized the sovereign equality of states.Footnote 15
The preceding discussions suggest that the US drafters were keen on voting equality, but they do not address why the designers floated a proposal for weighted voting. The next sub-section turns to that question.
The weighted voting proposal as insurance policy
Archival documentation suggests that the US delegates presented the idea of weighted voting ‘to obtain some greater degree of protection for United States interests in the Charter’, as the substantive negotiations evolved away from US preferences (ECEFP, 1947c: 1). The following quote captures the evolution of internal US discussions:
The United States, in its Suggested Charter, adopted the principle of ‘one country-one vote’ … The United States proposal received the approval of a strong majority of the delegates at the London Meeting. However, the British Empire delegations suggested that weighted voting should be given further consideration … [later, however,] The United Kingdom delegate indicated informally that the United Kingdom would probably not press for weighted voting [in Geneva] … On the other hand, the view has been expressed that the United States delegation in Geneva should seek to obtain some greater degree of protection for the United States interests in the Charter, either by weighted voting or by some other similar provision. (ibid: 1, emphasis added)
The internal communication among State Department officials points to the balance of payment (BOP) issue and its connection to quantitative restrictions on trade as the most critical aspect of the Charter where US interests would need to be protected (ECEFP, 1947b: 3, 1947d). The US negotiators regarded exchange rate restrictions, which were under the IMF's authority, and quantitative restrictions on trade, which would have been under the purview of the ITO, as two sides of the same coin in countries' efforts to deal with BOP problems (e.g. ECEFP, 1944; UN, 1947d; ECEFP, 1947a: Appendix III). Through the ITO, the US designers wanted to circumscribe others' ability to use quantitative restrictions, while having the members also observe the exchange rate rules the IMF set. The drafters, thus, foresaw that all members of the ITO would be members of the IMF, and failing that would sign ‘special exchange agreements with the organization’ (US Department of State, 1945; US Senate, 1947).
Initial vision
In their initial plans, the US drafters pursued the protection of US interests on BOP in a number of ways. In the Proposals, the designers stipulated that the members could impose trade restrictions to correct BOP problems for a ‘transitional period’, but these restrictions ‘in no event would be more restrictive of such trade than the principles applicable under Article XIV of the [IMF]’ (Section C2). Further, the Proposals envisioned that the transitional period would be determined by a ‘procedure analogous to Article XIV’ of the IMF (Section F). Similarly, Article 22 of the Suggested Charter on quantitative restrictions emphasized that these restrictions would conform to Article VII of the IMF's Articles of Agreement, which outlined the possible exchange rate restrictions in the case of a currency being deemed scarce. Interestingly, the Proposals and the Suggested Charter did not explicitly recognize the IMF's authority over the ITO's and merely emphasized the need for congruence across the two institutions as well as overlapping membership.Footnote 16
Change through negotiations
As the negotiations evolved, the US compromises especially during the London conference, directly or indirectly, affected the US designers' thinking on the BOP issue (Irwin et al., Reference Irwin, Mavroidis and Sykes2008; see also Odell and Eichengreen, Reference Odell, Eichengreen and Krueger1998). For instance, the US negotiators conceded to the imposition of trade restrictions that were aligned with the domestic provision of full employment (Wilcox, Reference Wilcox1949a: 536). Moreover, yielding to requests from developing countries, the London negotiations produced a draft with a section on economic development (Wilcox, Reference Wilcox1949a). As William Clayton noted, this chapter ‘was added because a number of underdeveloped countries felt that provisions for dealing explicitly with this subject are a necessary and proper part of an International Trade Charter’ (US Senate, 1947: 7). These changes permitted the possibility of trade restrictions beyond the US intentions. Further, the London Draft explicitly foresaw ‘greater flexibility in the exceptions permitted’ on the question of the discriminatory application of quantitative restrictions (Wilcox, Reference Wilcox1947a: 534). Particularly, it specified the period for countries to transition out of quantitative restrictions as 1951, whereas in the Suggested Charter it was 1949 (Wilcox, Reference Wilcox1947a; US Department of State, 1946a: Article 20).
