This evaluation assesses how well IMF-supported programs have helped to sustain economic growth while delivering adjustment needed for external viability. It focuses on IMF financing arrangements over the period 2008–19, under both the General Resources Account (GRA) and the Poverty Reduction and Growth Trust (PRGT).1 While the evaluation does not assess the experience during the COVID-19 pandemic, its lessons have become even more relevant as many countries now face strong headwinds to growth as they seek IMF-supported programs for achieving durable recoveries.
Findings and Lessons
Under the Articles of Agreement, the IMF lends to countries to help correct balance of payments (BOP) problems without resorting to measures destructive of national prosperity. While IMF-supported programs give primary place to restoring external viability, attention to supporting activity during a program and fostering medium-term growth has increased over time, and particularly since the Global Financial Crisis (GFC) in 2008.
This increasing attention to the growth consequences of IMF-supported programs seems to have delivered some positive results. The evaluation does not find evidence of a consistent bias towards excessive austerity in IMF-supported programs during the evaluation period. Except in the crisis context, IMF-supported programs were in most cases able to sustain output broadly in line with historical norms while still delivering needed adjustment. Indeed, cross-country evidence suggests that programs have yielded growth benefits relative to a counterfactual of no Fund engagement and that stabilization and reforms implemented in the program context boosted post-program growth performance. Historical data over a longer time horizon suggest a positive role of IMF-supported programs at initiating sustained growth surges.
Notwithstanding these positive findings, program growth outcomes consistently fell short of program projections, more so in GRA programs than in PRGT programs, broadly consistent with findings of the 2018 Review of Program Design and Conditionality (ROC). Of the programs covered in the evaluation, around one-half experienced an average growth shortfall (relative to initial program projections) during the program period of ½ percentage points or more, while one-fourth had a growth shortfall of over 1½ percentage points. Macro modeling errors, particularly those related to fiscal multiplier assumptions, seem to have been a source of such growth optimism, especially in GRA programs outside of a crisis context. While fiscal multiplier assumptions were broadly in line with the “bucket approach” suggested by guidance given to staff, they were not discussed widely in program documents and their adaptation to country circumstances was limited. Case study evidence suggests that political economy factors in program negotiations also played a significant role in motivating ambitious growth projections and there was limited attention in the program design stage to contingencies to respond to possible growth shortfalls.
Persistent growth optimism raises serious concerns because growth outcomes below program projections in the macroeconomic framework imply slower than intended progress in increasing incomes and strengthening the public balance sheet, undercut program ownership, and fuel rising adjustment fatigue and public opposition to reforms. While greater scrutiny of the realism of program projections as recommended by the 2018 ROC could help to reduce growth optimism, it seems even more relevant to consider whether IMF-supported programs can achieve stronger growth outcomes more in line with the program’s macroeconomic framework by paying greater attention to growth-friendly policies in program design and implementation.
To shed light on this question, the evaluation examined to what extent different policy instruments were used to support program growth objectives during the evaluation period and how they could be applied to foster stronger growth outcomes.
Fiscal policies typically incorporated growth-friendly measures, but with mixed success. Tax mobilization improved in PRGT programs allowing higher capital spending than otherwise, while GRA programs were able to make the tax structure more growth promoting in the post-program period. However, GRA programs relied heavily on spending cuts to achieve adjustment; efforts to protect low-income and vulnerable groups often fell short of their goals; health and education spending did not increase significantly in either PRGT or GRA programs; and a number of case studies raised concerns that growth benefits of higher public investment were limited by poor project selection and wasteful implementation. To help address such concerns, more attention is needed to building better public financial management and governance, building on recent initiatives in this area, and to strengthen monitoring and reporting of the social and distributional impact.
Structural conditionality included in programs generally played a positive role in promoting reforms and growth, but the potential growth benefits of structural reforms were not fully realized. Implementation of structural conditions (SCs) was positively associated with independently measured progress in structural reforms and helped to boost growth within and after the program, with a stronger growth impact for SCs with higher depth and growth orientation. However, the bulk of SCs was oriented to stabilization rather than promoting growth and the average depth and growth orientation of SCs was relatively low. While capacity development (CD) assistance was provided to support SC implementation and was broadly appreciated by country officials, it does not seem to have been delivered more to countries with weaker capacity nor been consistently effective in strengthening SC implementation. Moreover, implementation was significantly weaker for SCs in areas outside of Fund expertise and where collaboration with partners was sought. Country officials felt that Fund teams sometimes paid too little attention to growth-oriented reforms, relying too heavily on partner institutions, even for reforms crucial to program success. Overall, this evidence suggests the need to increase the focus on promoting deeper, more growth-supporting reforms, supported by steps to more closely integrate program and CD work and to strengthen collaboration with partners.
