Chapter

5 Executive Boards in International Organizations

Author(s):
Ruben Lamdany, and Leonardo Martinez-Diaz
Published Date:
July 2009
Share
  • ShareShare
Show Summary Details
Author(s)
Leonardo Martinez-Diaz

To identify ways to strengthen the IMF’s Executive Board in its various functions, this paper compares and contrasts that governing body with the executive boards of eleven other inter-governmental organizations (IGOs). The paper identifies four key roles that IGO executive boards are expected to play—those of political counterweight, performance police, democratic forum, and strategic thinker—and assesses how well the boards of the eleven organizations are equipped to play these roles. The exercise allows us to identify three “models” of governance, each with different strengths and weaknesses. The paper concludes that the twin crises of relevance and legitimacy that the IMF is currently facing are closely related to the Fund’s adherence to a particular model of governance. This model gives major shareholders close control via the Executive Board over the use of the financial resources they provide, but this control is maintained at the expense of the Board’s capacity to act as strategic thinker, performance police, and democratic forum. The paper offers recommendations on how to strengthen the Board’s capacity to play these other roles.

Rethinking IMF Governance Reform

In recent years, the debate on reforming the governance of the International Monetary Fund (IMF) has largely focused on the issue of quotas and voting power. Reforms in this area seek to protect the voting power of the institution’s smallest shareholders from further erosion and to augment the voting power of countries whose growing weight in the global economy is not reflected in their quotas and votes.1 But these adjustments, by themselves, are unlikely to address the institution’s most serious shortcomings in effectiveness, efficiency, accountability, and member representation. Also needed are reforms to the Fund’s internal governance—reforms that might improve how the institution thinks, makes decisions, and relates to its members and stakeholders. This type of reform means examining closely how the Fund’s governing bodies—and the Executive Board in particular—function.

Why focus on the Executive Board? From the Fund’s inception, the Board of Governors (the institution’s highest governing body) delegated to the Executive Board most of its powers. Charged with conducting “the business of the Fund” and with exercising “general control” over the Managing Director, the Executive Board was meant to be the locus of decision making and oversight in the institution (IMF Articles of Agreement, Article XII, Sections 3–4). The Board is also the principal forum in which the representatives of member governments interact with the technical experts that staff the institution and where political authorities give legitimacy to the staff’s technical judgments. And third, the Board is the main organ for providing voice and representation to the Fund’s near-universal membership.

The aim of this paper is to illustrate how the different roles of the Fund’s Board could be strengthened. Its method is comparative analysis—comparing and contrasting the Fund’s Board with the executive bodies of other IGOs. The paper attempts to show three things: (1) that the arrangements that govern the IMF’s Executive Board today are part of a larger universe of possible governance models, and that each of these models has a different set of strengths and weaknesses; (2) that changing how the Fund’s Board operates necessarily involves trade-offs among roles; and (3) that specific governance mechanisms imported or adapted from other governance models can help strengthen the Board and the Fund’s internal governance more broadly.

The paper has four parts. The first section identifies four generic roles that executive boards of IGOs are expected to play and proposes a series of indicators to measure these characteristics. The second section uses these indicators to evaluate how the Fund’s Board performs each of these roles. The third section does the same, though more superficially, for eleven international organizations. This assessment allows us to categorize the organizations according to their respective “governance models” and to compare them with the IMF. The final section draws conclusions from this comparative exercise and identifies governance mechanisms that might be helpful when thinking about IMF governance reform.

Executive Boards in International Organizations

At least a century ago, governments began to establish intergovernmental organizations to address transnational problems that they could not cope with on their own.2 IGOs offer governments several advantages, including a vehicle to engage in sovereignty-sensitive activities, such as surveillance and dispute resolution, which required a neutral agent that could be trusted to treat all countries equally. They also offer governments a way to participate at arm’s length in activities—such as development assistance and peacekeeping—that required some separation from domestic politics in order to generate legitimacy and trust (Abbott and Snidal, 1998; Hawkins and others, 2006).

Having decided to create IGOs and to delegate power to them, the problem for governments became how to exercise control over these organizations while preserving their capacity to produce global public goods. Member states faced a principal-agent problem, with national governments in the position of principals and IGOs as their agents.3 How much power should be delegated to the IGO? What mechanisms should be in place to ensure that the incentives of the organization’s management and staff were aligned with those of member states?

Many of the most important IGOs were given the same basic structure, outlined in Figure 1. In the typical structure, the highest governing body is usually an assembly or board of governors—a political body in which every member state has a seat at the table. Under this plenary body is typically an executive board or equivalent; this can be either a plenary body or one limited to a subset of the membership. (In some IGOs, such as the OECD, the executive-board equivalent is known as a “Council;” this should not be confused with the Council mentioned in the IMF’s Articles of Agreement, which would be a ministerial-level body.) Below the executive board is the chief executive officer (CEO) of the institution, variously referred to as director-general, president, or managing director. The CEO, usually appointed by the executive board, is in charge of the day-to-day management of the organization, subject to the board’s oversight. As head of the organization, the CEO is in charge of the staff and is ultimately responsible for its work. In many institutions, the CEO is embedded in a larger management structure, composed of a number of vice-presidents, deputy managing directors, or their equivalents.

Figure 1.Typical Governance Structure of an Intergovernmental Organization

Governments and citizens soon came to demand several things from IGOs: effectiveness (fulfilling their mandate), efficiency (fulfilling the mandate in a cost-effective way), voice (giving members adequate representation in decision making), and accountability (the right to hold IGOs to a set of standards and to impose sanctions when these standards are not met). The executive board or equivalent in each organization was central in helping the IGO meet these expectations.

Four Roles of Executive Boards

I argue that the executive boards of IGOs are expected to play a combination of four roles. Two of these—I call them performance police and strategic thinker—are roles executive boards play in other organizations, including private corporations. The other two—labeled here as political counterweight and democratic forum—are particular to IGOs. I describe each, in turn.

The Board as Political Counterweight

Executive boards in IGOs can serve as a “political counterweight” to the technical decisions made by the organization’s management and staff, as a political check by member governments on the organization’s actions and policies.4 This involves reviewing every staff decision of importance, judging whether these are consistent with the national interest of the country (or countries) that each executive director represents and, when they are not, taking action to bring them into line. The role of political counterweight assumes that executive directors act primarily or exclusively with their national interests in mind, as defined by the governments that appointed or elected them.

For a board to perform this role effectively, it must have several characteristics. First, board directors must owe their primary allegiance to their national authorities. Board members must have relatively little room to act autonomously from their political masters. Frequent turnover and short tenures for board directors help ensure their loyalty to capitals and keep the directors from “going native” and identifying too closely with the organization’s interests. To exercise political control, directors must also have adequate access to information about what is happening inside the institution. The board must have a bureaucratic machinery of its own, including a secretariat and advisors who can collect, process, and interpret information regularly. Finally, the board needs to be closely involved in all aspects of the organization’s business so it can monitor and intervene at a detailed level when political imperatives demand it.

The Board as Performance Police

The second role an IGO executive board is called upon to play is as “performance police”—as monitor and overseer of whether and how management and staff are carrying out the organization’s tasks in accordance with some standard collectively agreed by the organization’s members. In contrast to the political counterweight role, directors make judgments based on performance standards that are set out ex ante by the whole membership, instead of on their individual national interest. Indeed, performance standards may or may not be compatible with members’ narrow national interests at a particular point in time. In this role, the board is responsible for setting the standards against which management’s performance will be assessed periodically, and ensuring that policies set by the board are implemented fully and in a timely manner. When performance is found to fall short, the board is charged with taking corrective action.

An executive board can serve as an effective performance police only if certain institutional conditions are in place. First, responsibilities and actions of the CEO must be distinguishable from those of the board. If the behavior of CEO and board cannot be observed independently of each other, then the lines of accountability become blurred and the board can no longer evaluate the CEO’s performance without also passing judgment on its own performance, generating a conflict of interest. Second, performance standards or benchmarks must be established by the board itself or some outside authority. In addition, the board must have sufficient access to information to assess regularly the performance of the CEO and staff. At the very least, this means reporting requirements for the CEO. Finally, the board must be able to reward or punish management on the basis of performance evaluations, including dismissing the CEO in cases of serious underperformance or personal misconduct.

In the private and non-profit sectors, the performance police role is a fundamental responsibility of executive boards. CEO evaluation by the board has become central to board activities—for instance, 96 percent of S&P 500 firms have a formal process to evaluate the CEO’s performance and do so on an annual basis (Spencer Stuart, 2006a: 7). Eighty percent of non-profit executive boards in the United States follow the same practice (BoardSource, 2004: 9). CEO performance evaluation is no longer just the responsibility of a specialized committee—it is fast becoming a responsibility involving the full board.

The Board as Strategic Thinker

Boards are also expected to play the role of “strategic thinker.” This entails anticipating how the organization’s goals and instruments will be affected by changes in the external environment, formulating strategies for adapting goals and instruments to the changing environment, drawing lessons from experience, and feeding this knowledge back into the organization. In IGOs, “strategic thinking” also entails a larger responsibility not relevant to private-sector firms—directors must also ensure that the organization (and the board itself) is functioning effectively as a catalyst for cooperation among member nations.