By the time the US presented its weighted voting proposal (on BOP questions mirroring the IMF's voting and on selection to the Executive Board), key US officials had internally concluded that in the original draft (the Suggested Charter), the ‘obligations imposed upon Members were explicit and unambiguous and less discretion was left to the organization’, and in that context voting equality was appropriate (Paper on Organizational Chapters, 1947: 6). Yet, subsequent drafts foresaw more flexibility and ‘[t]his increased discretion … has increased the importance of the vote … Accordingly, the US has given more serious consideration to the idea of a weighted vote than was felt to be required at first’ (ECEFP, 1947e: 5‒6). Internal discussions emphasized the possibility of weighted voting where the ITO intersected with the IMF:
Section C of Chapter V on Quotas and Exchange Restrictions is a case where important interests of the United States might, under the present draft, be voted down as a result of equal voting…Moreover, there is a logical case for weighted voting with respect to this one issue, because of the existence of the International Monetary Fund. Quotas and exchange restrictions are two aspects of the same problem…It is believed that in view of the logic of the situation, the United States delegation may be able to persuade other delegations to accept this view. (ECEFP, 1947c: 2, 3)
In this context, Wilcox also specifically emphasized that the ITO‒IMF linkage was the proper counterweight to the London Draft's relaxation of quantitative restrictions (Wilcox, Reference Wilcox1947a; US Department of State, 1946a: Article 20).
Despite seeing an important interest in and finding an institutional logic for weighted voting, the US designers were not wedded to it. In fact, they thought two paths were possible: ‘Chapter V calling for action by the ITO in consultation with the Fund to be amended as to provide that the Fund's view in the matter shall be controlling; or … under which voting with respect to such decisions in the ITO shall be in accordance with the Fund's voting formula’ (ECEFP, 1947c: 2, emphasis added). The former path (of no weighted voting but establishing IMF dominance over the ITO on BOP), the designers noted, offered a number of advantages: it would keep to the USA's original emphasis on voting equality; it would not make the USA vulnerable to defeating coalitions on issues such as restrictive business practices, as mentioned earlier; and it would be relatively more pleasing to other states (ECEFP, 1947c: 6). Additionally, the US drafters did not want asymmetric voting on the Executive Board and thought that their weighted voting proposal would necessitate weighted voting there also:
The argument in favor of the first alternative (to make the Fund decision controlling) is that it would avoid the necessity of providing for weighted voting in the Executive Board of the ITO. This would be required under the second alternative since decision taken in the Executive Board should be made on the same basis as decisions taken in the Conference. (ECEFP, 1947c: 3)
The logic of closely tying the ITO to the IMF on BOP issues in a way that met expected US interests underwrote the logic for the US weighted voting proposal, even though the negotiators did not regard this proposal as the only possible means to achieve their goals.
Back to the original position
When the Havana Charter ended up placing the safeguards that the USA desired, the US internal discussions, not just the US negotiators' external stance, decisively returned to the original position of no weighted voting (ECEFP, 1947c; US Department of State, 1947b). The Havana Draft, unlike the Suggested Charter, included an entire section on the IMF-ITO inter-relationship (Article 24). Crucially, this section stated clearly that the IMF, where the USA formally dominated through weighted voting, would have the ultimate say on BOP matters:
In all cases in which the [ITO] is called upon to consider or deal with problems concerning monetary reserves, balance of payments or foreign exchange arrangements, the Organization shall consult fully with the Fund. In such consultation, the Organization shall accept all findings of statistical and other facts presented by the Fund relating to foreign exchange, monetary reserves and balance of payments, and shall accept the determination of the Fund whether action by a Member with respect to exchange matters is in accordance with the Articles of Agreement of the [IMF]. (UN, 1948: Article 2 Paragraph 4, Emphasis Added)Footnote 17
By their own account, the US drafters ‘fought for and obtained in Geneva’ the ITO‒IMF linkage they desired and managed to sustain this linkage at the Havana Conference, upgrading the connection between the two from being one of consultation to a ‘provision that says flatly that the ITO must accept the determination’ of the IMF (Wilcox, Reference Wilcox1947b; also Clayton and Wilcox, Reference Clayton and Wilcox1947).Footnote 18 In this regard, on the ‘most important decision that the trade organization would make’, the Americans had settled their concerns (Wilcox, Reference Wilcox1947b: 19).
Additionally, the Havana Charter included two-third majority requirements on issues of key concern to the US. On the key opt-out states could seek from their ITO obligations, which concerned the Conference's ability to waive the obligations imposed on a member, the Havana Charter required a two-thirds majority with more than half the members present, whereas in earlier drafts, a two-thirds majority was necessary only to determine the procedures for the waiver (Article 55 of the Suggested Charter and the London Draft and Article 77 of the Havana Charter; United States Delegation, 1947a: 7).Footnote 19 Further, as noted previously, the Havana Charter's endorsement of the representation of most important economies on the Executive Board met US plans.
