Full text of DOI:10.1111/1468-2346.00179

Making the IMF and the World Bank
*
more accountable
NGAIRE WOODS
1
The International Monetary Fund (IMF) and the World Bank (the Bank) are
now regularly accused of being secretive, unaccountable and ineffective. Not
only radical non-governmental organizations (NGOs) but equally their major
shareholders are demanding that the institutions become more transparent,
more accountable and more participatory. Accountability, in particular, has
become the catchcry of officials, scholars and activists in discussing the reform of
the institutions. Yet few attempts have been made to dissect the existing
structure of accountability within the international financial institutions (IFIs),
to explain its flaws and to propose solutions. That is the aim of this article.
The first section examines the structure of accountability planned by the founders
of the IMF and the World Bank. The second section discusses the defects in this
structure. Section three analyses recent attempts to make the institutions more
accountable. The conclusion offers some recommendations for improving the insti-
tutions, and a warning about the limits of accountability at the international level.
How are the IMF and the World Bank accountable to their
government members?
Like many international organizations, the IMF and the World Bank face com-
plex problems of accountability which begin when one tries to answer the
simple questions: to whom should they be accountable, and how? Within
democratic political systems, several mechanisms exist to prevent the abuse or
2
misuse of political power. These measures range from elections through to the
*
Any errors and misjudgements in this paper are of course my own. However, I would like to thank the
G-24 for sponsoring my research on this question, Devesh Kapur, Stephen Haggard and Peter
Oppenheimer for detailed comments and ideas which have substantially improved this paper, and also
Amar Bhattacharya, John Hicklin, David Peretz, Abbas Mirakhor and Gerry Helleiner for their
particularly useful input and comments.
The terms ‘World Bank’ and ‘the Bank’ are widely used in this article and also to refer to the World Bank
Group of institutions which comprises the International Bank for Reconstruction and Development
1
2
(
IBRD), the International Development Association (IDA) and the International Finance Corporation (IFC).
Andreas Schedler, Larry Diamond and Marc Plattner, The self-restraining state: power and accountability in
new democracies (Boulder, CO: Lynne Rienner, 1999).
International Affairs 77,  () ‒
83

Ngaire Woods
appointment of ombudsmen and judicial review. The aim is to ensure that poli-
tical actions are predictable, non-arbitrary and procedurally fair, that decision-
makers are answerable for their decisions, and that rules and limits on the
exercise of power are enforced. For all these reasons, accountability within
public institutions, whether national or international, is a desirable thing.
Unlike democratically elected governments, international institutions cannot
claim that voters elect them and can vote them out of office. Nor, in the past,
have the institutions been subjected to the normal restraints politicians face
from the checks and balances of government, including the role played by
judges, ombudsmen and other such figures. Rather, international organizations
grapple with an unwieldy structure of government representation which makes
ensuring their own accountability extremely difficult. In the past, when such
institutions were required to perform a narrow range of technical functions, the
problem was less acute. Today, however, the international financial institutions
are being required to perform a much wider range of tasks directly affecting a
wider range of people; and the question of their accountability assumes
correspondingly greater importance.
The basic structure of accountability in the international financial institutions
as they are at present constituted works through representatives of governments.
At the top of the system are the Boards of Governors—the ministers of finance
or development, central bank governors and equivalents—who meet just once
a year and are supposed to maintain overall oversight and control of the institu-
tions. The day-to-day operations and main work of the institutions are
overseen by representatives of member states who sit on the Executive Boards
of each institution. The Executive Directors (as they are called) have a dual
role: to represent a country or a group of countries (this is further described
below); and collectively to manage the organization. Executive Directors
appoint and can dismiss the head of each organization, who in turn controls the
management and staff.
The chain of representative accountability described is in practice a long and
imperfect one. Flaws in each link highlight how weak the relationships are
between most member governments and the IMF and the World Bank. Simply
put, member governments (with the obvious exception of the United States)
are too far removed from the workings of the representative body (the Executive
Board), which in turn exerts too little control over the staff and management of
the institutions for its role to be described as an exercise of vertical accountability.
Why do the IMF and World Bank not seem accountable?
The core of the problem of accountability lies in the flawed representativeness
of the Executive Board in the Fund and in the Bank. The Board should reflect
the membership of the organization, that is to say, the governments of its
constituent countries. Yet representation is inadequate in two respects. In the
first place, the Board does not adequately represent all members, and it parti-
84

Making the IMF and the World Bank more accountable
cularly fails adequately to represent the countries with the most intensive
relationships with the institutions, for example the African members. Second,
the representatives on the Boards of the institutions are too distant from the
governments they represent and the stakeholders most affected by the work of
the institutions. Let us examine each of these arguments in turn.
Representation on the Executive Boards is too unequal
The Board of Executive Directors (‘the Board’) is the vital link between
countries (and voters) and the IMF and World Bank respectively. Yet only the
largest member countries (the United States, Germany, France, Japan, the UK,
Saudi Arabia, Russia and China) are directly represented by their own Execu-
tive Directors. All other economies are grouped within constituencies each
represented by just one Executive Director. This means that most national
governments have only the weakest link to the formal deliberations and
decision-making processes of the institutions. For example, the 21 anglophone
African members of the IMF, at least 11 of which have an ‘intensive care’
relationship with the institution and all of which are deeply affected by its work,
are represented by just one Executive Director and have a voting share of 3.26
per cent. In the World Bank, the same group of countries plus the Seychelles
again are represented by just one Executive Director and have a voting share of
4.07 per cent. For these countries, the lack of representation is exacerbated by
their lack of influence in the informal processes of consultation and decision-
making with both the IMF and the World Bank.
In the Board’s decision-making process, the voting power of each country is
exercised by the Executive Director representing it. What this voting power is
depends upon the country’s ‘quota’, which is determined by a formula which
attempts to translate relative weight in the world economy into a share of
contributions and votes (and, in the IMF, access to resources) in respect of each
institution. The formula has been criticized for some time and its technical
elements have recently been reviewed for the Managing Director of the IMF.3
The real problem with quotas and voting power, however, is that they were
created to govern institutions with very different world roles from those played
by the IMF and the World Bank today.
Voting shares in the World Bank and IMF are allocated according to rules
formulated in 1944. At that time, members of both institutions were expected
to be both contributors and borrowers (with the exception of the United
States). European countries expected to draw upon the IMF for assistance in
crises and upon the International Bank for Reconstruction and Development
3
Report to the IMF Executive Board of the Quota Formula Review Group (Washington DC: IMF,
2000). For earlier critiques see Nancy Wagner, A review of PPP-adjusted GDP estimation and its potential use
for the Fund’s operational purposes, IMF Working Paper (Washington DC: IMF, 1995); Ariel Buira, ‘The
governance of the International Monetary Fund’, in Roy Culpeper and Caroline Pestieau, eds,
Development and global governance (Ottawa: International Development Research Centre, North–South
Institute, 1996).
85