For a board to play its role as strategic thinker, it must provide an environment that supports frank and constructive deliberation among board directors. In practice, this means relatively small boards. Corporate governance experts suggest that executive boards should have no more than ten members, with twelve as the absolute maximum (Carter and Lorsch, 2003: 89–91). Once boards get larger than a dozen members, the quality of participation declines, decision making becomes cumbersome, free-rider problems increase, and the effectiveness of the board deteriorates. Private sector firms seem to adhere closely to this principle.5 The tendency toward small boards is also evident in the non-profit sector.6

A board that can formulate strategy effectively also requires a high level of expertise, institutional memory, and experience. This generally means relatively long terms of office for board members and the recruitment of directors with considerable experience.7 Experts believe that in the private sector, directors should be expected to serve at least two three-year terms (Higgs, 2003: 5). The strategic-thinking board should also keep some distance from the day-to-day operations of the organization. If it is submerged in detail, the board will lose sight of strategic priorities and direction. For this reason, corporate boards tend to meet only a few times per year. For example, the typical board of a major business corporation meets six to eight times per year (Spencer Stuart, 2006a: 21).8

Finally, a board that is effective at strategy formulation can benefit greatly from the voices of independent directors. Independent directors are described as figures “free from any business or other relationship which could materially interfere with the exercise of their independent judgment” (Combined Code on Corporate Governance, 2006, A.3.2).9 Their main contribution is to bring an outside, more objective view to the board’s deliberations, and to reduce the possibility of conflicts of interest. In the private sector, independent, “non-executive” board directors have become the norm.10 Independent directors tend to dominate sensitive board committees, especially audit and remuneration committees.

The Board as Democratic Forum

Finally, an IGO board is also called upon to serve as a forum for giving voice to the views of individual members. In this role, process matters more than outcome—decisions are judged legitimate only if they are arrived at through a process of deliberation in which all voices can be heard and considered. The use of the word “democratic” here does not imply that members necessarily have equal voting or political power, but that they enjoy an equal right to speak and be heard.

If a board is to perform its role as democratic forum, it must be inclusive: it must have adequate mechanisms for representing, directly or indirectly, the entire membership, and for giving member states a channel to have their voices heard. The board’s rules should safeguard the right of all members to participate meaningfully in the body’s deliberations and should guarantee that dissenting views can be expressed and recorded. Board records should accurately reflect the degree of agreement behind decisions, and rules should limit situations in which a minority of the membership can force a controversial decision with little or no board deliberation.

A board with a one-country-one-vote system most closely conforms to the ideal of a democratic forum. Under an egalitarian voting system, board members can interact as equals, and they are compelled to consider the views of their colleagues (or at least of a majority of them). A board may play this role even if it operates on the basis of weighted voting, but its character as a democratic forum declines as voting power becomes more concentrated. At the extreme, when decisions can be pushed through by only a small fraction of the membership, then the largest vote-holders have few incentives to consider the views of the rest of the membership.

Formal rules aside, the culture of the board should encourage meaningful participation, debate, and the voicing of dissenting viewpoints. The chairman should have an explicit mandate and incentives to stimulate and facilitate board debate, as well as to protect the rights of minorities or dissenting voices. Also, members of the board should be able to dissent without fear of retribution—in boards where a “chilling effect” is present, formal guarantees of open debate count for little.

Trade-Offs

Tensions exist among each of the four roles outlined above, because the characteristics required for a board to perform each of the four roles sometimes conflict. For example, a board that functions as an effective strategic thinker prizes debate, expertise, distance from day-to-day management, and independence, but it sacrifices voice and representation by requiring a small number of directors and a lean decision-making structure. A board that serves effectively as political counterweight values close involvement in day-to-day management and a close relationship between the board and political authorities. All this comes at the expense of independence and the distance necessary to think strategically. Meanwhile, a board that serves effectively as a democratic forum prizes open debate, voice, and representation, but sacrifices a significant measure of decision-making efficiency. Finally, a board that serves as a good performance police, in its pursuit of institutional accountability, may reduce the political maneuvering room that members require to align the organization’s policies with their own national interests.

These tensions among the four roles of the board suggest that no unitary executive board can perform all four roles effectively at the same time. Trade-offs are inevitable, and therefore organizations trying to balance effectiveness, efficiency, accountability, and representation must make choices that inevitably strengthen some board roles but weaken others.

Measuring Board Capacity to Play Its Roles

How can we evaluate which roles an organization’s executive board is best equipped to play effectively? In this section, I develop a set of indicators to measure the institutional characteristics necessary to support each role. The proposed indicators and the rationale for their selection are listed in Table 1. These indicators can now be used to make judgments about whether international organizations, including the IMF, are well structured to perform the four roles outlined above. However, they are not meant to measure actual performance, but whether institutional characteristics support certain board functions.

Table 1.Indicators for Measuring Capacity to Play Board Roles
Role of the BoardIndicatorRationale
Political counterweightNumber of single-country directors (or directors of multicountry constituencies in which a single country is dominant) as percent of total.Single-country directors are more likely than multiple-country directors to be influenced by their governments, regardless of whether they are appointed or elected. Therefore, the larger the share of single-country directors, the greater the degree of direct political control by shareholders and the lower the degree of board autonomy.
Mandated term length.The shorter the term of office, the lower the probability that directors will develop the knowledge and credibility within the institution to operate autonomously from their capitals; also, the lower the probability that directors will “go native” and side more closely with management and staff.
Actual length of directors’ terms of service.Same as previous.
Average age of directors.The younger the director, the longer their future career back home after board service is over, and therefore the more sensitive they will be to pleasing their bosses back home; younger directors are likely to have less room for independence than more senior ones.
How can directors be removed by their national authorities?The more easily directors can be removed, the more sensitive they have to be to the interests of their authorities, and the more closely they will represent the views of their capitals.
Are qualifications for directors specified?The more specific are director qualifications, the more difficult it is for governments to appoint directors purely on the basis of political loyalty.
Staff size of directors’ offices.The more staff and resources directors have, the greater their capacity to gather and process information about the activities of management and staff.
Annual cost of running the board (as a per cent of net administrative budget).Same as previous.
Democratic forumRatio of board size to total membership.The closer this ratio is to 1, the greater the capacity of any one member to participate directly in board discussions.
If there are multi-country constituencies, average number of countries per constituency.The larger the constituency, the greater are the demands placed on the director’s time and resources to consult constituency members, and the more difficult it is to represent their interests and make their voice heard effectively on the board.
Is there a rotation system within constituencies?Rotation schemes give members in constituencies more opportunities to have direct representation on the board.
Voting system (egalitarian or weighted).The more egalitarian the voting system on the board, the greater the incentive members have to consider the views of their peers, as they need to build majorities to make decisions.
Minimum number of countries/board directors needed to secure a majority of the required voting power, as a percentage of total membership/total directors.The larger the required minimum, the greater the incentive members have to consider the views of their peers, as they need their vote to secure a decision.
Are special majorities required for certain kinds of decisions?The higher the special majorities (e.g., 60%, 75%, 85%), the greater the incentive members and directors have to consider the views of their peers.
Does board take formal votes or does it operate on the basis of consensus?If the board operates on consensus, the greater the incentive members have to consider the views of their peers.
Strategic thinkerBoard size.The smaller the board, the higher the quality of interaction among directors and the more efficient the decision-making process, which makes strategy formulation easier.
Meeting frequency.The less frequently the board meets, the farther removed it is from the day-to-day business of the institution, and the better its vantage point for strategic thinking (though at some point, lack of familiarity with the institution becomes a problem).
Mandated term length.The longer the term of office, the higher the probability that directors will have institutional knowledge and expertise, both necessary for effective strategy formulation.
Actual length of service of directors.Same as previous.
Average age of directors.Same as previous.
Are qualifications for directors specified?The more specific are director qualifications, the greater the probability that directors will be appointed based on merit and expertise.
Annual cost of running the board (as a per cent of net administrative budget).Same as previous.
Performance policeCEO is also chairman of the Board?If the two roles are fused, lines of responsibility becomes blurred and evaluation of the CEO by the board becomes more difficult.
Does the board have a formal review process for the CEO’s performance?Board must have access to information about the CEO’s performance to evaluate performance.
Are there performance standards for the CEO?CEO must be aware of the standards by which he/she will be judged.
Is the CEO required to report on his/her performance to the board?Same as previous.
Can the board reward/penalize CEO for his/her performance?Board must be able to create incentives for good CEO performance.

Assessing the IMF Executive Board

In this section, I turn to the IMF’s Executive Board and apply the indicators just identified. The argument here is that as originally designed, the IMF Board was best equipped to serve the roles of strategic thinker and democratic forum. The Board was less well equipped to serve as a political counterweight, and it was least equipped to play the role of a performance police. Over the succeeding 60 years, however, its capacity to serve as strategic thinker and democratic forum weakened steadily, while the Board’s capacity to serve as political counterweight strengthened significantly. The Board’s potential to act as performance police—never strong—did not improve over time.

