Structural Adjustment for the IMF
Briefing||19 January 2001|
Options for Reforming the IMF's Governance Structure
By Angela Wood
Three forces are shaping change at the International Monetary Fund (IMF). The first is in relation to the financial crisis that swept across the globe in 1997 and 1998, which has led to pressure to restructure the IMF to address the demands of a liberalised global financial market. Attention has focussed on increasing the IMF's surveillance function, giving it new responsibilities to monitor and advise governments on financial sector restructuring and enabling it to operate as a quasi-lender of last resort.
The second has arisen from the pressure for debt cancellation to be linked to poverty reduction objectives and the acceptance by donors and the International Financial Institutions (IFIs) that two decades of structural adjustment have not achieved lower levels of poverty. Reform effort is focused on re-examining the IMF's policy prescriptions, limiting the extent of its conditionality, facilitating its engagement with governments and civil society and reviewing its operational procedures.
The third is a highly critical force coming up from the grass roots calling for far more radical reform, if not closure of the IMF. This force is critical of the IMF's ability to act as an institution of global governance. A major objection is that it is controlled by the interests of a few elite, prohibiting it from taking into account the views of a wide range of citizens. The focus is on improving its engagement with, and accountability to, civil society at large.
These forces and the reforms they are proposing have several implications for the IMF's governance structure. In particular they raise questions about:
Whilst the voices of the rich governments have been heard very clearly in the financial architecture and poverty reduction debates, poor country governments, particularly the poorest, have remained silent. This silence reflects their lack of power in the global economy, which is reinforced, by the lack of democratic governance at the IMF.
This has allowed the G7 to foister limited "solutions", which suit their own interests, onto already overstretched developing country governments. The mechanism for doing so has been the IMF. Their dominance of the IMF has allowed them to dictate the agenda of the IMF Board, even as the new Managing Director, Horst Köhler, boasts that the IMF is a vital institution because it represents all interests.
Creating space for developing country governments to engage in dialogue and put forward their views is vital but it is also necessary to facilitate the involvement of a wider set of actors in the IMF's governance structures. The IMF's policy prescriptions have a much wider impact at the national level than previously thought, yet it is doubtful that the Board is effective at holding the staff to account, and often governments are more accountable to the IMF than they are to their citizens. Ways need to be found to ensure that the IMF is directly accountable to civil society. Also, civil society can help strengthen governments' capacity to hold the staff to account.
This paper identifies a number of reform proposals to enhance the IMF's governance structure and accountability to a growing number of stakeholders. Part 1 looks at measures to improve the IMF Board's decision making process; part 2 identifies reforms to improve the accountability of IMF management to the Board, and Board accountability to the public; and part 3 looks at how to improve management and staff's accountability to the public.
Three aspects of the IMF's governance structure need to be adjusted so that it can operate effectively as an institution of global governance in the new global architecture. These are: representation and power in the IMF's governing bodies; transparency and accountability of the Board; and transparency and accountability of the staff and management. Action must be taken in all three areas. Whilst some of the suggested recommendations will have to be implemented over the longer-term, because, for example, they require a revision of the Articles of Agreement, a number of the suggestions could be carried out more quickly. All that is preventing them is the lack of political will.
Improving representation in the IMF's governing bodies
Short term actions:
Increasing transparency and accountability of the Board
Transparency and accountability of the staff and management
Improving the decision making structures
"Developing countries are entitled to a stronger and more effective voice in all of these international institutions." (Department for International Development, 2000)
A major problem with the current IMF reform process is that developing countries are marginalised within it because they are not well represented in its decision-making bodies. This is partly attributable to the fact that the IMF's governance structures have not changed to accommodate the changes in its clientele. Developing country governments, and the people they represent, are penalised firstly, because they do not have many seats on the governing bodies, which means their interests are not represented; and secondly, they do not have an equitable proportion of the votes so their power to influence decisions is minimal.
The IMF was established in 1944 when the US was the most powerful economy and Europe and Japan were the IMF's principle clients. At this time developing countries were still colonies and excluded from the negotiations. As a consequence a power structure was developed which heavily favoured the US with less power to Europe and Japan. Since that time, despite significant changes in the global economy and in the IMF's principal clients, the power structure within the IMF has remained virtually unchanged. Today the situation is one whereby the rich countries control the institution but don't borrow from it whilst the middle income and poor countries borrow but have no effective input into decision making. (De Gregorio et al, 2000)
Control and influence over the IMF management and staff is exercised through the Executive Board (EB) and the International Monetary and Finance Committee (IMFC). They are poor tools for maintaining accountability, because the influential Executive Directors (EDs) and governors, those from the rich, creditor countries, only take an active role on issues of interest to them, i.e. the institution does not respond to a "clients" agenda. More generally, the Board is very reliant on the IMF management, which is able to dominate the agenda setting processes and control information flows to the Board.
The rich countries justify their dominance of the governance bodies on the basis that they provide the most money to the IMF and they are strategically important in the global economy. This argument has yielded a double dividend for them, firstly, they have linked voting rights to the share of the IMF's finances (quota) each country contributes; and secondly, they have also granted the wealthiest countries permanent seats on the Board and IMFC.
There is little justification for this concentration of power amongst the rich when smaller countries such as Thailand can have a major impact in the global economy. Although developing countries provide less in the way of the IMF's general resources they are contributing to the replenishment of the Poverty Reduction and Growth Facility (PRGF) in addition to their quota. This is significant because it is the only facility from which most of the poorest can borrow; it is also the facility which generates the majority of the IMF's programme work. Moreover, it is the interest earned on the IMF's loans , which finance its day to day running costs. More fundamentally, no national democracy is founded on a dollar per vote system.
