Rethinking the Governance of the International Monetary Fund

WP/06/273


Rethinking the Governance of the
International Monetary Fund

Abbas Mirakhor and Iqbal Zaidi

---

---

© 2006 International Monetary Fund
WP/06/273





IMF Working Paper


Office of the Executive Director for the Islamic Republic of Afghanistan,
Algeria, Ghana, the Islamic Republic of Iran, Morocco, Pakistan, and Tunisia

Rethinking the Governance of the International Monetary Fund

Prepared by Abbas Mirakhor and Iqbal Zaidi

December 2006



Abstract

This Working Paper should not be reported as representing the views of the IMF.
The views expressed in this Working Paper are those of the author(s) and do not necessarily represent
those of the IMF or IMF policy. Working Papers describe research in progress by the author(s) and are
published to elicit comments and to further debate.

This paper attempts to set out the principal issues that need to be resolved in formulating a
proposal for quotas and voice reform in the IMF that could command broad support. Following
John Rawls, we argue that “justice is the first virtue of social institutions,” and we use his theory
of justice to provide a method for understanding what should be the case, in the context of voice
and voting shares, before international institutions, such as the IMF, are to be justifiable to their
members. The implementation of this process suggests, among other things, that a major revision
of the quota formulas is long overdue, and leaving this unaddressed raises serious questions
regarding the IMF’s governance which could develop into a core mission risk and jeopardize the
relevance of the institution.

JEL Classification Numbers: F02, F33, F53

Keywords: Governance, International Monetary Arrangements and Institutions, International
Monetary Fund

Authors’ E-Mail Addresses: amirakhor@imf.org; izaidi@imf.org

---

2
Contents Page


I. Introduction ........................................................................................................................... 3

II. Quotas, Voice, and Rawls .................................................................................................... 7
A. Quota Formulas................................................................................................................. 7
B. Original Bretton Woods Formula Versus Original Position ........................................... 10

III. Rawls’s Method ................................................................................................................ 14
A. Quota Formula Review Group (QFRG).......................................................................... 14
B. Variables in the Quota Formulas..................................................................................... 16
C. Reflective Equilibrium .................................................................................................... 18
D. Basic Votes ..................................................................................................................... 27

IV. The Way Forward ............................................................................................................. 29
A. Adjusting Voting Power and Quotas .............................................................................. 30
B. Aligning Quotas and Basic Votes with Justice as Fairness............................................. 32

V. Conclusion ......................................................................................................................... 36

References............................................................................................................................... 43

Text Table 1. Basic Votes and Variables for Quota Formulas ............................................... 31

Appendix Table 1. Formulas Used for Quota Calculation: 2006 Update............................... 38

---

3
I. INTRODUCTION

Just as national regulation was broadened in the 19th and 20th centuries to protect workers and
consumers (e.g., anti-trust legislation, health standards, corporate governance, bank
supervision) from the excesses of free markets, there is now a general recognition that
globalization needs a regulatory framework in the 21st century that is less fragmented than
what exists today, and international financial institutions—in particular, the International
Monetary Fund (IMF)—can be expected to have major roles in this area. In fact, the IMF has
already taken steps in this direction (e.g., evaluating countries’ compliance with international
data standards, moving in the direction of setting a new surveillance remit, and launching a
multilateral consultation on addressing global imbalances). However, for the IMF to play an
important role in global governance, it is essential to enhance its credibility as an
international cooperative institution: there is widespread recognition that the quotas (IMF
capital shares), voting rights, and voice imbalances have become progressively worse.1 The
effectiveness of the IMF has been questioned both inside and outside the institution not only
because members’ quotas have become increasingly out of line with countries’ economic
weight (measured by GDP) in the global economy, but also because there is a growing
recognition that some important aspects of members’ economic weight and other variables
that should have a bearing on voting rights are not captured in the current quota formulas.

