2011 Review of Conditionality - Technical Appendices

This section provides the background studies relating to dimensions of Fund policy on conditionality. Appendix 1 provides a review of Fund experience with coordination, both in a low-income country (LIC) setting (in African programs) and in an emerging market and advanced economy setting in the European Union (EU) and Euro Area (EA). Appendix 2 summarizes the recent changes to debt limits in LICs and provides an assessment of the implementation of this policy in the early stages (up to mid-February 2011). Appendix 3 reviews the experience of countries with the Flexible Credit Line (FCL) and Precautionary Credit Line (PCL)-supported programs. Appendix 4 examines the impact of the 2009 Special Drawing Rights (SDR) allocation on program design


This section provides the background studies relating to dimensions of Fund policy on conditionality. Appendix 1 provides a review of Fund experience with coordination, both in a low-income country (LIC) setting (in African programs) and in an emerging market and advanced economy setting in the European Union (EU) and Euro Area (EA). Appendix 2 summarizes the recent changes to debt limits in LICs and provides an assessment of the implementation of this policy in the early stages (up to mid-February 2011). Appendix 3 reviews the experience of countries with the Flexible Credit Line (FCL) and Precautionary Credit Line (PCL)-supported programs. Appendix 4 examines the impact of the 2009 Special Drawing Rights (SDR) allocation on program design

I.Fund Policy on Conditionality

1. This section provides the background studies relating to dimensions of Fund policy on conditionality.1 Appendix 1 provides a review of Fund experience with coordination, both in a low-income country (LIC) setting (in African programs) and in an emerging market and advanced economy setting in the European Union (EU) and Euro Area (EA). Appendix 2 summarizes the recent changes to debt limits in LICs and provides an assessment of the implementation of this policy in the early stages (up to mid-February 2011). Appendix 3 reviews the experience of countries with the Flexible Credit Line (FCL) and Precautionary Credit Line (PCL)-supported programs. Appendix 4 examines the impact of the 2009 Special Drawing Rights (SDR) allocation on program design.

Appendix 1: Coordination Coordination of conditionality in African countries

2. Numerous institutional arrangements have developed to meet the challenges of conditionality coordination in LICs in sub-Saharan Africa (SSA). These arrangements were evident in a survey of Fund country teams covering six countries with programs supported by Extended Credit Facility (ECF) arrangements (Burkina Faso, Ghana, Liberia, Mali, Sierra Leone, and Togo) and three countries with Policy Support Instrument (PSI)-supported programs (Mozambique, Senegal, and Uganda).

3.The coordination of conditionality among country authorities, the Fund, and development partners is generally close. It strives to follow the principles of the Paris Declaration on aid effectiveness and to promote ownership and implementation.

  • Conditionality is an institutional feature of donors’ operations in SSA countries. In addition to the Fund, other multilateral institutions, such as the World Bank, the African Development Bank (AfDB), and the EU, use policy conditionality in their operations, as do some bilateral development partners.

  • Most development partners strive for close coordination and, where feasible, harmonization of their activities. Many development partners include conditionality linked to their support, with coordination ensured through a comprehensive performance assessment framework (e.g. Liberia, Mali, Togo, and Uganda). The Fund is typically closely involved in such coordination efforts.

4. The coordination of conditionality with other development partners takes place at each stage of the program cycle, through both formal and informal channels. It includes formal bilateral meetings and information sharing as well as informal exchanges at the margins of program discussions in the field, during program document preparation at headquarters, at international donor conferences, and through other ad-hoc contacts.

  • The coordination of conditionality is particularly close with the World Bank. During policy note preparation, Fund missions usually give World Bank staff an opportunity to comment on the draft, discuss bilaterally the list of proposed structural conditions, and attend policy consultation meetings (the Fund interdepartmental meetings held before program missions). Bank staff often comments on the authorities’ draft Memorandum of Economic and Financial Policies (MEFP) and makes suggestions on the sections on structural reforms subject to its conditionality.

  • Development partners pay special attention to conditionality coordination in the areas of shared responsibilities. Public financial management (PFM), revenue and customs administration, expenditure management, fiscal transparency, budget preparation, public enterprise reforms, central bank operations, and financial sector reforms are the main areas of coordination.2 At the same time, the Fund usually relies on World Bank conditionality to assess progress in other structural areas, such as civil service and pension system reforms, privatization, labor market policies, private sector development, education, and health policy.

  • The “lead agency” is determined through an informal process based on the agreed on areas of primary expertise. In Ghana, for example, the Fund takes the lead on macro-critical conditionality (fiscal, monetary, financial sector, and debt management issues), while the World Bank focuses on other structural measures and sectoral reforms (e.g., poverty reduction, energy sector, and public sector reform). In the PFM area, the Fund provides extensive strategic guidance and technical support on critical reforms (namely, program-based budgeting, expenditure control, and arrears management), while the EU deals mainly with specific reforms that complement program conditionality.

  • The Joint Managements Action Plan (JMAP) supports the coordination of conditionality with the World Bank. The JMAP has been a useful framework for the consultation process (Burkina Faso) and for discussion of the best ways for each institution to support the authorities’ Poverty Reduction Strategy Paper (Togo).

5. Streamlined Fund conditionality aims to avoid duplication with development partners.

  • Beyond the Fund’s core areas of responsibility, structural reforms in Fund-supported programs often rely on conditionality set by development partners, usually the World Bank (Ghana, Liberia, Mozambique, Sierra Leone, and Uganda). Fund structural conditionality is limited to measures in core areas of Fund responsibility.

  • Several institutional tools are used to avoid duplicating conditions. Donors’ framework agreements on budget support have been the most useful coordination tool in Liberia, Mali, Togo, and Uganda. In some countries (Burkina Faso and Senegal), the government prepares a unified matrix of policy actions and expected outcomes. It may also establish working groups for priority areas (e.g., PFM, education, and health), which are used to align conditionality of development partners with the country’s reform priorities. In some countries, lead donors establish their own committees to coordinate lending and technical assistance activities (Ghana, Liberia, and Togo). In Liberia, the reform pillars established by the authorities include a clear indication for the leadership role for certain development partners. Coordination in the field, primarily through Fund resident representative offices, also remains an efficient way to keep track of conditionality discussed between the authorities and other development partners, and avoid duplication.

  • Development partners usually take into account Fund-supported program status in their operations. Maintenance of an adequate macroeconomic policy framework, as shown by an on-track program, can be a condition for a World Bank development policy operation, and successful completion of a program review usually is a necessary condition for the disbursement of AfDB budget support.

6. The Fund is working with other development partners and country authorities to further strengthen conditionality coordination.

  • Strong Fund field presence and a proactive resident representative office are essential. Development partners are generally more decentralized than the Fund, and it is important that the Fund engage fully in the coordination process at the country level. Typically, this is done through the resident representative. For example, in Mozambique, Fund and World Bank representatives are permanent members of the so-called G19, a group of core donors that provides general budget support.

  • The authorities should be closely involved in coordination initiatives and internalize conditionality as an important guide post for their reforms. In some countries, the government assumes full ownership of coordination by establishing donor coordination committees (Mozambique, Senegal, Sierra Leone, and Uganda). In others, unified matrices of policy actions, sectoral working groups, and donors’ framework agreements help focus the attention of development partners on the reform priorities where their expertise would be most valuable. Increasingly, country authorities and Fund staff hold conferences with a range of stakeholders, including development partners, to discuss economic policy priorities before designing new programs. This initiative has promoted both ownership and coordination.

Fund/EU collaboration in programs with emerging EU members

7. In October 2008, Hungary faced an acute balance-of-payments (BoP) crisis and approached the Fund and the EU simultaneously for BoP assistance. The Hungary case set a precedent, resolving earlier ambiguity on how a BoP crisis in a non-euro EU country would be handled. The ambiguity had arisen from the EU’s (long dormant) small facility to support non-euro members in case of a BoP crisis, and from the legal requirement that EU members consult with the European Commission (EC) prior to approaching the Fund.3 From the Fund’s perspective, EU members have the right to approach the Fund independently. In the event, communication between the institutions facilitated an interpretation that allowed a simultaneous request from Hungary to the EU and the Fund.

8. Following the Hungary precedent, the Latvia and Romania programs were supported by both the Fund and the EU. All Fund/EU arrangements/agreements followed similar modalities, developed in the case of Hungary. They included early consultation, overlapping missions, and ongoing exchange of information. However, in view of differing internal procedures, reaching understanding on program revisions in the context of reviews was at times challenging, and in at least one case—Latvia—the EU agreed to a program review while there was still ongoing discussions with respect to the review under the Fund arrangement. Since then, information exchange during missions has been developed further, with a view to avoiding “mixed messages.” More recently the Fund and the EU approved arrangements and agreements for certain euro area members, under procedures that closely matched those developed earlier with non-euro area EU programs (see next section).

9. As to program contents, the institutions cooperated closely on the macro-framework. While the Fund played a somewhat greater role on financial issues, the EU took center stage on structural issues. This dividing line, however, was not watertight: the EU played a role in financial sector issues, not least because the EC’s Directorate-General for Competition (DG Comp) needed to approve government aid to the financial sector when such measures became part of crisis resolution. At the same time, the Fund dealt with structural issues of macro-critical importance—including, for example, tax administration reform, budgetary institutions, and banking sector restructuring,

10. Financing shares under the programs differed, depending on Fund access limits and the involvement of other institutions. In Hungary and Romania, over 60 percent of the programmed financing package originated from the Fund, while in Latvia the share was only 23 percent, which—given the size of the financing gap and Latvia’s small quota—was nevertheless a program involving “exceptional access” for Fund financing. The financing share of the EU ranged from 24 percent in Romania to 41 percent in Latvia compared with around two thirds in the Greece, Ireland, and Portugal programs.4 Other IFIs and bilateral donors contributed the remainder of the needed financing (Table 1).

Table 1.

Financing packages for Emerging European Member Countries Under Fund-Support programs

(As of June 2011, billions of Euros)1

article image
Source:Staff calculations

Figures are for original programs,excluding folloe-up precautionary stand by arrangement for Romania.

Disbursment as of end-june 2011.For Romania the European Inverstment Bank for Reconstruction and Developement/International finance corporation disbursed more than committed at the begining of the program.

Fund/EU collaboration in programs with Euro Area countries

11. Fund/EU collaboration in conditionality design and program review in euro area program countries constitutes an unprecedented experience. The Fund’s approach to crisis management is being combined with the EC’s drive to close longstanding gaps in the adoption of EU laws and regulations (the acquis communautaire) in the affected economies. And, as a third independent player, the European Central Bank (ECB) monitors financial stability in the Euro area and the transmission of its monetary stance.

12. Greece, Ireland, and Portugal were confronted with common challenges—acute fiscal imbalances mostly due to pro-cyclical bias, insufficient fiscal transparency, and costly entitlements; and a considerable degree of financial fragility—affecting public finances adversely and threatening contagion.5 Greece and Portugal also suffered from lackluster potential growth and inadequate market flexibility. By implication, comprehensive action was required on three fronts to achieve sustainable fiscal adjustment (supported by structural fiscal measures), allow orderly deleveraging of the banking system, and advance deep structural reforms to generate growth. And with the three countries being members of a monetary union (the European Monetary Union-EMU), ensuring consistency between these policy areas was a challenge, also taking into account the ramifications of the actions for EMU and the ECB.