1. Representation on the Board and International Finance and Monetary Committee
"In effect, the IMF is not a "revolving fund", as its staff like to claim, but a financial institution comprising "structural creditors", who dictate loan conditions, and structural debtors, who accept the said conditions." Oxfam, 2000 p4
Only 24 Executive Directors represent 182 member countries on the IMF's executive board. Eight countries have their own ED: China, France, Germany, Japan, Russia, Saudi Arabia, the UK, and the US. Other countries are grouped into constituencies, each one represented by an ED who votes on behalf of the constituency. These constituencies tend to be represented by the most economically powerful country in the group , for example Canada and Italy both represent their constituencies, which means that all the G7 countries have a representative on the Board: as do Belgium, the Netherlands, Spain, Denmark, Australia and Switzerland. This leaves only 8 seats (plus China) held by developing country representatives. In the worst case, 43 sub-Saharan African countries are grouped into only two constituencies and are represented by only 2 EDs.
Similarly, although each country appoints a governor to the IMF, typically the Finance Minister or Central Bank Governor, the International Monetary and Finance Committee (formerly the Interim Committee) is comprised of only 24 governors which represent the entire membership. Representation is constituted in the same way as for the Executive Board.
The constituencies vary in size, from 4 to 23 countries, and tend to be loosely organised along regional lines. There is little coherence in terms of grouping countries according to whether they are low income, middle income or rich countries: some constituencies have a mix; some tend to be middle and low income countries; and some are mostly rich countries. How countries are grouped is important in terms of how effectively their interests are represented on the Board and in the IMFC.
Although EDs representing constituencies are supposed to represent all interests equally, they tend to prioritise those of their own country. In some cases involvement of other members in the decision making process is regarded as "interference". In the case of the industrialised country groupings this may not be such a problem because there is a tendency for them to think alike. However, when it comes to mixed constituencies, in which the richer countries invariably occupy the ED seat, there is concern that the ED representing them is not effectively raising developing countries' points of view and concerns. Although one ED has suggested that being in a very mixed constituency has enabled him to understand better the needs of developing countries.
Whilst there is a question about the willingness of EDs to represent the views of other constituency members there is a related issue about whether all EDs have the same capacity to effectively represent their constituencies. For example, the two African constituencies include the largest number of countries but the African delegations have relatively fewer resources than the other EDs, particularly compared to those EDs that represent only one country. This lack of capacity is reinforced by the fact that the ED seat tends to revolve every couple of years in constituencies comprising mostly middle and low-income countries. This limits the opportunity for institutional learning, "the result is that, de facto, the G7 countries control the policy agenda in the IMF." (Bradlow, 2000).
It has been commented that developing country EDs often address the Board so there can't be a representation problem. To some extent this is true because they are frequently required to represent their constituents who are the IMF's principle clients. However, when it comes to general discussions of policy or direction, it is often remarked that the developing country EDs tend to make very broad statements and do not address the details, which is often where the devil lies. The problem would seem to be that they don't have sufficient capacity to brief themselves to do both jobs.
One option is to increase the capacity in larger constituencies by increasing the number of staff. The number of delegation staff has recently been increased , whilst this is welcome, the larger constituencies will still require more. One option could be to transfer staff from the single ED offices and smaller constituencies to the larger ones and/or employ more staff, or alternatively constituencies could be reduced in size.
1a. Reformulating the size and structure of constituencies
Firstly, a limit could be placed on the number of countries that a single ED can represent. With 24 EDs and 182 countries, each constituency could be limited to no more than 8 countries (originally EDs represented no more than 5 countries), with no single ED representing just one country. This will have implications for how countries are grouped. At the moment countries are able to choose which constituency they want to be in, under a more "equitable" system constituencies may need to be allocated.
Alternatively, constituencies could be reduced in size if there was an increase in the number of EDs on the Board. One option would be to raise the number of developing country EDs so that there are equal numbers of developing and developed country representatives. Since 1944, the number of members has increased from 39 to 182 countries whilst the number of EDs has only increased from 12 to 24 (Bradlow 2000). Given that it has largely been middle and low income countries that have joined the IMF in recent years, this would suggest there is a good case for doing so.
1b. Allocating constituencies
Constituencies could be reallocated along various different lines. For example, they could be grouped:
Ideally, it would be preferable if constituencies could include a number of these features, but this is probably not easily achieved and some trade-off will need to be accepted. It seems sensible to group countries along regional lines because that will help to ensure a more common perspective and cohesion. This may conflict with the first option, which would bring the most benefit in terms of addressing the range of interests represented on the Board, although it would not address their capacity problems. The second and third options could provide a better mix in each constituency but may not achieve better representation unless EDs were selected more democratically and mechanisms could be developed to ensure all views are represented (voting reform is one option, see below).
The IMFC is equally as biased as the EB. There is no legitimate reason why it has more developed country members than developing country members; thus it should be reformed along similar lines. The most equitable of which would be to ensure an equal number of representatives from rich, middle and low-income countries.
1c. Creating sub-boards.
Woods (2000a) suggests establishing a sub-board, comprising those governments particularly affected by IMF decisions and actions, i.e., developing countries broadly, emerging market countries and those countries which have opened up their capital account. This should allow those affected directly by IMF policies to be better involved in the decision making process.
An alternative, would be to create two different boards (although some members would be on both) to reflect the different roles of the IMF. One could deal with financial stability and short-term emergency situations in all members, and the other could focus on longer-term development and poverty reduction objectives in the poorest countries. The composition of the boards should reflect the countries most strategically involved in the issues, although equity should also remain a priority; i.e. there should be representation from all country income groups on both boards but perhaps with a bias towards the principle users. In the latter case, the Board could comprise all developing country EDs reflecting both the fact that these countries are the users of the Poverty Reduction and Growth Facility funds and contributors to it. Of course, there could be problems if one Board is considered to be marginal to another, care would have to be taken to ensure that they had equal power and status.
Another option could be to have one board representing developing countries and the other representing developed countries. Each would be required to take votes and actions would only be taken if both Boards were in agreement. Alternatively, one board could comprise EDs from finance ministries and the other from development ministries, to reflect the broader range of issues the IMF now engages with. This would not necessarily increase the number of staff because each ED already has an alternate. The point would be to give an increased role to the alternates. Two boards may seem unwieldy but it is quite typical at the national level to have two decision making bodies, such as the Congress and Senate in the US and the House of Commons and House of Lords in the UK.
