IFI governance


Bridging the democratic deficit

Double majority decision making and the IMF

2 February 2007 | Report

By Peter Chowla, Jeffrey Oatham and Claire Wren

Also available in PDF format. A summary version in Spanish is available from Choike: versión en español

Executive summary

In order to improve its legitimacy and effectiveness, the International Monetary Fund (IMF) needs to consider new reform proposals. The current voting system based on unequal weights puts weaker members at considerable disadvantage in decision making. As a result, the IMF faces further loss of confidence and disengagement by member countries. Quota formula recalculation even if accompanied by an increase in basic votes, risks further reducing developing country influence in decision making. Instead, the IMF should implement a double majority voting system that requires the achievement of two separate majorities—one based on one-country one-vote and the other on economically weighted quotas—for any decision to be made. This paper describes this system as a state-weight double majority, reflecting the two components of the suggested approach.

Reforms must empower those sidelined by the existing system and build in important checks to abuse of power. For progress to be made, however, they should not remove the ability of current power holders to look after their interest through traditional voting weights. By accepting the political realities inherent in the relations between countries, incremental reforms can achieve more voice for developing countries without affecting quota allocation.

A state-weight double majority will increase developing country influence in decision making by creating a situation where they can build coalitions with like-minded states. The increased dialogue that this would generate between members and the incentives it will create to build consensus will result in more stable and effective decisions. The paper addresses the issues facing the IMF and the concept of double majorities, as well as offering ways to implement such a system to increase IMF effectiveness and legitimacy. The model we propose for the IMF is equally applicable to other international institutions that have a weighted voting system.

Summary of recommendations:

  1. The IMF should institute a double majority system at both the level of the Board of Governors and the Executive Board. Decisions would require the requisite majorities of both the number of IMF members and their voting weight
  2. The thresholds for decision should be equal for both types of majorities. Simple majority decisions would then require approval by more than 50% of the voting weight and more than 50% of the membership. Super majority decisions would require either 70% or 85% of both the voting eight and membership
  3. At the Executive Board during every decision an Executive Director will cast votes for his constituency based on the sum of the voting weight of the members of the constituency and the sum of the number of members of the constituency
  4. The Executive Board should commit to following this procedure immediately while the necessary amendments to the articles of agreement are approved by the Fund’s membership
  5. This change to the voting structure should accompany other changes to the governance of the IMF, including increased transparency of Executive Board discussions and decisions, reforms to the structure of the board and accountability of Executive Directors, and the implementation of a transparent and merit-based process for the selection of the Managing Director

1. Introduction

Decision making in many international institutions is characterised by stark differences in the power different members hold over votes. The use of weighted voting that provides some member countries with more input into decisions than others causes a serious democratic deficit, as often those most affected by the results have effectively the least say in them. When the distribution of votes reflects economic performance, developed countries maintain more influence on decision making and control the agenda of the institution. This can lead to policies that do not enjoy the support of poorer countries, thereby undermining the legitimacy of the institution’s actions.

The International Monetary Fund (IMF) has initiated a series of reforms that some say will act to dampen these power discrepancies. However, the proposed reforms, if implemented as currently being discussed, will actually prove regressive for developing country interests. Altering the quota formula using the proposals put forward by the US or EU and simultaneously trebling the basic vote would actually weaken the voting shares of developing countries. The IMF is losing its legitimacy as a truly international institution because it fails to provide developing countries with enough control over the decision making process.

Bridging the democratic deficit therefore requires a fundamental shift in the debate over governance reform. Some wish to see weighted voting replaced by a so called Westphalian one-country, one-vote system, but such a sweeping proposition will never gain sufficient support to be implemented. Any reforms which are to be successful need to follow a more incremental approach and consider the political realities. Wealthy countries, which believe their economic strength should be represented in the way decisions are made in institutions governing global economic and monetary stability, are unlikely to allow their power to be strongly diminished.

As the situation in the IMF currently stands, there are three possible outcomes. Firstly, the do-nothing option risks a further loss of confidence and disengagement by a number of member countries. The second option is to reform the institution through the zero-sum game of quota reform, a divisive and potentially regressive process that has unlikely chances of success and is unlikely to achieve a significant change in power relations. Third, is to implement a double majority voting system. Only double majority has the chance of being accepted by middle- and low-income countries wanting more power and wealthier countries not to lose too much power.

This concept of using multiple majorities is not new, and is already used by many decision making bodies to balance the views of different consistencies. We believe that a state-weight double majority system, that requires a majority of states to support decisions alongside the majority of weighted votes, can be used at the IMF to effectively balance the competing interests during the current stage of the institutional reform process. Such a state-weight double majority, balancing Westphalian and economic principles, would effectively enhance the voice of developing countries and improve consensus decision making at the Fund, while maintaining the opportunity for those who hold the majority of weighted votes to express their interests. While a state-weight double majority may not address all the problems challenging international institutions, such as the overarching issues of political economy in international relations, it provides an alternative starting point from which to launch further necessary reforms.

We first explore the need for a fundamental shift in the debate, describe the context at the IMF and outline why the current reform proposals will not achieve their aims. We then discuss the concept of multiple majorities in decision making and provide examples of its use in other institutions. Finally, we explore how a state-weight double majority system, that allows developed countries to maintain their weighted votes while empowering developing countries to have more say in decisions that critically affect them, would work in the IMF, and propose a way to implement it.

2. Background and context at the IMF

  • The IMF is losing its legitimacy by continuing to operate with unequal weighted voting, failing to mitigate financial crises, and not addressing the macroeconomic policy of wealthier members.
  • Current reform proposals — quota formula recalculation using GDP at market exchange rates and an increase in basic votes — would actually harm developing country influence in decision making.