13. Detailed discussions on conditionality were needed to achieve coherence between the documents outlining the economic policy intentions of the authorities—the Memorandum of Understanding (MoU, EC) and the Memorandum of Economic and Financial Policies (MEFP, Fund). The distinctive character of both documents was nevertheless preserved, with the MoU generally including a wider and more detailed set of structural measures and Fund conditionality focusing on a smaller number of priority actions seen as macro-critical. Useful complementarities emerged in various policy areas, such as fiscal policy, where the EC contributed more on health care and pension reforms and the Fund’s expertise was crucial to outline a credible fiscal adjustment path (allowing time for structural reforms to support consolidation efforts), tackle fiscal risks, and design structural reforms. Financial sector conditionality generally aimed to encourage deleveraging through market-based mechanisms, supported by back-up facilities. The Fund brought expertise from its involvement in numerous banking crises, the ECB contributed its expertise on the provision of liquidity support to affected institutions, and the EC Competition Authorities advised on the application of competition and internal market rules. Reflecting a legacy of prolonged competitiveness losses (especially in Greece and Portugal), an ambitious structural agenda aimed at improving market flexibility (absent using the nominal exchange rate), appropriately distilled focus areas of critical importance for growth in the Fund’s MEFP, and converged in a few Fund structural benchmarks.

14. Unlike conventional Fund-supported programs, additional stakeholders involved represented an extra layer of complexity to conditionality design, often resulting in longer program discussions and less effective decision making. Differing views on program conditionality largely reflected differences in approach rather than fundamental dissonance on ultimate objectives. The sometimes lengthy discussions were nevertheless considered necessary to align positions on a number of sensitive items (e.g., the size and duration of ECB’s liquidity support and the pace of fiscal consolidation in all three countries). Building confidence within teams and creating a common sense of ownership proved a valuable asset to monitoring the implementation of conditionality, as program reviews require constant crosschecking and a gradual adaptation between the MoU and the MEFP.

15. While generally anchored in the Fund’s traditional areas of responsibility, conditionality in these programs was tailored to address country-specific challenges. For instance, far-reaching judicial reforms, clearly outside Fund’s core area of responsibility, were considered critically important for achieving program objectives in Portugal, with responsibility shared between the EC and the Fund. Critical legal reforms affecting bank resolution (in close collaboration with the ECB) and corporate and household balance sheet restructuring placed the Fund in a prominent position.

16. Conditionality follow-up relies on the principles of mutual trust and reciprocal evaluation rather than a strict division of labor within teams. There is no formal distribution of roles in discussions with the authorities, with each team steering discussions toward macro-critical conditionality independently of its own area of expertise. Reflecting this partnership approach, every team assigns dedicated staff to the key policy areas addressed by the programs, also beyond their traditional expertise.

17. Given the vast policy agenda, Fund/EU teams have aimed to establish monitoring tools to assess implementation efforts. Reflecting evolving circumstances, conditionality is adjusted, if needed, to reflect program objectives. For instance, conditionality on structural reforms in Greece was reviewed to reinforce reforms in labor and services markets, considered critical for regaining competitiveness. Reinforced measures for accelerating the transition to a healthy banking sector were added to the Irish program in March 2011.

Appendix 2: Debt Limits in Fund-Supported Programs in Low-income Countries

18. Fund-supported programs with LICs typically include external debt limits. These seek to prevent the build-up of unsustainable debt, while allowing for adequate external financing. The main component of these limits is a concessionality requirement applying to external debt contracted or guaranteed by the official sector. The debt limits policy applies to all members with Fund-supported programs, but provides specific considerations for members with access to concessional financing. This note therefore focuses on its implications for LICs.6

19. The Fund introduced new guidelines on debt limits in December 2009.7 The new framework moves away from a single design for concessionality requirements toward a menu of options. Accordingly, the new framework takes better account of the diversity of challenges faced by LICs with regard to their debt vulnerabilities and their macroeconomic and PFM capacity (“capacity”), which are key determinants of a country’s ability to borrow sustainably and productively.

20. This appendix is structured as follows. It first describes the main changes to the debt limits policy, reviewing its implementation up to early 2011. It further reports on related structural conditionality in the debt area, as well as on coordination with the World Bank.

What has changed with the new guidelines?

21. Before the introduction of the new guidelines, most programs with LiCs shared the same overall design for external debt limits. The general practice was to restrict non-concessional external borrowing (NCB) without constraining concessional borrowing. Typically, program conditionality included a minimum concessionality requirement of 35 percent; this was, for instance, the case in 26 out of 34 Poverty Reduction and Growth Facility (PRGF) and PSI-supported programs as of end-August 2006. Debt limits tended to be tighter—for instance, through a minimum concessionality requirement higher than 35 percent—in countries with higher debt vulnerabilities, although this was not systematic.8 Some flexibility was exercised by allowing for looser debt limits in some cases, e.g., through an exception to the zero NCB ceiling.9

22. By contrast, the new framework for concessionality requirements is based on a menu of options, which takes better account of the diversity of situations faced by LICs. A country where debt vulnerabilities are relatively high should indeed adopt tight concessionality requirements. Conversely, if debt vulnerabilities are relatively low, looser requirements can be considered. Similarly, the higher a country’s PFM capacity, the better it will be positioned to implement and benefit from more flexible but also more technically demanding approaches to concessionality requirements. Under this framework, each of the two above-mentioned factors can take two values, “lower” or “higher.” Thus, this framework results in four different types of concessionality requirements. Unless debt sustainability is a serious concern (“higher” value) and capacity is limited (“lower” value), the applicable concessionality requirements normally allow for NCB and, hence, provide generally more flexibility than the previous design of concessionality requirements. The various options, as well as the eligibility of program countries for these options, as of mid-February 2011, are presented in Table 2.

Implementation of the New Policy

23. Based on the 2010 capacity assessment of program countries, 25 out of a total of 39 program countries were eligible for more flexible external debt limits (Table 3).10 Four of these countries had higher capacity, and are therefore eligible for the more advanced (and new) options for concessionality requirements. The other 21 countries have lower capacity and lower debt vulnerabilities; accordingly, space for NCB can be considered in their program. The remaining 14 program countries have been classified as having lower capacity and higher vulnerabilities, and for them, NCB is expected to be truly exceptional, as under the previous policy.11

Higher-capacity countries

24. Half of the higher-capacity countries have made use of the more advanced options.

Armenia’s ECF arrangement includes, as an indicative target, a 30 percent floor on the average concessionality of new external borrowing. Georgia’s Stand By Arrangement (SBA) includes an indicative ceiling on total public external borrowing, which replaced a ceiling on non-concessional public external borrowing. The Moldova and Cape Verde programs continue to rely on traditional concessionality requirements; their programs including a nonzero limit on NCB.12

Lower-capacity countries

25. Ten of the 21 program countries with “lower vulnerability/lower capacity” (LL countries) have space for NCB in their programs. The size of this space ranges from 1.8 percent of GDP (Kenya) to 10.4 percent of GDP (Lesotho). It is still too early to assess the extent to which the incidence of nonzero limits has increased with the new policy, as a large number of eligible countries have not yet availed themselves of this flexibility. However, it is already clear that the average size of nonzero limits has increased compared with recent years (to an average of about 4.1 percent of GDP under the current policy).

26. The absence of nonzero limits in the other 11 LL countries’ programs may reflect a number of factors, including: (i) the absence of public investment projects, for which non-concessional financing sources could be tapped; (ii) the availability of large amounts of concessional resources (either from traditional donors or emerging creditors); and (iii) varying degrees of caution (or risk appetite) across countries.

27. In several programs, untied external debt limits have been a vehicle for exercising flexibility. Untied external debt limits—which are not linked to a specific project or type of expenditure—have been used in countries with relatively higher capacity. So far, six countries (Honduras, Kenya, Mozambique, Tanzania, Uganda, and Zambia) have used untied external debt limits, two (Lesotho and Rwanda) used tied limits, and two (Ghana and Senegal) employed a mix of tied and untied limits. In all the cases with untied external debt limits, non-concessional resources have been expected to finance expenditures in specific sectors, usually identified as priority areas in the government’s public investment program, for which concessional financing is typically not available. Untied external debt limits, when particularly large, have occasionally been complemented with safeguards on phasing disbursements to address possible absorptive capacity issues.13

28. External debt limits tied to specific projects were supported by an assessment of profitability. These assessments are critical to ensure that project returns are sufficiently high to cover the costs of the associated NCB. Even in the case of untied external debt limits, some country authorities conducted feasibility studies for projects chosen to be financed with non-concessional resources and considered priority in their respective public investment strategies (e.g., Mozambique). Third party assessment was not required when not available for smaller projects, and was replaced by a profitability assessment from the authorities (e.g., Senegal).

Operational Issues

29. The implementation of the new debt limits policy has raised some operational issues. These include (i) coverage of state-owned enterprises (SOEs) in the definition of external debt limits, reflecting the difficulty to move to broad coverage due to inadequate information and monitoring systems; (ii) the use of the currency of debt denomination when assessing concessionality; (iii) country eligibility for untied debt limits; and (iv) assumptions of NCB beyond the program period in debt sustainability analysis (DSA). While temporary solutions for these issues have been found, they will need to be reconsidered in the context of the forthcoming review of the debt limits policy.

Structural Conditionality

30. Programs with space for NCB, particularly those with untied external debt limits, have usually included capacity-enhancing reforms. Seven of the 14 countries with room in their programs for NCB have also included structural benchmarks aimed at enhancing debt and PFM capacity (Table 3).14 The completion of a medium- term debt management strategy and/or preparation of a DSA by the authorities have been the most common measures included in programs.

Coordination with the World Bank

31. The new debt limits policy has been implemented in close coordination with the World Bank, ensuring consistency of both institutions’ NCB policies. Following the adoption by the Fund of new guidelines, the Bank decided to revise its NCB policy to preserve alignment with the Fund. Two critical assessments under the policies are made either jointly or in close coordination—DSAs and the assessment of capacity. So far, the latter has always coincided for countries assessed by the two institutions, ensuring that member countries to which the NCB policies of both the Fund and the Bank apply are subject to the same requirements.

Table 2.

Concessionality Requirements: Eligibility of Program LICs for the Various Options1

(as of February 15, 2011)

article image

While capacity is assessed once a year, the distribution of countries may change depending on the latest DSA results. The authorities could choose to opt for tighter debt limits than implied by the concessionality matrix

Table 3.

PRGT-Eligible Countries: External Debt Concessionality Requirements Under The New Policy As of February 15, 2011

article image

Tied, untied, and mixed indicate countries with non-zero ceiling on NCB, while zero ceiling indicates no roomfor NCB.

'X' indicates the use of structuralconditionality on debt management since the implementation of the new policy.

NCB limit designed using advanced options under the debt limits policy.

During the first review of the ECF (Nov. 2010), roomfor NCB for a specific project was granted as an exception to the normal NCB zero ceiling. This, in view of the need for the project’s continuity and due to its significance in enhancing the potential ofthe agricultural sector.

In the approval of Maldives’ SBA (Dec. 2009), a ceiling for non-concessional borrowing was introduced, applicable through Dec. 2010. This was on the grounds of debt management considerations (substituting external debt for more costly domestic debt) and the use of resources in economically viable projects.

Appendix 3: Flexible Credit Line and Precautionary Credit Line15

32. Countries are assessed against a set of qualification criteria—ex-ante conditionality—before entering a Fund arrangement under the FCL and the PCL.16 An FCL arrangement is only subject to ex ante conditionality, while a PCL arrangement is subject to both ex ante conditionality and limited ex post conditionality to address vulnerabilities identified during the qualification process. The rationale for the ex ante approach is that as the member’s fundamentals and policy framework are appropriate for actual or potential financing pressures it faces, no new measures are required to resolve its balance of payments (BoP) difficulties. Moreover, the member’s strong track record gives the Fund confidence that in cases of unexpected developments, the member would take additional measures as needed. Ex ante conditionality also helps reduce perceptions of stigma associated with Fund lending.