As well as increasing the level of representation, another advantage of having two boards is that it would increase the EDs' capacity to monitor staff and to take informed decisions on programmes etc. (see 3d below).
2. Voting Power and the Quota Formula
"The current quota distribution also makes it possible for the industrial countries in the G7 to take decisions which they are able to push through in the Bretton Woods Institutions. In that sense, the, BWI serve a "legitimizing" function for decisions already taken by a small sub-set of the membership." (Mohammed 2000)
Reforms, which allow developing countries a greater voice in Board discussions, may not lead to a change in the general direction of decisions if the allocation of votes remains heavily skewed in favour of the rich countries.
Voting power is tied to the subscriptions (quotas) paid to the IMF by member countries. The more a country pays in the more votes it has; and the more it is eligible to borrow. Subscriptions are worked out according to a formula based upon a country's relative economic and political strength. Occasionally the formula is reviewed, the latest review - known as the Cooper Report - has been criticised for failing to fundamentally re-examine the quota formula.
The next allocation of quotas is likely to happen in a way that will allow a few "emerging market" countries to have a greater share of the votes in recognition of their economic strength in the global economy. It will not achieve an allocation in favour of the poorer countries.
2a. Reviewing the current formula
Many EDs argue that voting distribution should not be a concern because formal votes are rarely taken, instead decisions are reached by consensus. This is misleading because in fact votes are implicitly taken, i.e. EDs indicate how they would vote and the votes are informally totalled. So the allocation of votes is important because it affects the way that countries perceive decisions are going, "decision-making in international organizations is often done by "consensus" rather than via formal voting, but even then the underlying voting power influences outcomes profoundly." (Helleiner, 2000 p34)
The composite quota formula is calculated on the basis of several factors, some of which can not be accurately measured or clearly defined. Thus, the formula has been open to a certain amount of manipulation, which has given advantage to the richer countries who have been able to maintain a politically favourable allocation of seats. In fact, the "quota formulas themselves have had only an indirect influence on actual quotas" , their results have been used merely as guidelines during negotiations. The problem for the IMF is that it is difficult for it to claim to be representative if it cannot show on a rational basis how its members are represented on the Board.
2b. Improving the quota formula
"It [the IMF] should move over time toward a governing structure that is more representative and a relative allocation of member quotas that reflects the changes under way in the world economy -so that each country's standing and voice is more consistent with their relative economic and financial strength." (Summers, 1999)
To improve the quota formula, alternative variables have been suggested which would lead to a more equitable allocation of quotas and therefore voting rights amongst members, including population size, national income, and access to capital markets.
Improving the formula in terms of aligning it more closely with national income or capital market access, as Lawrence Summers suggests, will not give more weight to the poorest countries (ODC 2000). A better option would be to include population size since many of the poorest countries have the largest populations. This would ensure that "people", and therefore developing countries, were better represented.
A review of the quota formula has just been completed. Unfortunately, the Quota Formula Review Group's report merely proposes to simplify the existing formula whilst ensuring it produces results which more closely match the current allocation of votes. In fact, rather than giving a more equitable distribution of votes, if the proposed formula was to be adopted, it would increase the share of the US and decrease the number of votes to the poorest countries.
Even if the formula was revised considerably to the advantage of poorer countries, rich countries' quotas will be largely protected if quota increases continue to be "general" rather than "selective". For a general increase, the bulk of the quota increase is calculated as a uniform percentage of existing quotas. Only a small amount is available for selective distribution, "the balance between the equiproportional and selective elements in a quota review has largely been a matter of judgement by the Executive Board. The proportion of all quota increases allocated to the equiproportional and selective elements has, on average, been 70 percent and 30 percent respectively." If revising the quota formula is to make a difference, quotas will have to be reallocated selectively.
Another suggestion (Woods, 1997; Bradlow 2000) is to increase the proportion of the fixed "basic" vote each country receives. Each country is allocated 250 basic votes on top of which they receive additional votes according to the quota formula. Whilst the proportion of votes attached to the quota formula has increased the basic vote has remained unchanged. Thus the "equity" function of the basic vote has diminished. Although raising the basic vote would increase the voting power of the poorer countries it would do so only marginally, unless there was a very significant increase.
2c. Multiple formulas
Another suggestion is to break the link between the quota, ability to borrow and votes, "there is no longer a good rationale to keep the three tied together. On the contrary, there are good reasons to separate them to better to meet new standards of accountability, transparency and governance." (Woods, 2000b). Thus, the alternative would be to develop different formulas for the different aspects. At the very least, there is a case for de-linking the allocation of voting rights from the quota formula as Jha and Saggar (2000) suggests.
2d. One country One vote
In the past there have also been suggestions that since the IMF is part of the UN family it should adopt the UN's "one country one vote" system. Whilst this is obviously the most democratic way of allocating votes, many officials in developed countries object to this on the grounds that reaching agreement would be stymied. They perceive that the constituency, "consensus-based" system is a more effective way of making decisions. However, it would be possible to maintain the current system of constituencies, but require that each country casts a vote. For day to day decisions a simply majority should be sufficient with a super majority required for more fundamental decisions, as the current rules specify.
2e. US veto vote
It is widely regarded as unacceptable that the US should have the power to veto key decisions. The rules currently state that important decisions require an 85% majority, i.e. a "super-majority", of the Board to be in agreement. This means that the US, which controls 17.56% of the votes, can veto key decisions.
Oxfam (2000 p4) argues that "Whilst it is inevitable that the major shareholder will carry more weight than others, restricting the vote of any one country to 10 per cent would help to correct the current imbalance." Ironically, in order to change the rules so that no single country can have such a veto, there would need to be a super-majority. However, there is a case to be made for changing the rules. Both the World Bank and the Inter-American Development Bank have an arrangement by which the largest shareholder is not able to exercise a veto and this appears not to have harmed US interests in these institutions (ODC 2000). Moreover, in relative terms, as quotas are currently allocated, the US has a greater share than it should do (De Gregorio et al, 1999).