When considering the strengths and weaknesses of its Board, the IMF’s mandate should be kept in mind. Originally set up as guardian of the postwar system of fixed exchange rates, after 1971 the Fund’s main activities were three—lending members Fund resources to overcome balance-of-payments difficulties, conducting regular surveillance of members’ economic policies (through so-called Article IV consultations) and of the world economy, and providing technical assistance to members.

Political Counterweight and Strategic Thinker

From its inception, the IMF’s Executive Board was meant to serve as the institution’s primary locus of decision making. Under the Fund’s Articles of Agreement, the Board was made responsible “for conducting the business of the Fund” and for exercising the powers delegated to it by the institution’s highest governance organ—the Board of Governors (Articles of Agreement, Article XII, Section 3(a)). At their first meeting in 1946, the Governors delegated to the Executive Board almost all their powers.11 The Managing Director, who is the chief executive officer of the institution, acts under the “general direction” of the Executive Board.

Figure 1 in Chapter 1 of this compendium illustrates in a stylized manner the governance structure of the IMF, including its key formal and informal governing bodies. The Board of Governors, at the top, is the highest governing body. The International Monetary and Financial Committee (IMFC), composed of a subset of 24 governors, is an advisory body to the Board of Governors. The IMFC (in its previous incarnation, the Interim Committee) was not part of the original governance structure, but was established in the 1970s. At the center are the Executive Board and the Managing Director, who chairs the Board and is in charge of the staff. The membership is represented in the Board of Governors, the IMFC, and the Executive Board. On the left are informal country groupings (the so-called “Gs”), which have played an important but informal role in steering policy and strategy and the IMF.

Not surprisingly given its position in the governance structure, the character of the Executive Board was controversial among the Fund’s founders. Would executive directors would be government representatives tasked with ensuring that all Fund decisions were in accord with their national priorities, or would they be relatively independent “wise men,” overseeing the institution from a distance but leaving most of the Fund’s work to the staff’s technical expertise? Keynes, who represented the British Treasury at Bretton Woods, endorsed the latter option:

  • Some of us . . . had been hoping that the officials of the two bodies [the Fund and World Bank] would, in the course of time, come to regard themselves as primarily international officials, taking a world objective outlook, and only where clearly necessary grinding their own national axes. So one would have wished to minimize rather than maximize, their national representative character and their position as delegates from outside authorities. (Quoted in Hexner, 1964: 84.)

Keynes hoped to endow the Board with some of the characteristics we have already identified as necessary for the “strategic thinker” role and to minimize its character as political counterweight. He lobbied hard for a non-resident, high-level board, composed of senior officials from national treasuries and central banks. They would be “deputy governors of central banks” or “very responsible people in the heart of their own institutions” (Boughton, 2001: 1032). Directors would only serve the Fund on a part-time basis and would not be immersed in the day-to-day operations of the institution; they would be close to policymaking in their own capitals, but would be senior enough to be able to take independent stances when necessary.

However, the U.S. Treasury preferred board characteristics that accorded more closely with those of a political counterweight, and in the end, this vision prevailed. The result was a resident, twelve-member board based in Washington, D.C., and meeting “in continuous session.” It was composed of full-time executive directors who met regularly some three times per week, on average. Because they would be based in Washington and occupied full time at the Fund, directors would not be senior officials in their governments (though they could be former senior officials). While the Articles of Agreement specified that the Managing Director and members of the Fund’s staff “shall owe their duty entirely to the Fund and to no other authority,” there was no requirement that individual Directors owe their allegiance entirely or partially to the Fund (Articles of Agreement, Article XII, Section 4(c)).12 The Board was charged with making all decisions on bilateral surveillance (Article IV consultations) and the use of Fund resources.

In addition, the five members with the largest voting shares—the United States, United Kingdom, France, India, and China—were given the right to appoint their own executive directors. (India and China were later replaced by Germany and Japan in exercising this privilege.) These five directors served at the pleasure of their governments and could be dismissed at any time for any reason. The remaining seven directors represented the rest of the Fund’s 39 member countries, which were organized in multi-country “constituencies.” Directors representing constituencies were elected by the group for renewable two-year terms, and legally could not be removed until their term expired. No qualifications for executive directors were specified in the Articles of Agreement.

Despite the success of the political counterweight model, the Executive Board in its early years had several characteristics of the strategic-thinking board. With twelve directors, it resembled in size today’s corporate boards. Also, the first generation of directors was a very experienced group; its members had left very senior posts in their governments before coming to the Fund.13 Their attendance at the Board was poor—which suggests that, in practice, the early Board resembled the non-resident board that Keynes had envisaged.14 Finally, thanks to the relatively impractical and expensive communications technology of the time, directors enjoyed considerable autonomy from their capitals.

Over the next 60 years, the character of the Board changed considerably. The Fund’s membership quadrupled to 185, while the size of the Board doubled to 24 directors. The five largest shareholders retained their own directors, and three additional members—Russia, China, and Saudi Arabia—chose to elect a director to represent them alone.

Technology changed rapidly as well. The advent of fax machines and eventually cellular telephones and e-mail strengthened the capacity of governments to monitor and steer the activities of their directors in Washington. Capitals could now communicate instantly with their directors and could also review electronically—in real time—the same Board documents their directors were reading. This reduced directors’ latitude to act autonomously.

As the membership grew, the volume of the Fund’s surveillance, technical assistance, and lending work multiplied. The Board gradually shifted from a decision-making, “executive” body into one that could only review and approve decisions by Management and staff on the basis of relatively superficial analysis and discussion. The Board was forced to devote more and more of its time to the day-to-day business of the Fund and less to strategy formulation and to monitoring policy implementation. Constantly immersed in detail, the Board lost some of the perspective needed to think about the “big picture” issues confronting the Fund in a changing world economy.

What about the length of directors’ terms of service? In the past two decades, actual terms of service have fluctuated considerably, but the mean “age” of the Board—the average amount of time directors have served on the Board at a given point in time—has declined by nearly a year to just under 40 months, as shown in Figure 2. These numbers are skewed by a handful of directors who have remained on the Board for extraordinarily long periods, however.15 If we take out these outliers and look at median tenure, the number is about 23 months in the 1990-2007 period.16 This means that although directors’ terms are renewable, in practice few countries or constituencies keep their directors in place for more than their initial two-year terms. As we will see, these terms are shorter than those of directors in most other IGOs studied here.

Figure 2.IMF Directors’ Length of Time in Office, 1990-2007

Source: Based on data from IMF Secretary’s Department.

This degree of Board turnover means that by the time directors have mastered the complexities of Fund operations, they have little time left to put their expertise to productive use. It also means that the Board depends heavily on two or three long-serving directors who are repositories of institutional knowledge, and that when they leave, the Board suffers a sharp decline in expertise and human capital. All this hinders the capacity of the Board to think strategically about the direction of the institution. It also makes directors more dependent on instructions from capitals and on the views of staff and management.

Democratic Forum

What about the Board’s role as a democratic forum? The Board began as a compact body where aggregating and voicing members’ positions was relatively easy—a dozen directors represented 44 member countries, and multi-country constituencies represented, on average, around 5.6 countries. With the quadrupling of the Fund’s membership and the doubling of the Board’s size, voice and representation became more difficult. The ratio of Board to membership size fell from 0.27 in 1946 to 0.13 today. The average size of a multi-country constituency grew to 10.8 countries, and the median size to nine (the range is four to twenty-four countries). The problem of crowded constituencies was compounded by the increase in the number of single-country constituencies from five to eight—a third of the Board’s seats.

When the Fund was founded, the distribution of voting power among individual chairs was highly unequal.17 Just three chairs (those controlled by the United States, United Kingdom, and France), or about 6 percent of the membership at the time, held over 50 percent of the voting power. Today, voting power is less concentrated, but remains very unequal. Voting power ranges from 16.9 percent for the U.S. chair to 1.4 for the largest African constituency. Assuming everyone casts a vote, support of at least eight chairs representing about a fifth of the total membership is enough to secure a majority of the voting power. While special majorities (of 70 and 85 percent) are needed for some decisions, a simple majority is sufficient for most decisions, including many of the most important ones involving the ordinary business of the Fund, such as the use of the institution’s resources.18

Because most countries are represented on the Board as part of multi-country constituencies, the practices within constituencies are critical to the quality of representation (Woods and Lombardi, 2006; Martin and Woods, 2005.) Whether the words and actions of a director representing a multiple countries faithfully reflect the views of the governments represented depends on a variety of factors, including the number and diversity of the countries in the constituency, the distribution of voting power within the constituency, and the “culture” of the constituency—that is, the formal and informal consultation mechanisms that have developed over time among the members.