33. The nine qualification criteria for FCL arrangements cover areas of critical importance to macroeconomic, financial, and external stability, as well as data quality. Qualification for PCL arrangements is also assessed in light of these criteria. Specifically:

  • An FCL arrangement would only be approved for members that are very strong performers. While countries are reassessed on the qualification criteria at a 12-month review requirement for a 24-month arrangement, FCL arrangements are not subject to prior actions, performance criteria, or other forms of ex-post program monitoring.

  • A PCL arrangement would only be approvedfor members that are strong performers, but require some ex post conditionality—prior actions, performance criteria, indicative targets, or structural benchmarks. The extent of policy conditionality and monitoring is tailored to the member’s circumstances as informed by the initial assessment under the qualification framework and by subsequent semi-annual reviews. Semi-annual reviews are required to assess whether the country still meets the qualification requirements for the PCL and whether the program remains on track.

34. Thus far, three countries have received FCL arrangements: Colombia, Mexico, and Poland.17 None of these countries have made drawings under the FCL. Ex ante conditionality has been consistently applied across the three FCL cases:

  • All three countries received very positive assessments of policies in their most recent Article IV consultations.

  • Against the nine detailed qualification criteria, all three countries performed very strongly in most areas while there were no significant shortcomings in any of these criteria (Table 4).

  • Country circumstances were given careful consideration. For example, although Poland’s external debt appeared to be relatively high, detailed analysis suggested that debt sustainability was robust to stress scenarios. Moreover, the current account deficit was financed to a large extent by FDI and transfers of EU funds, helping to support a sustainable external position. For Colombia, the relatively low sovereign rating was compensated by the fact that Colombia’s bond spread was comparable to those with higher ratings and it continued to have market access on reasonable terms during the recent global crisis.

35. In the only PCL arrangement, Macedonia, ex post conditionality focused on the vulnerabilities identified during the qualification process.18 Moderate vulnerabilities in the external sector and data quality were identified as two areas that required improvement. Indicative targets on the overall government deficit and net international reserves were set to safeguard the external position, and the authorities committed to improving data quality. Macedonia drew on the PCL in March 2011, after unforeseen political shocks. The first review was completed in September 2011, and staffs assessment at the time the PCL was approved— that Macedonia performed strongly in three of the five PCL qualification areas and moderately underperformed in the remaining two—remained unchanged.19

36. In their requests for an arrangement, the member countries described clearly their broad policy goals and strategies during the proposed FCL and PCL arrangements. They explained how they would remain committed to strong economic policies and respond appropriately to any shocks that may arise. Although these written communications are not a Letter of Intent or Memorandum of Economic and Financial Policies for the purpose of monitoring policy commitments, they provide further assurances from the authorities that they are fully committed to implement appropriate policies.

37. The Executive Board had early involvement in the qualification process. The Board was promptly consulted in an informal Board meeting, as soon as interest in obtaining financing under the FCL (PCL) was confidentially communicated to the Fund and Fund management decided that access to Fund resources under the FCL (PCL) might be appropriate. A concise staff note, which included a preliminary assessment of qualification, was circulated to the Board before the informal meeting. A formal Board meeting followed at a later point to consider and approve an FCL (PCL) arrangement based on the authorities’ written request and the staff report, which included a rigorous assessment of qualification. In addition, the Board was also consulted in advance of the formal discussion on Macedonia’s request for a PCL (PLL), consistent with the Fund’s exceptional access framework.

38. There are not enough observations to prove or disprove that ex ante conditionality helped reduce stigma. The introduction of the FCL and its subsequent enhancements were generally recognized as a significant improvement to the Fund’s toolkit. The introduction of the PCL was also considered a useful tool for countries with sound policies but not qualified for the FCL. For the countries that requested FCL arrangements, the authorities probably would not have otherwise approached the Fund for financing in the absence of the FCL, for which no ex post conditionality is part of the attractive features. The limited demand for FCL during the recent crisis may, to some extent, reflect positive “externalities” from the introduction of the instrument—markets may have perceived that strong emerging market performers would qualify for the FCL, since spreads narrowed for a set of countries well beyond the three members using the instrument (Figure 1).

Figure 1.
Figure 1.

Emerging Market Spreads, 2008-20111

Citation: Policy Papers 2012, 046; 10.5089/9781498340397.007.A001

Source: Bloomberg.1/ Based on investment bank reports.EMBI - Emerging Market Bond Index.
Table 4.

Comparisons of FCL Arrangements 1

article image
Sources: WEO, staff reports, and staff calculations.

All calculations are based on the 2009 April WEO or the latest available information in April 2009, when qualification for the FCL was first assessed for Colombia, Mexico, and Poland. Subsequent reviews and successor arrangements all reaffirmed the initialassessments.

See Assessing Reserve Adequacy (IMF 2011) for more details.

Appendix 4: The 2009 SDR Allocation and Program Design

The analysis in this appendix concludes that the magnitude of the 2009 SDR allocation was significant, particularly for LICs. Many program countries used much of the allocation to build buffers in the form of reserves rather than using the policy space created by the allocation for additional absorption.

39. The 2009 SDR allocation was very large relative to a variety of metrics.20 The allocation resulted in a more than ten-fold increase in SDR holdings worldwide. It was equivalent to about half of all Fund disbursements in 2009-10 and amounted to approximately 1.6 percent of GDP in the median program country. Given their larger quota sizes, advanced economies received the bulk of the SDR allocation, accounting for 62 percent of the total (Figures 2 and 3). By contrast, when measured against the sizes of their economies, the allocation was largest for LICs, followed by emerging markets (EMs). The allocation contributed to a significant increase in reserve coverage across all groups of member countries. This effect was particularly large for LICs where reserves increased by nearly one month of imports. Moreover, reserve coverage generally increased in most EMs and LICs at the height of the crisis, suggesting that at least part of the allocation was saved (Figure 4). The allocation had also an important impact on the currency composition of countries’ reserves and reserve management decisions21.

Figure 2.
Figure 2.

The 2009 SDR Allocation 1(in percent of total)

Citation: Policy Papers 2012, 046; 10.5089/9781498340397.007.A001

1Based on PRGT eligibility criteria adopted in January 2010.
Figure 3.
Figure 3.

The 2009 SDR Allocation, GDP and Reserves

Citation: Policy Papers 2012, 046; 10.5089/9781498340397.007.A001

Source: Staffs calculations.
Figure 4.
Figure 4.

SDR Holdings since the 2009 Allocation

Citation: Policy Papers 2012, 046; 10.5089/9781498340397.007.A001

Source: Staff calculations.

40. It is difficult to assess the extent to which the SDR allocation was used to build reserves or to smooth policy adjustment during the crisis. Since reserve holdings are fungible, whether or not a country’s SDR holdings changed does not necessarily imply a change in its macroeconomic policy stance. Conceptually, the macroeconomic impact of the allocation would be the difference between the level of reserves after the allocation and the level of reserves that would have been in place had the allocation not occurred; the latter counterfactual scenario is not directly observable. Moreover, the discussion in staff reports of the impact of the SDR allocation on the macroeconomic policy mix for Fund-supported programs was limited (Box 1).

How was the SDR Allocation Addressed in Program Documents?

There was only limited discussion of the SDR allocation in staff reports of program countries.1,2 The allocation was approved by the Board in July 2009, with the general and special allocations in August and September 2009, respectively. Almost half of program staff reports did not discuss the impact of the SDR allocation on 2009 targets, and nearly 70 percent did not discuss its impact on 2009 outcomes. The discussion was particularly sparse for program documents on emerging market countries. This outcome was not consistent with Fund guidance that staff reports should describe the implications of the SDR allocation for each country concerned. In those staff reports that did discuss the allocation, most indicated that the likely impact was to bolster reserves, whereas a few other cases mentioned the possibility of absorption or spending, or a combination of reserve accumulation and absorption/spending.

Very few program countries introduced adjustors to performance criteria (chart below). When the allocation was announced in April 2009, its timing was still uncertain, and country teams were advised that changes to program targets could be introduced via adjusters to the performance criteria. However, only 15 percent of programs had introduced adjustors by August 2009. In some cases, the lack of adjustors was explained by the fact that no program reports were issued between the announcement and approval of the allocation (April to July 2009) or no performance criteria were going to be affected; in the other cases, however, it is not clear to what extent the non-inclusion of an adjustor was a deliberate decision or simply a lack of understanding of the guidance.

Of countries where it is possible to compare pre- and post-allocation performance criteria for end-2009, most opted for accumulation of reserves, at least initially. Nearly 33 percent of program countries in August 2009 opted to increase NIR targets in line with the allocation, or kept them unchanged when the allocation was not captured in the program definition of reserves. In almost all of these cases, program technical memoranda of understanding (TMUs) unambiguously described the impact of the allocation on performance criteria. In a few countries, the net international reserves target was actually reduced after the allocation. This may be related to other developments in the country beyond the SDR allocation.

Nevertheless, according to the survey of Fund staff for 2011 Review of Conditionality, a large majority of staff believe that the authorities have broadly acted in line with staff recommendations on the use of SDR allocation.The most frequently cited uses of the allocation suggested by the authorities were reserve accumulation, debt restructuring and boosting capital spending. In a few instances the allocation was used to increase current spending or to restore current spending levels resulting from lower fiscal revenues.

1Two memos were circulated making the case for a General SDR Allocation and discussing the implications for program design: “The Case for a General SDR Allocation,” FO/Dis/09/65, Aril 15, 2009; “The SDR Allocation — Implications for Program Design,” Office Memorandum, April 24, 2009.2Includes countries that had a financial arrangement (including Exogenous Shocks Facility—High Access Component or Exogenous Shocks Facility—Rapid Access Component) or PSI in August 2009 or negotiated a new arrangement/PSI by March 2010.

41. Analysis of projections before and after the SDR allocation shows a mixed picture in terms of its impact on macroeconomic targets for 2009 (Table 5).22

  • One-quarter to one-third of program LICs and EMs saw substantial upward revisions to their reserve targets that may well reflect the impact of the allocation23 For these countries (especially in the case of LICs), the decision appears to have been partly motivated by reserve adequacy considerations, with over three-quarters of LICs having relatively weak reserve cushions prior to the allocation.24 At the same time, many programs included some degree of additional within-year fiscal stimulus, generally without a concurrent widening in the current account balance, due to other factors, such as higher export growth and larger-than-expected import compression (especially among program EMs, but also program LICs).25 This could suggest that the larger reserve cushion was seen as a safeguard against the risks of heightened balance of payments pressures that could arise from a looser fiscal policy, even if the fiscal loosening was not expected to have a significant external impact in the baseline.

  • Most program countries did not raise the end-2009 reserve target by the full amount of the allocation, allowing some additional absorption and/or spending in 2009. In the case of program EMs in this group, two-thirds recast their programs (as of October 2009) by further loosening their fiscal stance. This was not, however, accompanied by wider current account balances (in fact the median current account balance projection for 2009 was revised upward after the allocation, possibly driven by import compression). The additional within-year fiscal relaxation in EMs added another 1 percent of GDP on top of the projected median fiscal stimulus of 1.5 percent of GDP for 2009 relative to 2008. For program LICs in this group, almost nearly half envisaged an increase in both fiscal loosening and absorption with another one-third planning either additional fiscal loosening or absorption. Program LICs had projected a median fiscal stimulus of 1.1 percent of GDP for 2009 relative to 2008, adding another 0.3 percent of GDP in projected fiscal loosening after the allocation.