Transparency and independence of the board
It is extremely difficult to hold either the staff or the Board accountable for decisions taken. Whilst the Board is nominally responsible for taking them, the EDs often state that they do not have access to all information and that they often defer to the staff's point of view i.e. the Board tends to merely "rubber-stamp" decisions advocated by the staff. However, the EDs also complain that they are overloaded and are not able to take an active interest in all aspects of the institution's work, thus they must rely on the staff to guide their decision making.
Accountability is further undermined by the fact that formal votes are rarely taken and few details of discussions are made public. The Board has taken a few minor steps to improve its transparency but on the whole there is much room for improvement.
3. Independence of the Board
There are two concerns about the Board's lack of independence. The first is that the rich countries, principally the US, are able to dominate the Board and prioritise their national objectives, i.e., the US government uses the IMF as an extension of its foreign policy tools. Examples of this are the US' insistence that the IMF should lend to Russia in the early 1990s, and its inclusion of trade conditions in South Korea's 1997-98 agreement with the IMF.
The second is that, Board members have only limited access to the information held by IMF management and staff (except the US). Thus the Board's decisions reflect the information the staff provides them with, i.e. the staff and management can control the agenda and debate. This limits the usefulness of the Board's function as an accountability mechanism. Moreover, based on the argument that the Board has limited capacity to engage in discussions about the content of individual countries' programmes, the staff never present the Board with a programme "that it cannot agree to". That is, staff disagreements are resolved internally and are not taken to the Board for discussion, "there is some complicity of the staff and management in the weak role played by the Executive Board. Staff and management seldom divulge internal disagreements to the Board, tending instead to attach "considerable importance" to presenting a unified view in Board discussions. The result is that real debates over policy and issues are conducted outside of the Board within the organization prior to Board meetings." (Woods, 2001)
A related problem is that "in addition to the Board structure, representation is compromised by the fact that Board members themselves are simply overwhelmed with information on the large number of country programs, making it difficult for them to focus on any one country", (ODC, 2000). Here the issue is that Board members have to make too many decisions about which they don't have full information, i.e., because of the weight of business, they are not able to brief themselves sufficiently well. Thus, they are very reliant on the staff's opinion. This lack of capacity leads to the suspicion that the Board is simply "rubber stamping" decisions which means that it is not exerting significant control over the IMF management and staff (Woods, 2001).
Several solutions suggest themselves:
3a. Independence from nation states
For De Gregorio et al, and many other critics, the fundamental problem is that the IMF is too responsive to the national interests of its principal shareholders, i.e. there is excessive "politicisation". This is something that the new Managing Director, Horst Köhler, himself has expressed a desire to avoid although he has made no concrete suggestions how to do so. De Gregorio et al, are concerned that measures which would increase the control of the Board over the IMF management would lead to a more powerful and therefore more politicised decision making process. Their solution is to make the Board independent of nation states. Whilst EDs would still be elected by governments, they would be prohibited from taking advice from them, "this would prevent Executive Board decisions from being influenced excessively by the parochial interests of national governments." p90.
In addition, they also propose that EDs should serve a maximum term of 5 years in office; EDs should be barred from moving into government or finance related jobs for a specified period; and the Board should have budgetary independence. The EB would remain accountable to the IMFC and would work to an explicit mandate.
De Gregorio et al are correct to argue that changing the structure of the Board will have little impact if the G7 governments actually exercise their influence through "informal" channels. Their solution might best address this problem although it is likely to create others and doesn't necessarily address problems such as the unequal representation of developing countries vis a vis developed countries. Whilst it is true that some governments are able to exercise excessive influence, on the whole most are not. Moreover, unless the capacity of EDs was increased then increasing the independence of the Board may just give the IMF management and staff more influence.
3b. Independent Managing Director
"In all of these [international] institutions, the UK government favours open and competitive processes for the selection of top management. This could include a definition of the competencies for the post, selection and search committees and a clear process for taking the final decision, in which competence would be put above considerations of natonality." DFID, 2000.
Another means for improving the independence of the IMF from its principle shareholders would be to do away with the arcane procedure which allows the European governments to pick the Managing Director (MD). It has always been the case that the Europeans (with the agreement of the US) have picked the head of the IMF, who has always been a European, and the US picks the head of the World Bank, who has always been an American. Given the lack of representation of poorer countries on the Executive Board and that much of the IMF's work is carried out in developing countries, it is particularly critical to ensure that high-level managers represent the interests of users of IMF resources equally with the interests of the major shareholders.
The undemocratic nature of this process was the centre of debate when the current MD was appointed following a drawn out process, which involved the rejection of the Europeans' first choice by the US. This raised questions about the suitability of candidates and the problem of selecting them from such a small pool.
More than 120 NGOs from 35 countries wrote to the IMF's Board of Governors calling for a transparent process, which should include the following elements:
After the public outrage about the secrecy and undemocratic nature of the process, a committee of EDs was established to investigate alternative, more democratic processes. The committee was due to report during the IMF's annual meeting in Prague 2000, however it did not. Few details about its proposals have emerged but it appears that the committee is advocating little change. It seems likely that it will simply propose that whilst the Europeans should continue to pick the candidates, there should be a more democratic and open process for choosing between them. This is not sufficient. It is unacceptable that candidates cannot be drawn from other regions
3c. Access to alternative information
Japanese Finance Minister Myazawa has suggested that governments should be represented at the Board when their programmes are being discussed. This would allow Board members to ask questions and receive answers directly and would increase the sense of ownership felt by the government. It would also enable the country to monitor the discussion (De Gregorio et al 2000).
Whilst in principle greater representation should be welcomed, by the time the programme has reached the Board level it is very unlikely to be revised and a borrowing government is very unlikely to raise any objections in case it jeopardises the Board's willingness to agree the programme and thus lending.