Ngaire Woods
later part of the World Bank group) for postwar reconstruction and develop-
(
ment. Hence ‘shareholding members’ were also ‘stakeholders’ in the work of
the institution. Furthermore, the accountability of the IMF was largely
‘
constitutionalist’, its duties and actions carefully prescribed and delimited by its
Articles of Agreement. Neither the Bank nor the Fund was given an explicit
mandate to enter into policy conditionality and to attempt to alter in a far-
reaching way the economic structure of a member’s economy.
These founding presumptions of the IMF and World Bank were rapidly
overtaken by events. The role of the World Bank (then actually the IBRD) was
transformed in 1947 when the Marshall Plan was announced to deal with recon-
struction in Europe. This left the World Bank lending exclusively to developing
countries. In the 1970s the IMF’s role changed dramatically when the Bretton
Woods system of exchange rates collapsed. By the 1980s, both institutions had
become heavily involved in conditionality and policy-based lending. And over
the course of this period the membership of both institutions more than trebled
as decolonization brought a host of new, independent states into their midst.
The result of these changes has been to expand sharply both the number and
the range of ‘stakeholders’ in these institutions. The original stakeholders made
large contributions to the basic capital of the institutions. A subsequent group of
stakeholders have become those who pay most of the running costs of the insti-
tutions through loan charges (non-concessional borrowing countries). A further
group of stakeholders (overlapping with the second) are those whose cooper-
ation is vital if the IMF and World Bank are to fulfil their respective purposes.
This now embraces a much wider, indeed all-inclusive, category of countries.
By 2000, in the wake of the financial crises of the 1990s, the powerful
industrialized members of the IMF and the World Bank had cast the institutions
in the role of ensuring ‘forceful, far-reaching structural reforms’ in the economies
of all members in order (among other things) to correct weaknesses in domestic
4
financial systems and ensure economic growth and poverty alleviation. The
result is not only that all members have a high stake in the institutions but that
equally the institutions have a high stake in gaining a deep political commit-
ment to change in all member countries. Yet the stake of those countries whose
commitment is now being sought is not reflected on the Executive Boards of
either the Fund or the Bank—indeed, it has even been diminished.
When the IMF and World Bank were created, there was a clear and explicit
concern to ensure some equality among members to reinforce the ‘universal’
and ‘public’ character of the institutions, as opposed to giving them a structure
which simply reflected relative economic and financial strength in the world
5
economy. This was achieved by giving every member of the institution 250
6
‘
basic votes’, to which were added weighted votes apportioned on the basis of
4
IMF, The IMF’s response to the Asian crisis (Washington DC: IMF, 1998).
Joseph Gold, Voting and decisions in the International Monetary Fund (Washington DC: IMF, 1972), pp. 18,
5
1
73–4.
6
Keith J. Horsefield, The International Monetary Fund, 1945–1965: twenty years of international monetary
cooperation, vol. 1 (Washington DC: IMF, 1996).
86

Making the IMF and the World Bank more accountable
the quota (mentioned above). In this way two kinds of stake were recognized:
to a degree every member had an equal stake in the institution, and to a degree
the different contributions and influence of states were also recognized. As I
have argued elsewhere, this balance, which was explicit in the foundation of the
institutions, has subsequently been dramatically eroded. The proportion of
‘
1
basic votes’ in total votes has diminished from its high point of 14 per cent in
955 to around 3 per cent in both the Fund and the Bank. Yet if anything, the
7
role of basic votes should have increased as the stakes in the institution changed
in the ways described above.
The Executive Boards do not adequately hold staff and management to account
The Executive Boards are the lynchpin of accountability of the Fund and Bank.
For this reason the quality and depth of their oversight and control of each
institution’s work is vital. In each organization the Board is expected to appoint
and oversee the senior management and work of the institutions. However, in
practice the Board rarely holds the management and staff of the institutions
tightly to account. Executive Directors often (although not always) rubber-
stamp decisions coming before them. This occurs for several reasons.
In the first place, it is difficult for members of the Executive Boards to
prepare positions on all countries, papers and issues brought to Board meetings.
Many Executive Directors are in the job only for a short time. Indeed, in some
multi-country constituencies there is regular and rapid rotation of the Executive
Directorship.
In the second place, the Executive Board is not assisted in playing an active
role by the staff and management of each institution who seldom divulge
internal disagreements to the Board, tending instead to attach ‘considerable
importance’ to presenting a unified view in Board discussions.8
A third feature of the Executive Boards is that many decisions are taken—or
agreement reached on them—prior to Board meetings. As reports on both the
Fund and the Bank aver, real debates over policy and issues are conducted
9
outside of the Board. Controversial cases and stand-off debates are rare. For
example, a loan that did not meet with US approval would seldom be presented
to a Board for discussion. Before getting that far, in most (but not all) cases staff
and management would have been in dialogue with those whose agreement
was necessary for the loan to go through.
Finally, it has been argued that the Executive Board’s limited oversight is also
attributable to Directors’ protectiveness towards the countries they represent,
and their consequent expectation (and reciprocation) of similar deference from
7
Ngaire Woods, ‘The challenge of good governance for the IMF and the World Bank themselves’, World
Development 28: 5, May 2000.
IMF, ‘External evaluation of IMF surveillance’, report by a group of independent experts (Washington
8
DC: IMF, 1999), p. 34; World Bank, ‘Report of the ad hoc committee on Board procedures’, 26 May
(
Washington DC: World Bank, 1992).
9
IMF, ‘External evaluation’; World Bank, ‘Report of the ad hoc committee’.
87

