IFI governance


Reform of World Bank governance structures

12 October 2007 | Briefings

A fully-formatted PDF version of this briefing is available on IFIwatchnet.

Table of Contents

I. Introduction

II. Reforms within the competence of Management and the Executive Board
A. Leadership Selection
B. President and Executive Board accountability
C. Increased support to developing countries Executive Directors offices
D. Cap the size of constituencies
E. Staff diversity
F. Transparency
G. Fit-for-purpose governance structures
H. Selective capital increase

III. Reforms requiring a change in the Articles of Agreement
A. Increase basic votes and/or membership shares
B. Double majority voting
C. Fit-for-purpose governance structures
D. Fit-for-purpose quota formulae
E. Giving up European chairs in the Executive Board

IV. Conclusion: Prioritising reforms

Annex 1: IBRD (the ‘Bank’) and IDA (the ‘Association’) governance structures
Annex 2: Related resources

I. Introduction

1. The World Bank Group faces a legitimacy crisis. Devised in a post-war colonial era, the Bank’s governance structures have not adapted to new international norms of democracy, transparency and accountability1. This paper will examine the options for reforming them.

2. This discussion is not new. For five years, the issue of what has euphemistically been termed “voice” has been on the agenda of nearly every spring and annual meeting of the World Bank and IMF. Calls have come from every quarter for a democratisation of the international financial institutions. A ‘voice’ progress report will be given to the Executive Board in September 2007, and the issue will once again be discussed at the annual meetings in October 2007.

3. There are two reasons to believe that the prospects for reform may have improved. The first is the highly public and institutionally embarrassing controversy surrounding the resignation of Paul Wolfowitz that highlighted the anachronistic nature of the Bank’s leadership selection process. The second is the increasing competitive pressure that threatens the Bank’s financing model. Developing countries and their supporters should seize this opportunity to forge structures that serve them; if the countries that currently dominate at the Bank are unwilling to change, then developing countries can and should go elsewhere.

4. It is unclear whether the governance reform discussions will have to await the completion of the current discussion of the Bank’s long-term strategy 2. Considering the inter-relation of form and function, this would seem a sensible approach. But this paper will discuss potential governance reforms with an understanding that some reforms become more or less imperative depending on the direction of the discussions around the long-term strategic direction of the Bank.

5. The reform options have been divided into those that are within the competence of Bank Management and the Executive Board, and those that require changes to the Articles of Agreement. This is not to suggest that those reforms that require changes to the Articles should be abandoned, as is sometimes argued. In contrast, some of the most needed reforms will require opening up the Articles. If the Articles are to be opened (as is likely if at least the ‘basic votes’ question is to be addressed), it makes sense to address a number of issues in parallel.

6. The paper ends with a brief discussion of what issues would be most strategic to highlight, considering potential impact and potential for leverage. The first annex gives a brief summary of the differences between the IDA and IBRD governance structures. Further reading suggestions and key links are at the end.

II. Reforms within the competence of Management and the Executive Board

A. Leadership Selection

7. The need for this should be clear after the Wolfowitz resignation, with the US disregarding any attempt to open the process up in their appointment of Zoellick to take over from Wolfowitz, and then Europe’s quid pro quo with their nomination of Dominique Strauss-Kahn at the Fund. An open, transparent and merit-based selection process is sine qua non, with emphasis on experience with development issues in a developing country context.

8. There are a couple of small sticking points in the details: Should candidates have to publicly declare their candidacy? This is what the board tripped over when they reviewed the Bank-Fund paper on the issue in 2001 (see ‘related resources’). There are genuine concerns that a public contest held in the full glare of the media might scare off some qualified candidates.

9. Secondly, should the board vote on the candidates? Placing the selection in the hands of some kind of expert panel has the advantage of de-politicising the process, but may diminish the democratic legitimacy of a candidate chosen by country representatives. If a vote of the Executive Board is held, should it be a secret ballot? While this would allay fears of low-income countries being pressured by industrialised countries, it runs counter to a broader appeal for greater transparency and accountability. In any event, the presidency of the World Bank is a position charged with the global public interest, and the Bank’s members have the right to be involved effectively in selecting who will hold the position. This will also make the President more accountable to the members.