Direct contact with country representatives should not be limited to the government. It would be helpful for the Board to get a fuller picture of a country's circumstances if it heard directly from civil society, private sector and academic experts too.
De Gregorio et al suggests that the Board should convene panels of Experts to offer independent analyses of pressing policy issues. This could be extremely helpful but it will be important to ensure that participants are drawn from a variety of backgrounds.
More generally, the Board should assist outside experts to informally provide alternative analyses by making timetables of board meetings public (see below).
3d. Improving Capacity to Monitor
The Board's lack of capacity means that it often simply "rubber stamps" decisions. The staff generally acknowledge that a programme is never sent to the Board, which members could not sign. All disagreements about programme content are settled between the staff and are not taken to the Board. This is problematic because it means that the Board is not party to the debates and is not subject to alternative points of view. Several suggestions have been proposed to help increase the Board's monitoring and decision making capacity.
Working committees are a well established mechanism in many national parliamentary systems for monitoring and holding accountable the government and its civil servants. The ODC Task Force recommends that the IMF Board should create committees or "sub-boards" to increase the Board's capacity to monitor members (and the staff) more effectively i.e. groups of directors would be responsible for overseeing specific groups of countries, thereby reducing the overall burden. The idea is based on the OECD's peer review system. To further increase capacity, the sub-boards could include alternate EDs. Periodically, the sub-boards would report to the full Board.
The ODC report suggests that sub-boards could monitor Article 4 processes and adherence to international standards. This remit could be extended to include monitoring country programmes and compliance with conditionality. By reducing the number of countries each ED is expected to take decisions on, it would allow them to build up more knowledge about each case and reduce their reliance on the IMF staff.
The IMF Board's recent experience with forming a committee of Executive Directors to oversee the process of developing an independent evaluation unit appears to have worked well. In particular, it was a key point of contact for outside stakeholders, which helped to facilitate a more effective and sustained engagement on the issue.
4. Transparency of the Board
Transparency of the Board's operations is important because it allows external stakeholders to see what issues were discussed and who was involved, i.e., transparency will help diminish the opportunity for influential members to dominate the Board. It will also reveal the quality of discussions and whether all relevant interests are being taken into account. This is important because some have argued against voting reforms on the basis that it will diminish the usefulness of the Board still further by forcing decisions to be taken in "smoke filled back rooms".
4a. Voting procedures
"An obvious way to enhance IMF accountability is by requiring more decisions to be taken by formal up-or-down votes and releasing the results. Directors' ultimate constituencies will then be in a position to judge whether or not their representatives supported or resisted a particular Fund policy." De Gregorio et al, 2000 p85.
It is rare for the Board to make a formal vote. Instead decisions are taken by consensus; i.e. voting occurs informally with large vote holders able to exert considerable influence over discussions. This informal procedure has limited the effectiveness of efforts to try to make individual EDs more accountable to citizens in their home countries. For example, the US ED is required to formally publicise how she votes, however, since she only rarely formally votes this has not aided transparency.
Countries should be required to reveal their positions taken during Board discussions and/or voting rules should be changed to require a formal casting of votes (Meltzer Commission).
The benefits of a formal voting process is that it allows vote splitting , i.e., instead of constituencies voting as a block each individual country would cast a vote. This is necessary if votes are to be made public. Publishing how each country voted as well as the overall outcome of the vote is essential.
4b. Publishing board agendas and minutes
"The IMF must be more open about its proceedings and give a greater voice to developing countries when discussing issues of direct relevance to their economic well-being. Hence we believe there is a case for publishing both the agenda and minutes of the executive board meetings." (Treasury Committee Third Report, 1999-2000 session)
The Board has recently started to publish a work programme on the IMF website. Whilst this is a step in the right direction the programme is in fact so devoid of detail it makes it practically worthless. Not only are key debates not mentioned, there is no timetable for when issues will be discussed or decisions taken.
Whilst it must be accepted that the management will be influential in setting the Board's agenda because of their intimacy with the issues and Fund's work, it is essential that agendas are published in advance of meetings. The benefits of doing so are that it would:
In addition, the Board should publish timetables of when it will be discussing country programmes, in order that interested stakeholders can make written submissions to the Board in advance.
In the case of policy discussions, the Chair's (which is usually the Managing Director) summing up of Board meetings is also published. Whilst this is a welcome step forward it only indicates what the final agreement was, it gives little flavour of the debate or why some countries objected. The Meltzer Commission and Gregorio et al, recommend publication of the summaries of all Board meetings, after a time lag.
Making the minutes of IMF Board meetings public would help to clarify the positions of the IMF Board and the Management. Whilst they remain unpublished, citizens of member countries cannot find out what was said on their behalf. Objections may be raised on the grounds that discussions would be less frank, however, the UK government's experience with publishing the minutes of the Monetary Committee's meetings is that the benefits of transparency outweigh the costs. The minutes should specify which Executive Directors have expressed support for the proposed decision under discussion, and which have explicitly withheld support or proposed amendments.
An immediate action would be to allow publication of ED statements to the Board. Whilst this could result in rather boring and bland statements, if it did so it might help to make it clearer where important decisions were being made.
Transparency and accountability of imf operations
"This situation of decision makers having power with accountability to people who do not have to live with the consequences of their decisions but without accountability to those most affected by their decisions is a situation ripe with potential for abuse." (Bradlow, 2000 p7)
Reacting to the overwhelming criticism following the crisis in South East Asia in 1997 the IMF has taken a number of steps to increase transparency of its operations including making more programme documents available and releasing public information notices. These small steps are welcome but larger ones still need to be taken.
In particular, more needs to be done to integrate external stakeholders - parliaments, civil society organisations and the private sector - into the decision making processes. As Woods (2000a) outlines, there are different levels of engagement, from consultation to active participation in decision making. Whilst it may not be for the IMF to dictate how involved stakeholders outside the government should be, staff should be helping to facilitate an inclusive process. For example, by ensuring resident representatives provide documents in time for NGOs to prepare themselves for meetings. The ad hoc nature in which stakeholders are currently involved suggests a need for clear rules of engagement. Moreover, given that the IMF often has a greater role in national decision making than parliaments or civil society, its imperative that processes are established that allow national stakeholders to hold the IMF accountable.
