The Web edition of the IMF Survey is updated several times a week, and contains a wealth of articles about topical policy and economic issues in the news. Access the latest IMF research, read interviews, and listen to podcasts given by top IMF economists on important issues in the global economy.


The Web edition of the IMF Survey is updated several times a week, and contains a wealth of articles about topical policy and economic issues in the news. Access the latest IMF research, read interviews, and listen to podcasts given by top IMF economists on important issues in the global economy.

In 2000-02, the IMF conducted a major review of the conditions it attaches to its lending—its conditionality—and I replaced guidelines dating back as far as 1979. The new policy chiefly stems from concerns that there had been a large expansion of structural conditionality in the 1990s and that disappointing implementation suggested relatively weak ownership of IMF-supported programs by national authorities. New guidelines highlight the need for greater focus and streamlining, but has there been real progress? In many areas, yes, says Tessa van der Willigen (IMF Policy Development and Review Department). She summarizes here the findings of a recent IMF evaluation on the application of the new guidelines.

Conditionality remains a subject of controversy and debate. Some argue that conditions should be done away with altogether, but from the IMF’s point of view conditionality is not optional. The IMF must be sure that its resources are supporting policies that help countries resolve their balance of payments difficulties and allow them to repay loans so that these resources can, in turn, be used by other countries. Conditionality also clarifies the terms on which future installments of IMF loans will be available, thus giving countries the confidence to embark on programs that could not be sustained without such support.

Of course, that conditionality is here to stay does not mean that it is, or has been, perfect. The new policy is the culmination of a long process of internal and external discussion that led the IMF to embark on streamlining. These efforts actually began in 2000, even before the new policy was formally in place, and in March 2005, the IMF’s Executive Board assessed how much progress had been made.

The 2002 guidelines are based on five key principles: national ownership of policy programs; parsimony and clarity in the application of conditions; tailoring of policies to circumstances; and coordination with other multilateral institutions. These principles are intended to reinforce each other and improve program design and implementation. The guidelines depart from earlier practice primarily in that they call for streamlining structural conditionality and formulating programs in a way conducive to national ownership of policies. These issues were the focus of the recent review. Ultimately, the new guidelines will be judged successful if they contribute to improved economic outcomes, but it is too early to gauge whether this has been the case. This review sought instead to serve as an interim checkpoint, focusing on whether the guidelines are being implemented, and how that implementation can be improved.

What’s being covered?

If structural conditionality is becoming more focused on those measures critical to program success, it should become more concentrated in especially relevant areas and it should be more strongly linked to the country’s initial economic conditions. This is exactly what the review found has happened. Structural conditionality has shifted out of “noncore” areas and into “core” areas that are likely to be most closely linked to the goals of Fund-supported programs. Specifically, structural conditionality has become more concentrated in areas related to economic management and vulnerability, and less dispersed across sectors. Moreover, econometric evidence suggests that the link between the numbers of conditions in a particular area (such as fiscal management) and economic circumstances (for example, the level of the fiscal deficit) has become stronger, suggesting a sharper focus on priorities.

At the same time, the 2005 review points to some scope for further streamlining coverage. First, in examining staff reports on IMF loans, the review found that program strategies for those areas of action considered crucial are not always set out clearly. While this could be only a presentational issue, there is no doubt that focusing on a few key strategies in the process of developing a policy program is conducive to parsimonious conditionality. Second, greater care needs to be taken to set “structural benchmarks” in only critical areas. These differ from “structural performance criteria” in that disbursements of loan installments are not automatically interrupted if benchmarks are not observed. This makes it tempting to use benchmarks for less-than-critical measures when, in truth, they should be used for small steps in a critical process of reform, where a failure or delay in implementation of one step is not sufficient to derail the entire process.

Of course, streamlining the coverage of structural conditionality is not without risks. Clearly, conditionality has shifted away from growth- and efficiency-related reforms—even in low-income countries, where a focus on growth is especially needed. And World Bank conditionality does not appear to have stepped in to fill this “gap.”