These concerns are reflected in the International Monetary and Financial Committee
communiqué of April 22, 2006, which stated that the IMF’s effectiveness and credibility as a
cooperative institution must be safeguarded and its governance further enhanced, and
emphasized the importance of fair voice and representation for all members. The IMF has
adopted a two-stage process for quota and voice reform, with initial ad hoc increases for the
clearly most underrepresented members in the first stage, and more fundamental reforms in
the second stage.2 While the specific reform proposals for the second stage are just beginning
to be discussed, there is already considerable concern among developing countries that the
discussion is being confined to an unduly narrow area, and important issues are not being
raised. These concerns are only magnified by the concerted efforts being made in many
quarters to validate the traditional approach of basing voting power in the IMF largely on
countries’ respective weight in the world economy, with the justification being provided in
terms of the mandate of the institution. However, representatives from developing countries
have rightly pointed out that the IMF mandate is not as narrow as some would have us

1 Quotas are currently calculated according to a member's gross domestic product, current account transactions,
and official reserves. The quota largely determines a member's voting power in IMF decisions and is reviewed
every five years, with the next review due in 2008 (see Section II).
2 A two-stage process with an ad hoc increase in the first stage is not consistent with the need for a
comprehensive review, and the Finance Minister of India was correct in saying during the IMF-World Bank
Annual Meetings in Singapore in October 2006 that “[b]y definition, a comprehensive reallocation of quotas to
reinforce legitimacy cannot be achieved by a short-term ad hoc approach." Equally valid is the point that the
under-representation of developing countries undermines the credibility and legitimacy of the Bretton Woods
Institutions, which hinders effectiveness and relevance of these institutions, as noted in the G-24 communiqué
during those meetings.

---

4
believe, and history bears this out. Furthermore, even in the discussions on the need to find
ways to enhance the representation of developing countries, the discussion seems to be
almost entirely directed to the admittedly important, but still only one area of voice reform,
namely, the need to arrest the declining role of basic votes since the IMF was established,
which has weakened the voice of smaller developing countries. However, the voice reform
should mean much more than just ensuring that small countries, whose share in the world
economy is small but for whom the IMF provides important policy advice and financing,
have adequate opportunities to participate in the governance of the institution. In particular,
the IMF also provides policy advice and financing to countries with large populations—and,
of course, it has important regulatory and supervisory functions that affect them—and there
will remain a “democracy deficit” if these countries are not adequately represented in the
governance structure.

These concerns are also not lessened by the fact that the Achilles Heel of the two-stage
approach adopted by the IMF is that the formula issue is to be addressed in the second stage,
which puts the cart before the horse because it was decided to correct the quotas of some
members in the first stage by using the same formulas that are widely regarded as faulty and
inadequate. The ad hoc increase in quotas in the first round was supposed to play an
important role in improving the distribution of quotas to reflect changes in the weight and
role of countries in the world economy. The quota formulas were used to find clearly
underrepresented countries, which turned out to be only four (China, Korea, Mexico and
Turkey) for ad hoc quota increases. However, it does not require rocket science to know that
a wrong metric will give a wrong result. In fact, instead of providing useful information for
deciding on the ad hoc increase in the first round, what the calculations confirmed, based on
the faulty formulas, was that the use of inappropriate variables (e.g., using market exchange
rates rather than purchasing power parity to derive GDP, collinearity amongst the variables,
the problem of multiple quota formulas, and the nonlinearities in the quota formula) tended
to skew results in the most troublesome ways.3

The proponents of the ad hoc increases had argued that the first round of increases would
cover only those countries that meet a robust standard of underrepresentedness, but ignored
the fact that a robust standard derived from a flawed formula is an oxymoron. It is hard to
understand, given the widespread dissatisfaction with the quota formulas, why these formulas
were used for deciding on the ad hoc increase in quotas. The preferred approach would have
been to look at the problem in an integrated framework, focusing, in particular, on the quota