B. President and Executive Board accountability

10. Systems should be put in place for the evaluation of the performance of the President by the Executive Board, and of the Executive Board by (likely a sub-committee of) the Board of Governors. This would mirror the recommendations of the panel on IMF board accountability (see link in ‘related resources’).

11. While such evaluations would clarify the internal lines of accountability, they do not address the accountability of Executive Directors to their constituencies. While this is clear in principle (if not in practice) in single-country constituencies, it is not at all clear how an African Executive Director, for example, could be accountable to the governments and/or stakeholders of 24 countries even if s/he wanted to. The issue of constituency size will be discussed below. Beyond the structural constraints, resources could be devoted to facilitating Executive Director interaction with national/regional parliaments in her/his constituency.

C. Increased support to developing countries’ Executive Directors’ offices

12. Barring fundamental reform in constituency representation, Executive Directors from large, developing country groupings should be provided with an additional alternate or additional senior advisor(s). (nb. An analytical trust fund supported by the UK’s Department for International Development was created in 2004 to allow African EDs at the Bank and Fund to commission a programme of research3.) There is a battle ongoing at the moment over how many advisors the Fund can afford to pay for the African Directors4. Increasing advisors (or research trust funds etc.) only requires additional financial resources. Increasing alternate EDs requires an amendment to the articles, the desirability of which depends on the package of reforms.

D. Cap the size of constituencies

13. Of the 24 seats on the board, 8 are single seat chairs, while the remainder range from four to 24-country groupings. Countries may decide to switch the constituency they are in at the bi-annual election of executive directors, with the unanimous approval of the members of the constituency they wish to join. Large constituency Executive Directors must use their limited time to focus on high-profile projects and programmes for their constituency, leaving little time for broader policy engagement, stakeholder outreach, etc. Large constituencies with mixed developed and developing country members may also make it more difficult for their Executive Directors to give equal attention or prominence to the concerns that may be raised by the weaker members of their constituency.

14. Ensuring a better match in constituency-constitution among countries who share certain basic common interests may be a better option to go along with capping the size of constituencies. This would mean that constituencies that may be composed of less well-resourced members will require the allocation of additional Bank resources to support the proper functioning of their respective Executive Directors.

E. Staff diversity

15. At a meeting in Mozambique in August, the African Consultative Group of the IMF , made up of finance ministers and central bank governors called for increased representation of African staff at the IMF5. The Bank has made more progress on this issue than the Fund6, however more progress needs to be made, especially at the management and senior technical positions. Diversity in academic background – both academic field and location of study – should also be encouraged.

F. Transparency

16. A positive step was taken in making available the agendas and the minutes of Executive Board meetings. They’re helpful but lack detail. To ensure the accountability of decision-makers, transcripts should be made available.

17. More generally, the Bank should adopt in its disclosure policy those principles recommended by the Global Transparency Initiative – namely, presumption of disclosure, strictly limited exceptions, and the introduction of an independent appeals mechanism (see ‘related resources’ for link to GTI IFI transparency charter). The Bank was to have conducted a full review of its disclosure policy in 2007, but this review is now in question as part of the work agenda that was put on hold with the Wolfowitz-Zoellick transition.

G. Fit-for-purpose governance structures

18. The reform of the Bank’s governance structures (IBRD and IDA) should be de-linked from reform of the IMF’s governance structures. The stylised histories of the reforms of the IMF and the World Bank Group mean that this is already partially the case (see table 1 below).

19. The need to de-link was recognised in the World Bank options paper on voice and representation (see ‘related resources’) – “the Bank has an independent need to deal with the voice issue and the considerations that it needs to take into account are different from the Fund” (p. 4). The Bank’s legal department has also made clear that IMF reforms “do not have direct legal consequences for the Bank”. If this implies changes in the quota formulae (see below), then it requires an amendment to the articles of agreement.

H. Selective capital increase

20. Unlike a general capital increase, which is used if the Bank’s capital stock is considered insufficient (which by most experts opinion it is currently not), a selected capital increase can allow the shareholdings of certain under-represented countries to be increased at the expense of others.

21. It is likely that, following on the heels of the IMF’s selective quota increase7, that there will be pressure for the Bank to offer a selective capital increase to countries that are ‘under-represented’. ‘Under-represented’ here means one of two things – either the country is under-represented at the Bank relative to its quota share at the Fund; or the country is under-represented at the Bank relative to its share in the global economy8.