Coupled with the strong IMF‒ITO connection in the Havana Charter, these aforementioned clauses convinced the US drafters that ‘under the present draft of the Charter, the interests of the United States will not require additional protection of weighted voting’ (United States Delegation, 1947b: 9; US Department of State, 1947c). An internal document summed up this matter: ‘The US believes that, in light of various requirements for qualified majority votes on crucial issues, and of the composition of the Executive Board, the essential interests of the US will be fully protected without weighted voting’ (US Department of State, 1948: 4; also US Department of State, 1947c).Footnote 20
Summary
Overall, this analysis offers two different answers to two distinct questions. On the first question of why the USA drafted a weighted voting proposal, the evidence suggests that it was an insurance policy for the Charter further eroding clauses that the USA saw as critical to the protection of its narrow (i.e. non-collective) interests as a surplus nation in the post-war period. Here, the BOP issue appears at the forefront of the negotiators' thinking. The second question is why the US negotiators kept the weighted voting proposal on the table until the eleventh hour, even though they knew an alternative path – establishing the dominance of the IMF over the ITO – would have been equally acceptable to them. The answer is that they acted in a tactical manner. Wilcox noted that ‘the delegation might for tactical reasons want to keep weighted voting in the picture until we saw how we were making out at Habana on the substantive provisions of the Charter’ (ECEFP, 1947e: 8; Wilcox, Reference Wilcox1947b).Footnote 21 Another internal document, similarly, suggested that the weighted voting possibility might ‘dissuade other delegations from attempting to further weaken the Charter’ (US Delegation, 1947a; see also ECEFP, 1947b).
The evidence in the preceding discussions also suggests that the British position did not significantly affect the drafters' calculations. As late as November 1946 (almost a year after the US Proposals), the US officials ‘had not seen any concrete British proposals for weighted voting’ and could not conceive of weighted voting being applied wholesale (US Department of State, 1946b). In addition to this lack of conversation between the two parties on voting, internal US documents indicate that the US drafters specifically disliked the British proposal because they saw it as designed to give disproportionate influence to the British Empire through extensive basic (equal) votes because a number of countries would have individually contributed to the Empire's views (US Department of State, 1947a; ECEFP, 1947e). In fact, the US negotiators preferred no weighted voting to the UK proposal (ECEFP, 1947e; ECEFP, 1947b). In any case, as one of the aforementioned quotations indicated, the US proposal on weighted voting ensued the secret UK admission that they were ready to forsake weighted voting.
Domestic audiences
Moreover, although key domestic audiences expressed a preference against voting equality, this preference does not contribute to understanding the US negotiators' tabling of the weighted voting proposal. During the 1947 Senate Committee on Finance Hearings on the ITO Charter, the voting equality enshrined in the Charter came under critical scrutiny (US Senate, 1947). For instance, the Chairman of the Finance Committee, Republican Senator Millikin of Colorado, indicated that there was a ‘substantial field of power’ in the ITO Charter in which the Organization, and not individual member states, had a say and thus ‘the question of one member one vote [became] very important’ (US Senate, 1947: 533). But, Clayton defended voting equality:
The Chairman [Millikin]: May we assume, Mr. Secretary, that it is a fact that the United States, so far, has been for one vote in the Executive Committee [Board] and in the Conference?
Mr. Clayton: Yes, sir; that is right.
Mr. Chairman: May we assume, for purposes of this inquiry, that that will continue to be a fact?