At the time of the IMF’s establishment in 1944, the United States was the world’s most powerful economy and the only country in a position to export capital to the rest of the globe. The formation of the institution was heavily influenced by debates between British economist John Maynard Keynes and US-based Harry Dexter White,1 but the organisational structure of the Fund was essentially decided unilaterally by the US Treasury.2

Despite significant changes in the global economy and in the IMF’s principal clients, notably the creation of many new independent countries through decolonisation and the collapse of the Soviet Union, the power structure within the IMF has remained largely unchanged. Today this has created a situation where the wealthier countries control the institution but don’t borrow from it; middle-income countries have some voice but seek to avoid borrowing from it; and poorer countries have no voice but are forced to borrow from it.

Loss of legitimacy

Continuing to function with a governance system designed during the colonial era prima facia reduces the chance of the institution being perceived as equitable and just. The 47 countries of Sub-Saharan Africa, despite counting for 25% of the Fund’s membership, hold just 5.6% of the vote and two seats on the executive board. On the other hand, the 27 members of the European Union hold 32.1% of the voting power and seven seats on the executive board, not including Switzerland’s chair and votes. The United States has maintained its grip on power at the institution, holding its own seat on the executive board and wielding 16.8% of the voting power. This makes the U.S. the only country that can singularly veto decisions on quota adjustment and changes to the Articles of Agreement. (See Table 1)

The Fund claims to use a process of consensus to come to decisions at the Executive Board in order to protect the rights of minority shareholders.5 This process, however, requires dedication to consensus building by both the Managing Director, as the chair of the Board, and the Executive Directors (EDs) representing the major shareholders. The IMF’s longest-serving board member, Abbas Mirakhor of Iran, describes the decline of consensus building: “Unfortunately there has been a continuous erosion in the strength of this time-honoured tradition in recent years. Significant and crucial decisions have been made with majority votes, that is, though the sheer voting power of a minority of chairs.”6 The undermining of the consensual process for decision making is a crucial factor in developing countries feeling disempowered by the way the institution works.

Table 1: IMF voting weights, economic size and demographic information for selected countries

Country Population 4 Share of global population GDP 5 Share of global GDP Quota Voting Weight
United States 298.2 4.6% 7,948,874 30.35% 371,493 16.78%
United Kingdom 59.7 0.9% 1,314,530 5.02% 107,385 4.86%
China 1,315.8 20.4% 1,190,158 4.54% 80,901 3.66%
Belgium 10.4 0.2% 219,132 0.84% 46,052 2.09%
India 1,103.4 17.1% 415,276 1.59% 41,582 1.89%
Malaysia 25.3 0.4% 75,908 0.29% 14,866 0.68%
Uruguay 3.5 0.1% 8,799 0.03% 3,065 0.15%
Mozambique 19.8 0.3% 3,642 0.01% 1,136 0.06%
Swaziland 1.0 0.1% 1,338 0.01% 507 0.03%

Sources: United Nations, World Bank, IMF

The IMF’s mistakes in handling the financial crises in Asia, Russia and Latin America over the last decade have also caused the Fund to further lose legitimacy. This has prompted anyone who is able to disengage from the IMF to do so. Asian countries have built up large stocks of foreign reserves and have begun regionally pooling reserves under the Chiang Mai Initiative to avoid having to borrow from the IMF.7 Likewise, Latin American countries have been repaying their loans early to divest themselves of the IMF’s conditionality and oversight of their economies and budgets.8 These problems are inextricably linked with the developing countries’ perceptions that the IMF is not responsive to their needs and that they have no input into decision making.

Meanwhile, industrialised countries repeatedly ignore the Fund’s advice, noticeably in addressing current account imbalances. The current governance system and the lack of any effective sanction on the Fund’s creditors means that Fund economic surveillance is ignored where it is most crucial: addressing the international impact of macroeconomic policy in systemically important such as the United States. The lack of implementation of Fund economic prescriptions by developed member counties cannot be tackled in a governance structure where the same developed countries control the decision-making process. The inability of the Fund to take on its powerful members threatens the global public good of a stable international economy.

Reform of the IMF’s governance and organisational structure has been a controversial issue for decades. Fights over quota rights were prominent in the 1970s during the boom of oil revenues for OPEC members and at the end of the 1980s after the phenomenal growth of Japan. Over the last six years the debates over quota have also highlighted the impact of the introduction of a common currency in the eurozone, as intra-eurozone trade is still included in calculated quotas but fluctuations in this trade cannot lead to any imbalances in the current account. While transparency, board structure, management selection and other issues are also pressing, the debate about IMF governance has of late focused on the quota formula.

The disagreement comes despite many previous efforts to improve the quota formula. The Fund now uses five separate formulae to calculate potential quotas, while a process of political horse-trading is required for the final agreement. The last time the Fund reviewed quotas and quota formulas, in the Twelfth General Review of Quotas, the member states could not reach sufficient consensus on a new quota formula or a quota increase, and concluded the review in early 2003 with no changes.9 It is unclear whether the latest efforts to reform the quota formula will be any more successful.

The reforms discussed in Box 1 face difficulties in implementation however. By setting the agenda for quota formula reform early, the major shareholders have succeeded in pre-empting many of the more innovative proposals that could be considered.12 The European Union has been unhelpful by confining itself to arguing over a few details of how the new quota formula should be devised. This situation is symptomatic of the very problem in IMF governance: that countries lacking voice are precluded from participating in the process.

Box 1 – Current reform proposals: quota formula and basic votes

Quota reform: The major Fund shareholders are pushing for a new quota formula to have fewer variables and reflect a country’s weight in the global economy. The G7 propose a new formula that gives preference to GDP at market exchange rates. This would boost the shares of fast-growing Asian economies, encouraging them to maintain their participation with the Fund. Alternatively, the G2410 suggest a formula based on the purchasing power parity (PPP). Using PPP to calculate quotas would provide middle-income countries such as China, India and Brazil with increases in their quotas. However, using PPP would run into measurement difficulties, as PPP factors are not uniform, and would also decrease the quota shares of the IMF’s powerful industrialised countries, something they are unlikely to accept.