Ngaire Woods
their colleagues on the Board. Hence, in the words of the External Evaluation
into Surveillance, what is supposed to be ‘peer pressure’ in fact becomes ‘peer
protection’.10
The heads of both organizations are selected by a non-transparent process which excludes
most member countries
A further, very obvious way in which the IMF and the World Bank seem
inadequately accountable to their membership lies in the way in which the head
of each organization is appointed by the Executive Boards. It is to these heads of
the organizations that all staff must eventually account. Yet in neither the Bank
nor the Fund is there an open and transparent process of appointment—
whether aimed at achieving political representation or technical excellence.
Rather, a 50-year-old political compromise means that in each organization the
head is appointed by convention according to the wishes of the United States
(in respect of the World Bank) or western Europe (in respect of the IMF). This
process came under scrutiny during the most recent appointment of the
Managing Director of the IMF when Germany’s first favoured candidate failed
to win support from other major shareholders, leading to much adverse press
and policy attention—focused not so much on the personalities involved as on
the lack of transparency and accountability in the process of selection. Although
both institutions have established committees to propose improvements in the
appointment procedure, until the United States and European countries are
prepared to give up their respective privileges, change in this procedure will not
occur.
The role of the IMF and World Bank has expanded; their accountability has not
The problems of accountability highlighted above have all been magnified by
the expansion and transformation of the activities of the IMF and World Bank.
Previously, the need to respect the sovereignty of member governments limited
their range of activities. However, both now reach deep into policy-making
within member governments, going well beyond the delicately respectful
boundaries set out in the original Articles of Agreement.
Specifically, as Devesh Kapur has argued, both the IMF and the World Bank
now embrace areas of policy it was inconceivable they would touch prior to the
1980s. The expansion can be tracked by measuring how the ‘performance
criteria’ on which loans are conditional have multiplied: in a sample of 25
countries, there were between 6 and 10 measures in the 1980s, as contrasted
11
with around 26 measures in the 1990s. Equally revealing, if not more so, is the
1
0
1
IMF, ‘External evaluation’, p. 34.
1
Devesh Kapur, ‘Expansive agendas and weak instruments: governance related conditionalities of the
international financial institutions’, Journal of Policy Reform (forthcoming); Devesh Kapur and Richard
Webb, Governance-related conditionalities of the international financial institutions (Geneva: UNCTAD, 2000).
88

Making the IMF and the World Bank more accountable
way in which the numbers of programme ‘objectives’ being included in loans
and programmes have increased, with countries now being required to under-
take actions such as to mobilize, redefine, strengthen or upgrade government
processes in an ever wider range of areas.
The new conditionality is dramatically deepening and broadening the pur-
view of the international financial institutions within member countries. No
longer are they engaged in monitoring specific macroeconomic policy targets in
the context of a crisis, or specific project loans and conditions. Both institutions
are now engaging governments in negotiations which cover virtually all issues
in economic policy-making—and beyond, with ‘good governance’ extending
into areas such as the rule of law, judicial reform, corporate governance and so
forth. This new, wide-ranging domain of advice and conditionality directly
affects a broader swathe of policies, people, groups and organizations within
countries.
Yet the IMF and the World Bank were neither created nor structured to
undertake or to be accountable for such far-reaching activities. They were
created to deal with a narrow, clearly stipulated range of technical issues. For
this reason, at their birth it was decided that they should deal with member
countries only through the treasury, finance ministry, central bank or like
institution of a country, and that only representatives of such agencies could sit
on the Boards of the institutions.12 This is still true today.
Meanwhile, the work of both the IMF and the World Bank has broadened
and deepened far beyond the purviews of the finance ministries or central banks
with which they are negotiating. This means that, through conditionality and
loan agreements, the Fund and Bank are making finance ministries or central
banks formally accountable for policies which should properly lie within the
scope of other agencies, and for which those other agencies are domestically
accountable. A policy affecting the distribution of health care, for example, we
would expect to be the responsibility of a minister of health, whom we would
expect to be answerable for it to voters and his or her society at large. Yet as the
Fund and Bank intrude further into these kinds of decisions, the risk is that the
line of accountability they establish with the Finance Ministry or Central Bank
will override other agencies and local or democratic accountability.13
A further implication is that, while in theory different agencies within govern-
ment compete for and debate competing priorities and goals, negotiations with
the Fund and Bank distort these debates, subjecting broad areas of policy to the
narrower focus, priorities and analysis of the central bank and finance ministry—
even though neither necessarily has the desire, mandate, accountability or
expertise to evaluate and formulate policy in respect of these broader issues. In a
subtle way, this point is underscored by a remark in the external evaluation of
1
2
3
World Bank Articles of Agreement, Article III, 6.2; IMF Articles of Agreement, Article V, section 1.
Of course, the external line of accountability does not always produce the outcomes desired by the IFIs,
as argued by Paul Collier, ‘Learning from failure: the international financial institutions as agencies of
restraint in Africa’, in Schedler et al., The self-restraining state, pp. 313–32.
1
89