Voting-power inequality within constituencies is significant. In three constituencies—those chaired by Italy, Canada, and India—voting power is highly concentrated; the largest vote-holding member has more than 75 percent of the constituency’s votes. In another six constituencies, the largest vote-holding member has between 40 and 75 percent of the votes, and in seven constituencies the vote distributions are more egalitarian (see Annex). In eight of the 16 constituencies, the largest member has more than twice the voting power of the second largest member.

Potential gaps in voice and representation are especially acute in the eight constituencies that mix countries that use Fund resources and those that do not. Here, the interests and preferences of member states are more likely to conflict. In these constituencies, the quality of representation for the Fund’s smallest (and often poorest) shareholders depends largely on whether the dominant countries in the group select a director who is interested in playing the role of active and fair representative. This can often be a matter of luck, rather than institutional design.

Another important factor in the quality of voice representation for the smallest members is the personal judgment of the chair of the Board (i.e., the MD) who plays a crucial role, as he is responsible for determining the “sense of the Board” during meetings and deciding when consensus has been reached on a particular decision. Thus, the MD’s role as protector of minority voices is key—the MD can force through decisions strictly on the basis of simple majorities, or he can work to build wide agreement or to postpone a decision until this emerges. Another crucial aspect of voice is the preparation of the “summing up,” as the main document that captures where the Board stands on a certain issue or decision. It is largely up to the Chair, assisted by the Secretary, to determine the extent to which minority viewpoints are reflected in the summing-up of a meeting.19

Performance Police

As performance police, the IMF’s Board is, and always has been, poorly equipped. According to the Articles of Agreement, the Managing Director operates under the “general direction” of the Executive Board. However, the Articles are silent on whether and how the MD’s performance should be evaluated. There are no performance standards, no reporting requirements, no formal performance review, and no performance contract.20

The only relevant innovations in this area have been the introduction of a codes of conduct for staff (1998) and for Board members (2000). The terms of appointment of the current MD specify that he must abide by the staff code of conduct.21 A Board Ethics Committee was also established to oversee the implementation of the Board’s code of conduct. The Board itself has no self-evaluation process, nor is its performance evaluated by any other body other than the extent to which members evaluate the performance of the Directors which represent them.

There are at least three reasons for this gap in Fund governance. The first is the relative difficulty of producing performance benchmarks for an institution with multiple functions as diverse as surveillance, lending, and technical assistance. Unlike for a business firm, there are no simple metrics such as price-to-earnings ratios or profits with which to measure Fund performance.

The second problem has to do with blurred lines of responsibility. The Board and the MD exercise “separate but closely related powers,” and the Board is ultimately responsible for determining the precise scope of the MD’s powers (Gianviti, 1999: 49). In practice, however, this is not a neat distinction. In his dual roles as CEO and chair of the Board, the MD does not simply take the Board’s decisions and execute them. The MD also helps shape those decisions, advises the Board, lobbies directors in private, has significant control over the Board’s agenda, and ultimately—as the chair of the Board—determines when a decision has been made. This overlap means that the Board cannot pass judgment on the MD’s performance without a conflict of interest, unless it evaluates Management in areas where the MD has sole responsibility.

The third problem is that, while the Board technically appoints the MD, in practice the selection process has historically been opaque and ultimately determined by negotiations among G-7 members and other European shareholders.22 Presumably, the removal of an MD would require a similar negotiation among major shareholders. This means that the Board is not in a position to objectively pass judgment on the MD nor to reward or sanction him for performance.

In conclusion, today’s Board is best equipped to serve as a political counterweight, and the characteristics that support that role for the Board have strengthened gradually since the Fund’s creation. The characteristics supporting the Board’s role as democratic forum have deteriorated over time, largely as a result of membership enlargement and the expansion of constituency size. Voting power has become more diffuse, but remains highly unequal. The characteristics supporting the Board’s role as strategic thinker have also eroded over time, and today this is one of two roles for which the Board is least prepared, largely as a consequence of its size and high turnover. Finally, the Board was never well equipped to serve as performance police, and today remains least well prepared to carry out this role.

The IMF in Comparative Perspective

Having examined the IMF’s Board in some detail, in this section I place the Fund’s governance arrangements in a wider context. I focus on a sample of eleven international organizations, chosen because they operate in the same or similar sectors as the IMF and because they share at least one of the Fund’s three institutional functions—surveillance, provision of technical assistance, and lending.

The sample includes six multilateral development banks (MDBs), including the World Bank, whose governance structure closely resembles that of the Fund. The sample also includes the Bank for International Settlements (BIS) and the Organization for Economic Cooperation and Development (OECD). Like the Fund, these two organizations are involved in the surveillance of international financial markets and national economic policies, respectively. Also included are three IGOs that, like the Bretton Woods institutions, have near–universal membership, though they operate in sectors other than international financial and monetary affairs. Two of these—the United Nations Development Program (UNDP) and the World Health Organization (WHO)—perform surveillance and provide technical assistance, like the Fund. The Global Environment Facility (GEF) is different from the other organizations listed here because most of its financing is disbursed as grants, not loans. However, the GEF is included because it offers one of the more innovative governance structures among IGOs. The full sample is shown in Table 2.

Table 2.Sample of Inter-Governmental Organizations and Functions Shared with IMF
Function
OrganizationPolicy AreaSurveillanceTechnical assistanceLending
International Monetary FundInternational finance
United Nations Development ProgramDevelopment, trade, and investment
Organization for Economic Cooperation and DevelopmentDevelopment, trade, and investment
World Health OrganizationGlobal health
Bank for Internationa SettlementsInternational finance
World BankDevelopment lending
African Development BankDevelopment lending
Inter-American Development BankDevelopment lending
European Investment BankDevelopment lending
Asian Development BankDevelopment lending
European Bank for Reconstruction and DevelopmentDevelopment lending
Global Environment FacilityEnvironmental protection1

The World Trade Organization (WTO) was not included in the sample because from a governance standpoint, it differs from the rest of the organizations studied here. In those organizations, member states delegate significant authority to an executive body and a CEO. In contrast, the WTO is a “member-driven” organization in which little authority is delegated to the Secretary General and the Secretariat. Instead, nearly all the WTO’s councils and committees—including the General Council, which handles WTO’s day-to-day operations—are plenary committees, which means that decision-making always involves representatives from each of the 150 members. The absence of a non-plenary executive body has been identified as one of the most important limitations on the capacity of the WTO to make decisions efficiently (Sutherland and others, 2004: Chapter VII). These unique characteristics make the WTO difficult to compare meaningfully with the rest of the IGOs in the sample, where the delegation of authority is a key feature. The WTO is therefore left out of the analysis, though references are made to it at several points.

Three Models of Governance

How to compare and contrast meaningfully this very diverse set of IGOs? I classify them based on the same executive-board characteristics that were outlined above and applied to the IMF. The result is that the eleven organizations fall into three categories, or “models” of governance, each with a different configuration of strengths and weaknesses. I call the three models the (1) delegate-and-control model, (2) the direct representation model, and (3) the constituency-based oversight model.

Delegate-and-Control Model

The organizations in this category include both the World Bank and the Fund, as well as major regional development banks—the Inter-American Development Bank (IADB), the African Development Bank (AfDB), the Asian Development Bank (AsDB), and the European Bank for Reconstruction and Development (EBRD). The pioneers of this model were the architects of the Bretton Woods institutions, but the model was adopted and replicated by the founders of regional development banks in the 1950s and 1960s.23

The central feature of this model is that power and representation are delegated to a relatively small executive board that exercises close control over the activities of the institution. Specifically, organizations based on the delegate-and-control model have the following characteristics: (1) a compact executive board (relative to the total membership size) whose members are elected or appointed by member countries, and which is in continuous session (resident board); (2) a system in which most members are represented indirectly through multi-country constituencies and share a single director; (3) a CEO who is also chair of the board, and (4) a decision-making system based formally on “consensus” but underpinned by weighted voting. Table 3 provides key indicators for the five IGOs in the sample that fall into this category, including the IMF.

Table 3.Selected Indicators for Inter-Governmental Organizations Following the Delegate-and-Control Model
IMFWorld BankAsDBAfDBIADBEBRD
Membership size (number of countries)18518566774763
Staff or secretariat size2,60013,0002,0001,0001,8501,570
Size of executive board242412181623
Ratio of board size to total membership0.130.130.180.240.340.37
Frequency of board meetings3/week2/week1-2/week1/week2/week1-2/mo
Annual cost of running the board (as a % of net administrative budget), 200616%3%7%n/a4%5%2
Mandated terms of office for directorsNo term limits for appointed EDs; 2 year, renewable terms for elected EDsNo term limits for appointed EDs; 2 year, renewable terms for elected EDs2 years, renewable3 years, renewable once3 years, renewable3 years, renewable
Average age of directors535354n/an/a55
Voting systemWeightedWeightedWeightedWeightedWeightedWeighted
Resident or non-resident board?ResidentResidentResidentResidentResidentResident
Number of single-country chairs as a % of the total33%33%25%5.5%12.5%35%
Average size of multi-country constituencies10.910.97.24.53.83.7
Highest number of countries represented by a single director242411979
Minimum number of countries needed for a simple majority of voting power, as a % of total membership318.1%18.1%41.8%36.2%10.6%9.5%
Minimum number of directors needed for a simple majority of voting power, as a % of total directors433%33%50%45%19%26%
CEO is also chairman of the board?YesYesYesYesYesYes
Performance standards for CEO?NoNoNoNoNoNo

While there are subtle differences among the five MDBs that adhere to the delegate-and-control model, some useful generalizations are possible. As the name suggests, executive boards following this model are best equipped to perform the role of political counterweight. Small boards and weighted-voting systems allow for efficient decision-making, and executive directors function primarily (and often exclusively) as representatives of their member countries.24 Communication and relations between directors and their capitals tend to be frequent and close. As members of resident boards, meeting one to three times per week, directors are closely involved in most aspects of their organization’s policy and operations. Directors in all MDBs also have their own staff, which increases their capacity to collect and process information about what is happening in the organization. This level of involvement is reflected in the resources the boards consume as a proportion of the organizations’ net administrative costs—between 4 and 7 percent, as shown in Table 3.