Basic vote reform: Each country in the IMF receives a set number of basic votes regardless of their economic size. Originally designed to recognise the equality of states, current basic votes are at the lowest levels in the IMF’s history: 2.1% as opposed to that of 11.3% upon the institution’s founding. Despite encouraging discussions on increasing basic votes to enhance the voice of low income countries, the reality is that even a trebling of basic votes (the most generous proposal on the table) combined with a new quota formula that primarily factors in GDP at market exchange rates would erode, not increase low-income country voting power (see Table 2).11 The motive behind the move to increase basic votes, after decades of complaint about their low levels, is likely to provide the necessary cover for changes to the quota formula that will as a by-product weaken the voting power of low-income countries.

Table 2: Effect of basic vote and quota changes on different categories of countries

Country Group

Status Quo

Trebled basic votes

Trebled basic votes and formula reform combined

2/3 GDP and 1/3 Variability

2/3 GDP and 1/3 Openness

Advanced economies 61.45% 59.61% 71.02% 68.04%

Major advanced economies (G7)





Of which US





Of which UK





Other advanced economies





Developing Countries 30.99% 32.55% 23.99% 26.39%











Middle East, Malta & Turkey





Latin America and the Caribbean





Transition Economies 7.56% 7.84% 4.99% 5.57%

European Union 27 32.09% 31.39% 36.98% 33.51%
Eurozone 22.63% 22.01% 27.07% 24.79%

Source: IMF, authors’ calculations

3. The multiple majority decision requirement

  • Multiple majority is a decision rule that requires the achievement of two separate majorities for a resolution to be accepted.
  • Multiple majority increases the incentive for coalition-building because it facilitates countries to band together and block decisions
  • International institutions, such as the European Union, already use double majority voting systems.

To address the loss of legitimacy in the IMF and other international institutions that use weighted voting, alternative processes for making decisions must be explored. There are two factors in any decision making process: the distribution of votes (how many votes one has) and the decision rule (the rules governing how a decision is made). These factors determine the power different groups have when voting. For example, changing how many votes a member has will change their power in making a decision. Alternatively, changing the rules of voting can also affect the influence a member has in the decision making process. Both factors can be used to ensure that different members, or groups of members, have influence within a decision making process.

In the context of an international institution, altering the distribution of votes among members can be done on, for example, an economic or population basis. In some cases, allocating enough votes to one group may provide them with the power to block a proposal or carry a decision. This is seen with the amount of votes prescribed to the United States in the IMF.

Changing decision rules can also shift influence without altering the distribution of votes. Thus, blocking power can also be assigned with decision rules. Altering power in decision making is usually done to reflect principles of fairness and equality or to ensure a system of checks and balances. The aim is to enable decisions to be made while acknowledging that particular groups may have a greater interest in the outcome, and at the same time not giving any one group an unfair amount of authority.

An example of how distributing votes can affect a particular group’s influence is employee representation in Germany. Legislation requires companies with more than 2,000 employees to have supervisory boards with equal representation of employees and shareholders.20 This uses the make up of the board and the distribution of voting rights to ensure that there is consideration of the views of a company’s workers in the oversight of management and the taking of any major decisions.

As an alternative, the decision rule can also give greater power to a particular group by the requirement of a super majority or unanimity. The U.S. Senate can override a Presidential veto by a two-thirds majority. In an institution where there is often weak party discipline and a slim majority this decision rule ensures that there is the opportunity for a significant minority to protect the President’s veto and makes sure that no one group can ignore the wishes of another. This decision rule also provides incentives for bipartisan coalition building to ensure that there is the sufficient majority to secure a bill’s passage in the face of a potential presidential veto.

A multi-majority requirement is another example of a decision rule that can shift the balance in power from the traditional sole simple majority requirement. Most often seen as double majority, it requires two separate majorities on different criteria for a measure to pass. This gives greater power to an alternative interest and means that a broader coalition is required to pass a resolution.21 Greater power is given to a second constituency without dramatically reducing the power of the first.22

The essence of a multiple majority requirement is that there are multiple potential blocking points, the result of which is that there are a number of potential groupings that have the ability to shape a decision. Voting power analysis shows that a double majority distributes voting power more equally, thus increasing perceptions of legitimacy.23 Power is spread more widely across players, which creates incentives for building broad coalitions to ensure that the decision is not blocked by another group. Through discussions, persuasion, negotiation and vote trading the proponents of a resolution will need to build a broad base of support.

With quota reform, in order for one country to gain, another country must lose the same amount. In contrast, for a double majority system, because there is a second vote with a different power distribution, it is not necessarily the case that gains equal losses. Countries that previously held power would experience a small reduction in power since they would no longer be able to so easily force through decisions not supported by the alternative majority. However, they would still maintain the initial majority of votes and therefore keep their influence over members, due to the power of that majority and the potential to block decisions in the first majority. While not a win-win, it is not a zero sum game, as the gain to one group outweighs the loss to another.

Box 2 – Multiple majority in practice: “triple majority” voting in the EU

In anticipation of the enlargement of the European Union, the Treaty of Nice reduced the number of issues within the Council to be decided on the basis of unanimity. This was to expedite the decision making process, which could frequently have been blocked with an enlarged membership.

During the negotiations of the Nice Treaty there were disagreements about the nature of the qualified majority system to be introduced. Consequently, all three proposals were incorporated creating a “triple majority” system for Council decisions. Thus, Council decisions require a 72% majority of weighted votes, a simple majority of member states and, at a country’s request, a 62% majority of the total EU population.25 As the final criteria is only at a country’s request this is sometimes also considered to be a “double majority plus safety net” system.