3 It is disconcerting to note that three of the four countries that received the ad hoc increase in the first round of
the quota and voice reform (Korea, Mexico, Turkey) had their quotas calculated using the new formulas, which
give a lower weight to GDP than the Bretton Woods formula (see Appendix Table). In other words, the clearly
most underrepresented members were supposedly the ones whose economic weight (i.e., GDP) was not
adequately reflected in the share of voting power in the IMF, but ironically, for these three countries, it is the
openness and external variability metrics that were relatively more important in the determination of the
calculated quotas than if the Bretton Woods formula had been used. As discussed in Section II, the Bretton
Woods formula has a coefficient of 0.01 in front of the GDP variable, whereas in the new formulas it is lower,
ranging from 0.0045-0.0065.

---

5
formulas and basic votes. By resolving or at least making substantial progress on the
revamping of the quota formulas in the first step, the IMF would have sent an important
signal that its governance reform is not focused on the quota problems of just a handful of
countries, but is meant to address the voice imbalances that exist across the membership.
After resolving the issue of the quota formula, which is the only way to come up with the
correct metric for discussing which countries’ quotas are most out of line as well as the
broader question of the overall distribution of quotas across regions and across analytical
country classifications, the IMF would have put itself in a position to grant significant ad hoc
increases to a few members, as well as selective increases that would be distributed across a
wide group of countries.

This paper does not focus on the flaws in the two-stage approach to quota reform, but instead
takes a forward looking approach and attempts to set out the principal issues that need to be
resolved in formulating a proposal for quotas and voice reform that could command broad
support. Following John Rawls, we argue that “justice is the first virtue of social
institutions,” and we think that his theory of justice provides an appropriate method for
understanding what should be the case, in the context of voice and voting shares, before
international institutions, such as the IMF, are to be justifiable to their members. Our analysis
is based on the Rawlsian notion of “justice as fairness” and, at the risk of oversimplification,
our conclusion is that justice in the IMF governance structure requires a distribution of voting
power that participants accept as the end-result of a fair process. The implementation of this
process suggests that a major revision of the quota formulas is long overdue, and leaving this
unaddressed raises serious concerns regarding the IMF’s governance. Furthermore, there is
no legitimate way to view these issues in isolation, and a holistic approach is required, which
would entail increasing basic votes sharply—to at least restore its importance at the inception
of the IMF—and revamping the quota formulas, with the latter allowing for selective quota
increases for a broad group of developing countries. However, we hasten to add that this
work should be regarded as advancing possible options for further discussion, and not as
constituting a specific proposal. In particular, the ranges given to demand and supply
variables, or for treating the democracy and Westphalian deficits, in the quota table in this
paper, are for heuristic purposes only and are not meant to be specific recommendations.

The IMF has a complex governance structure in which the constituency system attempts to
reconcile the legitimacy of an almost universal membership with efficient decision-making
and collegiality of a not-too-large Executive Board (24 Executive Directors). In the
constituency system, the five member countries with the largest quotas appoint an Executive
Director, while the remaining members elect the remaining Executive Directors. Questions
have been raised about the constituency system, particularly the point that the Executive
Director cannot split his/her vote even though there are instances in which the countries
within the constituency are divided on the issue being considered by the Board; this is
especially relevant in those cases in which there are mixed constituencies, industrial and
developing countries. Another governance issue has been the debate on converting the
International Monetary and Finance Committee (IMFC) into a decision-making council.
After long debates, this was turned down in 1999, attributed mainly to the concern that the
industrial country members of the council may not show the necessary patience and
willingness to work toward consensus decision-making, which is necessary to protect the

---

6
interests of minority groups. Yet another set of questions relates to the simple majority that
applies to many decisions and the special majorities of 70 and 85 percent for certain key
decisions. The special majorities help to protect sizeable minorities, but the 85 percent
majority gives veto power to one country.