  • In terms of the outturn, the number of countries accumulating reserves in 2009 was higher than expected right after the allocation (Table 6). End-2009 reserves exceeded the April projection by at least the amount of the SDR allocation for nearly two-thirds of program EMs and LICs, including countries with broadly adequate reserves prior to the allocation. This suggests that in many cases the allocation was not fully needed for additional absorption through the balance of payments, notwithstanding the more relaxed fiscal stance.

  • The aggregate picture corroborates the finding that the SDR allocation significantly increased reserve levels for both EMs and LICs. Two-thirds of program EMs and LICs substantially increased their reserves in 2009 and 2010. This may reflect a better than expected balance of payments situation; with current account balances improving for most countries implying no additional absorption in 2010. Nevertheless, fiscal loosening appears to have continued well into 2010 despite the continued reserve accumulation.

42. The allocation supported the crisis response by providing an additional buffer although few countries used the policy space created by it for immediate absorption. By raising the level of reserves overall, the allocation allowed countries to pursue countercyclical policies while retaining a buffer against future shocks. Moreover, in a few cases, in particular several LICs (e.g. Chad, Cape Verde, Cameroon, Djibouti, Malawi), there was a significant drawdown in SDRs accompanied by an apparent higher level of absorption and spending than envisaged before allocation. Overall, the immediate impact on external global flows has been very limited.

Table 5.

Derived Impact of SDR Allocation Based on 2009 Target Projections 1

article image
Source: Staff calculations.

Comparison of April and October 2009 WEO vintages. Iceland is grouped with EM programcountries in the analysis.

Table 6.

Derived Impact of SDR Allocation Based on 2009 Outturn 1

article image
Source: Staff calculations.

Comparison of April and October 2009 WEO vintages. Iceland is grouped with EM programcountries in the analysis.

II. Shareholders and Stakeholders’ Views on Fund Conditionality

43. The 2011 Review of Conditionality (RoC) draws on staff’s proactive efforts to engage with both Fund shareholders and external stakeholders. Several outreach events were undertaken in preparation for the Review, with non-governmental organizations (NGOs), civil society organizations (CSOs), and other stakeholders. Early Board engagement was sought, with a discussion of the concept note for the Review.26 The Review was also informed by structured interviews of Fund Executive Directors and a survey of Fund staff (mission chiefs and resident representatives), program countries authorities, and donors.

44. This section informs on the methodology and main outcomes of staff’s engagement with shareholders and stakeholders. Appendix 5 summarizes the outreach events that took place in the early preparation stage of the Review. Appendix 6 presents the main conclusions from structured interviews with the Fund Executive Directors. Appendix 7 sums up the results of the survey of country authorities, donors, and Fund staff.

Appendix 5: Summary of Outreach to Civil Society Organizations (CSOs) and Other Stakeholders

45. In preparation for the 2011 RoC, Fund staff undertook several outreach events with NGOs, CSOs, and other stakeholders, in order to gather input and reactions regarding the main themes for the review. This annex summarizes the outcome of the discussions.

46. Three outreach events were organized. The main themes were presented and discussed at the CSO Forum during the 2011 Fund-World Bank Spring Meetings. An outreach mission visited a number of stakeholders in Europe during April 28-May 5.27 A web-based public consultation process was started in March and ended on May 31.28

47. Counterparts welcomed the outreach initiative, and commented that the upcoming review covered the relevant issues. Some NGOs and CSOs however felt that outreach to populations in countries with Fund-supported programs had not been sufficient.

48. NGO submissions to the public consultation process focused on fiscal policy.

  • One contribution recognized that the design of fiscal policy in Fund-supported programs at the start of the global financial crisis had been appropriately supportive of growth. However, fiscal policy appeared to becoming tighter, with more constraints on spending. In that context, there was interest in an analysis of the quantitative design of Fund-supported programs, including through a study of the flexibility built into programs to account for unexpected developments (e.g., shortfalls in aid flows).

  • Comments widely highlighted the importance of supporting revenue mobilization in Fund-supported programs, as a means to ensure sustainable public spending financing. There was a concern that some revenue-raising measures embedded in programs focused on short-term gains in revenues while insufficiently addressing longer term issues such as tax evasion.

49. The streamlining of Fund conditionality was generally seen as positive. The parsimony of conditions was viewed as supporting more efficient program design, leading to greater ownership and enhancing coordination with other (multilateral and bilateral) donors with each institution focusing on its area of expertise. At the same time, streamlining raised questions related to the leverage of Fund conditionality. Some counterparts felt that some programs, while adequately designed to contain a global crisis, had left a large structural agenda unfinished. Structural reforms were seen as needed to enhance economic growth and thereby improve debt sustainability. While it was recognized that those reforms were not necessarily in the core area of Fund responsibility, it was felt that the Fund leverage had not been used fully in a few cases. The discontinuation of SPCs was seen by a few as reducing the leverage of the Fund.

50. Coordination between the Fund and other stakeholders was seen very positively, particularly in Europe.

  • The Vienna initiative was seen as a real success as the Fund had been critical in bringing clarity to the process, including in resolving coordination issues.

  • While the initial phase of the EU programs implied a process of learning, issues were quickly overcome.

  • Bilateral donors also expressed satisfaction, although some felt that there was room for more strategic coordination.

51. The increased focus on social issues in programs was appreciated, notably by NGOs and trade unions. Some, though, felt a disconnect between the policy announcements and the actions on the ground, which they saw as the result of potential contradictions between the need for large macroeconomic adjustments and enhanced social protection. In that context, NGOs and donors suggested that more be done ex ante to assess the impact of programs on social outcomes, including income distribution. One contribution suggested that in LICs the reduction of fuel or food subsidies was not timed adequately with the implementation of offsetting social protection measures.

52. Interlocutors observed that the FCL was valuable in providing confidence to the markets. They thought this effect largely outweighed risks, which had not yet materialized. The experience so far with the PCL was seen as inconclusive.

53. Fund outreach efforts in program countries were appreciated, and seen as supporting domestic ownership of programs. In LICs, some wished that the Fund could provide more frequent signals on the macroeconomic situation, in the context of infrequent semi-annual reviews. In emerging and advanced economies, some suggested that the Fund conditionality could be leveraged to foster data transparency.

54. Discussions highlighted some of the strengths and weaknesses of Fund-supported programs. On the positive side, for example, technical assistance was seen as indispensible and well integrated into the design of programs. On the negative side, some felt that fiscal, financial, and structural dimensions could be better integrated into program design.

Appendix 6: Structured Interviews with Fund Executive Directors

This note summarizes structured interviews with 21 Executive Directors, conducted during August-October 2011.29 The interviews complement the surveys of country authorities, donors, mission chiefs, and resident representatives, as well as the qualitative and quantitative analyses of macroeconomic and program data.

55. To what extent are the policies in programs consistent with program objectives (including fiscal, monetary, exchange rate, and financial sector and other structural reforms)?

Directors noted that proposed policies were generally consistent with program objectives. However, some Directors saw room for improvement in ensuring consistency between program goals and proposed financial sector reforms. In related comments, Directors expressed support for strengthening the clarity of program objectives in staff reports. Further, Directors thought that surveillance activities could be better geared towards determining what could be program objectives in potential future programs if economic conditions were to deteriorate.

56. To what extent do structural conditions in programs cover reforms that are critical for program objectives, design, and implementation? Was the pace of structural reforms appropriate?

Directors expressed a range of views. Some Directors thought that, possibly as a result of efforts to streamline conditionality and focus the Fund’s work on its core areas of expertise, a number of programs had left macro-critical structural issues unaddressed (e.g., reform of state-owned enterprises). Some of these Directors felt that as a result, the Fund had moved from being overly prescriptive on structural conditionality to the other end of the spectrum, now paying too little attention to it. In contrast, other Directors expressed concern about what they saw as a loss in recent years of programs’ traditional focus on macroeconomic and external stability.

57. To what extent has program design taken into account country circumstances, including administrative capacity?

Directors noted that staff had made valid efforts to take into account the specific circumstances of program countries. However, they observed that some programs had failed to consider the authorities’ limited administrative capacity possibly, resulting in weak program implementation. Accordingly, in countries with low administrative capacity, programs should focus on a limited number of objectives, complemented by capacity building and support for structural reforms from the World Bank and other agencies. Where appropriate, staff reports should highlight limited administrative capacity as a key risk to program implementation.

58. To what extent do Fund-supported programs have equivalent conditionality across countries with similar characteristics and qualifications?

Directors thought that programs were overall even-handed. However, some Directors mentioned the possibility that large countries and countries with a strong track record of program implementation may have benefited from lighter conditionality and greater access to Fund financing. In particular, some Directors thought that euro area countries had been given unusually high access to Fund resources. At the same time, Directors acknowledged that some of these countries were undergoing sharp macroeconomic adjustment.

59. In your view, what are the key criteria allowing to observe government ownership of the program? How would you characterize the ownership of Fund-supported programs in countries in your constituency or in general?

Directors agreed that defining and observing ownership is a challenge. They suggested relying on the following indicators: (i) track record of implementation of the current and of earlier programs, and historical economic performance; (ii) access to government officials (the higher the position, the more ownership); (iii) support from the parliamentary opposition and civil society. In related comments, Directors suggested that Fund staff could maximize ownership by collaborating closely with the authorities and by flexibly accommodating country characteristics and the authorities’ policy objectives.

60. To what extent do Fund-supported programs prove to be flexible to accommodate circumstances and events not under the control of the government?

While Directors agreed that programs should attempt to accommodate developments not under the authorities’ control to some extent, a number of Directors felt that the concept of flexibility had possibly been taken too far. Some of these Directors thought that some program types demanded too little conditionality at program start. Others thought that while many programs’ initial design seemed to include fairly demanding conditionality, substantial parts of this conditionality tended to be watered down at later stages through waivers. This group of Directors felt that it was difficult for them to distinguish between requests for waivers motivated by events not under the control of the government and those motivated merely by insufficient program implementation.

61. To what extent did the recent reforms to Fund conditionality (i.e. ex-ante conditionality in the FCL and more streamlining and focus of program conditions) help to overcome some authorities’ hesitations to seek Fund support?

Directors broadly concurred that the design of ex-ante conditionality in FCL programs had in some cases helped overcome hesitation to seek Fund support. They also thought that meeting the FCL eligibility criteria had helped countries demonstrate a solid track record that influenced market perceptions positively. At the same time, some Directors expressed concern that the reformed set of Fund instruments may segregate member countries into good performers and less successful ones. Some Directors also suggested that the transparency of the selection process for the FCL/PCL could be strengthened.

62. To what extent do the programs strike the right balance between fiscal adjustment and programmed financing?

Directors broadly concurred that the balance of adjustment and financing in programs was appropriate. A number of Directors warned, however, that the speed of adjustment (and the scope of structural reforms) needs to take into account political economy constraints. Other Directors suggested that excessively slow adjustment could jeopardize donor support.

63. To what extent do Fund-supported programs help catalyze additional financing from other financial institutions/donors and the private sector?

Many Directors saw Fund-supported programs as a powerful catalyst for market financing. Directors also noted that burden sharing between the Fund and other agencies and institutions had improved. In this context, some Directors highlighted the European Bank Coordination Initiative (“Vienna Initiative”) as an excellent approach to catalyzing public and private sector financing. By contrast, a few Directors noted a possible decline in the catalytic effect of programs, suggesting that donors rely on their own analysis and conditions to an increasing extent.