Mr. Clayton: Yes, sir, Senator. I think we ought to do it that way (ibid: 44).Footnote 22
Millikin was not alone, and other Senators also showed their concern with voting equality.Footnote 23 For instance, Senator Hawkes from New Jersey indicated that:
you are putting the future welfare of billions of dollars worth of industry into this picture, and certainly you and I [Millikin] , as individuals, would not make a deal with some fellow who had one-thousandth of part of what we had, to have an equal vote as to what was going to happen in an organization that we would belong to. (US Senate, 1947: 37)Footnote 24
Still, despite this Congressional skepticism of voting equality, the US negotiators were hesitant to discuss their weighted voting proposal with the Committee, even though it had already been tabled by the time of these hearings (ibid.: 44). For instance, upon Clair Wilcox's mentioning of the ‘possibility’ of weighted voting under questioning from the Committee, Senator Millikin enquired more about it, only to receive vague answers from Wilcox. To Millikin's question of ‘Has the formula [for weighted voting] been evolved?’, Wilcox replied ‘It has been explored, but there is no final position.’ The Senator then pressed, ‘Is it on paper?’, to which Wilcox simply replied, ‘We do have something on paper’ (ibid: 44). If the weighted voting proposal intended to please Congress, the negotiators would be expected to be less coy about discussing it during the hearings.Footnote 25
In addition to Congressional members, key US domestic businesses were also concerned about voting equality. Ultimately, the failure of the ITO Charter was closely related to the lack of domestic support from influential interest groups that represented export-oriented businesses (Diebold, Reference Diebold1952; Bidwell and Diebold, Reference Bidwell and Diebold1949: 231). These groups, such as the International Chamber of Commerce (ICC) and the National Foreign Trade Council (NFTC), were originally and in principle supportive of the ITO. Their support was critical in counteracting the protectionist business groups, which opposed the institution, and in lobbying protectionist Republicans and Democrats in favor of it. Yet, these business groups judged that the ITO Charter ended up tipping the balance between multilateral openness and domestic interventionism to provide for national publics too much in favor of the latter (e.g. Diebold, Reference Diebold1952: 14‒20; Ruggie, Reference Ruggie1982).
The voting rules influenced the way in which these groups interpreted the ITO Charter, especially in the context of other US compromises (see also Diebold, Reference Diebold1952). For instance, the ICC argued that:
In view of the inconsistencies, ambiguities, and differences in philosophies the voting procedure proposed for the ITO would be of utmost importance, for it is within the organization itself that differences in interpretation are to be compromised or resolved. And here is one of the great weaknesses of the [Havana] charter, for it is based on the one-nation-one-vote principle. (US Senate, 1947: 568)Footnote 26
Voting in the ITO was also problematic according to the NFTC. The group was particularly concerned about the rule that a two-thirds majority of the members could ‘waive an obligation imposed upon a member by the Charter’ (Loree, Reference Loree1950: 57). The NFTC reasoned that the waiver could be relatively easily attained, and it would allow states the option of adopting policies that would harm the USA. For instance, (unwise) inflationary policies by governments could facilitate BOP problems, for which states could seek a waiver. In sum, the voting equality enshrined in the ITO magnified the business community's worries about having a Charter in which the USA had compromised, from their perspective, too much. Had the voting distribution been different, it is plausible that the business community could have interpreted all the ‘opt-out's or compromises from free trade in the Charter with less skepticism.
It is difficult to state definitively why the domestic concerns on voting did not make a stronger impression on the US negotiators. A plausible explanation is that the drafters had difficulty juggling the various international and national considerations in the sense that they miscalculated their ability to convince Congressional members of the point that US interests would properly be protected at the ITO.Footnote 27 According to the two-level analysis (Putnam, Reference Putnam1988), negotiators bargain with each other on the international platform (Level I) for an agreement that will be acceptable on the domestic level (Level II). The ‘win-set’ denotes the agreements acceptable on Level II, and the intersection of states' win-sets is necessary for an international agreement. The framework, thus, expects the designers of multilateral institutions to integrate the demands of key domestic groups. The findings here do not weaken the two-level approach, as voting equality in the ITO contributed to domestic skepticism about the institution. They do, however, caution against taking it for granted that international negotiators are able to fully read or address key domestic audiences' relevant concerns.Footnote 28
4. Conclusions
The US drafters' choice of voting equality for the ITO is largely explained by their perception of what would best achieve the institution's intended purposes and functions. Specifically, through the institution, the negotiators not only aimed to create a collective mechanism for addressing trade barriers, but they also wanted to ensure that the USA, as a surplus country, would not be disadvantaged by other countries' restrictive trade and exchange rate policies in the post-war period. As a key example, the negotiators sought to ensure that the balance of payment issues would not be decided against US interests by subjecting ITO decisions on this issue to IMF ruling. As the paper shows, not just the connection between the ITO and the IMF, but also the differences across the Bretton Woods institutions and the ITO guided the US drafters' thinking. In this respect, the paper suggests the importance of the institutional setting in mediating the importance of the underlying distribution of economic power. The ITO's intended functions and purposes subdued the asymmetries in inter-state economic power in that tilting voting power in favor of the economic heavyweights would have detracted from institutional aims.