The issues raised in the preceding paragraph are just a few of the many outstanding issues in
the IMF governance debate, but this paper does not attempt to cover every conceivable area.
First, it does not discuss the merits or otherwise of voting majorities, or the efficacy and
representation of the constituency system. Nor does it express a view on converting the
IMFC into a council. The paper focuses on the quotas and voice debate, which is arguably
the overriding issue in the larger governance debate. Second, and at least equally important,
the paper does not discuss the question of increasing the independence and accountability of
the Executive Board. The Board of Governors is the highest decision-making body of the
IMF but the daily business is conducted by a resident Executive Board, which exercises
under delegation most of the powers. The main functions of the Executive Board include:
approving all policies of the IMF; discussing consultation reports with individual countries,
the world economic outlook and the global financial system to carry out its mandate on
bilateral and multilateral surveillance; and approving loans provided for adjustment programs
and reviewing the implementation of the conditions attached to those programs to decide on
whether to disburse the loan tranches. The question of political oversight by national capitals
of the business of the IMF has garnered attention from time to time, and in particular, some
rules and practices related to the appointment, election, and term duration of Executive
Directors have been challenged by some observers from the point of view of strengthening
the autonomy and accountability of the Executive Board. These issues are not discussed in
the paper not because they are simple and unimportant, but rather because these problems
have been extensively analyzed elsewhere and because they are quite distinct from the
questions raised in this paper on quotas and voice reform.4 Third, it should perhaps be made
explicit that the paper is not concerned with other areas, such as the desirability of further
enhancing the capacity of Executive Directors’ offices representing large numbers of African
members and of including more transparent selection procedures for the position of
Managing Director. These issues are not discussed not because they are unimportant, but
because they are simple. There is no question that the challenges faced by the two African
chairs, each of which represents more than 20 countries, are serious. However, this is not an
area that requires further deliberations, and it should be implemented expeditiously. The
point is that the resources involved for strengthening capacity of African Executive
Directors’ offices are not substantial in terms of the overall budget of the IMF, and by having
these problems linger on, they only serve to confuse the discussion about quotas, basic votes,
legitimacy, which are issues of a different kind than some small budgetary matters; in short, a
larger budget for an Executive Director’s office is no substitute for underrepresentation. On
the desirability of including more transparent selection procedures for the position of
Managing Director of the IMF, this too should have been done some time ago because there

4 See Kenen (2001), King (2006), Portugal (2003), Truman (2006), Van Houtven (2002) and Woods (1998,
2001) for comprehensive discussions of these issues. See Boughton (2001) for the recent history of IMF
finances, pp. 849-874.

---

7
really is no debate about it, at least ever since the discussion on the joint draft report of the
IMF’s Working Group to Review the Process of Selection of the Managing Director and the
World Bank Working Group to Review the Process of Selection of the President. In this
regard, specific procedures for ensuring this transparency should be developed soon, and
there is no need to wait for the two-year program of actions on governance reform. As with
strengthening capacity of African Executive Directors’ offices, this issue should not be
allowed to remain unaddressed because it needlessly complicates the more important issues
of quotas and voice reform.

II. QUOTAS, VOICE, AND RAWLS
Many policymakers, not to mention researchers, have commented that trying to understand
IMF quota formulas is a formidable undertaking, yet the mathematics involved is nothing
worse than the simplest algebra. One reason is that, even in the very first reading when one is
busy trying to understand the formulas, it is difficult not to get bogged down into disagreeing
with just about everything contained in the formulas and to start arguing why a particular
variable is used, why it has more weight than some other variables, or why there are five
distinct formulas, and so on. That problem could be overcome, to some extent, by practicing
in advance of the need, that is, the first step in understanding these formulas should be to just
peruse them without any comments, and then only afterwards go about disagreeing with the
formulas. Even with this practice, one may find IMF quota discussions rather confusing
unless one is careful in defining the objectives. The procedure followed in this section is:
first, we discuss what the IMF quota formulas are meant to achieve; second, we simply state
the formulas; and, third, we take issue with several aspects of the formulas, including the
choice of variables, multiple formulas, and nonlinearities in the formulas.5

