Chapter 2: What Is Prolonged Use and How Widespread Is It?
- International Monetary Fund. Independent Evaluation Office
- Published Date:
- April 2003
What Is Prolonged Use of IMF Resources?
1. There is no official or generally agreed definition of prolonged use of Fund resources (UFR). In previous studies, undertaken both within and outside the IMF, prolonged UFR has been characterized in several different ways, focusing either on the number of years spent by a country under IMF-supported programs—in some cases approximated by the number of IMF-supported programs entered into—or on the length of time a country has outstanding obligations to the IMF. Annex 1 discusses the most commonly used definitions in more detail.
2. In the current project, the definition used is based on the concept of “time under arrangement”: countries are defined as prolonged users if they have been under arrangements for at least 7 years out of any 10.1 This concept was preferred to other alternatives mainly because it is more homogeneous across lending instruments (since a single IMF “lending arrangement” can vary in length and repayment period, depending upon the particular facility used). It permits the use of a single threshold for all member countries, regardless of the types of facility used.2 Moreover, focusing on “time under arrangement” goes to the heart of issues such as program design, ownership, and conditionality.
3. The threshold of 7 years in any 10, which is higher than in previous reviews of the phenomenon, was chosen so that any country with just two three-year arrangements (EFF or ESAF/PRGF) in a decade, or a combination of a few Stand-By Arrangements with one medium-term arrangement, would be classified as a “temporary” user, not as a prolonged user.3 This relatively high threshold means that some countries classified as “temporary” users might have made relatively frequent and/or protracted use of IMF resources.
4. The 2000 Review of Fund Facilities made a distinction between “repeat users”—that is, countries with many programs but where effective use of IMF resources was relatively low because the programs went off-track quickly—and “prolonged users”—that is, countries that made greater effective use of the IMF resources. The former type is represented in our country case studies by Pakistan and raises the most serious questions about program implementation and ownership. The latter type is, to some extent, closer to the examples of the Philippines and Senegal and raises questions about the appropriate length of IMF program involvement in cases with protracted adjustment difficulties and about the revolving nature of IMF resources. In general, however, the empirical evidence suggests no sharp dividing line between the two types and, for the purposes of this evaluation, the terms “prolonged use” and “repeat use” will be used to mean the same thing.
5. Moreover, neither the choice of concept on which to base the definition of prolonged use, nor the choice of threshold, has a major impact on the general trends observed, either as regards the extent of the phenomenon of prolonged use or its evolution over the last three decades.4 Settling on a specific definition was necessary in order to identify a list of countries on which to conduct quantitative analyses, but the general conclusions reached do not seem very sensitive to the precise definition.
Extent and Evolution of Prolonged Use over 1971–2000
Trends in prolonged use over 1971–2000
Scope of prolonged use
6. Out of 128 countries that made use of IMF resources during the 1971–2000 period,5 51 countries meet our definition of prolonged use at some time during the period (44 countries if precautionary arrangements are excluded) (Table 2.1).6 Prolonged users are predominantly—but by no means exclusively—low-income countries with little or no access to private capital markets. For example, in the group of prolonged users excluding precautionary arrangements, 29 of the 44 prolonged users are eligible for the IMF’s concessional facility (hence-forth, PRGF-eligible) (Figure 2.1).
|Countries that would be classified as prolonged users excluding precautionary arrangements (44)||Countries that would be classified as prolonged users only if precautionary arrangements were included (7)|
|Congo, Dem. Rep. of||Mozambique||Uruguay|
|Kyrgyz Rep.||F.S.R. of Yugoslavia|
Figure 2.1 summarizes in graphic form the history of lending arrangements of the countries listed here. The vast majority of these countries were still under program at the end of the period. Of the 13 countries which were not, 4 countries (the Democratic Republic of Congo, Liberia, Somalia, and the Federal Socialist Republic of Yugoslavia) were no longer eligible to use IMF resources at the end of the period (the first three because of arrears to the IMF and the fourth because it was dissolved).
Countries in italics are PRGF-eligible. However, a large part of the resources made available to them over that period came from the General Resources Account (GRA).
Figure 2.1.IMF Arrangements with Prolonged Users During 1970–2000
7. Nearly 40 percent of prolonged users made an intensive use of the IMF’s general resources, in the sense that they had outstanding obligations to the General Resources Account (GRA) of over 100 percent of their quota for 10 years or more (Table 2.2). When both general and concessional resources are taken into account, more than half the countries had average outstanding liabilities over the 30-year period in excess of 100 percent of their quota, that is, they made both long and large use of IMF resources.
|Use of Fund credit (GRA) (Years with outstanding obligations)|
|Time under program 1 (years)||First year under program (In period)||Last year under program (In period)||Number of programs||Percent of partially disbursed programs1||Over 100 percent of quota||Over 200 percent of quota||Average outstanding credits and loans2||Outstanding credits and loans as of Dec. 20002|
|Country||Of which precautionary||GRA||Non-GRA|
|Congo, Democratic Republic of||10.7||0.0||1984||1990||9||1||90||13||4||137||103|
Time spent refers to the actual period covered by arrangements, whether or not a country was eligible to draw under the program. Only programs that were precautionary from the point of approval and were effectively treated as such are represented as precautionary.