Is this something to worry about? Not necessarily, as country ownership, rather than conditionality, ultimately drives policy agendas forward. Gaps in analytical work, policy advice, or technical assistance would thus arguably be more worrisome. Still, the issue clearly deserves to be kept under review and assessed in more depth once the outcomes of “streamlined” programs are known.

Numbers and clarity

Intuitively, streamlining could be expected to produce fewer structural conditions. In fact, there has been only a small decline in the number of conditions associated with Poverty Reduction and Growth Facility loans and none at all with regard to non-concessional loans. This is disappointing, but it is also necessary to recognize that numbers are a crude metric; indeed, sometimes a clearer focus on what is critical may bring with it a need for more conditions rather than fewer.

In fact, in the IMF’s nonconcessional lending there has been a large increase in conditions related to financial sector vulnerability, reflecting the growing understanding of the importance of this area. Other factors may also keep the numbers of conditions high, including detailed specification of conditions and a tendency to set more conditions when countries have weak track records. Both tendencies are in accordance with the guidelines—and indeed some borrowing countries prefer conditions set at a high level of detail, as they function as helpful guideposts—but they should not be allowed to get out of hand.

The recent record is unambiguously positive on the clarity of conditions. The 1990s had seen boundaries blurred between measures critical for disbursements to continue and others that merely signaled IMF encouragement or the authorities’ commitment. Five years into the streamlining initiative, program-related conditions are now almost always clearly specified and transparently distinguished from the rest of the authorities’ program.

Program implementation

Although it is too soon to judge whether IMF-supported programs under the new guidelines have contributed to better economic outcomes, some improvement is evident in program implementation. Early evidence suggests that programs now suffer fewer irremediable interruptions, although temporary interruptions and their counterparts—delays in completing program reviews—have not declined.

At first sight, developments in the implementation of structural performance criteria have been disappointing, although looking below the surface suggests a rather better picture. Performance criteria require waivers if they have not been implemented and if disbursements are to proceed. The review found that waiver rates have not fallen (and have even risen in nonconcessional loans). Tracking waived performance criteria through to the end of the arrangement, however, shows that an increasing proportion of these criteria is implemented eventually. This better implementation is in line with a greater focus on critical conditions, although waivers continue to be used to give the authorities leeway, in particular with respect to timing of implementation. Realistic timetables will be key to reducing waiver rates, while maximizing the assurance given to a borrowing country that it will be able to access IMF resources.

Evidence suggests that programs now suffer fewer irremediable interruptions, although temporary interruptions . . . and delays in . . . program reviews . . . have not declined.

Ownership and process

The 2005 review looked at the process of program development in 10 country cases. While the evidence is preliminary—and the IMF Independent Evaluation Office (IEO) upcoming review of structural conditionality will go into greater depth—indications are that IMF staff are making serious efforts to implement processes conducive to ownership: for example, by establishing an active dialogue with the authorities and accommodating their preferences where possible, seeking to involve all the key officials responsible for implementation, and helping the authorities work toward broad public ownership of the policy program.

Of course, good processes do not guarantee ownership and, indeed, gauging the level of ownership is, and will remain, a major challenge. Similarly, whether to proceed with a loan in the presence of uncertain ownership remains a delicate matter of judgment. Certainly, substituting conditionality for ownership is not the answer. Conditionality, especially prior actions, can be used as a device for governments to demonstrate their commitment. But the review advises caution. In its findings, programs with many prior actions tended to have worse implementation of subsequent conditionality than average, leaving one to wonder about the extent of ownership as a whole and the durability and quality of implementation of even the prior actions. In some of these cases, rather than loading programs with conditions and prior actions, it may be preferable to exercise greater selectivity and, where possible, make use of staff-monitored programs to establish a track record of implementation.

The IMF will continue its efforts to implement the new policy, guided by the findings of this review and, no doubt, by the recommendations of the forthcoming IEO evaluation of structural conditionality. A new review of the 2002 conditionality guidelines will be conducted in 2008. By that time data on multiyear economic outcomes of a number of “streamlined” programs will be available. That will allow the 2008 review to ask the key question that eluded this year’s review: have the new guidelines achieved their objective of helping borrowing countries reach better outcomes?

IMF Survey, Volume 34, Issue 15
Author: International Monetary Fund. External Relations Dept.