5. Improving Evaluation and Disclosure of Reports
In the light of recommendations made by the G22 Working Group on Transparency and Accountability the IMF has recently improved its disclosure policy. The following documents are now routinely made available, including :
5a. Disclosing reports
Whilst these steps are very welcome many internal staff documents are still not released, often on the basis that they contain sensitive information. Developing countries are particularly sensitive to this. Whilst some of the broader documents, such as the PRSP are routinely released some key documents, such as the policy matrix which identifies the conditions agreed between the government and the IMF may be more easily withheld in the future than they were previously. The policy matrix used to be published as part of the Policy Framework Programme (which has now been replaced by the PRSP) but it is now up to the government to disclose this if it so wishes.
The policy matrix should be in the public domain because it allows stakeholders to see what actions the government has agreed to take with the IMF. This is important in the context of the PRSP which strongly emphasises ownership. It is imperative, that interested stakeholders should be able to monitor that IMF programmes reflect the objectives of PRSPs.
To assist ownership, there is also a case for making programme documents, such as the Letter of Intent and policy matrix, available to national stakeholders before they are discussed by the Executive Board so that stakeholders can comment on them and express their views to their representatives. This would help improve the flow of information to the Executive Board who would then have a much more informed picture, at an earlier stage, on which to base their decisions.
At the present, all documents are only provided in English, which limits the number of people who can understand them. If the IMF is serious about increasing its transparency and enhancing participation (see section 6), then as a minimum it should release country documents in the majority language. All policy documents and evaluation reports should be released in French and Spanish as well as English.
A further step should be to minimise the jargon in the documents. At the moment, they are so technical that most people, even economists, can have trouble understanding the terms. IMF staff are currently looking into these issues and will report to the Board in the near future.
Many generic reports, which, for example, examine policies or issues more broadly, also remain unpublished because not all Board members agree with their substance. Current guidelines stipulate that the decision to published will be "based on whether the discussion has reached completion or the point where informing the public is deemed useful."
This raises interesting questions about what is the purpose of staff policy and issues papers and who has ownership of them. Should reports simply express a unified Board view or should staff be encouraged to publish reports, not attributable to the Board, which reflect a broader set of views? Given the wide impacts of IMF programmes it does not seem justifiable that the IMF should just reflect the views of a small number of finance ministers and central bankers. The IMF staff should be encouraged to reflect a wider set of views in its reports (indicating where viewpoints are attributable to the Board) and all reports should at least be published on the IMF's website.
5b. Improving internal evaluation
Internal evaluation can be an important learning process for staff. However, in order for this to be effective the staff need to ask the right questions and to implement the lessons learned. Whilst the staff may actually be good at this, there is little opportunity to judge them because rarely are internal reports made public. Unfortunately, when they have been, they have often appeared to be biased, have failed to address the most pressing issues identified by civil society and have been of poor quality.
Because internal evaluations are not made public it is hard to know what the staff do actually evaluate. For example, the staff argues that there are implicit evaluations of completed programmes as the next programme is negotiated, however, they do not produce "completion reports" which evaluate whether conditions were implemented or whether objectives were achieved. Thus, it is impossible to tell if there is systematic monitoring of whether or not objectives were in fact achieved. This needs to be rectified. Staff should be required to produce and publish completion reports on all completed programmes (and those that break-down).
Nor are ad hoc internal evaluations necessarily made public after Board discussions, although the rules have recently changed and they must now be disclosed after 5 years. Whilst this is an improvement on the former policy, the delay means that external stakeholders are unable to judge whether lessons have been effectively absorbed until often it is too late. For example, the 1995 evaluation of the IMF's handling of the crisis in Mexico might have yielded lessons for how to deal better with the crisis in Asia if those outside the IMF had access to the report. To facilitate learning and contribute to wider debate, all internal evaluations should be automatically disclosed upon completion.
5c. External evaluation
The Board has recently taken the decision to improve accountability to the general public by establishing an independent evaluation office (EVO). After pressure from critics it has now been agreed that:
Based on these terms, the EVO could be a useful means of accountability. This could further be enhanced if the transparency and openness to external stakeholders is improved by:
5d. Establish an ombudsman
"The IMF advocates that good governance requires that all decision makers in the policy making process should be held accountable for their actions and decisions. To date, however, it has not established any mechanism through which the citizens of its consuming countries can hold the IMF or its management accountable for their actions as decision makers in this process." (Bradlow, 2000)
The IMF cannot be held accountable for its policy advice at the national level. That is, it can not be tried in a national court of law. However the IMF has no procedure of its own to allow citizens to lodge complaints. Thus, whilst the IMF argues that policy makers at the national level should be held accountable for their actions there is not a similar requirement of its own staff. However, IMF staff are tremendously influential in the policy setting process in borrowing countries as Botchwey et al indicate, "At the root of all the concerns they [governments] reflect is a feeling of a loss of control over the setting of the policy agenda in reform programs." Botchwey 1998, p36.
The "loss of control" suggests that citizens cannot hold their governments accountable, in which case they should be able to hold the IMF accountable instead. There is a danger that the emphasis on "ownership" in the PRSP will give the IMF an argument for resisting this. However, it is widely perceived that in reality government "ownership" will be marginal because the IMF will continue to provide advice to them. More importantly, the IMF is still required to approve national strategies , thus the IMF will remain very influential.
This suggests that the IMF should set up on ombudsman to receive and investigate complaints from affected stakeholders. The World Bank established an Inspection Panel in 1994 and the International Financial Corporation established a similar complaints procedure in 1999. Whilst these are by no means ideal they demonstrate that there is a precedent for doing so.