Certain characteristics of this model suggest that directors have relatively little autonomy from the countries they represent. Directors are typically officials in their early fifties, which means that they still have future career plans that they must be concerned about when they return to their capitals. Mandated terms of service are short (two to three years), and many directors serve only one term. Qualifications are not specified in the charters or are described only in general terms, typically with the phrase “directors shall be persons of high competence in economic and financial matters”. This allows members wide latitude in whom they select. Also, up to a third of all directors represent only one country, which means that they are likely to be closely controlled by their capitals.

Executive boards in this category are not well suited to play the role of strategic thinker. While some smaller boards may facilitate high-quality interaction among directors, most boards are significantly larger, especially those of the Bretton Woods institutions. In addition, all of these resident boards are too closely engaged in the day-to-day business of the institution to have good strategic vantage point. Finally, low levels of board independence render these boards effective political counterweights, but because they are constantly focused on attending to the interests of their governments, directors have less time and freedom to think strategically from the perspective of the institution as a whole.

As democratic forums, boards in this category are also relatively ineffective. Because they are small relative to the overall size of the membership, the voice and voting power of small shareholders is diluted in multi-country constituencies, whose size ranges from 3.7 to 10.9 countries per constituency, on average. With the exception of the EBRD and the AsDB, where vote-splitting is allowed, countries in these constituencies must share a single director, who casts the constituency’s votes as a single unit.

Small boards and weighted voting mean that a few large shareholders may exercise considerable influence. Concentration of voting power is most dramatic in the IADB and EBRD, where a majority of total voting power is held by only 10 percent of the membership (or a fifth and a quarter of directors, respectively). To secure a simple majority in the Bretton Woods institutions requires support from as little as 18 percent of the membership. By contrast, in the African and Asian development banks, voting power is significantly more diffuse. To be sure, the boards of all of these MDBs operate on the basis of “consensus” and formal voting is rare; however, the consent of the largest shareholders is usually necessary, particularly on controversial issues, and the concentration of voting power still affects decision making, albeit in a subtle way.

The weakest role of these boards is as performance police. Their charters do not set forth an evaluation mechanism for the CEO, and in practice, none has performance standards for management or a formal process of evaluation. As already discussed in the case of the IMF, this is partly because identifying practical performance measures is difficult; the actions of the CEO and the board are not easily separable (especially since the CEO chairs the board) and because the CEO often is not chosen by the board in practice.

Direct Representation Model

Organizations in the second category follow what I call the direct representation model. Three organizations in our sample adhere to this model of governance: the European Investment Bank (EIB), the Organization for Economic Cooperation and Development, and to a lesser extent, the Bank for International Settlements (BIS). Selected indicators for these organizations are found in Table 4 above.

Table 4.Selected Indicators for Inter-Governmental Organizations Following the Direct-Representation Model
EIBOECDBIS
Membership size (number of countries)283055
Staff or secretariat size1,3302,500550
Size of executive board283121
Ratio of board size to total membership1.001.031.00 (founding members)
Frequency of board meetings10/year12/year6/year
Annual cost of running the board (as a % of net administrative budget), 20061>1%n/a1.4%
Mandated terms of office for directors5 years renewableAt the discretion of each government; in practice, ambassadors have served about 3.5 years, on averageThe 6 ex-officio directors are appointed for their terms as central bank governors; the rest are appointed for a renewable 3-year term2
Voting systemDouble-majority3Simple majority; one country, one vote; QMV for key issues4Simple majority; one board vote per board member5
Resident or non-resident board?Non-residentResidentNon-resident
Number of chairs representing single countries as a % of the total96% (one represents the European Commission)97% (one represents the European Commission)100%
CEO is also chairman of the board?YesYesNo
Performance standards for CEO?NoNoNo

Admittedly, these three institutions are very different from each other. The Luxembourg-based EIB is the world’s largest multilateral development bank, and it has adopted governance arrangements that vary in significant respects from those of its peers. The OECD is best described as a research organization and as an institutional platform that supports and coordinates an extensive web of technical networks and committees. Finally, the BIS—often called “the central bankers’ central bank”—was chartered as a private company and is best known today for its surveillance of the international financial system, its research and standard-setting activities, and for its role as a meeting place for central bank governors. These organizations are also diverse in terms of their governance arrangements. The EIB and BIS have non-resident boards composed of senior government officials, while the OECD has a Council composed of resident ambassadors.

But despite their differences, all three organizations share the basic elements of this governance model: (1) a “plenary” executive body in which all members are directly represented; (2) a board or equivalent that meets only a few times per year, typically monthly or bi-monthly; and (3) voting systems that either rely completely on the principle of one-nation-one-vote or combine it with some form of double-majority voting. The characteristics of the direct representation model weaken somewhat the board’s role as political counterweight, especially when compared with the delegate-and-control model. Meeting once per month at most, these boards are relatively distant from the operations of the institution and leave more of the day-to-day business to the management. This is especially true of the BIS, where the central bank governors who constitute the board come to Basel every two months and have little to do with the management of the institution; this is left to the General Manager, who reports regularly to (and does not chair) the Board. The EIB’s Board meets more frequently and takes a more active role in management, but much less so than in other MDBs—indeed, EIB is the only one of these organizations with a non-resident Board. The less intensive engagement of these boards is reflected in the costs of running them—the cost at both EIB and BIS is less than 1.5 percent of the administrative budget of each institution.

The OECD’s Council is more involved and considerably more costly. It has resident status and large ambassadorial support staff. However, with monthly meetings, the Council is not nearly as involved as the boards of the IMF or the World Bank.

Perhaps because member states in this model exercise less direct control over the institution at the board level, governments have devised other ways to exert control, usually further down the chain of delegation. The EIB exemplifies this point. At the EIB, the Board is non-resident and relatively removed from day-to-day affairs, and the business of the institution is conducted by a nine-member Management Committee composed of the President and eight vice-presidents. Management Committee members are elected by the Board of Governors, and they represent specific countries or constituencies of countries based on formal nationality requirements.25 Presumably, formal nationality rules mean that the members of the Management Committee are more likely to act on the basis of their governments’ national interests than are the members of organizations in which senior management figures do not face formal nationality quotas. This contrasts with the delegate-and-control model, where the board exercises political control, and the decisions of management and staff are less likely to become politicized.

Similarly, at the OECD, the Council may meet only on a monthly basis, but national politics penetrate more deeply into the structure. Much of the organization’s work is prepared by staff working closely with committees, which are composed of representatives from capitals; government officials from member countries are present at the organization’s working level.

Two factors make these organizations better equipped for strategic thinking compared to those following the delegate-and-control model. First, greater distance from day-to-day management allows their boards to focus better on strategic issues. Second, board members stay longer in their posts, which gives them more expertise and institutional knowledge. EIB directors serve renewable five-year terms (in practice, they tend to serve for more than five years). The core members of the BIS board (more on what this means below) are elected for the entire duration of their terms as central bank governors, which in practice can exceed a decade, and the elected members of the BIS board have renewable, three-year terms. At the OECD, ambassadors serve at the pleasure of their governments, but in practice, OECD ambassadors remain at their posts for long periods—since the mid-1980s, the average term of an OECD ambassador has been 41.4 months, or almost three and a half years.26 However, there is a trade-off between direct representation and strategic thinking. At between 21 and 31 members, these boards are too large for efficient decision-making and strategic planning.

At the same time, the boards of these institutions are well suited as democratic forums. In the OECD and EIB, all members are directly represented at the board, and double-majority voting (DMV) schemes magnify the voice of smaller shareholders and guard against powerful minorities pushing through decisions opposed by the majority of the members.27 Double-majority voting is a recent innovation in both institutions. Through DMV, the members hope to keep decision making efficient despite the addition of new chairs, while preserving a degree of representation and ownership. To date, the mechanism has not yet been invoked at either organization, but its existence—and the possibility that a vote might be called—has reportedly changed the dynamics of decision making by forcing the biggest financial contributors to take into account the voices of other countries.