Without the additional requirements of a majority of member states, it would be possible for a decision to be passed with a qualified majority of weighted votes without a majority of states supporting the position. This had not been possible under earlier qualified majority voting criteria.26 The demographic requirement gives greater power to Germany who received a disproportionately small number of weighted votes for the qualified majority criterion.27

The introduction of these rules has allowed for a balance to be found between Westphalian one-state-one-vote system while also making allowances for the relative sizes of countries’ populations and the need for a compromise between the interests of the different member states. In adopting the multiple majority voting system the EU was able to overcome a stalemate in negotiations over voting rights.

The significance for developing countries is that their influence rises through the ability to build coalitions with like-minded states. Since a multiple majority requires decisions to have a broad base of support to pass (because two different majorities of various groups must lend their support), coalitions must be developed in order to reach a majority. Developing countries can gain influence in decision making during the building of coalitions by grouping with like minded members.

Building broad coalitions also persuades members to support measures through the potential to trade votes. Groups that would not otherwise have the power to promote an issue, gain the opportunity to trade their support for a resolution in return for support on their own initiative. Therefore, previously disempowered groups would have the potential to more effectively oppose unpopular decisions and to trade votes for more preferred outcomes.

Use of multiple majorities

The logic of a double majority system is institutionalised in most bicameral legislatures throughout the world.24 In bicameral systems proposed legislation must be passed by two different bodies each designed to represent different interests. This is clearly shown in the US Congress where the House of Representatives reflects population and the

Senate the individual states. Bicameral systems are effective when the group is divided by one major conflict, but have several common interests. This holds true for many international organisations where countries are united by the common interest of the organisation—for example a stable global economy, international peace and security or shared access to common resources.

The multiple majority requirement does not have to be institutionalised into two separate bodies as is the case with bicameral legislatures. It can be a criterion established for a single group of voters. This is the case for the EU voting arrangements for the Treaty of Nice (see Box 2).

The Council of the Global Environment Facility (GEF) has a double weighted majority requirement for when there is a failure of consensus. This requires that:

decisions requiring a formal vote in council shall be taken by a double weighted majority, that is, an affirmative vote representing both a 60 percent majority of the total number of Participants and 60 percent majority of the total contributions.28

While there are countries that are both significant funders and participants, in introducing the double majority requirement the GEF avoids overemphasising the financial interest of members. Including the contribution criterion highlights the practical need for funding to be reflected in the voting structure.

The GEF’s formal decision making process is cumbersome, but there has never been a vote under the double majority. However, the existence of the double majority works towards creating consensus as the possibility of the vote under the double majority rule means that all interests are represented as the resolutions are negotiated. Although there could be the concern that such a governance structure would reduce efficiency, according to Cord Jakobeit the double majority requirement has “not hampered the development, effectiveness and performance of the institution”.29

The International Seabed Authority also has a multiple majority voting system whereby the approval of four groups is required for a resolution to pass. These groups (consumers, investors, net exporters and developing countries)30 reflect the various interests in the work of the Authority as the international organisation charged with controlling activities on ocean floors beyond national jurisdictions. As this may one day include mineral extraction there is a significant potential for divisions over regulation. The quadruple majority required should help achieve decisions that are stable with a wide base of support since agreement from all major interests is developed during the decision making process.

A similar system is in operation at the African Development Bank for amendments to the Charter which requires two separate double majorities. Standard voting power is comprised of basic votes assigned to all members and additional paid in shares. However, an amendment to the Charter must be approved first by 66% of all members with 75% of voting power but also 66% of regional members (that is African countries) and 75% of this regional voting power. So there is both a double majority requirement for the full Board, and also a double majority requirement for a subset of the Board ensuring that all interests are represented.31

Efficient and legitimate governance requires providing all groups with the ability to influence decision making. Bicameral institutions facilitate agreement by providing groups with the power to influence decisions in different voting bodies. Since bicameralism is not practical in most international institutions, efficiency and legitimacy can be gained by utilising multiple majorities with different voting distributions. Doing so enables a unicameral body to achieve the stability inherent in bicameralism. In each majority, some groups will have more influence than others, but by requiring both majorities to pass resolutions the groups will have to compromise. Therefore, multiple majorities enable an institution to recognise the importance that all members have in the decision making process, while achieving decisions in a more legitimate way.

4. Using a multiple majority in the IMF

  • The IMF should look at different decision rules as a way to be.er balance the diverse interests of its membership.
  • A state-weight double majority would encourage dialogue between member countries and help build consensus.
  • The Executive Board should use a state-weight double majority voting system where an ED would cast all the shares available from constituent member countries and cast the votes for the number of countries in the constituency.

Given the ability of multiple majority voting systems to better balance the interests of disparate members of an institution who have a common goal, in this case a stable global economy, and strengthen institutional legitimacy, a double majority voting system presents a tenable alternative for the IMF. We put forward here a clear proposal for the use of such a system as an interim solution to the IMF’s problems of legitimacy and effectiveness. It should be noted that this proposal could be adopted for use in other international organisations, particularly those organisations using weighted voting systems, such as the World Bank.

Interestingly, the IMF already uses a form of multiple majority, called a qualified majority rule, which it enshrined in its Articles of Agreement. Article XXVIII states that amendments to the Articles are considered approved “when three-fifths of the members, having eighty-five percent of the total voting power, have accepted the proposed amendment”.32 The idea of using a qualified majority was not alien to those at the Bretton Woods conference. We suggest extending such a rule to all decisions made by the Fund, including Executive Board decisions requiring qualified and super majorities.

Issues with a state-weight double majority 35

Implementing a qualified majority system that balances Westphalian principles of national sovereignty and equality of representation with the current system of economically weighted voting would neatly bridge some of the divides in the membership while avoiding some of the pitfalls with a creditor-debtor split. The rift between developed and developing economies is the biggest stumbling block in Fund legitimacy. The current system essentially works to privilege the interests of the large economies at the expense of the small ones. As the current membership of the Fund is made up of 30 members of the OECD and 154 non-OECD members, assigning each country an equal vote would significantly empower developing countries.