64. To what extent does the focus of Fund-supported programs on macroeconomic adjustment come at the expense of progress towards the country’s social policy objectives?

The majority of Directors thought that the Fund had succeeded in preserving social objectives in programs. They also noted that in recent years, greater outreach efforts had helped strengthen the dialogue with broader audiences, enhancing the understanding of Fund policies on social issues. Nevertheless, Directors called for better reflection of social policies in program documents, particularly as concerns issues related to income distribution and employment.

65. To what extent does Fund-supported program conditionality help to achieve growth and poverty reduction beyond the life of the program?

Directors concurred that well-designed and implemented conditionality will contribute to growth and poverty reduction beyond the duration of programs. Some Directors observed that the Fund’s capacity building role also contributes to achieving these objectives. Given the longer-term nature of the growth and poverty reduction, Directors suggested that following the completion of a program, surveillance could pay greater attention to how to achieve lasting poverty reduction. However, some Directors cautioned against a possibly excessive focus on poverty reduction and asked that the Fund remain focused on its traditional core competencies.

Appendix 7: Summary of Surveys with Country Authorities, Donors, and Fund Mission

Chiefs and Resident Representatives

A survey of stakeholders provided support for the many positive findings of the Review of Conditionality. However, the results should be interpreted with caution, in particular since the survey was conducted in May-June 2011, before the issues raised by later programs (especially in the euro area) were fully apparent. Moreover, specific divergences of opinion among Fund staff, donors, and country authorities on a few important issues highlighted the value of further communication efforts in support of Fund-supported programs.

66. A survey of the key stakeholders was conducted to complement other analysis in the Review of Conditionality. The survey helped to gain insights on Fund-supported programs from stakeholders comprising country authorities (Central Banks and Ministries of Finance), donors in program countries, Fund mission chiefs, and Fund resident representatives.30 The survey covered a wide range of issues from objectives of programs to program design and implementation. In particular, it looked at views on the consistency of program objectives with program strategies, program flexibility, evenhandedness, the reform of Fund conditionality, and coordination with key donors. It also sought views on the main characteristics of program ownership, the mix of financing and adjustment, and factors affecting program implementation. Given the multidimensional results of the survey, each paragraph of this note focuses on one or more thematic aspects of the Review, where: (i) the responses of groups of stakeholders had statistically significant differences; and (ii) the distribution of responses and stakeholders’ comments offered interesting perspectives.31, 32

67. The majority of all stakeholders (more than 90 percent) agreed that the program objectives and underlying strategies were consistent. The degree of agreement was particularly strong in Fund core areas, including monetary, exchange rate, and financial sector policies, but less so on other structural reforms. Based on comments, country authorities and donors perceived that programs were geared to promote macroeconomic stability and sustainable public finances, which were essential to promote economic growth. Authorities also thought that the consistency in program macroeconomic policies with the priorities of governments helped to sustain program implementation and economic performance. Despite broad agreement that quantitative macroeconomic conditionality (QPCs) were appropriate and consistent with program objectives, country authorities were less likely than mission chiefs and resident representatives to hold this view. Among the key reasons for disagreement cited by the respondents were frequent revisions to the QPCs and an unnecessarily large number of adjustors applied. The latter made the understanding and the implementation of targets more difficult.

68. The policies in Fund-supported programs were perceived by 94 percent of all respondents to be tailored to country circumstances and flexible. Fund staff commented that programs provided a sufficient range of policy options for successful implementation. Among the key reasons inhibiting program implementation, staff cited weak implementation capacity and unexpected developments/exogenous shocks (Figure 5). However, country authorities and donors perceived that Fund staff could have paid more attention to local capacity constraints.

Figure 5.
Figure 5.

Sources of Missed Conditionality

Citation: Policy Papers 2012, 046; 10.5089/9781498340397.007.A001

69. While country authorities generally agreed that the pace of structural reforms in programs was appropriate, they did so less often than Fund staff (Figure 6). There were statistically significant differences between the proportion of mission chiefs agreeing, on the one hand, and country authorities and donors on the other. Comments from the authorities and donors noted that some structural benchmarks were not specific enough. Some respondents thought that the timing for implementation of structural benchmarks was somewhat tight and did not take into account exogenous factors beyond the control of the government (e.g., parliamentary calendars).

Figure 6.
Figure 6.

Appropriateness of Structural Reforms

Citation: Policy Papers 2012, 046; 10.5089/9781498340397.007.A001

70. Nevertheless, reforms of Fund conditionality were perceived to have some benefits, without major negative implications. In particular, most respondents believed that the discontinuation of structural performance criteria (SPCs) did not diminish program implementation, monitoring, or the negotiating leverage of staff (Figure 7). While only about a third of staff respondents agreed that discontinuation of SPCs reduced stigma (Figure 7), more than three quarters of country authority respondents believed the reforms in conditionality have helped reduce stigma (Figure 8). A majority of staff did not perceive that conditionality reforms had strengthened ownership of structural conditionality. Staff also thought that the principle of parsimony had helped focus conditionality on macro-critical reforms.

Figure 7.
Figure 7.

Impact of the Discontinuation of SPCs

Citation: Policy Papers 2012, 046; 10.5089/9781498340397.007.A001

Figure 8.
Figure 8.

Stigma to Borrow From the Fund

Citation: Policy Papers 2012, 046; 10.5089/9781498340397.007.A001

Source: Country authorities.

71. Stakeholders’ perceptions of evenhandedness in Fund-supported programs were rarely negative, but country authorities and donors often felt unable to judge (Figure 9). Majorities of mission chiefs and resident representatives agreed that programs were even-handed (i.e., equivalent conditionality across countries with similar characteristics and qualifications), but around the half of the authorities and donors had no position on issue, and their comments focused on the lack of sufficient information on other countries’ programs.

Figure 9.
Figure 9.

Perceived Evenhandedness of Program Conditionality

Citation: Policy Papers 2012, 046; 10.5089/9781498340397.007.A001

72. The respondents perceived that the programs struck the right balance between adjustment and financing and played a catalytic role in financing (Figures 10 and 11). At least 90 percent of all groups of respondents agreed on the right balance, while a smaller majority of respondents believed that programs played a catalytic role in mobilizing private sector financing. The stakeholders also thought that Fund-supported programs involved extensive donor coordination (Figure 12). Among the key reasons cited for missed donor financing were failure to meet donor conditionality and slippages in programs.

Figure 10.
Figure 10.

Financing and Adjustment Balance 1

Citation: Policy Papers 2012, 046; 10.5089/9781498340397.007.A001

1Average of all respondents (donors, country authorities, and Fund staff.
Figure 11.
Figure 11.

The Catalytic Role of Fund-Supported Programs

Citation: Policy Papers 2012, 046; 10.5089/9781498340397.007.A001

Figure 12.
Figure 12.

Donor Coordination1

Citation: Policy Papers 2012, 046; 10.5089/9781498340397.007.A001

1Respondents include donors and Fund staff.

73. Stakeholders all agreed on some degree of ownership in Fund-supported programs, but to varying amounts (Figure 13). For instance, all respondents agreed that the authorities provided written input into the design of conditionality (drafts of the Memorandum of Economic and Financial Policies, or MEFP), but this feeling was more predominant for the authorities (with a statistically significant difference). All stakeholders generally perceived that civil society did not play an active role in authorities’ discussions of programs, although donors more often thought that CSOs were involved. At the same time, some donors also commented that in many circumstances governments were committed to implement reforms but were unable to because of a lack of political consensus.

Figure 13.
Figure 13.

Program Ownership

Citation: Policy Papers 2012, 046; 10.5089/9781498340397.007.A001

74. Concerning outcomes, stakeholders also had positive views on the impact of programs on growth and poverty reduction in LICs (Figure 14). However, a few country authorities and donors had no position on this; their comments focused on the lack of sufficient information and the fact that growth and poverty reduction go well beyond the timeframe of programs.

Figure 14.
Table 7.

Summary of Respondent Characteristics1

article image
Source: 2011 Review of Conditionality Survey.

The survey covers program countries including GRA and PRGT-supported programs, as well as PSI arrangements. The samples of mission chiefs and resident representatives are different, given that in a few instances there are no resident representatives in program country.

43 of 71 program country mission chiefs and 41 of 60 resident representatives responded to survey.

Authorities from 40 of 71 countries responded to the questionnaire. In 25 cases, only one institution (Ministry of Finance or Central Bank) responded. In 15 cases, both responded. The number of responding countries appears in parenthesis.

33 of 71 countries responded to the donor survey, while 22 countries had more than one donor response.

III. Qualitative Assessment of Conditionality and Program Design

75. Structural changes brought by individual conditions were systematically assessed. The impact of structural conditionality depends not only on numbers of conditions but also on the depth of those conditions, where depth is defined as the degree and durability of structural change that would likely occur if the condition was implemented. Appendix 8 presents the methodology for assessing the depth of structural conditionality, and provides detailed analysis of the depth of conditionality.

76. The Review also performed a systematic assessment of macro-social structural conditions. The conditions set between 2002 and 2011 on three sectors closely related to macro-social issues were analyzed and their likely impact assessed. The main conclusions are provided in Appendix 9.

77. The Review draws on existing assessments of completed programs. Conclusions from the recent ex-post assessments (EPAs) and ex-post evaluations (EPEs) were reviewed. The main takeaways are summarized in Appendix 10.

78. A set of representative countries was also used to perform qualitative analysis. Countries were selected to be representative of the whole set of program countries on the most recent part of the review period (2006-2011). The methodology for designing the representative sample is presented in Appendix 11.

Appendix 8: Depth of Structural Conditionality

79. This appendix provides an analysis of differences in degree and durability of structural change brought about by structural conditions. This degree defines the depth of structural change implied by each condition set and conveys a notion of durable structural change for each condition set. It is important to underscore that the notion of depth does not imply any assessment of (1) the difficulty of implementing conditionality on the part of authorities; or (2) the criticality of structural measures agreed upon in a program. Indeed, depending on country capacities or political circumstances, the same condition may be relatively easy or quite difficult to implement; similarly, depending on country circumstances, a condition with limited structural impact—such as identification of the main stages for a major reform—may be just as critical as a condition with expected long-lasting effects, such as the enactment of a legislation. The following elaborates further on the methodology, and presents the main results.


80. The assessment of structural depth comprises 2,226 program measures in 99 programs, which were reviewed by the Executive Board between 2006 and the cut-off date for this exercise of September 30, 2011. The exercise builds on the 2007 IEO Structural Conditionality Evaluation, which applied comparable categories.33 For the purpose of assessing the degree of structural change that a condition would likely bring about, this review distinguishes between program conditions along three categories (see Table 8 for examples):

Table 8.

Examples of Categories of Structural Depth

article image
Source: MONA.

While the adoption of a budget usually involves legislation, the budget for any given year would, under normal circumstances, not be expected to trigger lasting structural change.

  • 1. Low structural depth: the condition could be a stepping stone for significant reforms, but would by itself not trigger significant structural change. In the classification of this review, 47 percent of measures fall into this category.

  • 2. Medium structural depth: this category consists of one-off measures with immediate and possibly significant effect that need to be followed up or repeated in order to entail substantial structural change. 29 percent of conditions fall into this category.

  • 3. High structural depth: conditions in this category have, by themselves, the potential to bring about long-lasting change to the institutional environment. 24 percent of conditions are placed in this category.

