Indispensable or unworkable? The IMF’s new approach to conditionality

28 January 2003 | At Issue

by Vivien Collingwood

In September 2002 the IMF‘s Executive Board finally approved a new set of guidelines on conditionality. These set out how its structural conditionality can be made more effective, with provisions on improving the clarity and focus of conditions, increasing recipient-country ‘ownership’ of reforms, and coordinating IMF conditionality with that of other organisations.

While any review of IMF conditionality is welcome, the new guidelines fail to address the deeper problems and flawed assumptions plaguing conditionality. Questions of legitimacy have been sidelined in the drive for efficiency. Effectiveness and legitimacy cannot be separated, however, and if IMF conditionality is perceived by recipient-country governments and peoples to be illegitimate, then it will not be successfully implemented.

By May 2000, when the streamlining process was first announced, IMF conditionality had expanded significantly, and had become a target of criticism from external analysts and staff members. IMF conditionality was accused of being intrusive, confusing, and occasionally inappropriate. The IMF‘s 2001 Annual Report stated that streamlining aimed to make ‘conditionality more efficient, effective, and focused.’ The new guidelines focus on three core areas:

1. Simplifying conditionality

Conditionality should only be applied to those policy areas deemed to be critical to achieving the goals of Fund-supported programmes. This means fiscal, financial, and exchange rate policies, and in some cases privatisation, governance, and public sector reform. Conditionality attached to non-critical policies should be avoided. The IMF intends to achieve greater clarity and consistency by explaining the different types of conditionality, and their purposes, in loan agreements.

2. Increasing recipient-country ownership

IMF conditions must be used in ways that enhance, rather than undermine, ownership. Recipient countries must initiate reforms and IMF staff should be flexible in its approach and tailor reforms to suit local conditions and capacities.

3. Improving IMF-World Bank relations

The IMF should coordinate its conditionality with that of other organisations. In Bank-IMF relations, the organisation with expertise in a particular area should take the lead (the ‘lead agency’ principle), and its assessments should be included in the other organisation’s reports. In addition, conditionality should be devised within a ‘common framework’, such as the PRSP, so as to avoid policy conflict and increase country ownership.

It is difficult to judge the initial impact of the new guidelines, particularly given that the evidence varies significantly from case to case. Despite this, the IMF has claimed a number of positive changes:

  • The average number of conditions has been reduced overall, and the IMF claims there is better differentiation between ‘critical’ and ‘non-critical’ conditions.
  • There has been better prioritisation and a decrease in the number of conditions on low-income country loans. The average number of conditions on 27 new PRGF programmes has been reduced by one-third, and two-thirds of all PRGF conditions focus on core areas.
  • There is a clearer division of labour with the Bank, with the IMF withdrawing in some areas to let the Bank take the lead.

However there is plenty to question in the IMF‘s current approach:

No alternative model of conditionality

Despite the new rhetoric of ‘clarity’ and ‘parsimony’, changes brought in by the guidelines are extremely limited. Whether conditionality should be abandoned was never in question-the IMF considers conditionality to be ‘indispensable’-but that said, this was a missed opportunity to explore alternative kinds of conditionality. The review merely clarifies the status quo while reiterating the IMF‘s commitment to the principle of ownership. Conditionality on general goals of IMF programmes rather than the means to achieve them would have deserved better consideration for a start.

Few limits on conditionality

The guidelines are imprecise and give the IMF considerable scope in interpreting what ‘critical’ means. There are no limits on the number of conditions that can be attached to loans, and nothing prevents the IMF from using large numbers of structural conditions outside its ‘core’ areas. Given that the IMF is cautious about granting recipient countries a ‘long leash’, it is likely that the IMF will find additional ways to try to control recipient-country actions, such as increased monitoring and surveillance.

Isolated review

The IMF failed to link the review sufficiently to other issues, such as PRSPs or the IMF‘s use of governance conditionality. This risks limiting the positive effects of streamlining.

Effectiveness takes precedence over ownership

The IMF is more concerned with questions of effectiveness and the preservation of Fund resources (that is, loan repayment) than with improving genuine ownership and participation. For this reason, the Fund is likely to err on the side of caution and opt for more conditionality when doubtful about a government’s commitment to reform. The Fund assumes that its current policies are appropriate: the review does not cover macroeconomic policy content, despite widespread failure to promote growth and reduce poverty in recipient countries. It aims at making recipients internalise and accept IMF policies, rather than giving them flexibility to design their own. This is particularly the case with the continuing use of prior actions, which must be implemented before a country receives assistance.

Need for more capacity building and empowerment

Ownership does not just mean recipient-government accordance with Fund objectives; it also means having the capacity to engage a range of actors, governmental and non-governmental, in forming and implementing policies. Many recipient countries cannot compete with the Fund’s knowledge base and resources, and lack the technical knowledge to propose alternative policies. Until capacity-related issues are addressed, recipient-countries will find it difficult to assume the initiative for reform.

Thinner Fund, fatter Bank?

There is a danger that IMF conditions will simply be moved to Bank agreements (which have not been streamlined). If this happens, the overall impact of conditionality is likely to stay the same. Increased Bank-Fund cooperation could result in less policy leverage for recipient-countries. The ‘lead agency’ and ‘common framework’ principles bind Bank and Fund prescriptions closer together. While this might reduce confusion, it also increases the pressure on developing countries to comply and reduces scope for policy choice.

The IMF is unlikely to achieve its objectives unless deeper problems concerning the content of IMF policies and the nature of the IMF‘s relations with recipient countries-whether these are relations of domination or empowerment-are addressed. The long-term aim must be for conditionality to be replaced with a form of financing that gives priority to local needs and greater emphasis on democratic decision-making. The challenge is to establish how orthodox IFI conditionality might be minimised, and with regard to this, the new IMF guidelines do not go far enough.

Vivien Collingwood is at Nuffield College, Oxford University, and is writing her doctoral thesis on governance conditionality. Her article on the ethics of conditionality, ‘Assistance with strings attached’, will be published in the Spring 2003 issue of Ethics & International Affairs.