The immediate attraction of a state-weight double majority is that it would make the divisive debate about what variables to use in a revised quota formula less intractable and less important. While the current debate has been unhelpfully limited by the major shareholders and the management of the Fund,36 enhancing the voice of low-income countries can come through empowering each country with an equal vote in one aspect of the institution’s power structure instead of through a different constellation of percentages for various variables.

By combining a one-country one-vote system with a one-share one-vote system, no small interest group or set of countries could push through a measure against other members’ wishes. The G7 as a block would still maintain an effective veto in terms of the economically-weighted voting, but developing countries could muster a veto on issues of concern through the one-country, one-vote system.

While some argue that a double majority proposal would not win sufficient support, we believe this judgement is premature. The dynamics of the double majority are significantly different from a move to reallocate quota more justly. Because reallocating quota is a zero-sum game, quota formula changes are deeply divisive. But adding an additional requirement for agreement would be less threatening to those that already hold power because their opportunity to look after their own interests based on economically determined voting weight, effectively the power to veto, would be maintained.

Once implemented this double majority system would encourage greater dialogue amongst the Fund’s membership and greater consensus building efforts. Proposals would need members to build coalitions to agree solutions, instead of decision making relying on the narrow support of a select group of countries that wield high shares of the vote. Such a system would promote more dialogue as policy decisions would require supporters to win over those with different interests, much as may have been done on contentious issues in the early years of the institution.

Box 3 – Creditor-debtor double majority

A common practice in regional development banks is for the membership to be divided between the lenders and borrowers. This may often be characterised as “regional” versus “non-regional” members.

However, at the IMF instituting a double majority decision system along the creditor-debtor split33 would be difficult to implement because of the revolving nature of Fund resources. For example, in early 1997, few would have considered South Korea to be a potential debtor of the organisation. In the current benevolent global economic climate, nearly every member of the Fund would fall in the creditor category, meaning that it would be hard for such a system to accurately reflect the varied interests of member countries who occupy different positions in the global economy.

Additionally, such a provision would fall foul of the Fund’s long standing principles of universality and uniformity of treatment of members.34 The Fund is required to treat all members equally and only puts its members in categories for analytical purposes such as the tables in its flagship publication, the World Economic Outlook. Crafting subdivisions into Fund membership undermines this practice and might contribute to damaging the consensual character of the Fund’s work.

For example, a decision to change the rate of charge (interest paid by borrowers) or the rate of remuneration (interest paid to creditors) requires a 70% majority, meaning that with the support of a few middle-income countries, the major creditors can push through changes. Over time creditors have increased the rate of remuneration and are now remunerated at market interest rates.37 This has increased the expenses of the Fund, causing it to raise the rate of charge to compensate, thereby increasing the costs to borrowers of using IMF resources.

Under the current system, remuneration rate increases could only be opposed by a coalition of the 48 largest non-OECD countries. Such a blocking coalition necessitates the support of both Saudi Arabia and China, the two largest non-OECD vote holders. If either of those two countries, both of which have large reserves and surpluses and would benefit from an increase in the rate of remuneration, defects, it would not be possible for the remaining 153 non-OECD countries to prevent the decision. Under a double majority system, on the other hand, the proposal would require the support of more than 125 member states, meaning that any coalition of 60 countries could block the move. The double majority system would facilitate the formation of coalitions and prevent industrialised countries from pushing through measures that do not enjoy broad-based support.

Similarly, normal decisions at the Executive Board of the Fund require a simple majority to be approved. Under the current system, non-OECD countries cannot block any proposal, as the OECD countries control nearly 70% of the votes at the board. In fact, if the EDs representing the G7 countries all agree to a proposal; they only need to attract the support of the Belgian ED to achieve a simple majority. These eight EDs, representing 35 countries, can unilaterally make decisions. To oppose normal decisions, developing country EDs would, at a minimum, need the support of either four developed country EDs or the US and one other developed country ED. A state-weight double majority on the other hand would allow the eight developing country EDs that lead constituencies to block a decision. This significantly lowers the barriers to successful coalition building.

The end result of a double majority system should be compromise and cooperation, forcing sufficiently large coalitions to be built to support any proposal. Given the differences over quota reform, implementing a state-weight double majority is the only feasible outcome that would have the desired effect of enhancing the voice of developing countries in decision making. There are two possible drawbacks to such a double majority: no representation based on population and the continued under-representation of fast growing middle-income economies. We discuss these drawbacks in turn, but find them unconvincing as a case against use of the double majority system at the present time.

Such a system still ascribes voting power based on economic weight, which is fundamentally opposed to democratic principles of equality. A truly democratic structure would seek to balance the principles of equality of representation of nations with the equality of people.38 While the double majority is not a perfect solution, it is a move in the correct direction; that is away from having voting rights decided by purely economic criteria. In the long-run consideration should be given to how population can be appropriately incorporated into calculating IMF quotas, but in the meantime, as a step towards the end goal of democratic governance at the quotas, but in the meantime, as a step towards the end goal of democratic structure that fully balances economic criteria with other relevant considerations, would be highly improbable.

The current proposals for reform seek to assuage the concerns of fast-growing economies, particularly those in Asia, over their lack of input into decision making. The double majority system proposed would not give greater voice to those countries. However, implementing the double majority would not preclude quota adjustments to raise the voting share of those countries. Implementing a double majority may take some of the pressure off the bitter debate over quota formula changes, facilitating compromise on representation for those fast growing countries. Additionally, the fast-growing economies typically are elected to the Executive Director chairs for their constituencies, so the implementation of the double majority at the executive board would enhance their ability to influence debates. Finally the double majority system would also help enhance the voice of developing countries as a whole, ensuring that any quota formula changes are well supported by the entire membership and reflect the needs of diverse constituencies.