81. In some cases, the classification of measures along the three categories of structural depth leaves room for discretion (see Box 2 on classification challenges). In those cases, where a condition was found to be on the borderline between two categories, this review tried to err on the conservative side, placing the measure in the lower of the two possible categories.34

Depth of Conditionality: Classification Challenges

The classification of the structural depth of conditionality aims to be comprehensive and exclusive, and accordingly, each measure has been assigned only one out of three categories. However, in some cases, the categorization involves judgment calls in grey zones. While the assigned categories are clear-cut for the large majority of measures, particular classification challenges arise from

  • Multi-step processes. This review assesses each measure as a singular event, irrespective of other, possibly connected conditions. The approach results in a downward bias in the assessment of structural depth, as each step in a chain of policies may have a low structural impact, while the sum of measures could be systemic.

  • Comparability among programs and countries. The structural depth of program conditions is context-specific, depending on the structural environment at the outset of a program, the political economy of any given program country, and the time period. Context specificity implies that measures that have a deep structural impact under one program may have a more limited impact under another one.

  • Odd conditions. Outliers include measures referring to the adherence to existing rules without implementing new policies, the abstention from introducing certain policies, and measures with a quantitative rather than a qualitative objective. These outliers (limited in number) were classified by adopting the classification of the closest possible proxy measure.


82. Key conclusions based on the categorization of structural depth are as follows:35

  • Structural depth increased over time during 2006-10 (Figure 15). For both PRGT and GRA programs, the share of measures falling into the category of high structural depth increased significantly. This trend was reversed slightly in the first nine months of 2011.

Figure 15.
Figure 15.

Structural Depth by Year

Citation: Policy Papers 2012, 046; 10.5089/9781498340397.007.A001

Source:MONA and staff calculations.
  • Structural depth by review was most pronounced at Board approval and in the later reviews for GRA programs, but does not show a clear trend line for PRGT programs (Figure 16).36

Figure 16.
Figure 16.

Structural Depth by Review

Citation: Policy Papers 2012, 046; 10.5089/9781498340397.007.A001

Source: MONA and staff calculations.
  • On average, structural assessment criteria (SACs) were structurally somewhat deeper than other categories of measures. (Figure 17).37

Figure 17.
Figure 17.

Structural Depth by Type of Measure

Citation: Policy Papers 2012, 046; 10.5089/9781498340397.007.A001

Source:MONA and staff calculations.
  • Structural depth varied among economic sectors (Figure 18). Conditionality related to labor market policies, a non-core area of Fund-supported programs, was limited, and most conditions fell into the lowest category of structural depth. By contrast, conditionality linked to public enterprise reform, pensions and other social sector reforms, other structural reforms, trade policy, and, in particular, exchange systems showed a high degree of structural depth.

Figure 18.
Figure 18.

Structural Depth by Sector

Citation: Policy Papers 2012, 046; 10.5089/9781498340397.007.A001

  • Overall, structural depth was more pronounced in GRA programs than in PRGT programs (Figure 19). Over time, both PRGT and GRA conditionality was marked by a trend toward higher structural depth.

Figure 19.
Figure 19.

Average Structural Depth by Arrangement Type Group

Citation: Policy Papers 2012, 046; 10.5089/9781498340397.007.A001

Source: MONA and staff calculations.
  • There was significant variance among countries and regions with respect to structural depth of conditionality (Figures 20). GRA programs with European countries had the highest structural depth. For a few countries, conditionality fell into only one or two of the three categories of structural depth, typically due to the overall low number of conditions.

Figure 20.
Figure 20.

Structural Depth by Country Grouping

Citation: Policy Papers 2012, 046; 10.5089/9781498340397.007.A001

Source: MONA and staff calculations.
  • The analysis of structural depth by country grouping (or income level) partly corresponds to the analysis by region (Figure 21 and 22). On average, programs with advanced market economies have been structurally deeper than those with emerging markets. While programs with LICs accounted for the largest number of conditions, structural depth in these programs was lower than for the other two country groupings.38

Figure 21.
Figure 21.

Structural Depth of GRA Programs

Citation: Policy Papers 2012, 046; 10.5089/9781498340397.007.A001

Source: MONA and staff calculations.
Figure 22.
Figure 22.

Structural Depth of PRGT Programs

Citation: Policy Papers 2012, 046; 10.5089/9781498340397.007.A001

Source: MONA and staff calculations.
  • The structural depth of recent Euro area programs exceeds the average of structural depth for other programs (Appendix Figures 23 and 24). This primarily reflects the Greece program. Programs with Ireland and Portugal are marked by lower structural depth.

Figure 23.
Figure 23.

Structural Depth in Euro Area Programs

Citation: Policy Papers 2012, 046; 10.5089/9781498340397.007.A001

Source: MONA and staff calculations.
Figure 24.
Figure 24.

Structural Depth in the Greece, European High Access, and SBA Programs

Citation: Policy Papers 2012, 046; 10.5089/9781498340397.007.A001

Source: MONAandstaff calculations.
  • Conditions that were either not met or waived were modestly deeper than those that were met or met with delay (Figure 25). This confirms some implementation challenges to policy makers from deeper structural conditions.

Figure 25.

Structural Depth by Implementation Status

Citation: Policy Papers 2012, 046; 10.5089/9781498340397.007.A001

Source: MONA and staff calculations. 1 One observation is classified as Modified/Met, and another is unclassified.

    Appendix 9. Macro-Social Structural Conditionality39

    In line with the call for parsimony and focus on macro-critical conditions in Fund-supported programs, structural conditionality directly related to macro-social issues represents a small share of all structural conditions. This share declined during 2005-2009, but increased during the global financial crisis. The decline in the number of labor market related measures since the mid 2000s represents a shift away from measures that are more likely to have higher short-term costs.

    83. Macro-social structural conditions amount to a minority of all structural conditions, peaking at 10 percent in 2006 and declining to 5 percent in 2009 (Figure 26). Most recently they account for about 8 percent of all structural conditions in 2011. Pension and other social sector reforms, and civil service and public employment reforms are more common, with fewer structural conditions on labor markets beyond the public sector.

    Figure 26.
    Figure 26.

    Macro-Social Structural Conditions (All programs, share of total structural conditions)

    Citation: Policy Papers 2012, 046; 10.5089/9781498340397.007.A001

    Source: MONA and staff calculations.

    84. GRA programs have typically included more macro-social conditions, with the relatively high frequency of such measures in the mid-2000s reflecting specific issues in former transition economies programs. Conditions in GRA programs have a greater relative focus on pension and social sector, and labor market reforms (Figure 27). By contrast and also reflecting labor market characteristics in these countries, civil service reforms are relatively common in PRGT programs, where there is virtually no conditionality related to pensions or labor markets outside public employment.

    Figure 27.
    Figure 27.

    Macro-Social Conditions, by Type of Fund-Supported Program

    (Share of total structural conditions)

    Citation: Policy Papers 2012, 046; 10.5089/9781498340397.007.A001

    Source: MONA and staff calculations.

    85. Several programs include measures aimed at lowering fiscal expenditure in the short-term and/or improving productivity through structural reforms in labor markets.

    While these measures contribute to resolving balance of payments difficulties and are likely to result in longer term gains, they may also have short-term labor market costs. A closer look at labor market related structural conditions suggests that:

    • Many labor market conditions are likely to have short-term costs. A qualitative assessment of structural conditions related to labor markets excluding public sector employment suggest that about 75 percent of these measures are likely to result in higher unemployment, lower wages, or lower worker bargaining power in the short-term. The decline in the number of labor market measures since the mid 2000s therefore represents a shift away from measures that are more likely to have higher short-term costs.

    • Conditions often involve social partners and have longer term benefits. Many labor market measures involve social partners in the decisions-making process (e.g., requiring social partner agreement to increase work time flexibility in Bulgaria or involving social partners directly in the “Committee to Promote Wage Restraint” in Latvia). In addition, many of the labor market measures are likely to improve the functioning of labor markets and make them more inclusive over the longer term: for example, structural measures aimed increasing working time flexibility and opening up regulated professions are likely to allow broader labor market participation.

    • The recent increase in GRA programs is driven by conditionality in a couple of programs.40 Four structural conditions in the Greece and Portugal programs explain all of the increase in labor market (excluding the public sector) conditions in 2011. These measures are aimed at increasing differentiation in wage setting and opening up regulated professions in Greece and reducing severance costs and labor taxes in Portugal.41 Greater focus on structural conditions in these programs reflects both significant labor and product market rigidities and the need to address competitiveness issues through domestic measures (in the absence of exchange rate flexibility).

    86. In addition to direct labor market measures, civil service reforms are likely to matter for macro-social outcomes. Out of all structural conditions set in relation to public employment, some measures (about one quarter) are likely to result in lower public employment or lower wages in the short-term. Many of these measures were included in programs in former transition economies with large public sectors. Others may affect macro-social outcomes through enhanced productivity of the public sector, for instance, measures that aim to initiate broader public sector reform (with no explicit goal for reducing/limiting public sector employment or wages) or focus on improving governance and transparency (for example through public employment censuses). These measures are more common in PRGT programs.

    87. The share of pension and other social sector reforms, aimed at improving the safety net, has also increased over the last two years in both GRA and PRGT programs. Over the period, almost half of these conditions aimed at improving the social safety net either through higher social spending or better targeting of benefits. In some cases, conditions aim explicitly at mitigating the impact of higher energy prices (in the context of energy subsidy reform) on the poor. In addition to conditions that explicitly focus on higher benefits or better targeting, several structural conditions aim at broader social sector reform.

    Appendix 10: Assessment of Conditionality in EPAs and EPEs

    Fund staff have analyzed program conditionality in the context of Ex Post Assessments of Longer-Term Program Engagement (EPAs) and Ex Post Evaluations of Exceptional Access (EPEs). This section reflects the assessment of program design and conditionality in 13 EPAs, eight EPEs, and two combined exercises conducted during 2007-March 2012. 42

    EPAs and EPEs: Summary of Lessons Learned

    • Appropriate diagnostics of the underlying challenges will enhance the design of program conditionality, which needs to take into consideration institutional weaknesses, the political economy, and the medium-to-long term reform agenda.

    • Conditionality needs to address follow-up reforms.

    • Programs should aim to address data weaknesses in order to facilitate the monitoring of conditionality.

    • Flexibility in the design of conditionality in response to changing circumstances and/or implementation challenges can enhance ownership and may be critical for the success of a program.

    • Ownership, including at the highest political levels, is key to program success.

    • EPAs/EPEs note a positive correlation between parsimony of program conditionality and ownership by country authorities, since programs relying on overly complex conditionality can be derailed by limited implementation capacity.

    • Parsimony of Fund-supported programs can be supported by coordination with other international organizations that apply complementary conditions in their areas of expertise.

    • Collaboration with other international financial institutions (IFIs) and with donors, building on a clear delineation of responsibilities, is instrumental for program success and reduces the risk of stretching institutional capacities.

    • Technical assistance (TA) plays a critical role in assisting the authorities with the implementation of structural conditionality.

    Design of Conditionality

    88. Diversity in programs covered by EPAs and EPEs is reflected in a wide range of program objectives and strategies, including reinforcing macroeconomic and financial stability, strengthening public finances, enhancing growth and competitiveness, and achieving or maintaining debt sustainability. For LICs, re-orienting expenditure to poverty reduction was an additional important objective. In some of the transition economies as well as in LICs, program objectives comprised streamlining government operations and reforming the civil service; creating a business-friendly environment; and addressing corruption and governance issues. Since exceptional access programs aimed to address the most pressing vulnerabilities, macroeconomic policy adjustment and structural reforms under these programs were designed to achieve better resilience to external shocks and the restoration of market confidence in an adverse global environment.