As a percentage of quota.
Evolution of prolonged use over time
8. Prolonged use is neither a recent nor a rare phenomenon and contrary to what previous internal reviews, which focused on the use of the IMF’s general resources, have suggested, it is also not one that is diminishing in importance. As indicated in the 1991 and 2000 reviews, prolonged use started to build up in the second half of the 1970s and accelerated sharply in the first half of the 1980s as a result of the debt crisis. Thereafter, the establishment of concessional facilities resulted in a large shift of prolonged users from the general to the concessional window, so that by 1990 the number of prolonged users of general resources had fallen dramatically. Prolonged use, thus defined, rose again slightly in the 1990s, partly due to the intensive involvement of the IMF in “transition” countries.
9. However, an analysis of prolonged UFR that does not limit itself to the General Resources Account (GRA) leads to different conclusions. First, the decline in the scope of the phenomenon that occurred in the late 1980s is largely cosmetic, since five of the countries previously characterized as prolonged users fell into arrears and were declared ineligible for further borrowing from the IMF,7 and most of the others had become prolonged users of concessional resources. In fact, prolonged use, according to the definition used in this evaluation, has consistently expanded since the late 1970s. In terms of the number of countries, most of the expansion was in PRGF-eligible countries, but in terms of financial obligations, the expansion in prolonged use of general resources was greater (Figure 2.2).8
Figure 2.2.Prolonged Use by Number of Countries and by Outstanding Exposure to the IMF
Sources: IMF Treasurer’s Department and IEO calculations.
10. Furthermore, the increase in the overall number of prolonged users exceeded the pace of increase of the number of countries making use of the IMF’s resources. As a result, while prolonged users represented a little over 10 percent of users of IMF resources in the late 1970s, that proportion had gone up to over 30 percent in 2000, with an even higher proportion of PRGF-eligible countries (40 percent, against a little over 20 percent for GRA-only borrowers). The share of prolonged users in total resource commitments has also tended to increase over time, reaching 60 percent of PRGF resources and a little over 20 percent of general resources in 2000. (See the figures in Annex 1 for more details.)
Persistence of prolonged use
11. The persistence of prolonged use is substantial: almost 40 percent of countries that became prolonged users in the second half of the 1980s were still in that group in 2000, and 60 percent of countries that were prolonged users in 2000 had joined that group prior to 1995. Moreover, over 1970–2000, the average length of episodes of prolonged use was 10 years, and even this indicator underestimates the eventual length since the majority of countries characterized as prolonged users were still in that category in 2001.9
12. Further evidence of the persistence of prolonged use is provided by the small number of “graduators,” that is, prolonged users that have ceased to make use of IMF resources. Of the 51 countries that have been prolonged users at some time between 1971 and 2000, just 12 had less than 25 percent of quota outstanding at the end of May 2002.10 It is also striking that only 15 of the 51 prolonged users did not have active arrangements with the IMF at some point in 2001–02 and this includes 3 countries that were in substantial arrears to the IMF and were therefore ineligible for such arrangements.
13. Another indication of the extent of prolonged use is found in Jeanne and Zettelmeyer (2001), who measure the time frame in which IMF members eliminate outstanding obligations to the Fund—after taking account of new lending.11 They find that, for developing countries, about 40 percent of all the lending cycles initiated since the creation of the IMF were not completed at end-2000, and that the average length of such “incomplete” cycles is 18 years (see Table 1.1 in Annex 1). Not surprisingly, the proportion of uncompleted lending cycles and their length is even higher for PRGF-eligible (and HIPC) countries. But they also find that 30 percent of emerging market countries12 initiated a lending cycle prior to 1991 that was still incomplete at the end of 2000. The average length of time these countries had outstanding obligations to the IMF is about 21 years. Clearly, for these countries, IMF resources are revolving very slowly.
Prolonged use in 200113
14. A look at the extent of prolonged use in 2001 using the dynamic definition suggests that the incidence of prolonged use is important for both concessional and GRA resources. The largest number of prolonged users (22 out of 31) are accounted for by countries eligible for concessional resources. However, the exposure to prolonged users in the General Resources Account is much larger than in the PRGF Trust Fund (SDR 20.6 billion compared with SDR 3.5 billion).14 Within each resource category, the incidence of prolonged use varies. Prolonged users account for 62 percent of the commitments of concessional resources. The proportion of GRA commitments absorbed by prolonged users in 2001 is lower at 37 percent but this is still substantial (Figure 2.3).