Implementing a double majority system

One concern is how such a system could work in practice. So far no detailed, concrete proposals have been put forward. Below we present some options and recommendations on the implementation of a double majority system.

There are two levels of decision-making at the IMF, both of which could implement a state-weight double majority system. At the highest level is the Board of Governors, a body comprised of the finance ministers of the IMF’s member nations. In practice, the governors make few decisions, as they meet only twice a year and do not discuss finer points of IMF policy and practice. Implementing a double majority system would be very straight forward, as decisions at this level are accompanied by formal votes usually communicated to the board secretariat via electronic means. The votes could easily be tallied along two criteria: support by member states and support by voting weight.

The remaining issue is at which level the threshold for approval should be set. In practice the Governors’ decisions are either mundane (e.g. approval of accounts and budgets, salaries of EDs and dates of accounts and budgets, salaries of EDs and dates of meetings) or require 85% of voting weight for approval (e.g. quota changes). We recommend that the threshold of approval be equal for weighted and state-based votes as there is no rationale for setting them differently. For controversial issues, this can work in practice, as seen by the resolution on ad hoc quota adjustments approved during the 2006 annual meetings in Singapore. The resolution passed with the support of over 90% of the voting weight and over 87% of the membership. Matching the one-countryone-vote threshold to the threshold for majorities or super majorities of voting weight would be the most logical way to ensure that the interests of the diverse membership of the Fund are met.

The issue becomes more complicated at the level of the Executive Board, where there are two possibilities for implementation of a double majority system; one based on the equality of directors and the other on the equality of member countries. In the first, each ED would have an equal vote in terms of the second majority. If the thresholds for a decision were matched for both of the majorities, a decision that needed a simple majority would require support of directors accounting for more than 50% of the voting rights and would need to carry the support of 13 of the 24 directors to be approved. Similarly decisions requiring 70% of the voting weight would need approval by 17 directors.

Treating each ED equally, however, would not conform to the notion of national equality. Instead, it would serve to continue valuing the opinion of certain members, particularly the five largest economies that automatically hold their own chairs on the board, above those of other countries that are part of constituencies. Abstracting from the question of intra-constituency decision making and assuming equal treatment of constituency members, such a system would accord a country in the Africa group II constituency 1/24th, or 4.16%, of the power given to the United Kingdom. That is even less than the ratio of current voting power between the Democratic Republic of the Congo (the largest member of Africa group II) and the UK, currently at 5.18%.

The second option would be for the Directors to be accorded a number of votes equivalent to the number of members of their constituency. Thus the Executive Director for the Africa II group would cast 1.39% of the weighted votes but 13.19% of the state-based vote tally corresponding to 24 out of the 182 active members of the Fund.39 The five largest countries that hold their own chairs on the board would each count for only 1 vote, 0.55% of the state-based vote tally.

This second system would preserve the equality of nations within the Fund, upholding the principle of parity of members, and do more to redress the unequal voice in decision making. Given that the 30 OECD countries hold up to 13 of the 24 chairs at the Executive Board at any one time, using the former system of weighting each chair’s vote equally would not alleviate the perceptions of inequality in decision making nor enhance the legitimacy of Board decisions. It would also not increase the incentives for dialogue and coalition building much above the current arrangements for board decision making. Only if the large constituencies of developing countries hold more voting power would it counter-balance the marginalisation of holding fewer shares in terms of economically weighted voting. The latter system of maintaining equal weight for each member nation would more effectively generate discourse between different interest groups and compromise.

This scheme would unfortunately not solve the problem of voting in mixed constituencies. Even under this arrangement developed country EDs would wield the state-based votes for their whole constituency, for example in the case of the Canadian ED it means 12 votes or their whole constituency, or 6.5% of the total state-based vote tally. This is clearly a shortcoming, but with the double majority as one aspect of a comprehensive reform proposal, this can be overcome. Constituency realignment, which can be done voluntarily, may help assuage some of the concerns along these lines, with developing countries banding together to hold more chairs on the board or increase the number of state-based votes accorded to developing country EDs.40

Implementing these changes to the voting system of the Board of Governors and Executive Board would require amendments to the Articles of Agreement to be binding. The textual changes would generally be uncomplicated, simply adding the clause “and the membership” after each location in the articles that specifies a super majority, and modifying Article XII Section 5 (c) to indicate that agreement by a majority of the membership is also required for decisions to be made.

However, amendments to the articles can take years,41 and some feel that there may be insufficient support for such a radical change to pass the hurdle of approval required by an amendment. While quota reform can be decided on by the Board of Governors, without resorting to national-level legislative approval, Fund members have already committed themselves to a proposal to amend the Articles of Agreement in order to boost the level of basic votes. Additionally, the idea to make quota self-adjusting based on current economic factors would also require an amendment to the articles, yet this has achieved significant support.

While an amendment is moving through the hurdles of the administrative process, the Executive Board should come to an understanding that its internal decision making processes would reflect the need for a double majority in the interim. The Board seeks to act by consensus and votes are rarely taken. Though we believe the taking and publication of votes is necessary to promote accountability, for purposes of discussion and the Managing Director’s chairing of Board meetings, consensus should be defined as being achieved only when the double majority requirement is fulfilled. This would set firmer criteria to judge whether consensus was achieved, much as was the practice at the Fund until the end of the 20th century. An understanding to work on the basis of a state-weight double majority can be put into operation before the amendments to the Articles are fully adopted if the proposal has good faith support from the EDs and the Managing Director.

5. Conclusion

International organisations have always faced difficult challenges in terms of organisational design. Managing the representation of a diverse group of member states-including small island off-shore banking centres, poor and populous developing countries, emerging global powers, and industrial economies-is very difficult under simple decision making rules. To better reflect the membership’s interests and ensure the ability of all members to participate fully and equally, international organisations need more evolved mechanisms for decision making.