    89. Many EPAs and EPEs conclude that the programs under review were characterized by broadly appropriate design of conditionality.43 For example,

    • The El Salvador EPE notes that conditionality included the appropriate crisis prevention measures.

    • In the programs with Georgia, conditionality focused on accommodating a strong fiscal countercyclical response, protecting international reserves, and enhancing the banking sector’s ability to mitigate potential vulnerabilities.

    • The EPE for Hungary finds that fiscal policy balanced fiscal sustainability concerns with the need to avoid a sharper downturn.

    • The Iceland EPE notes that the program struck an appropriate balance between short-and medium-term objectives, focusing on restoring early confidence in the króna and putting public finances on a sustainable path.

    • The Kenya EPA concludes that policy design and implementation were generally sound, with governance being a central point of Fund engagement in a sequence of programs.

    • The Kyrgyz Republic and Mali EPAs note that program design and conditionality supported program objectives, and exogenous risks to program implementation were identified correctly.

    • The Niger EPA notes the realistic, gradualist approach to program design in a fragile context, reflecting the lessons from the 2004 EPA and an appropriate ex-ante recognition of risks.

    • Reflecting the view that emergency programs should focus on preserving stability, conditionality in Serbia’s SBA centered on safeguarding the financial system and ensuring fiscal consolidation.

    • The Romania EPE finds that program conditionality appropriately aimed to address the areas of greatest vulnerability, in order to restore economic stability and investor confidence.

    • Financial sector policies in the SBA with Guatemala focused appropriately on correcting weaknesses in the banking system’s legal and regulatory framework.

    90. However, some programs were affected by weaknesses in the design of conditionality. In the programs with Benin, many structural conditions focused on preparatory steps for broader reforms rather than on the implementation of concrete actions, also reflecting the expectation that the World Bank would take the lead in their areas of expertise. The EPA for Sierra Leone diagnoses room to better align the reform agenda with program objectives, and to sequence reforms more consistently in order to overcome capacity constraints. In Mauritania, better targeted strategies for social indicators would have been useful. While conditionality under Burundi’s first PRGF focused on addressing macro-critical structural gaps, it also comprised elements of coffee sector reforms, going beyond the Fund’s core area of responsibility. The effectiveness of this conditionality was hampered by limited alignment between the Fund and World Bank on the appropriate pace of reforms.

    91. Some evaluations conclude that critical conditionality did not form part of the program. In Benin, the absence of conditions related to the financial sector is explained by arrangements in the WAEMU and the expectation that an FSAP would take place. The Georgia EPA notes that programs in later years could have been more forceful in addressing risks associated with banking sector vulnerabilities and overheating. Although both programs with the Kyrgyz Republic emphasized fiscal transparency as a means to strengthen public sector governance, strategies to address corruption did not translate into tangible and decisive measures. The Mali EPA questions the appropriateness of the design of conditionality targeted at addressing key problems in the cotton sector. In the SBA with Serbia, the macro-criticality of enterprise restructuring and privatization should have been reflected in conditionality, with a clearer definition of measures to provide a better sense of direction. For Ukraine, the EPE notes room to increase the number of structural conditions in the fiscal area, including prior actions, in order to enhance program success. The Guatemala SBA could have been strengthened by attaching conditionality to revenue-generating tax reform, as a commitment to medium-term fiscal solvency. While overall positive in its assessment of program design and conditionality, the Niger EPA notes that financial sector reforms could have been better-tailored to the country’s specific circumstances.

    92. Inappropriate diagnostics of underlying problems reduced the effectiveness of conditionality in some programs. In the Belarus SBA, underestimation of the macroeconomic importance of lending under government programs (LGP) resulted in conditionality that was easily circumvented by the authorities who adhered to the letter rather than the spirit of the program. The EPE notes that direct quantitative ceilings on LGP and central bank non-market-based lending to banks could have been more effective than conditionality on government deposits. With respect to exchange rate policy, the EPE finds that a larger upfront devaluation would have been meaningful. In Ghana, the macro-relevance of public sector reform became clear and was reflected in conditionality only toward the end of the 2003-06 program. A stronger focus on improving fiscal institutions at an earlier stage could have prevented some of the most critical slippages. In Ukraine, political resistance and institutional capacity constraints triggered compromises based on limited and qualitatively uneven diagnostics. The EPE suggests, in particular, more selectivity in the recapitalization of commercial banks. In the Kyrgyz Republic, governance concerns and weaknesses in domestic financial sector supervision were not identified as sources of program risk at the outset of the program.

    93. Some EPAs diagnose a lack of appropriate follow-up conditionality. For example, the EPA for Mauritania notes that programs lacked a strategy for the completion of earlier interventions. Some structural reforms, such as public enterprise restructuring, were expected to continue, but were not covered by formal conditionality. In response to misreporting, proper corrective actions were identified but strong follow-up actions were lacking. Rather than addressing the underlying problems, formal conditionality relied excessively on measures with an immediate impact. In the Gambia, following extensive up-front conditionality at the outset of the program, subsequent weaknesses in revenue mobilization were not addressed through appropriate corrective measures. The Mali EPA finds that in some cases, reform objectives appear to have been abandoned without obvious reason. In Nicaragua, mixed implementation of the reform program reflected the focus of conditionality on relatively short-term measures. While a short-term focus had been appropriate to address short-term risks, this should not come at the expense of the medium-term perspective.

    94. The design of fiscal sector conditionality could have been enhanced in some programs. In Hungary, a more comprehensive set of fiscal targets would have been needed for effective program monitoring. In the Mongolia SBA, the necessary emphasis on the underlying fiscal effort could have been addressed by quantitative performance criteria (QPCs) on the primary deficit or the non-mineral deficit. The Gambia EPA notes that program targets for each fiscal year, in line with the budget cycle, would have been more appropriate than cumulative targets from the beginning of the program, in order to facilitate monitoring by the authorities and avoid carry-over of one-time slippages. In the El Salvador program, a cyclically-adjusted measure could have helped to address the significant growth surprise. The EPE notes, however, that the computation of cyclically-adjusted QPC could have proven difficult in a LIC context.

    Program Implementation

    95. Conditionality played a critical role in program implementation. The Serbia EPE finds that the SBA focused appropriately on fiscal adjustment, achieving its targets and contributing to the primary objective of avoiding a financial meltdown. In Ukraine, despite complex political challenges, the SBA also helped to prevent a financial meltdown and contributed to re-establishing confidence in the banking system. The Burundi EPA finds that Fund involvement played a key role in economic and social stabilization, also paving the ground for structural reforms. In Mongolia, a cautious approach to macroeconomic projections was found to have supported key program objectives by avoiding significant deviations between targets and outturns. In Sierra Leone, the program achieved its main objectives by providing a sound macroeconomic framework that allowed the economy to build up over time adequate resilience to external shocks.

    96. Many evaluations conclude that performance against QPCs was superior to implementation of the structural reform agenda, which suffered frequent delays in both LICs and more advanced economies. In Serbia, quantitative performance under the SBA was broadly satisfactory; however, many structural benchmarks were met only partially or with delays. In Georgia, QPCs were almost always met, in many cases by wide margins, while various structural measures were implemented with delay. The EPA/EPE for Turkey finds that structural conditions were often met with significant delay, and that staff dealt appropriately with these difficulties by delaying reviews in order to maintain engagement. The Ghana EPA notes repeated delays in structural reform, particularly in the energy and public sectors. While QPCs were generally met in Mali, the observance of structural conditionality was only partially successful in addressing utility sector problems and the challenges associated with raising growth. In Mauritania, macroeconomic stability had been secured, but higher and sustainable growth remained elusive, and the impact of the programs on social indicators remained disappointing with limited progress in structural areas.

    97. The design and monitoring of conditionality was sometimes impaired by data weaknesses. The Belarus EPE, in particular, mentions that the scale of the lending-under-government-programs had not been realized adequately due to lack of comprehensive data, resulting in weak conditionality. The Mauretania EPA notes that formal compliance with program conditionality does not necessarily indicate achievement of the medium- and long-term objectives of the programs. Taking into account this limitation, the overall structural reform record was less than impressive.


    98. Several programs demonstrated a high degree of flexibility in response to external shocks. Programs with Mauritania were initially based on ambitious macroeconomic targets, which had to be adjusted to more realistic levels in response to external shocks. In Georgia, the modification of some measures mostly reflected progress toward program objectives and new macroeconomic challenges. In Hungary, the initial expectation of a short-lived liquidity crisis translated into a relatively short program with large frontloaded access and streamlined conditionality. As the onset of a deeper global crisis became evident, program targets and conditionality had to be adjusted. In the programs with Burundi, fiscal flexibility facilitated the targeting and full use of donor aid, which, however, came at the expense of fiscal consolidation and debt sustainability. In Nicaragua, flexibility in adjusting conditionality to domestic constraints enabled the authorities to meet program targets when the “first best” outcome could not be achieved. The Romania EPE highlights the flexibility exercised in response to the deeper-than-expected downturn, which resulted in revised fiscal deficit targets, but did not concede the overall program objectives. Under the program with Iceland, fiscal targets were eased appropriately, intended to achieve a balance between securing sustainability, supporting the economy, and preserving social objectives.

    99. Flexibility was also essential to accommodate an uneven implementation record and institutional constraints. In the Serbia SBA, staff addressed delays in implementing critical fiscal measures by introducing PAs that would ensure compliance with structural conditionality. In Sierra Leone, extensive structural targets in the initial program were reduced in the course of the program, changing the focus of the structural agenda to reflect the evolution of the government’s reform agenda. Under the Hungary SBA, the structural benchmark on the bank resolution regime was changed to address legal issues involving shareholders’ rights and the protection of property rights under the constitution.

    100. In several programs, flexibility played a central role in meeting conditionality. In the case of Mongolia, the EPE noted flexible adaptation throughout the program to accommodate economic developments and political gestation periods. While focused on core policy measures, principled flexibility had helped to enhance ownership. The Serbia EPE notes that the success of the 2009 SBA could be attributed in part to the adaption of the program to the prevailing circumstances. In view of a worse-than-projected macroeconomic environment and possible social implications, the relaxation of conditionality and a more generous timeline were considered appropriate. In Burundi, the built-in flexibility of the program was critical for success in a fragile and risky environment, allowing for second-best rather than orthodox solutions in a challenging transition.

    101. A few programs were characterized by constant revisions to conditionality.

    • In the case of Belarus, substantive program changes on an ongoing basis reflected the need to adjust conditionality related to the exchange rate regime and LGP, which was initially less than effective. However, despite continuous program discussions, problems related to LGP persisted. The EPE notes that tighter conditionality from the outset could have helped render this program more robust.

    • The Ukraine SBA demonstrated a large degree of flexibility, in particular in the fiscal area, in view of the exceptional degree of macroeconomic uncertainty. Continuous changes in conditionality with increasing reliance on PAs aimed to address both changes in priorities and the attempt to ensure ownership in a polarized political environment. However, the continuous urgency of program discussions made it difficult to assess whether conditionality was implemented in a durable manner.

    • In response to uneven ownership and program implementation, the programs with Kenya underwent frequent discussions. Added, extended, and modified measures in response to incomplete program implementation and new developments triggered the perception on the side of the authorities of shifting goalposts and micromanaging reforms. Conditionality in the area of governance sometimes became reactive, and some measures appeared to have stretched beyond core Fund expertise.