22. Other than under-represented middle-income countries, there could be a push to give a selective capital increase to poor countries. However, this would be tricky since this is about the last thing these countries would like to spend money on. This could be resolved by either (a) setting up a donor trust fund to pay for their capital increases, or (b) setting the price of their shares at a nominal rate (this has been done before).

23. This does not require an amendment to the Articles of Agreement, but would require ‘non-subscribing members to agree not to exercise their pre-emptive rights’9. There would obviously be some political difficulty in deciding who gets how much of an increase.

Table 1: Voting weight (% of total) at IMF, IBRD, and IDA (indicative only)

Country IMF*** IBRD** IDA**
United States 16.79 16.38 12.94*
Germany 5.88 4.49 6.45*
France 4.86 4.30 4.12*
United Kingdom 4.86 4.30 5.14*
Belgium 2.09 1.81 1.16*
Canada 2.89 2.78 2.77*
Russian Federation 2.70 2.78 0.31*
Saudi Arabia 3.17 2.78 3.34
China 3.66 2.78 1.99
South Korea 1.33 0.99 0.64
Bangladesh 0.25 0.32 0.60
Mexico 1.43 1.18 0.62
Bolivia 0.09 0.11 0.31
Egypt 0.44 0.45 0.48
Ghana 0.18 0.11 0.40
South Africa 0.85 0.85 0.30

* Part 1 (donor) countries in IDA

** As of 24 July 2007

*** Updated 21 August 2007

III. Reforms requiring a change in the Articles of Agreement

A. Increase basic votes and/or membership shares

24. Returning the basic vote to its original levels of 10.78% (currently at 2.8%), would increase the total votes of developing countries at the IBRD from 40% to just over 43%.

25. There will likely be movement on this -the question is how much. Options examined include doubling the IBRD ‘membership share’ 10 to 500 and adding 250 ‘basic votes’11 – effectively this ‘tripling’ would reduce high income countries share from 60% to 57.79%. Any changes should be accompanied at the IBRD with the introduction of a mechanism to automatically maintain the ratio of basic votes in future quota increases (this is already the case at IDA). While increasing basic votes requires an amendment to the Articles of Agreement, increasing membership shares only requires approval of the Board of Governors.

B. Double majority voting

26. For the IMF, there have been proposals from both governments (notably Germany12 and supported by the UK13, France, and China) and NGOs (such as the Bretton Woods Project and One World Trust) suggesting that both simple majority (50% + 1) and super majority (66%, 85%, etc.) decisions require majority backing both by economically weighted votes and number of countries (for report, see link in ‘related resources’ below).

27. The Bank’s options paper on voice and representation (2001 – see ‘related resources’) mistakenly posits an obstacle to double majority. The paper’s authors argue that it would be difficult to reach agreement on the legal definitions of a new category of members (assuming that double majority would imply majorities of both ‘developed’ and ‘developing’ countries, or in the Bank’s language, part I and part II countries). This is not necessary. The requirement of a simple numerical majority of countries would serve similar political purposes in re-balancing North-South relations on the Board.

28. Furthermore, double majority (and other structural reforms proposed below) will not affect the Bank’s AAA credit rating, i.e. institutional investors will not think that the ‘madmen are running the asylum’. 14

C. Fit-for-purpose governance structures

29. Reform of IBRD governance structures could be separated from that of IDA, based on the notion that they are serving different clients (middle-income vs. low-income); operating in different ways (loans vs. increasingly grants); and financed differently (bond issues vs. donor replenishment). This is already partially the case (see Annex 1 below, and Table 1 above).

30. However, splitting IBRD and IDA posts would represent increased burden both financially and in human resource terms. Would separate governors/directors/presidents be desirable? Who would bear the costs? There is also the question of whether or not this risks dividing Southern solidarity between middle and low-income countries15.

D. Fit-for-purpose quota formulae

31. At the IMF, the quota formula16 is used to determine contributions17, access levels and decision-making power. At the IBRD, the formula18 is used to determine contributions and decision-making power. At IDA, the formula is used to determine decision-making power only.