Bradlow (2000b), suggests the following key features for an ombudsman:
It could be argued that because the IMF does not have operational directives an ombudsman is impractical. This is not an argument against an ombudsman but for developing operational directives. In the absence of operational directives, the staff and management have great discretion in carrying out their work. Rectifying this should be a priority. In the mean time, Bradlow suggests that the ombudsman could base his/her investigations on the Board's decisions, which seek to operationalise the IMF's mandate and are intended to guide the staff and management in their operations. As well as using other applicable staff rules, guidelines and policy statements.
In addition to developing operational directives, a more immediate action would be to initiate social impact analyses of programmes before they are implemented (ex ante social impact analyses). These should be published. Ex ante impact analyses could provide a useful benchmark in case of complaints because they would detail what the anticipated impacts were and what actions should have been taken to mitigate them. Where such actions failed or where social impacts were larger than anticipated, there may be cause for complaint.
6. Accountability to a Wider Range of Stakeholders
"Just as the institution need[s] 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." Lawrence Summers, December 14th 1999
New IMF roles mean that it will be able to reach much deeper into countries' domestic affairs than it previously could do. It now has the remit to correct weaknesses in domestic financial systems and to achieve poverty reduction and governance goals. Yet the IMF's articles of agreement only require that it consults with finance ministers and central bank governors. This may have been appropriate when the IMF's objective was simply macroeconomic stability and its conditionality was very narrow but it is no longer so. Today's broader policy reforms have a much wider social impact than previously, which suggests that the IMF should consult far more widely than it previously has done and be accountable to a wider range of stakeholders.
The UK government has committed itself to "work with civil society to strengthen the capacity of poor people to hold governments and international institutions to account for progress on poverty reduction." DFID, 2000. The need for more inclusion is recognised in the Poverty Reduction Strategy process, which for the first time stipulates that participation should be a requirement, although how participation should be carried out and who should be involved is not defined. This could be a significant step forward, however, the IMF appears to regard efforts to include civil society, the private sector and parliaments in the process as simply a job for governments and it has not established mechanisms to support their efforts.
The lack of institutional commitment to participation is reflected in ad hoc, poor quality consultations. For example Botchwey et al remark that, "What is interesting though is that the Fund continues to be berated for its secrecy although Fund missions have, as the country studies show, for some time now been dealing with a broader cross section of government officials and departments. When asked to explain this apparent paradox, many interviewees argued that meetings with people outside the core ministries an allied agencies were seldom truly consultative since only the mission leader spoke - and even he usually did no more than seek view points - while other mission leaders simply took notes." (Botchwey et al, 1998, p36).
Bradlow (2000) proposes, as a minimum, the IMF should consult with all government departments which are affected by specific actions that it is advocating; with legislators where changes to laws are required; and with civil society stakeholders who have the ability to determine the success or failure of the policies.
6a. Rules of engagement
"The IMF's failure to establish formal consultation procedures and to engage in meaningful consultations contradicts the principles of participation and the need for transparent governance procedures that it advocates to its member states." Bradlow, 2000a p9.
Consultation and participation is very new for the IMF thus it is essential to provide guidelines to staff to assist them as they internalise the new approach. Rules of engagement will clarify for civil society representatives what the IMF's role is in participation and what they can expect to receive from the institution. Also, it is important that the IMF is able to demonstrate that it has considered civil society's points of view in the policy making process. Some NGOs have called for the IMF to detail in its programme documents the various views of different stakeholders and to justify why it has chosen a particular policy. The issue is that civil society must be assured that they are involved in a effective process to ensure they have ownership.
Whilst participation is accepted in the PRSP process, this only applies to low-income countries. Thus it is important to establish that participation is essential in all IMF programmes. The most effective way to do so would be to develop an operational directive laying out actions staff must take.
Whilst it would be inappropriate to define a comprehensive participatory mechanism which would include actions to be taken by the government , the operational directive should detail what actions IMF staff will be required to take, such as:
6b. Systematic engagement
The IMF argues that in a crisis situation it is not possible to consult widely because decisions must be taken quickly. This suggests that the IMF should establish mechanisms that allow it to engage with civil society on a systematic and sustained basis so that they are aware of civil society issues and concerns in advance of any crisis and are able to take these viewpoints into consideration when advising governments.
Edwards (2000), proposes that the IMF should hold a World Financial Forum every five years, in which the IMF could engage civil society and other actors in broad debate about financial policy, and to assess the evidence and draw out the lessons from particular policy interventions. Such a mechanism could prove very useful, but it won't help to institutionalise consultation with the IMF at the national level, thus national level mechanisms should also be developed.
Non of these reforms will be sufficient by themselves to ensure effective governance. Measures to increase the voice of developing countries, particularly reforms to constituencies, without increasing their vote may not improve the willingness of developing countries to raise critical issues and therefore to raise the quality of the debate. This is because they are unlikely to risk jeopardising support for their own lending programmes. Efforts to increase voice and vote must be taken together as a package, and indeed its unlikely that they wouldn't be. But they are likely to take considerable time.
In the meantime ways must be found to increase the Board's access to information both to increase Executive Director's awareness of wider concerns and to enhance its ability to function as an accountability mechanism. Establishing expert panels will be helpful but there should also be clear channels for all external stakeholders to submit comments to the Board on a continuous basis. This need not be a particularly formalised process but it will be important to ensure that the Board facilitates the flow of information by giving advance notice of agendas and the focus of debates.
To increase accountability of the Board to the general public, as well as encouraging an improved flow of information, the Board should release minutes from Board meetings, take votes and publish the results, and release Board statements.
It is also essential to improve governance of the staff and management. The Board's access to information is limited and so too is its capacity to effectively fulfil this role. This is particularly the case today because the IMF has expanded its involvement in member countries' beyond its traditional remit.
Improved oversight of staff activities could be achieved by increasing the transparency of IMF operations and external stakeholders participation in them. For example, by making staff reviews and internal evaluation documents public. It will also be necessary to enhance evaluation processes by establishing an independent evaluation unit which is broad in its remit and independent from staff and management.