Ngaire Woods
IMF surveillance, the evaluators finding that ‘the most favourable appraisals came
from those whose lines of work bore close similarities to the Fund’s—central
14
banks, and, to a lesser extent, finance ministries.’ Turned on its head, this state-
ment emphasizes the degree to which the Fund’s core mandate remains that
shared with central bankers. Yet at the same time the institution is now
formulating directions for policy in areas outside this formal mandate and
expertise.
In the extreme, the problem becomes that succinctly expressed by Martin
Feldstein in response to the IMF’s intervention in East Asia: ‘The legitimate
political institutions of the country should determine the nation’s economic struc-
ture and the nature of its institutions. A nation’s desperate need for short-term
financial help does not give the IMF the moral right to substitute its technical
15
judgements for the outcomes of the nation’s political process.’ The accounta-
bility problem underscored here may be summed up in three questions: Who
makes particular policy decisions? By whose rules? And under whose scrutiny?
Making the IMF and World Bank more accountable
Aware of the criticisms they face, and also frustrated by their limited effective-
ness in implementing wider policy reform, both the IMF and the World Bank
have begun more explicitly to recognize a wider range of stakeholders in their
work. Both institutions have undertaken a number of steps to make themselves
more accountable to such stakeholders, including more transparency, new
mechanisms of horizontal accountability and working more closely with non-
governmental organizations. The implications of these steps for accountability
are worth examining.
Transparency
First and foremost among the measures taken by the institutions in the attempt
to improve their own accountability is an increase in transparency. Both the
IMF and the World Bank now publish a large amount of their own research and
explanations of what they are doing (and to what effect) on their websites. They
are also pressing governments to permit greater disclosure and publication of
policies and agreements made with governments (which are confidential if the
government so wishes).
Transparency is crucial to improving the accountability of both institutions.
It makes it possible to hold them to account by opening up to observers their
express objectives, advice and agreements. At the same time, however, some
important limitations should be noted. First, to date there are still many gaps in
what is publicly available. For example, the World Bank has an excellent,
1
4
5
IMF, ‘External evaluation’, p. 35.
Martin Feldstein, ‘Refocusing the IMF’, Foreign Affairs 77: 2, March/April 1998, pp. 20–33.
1
90

Making the IMF and the World Bank more accountable
independent unit of evaluation—the Operations Evaluation Department—yet
not all of its publications are available. In the IMF, internal review documents
are not available. These omissions are important, for the outside scrutiny of such
documents not only adds to the external accountability of the organizations, but
also ensures that such reviews are taken seriously within the institutions
themselves.
A second issue raised by the new penchant for transparency is its cost. The
high costs of collecting, editing and publishing information are often under-
estimated by major shareholders pushing for greater transparency. Such costs are
borne in large part by borrowing members of the institutions, since they add to
the running costs of the institutions and thereby to their loan charges. Finally,
transparency of selected data, policy or considerations—and not of others—can
distort decision-making or perceptions of it. Careful consideration and balance
in what is made available is essential.
Transparency provides the most important and necessary, but not the suffi-
cient, means for the international financial institutions to be held accountable.
Ensuring member governments are accountable to their own people for policies agreed with
the IFIs
A second element of the new transparency policies of both the Fund and the
Bank has been to promote transparency within countries with which they
work. The rationale here is that the IFIs advise and assist member countries, but
it is governments in those countries that are accountable for all policies both to
global markets and to their own people. Of course, this formulation glosses over
the fact that some governments feel they have little choice but to accept Fund
or Bank advice and assistance—having resorted to the IFIs for assistance
precisely because they have no alternatives.
The transparency being pursued by the IMF involves releasing their
agreements with countries and other documentation, with the agreement of the
member governments concerned. The result has been the publication of
information such as: Public Information Notices (PINs) following about 80 per
cent of its Article IV consultations, and publishing Letters of Intent (LOIs) and
related country documents that underpin Fund-supported programmes with
respect to about 80 per cent of requests for or reviews of Fund resources.
Going yet further, in some cases the IMF and World Bank now require
governments to consult more and to be more actively accountable to their own
people. At the behest of their largest contributors, both IFIs are requiring
governments wanting enhanced debt relief under the Heavily Indebted Poor
Countries initiative (HIPC) to produce plans setting out how they intend to
reduce poverty. The plan (labelled the Poverty Reduction Strategy Paper or
PRSP) must be ‘nationally owned’ and produced in consultation with ‘civil
society’. Uganda and Mauritania were the first to qualify for enhanced debt relief
under this programme, having each had a pre-existing well-developed plan for
91