32. Therefore, unlike the IMF, where arguably three formulae are needed, perhaps only two are needed at the Bank. The formula to determine contribution would rightly emphasis GDP/capita (based on the principle that those who can better afford to pay should pay). The formula to determine decision-making power should rightly emphasise democracy – see Woodward’s proposals in a G24 technical briefing (see link in ‘related resources’ below) to use population-weighted figures. Woodward argues that the standards applied should be those commonly accepted in democratic processes at the country level and concludes that voting rights should be determined exclusively on democratic (ie one-country-one vote and/or population-related) principles, in such a way that votes increase less than proportionally with population.

E. Giving up European chairs in the Executive Board

33. Since the US is likely to veto any increase in the size of the Board, the obvious solution to the inequitable representation at the Board is to reduce the disproportionate number of Europeans at the table 19. European civil society organisations have already called for a reduction in the number of European seats on the boards.

34. Edwin Truman of the Institute for International Economics has been one of the few to wade into detailed proposals for European consolidation on the board of the IMF. He has argued for the consolidation of EU representation into seven EU-majority seats in the next election of executive directors, ultimately moving towards two EU seats and one non-EU European seat. While the reasoning works out differently at the Bank (membership in the Euro-zone is less an issue at the Bank than it is at the Fund), European NGOs will need to push European governments to start putting forward proposals. UK NGOs, for example, have been pushing their government to voluntarily concede its appointed chair.

IV. Conclusion: Prioritising reforms

35. Despite frustration over the Zoellick appointment and the on-going selection process for the next IMF Managing Director, pressure must be kept up on the leadership selection issue at both the World Bank and the IMF. The Wolfowitz issue will increase attention to the accountability of the president and the board. Generally transparency is better dealt with during reviews of the disclosure policy, but pushing on board transparency and accountability usually gets overlooked, so it would be useful to push on this here.

36. While double majority voting makes political sense at the IMF, the imperative to give precedence to democracy over economics at the development institution of the BWI twins seems even stronger. However, this does not change the political reality that the US is not likely to agree to a formula based on purely democratic principles (especially if this is in combination with progress on the leadership selection issue) that would lessen its influence.

37. Therefore, a double majority might be more likely to get past20. Double majority can be seen as a medium-term practical step, moving towards the longer-term resolution of democratic representation in the BWIs.

38. Global governance institutions, including the Bretton Woods institutions, the World Trade Organization, and the United Nations, need to adapt their governance and decision-making structures and processes to reflect the growing influential role of developing countries in the global political and economic arena. The BWIs, as global economic governance institutions, have to change. Reform is necessary not only for developing countries but also for the continued legitimacy and global relevance of these institutions

39. The issue of democratic representation and effective representation by developing countries in the governance structures of both the World Bank and the IMF will hence not go away for as long as their concerns have not been adequately addressed. Governance structures of institutions that are supposed to assist developing countries improve their development prospects need to respond to and be reflective of the priorities of their developing country constituency, otherwise they will risk losing both relevance and legitimacy.

40. In the long-term, developing countries should push for the following principles to be reflected in any reform of the BWIs, as has been suggested previously by the South Centre21:

  • a. Borrowers should be included in decision-making at the top, … by creating incentives for the most powerful members to consult and build coalitions across a wide range of members. For example, the proposals for increasing the scope for double majority voting would ensure that developed countries cannot act or adopt decisions without considering the views of developing countries.
  • b. Voting power is very important in the BWIs both explicitly and implicitly, because it underlies the consensus decision-making. … the underrepresented members should be given what is rightly theirs: voting power that reflects their position in the world economy. The suggestions for quota increases are in line with this principle, as are the G-24 proposals for a substantial increase in basic votes and a new quota formula that reflects the relative economic size of developing countries in the world economy.
  • c. The fact that the executive directors make the decisions and cannot split their vote to reflect differences in the views of constituency members means that developing countries, and especially those in mixed constituencies, do not affect the decision-making at all. The system could be changed to ensure that each country’s vote is reflected.

41. Finally, institutional reforms of the BWIs need to be linked to a broader understanding of the need for structural changes in global economic and political power relations, in how developed and developing countries relate among and between themselves and their peoples. The underlying power imbalances between developed and developing countries that so far mark the international system need to be addressed if institutional changes are to result, in the long term, in reforms that effectively reconfigure the philosophical and institutional perspectives and actions of the BWIs.

