Many observers were sceptical when a few years ago the IMF embarked on a mission to respond to the "cries of the poor", in the words of its then head Michel Camdessus. Little has been delivered since then on what was supposed to be a new, more flexible approach and the time has come for an in-depth reconsideration of the Fund's role in low-income countries.

The turning point in the IMF's commitment to contribute to poverty reduction was a 1999 statement by Michel Camdessus that member countries had endorsed "a clear mandate for the Fund to integrate the objectives of poverty reduction and growth more fully into its operations." The IMF has however delivered very little. It is currently trying to address deep contradictions and tensions over its role in low-income countries and is struggling to bridge the gap between rhetoric and implementation. An internal review is underway, based on three papers. NGOs have started to weigh in, in particular Oxfam with a paper calling for the IMF to not stand in the way of, and in fact actively contribute to achieving the Millennium Development Goals.

Several key entangled issues need to be resolved when trying to understand what role the IMF should play, if any, in poverty reduction. Does the IMF have the mandate and legitimacy to be an actor in poverty reduction? Does the IMF have the financial and human capacity to effectively understand and act on poverty? If not, should this be strengthened, or should poverty reduction be left to other institutions? What should be the specific role of the IMF in low-income countries? Short term assistance to face temporary problems and shocks? Longer term involvement to help resolve deeper problems and address broader, sometimes systemic issues?

A casting mistake?

While the IMF was clearly not originally designed as a development institution, it could be argued that the first of its Articles of Agreement defining its purposes can be interpreted to include this. It mentions contributing to "the promotion and maintenance of high levels of employment and real income and to the development of the productive resources of all members as primary objectives of economic policy".

However long-time critics of the IMF were defiant when it announced that it would now officially aim to actively contribute to poverty reduction and would adapt some of its instruments accordingly. Many of them felt the IMF had no legitimacy, credibility and capacity to do so and considered the pronouncement ironic, pointing to the devastating impact of 20 years of adjustment orchestrated by the Fund. An important factor in the deficit of credibility of the Fund is the fact that its human capacity is limited. The Fund recently doubled the number of its social advisors. However this only brings the total number of poverty specialists to two, out of 2,600 staff members.

Is the IMF's toolbox adequate?

Beyond the Fund's usual surveillance activities and a technical assistance that increasingly focuses on low-income countries, IMF lending instruments have also evolved over time (see Box). The major change accompanying pronouncements on poverty reduction in 1999 was the transformation of the Enhanced Structural Adjustment Facility (ESAF) into a new Poverty Reduction and Growth Facility (PRGF). The change of name was intended to signal that IMF lending for poor countries would now be embedded in a broader development agenda: Poverty Reduction Strategies. Some of the expected "key features" of PRGF are: pro-poor budgets, more flexibility on fiscal adjustment, more selective conditionality and social impact analysis.

While the IMF claims some progress on making budgets more pro-poor, it is clear that it has failed to deliver on many aspects. For example while overall the IMF has reduced the number of conditions attached to its programmes in a 'streamlining' process, disparities across countries and takeover by the World Bank make that change meaningless for some of them. The European Network on Development and Debt (Eurodad), using a matrix to monitor trends across countries, has taken the Fund to task for failing to deliver on the key features of the PRGF. Eurodad suggests that beyond the shortcomings in the implementation of the PRGF, some of the assumptions and building blocks also need to be changed if the IMF is to contribute actively to the MDGs.

The missing link

The Fund itself is concerned with the fact that the PRGF programmes -especially the macroeconomic frameworks- are disconnected from supposedly country-owned, participatory PRSPs. The IMF has therefore proposed to better align national budgets, PRGF programmes and PRSPs. This has raised concerns among observers that the aim is to make PRSPs more 'realistic' - while in practice many civil society groups were already concerned that the macro framework in PRSPs was not flexible enough to allow countries to reach social targets.

Instrument	Purpose	Terms
Poverty Reduction and Growth Facility (PRGF)	Concessional facility with greater emphasis on pro-poor growth	Longer repayment periods, lower interest rate (0.5%) than standard facilities
Emergency assistance for natural disasters and post-conflict	Similar to, and sometimes combined with, standard arrangements, but with fewer conditions attached.	Limited volume. Expensive and therefore rarely used (33 countries have used it since 1962). Expanded to post-conflict countries in 1995
Compensatory Financing Facility (CFF)	Designed to address commodity price shocks	Expensive and therefore gradually supplanted by Enhanced Structural Adjustment Facilities (ESAF) and now PRGF - has not been used since 2000

A related problem is a tendency towards 'aid pessimism' at the IMF, which can become a self-fulfilling prophecy: assuming that aid might not be forthcoming leads to frameworks designed to accommodate only limited amounts of aid. In effect this has led to controversy over whether the IMF is actually blocking aid to countries, by fear of a so-called Dutch disease effect - the inflationary pressure that could accompany massive inflows. For example in the past this logic has led the government of Uganda to oppose increased donor support to the health sector.