The BIS is the least well equipped to act as a democratic forum. In practice, the BIS implicitly retains a three-tiered membership structure, with each tier enjoying a different level of representation on its Board. Permanent direct representation (and a majority of the votes) is guaranteed only for the six founding (“ex-officio”) members. Countries in a second tier (up to nine) are elected to the Board for three-year renewable terms. The other 38 central banks that are members of the BIS are in a third tier and do not have representation on the Board. The BIS thus fits under the direct-representation model only to the degree that its founding members enjoy direct representation.

In terms of policing performance, IGOs following the direct representation model are in some respects better positioned than their MDB counterparts to evaluate and judge management’s performance, because their lines of accountability are clearer. At the BIS, the separation of the roles of CEO and chairman, complemented by regular reporting by the CEO to the Board, the arms-length involvement of the Board in management, and the seniority of board members, renders the CEO relatively accountable. At the OECD and EIB, the CEO and board chair positions are fused, but the distance of the Council and Board from management makes the actions of the CEO more easily separable from those of the board. However, none of these institutions uses performance measures for the CEO.

The direct representation model makes most sense for “peer group” organizations—IGOs with memberships of relatively few, like-minded states. Small peers groups can afford to have everyone represented on the executive body without risking paralysis. The three organizations just discussed reflect this: their relatively small memberships consist of advanced or transition economies, largely or exclusively from Europe.

Constituency-Based Oversight Model

This model of governance is common among United Nations agencies with large memberships (more than 170 member states), such as the United Nations Development Program (UNDP) and the World Health Organization (WHO). Some organizations outside the UN system, such as the Global Environment Facility, have also adopted it. As in the delegate-and-control model, member states delegate power to a non-plenary board, and members are represented through constituencies. However, these organizations have several distinguishing features: they have (1) executive bodies that are large in absolute terms but small relative to the size of the membership; (2) non-resident boards that meet only two or three times per year; (3) board directors who represent constituencies with rotation schemes; (4) one-nation-one-vote or double-majority voting systems; and (5) separate CEOs and board chairs. Table 5 shows selected indicators for the organizations following the constituency-based oversight model.

Table 5.Selected Indicators for Inter-Governmental Organizations Following the Constituency-Based Oversight Model
WHOCEFUNDP
Membership size (number of countries)193177192
Staff/secretariat size8,0006017,000
Size of executive board343236
Ratio of board size to total membership0.170.180.19
Frequency of board meetings2/year2/year3/year
Annual cost of running the board (as a % of net administrative budget), most recent year availablen/an/an/a
Mandated terms of office for directors3 years, renewable3 years, renewable3 years, renewable
Voting systemOne country, one vote2Double majority3One country, one vote4
Resident or non-resident board?Non-residentNon-residentNon-resident
Number of directors representing a single country as a % of the total0%31%0%
Average rotating constituency size5.657.665.37
CEO is also chairman of boardNoOn occasionNo
Performance standards for CEONoNoNo
Mandated reporting by CEONoYesNo

How does this governance model affect the board’s role as political counterweight? Directors in organizations following this model are nonresident and there is no requirement that they owe their primary loyalty to the organizations. Some of these organizations have explicitly recognized that directors are delegates representing their national governments.28

Despite the proximity of directors to capitals, several characteristics significantly weaken the political counterweight role of these boards. The institutions’ non-resident boards, meeting twice or thrice per year, are too far removed from the day-to-day business of the organization to be able to focus on anything but the most strategic, highest-level issues. Without staff or offices, the directors have little capacity to collect or process information about the organization’s work. Directors are elected, not appointed, by single governments which weakens the degree of political control that any single capital can exert over them.

Yet, the characteristics that weaken the political counterweight role do not result in a strong strategic-thinking role. At between 32 and 36 directors, these boards are larger than those in the organizations covered thus far—too large to serve as effective forums for strategic thinking. Also, while the official tenures of directors are longer than in the Bretton Woods institutions, turnover is in fact higher because of mandated rotation schemes. This contrasts with the IMF and the World Bank, where a handful of directors tend to stay on for very long tenures and become repositories of institutional knowledge. In practice, the boards in the constituency-based oversight model must rely heavily on the CEO to think about strategy and make concrete proposals to the board.

As democratic forums, these boards are more effective at accommodating near-universal memberships than those in the delegate-and-control model. With larger boards and relatively few or no single-country chairs, members are part of smaller constituencies (between 5.3 and 7.6 countries per constituency, compared with 10.9 for the IMF and World Bank). Also, formalized rotation schemes provide regional balance and give every member a chance to serve on the board. Most importantly, the one-country-one-vote system of the WHO and UNDP, as well as the double-majority voting system of the GEF, ensure that the voices of all or most members count.

Finally, the board’s role as performance police in organizations following the constituency-based oversight model is potentially more effective than in the delegate-and-control model. The separation of the CEO and board chair roles and the arms-length engagement of the board produce clear lines of responsibility, with the board instructing and supervising and the CEO implementing. In practice, however, the IGOs do not have a formal process for evaluating the CEO. There are periodic reports by the CEO to the board (the GEF, in particular, requires the Secretariat to report to the Assembly and to the Council), but no performance criteria or formal review process.

Looking Across Models

Summarizing the main characteristics of all three models in a single table (Table 6), we can now compare the relative strengths and weaknesses of the models in terms of the four roles that boards can play. The delegate-and-control model is the strongest when it comes to the board’s role as political counterweight, with the direct representation model in second place. As democratic forums, the direct representation and constituency-based oversight models have the most potential, though they were conceived for two different orders of magnitude in membership size. In terms of strategic thinking, the direct representation model is the least inadequate. Performance police is not a role that IGO boards perform well in general, but among the three models, the least poorly suited for this role are the direct representation and constituency-based oversight models.

Table 6.Rating the Roles of the Executive Board by Governance Model
Role of the Board
Political counterweightDemocratic forumStrategic thinkerPerformance police
Governance ModelDelegate-and-controlStrongMediumWeakWeak
Direct representationMediumStrongMediumMedium
Constituency-based oversightWeakStrongWeakMedium

Conclusions and Lessons for the IMF

What does this comparative exercise tell us about the governance of the IMF? First, it helps us place the IMF in a larger constellation of IGOs, both with similar and different models of governance. The main findings are the following:

  • The IMF’s Board has characteristics that, at least in theory, make its decision-making relatively efficient among IGOs with large memberships. Of the five organizations in the sample with near-universal memberships, the IMF and the World Bank have the smallest boards. Also, the Fund and the Bank have the lowest ratio of board to membership size of any IGO in the sample.

  • The features that facilitate decision-making come at a cost in terms of the quality of representation and voice for at least some of the Fund’s member countries. Among IGOs that have constituencies, the World Bank and the IMF have the most single-country directors and the largest average constituencies; this dilutes the extent of direct representation that members enjoy on the board.

  • At around six percent of general administrative costs, the cost of running the IMF’s Board is relatively high when compared with other IGOs with resident boards, though not significantly out of line with that of peer institutions (the range is four to seven percent). These numbers should be interpreted with caution, given the different mandates and membership sizes of each organization.

  • The tenure of IMF directors is relatively short. Along with the World Bank and the AsDB, the IMF has the shortest mandated terms for directors, and at 25 months, the actual median term of office for IMF directors is also one of the shortest. This high turnover is partly offset by the experience that some IMF directors accumulate while serving as alternate directors.

Issues of Institutional Design

The above comparative exercise also raises two larger issues of institutional design. This paper has shown that the IMF’s governance arrangements are part of a larger universe of governance models, and that the choice of model affects the capacity of the organization’s board to perform key roles. In the case of the IMF, a key question is whether governance should remain closely wedded to the delegate-and-control model. This model makes sense for multilateral lending institutions because those who contribute the bulk of the financial resources will only do so if they can be assured a certain degree of control over their use. Not surprisingly, all of the other IGOs that use the delegate-and-control model are multilateral development banks.

But there are reasons to question the IMF’s complete adherence to the model. The Fund’s near-universal membership (as opposed to the regional memberships of most MDBs), the changing weight of some member countries in the world economy, and the Fund’s current crisis of legitimacy suggest that importing governance innovations from other models, if not a total departure from the existing model, may be in order. Also, the Fund has two other “lines of business” in addition to lending: the provision of technical assistance (a responsibility it shares with MDBs) and surveillance (which no MDB practices to the same degree). These two lines of business are arguably better served by governance models other than delegate-and-control. Surveillance, in particular, may be better served by a system in which the political counterweight role of the board is weaker, reducing political interference that has been known to water down staff analysis of member states’ economic policies and conditions. To try to undertake all three lines of business with an board that is structured to exercise political control over lending may not be the best way to operate effectively and with legitimacy.

What governance mechanisms could the Fund borrow from other models? The answer depends on how one wishes to change the configuration of strengths and weaknesses in the Board’s four roles. I consider several mechanisms below.

Strategic Thinking

If the goal is to strengthen strategic thinking at the IMF, there are two general directions. One to outsource this role to a ministerial body such as the International Monetary and Financial Committee (IMFC). The second general direction is to increase the Board’s autonomy and capacity, and to promote the board characteristics that support its role as strategic thinker.