Figure 2.3.Distribution of Current Commitments of IMF Resources and Current Arrangements
Sources: IMF Treasurer’s Department and IEO calculations.
This definition is applied in two different ways in the report, depending on the type of analysis undertaken and the nature of the question being addressed: in one case, prolonged users (PUs) are treated as an invariant group that includes all the countries that met the 7 out of 10 years criterion at least once over the 1971–2000 period (i.e., a “fixed” definition); in the other case, the composition of the PU group varies each year, as it includes only the countries that met the criterion in that particular year (i.e., a “dynamic” definition).
Short-term arrangements (SBAs) typically span 12 to 18 months, with no minimum and a maximum length of three years, whereas medium-term arrangements (under the EFF, SAF, and ESAF/PRGF) cover a three-year period. Repayment periods have been markedly differentiated across lending facilities, in order to take account of the different speeds at which various countries can be expected to return to balance of payments viability: from 2½ years for the SRF to 10 years for the EFF and PRGF.
However, unlike the definition used in previous IMF reviews, the proposed definition does not impose any threshold on the outstanding use of IMF resources at the end of the period. This is in order not to exclude countries that have repaid all or most of their outstanding obligations to the IMF and in that sense have subsequently “graduated” from IMF support. The terms “prolonged user” and “temporary user” (i.e., all those not counted as prolonged users) are employed as a convenient terminology for purposes of this evaluation. Under the IMF’s Articles of Agreement, all use of IMF resources is supposed to be temporary.
We reached this conclusion after mapping out prolonged use based on the alternative concept of number of IMF arrangements entered into, by including or excluding precautionary arrangements within the total for each country, and by raising or decreasing the “time under program” threshold by one year. Figure 2.1 below as well as the figures presented in Annex 1 broadly illustrate that point.
This excludes countries whose UFR consisted exclusively of “outright purchases,” that is, was not associated with a lending arrangement or program (e.g., first credit tranche purchases).
So-called “precautionary” arrangements are the same as other IMF arrangements except that the authorities have indicated that they do not intend to draw on the resources made available. Even though that commitment is never binding, to the extent that it is observed, these programs do not reflect any actual balance of payments need nor involve any actual use of IMF resources. However, the program negotiation and Board approval process is identical, and the committed resources can be drawn upon if the authorities so wish, and indicate they have a balance of payments need; therefore the resources are not available for lending to other members. A significant number of precautionary arrangements were eventually drawn upon. Therefore, for completeness, the tables in this chapter identify those countries that would be classified as prolonged users if precautionary arrangements were included in the definition, but not otherwise. Arrangements under the ESAF/PRGF cannot be precautionary.
These are the four countries mentioned in footnote 1 of Table 2.1, plus Zambia, which has subsequently regained eligibility.
Outstanding obligations are shown according to whether or not a country is PRGF-eligible, not according to the nature of resources at stake. The figures are based on the “dynamic” definition of prolonged use, so that prolonged users’ outstanding obligations in each year are taken into account only if they qualify as prolonged user in that year. For example, Ecuador’s obligations in 1999 are not taken into account because it did not qualify as a prolonged user in that year.
This average is the same in the GRA-only and PRGF-eligible groups.
Bangladesh, Costa Rica, Egypt, El Salvador, Equatorial Guinea, Haiti, Jamaica, Korea, Latvia, Morocco, Panama, and Mexico. Of these, Costa Rica, Egypt, El Salvador, Korea, and Latvia only qualify as prolonged users if precautionary programs are taken into account.
In the absence of subsequent programs, the length of this time frame, which they call a “lending cycle,” should be equal to the sum of the program and the repayment period, that is, a maximum of 13 years for an EFF or an ESAF/PRGF arrangement and 6½ years for an 18-month SBA. This concept would not be entirely satisfactory as the main basis for the definition of prolonged use, because it does not distinguish the respective contributions of repeat use and of the length of the repayment period (which can and has been increased by policy decisions).
Defined as countries whose sovereign bonds are tracked in the J.P. Morgan Emerging Market Global Bond Index, which is an indication that they would normally be expected to have access to private market financing.
Data reported in this section refer to the situation as of December 31, 2001.
Prolonged users in 2001 (i.e., countries that had IMF arrangements for seven or more years during the previous decade) were (i) Argentina, Bulgaria, Gabon, Jordan, Panama, Peru, the Philippines, Romania, and Ukraine (GRA-only borrowers); and (ii) Albania, Benin, Bolivia, Burkina Faso, Cameroon, Côte d’Ivoire, Ethiopia, Ghana, Guinea, Guyana, Honduras, the Kyrgyz Republic, Malawi, Mali, Mauritania, Mongolia, Mozambique, Nicaragua, Pakistan, Senegal, Tanzania, and Uganda (PRGF-eligible members).