The paper also offers insights into rational, deliberate design, though a fully fledged discussion is not possible on this point. The findings here suggest that the US negotiators followed the planned functionality of the institution in engendering its voting design. Hence, they were indeed deliberate. Yet, unlike some of the suggestions in the ‘rational design’ literature, the US designers' institutional design choices were more generally conceived visions as opposed to finely calculated intended solutions to differentiated collective action problems, namely distribution, enforcement, uncertainty, and the number of actors (Koremenos et al., Reference Koremenos, Lipson and Snidal2001a, Reference Koremenos, Lipson and Snidalb). Instead of thinking of distinct collective action problems and their corresponding institutional solutions, the designers appear to have pursued broad aims – both of a collective (order and peace through freer trade) and specific (protecting US interests in the post-war world) nature. Their design choices, in turn, were broad-stroked and not fine-brushed in light of these aims as well as the comparative position of the institution vis-à-vis the existing two multilateral economic institutions. Further, the paper's discussions suggest the importance of focusing on the evolution of inter-state negotiations in understanding design outcomes. The USA's tabling of a weighted voting proposal was prompted by the progression of negotiations away from the US wishes on the question of balance of payments. Likewise, the US negotiators' eventual move back to voting equality had to do with the negotiations endorsing the supremacy of the IMF over the ITO on this specific matter and providing the USA with other guarantees, such as special majorities. In this respect, analyzing the process of inter-state negotiations during the crafting of the institution is important – that very process could significantly alter the connection between planned functionality of the institution, on the one hand, and the design outcome, on the other (see also Thompson, Reference Thompson2010).
This study also offers a good illustration of how the US drafters specifically showed ‘restraint’ in the drafting of the post-war multilateral order. Ikenberry (Reference Ikenberry2000) has argued that in order to placate weaker states' concerns for ‘abandonment’ and ‘domination’, the US leaders agreed to bind themselves to rules-based institutional frameworks at the dusk of war.Footnote 29 From a different angle, Goldstein and Gowa (Reference Goldstein and Gowa2002) have identified restraint in both the US willingness to pursue an institutionalized post-war trading regime and the US compromises within that regime. Such restraint was necessary for the USA to make its ‘commitment to free trade credible’, as the USA's economic preeminence placed it in a position (at least in the eyes of others) to easily renege on its promises (ibid.: 154).Footnote 30 In showing that the US designers sought institutional logic for their proposals on voting (both the original equal voting and the subsequent weighted voting suggestions), the paper gives credence to arguments about restraint. For instance, the US designers thought the ITO did not offer a clear institutional logic for weighted voting – there would not be capital subscriptions. And, when they made their weighted voting proposal, they ensured to limit weighted voting to instances where there could be an institutional justification – they considered trade quotas and exchange rate restrictions as two sides of the same coin, which made it plausible to potentially seek the IMF's weighted voting at the ITO on BOP discussions. Given other states did not rule out some role for weighted voting, the US designers could have sought to persuade them of the necessity for some weighted voting or face the prospect of a US exit. Yet, they did not seek such persuasion.Footnote 31 Rather, they sketched out an alternative to the mirroring of IMF weighted voting on BOP issues at the ITO – IMF decisions on BOP superseding those of the ITO's. This behavior suggests that the drafters were not only sensitive to there being an institutional rationale for their design proposals, but they also anticipated others' views in that they knew voting equality would have greater appeal.
The paper's historical analysis raises some relevant current questions. While today both the legal and the economic basis for linking the IMF and the WTO on questions such as putative under-valuation of currencies is unclear despite growing calls for such a linkage (Staiger and Sykes, Reference Staiger and Sykes2010), the paper's evidence suggests that the designers indeed intended for a strong connection between the two institutions. Another area of contemporary interest concerns reform of the WTO's decision-making system – for instance, some authors have suggested the establishment of an EB for the WTO, while others have called for weighted voting in the institution (see discussions in Pauwelyn, Reference Pauwelyn2005). The discussions here reveal the kind of objections/endorsements that might arise for these arrangements as well as offer alternatives – for example, members might look more favorably upon an EB chosen by all member states of the WTO with due attention to regional representation, economic importance, and levels of development. Would such a Board provide a mechanism for ensuring that the voices of different constituencies within the WTO are heard and overcome the kind of deadlocks that plague multilateral trade negotiations at the moment? It seems that the drafters of the ITO thought so, given that they envisioned that the Board would execute the Conference's work. Whether such reforms can address the WTO's current legitimacy problems remains open to debate (e.g., Esty, Reference Esty2002). Regardless, the questions that occupied the drafters of the ITO, ranging from the appropriateness of voting equality to the inter-connections across the three multilateral economic institutions, turned out to be pressing problems of our time.