Given the Fund's significant influence in the policy framing process, the IMF must be answerable to external stakeholders directly. The most effective way to do so is to include them in processes for identifying and setting policies. Early involvement will be an important preventive mechanism. However, where there are negative unanticipated impacts on particular sections of society, the affected people should be able to hold the IMF accountable. This suggests the need for an ombudsman function.
References and resources
Kewsi Botchwey et al, 1998, Report of the Group of Independent Persons Appointed to Conduct an Evaluation of Certain Aspects of the Enhanced Structural Adjustment Facility, IMF: www.imf.org/external/pubs/ft/extev/esaf1.pdf
Danny Bradlow, 2000a, Rapidly Changing Functions and Slowly Evolving Structures: The Troubling Case of the IMF, presented at the Annual Meeting of the American Society of International Law, Washington DC, April 7, 2000.
Danny Bradlow, 2000b, A Proposal for an IMF Ombudsman, unpublished. firstname.lastname@example.org
Mike Edwards, 2000, NGO Rights and Responsibilities, a New Deal for Global Governance, Foreign Policy Centre and NCVO. Tel: +44-(0)20-8986-5488.
Jose De Gregorio et al, 2000, An Independent and Accountable IMF, Geneva Reports on the World Economy 1, International Center for Monetary and Bank Studies and Council for Eco-nomic Policy Research. www.cepr.org
Department for International Development (DFID), 2000, Eliminating World Poverty: Making Globalisation Work for the Poor, White Paper on International Development. www.dfid.gov.uk
Gerry Helleiner, 2000, Global Financial Governance, in Global Financial Challenges, Towards a Responsible Financial Architecture for the Developing Countries, Norwegian Forum for Environ-ment and Development. email@example.com; www.forumfor.no
Raghbendra Jha and Mridul Saggar, 2000, Towards a More Rational IMF Quota Structure: Suggestions for the Creation of a New International Financial Architecture, Development and Change, Vol.31 No.3 pp579-604.
Allan Meltzer, 2000, "Meltzer Report", paper prepared by the International Financial Institution Advisory Commission phantom-x.gsia.cmu.edu/IFIAC/Report
Aziz Ali Mohammed, 2000, The Future Role of the IMF: A Developing Country Point of View, in Reforming the International Financial System Crisis Prevention and Response, edited by Jan Joost Teunissen, Fondad. firstname.lastname@example.org; www.fondad.org
ODC Task Force Report, 2000, The Future Role of the IMF in Development, Overseas Development Council, Washington DC. www.odc.org/whatsnew/index
Oxfam, 2000, Reforming the IMF, Oxfam International. email@example.com
Andrew Simms et al, 2000, Its Democracy Stupid, The Trouble with the Global Economy - The United Nation's Lost Role and Democratic Reform of the IMF, World Bank and World Trade Organisation. www.neweconomics.org
Lawrence Summers, The Right Kind of IMF for A Stable Global Financial System, 14th 1999.
Treasury Committee, 2000, The International Monetary Fund, Report, Proceedings of the Committee, Minutes of Evidence and Appendices, Third Report, 1999-2000 session.
Ngaire Woods, 2001, Making the IMF and World Bank more Accountable, International Affairs, January. users.ox.ac.uk/~ntwoods
Ngaire Woods, 2000a, Rethinking Accountability and the Reform of the International Financial Institutions, presented at 'Reforming the Architecture of Global Economic Institutions' conference, Bank of England, London Friday 5 May & Saturday 6 May, 2000 (ESRC Global Economic Institutions Programme). users.ox.ac.uk/~ntwoods
Ngaire Woods, 2000b Global Institutions and Developing Countries, Discussion Notes for Presentation, Overseas Development Administration, 8/6/2000, unpublished.
Ngaire Woods, 1997, Governance in International Organizations: The Case for Reform in the Bretton Woods Institutions, International Monetary and Financial Issues for the 1990s, Volume XI (Geneva, UNCTAD, 1998).
1 This does not include concessional lending.
2 Sometimes the seat "revolves" amongst the constituency members.
3 Developing countries refers to all middle- and low-income countries.
4 The IMf pays the salaries of delegation staff not member countries.
5 Whilst 2 boards might better reflect the reality of the IMF's roles, it does indicate that the IMF is doing two very different jobs which is probably not optimal. This raises questions, as other critics are doing, about whether the IMF should be involved in long-term programmes in the poorest countries.
6 The "Cooper" report is basically concerned with looking at how to adapt the formula so that it better matches how quotas are currently distributed. It does not try to examine how the system could be improved to be more equitable, ie give more voting power to developing countries.
7 A reallocation of the quota will be most costly to European countries because they are already too heavily represented (according to the US), although the US is obviously the most influential in practice because the Europeans rarely combine their votes.
8 Quota Formula Review Group, Report to the IMF Executive Board, April 2000 p4.
9 Quota Formula Review Group, Report to the IMF Executive Board, April 2000 p26.
10 The voting rules would have to change to let each country vote, at the moment the ED votes for the whole constituency, see section 4a below.
11 This may require a change to the Articles of Agreement.
12 Some are automatically published whilst others are published at the country's discretion, but with a presumption towards publication. For details see Pubic Information Number 00/81
13 IMF Public Information Notice No. 00/81
14 In the case of Poverty Reduction Strategies, the IMF's remit has actually increased because it is now required to approve an entire poverty reduction strategy not just the macroeconomic element of it.
15 For more discussion see Woods 2001.
16 Woods 2001 warns of the danger that a "new "Washington Consensus" on the politics of participation and consultation will be forged…"
This text may be freely used providing the source is credited.
Published: 19 January 2001 , last edited: 27 May 2010
Viewings since posted: 7004
Climate Investment Funds Monitor 7: April 2013 25 April 2013
Working paper: The private sector and climate change adaptation: International Finance Corporation investments under the Pilot Program for Climate Resilience 24 April 2013
The UK's role in the World Bank and IMF: Department for International Development and HM Treasury 13 March 2013
The World Bank and industrial policy: Hands off or hands on? 6 December 2012
Climate Investment Funds Monitor 6: October 2012 26 October 2012