Ngaire Woods
poverty reduction. Bolivia also qualified early by producing an interim PRSP
on the basis of a ‘national dialogue’ already undertaken which sets out future
plans for reducing poverty, and for engaging civil society in the formulation
of its full PRSP.16
The new disclosure and consultation measures highlight the sensitivity of the
IFIs to concerns about accountability not just within the institutions, but within
countries with which they are working. Significantly, neither the Fund nor the
Bank any longer describes its interlocutors in member countries exclusively as
national authorities’. Rather, the World Bank writes of ‘development partners’,1
7
‘
and the IMF of ‘authorities and civil society’, and of the need for its pro-
18
grammes to enjoy ‘ownership by the societies affected’. Along with this recog-
nition have come new mechanisms through which these groups can question or
probe the legitimacy of the IFIs’ assumptions and recommendations.
Agencies of horizontal accountability
The terms ‘constitutionalism’ and ‘democracy’ have been used to contrast dif-
ferent kinds of accountability.19 For example, in the US political system, while
democracy is served by the US Congress, constitutionalism is served by the
Supreme Court. This simple contrast usefully highlights the way ‘horizontal’ or
‘
sideways’ accountability can add to vertical accountability by contributing
agencies and processes which exist to monitor and to enforce the mandate,
obligations, rules and promises of institutions.
Within the IFIs, several agencies and processes have recently emerged with
the aim of enhancing horizontal accountability. For example, the IMF has
commissioned three independent external evaluations in the past decade and has
published their reports, and is now in the process of creating an office for
independent evaluation.20
Much more established is the independent evaluation unit in the World
Bank (the Operations Evaluation Department or OED, referred to above),
which reports to the Executive Board. The OED rates the development impact
and performance of all the Bank’s completed lending operations, as well as the
Bank’s policies and processes, and reports its findings to the Board. In 1993, in
the context of a broad review of the Bank’s disclosure policy, access was opened
up to the OED’s ‘Annual Review of Evaluation Results’ and summaries of
evaluation reports (‘Précis’) for selected projects. Since that time, much more of
the OED’s work has become publicly available.
1
1
1
6
7
8
For an updated list of countries preparing PRSPs see www.imf.org/external/np/prsp/.
World Bank, Partnership for development: from vision to action (Washington DC: World Bank, 1998).
IMF, ‘The IMF in a changing world’, remarks by Horst Köhler, Managing Director, given at the
National Press Club, Washington, DC, 7 Aug. 2000.
See comments by Richard Sklar on Guillermo O’Donnell, ‘Horizontal accountability in new
democracies’, in Schedler et al., The self-restraining state, pp. 29–51.
1
9
0
2
See IMF, Report to the IMFC on the Establishment of the Independent Evaluation Office and its Terms
of Reference (Washington DC, 12 September 2000).
92

Making the IMF and the World Bank more accountable
A more powerful and unprecedented step towards greater horizontal
accountability was taken in the World Bank in 1993 when an Inspection Panel
was created by the Executive Board. The establishment of the panel opens up
the possibility for complaints to be made by any group able to show that: (1)
they live in the project area (or represent people who do) and are likely to be
affected adversely by project activities; (2) they believe that the actual or likely
harm they have suffered results from failure by the Bank to follow its policies
and procedures; (3) their concerns have been discussed with Bank management
and they are not satisfied with the outcome.
The three-person Inspection Panel has powers to make a preliminary assess-
ment of the merits of a complaint brought by a group, taking into account Bank
management responses to the allegations. Subsequently, it can recommend to
the Board that a full investigation be undertaken, and make recommendations
on the basis of such a full investigation. The Executive Board retains the power
to permit investigations to proceed, and to make final decisions based on the
panel’s findings and Bank management’s recommendations. The Inspection
Panel thus enhances the power of the Executive Board, as well as of a wide
group of affected ‘stakeholders’ in the Bank’s work.
The most highly publicized of recent cases is that of the Western Poverty
Reduction Project in Qinghai, China. The case resulted from a complaint filed
by the International Campaign for Tibet (ICT), a US-based NGO, acting on
behalf of local people affected by the project, claiming in particular that it
would harm Tibetan and Mongolian people. The final report of the inspection
panel found that the Bank had failed to comply with some of its own policies,
including those on environment, indigenous peoples and disclosure of inform-
ation.21 The case became a notorious one in the media, mobilizing US and
international Tibet campaign groups as well as environment lobbies and
supporters of indigenous groups.
Behind the media glare of such cases, a couple of serious questions of govern-
ance emerge as to the role and implications of such inspections, as highlighted
in retrospect by one of the expert consultants who advised the Inspection Panel
in the China case. Robert Wade points out that an initial problem with the
panel is that its ‘image of success is to find projects out of compliance’; he then
writes that ‘since almost any project can be found to be out of compliance if one
pushes hard enough, and since there is no limit to the cases that affected groups
can bring—assisted by Washington-based NGOs—the Bank is likely to be
deluged with Inspection Panel investigations.’22
The problems hinted at here are worth examining further, for they touch on
the core question of how widely or narrowly ‘accountability’ should be defined,
2
1
2
World Bank, China: Western Poverty Reduction Project Inspection Panel report (Washington DC: World
Bank, 2000); and see <www.worldbank.org/eap/eap.nsf/>.
Robert Wade, ‘A defeat for development and multilateralism: the World Bank has been unfairly
criticised over the Qinghai Resettlement Project’ (full manuscript unpublished, short version published
in Financial Times, 4 July 2000).
2
93