Reducing the size of the Board or shifting to a non-resident Board are unattractive options, as these measures would further weaken the board’s role as a democratic forum. A move toward a non-resident board would likely be accompanied by demands that the political counterweight role move down the delegation chain into Management and staff, as it has at the EIB, and that Management become more representative of the membership. Management would have to expand, and formal nationality quotas and rotation mechanisms might be necessary. These measures would reduce efficiency and would increase the politicization of decisions by Management and staff.

There are, however, some intermediate measures that could strengthen the Board’s role as strategic thinker without drastic structural change. I am not recommending the adoption of all of these measures, but laying out a menu of the most promising options.

Independent/Outside Directors

Independent directors can bring external expertise to an organization, improve the objectivity of board decisions, and reduce conflicts of interest. There is only one relevant case in our sample of IGOs. In 2004, the EIB amended its Statute to allow for the addition of up to six outside experts (three non-voting directors and three alternates) to the Board.29 These experts participate in all Board meetings in an advisory capacity, without voting rights, and like other directors, they are appointed for renewable five-year terms. The stated purpose of adding outside directors is to broaden the Board’s expertise in certain fields. Interviews at the EIB suggest that the independent directors have added value to the Board’s decisions.

The introduction of independent directors to the IMF Board might offer similar benefits. Outside directors could be a mix of senior academic economists, former policymakers, and private-sector figures. They would sit on the board in a personal capacity, serving no government but only the institution as a whole. Free from influence from capitals and already at the peak of their careers, these directors would be able to provide frank opinions about country and policy issues. They could also bring much-needed expertise in specialized areas, such as financial sector policy.30 Directors from the private sector could prove especially valuable by providing the Board—which generally lacks private-sector experience—with insights about how the markets might react to Fund policies.

An alternative to introducing independent experts into the Board itself is to create an advisory council of eminent experts with whom the Board would meet periodically to receive advice. The experts would not be directors—they would be separate from the Board and not participate in Board deliberations—but the Board would still benefit from their guidance and specialized knowledge. The quality and nature of the advice the experts can provide would naturally be more limited and of a different character than if the experts were full participants in the Board discussions.

Meeting Frequency

The Board could strengthen its strategic role by delegating more to Management and distancing itself from the details of the Fund’s business. Without resorting to a non-resident Board, the Fund could cut down on the Board’s meeting time, following the examples of the EBRD and OECD. The question, of course, is what to cut.

Under the Articles of Agreement, the Board cannot delegate its powers to any other body, either within or outside the Fund. Article IV discussions would be especially difficult to delegate, because surveillance is a key function of the Board; changing this would require amending the Articles. The introduction of written statements in place of oral interventions at the Board has cut down on meeting time, but there is a limit on how much more could be gained from similar measures.

Unfortunately, other IGOs offer few good examples of how to reduce Board meeting time significantly. One idea, recently introduced at the OECD, is to give Board committees decision-making power and make it difficult for the Board to re-open issues once they have been decided by committees. But, given the Fund Board’s traditional antipathy to working in committees, this idea is unlikely to work unless the Board changes its attitude toward committees and makes more active use of them.31 The Board could also rethink the modalities through which it provides input for bilateral surveillance and for decisions involving the use of Fund resources.

Term of Office

As mentioned, IMF directors serve comparatively short terms of office. One of the simplest and most effective ways of increasing Board capacity and autonomy would be to extend the terms of office to at least three years. One downside of lengthening terms of office is that it would lengthen the time that members must wait to get leadership positions. Another issue relates to accountability. At the moment, elected directors cannot be removed during their terms of office; if the terms were to be lengthened, more robust accountability mechanisms should be introduced in parallel.

Democratic Forum

In effect, the Fund has chosen to sacrifice some of the Board’s role as democratic forum in exchange for a Board that is smaller and more efficient. To strengthen the Board’s ability to be a democratic forum, four options are especially promising. These are not mutually exclusive.

Board Enlargement

The first is to add more seats to the Board. This would inevitably erode the Board’s role as strategic thinker and increase transaction costs. On the other hand, having long passed the ideal number of 10 to 12 Board members, the marginal efficiency loss of adding one or a few more chairs might be outweighed by the gains in voice and representation.

Rotation Schemes

The second option, drawn from the constituency-based oversight model, would be to establish egalitarian rotating schemes in many or most constituencies. Director and alternate chairs would no longer be held exclusively by the largest vote-holding members of the constituency but would rotate equally among all members, regardless of voting power. The main advantage of the scheme would be a much-enhanced voice and sense of ownership of the institution by small shareholders. At the same time, the largest economies in the constituency would continue to provide much of the expertise and input, given their greater capacity to contribute. Of course, less drastic, intermediate rotation schemes that would not require the largest shareholders to surrender all of their chairs are also possible.

Reducing Single-Country Seats

As this study has shown, the Bretton Woods institutions have the largest executive board constituencies, on average. This is not only because of their small boards relative to their total memberships, but also because of the relatively large number of single-country chairs. One way of relieving this “overcrowding” would be to impose a cap on the number of countries that can be represented by a single director, forcing countries to migrate to smaller constituencies and relieving the burden of representation on the most crowded chairs.32 This approach would work best if combined with efforts to reduce the number of single-country constituencies. This would involve a delicate political deal whereby all, if not most, of the top shareholders would agree to open their constituencies to other member countries. The first step in this direction would be to abolish appointed chairs on the Board, which would open the door to the formation of new multi-country constituencies where currently there are only single-country chairs.

Double-Majority Voting

A fourth option is to introduce a double-majority scheme similar to those at the EIB, OECD, and GEF. Already, double-majority voting (85 percent of the voting power and 60 percent of the members) is required of the IMF’s Board of Governors to amend the Articles of Agreement or to expel a member from the organization. A similar scheme could be introduced at the IMF Board for certain kinds of decisions (for example, on policy but not on country issues); a more ambitious scheme would require double majorities for most decisions, exempting only a narrow category of decisions.

Performance Police

As we have seen, the boards of IGOs are not well suited to play the performance-police role of private-sector boards. Performance monitoring and evaluation often take place through separate evaluation offices or units, or through ombudsmen like the World Bank’s Inspection Panel that accept and follow up on grievances from stakeholders. In some organizations, the CEO is required to report to the Board on a regular basis.

The IGOs studied here do not offer useful insights to help strengthen the IMF’s role as performance police. What is clear is that for political reasons, such an undertaking would have to be approached delicately, possibly in parallel with a process of Board self-evaluation. This would demonstrate the Board’s commitment to evaluating its own performance as well as the MD’s. Also, the MD’s “report card” would need to be disaggregated into a variety of specific dimensions, such as managing relations with shareholders, chairing the Board, and managing and recruiting the staff. In contrast to the private sector, where performance is often linked directly to CEO compensation, CEO evaluation at the Fund would be the beginning of a constructive dialogue between the MD and the Board.

Conclusion

The central point of this paper is that the twin crises of relevance and legitimacy that the Fund is facing today are partly related to the organization’s adherence to the delegate-and-control model. The model has proven to be an effective way to ensure a strong political counterweight role for the Board and to guarantee major shareholders that they will have control over the use of the resources they provide. This has ensured sustained support for the institution by the largest economies. However, this has come at the expense of the Board’s capacity to play other important roles—as strategic thinker, as performance police, and as democratic forum.

Today, more than ever, the IMF needs its Executive Board to play these other three roles effectively. Governance reform should mean shifting away from the delegate-and-control model and importing or adapting governance mechanisms from other models to strengthen the Board’s other roles. Which roles are to be strengthened—and with which governance mechanisms—are political decisions that must be taken by the Fund’s stakeholders. This decision will affect the balance of power within the institution, how the IMF functions, and whether it will be able to remain relevant and effective in coming decades.