In the case of the IMF, it is clear that the quota formula method is an ineffective tool. Not only is it overburdened with different roles, but the quota formula alone cannot balance the competing demands for representation. This has lead to a democratic deficit in the IMF structure and helped to reduce the legitimacy of an institution which should provide global public goods.

Wholesale reform of the IMF’s governance is not possible without the support of the current power holders, which complicates the search for an ideal way to democratically represent the Fund’s member states. As a step towards a more democratic structure and an empowerment of voices that have long been silenced in the governance of the global economic system, a double majority has the most promise of success.

The shortcomings in the IMF’s governance extend beyond just decision rules and the quota formula. Attempts at reform, if they are to create an institution that is both effective and accountable, will have to address not just the voting rights of countries, but also the structure and procedures of the executive board, the transparency of decision making processes, the overburdening of quota with too many functions, and the method for the selection of the managing director. Comprehensive reform is the only way to capitalise on the synergies that can be created by equity, transparency and accountability.

Summary of recommendations:

  1. The IMF should institute a double majority system at both the level of the Board of Governors and the Executive Board. Decisions would require the requisite majorities of both the number of IMF members and their voting weight
  2. The thresholds for decision should be equal for both types of majorities. Simple majority decisions would then require approval by more than 50% of the voting weight and more than 50% of the membership. Super majority decisions would require either 70% or 85% of both the voting eight and membership
  3. At the Executive Board during every decision an Executive Director will cast votes for his constituency based on the sum of the voting weight of the members of the constituency and the sum of the number of members of the constituency
  4. The Executive Board should commit to following this procedure immediately while the necessary amendments to the articles of agreement are approved by the Fund’s membership
  5. This change to the voting structure should accompany other changes to the governance of the IMF, including increased transparency of Executive Board discussions and decisions, reforms to the structure of the board and accountability of Executive Directors, and the implementation of a transparent and merit-based process for the selection of the Managing Director


1 Among many others see: Kapur, Devesh, John Lewis and Richard Webb, The World Bank: Its First Half Century, vol.1, pg. 9; Woods, Ngaire, The Globalizers: The IMF, the World Bank and Their Borrowers, Ithaca: Cornell University Press, 2006.

2 The development of the Fund’s quota system has been amply described by those involved in its creation for example, see “The Bretton Woods Debates: A Memoir”, Raymond F. Mikesell, Essays in International Finance, No. 192, March 1994, Princeton University.

3 Millions of people, projections for 2005 from the medium variant of Population Division of the Department of Economic and Social Affairs of the United Nations Secretariat, World Population Prospects: The 2004 Revision and World Urbanization Prospects: The 2003 Revision, http://esa.un.org/unpp, Monday, January 22, 2007.

4 At market exchange rates, three year average (2002-2004)

5 Van Houten, Leo, Governance of the IMF: Decision making, institutional oversight, transparency and accountability, pamphlet series No. 53, Washington: International Monetary Fund, 2002.

6 Mirakhor, Abbas, “Interview with Mr. Abbas Mirakhor, the Dean of the IMF Executive Board”, African Group I Constituency Newsletter, Vol. 6, No. 2, May 2006.

7 Injoo Sohn, “East Asia’s Counterweight Strategy: Asian Financial Cooperation and Evolving International Monetary Order”, G24 technical paper, September 2006, http://www.g24.org/sohn0906.pdf or Yung Chul Park, Economic liberalization and integration in East Asia : a post-crisis paradigm, 2006, Oxford: Oxford University Press.

8 “Latin America sends IMF packing”, Bretton Woods Update, No. 49, The Bretton Woods Project, http://www.brettonwoodsproject.org/art. shtml?x=507679.

9 The September 2002 notice on the board discussion describes this failure diplomatically, stating that the EDs had “a useful opportunity for a further exchange of views”. By January 2003 the twelfth quota review was concluded without any progress. See Köhler, Horst, “Twelfth General Review of Quotas: Alternative Proposal”, Memorandum to Executive Directors, 20 December 2002, http://www.imf.org/external/np/tre/quota/2002/ eng/122002.htm; IMF, “IMF Executive Board Discusses Further Considerations in the Twelfth General review of Quotas”, Public Information Notice 02/105, September 20, 2005, http://www.imf.org/external/np/ sec/pn/2002/pn02105.htm.

10 G24 members include Algeria, Argentina, Brazil, China, Colombia, Congo, DR, Cote d’Ivoire, Egypt, Ethiopia, Gabon, Ghana, Guatemala, India, Iran, Lebanon, Mexico, Nigeria, Pakistan, Peru, Philippines, South Africa, Sri Lanka, Syria, Trinidad & Tobago, Venezuela

11 The Bretton Woods Project, “IMF quota reform poses risks to developing countries”, Bretton Woods Update, No. 53, http://www.brettonwoodsproject.org/art.shtml?x=545872.

12 Mirakhor, Abbas and Iqbal Zaidi, “Rethinking the Governance of the International Monetary Fund”, IMF Working Paper, WP/06/273, December 2006.

13 The United States has committed to not increase its quota beyond its current shareholding in the IMF. We have assumed that the United States maintains its same level of quotas and that the overall quota level remains the same, and then distributed the remaining quota according to the formula specified.

14 The chosen ratio is for illustrative purposes and does not indicate an expected outcome. This mimics the proposal of the Cooper commission, (QSFR 2001) which suggested that a new quota formula should be based two-thirds on GDP and one-third on variability.

15 EU Member states do not strongly support the use of variability, as proposed by the Cooper commission, but strongly support the use of openness as a factor. We have maintained GDP as 2/3 of the weight for illustrative purposes only.