Ngaire Woods
and what kind of breach should trigger an enforcement action. If the ‘triggering
mechanism’ for inspection is unlawfulness, this not only presumes that actions
that are legal must also be legitimate in the eyes of the citizenry, but opens up
the risk that minor ‘legal’ infractions can be used as a weapon for much larger
political purposes. This point has been made in a study of accountability in
which one scholar, alluding to the Clinton/Lewinsky scandal, reminds us that
‘
minor legal infractions can be used by partisan opponents to thwart the clearly
23
expressed preference of the public-at-large.’ In other words: beware of the
fact that agencies of horizontal accountability can be abused.
The warning about the abuse of inspections or horizontal accountability
agencies is an important one, because inspections cost money and take time.
Wade reports that the East Asia region of the Bank spent about $3 million on
work responding to the panel’s investigation of Qinghai, in addition to the
extra costs incurred by the Chinese government. The cost of the extra work
that the East Asia region proposed be done came to another $2.5 million. The
cost of the extra work that the panel report calls for is estimated at around $4
million, or 10 per cent of the total loan. Wade reports that this cost (and the fear
of an inquisitorial process) means that Bank staff are now refusing to contem-
plate projects involving either involuntary resettlement or indigenous peoples,
because they cannot compete with other sources which do not have to take into
account such high additional costs. This, he implies, is ultimately to the detri-
ment of the disadvantaged groups the Bank is setting out to assist.24
An alternative model of accountability, which avoids some of the problems
sketched above, is that provided by the new Office of Compliance Adviser/
Ombudsman (CAO) of the IFC and Multilateral Investment Guarantee Agency
(MIGA). This new ombudsman’s office was created in June 1999, after
consultations with shareholders, NGOs and members of the business com-
munity aimed at finding a workable and constructive approach to dealing with
environmental and social concerns and complaints of people directly affected by
IFC and MIGA financed projects. The CAO or ombudsman and her staff are
independent of the Bank and IFC and report directly to the President of the
World Bank. The emphasis of the office’s work, however, is very much on
dialogue, mediation and conciliation. Other than the power to make recom-
mendations, the CAO has no formal powers. Indeed, the draft operational guide-
lines of the office state: ‘The ombudsman is not a judge, court or policeman.’
In the absence of enforcement powers, one must ask whether an ombudsman
can really be considered a mechanism of accountability. Clearly the CAO office
provides for transparency and monitoring, and these are vital to accountability.
It also provides for a very light form of indirect enforcement. For this reason, it
avoids the costs highlighted in respect of the Inspection Panel above, and
possibly also the incentive for users to abuse the process of accountability in the
2
3
See comments by Philippe Schmitter on Guillermo O’Donnell, ‘Horizontal accountability in new
democracies’ in Schedler et al., The self-restraining state, pp. 29–51.
Wade, ‘A defeat for development’.
24
94

Making the IMF and the World Bank more accountable
pursuit of other goals. However, it remains to be seen whether this mechanism
has enough clout to hold decision-makers to account. More generally, such
mechanisms cannot be seen as sufficient in and of themselves to patch up the
accountability of international institutions. They operate alongside the vertical
accountability already outlined above, affording another ‘check’ on IFI officials.
The experiments in compliance enforcement being undertaken in the World
Bank and IFC highlight how little horizontal accountability exists for the inter-
national financial institutions. Obviously the primary agencies which should
hold the institutions to account are their member governments, through bol-
stered and improved forms of vertical accountability. However, both IFIs are
now working in a world political system in which groups both within and across
countries are becoming more effective at demanding more account of the work
of international organizations, both through governments and directly from the
organizations concerned. It is for this reason that horizontal accountability has
become a large plank in both IFIs’ responses to those who criticize their unac-
countability. A further part of this response has been to engage more directly
with their critics, and in particular with non-governmental organizations.
Engagement with non-governmental organizations
In recent years both the IMF and the World Bank have begun to recognize non-
state actors and non-governmental organizations (NGOs). The trend has been
acknowledged by the US Secretary of the Treasury in discussing the Fund’s
modus operandi:
it should become more attuned, not just to markets, but the broad range of interests and
institutions with a stake in the IMF’s work. Just as the institution needs to be more
permeable for information to flow out, so too must it be permeable enough to let in
new thoughts—by maintaining a vigorous ongoing dialogue with civil society groups
and others.25
Both the Fund and Bank make much more information and analysis available to
NGOs than was formerly the case. The World Bank’s NGO–World Bank
Committee (established in 1982) has become more active. Both the Bank and
Fund now consult with lobbying organizations in Washington DC and with
grassroots organizations in member countries: trade unions, church groups and
other such bodies. These contacts are taking place at regional, country and local
levels. World Bank regional directors and IMF resident representatives are
being told to seek out and maintain such contacts. At the annual and spring
meetings both institutions have been actively involved in more dialogue and
meetings with a select group of transnational NGOs. In addition to these
measures, which increase transparency and consultation, the institutions have
2
5
Lawrence Summers, ‘The right kind of IMF for a stable global financial system’, remarks to the London
Business School, 14 Dec. 1999 (US Treasury: LS-294).
95

Ngaire Woods
also more recently begun to require some level of local participation by non-
state actors, such as in the poverty reduction strategy papers being asked of
countries seeking enhanced debt relief.
It is worth clarifying that NGOs have not taken a place as major ‘stakeholders’
in the institutions: they have not acquired control, nor a formal participatory
role in decision-making (except at the behest of their own governments).
However, where ‘Northern’ NGOs have allied with or used political leverage
in major shareholding countries—at any rate in the United States—they have
exercised considerable informal power and influence. Indeed, in such cases the
position of some NGOs starts looking much stronger than that of many smaller
developing countries—whose formal right to participate in decision-making is
diluted by the problems of representation described earlier. For this reason, the
recognition of NGOs as stakeholders has led to a vociferous debate—in
particular among developing country governments—about the accountability
and legitimacy of the NGOs themselves.
In further analysing this debate it is essential to distinguish local or ‘Southern’
NGOs, within borrowing countries, from transnational or ‘Northern’ lobbying
organizations, usually based in Washington DC or one of the G-7 capitals. The
implications for the accountability of the IFIs of developing relations with each
are somewhat different.
Engagement with Southern NGOs Local or Southern NGOs are stakeholders in a
direct sense of the term: they represent groups directly affected by the pro-
grammes and policies of the IFIs. Their inclusion in discussions and strategy
formulation is required because of the way in which the activities of the IFIs
have broadened. Both IFIs recognize that, to quote the World Bank, ‘policy
26
reform and institutional development cannot be imported or imposed’. In
countless publications, both IFIs recognize that wider participation and owner-
ship are required for policies to be successfully implemented.27 For these
reasons, the IMF and the World Bank are encouraging both their own local
representatives and government officials (for example, in the PRSP process
outlined above) to develop consultative links with local NGOs. At the same
time, NGOs now have access to the complaints procedures described above
(World Bank Inspection Panel and IFC ombudsman).
However, new relations with Southern NGOs do not resolve the problems
of accountability faced by the IFIs, even if they add some positive elements to
the picture. Neither the Fund nor the Bank has been structured as an agency of
2
6
7
World Bank, Annual Review of Development Effectiveness (Washington DC: World Bank, 1999).
See World Bank, Sub-Saharan Africa: from crisis to sustainable growth (Washington DC: World Bank, 1989),
Governance and development (Washington DC: World Bank, 1992), Governance: the World Bank’s experience
2
(
Washington DC: World Bank, 1994); The World Bank participation source book (Washington DC: World
Bank, 1996), and Annual Review; IMF, Good governance: the IMF’s role (Washington DC: IMF, 1997);
External evaluation of the ESAF,’ report by a group of independent experts (Washington DC: IMF,
998); A guide to progress in strengthening the architecture of the international financial system (Washington DC:
IMF, 2000).
‘
1
96