Annex. Representation and Voting Power in IMF Multi-Country Constituencies
Multi-Country ConstituencyNumber of ContriesCompositionNumber of Members Under an IMF Program (2006)Configuration of Voting PowerLeaders’ Share of Constituency’s Voting Power (%)ED Selection Arrangements (at Time of Writing)
India4Developing countries.2Single head.India80.3India always ED; Sri Lanka always Alternate.
Italy and Southern Europe7Mixed.1Single head.Italy77.8Italy always ED; Greece always Alternate.
Southern Cone6Developing countries.4Single head with junior partner.Argentina49.4Both positions rotate among all members, but Argentina ED more often than others.
Chile20.3
Australia/Korea14Mixed.1Single head with junior partner.Australia45.0Until 2004, Australia always ED; now, rotation with Korea; Alternate chair rotates among Australia, Korea, New Zealand, and the Philippines.
Korea16.6
Belgium, Turkey, and Eastern Europe10Mixed.1Single head with junior partnerBelgium41.5Belgium always ED; Austria always Alternate.
Austria17.0
Brazil9Developing countries.4Single head with junior partner.Brazil57.0Brazil always ED; Alternate rotates among Colombia, Ecuador, Panama, and Trinidad and Tobago (though by agreement, not cyclical).
Colombia14.9
Canada, Ireland, and the Caribbean12Mixed.1Single head with junior partner.Canada79.3Canada always ED, Ireland always Alternate.
Francophone Africa24Developing countries.11Single head with junior partner.Congo, DR18.1Rotation system includes all members; each member serves four consecutive terms, first two terms as Alternate and then two terms as ED.
Netherlands and Eastern Europe12Mixed.3Single head with junior partner.Netherlands49.2Netherlands always ED; Ukraine always Alternate.
Ukraine13.4
Switzerland, Poland, and Central Asia8Mixed.4Single head with junior partner.Switzerland56.3Switzerland always ED; Poland always Alternate.
Poland22.63
Anglophone Africa19Developing countries.9Dual head.S. Africa29.0Rotation system includes all members; each member serves two consecutive terms, first as Alternate and then as ED.
Nigeria27.3
Egypt and Middle East13Mixed.1Multiple heads.Kuwait22.0Egypt has always been the ED; Alternate elected from among other members.
Libya18.0
Egypt15.1
Iran, Pakistan, and Northern Africa7Developing countries.2Multiple heads.Iran28.4Iran always ED; Morocco always Alternate (Algeria and Pakistan and get the World Bank).
Algeria23.9
Pakistan19.7
Nordic/Baltic8Mixed.0Multiple heads.Sweden31.7Both positions rotate among top five (Denmark, Finland, Iceland, Norway, Sweden).
Norway22.2
Denmark21.9
Southeast Asia12Mixed.2Multiple heads.Indonesia21.0ED and Alternate rotate among Indonesia, Malaysia, Singapore, and Thailand.
Thailand16.0
Malaysia15.1
Spain and Central America8Mixed.2Multiple heads.Spain33.1Both positions rotate among the three heads.
Venezuela28.9
Mexico28.0
References

A version of this article was published as “Boards of Directors in International Organizations: A Framework for Understanding the Dilemmas of Institutional Design,” in Review of International Organizations, May 2009.

I am thankful to Alisa Abrams, Jeffrey Levine, Borislava Mircheva, and Roxana Pedraglio for their excellent research assistance. Thanks also to Tom Bernes, Ralph Bryant, Jeff Chelsky, John Hicklin, and Sandy Mountford for their helpful comments. For her editorial assistance, thanks to Rachel Weaving; for administrative assistance, thanks to Arun Bhatnagar and Jeanette Abellera.

Agreement on this issue was reached by the Executive Board in April 2008 and endorsed by the IMFC. For a critical view, see Bryant (2008).

Between 1909 and 1999, the number of conventional intergovernmental organizations grew from 37 to 251. Union of International Associations, see www.uia.org/statistics/organizations/ytb299.php (accessed September 2007).

Two factors make this a particularly thorny principal-agent problem. For one, IGOs are single agents, but they receive instruction and oversight from a “collective principal”—multiple states, which do not always agree with each odier. Another complication is the long delegation chain ultimately connecting the citizens of the member countries with the staff who actually carry out the organization’s mandate. Agents at each link in the delegation chain have incentives to follow most closely the directives of the principal immediately above them, rather than those of more distant principals. The longer the delegation chain becomes, the greater die probability of “agency slack”—that is, of the agent diverging from the preferences of principals.

This is not to say that the decisions of an IGO’s staff are always apolitical and based solely on technical considerations. However, the legitimacy of the staff’s influence is based solely on the claim to superior knowledge and technical rationality, and their decisions and advice are provided as if they emanated solely from this source.

Among major U.S. companies (S&P500), the average board size is 10.7; among the U.K.’s top 150 companies, it is 10.8, and among Italian blue-chip companies, the average is 10.7 directors. (Spencer Stuart, 2006a: 10; Spencer Stuart, 2006b: 5; and Spencer Stuart, 2006c: 7.) Among the top 50 Japanese companies, average board size is 13 directors. (Forbes, “The Global 2000,” 2007.)

The median board size among the nearly 400 U.S. non-profits participating in a recent survey declined from 17 members in 1994 to 15 in 2004. (BoardSource, 2004: 4.)

Again, private sector boards exemplify this point well: the average board member in an S&P500 firm was 61 years old and in top U.K. firms, executive directors were 50 and nonexecutive directors were 57 years old, on average. This suggests work experience of 25–30 years. Directors also tend to stay several years; in top U.K. firms, the average length of service for non-executive directors as of 2006 was 3.8 years. (Spencer Stuart, 2006b: 6.)

The largest number of meetings reported for an S&P500 corporate board in 2006 was 39.

For example, independent directors should not have been former employees of the company in the previous five years, should not have a material business relationship with the company, should not be or represent a significant shareholder, should not have close family ties with any of the company’s directors or senior staff, and should not have significant links with other directors through involvement in other companies.

The shift has been dramatic: in S&P500 firms, the percentage of independent board directors has increased from 27 percent in 2001 to 81 percent in 2006. In the U.K., some 62 percent of boards are made up of non-executive directors, nearly all of whom are independent. (Spencer Stuart, 2006b: 5.)

The governors retained the power to approve quota increases, SDR allocations, membership applications, and amendments to the Articles of Agreement and By-Laws. Voting on these issues generally takes place by mail ballot, rather than during the Annual Meetings.

However, some scholars have argued that the fact that executive directors are granted legal immunity by the IMF with respect to acts performed in the exercise of their official duties, and that this immunity can only be withdrawn by the Fund (not by their governments), is evidence that they are officials of the Fund rather than delegates of their governments. See Gianviti (1999).

The first generation of directors included one former vice-minister of finance, one under-secretary of state for finance, and three directors, two commissioners, and one general manager, all from central banks (Horsefield, 1969: 138).

According to a survey of Board attendance in the 1940s, only three executive directors were present at more than 75 percent of the meetings and three directors attended less than 25 percent (Horsefield, 1969: 167).

For example, Brazilian Executive Director Alexandre Kafka served on the Board for 32 years. When he retired in 1998, the average “age” of the Board dropped precipitously from 47 months to 25.

Calculations are based on data provided by the Secretary’s Department. Directors’ length of service rises if we include in the calculation the time that some spent as alternates before becoming directors. Including the time served as alternates, the average time on the Board between 1990 and 2007 increases to 54 months, while the median rises to 39 months.

Voting power is allocated according to each member’s quota. While formal voting is rare and the Board operates on the basis of “consensus,” Board decisions are determined by a preponderance of the weighted votes, even if no votes are formally cast. Whether decisions are reached with unanimity, with broad agreement, or only with a simple majority of the voting power depends largely on the judgment of the Managing Director, who chairs the Board. Voting weights also affect representation within constituencies. In some constituencies, voting power determines which country or countries sit in the director’s chair, which fill the position of alternate, and which countries are to get staff positions as senior advisors and advisors to directors.

Special majorities are required for some 39 types of decisions. Decisions requiring special majorities are not necessarily the most sensitive or important (see Mountford, Chapter 2 in this volume).

“On this point, see Chelsky (Chapter 8 in this volume).

Executive directors have committed to devising a performance contract for the current MD.

Terms of Appointment of Dominique Strauss-Kahn as Managing Director of the International Monetary Fund, November 2, 2007.

On the selection process, see Peretz (Chapter 11 in this volume).

For an informative history of multilateral development banks, see Kapur and Webb (1994: 229-50).

The World Bank’s Intranet states that “An Executive Director (or Alternate) fulfills a dual function, as an official of the Bank and as a representative of the member country or countries that appointed or elected him.” However, as for the IMF, this dual role is not reflected explicitly in the Articles of Agreement or By-Laws.

Four vice-presidents always come from each of the Bank’s four largest shareholders (Germany, France, Italy, and the United Kingdom), and the rest come from specified constituencies, each with its own scheme for regular rotation. In addition, great care is taken to ensure that the nationalities of the Bank’s staff reflect the shares of member countries’ contributions to the Bank’s capital.

Author’s calculations based on data provided by the OECD.

The introduction of DMV in these two organizations is particularly important given the large inequalities in the members’ financial weight. For instance, at the OECD, two members (Japan and the United States) alone provide some 42 percent of the total contributions that make up the bulk of the organization’s budget. At the EIB, the “big four” (France, Germany, the United Kingdom, and Italy) represent 65 percent of the Bank’s subscribed capital.

For example, since 1998, the WHO explicitly recognized its directors as government representatives, after years of pretending that they served only in their personal capacities and owed their allegiance only to the medical profession. On this point, see Burci and Vignes (2004: 57-58).

Currently, six experts are in place; the directors are from France, Italy, and the U.K., while the alternates come from Spain, Poland, and Germany. The three directors are senior, private-sector bankers, usually with experience in project finance.

On this point, see Bossone (Chapter 12 in this volume).

On this point, see Chelsky (Chapter 7 in this volume).

This has been suggested for the World Bank by the South Centre (South Centre, 2007).

    Other Resources Citing This Publication