42. Development continues to be the primary challenge facing the global community. The process of globalization, in which both Bretton Woods institutions were key drivers, has resulted in increased development inequality and global inequity. The failure of past policy prescriptions based on a single development policy paradigm should give way to a fresh approach that emphasizes development policy choice and flexibility for developing countries. Global institutions that are supposed to foster the development of developing countries should be open to fresh and innovative ideas to enable them to respond better to developing countries’ development needs and priorities, and this should include taking a fresh look at how their governance structures should be reformed in order for them to become more responsive to the demands of their developing country constituency. With the potential for stronger South-South integration and cooperation for development now much greater than ever and constituting a viable alternative that could be explored by developing countries, the BWIs may, in fact, need to take a closer look at themselves to see whether they continue to be relevant as development actors.

Annex 1: IBRD (the ‘Bank’) and IDA (the ‘Association’) governance structures


Governors: Governors and alternates serve for five year terms and may be reappointed. Board of Governors shall select one governor as chair.

Directors: Twenty-four executive directors22, five appointed by the five members having largest number of shares. Remainder elected by the board of governors. Elections will take place every two years23. Directors may be reappointed. Rules for rotating the director post in elected constituencies are decided by each constituency. (nb. no provisions for evaluation of performance)

President: The executive directors select a president, who will cease to hold office when the directors decide. (nb. no provisions for evaluation of performance)

Advisory Council: Not less than 7 wise persons, broadly representative, to advise the Bank on general matters. Councillors shall serve for two years and may be reappointed. (nb. never enacted)

Location: The principal office of the Bank shall be located in the territory of the member holding the greatest number of shares.


  • 250 basic votes24 plus one additional vote for each share of stock held. High-income countries hold 60% of votes; low and middle income countries hold 40% (two African chairs hold 5.31%).
  • Subscription votes (20% paid-in; 80% subject to call). Due to repeated selective capital increases, the shareholdings are out of line with those at the IMF (the formula for share allocation at the IMF is supposed to guide IBRD allocation). Most developing countries (virtually every African country) are over-represented at IBRD according to the IMF quota formula.

Amendments to the articles: three-fifths of the members having eighty-five percent of the total voting power


Posts: Governors and alternates / Chair of the board of governors / Executive directors and alternates / President of the Bank shall ex officio serve in the same positions of the Association.

Location: Principal office of the Association shall be the principal office of the Bank.


Originally, 500 basic votes plus one vote / $5000 of its initial subscription (‘subscription votes’). Through successive replenishments there are now a possible 39,700 basic votes per member (for each member who has subscribed to all allocations made).

After IDA replenishments, subscription votes are re-allocated based on three principles:

  1. votes of rich countries (so-called ‘Part 1’) should correspond to its share of total cumulative Part 1 resources contributed to IDA;
  2. the relative voting power of the Part II countries as a group should be maintained by conferring subscription votes on part II members at a nominal cost (currently one vote for each $25 subscribed);
  3. membership votes (basic votes) should be increased to preserve the voting share of the smaller countries.

When any additional subscription is authorized, each member shall be given an opportunity to subscribe, under such conditions as shall be reasonably determined by the Association, an amount which will enable it to maintain its relative voting power, but no member shall be obligated to subscribe. (Article III, section 1)

Amendments to the articles: three-fifths of the members having four-fifths of the total voting power

Annex 2: Related resources

World Bank options paper on voice and representation (April 2007)

Open CSO statement on steps to democratise the World Bank and IMF (April 2003)

UK NGO open statement on governance reform at the IMF (July 2006)

Joint WB-IMF report on the process for selection of the president and MD (2001)

Global Transparency Initiative (GTI) charter for IFI transparency

High-level panel on IMF board accountability, New Rules for Global Finance

Bridging the democratic deficit: Double majority decision-making at the IMF, Bretton Woods Project

IMF voting reform: need, opportunity and options, David Woodward for the G24

IBRD articles of agreement

IDA articles of agreement

Long due reform? The IMF, the World Bank and global economic governance 60 years later

Comparative Assessment of Developing Country Participation in the Governance of Global Economic Institutions (South Centre SC/GGDP/AN/GEG/3/Rev.11Sep, August 2006)