Making the IMF and the World Bank more accountable
forceful, far-reaching domestic reforms. Their governance structure gives them
none of the necessary elements of either legitimacy or accountability for such
tasks. They are being forced for practical and political reasons to look beyond
their traditional and narrow points of contact with ‘national authorities’ (finance
ministries, central banks) to the wider ‘civil society’; but such relations should
not divert attention from the core lack of accountability to developing country
governments. Are the IFIs beginning to step around governments and them-
selves attempt to persuade societies to support certain kinds of reform? The
problem here is that the government itself should be the agent of persuasion;
but borrowing governments will not be able to persuade their own societies to
accept changes required by international institutions in which they (developing
country governments) cannot claim to have an effective voice.
Relations with NGOs also bring problems of accountability and legitimacy
of their own. Foremost is the question of which ‘civil society groups’ ought to
be consulted or recognized. As Jan Aart Scholte has documented in the case of
the IMF, some non-governmental groups (such as business groups) are being con-
sulted much more than others.28 This poses the question: To whom are local
NGOs accountable, and for what? Until now, both the IMF and the World
Bank have been pursuing relations with NGOs in an ad hoc and reactive way.
However, as Charles Abugre and Nancy Alexander have noted, more systematic
consideration need to be given to criteria such as the effectiveness, representa-
tiveness, internal decision-making structure, membership and accountability of
groups being so consulted.29
The problem of accountability of local NGOs is further heightened where
they are given a more active role in the formulation or implementation of
policy. In such cases, the accountability of local NGOs ought to be compared to
that of local government agencies. In the 1980s and early 1990s it was some-
times too readily assumed that the former were preferable; and in the World
Bank’s work in Africa this led to criticisms of the Bank for undermining the
capacity of governments in the region.30 A decade later, it has become more
obvious that using NGOs to bypass government institutions risks thwarting the
desired processes of ‘institution building’ and ‘state modernization’.
A final problem concerning relations with Southern NGOs is the question of
who shapes and influences the modalities and processes of consultation with
such groups. In the PRSP process described above, no framework was spelled
out. There is nevertheless a risk that certain kinds of consultations become
recognized in a de facto way as preferable to others, not on the grounds that they
enhance existing mechanisms of local accountability, but rather on the grounds
that they please the political sensitivities and preferences of major shareholders
2
8
9
Jan Aart Scholte, ‘The IMF meets civil society’, Finance and Development 25: 3, Sept. 1998, pp. 42–5.
Charles Abugre and Nancy Alexander, ‘Non-governmental organizations and the international,
monetary and financial system’, International monetary and financial issues for the 1990s, vol. 9 (Geneva:
UNCTAD, 1998), pp. 107–25.
2
30
Ibid., p. 114.
97

Ngaire Woods
in the IFIs. In other words, there is a risk that a new ‘Washington consensus’ on
the politics of participation and consultation will be forged, and that this will
pay insufficient attention to the complex social and political arrangements
which give life to accountability at the local level.31
Nonetheless, the vociferous debate about NGOs and their lack of accounta-
bility risks being overplayed in the context of Southern NGOs. Certainly they
now have access to more information; and transparency is a powerful step
towards holding governments and institutions accountable. NGOs are also,
with government approval, consulted more regularly. In the World Bank and
IFC they have the right (although not necessarily the resources) to access the
new complaints procedures. These steps, however, do not amount to a transfer
of decision-making power or substantial influence. Indeed, the argument has
been made that Southern NGOs should be strengthened and more strongly
used by developing country governments, (1) in order to enhance their own
information and analysis about the IFIs; (2) as a bargaining counter (in alliance
with Northern NGOs) to pressures by major shareholders, who face demands
from their own publics to pay attention to NGOs; and (3) to counter the power
of Northern NGOs.32
New consultation mechanisms with NGOs in borrowing countries do not
resolve the deeper problems of accountability both within the IFIs and in
relation to their government interlocutors. Indeed, these problems may well be
exacerbated by new developments in the IFIs’ relations with the other category
of new lobbyists and stakeholders: Northern NGOs.
Engagement with transnational and Northern NGOs Transnational NGOs do not
have the same stake in the IFIs as directly affected local groups. In defining their
‘
stake’ in international organizations, one needs to consider their broader place
in global governance. Even there, they are not part of the state-based system of
representation in world politics. But politics is not just about representation; it is
equally about debate, and in international politics, NGOs open up and contri-
bute to an active debate about the IFIs and their policies. Their role is to speak
for different views and interests which are not necessarily expressed through the
formal channels of representation, i.e. to act as lobbyists for particular interests
relevant to notions of justice, development, the environment and so forth.
In more academic terms, the argument for the place of NGOs in an inter-
national ‘deliberative democracy’ is voiced by theorist James Bohman: ‘In the case
of a political setting where there is no public to whom appeal can be made or
institutions in which voice is important, international institutions and the NGOs
31
For the original statement of the Washington consensus see John Williamson, Latin American adjustment:
how much has happened? (Washington DC: Institute for International Economics, 1990). For subsequent
analysis see Moises Naim, ‘Washington consensus or Washington confusion?’, Foreign Policy, no. 118,
2000, pp. 86ff; Joseph Stiglitz, More instruments and broader goals: moving towards the post-Washington
consensus (Helsinki: WIDER/United Nations University, 1998); and Robin Broad and J. Cavanagh, ‘The
death of the Washington consensus?’, World Policy Journal 16: 3, 1999, pp. 79–88.
Abugre and Alexander, ‘Non-governmental organizations’.
32
98