Also disputed is whether the Fund accepts more flexibility when negotiating fiscal adjustment targets. A study by the IMF's evaluation office argues that there is no one-size-fits-all approach on this and that austerity is not systematic. However research by Oxfam and Eurodad argues that IMF fiscal inflexibility is still the rule in low-income countries, leading for example to inflation targets below 5 per cent without clear justification.

This gap between what the IMF has committed to (help achieve the MDGs) and the straightjacket that it still imposes on most countries reflects a problem with the Fund's analysis of problems faced by low-income countries. Most of the time the Fund still sees the problem in terms of stabilizing countries and getting their macro policies right to create the right environment for pro-poor growth. However the concept of stabilization in these countries is debatable. Some countries in Africa for example have been stabilized in stagnation and/or a dependence on factors that they have no control on, such as the price of raw commodities. The analysis of the problem determines the solutions envisaged but also the modalities of the Fund's interventions, including financial conditions.

Is IMF lending to poor countries unsustainable?

Beyond the question of delivering on the rhetoric, an important factor in the debate is the level of resources available for IMF involvement in low-income countries, and the related question of the terms of the PRGF, ie the interest rate and repayment conditions.

Undoubtedly most low-income country governments are keen to have access to various kinds of concessional finance because they have limited or no access to private capital. However, following recent debates on whether assistance to poor countries should be delivered in the form of grants or loans, it is legitimate to ask if the terms of the PRGF are adequate. PRGF is more expensive than loans from the International Development Association (IDA) for example, and the repayment periods are shorter.

A further complication is that there is little room to make the terms of PRGF more suitable to the needs of low-income countries. Donors provided initial resources for the PRGF through voluntary contributions and it was expected that interest repayments would rapidly make the operation self-sustaining. But in practice resources available for this form of lending will be halved to just $870 million in 2005.

IMF as a gatekeeper: power and donor politics

One of the most debated aspects of the IMF's role in poor countries is the gatekeeper function. In other words, having an IMF programme in place in a country sends a signal that economic policies are sound and therefore acts as a green light for other donors. Conversely the suspension or cancellation (when a country gets 'off-track') of this programme often creates a snowballing effect, all donors suspending their aid. This is particularly problematic for heavily aid-dependent countries. It can also lead to situations where countries which do not need an IMF programme are keen to keep the Fund involved just for its signalling effect.

There seems to be some consensus both inside and outside the Fund that the institution should not have the implicit power to interrupt all types of foreign aid. Some donors arguably seem sometimes keen to retain this function, possibly because it allows them to deflect responsibility. However it has happened that donors officially voiced discontent about the Fund suddenly declaring a country off-track, as in the case of Bolivia in April 2002, over a failure of parliament to approve a tax law the IMF had made a condition.

The role of the Fund as a gatekeeper needs to be reduced. One evolution would see more shared power between donors, and the Fund get out of countries which do not need financing. It could perhaps provide some other form of signal (through surveillance possibly). A real alternative however would require the creation of regional monetary funds to counterbalance the influence of the IMF. Additionally other mechanisms, mostly based on genuine peer review, would also provide a workable alternative.

What role for the IMF? First do no harm

A question underlying the whole debate on the IMF in low-income countries, as often with the Fund and the World Bank, is: should the IMF do less (do no harm), or more (do good)? This implies but does not necessarily equate to: should the Fund have less power, or more power? In hindsight, it could be argued that some of the Fund critics, by blaming the IMF for not paying attention to the consequences of adjustment on poor people may have contributed to the new IMF emphasis on poverty, and therefore a form of mission creep.

There seems to be a broad consensus that the IMF is not a development institution. The staff and the Executive Board of the IMF have repeatedly stated this and said involvement in low-income countries should be limited to the Fund's areas of expertise. Efforts to increase coordination and limit overlap with the World Bank have been stepped up, as the division of labour has become increasingly unclear.

For reasons mentioned above, in particular capacity issues and problems with the prism through which the Fund analyses problems in low income countries, IMF lending is not adapted to allow countries to face long-term challenges partly induced by systemic issues such as the international trade regime. However low-income countries do face short term imbalances and are exposed to sudden shocks that may require quick disbursing lending facilities. Therefore there is a case for highly concessional finance to be channeled through an institution like the IMF for genuine short-term purposes - but any development work should remain in the remit of other multilateral agencies.

Efforts should therefore probably be concentrated on ensuring that the IMF's does not stand in the way of attempts to reach the MDGs, without placing high hopes on the active contribution it can make to development. Challenging core IMF prescriptions on inflation or deficit targets is particularly useful to give more space to governments to elaborate their own policy mix. But beyond this the battle of ideas also needs to force a fundamental rethink of broader systemic issues that keep low-income countries in a vicious circle of indebtedness, impoverishment and environmental destruction. The IMF is part of this system - therefore not the best interlocutor to discuss fundamental changes.