A. Quota Formulas
Quota subscriptions generate most of the IMF's financial resources, and total quotas at end-
September 2006 were SDR 213 billion. Quotas perform several functions, including
delineating basic aspects of members’ financial and organizational relationship with the
IMF:
Subscriptions. A member’s quota subscription determines the maximum amount of
financial resources the member is obliged to provide to the IMF. A member must pay its
subscription in full upon joining the IMF: up to 25 percent must be paid in SDRs or widely
accepted currencies (such as the U.S. dollar, the euro, the yen, or the pound sterling), while
the rest is paid in the member’s own currency.
Voting power. The quota largely determines a member’s voting power in IMF decisions.
Each IMF member has 250 basic votes plus one additional vote for each SDR 100,000 of
quota.

5 Readers familiar with IMF quota formulas may wish to skip subsection II.A.

---

8
Access to financing. The amount of financing a member can obtain from the IMF (its access
limit) is based on its quota. Under Stand-By and Extended Arrangements, for instance, a
member can borrow up to 100 percent of its quota annually and 300 percent cumulatively.
However, access may be higher in exceptional circumstances.
SDR allocations. A members’ share of general SDR allocations is established in proportion
to its quota.
Since quotas serve multiple purposes, the quota formula necessarily has to balance
sometimes competing considerations about what variables to include in the formulas and the
weights to attach to each variable. The formulas are overburdened by the multiple roles of
quotas, and there is no particular need to have a rigid relation between financial contribution,
access to Fund resources, voting power, and share of SDR general allocations. In fact, there
are good reasons why the share of SDR allocations should not be linked to quotas.6
Regarding access, it is already the case that waivers are permitted to allow access outside the
quota-based limits specified under the Articles (see Article V, Section 3 (b) (iii) and 4, under
which use of Fund credit is limited to 100 percent of quota, unless waived). In our view, a
cogent case can be made for fully delinking quotas from all decisions on access, but we
realize that such an approach would raise a number of issues that would take us far away
from our main purpose. Also, it would require an amendment to the Articles, and since this
approach has received little support, we do not pursue it in this paper.
A formula used in 1944 when the IMF was established has become known as the Bretton
Woods formula.7 This formula contained five variables: national income, official reserves,
imports, export variability, and the ratio of exports to national income. This single formula
was replaced by a multi-formula approach in the early 1960s, when the original formula was
supplemented with four more formulas containing the same basic variables but with larger
weights for external trade and external variability. However, this was not the end of the
problem because two different data sets were used; there were in effect ten formulas. The
quota formulas were simplified in 1981–82, including the following changes: (i) eliminating
five of ten formulas by focusing on only one data set; (ii) replacing nominal income with
GDP, which was viewed as a more comprehensive and readily available measure of national
output; (iii) broadening the measure of reserves to include holdings of SDRs, ECUs, and
IMF reserve positions and calculation of the holdings as a 12-month average rather than an
end-of-period total; and (iv) reducing the coefficient of variability in the four derivative
formulas by 20 percent to moderate the impact of the very sharp increases in the prices of
certain commodities, especially the increases in oil prices in 1973/74 and 1979. These new
formulas were supposed to help the developing countries because of their vulnerabilities to
terms-of-trade shocks and reliance on a narrow range of exports. However, it is interesting to
note that there are several industrial countries—and one G-7 country—that have calculated
quotas determined by the new formulas. Also, the 60 year old Bretton Woods formula is
used for more than one-third of the members, including many developing countries (see

6 See Yaqub, Mohammed, and Zaidi (1996).
7 For a more comprehensive overview of quota formulas, see International Monetary Fund, 2001b.

---