Making the IMF and the World Bank more accountable
that attempt to influence their procedures and standards themselves can function
as a public to whom equal access of political influence is guaranteed and open.’33
The accountability problem posed by transnational NGOs (TNGOs) lies first
and foremost in the question: To whom are they themselves accountable? Most
TNGOs are accountable in various ways to at least three constituencies: their
membership (actual and potential, and predominantly Northern); their major
funders and/or clients (which include governments and corporations); and in
many cases NGO partners (some in the South). Overall, there is a tendency for
such groups primarily to be accountable to Northern groups, funders and part-
ners. For this reason, a long-standing concern about TNGOs has been that they
magnify Northern views—both outside governments, and through govern-
ments—in the international organizations, adding yet another channel of influ-
ence to those peoples and governments who are already powerfully represented.
The risk here is that TNGOs further distort the inequalities of power and
influence already emphasized in this article. As Abugre and Alexander found:
‘
Activism by United States NGOs has probably expanded the already dispro-
portionate role of the United States in the international financial institutions,
especially the World Bank.’34 Similarly, where TNGOs deploy their consider-
able resources and expertise in representing and assisting local groups bringing
cases to the Inspection Panel, it is not always the case that the interests of those
being represented are the same as the interests of those to whom the TNGO is
primarily accountable (members and funders). The TNGO has an incentive to
favour an outcome which maximizes publicity and support for itself; yet in
some cases the quieter, compromise decision may well do more for the group it
claims to represent.
The difficulty for critics is that, while the work of many TNGOs has un-
doubtedly magnified Northern influence in the IFIs, other TNGOs have used
their influence with the US government (both Congress and the Executive) and
other G-7 governments, effectively to campaign for greater transparency,
disclosure and new forms of horizontal accountability, which are of interest to
all stakeholders. Thus some TNGOs have assisted in enhancing the accounta-
bility of the IFIs, while at the same time further undermining the relative power
and participation of both developing countries and Southern NGOs.
The devil with TNGOs lies not so much in the detail as in the objectives,
priorities and constitutions of the organizations. Some TNGOs working on and
with the IFIs are fully aware of the risks mentioned and steer clear of them by
prioritizing multilateralism, local accountability and capacity-building in develop-
ing countries. In so doing such groups are implicitly respecting the claims of
more directly affected stakeholders to have priority in holding the institutions to
account. TNGOs that do not so delimit their priorities, it is argued here, are
wrong to claim that their stake in the institutions gives them a right of account.
3
3
James Bohman, ‘International regimes and democratic governance: political equality and influence in
global institutions’, International Affairs 75: 3, 1999, pp. 499–513, at p. 511.
Abugre and Alexander, ‘Non-governmental organizations’, p. 116.
34
99

Ngaire Woods
Conclusion
Improving the accountability of the IFIs has become a mantra for economic
policy-makers in the G-7. In the present debate this has come to mean making
the institutions more transparent and opening up dialogues with new groups of
stakeholders. This article has shown that these measures may improve
governance, but do not solve the core accountability deficit in the IFIs. The
institutions are supposed to be directly accountable to their member govern-
ments through their Executive Boards. Yet there are deep flaws in this
structure. Furthermore, as the activities of both IFIs expand to embrace wider
and deeper conditionalities, the accountability deficit is further widened. This is
because, just as the activities of the Fund and Bank are widening, so too are the
categories of affected groups who might legitimately claim to be stakeholders in
the institutions and therefore to have some right to hold them to account.
The implications for the institutions are several. First, the flaws in the vertical
structure of accountability need to be addressed. This means working to put
into place:
an open and legitimate process of appointment for the heads of the institutions;
a stronger role for the Executive Boards in overseeing the work of the
institutions;
•
•
a structure of representation which better reflects the stakes of all state-
members; and
•
measures which assist in enhancing the accountability of Executive Directors
to their governments and voters.
•
These measures, along with positive steps towards independent evaluation,
inspection and transparency, would doubtless improve the accountability of the
IMF and World Bank. However, ultimately there are limits to how accountable
the IMF and World Bank can be to the governments and peoples most affected
by their lending and policies. This raises a more profound issue as to how far-
reaching the activities of relatively unaccountable agencies should be.35 In his
essay on whether international institutions can be democratic, political theorist
Robert Dahl warns that we should be ‘wary of ceding the legitimacy of demo-
36
cracy to non-democratic systems’. His point is that domestic political systems
have a potential to be democratically accountable in a way that international organ-
izations cannot. The implication is that the IMF and the World Bank should be
reined in from far-reaching policy conditionality. Their activities should be
limited to those for which they can claim to be effectively accountable.
3
5
6
I am grateful to Devesh Kapur for spurring me to think about this issue.
Robert A. Dahl, ‘Can international organizations be democratic?’ in Ian Shapiro and Casiano Hacker-
Cordon, eds, Democracy’s edges (Cambridge: Cambridge University Press, 1999), pp. 19–36, at p. 33.
3
100