    102. EPAs and EPEs stress the critical role of ownership and its interdependence with conditionality. In Georgia, the authorities’ strong ownership of the reform agenda was considered essential for program success. The Mongolia EPE notes that ownership benefitted from focused conditionality, in combination with flexible implementation and extensive outreach efforts. In Mali, the improved observance of conditionality under the recent program was found to reflect greater ownership, supported by more effective consensus-building outreach efforts, as well as the streamlining of conditionality. For Benin, the EPA notes the usefulness of front loading of measures, through the use of PAs, as a signal of the authorities’ commitment. In later programs with Kenya, a more pragmatic approach from the Fund was mirrored by enhanced ownership of the structural and institutional reforms. The Serbia EPE observes that the interdependence between conditionality and ownership implies that programs can serve as a coordination device for the authorities, and that a correctly specified incentive structure helps enhance ownership. The Iceland EPE attributes the success of the program to strong ownership by the authorities, who were determined to preserve the country’s social model in spite of fiscal adjustment. In Romania, strong ownership—in combination with large financing, flexible program design, and appropriate reform prioritization—was central to achieving program objectives. The authorities demonstrated ownership by taking difficult and sometimes unpopular decisions in order to restore medium-term sustainability.

    103. At the other end of the spectrum, the reports note a lack of ownership in several programs. Greater recognition of ownership issues at the highest levels in Belarus would have entailed a different program design in critical areas, such as stricter conditionality on LGPs and exchange rate policy. In the absence of such high-level endorsement, greater reliance on PAs could have enhanced program success. In Ukraine, success was reduced by weak ownership despite many PAs. However, in a complex political environment, conditionality often reflected trade-offs between the conflicting attempts to ensure success of the program and to obtain political support. Weak ownership, in combination with governance issues rendered conditionality in support of necessary structural reforms less than successful the Kyrgyz Republic. Beyond unforeseen external shocks, the lack of program performance was attributed to ownership erosion also in the Gambia. Despite strong overall ownership of the reform agenda on the part of the Guatemalan government, political resistance thwarted the approval of critical banking legislation and regulations.

    104. Lack of ownership was sometimes exacerbated by the complexity of measures and capacity constraints in public administration. In Sierra Leone, slow implementation of structural reforms reflected both capacity constraints and limited ownership in implementing some of the “stronger” structural measures. The EPA finds that a more prioritized approach in defining the reform agenda could have been useful.

    105. Interdependence between ownership and political agendas was strong. As demonstrated by the Mauritania programs, conditionality was less effective in the face of a deteriorating political climate. In Burundi, limited consideration of political and capacity constraints reduced the usefulness of conditionality. The EPE for El Salvador notes that the prescribed timing of key financial sector legislation should have been better aligned with political realities. In Ukraine, implementation risk was acknowledged from the outset, but could have been addressed more prominently in view of complex political realities. The Kyrgyz Republic EPA suggests structural conditionality consider national development strategies and implementation capacity. The Nicaragua EPA stresses the importance of taking into account legal and institutional constraints, while targeting a realistic timeline.

    106. Coverage of ownership in EPAs and EPEs is uneven and does not seem to be based on a clear-cut definition. While some EPAs and EPEs discuss ownership extensively (see above), others only address the need for enhanced ownership in future programs. In several reports “lack of ownership” seems to correspond to the residual explaining a low degree of program implementation, at the expense of analysis of other possible explanations.


    107. The emphasis on streamlining and the macro-criticality of measures were reflected in increasingly parsimonious program conditionality. In Georgia, the success of reforms and the authorities’ credibility provided a basis for limiting structural conditionality. In Mongolia, structural conditionality was also significantly more focused than in earlier arrangements, with all measures being well aligned with program objectives and the Fund’s core expertise. Streamlined and flexible conditionality had facilitated negotiations and strengthened ownership of the program, despite unexpected developments in the global economic environment. The Ghana EPA notes that streamlined conditionality and more limited PAs in the 2003-2006 PRGF resulted in fewer waivers than in earlier programs. In Burundi, structural conditionality was lighter than in other comparable programs, with a focus on macro-critical PFM measures with strong positive spillover potential into other policy areas. The trend toward more parsimonious use of structural conditionality was also noticeable in the programs with El Salvador and the Gambia. In the Serbia SBA, conditionality remained overall light. The Mauritania EPA diagnoses excessive streamlining of structural conditionality in the 2006 PRGF, in particular in the safeguards area.

    108. Parsimony was supported by the involvement of the World Bank and other organizations in key structural areas. The Kyrgyz Republic EPA finds that conditionality under the 2008 ESF was more streamlined than under earlier programs, allowing staff to focus on macroeconomic policy. Parsimony was supported by a clear delineation of responsibilities between the Fund and the World Bank, reducing the risk of overstretching capacity. In Hungary, fiscal conditionality was streamlined in view of the authorities’ ongoing efforts and commitments to the EU targets. In Burundi, the second PRGF was characterized by considerably lighter conditionality than the first one, focusing on areas of Fund core expertise and supported by technical assistance (TA).

    109. The trend toward parsimony was occasionally reversed. In the later stages of the programs with Kenya, the track record of structural measures improved, reflecting the Fund’s focus on streamlining and macro-criticality, facilitated by World Bank involvement in governance. However, following the earlier trend towards parsimony, the Fund reverted to a broader governance agenda in 2005-06, invoking accusations of shifting goalposts when measures were extended or modified due to incomplete action or in response to a changing environment. Although anti-corruption conditionality was often justified by the potential impact of specific measures on donor financing, the extensive reliance on PAs to test ownership faced capacity constraints and reduced transparency. While conditionality was originally streamlined in the Ukraine program, reviews relied on an increasing number of PAs in an attempt to enhance implementation.

    110. EPAs and EPEs caution against a simplistic concept of parsimony. In the Turkey SBA, the number of conditions was reduced sharply compared to earlier programs, but structural conditionality nevertheless exceeded the average for programs with exceptional access. However, most conditions fell under the Fund’s core areas of expertise, and the relatively large number of conditions reflected the need to break down broad measures into smaller tasks, in order to make them easier to operationalize. The Burundi EPA notes that an environment characterized by low institutional capacity requires technically simple reforms with potentially large spillovers, where visible impact should take priority. The EPA cautions against a simplistic approach to parsimony, since an overly strong focus on the number of measures could distract from the need to break down complex measures into a larger number of more limited measures that allow for easy monitoring and adequate sequencing. The Nicaragua EPA stresses the need for flexibility and parsimony to be balanced with a firm stance in areas where significant deviations from the reform agenda result from political constraints and a lack of determination. However, the definition of intermediate conditionality should aim to avoid the perception of micromanaging the reform process.

    Coordination with other IFIs and Donors

    111. Collaboration with the World Bank, other IFIs, and donors remains essential. In the Nicaragua PRGF arrangement, conditionality focused on areas close to the core areas of responsibility of the Fund, while the World Bank and the IADB took the lead in the other main reform areas. The narrowing of conditionality and the division of labor was considered appropriate, preventing the Fund from spreading its resources and attention too thin. The EPAs and EPEs for Ghana, Mongolia, and Sierra Leone also stress the need for coordination with the World Bank and other institutions in macro-critical structural areas beyond the expertise of the Fund, such as decentralization and energy sector reform. The Romania EPE notes that the program embodied successful cooperation between the EU and the Fund.

    112. EPAs and EPEs also emphasize the critical role of TA in assisting the authorities with the implementation of structural conditionality. In Mali, Fund TA appropriately supported program objectives. Policy advice on structural reforms provided in close collaboration with the World Bank contributed to building technical capacity and strengthening domestic institutions, which in turn facilitated the implementation of conditionality. In Mauretania, programs helped the authorities delineate the policy agenda and establish priorities, providing the basis for the provision of much needed TA, which was key in overcoming technical and capacity constraints. In Niger, Fund TA was found to have been well-tailored to program objectives. The Burundi EPA underlines the criticality of coordination among development partners, in particular World Bank TA support to macro critical structural reforms outside the Fund’s core area of responsibility. In a fragile environment with severe capacity constraints, large-scale early TA with a focus on capacity building and the ability to reach out to decision makers remains essential. Following limited TA under the first PRGF arrangement, Burundi’s second PRGF arrangement benefitted from intensive and timely TA, which helped to address capacity constraints and enhance ownership. Financial sector conditionality in the SBA with Guatemala built on extensive Fund TA in this area.

    Appendix 11: Design of the Sample for Country Case Studies

    113. The qualitative analysis is partly based on case studies of a representative sample of countries over the period 2006-2011.44 The sample of 18 countries accounts for a quarter of the 73 countries that benefited from a Fund-supported program during 2006-2011.45 The composition of the sample is as following:

    114. The sample was chosen by taking into account the following:46

    • Equal number of countries with GRA and PRGT programs, implying a relatively higher percentage of GRA (9 of 30) than PRGT ones (9 of 43);

    • Broad geographic distribution, resulting in at least one GRA and one PRGT program from each area department, other than APD (only GRA);

    • Full range of the facilities/instruments for PRGT programs;

    • Mix of countries where program implementation was viewed as difficult and others where it was viewed as good;

    • A range of institutional characteristics of the program countries (such as EU/EMU membership, franc zone membership); and

    • CSO/donor interest.

    These criteria limited the number of potential candidates. In the case of PRGT programs in Africa, however, the room for selection was sufficiently large.

    115. The geographic distribution of the countries by respective area department in the sample closely matched that of actual GRA program countries.47 Once the geographic distribution was chosen, candidates for the sample were narrowed further by eliminating very recent programs, i.e., those where information was only available through Board approval or the first review at the time of the sample selection. Finally, countries were selected based on a judgment by the team, also based on specific issues of general interest, also to civil society. Where available, alternatives were identified. Overall, the final sample is tilted somewhat more toward programs in the early crisis years.

    116. For PRGT countries, the sample also closely matched the geographic share of PRGT program. Two-thirds of the countries in the sample were from AFR, broadly reflecting their weight in the full sample. The selection of countries also reflects the relative shares of franc zone (one-third) and non-franc zone (two-thirds) countries in the AFR sample. There are no APD countries in the sample. In addition, the sample of PRGT countries includes different exchange rate regimes and different program types (including PSIs). The proposed sample is broadly in line with the temporal distribution of programs.

    IV. Data

    117. This section provides information on the data underlying the analysis in the Review of Conditionality. Appendix 12 provides a short presentation of the MONA database. Appendix 13 comprises Tables 9 and 10, which list Fund-supported programs covered by the review.

    Appendix 12: The Monitoring of Fund Arrangements (MONA) Database48

    118. The MONA database contains information on Fund-supported programs since 2002. It tracks program objectives, strategies, conditionality (quantitative and structural), and implementation. It also includes scheduled purchases and reviews and selected macroeconomic indicators.

    • Data are collected at the time of arrangement approval and following each review. MONA data are disseminated within a few weeks following Executive Board meetings, resulting in a cumulative history of Fund-supported programs from Executive Board approval through program completion.

    • MONA comprises information on conditions set for GRA and PRGT-supported programs during 2002-2011, through prior actions, SBs and SPCs, and QPCs. The database also covers indicative targets for programs starting in 2006 (ITs).

    119. MONA is the only electronic database on program design, implementation, and economic objectives and developments, making it a corner stone of the Fund’s institutional memory. The database contributes to the Fund’s capacity to respond in a timely manner to queries about country-specific as well as cross-country aspects of Fund-supported programs. Moreover, MONA is key source of information for the ex post assessments and evaluations of programs.

    Appendix 13: Fund-Supported Programs during 2002-2011 Table 9. List of GRA Programs, 2002-2011

    Table 9.

    List of GRA Programs, 2002-2011

    article image