16 This includes Singapore and South Korea as per current IMF conventions.

17 Libya and Egypt are considered Middle Eastern countries as per current IMF conventions.

18 Includes the votes of Liberia and Zimbabwe which are currently suspended.

19 Including Slovenia

20 This is under the “Mitbestimmungsgesetz”, the German “Co-determination Act” of 1976 that expanded on the previous one-third requirement.

21 Harstad, Bård, “Majority Rules and Incentives”, Quarterly Journal of Economics, 2005, 120(4) pp 1535-1568.

22 Rapkin, David P. and Jonathan R. Strand, “Reforming the IMF’s Weighted Voting System”, The World Economy, 2006, 29(3), p. 322.

23 Strand, Jonathan R and David P Rapkin, “Voting Power Implications of a Double Majority Voting Procedure in the IMF’s Executive Board”, in Reforming the Governance of the IMF and the World Bank, Ariel Buria ed., London: Anthem, 2005.

24 Bräuninger, T., “When Simple Voting Doesn’t work: Multicameral Systems for the Representation and Aggregation of Interests in International Organisations”, British Journal of Political Science, 2003, 33(4) pp. 681-703.

25 Tsebelis, George and Xenophon Yataganas, “Veto Players and decision-making in the EU After Nice: Policy Stability and Bureaucratic/Judicial Discretion”, Journal of Common Market Studies, 2002, 40(2) p. 283.

26 Ibid. p. 287.

27 Ibid.

28 Instrument for the Establishment of the Restructured Global Environment Facility, May 2004, para. 25(c)(i).

29 Jakobeit, Cord, “Enhancing the Voice of Developing Countries in the World Bank”, in Buira, Ariel, (ed.) Reforming the Governance of the IMF and the World Bank, London: Anthem Press 2005, pp. 213-233.

30 Bräuninger, op. cit.

31 Article 60, Agreement Establishing the African Development Bank, 2002. Available at: http://www.afdb.org/pls/portal/url/ITEM/F56CEF0DF7349109E0 30A8C0668C4DD4

32 International Monetary Fund, Articles of Agreement, Washington: IMF, 1998.

33 Bradlow, Daniel, “Stuffing old wine into new bottles: The troubling case of the IMF”, Journal of International Banking Regulation, Vol. 3, No. 1, August 2001.

34 Guitián, Manuel, “The Unique Nature of the Responsibilities of the IMF”, Pamphlet Series No. 46, Washington: IMF, 1992, http://www.imf.org/external/ pubs/ft/pam/pam46/pam46con.htm.

35 This proposal has been made by Cord Jakobeit, Jonathan Strand and David Rapkin, as well as Ngaire Woods. See Jakobeit C, op. cit.; Rapkin and Strand, op. cit.; Woods, op. cit.

36 Mirakhor and Iqbal, op. cit.

37 See Woods, op. cit., pp 194-5.

38 Woodward, David, “’Democracy’ in the IMF: the threat to the global economy from a colonial era relic”, nef briefing, the new economics foundation, http://www.jubileeresearch.org/news/IMFBWP1d.pdf.

39 The Fund has 184 current members, but two of those, Zimbabwe and Liberia, have had their voting rights suspended because of arrears – see IMF Executive Directors and Voting Power, The International Monetary Fund, November 07, 2006, http://www.imf.org/external/np/sec/memdir/eds.htm.

40 Such a system would introduce a mildly perverse incentive for large constituencies, which have been pointed to as an impediment to the effective exercise of voice by developing country EDs. This is addressed by reforms to the system of allocation of resources to ED offices and the number of alternate EDs allowed. Larger constituencies should have proportionally larger resources and more alternate EDs to handle both the higher number surveillance documents and the higher load of IMF programmes. Efforts to reform both have already commenced, see: de Rato, R. “Statement on the Work Program of the Executive Board”, press release no. 06/267, International Monetary Fund, November 30, 2006, http://www.imf.org/external/np/sec/pr/2006/pr06267.htm.

41 The fourth amendment to increase the allocation of SDRs, approved by the Board of Governors in 1997, has still not been ratified by members representing enough of the voting weight to enter into force. See http://www.imf. org/external/np/exr/facts/sdr.HTM


The authors would like to thank the following people for their comments and suggestions: Monica Blagescu, Daniel Bradlow, Ariel Buira, Jo Marie Griesgraber, Michael Hammer, Dennis Leech and Jeff Powell. Any remaining errors and omissions are solely the responsibility of the authors.

Published by the One World Trust and Bretton Woods Project © One World Trust and Bretton Woods Project, February 2007

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopy, recording or otherwise, without the prior permission of the One World Trust and Bretton Woods Project.

About the organisations

One World Trust

The One World Trust promotes education and research into the changes required within global organisations in order to achieve the eradication of poverty, injustice and war. It conducts research on practical ways to make global organisations more responsive to the people they affect, and on how the rule of law can be applied equally to all. It educates political leaders and opinion-formers about the findings of its research.

Its guiding vision is a world where all peoples live in peace and security and have equal access to opportunity and participation.

One World Trust
3 Whitehall Court
London SW1A 2EL
United Kingdom
Tel: +44 (0)20 7766 3470
Email: accountability@oneworldtrust.org
Website: www.oneworldtrust.org
Charity Number 210180

Bretton Woods Project

The Bretton Woods Project works as a networker, information-provider, media informant and watchdog to scrutinise and influence the World Bank and International Monetary Fund. Through briefings, reports and the bimonthly digest Bretton Woods Update, it monitors projects, policy reforms and the overall management of the Bretton Woods institutions with special emphasis on environmental and social concerns.

The Project works towards institutions at the local, national, regional and international levels that operate on the principles of equity and sustainability and that are transparent and responsive to the poorest and most vulnerable citizens.

Bretton Woods Project
Hamlyn House
Macdonald Road
London N19 5PG United Kingdom
Tel: +44 (0)20 7561 7610
Email: info@brettonwoodsproject.org
Website: www.brettonwoodsproject.org