Use of the exchange rate as a policy tool to support growth and external adjustment during programs was quite limited. Exchange rate regime transition was infrequent and more often toward greater fixity, reflecting in part fear of floating. Efforts were typically made to correct significant pre-program overvaluation, although more generally the impact of nominal exchange rate movements on the real effective exchange rate (REER) was partially muted by pass-through to prices. There was also a tendency towards a loss of competitiveness in PRGT programs relying on the exchange rate as a nominal anchor. Where it did occur, significant REER depreciation seems to have supported external adjustment and growth particularly in PRGT programs. This experience suggests that there could be greater scope to use the exchange rate as a policy tool in program design subject to the principle that the exchange rate regime choice is ultimately the authorities’ decision and to members’ obligations under Article IV to avoid manipulating exchange rates to prevent effective BOP adjustment or to gain an unfair competitive advantage. However, doing so successfully would require early attention to providing a supporting policy framework, including to securely anchor inflation expectations, to develop a deep foreign exchange market, and to alleviate supply-side impediments to exports.
In a number of cases, market debt operations were useful to restore debt sustainability and provide the basis for renewed market access, supporting a return to growth. However, debt operations were sometimes too little and too late, and thus had only mixed success in strengthening debt sustainability and improving the balance of payments position. Debt operations with principal haircuts and upfront fiscal adjustment were more successful than those with just debt reprofiling and lower coupons. This experience suggests that while respecting the neutrality principle, the IMF should consistently seek to ensure ambitious debt operations upfront to address debt sustainability concerns to qualify for access to Fund financing, based on careful application of the recently modified debt sustainability analysis (DSA) frameworks.
Tree more general lessons are worth emphasizing.
First, the diverse experience in the case studies underlines that there is no simple recipe for delivering better growth outcomes in IMF-supported programs given the variety in country circumstances and preferences, the underlying causes and contexts of the BOP problems, and the potential scope for policy action. The need for careful tailoring for country conditions is underlined by the case study experience showing the importance of ensuring that the adjustment and growth strategy is fully owned by the government and broadly supported.
Second, the groundwork for a successful policy response to cushion the output consequences of an exogenous shock should ideally be laid well in advance through surveillance and CD work. The case studies repeatedly show that meaningful reforms to strengthen such growth resilience take many years to put in place and become effective, even with strong efforts to provide CD support.
Third, growth and reform strategies envisaged in program design should pay adequate attention to social and distributional consequences. While the focus in this evaluation has been largely on aggregate outcomes, fair distribution of the burden of adjustment and the rewards of recovery are of prime importance, both in their own right to meet national goals and to ensure continued public support for program implementation.
Recommendations
Notwithstanding the IMF’s increasing attention to growth in program design and the generally positive role played by IMF-supported programs in promoting growth, the IMF should consider a number of actions to further enhance program countries’ capacity to sustain activity while undertaking needed adjustment during the program period and to enhance growth prospects beyond the program.
Recommendation 1—Attention to growth implications of IMF-supported programs should become more thorough, systematic, realistic and sensitive to social and distributional consequences.
▸ Board papers supporting GRA as well as PRGT programs should clearly explain the program’s growth implications both during the program and over the medium term. They should discuss how program design reflects the country’s growth strategy and how growth considerations have been taken into account in the macroeconomic framework, ideally based on a well-calibrated country-specific model. Documents should provide more systematic coverage of the quality dimensions of growth, including the distributional consequences of adjustment and reform policies such as how low-income and vulnerable groups are affected during the program period and how they would share in growth over time.
▸ In discussing the program’s macroeconomic framework, particular attention should be paid to discussion of fiscal multiplier assumptions, especially where available country-specific modeling is limited, which should be further fine-tuned to country circumstances based on available evidence and informed judgement.
▸ Program design should pay more consistent attention to contingencies for growth shortfalls, based on scenario analysis, which should help fend of negative perceptions of the Fund’s austerity bias.
▸ Efforts to pay greater attention to distributional aspects may require more granular approaches to conditionality and monitoring. Subject to data availability, strengthened monitoring of key social and distributional metrics would help to measure progress and signal emerging issues for program reviews.
▸ Revisions to the 2002 Guidelines on Conditionality and the 2014 Operational Guidance Note on Conditionality should be considered to give further guidance on the role of Fund-supported programs in fostering favorable growth outcomes. The update to the Guidance Note now under way can provide an opportunity to advance this work, while revisions to the Guidelines on Conditionality could be considered in the next Review of Program Design and Conditionality.
Recommendation 2—IMF-supported programs should pay greater attention to supporting deep, more growth-oriented structural reforms with more effective capacity development support and more effective collaboration with partners in areas outside the Fund’s core mandate and expertise.
▸ The structural reform strategy should be geared to what is important and not what is most easy to agree on or monitor or where the IMF has core expertise, subject to careful consideration of the country’s implementation capacity and the program’s goals. Structural conditionality should be parsimonious enough to avoid overtaxing country capacity but also more focused on correcting underlying distortions and removing structural impediments critical to achieving sustained and inclusive growth.
▸ The Fund should seek ways to strengthen collaboration with the World Bank and other relevant partners in design and implementation of structural reforms in shared and non-core areas. A useful step could be preparation of a Board paper reviewing experience with Bank-Fund collaboration in Fund-supported programs.
▸ The Fund should revisit how CD support is integrated with program design and implementation aimed at promoting deeper and more successful reform efforts in the program context.
Recommendation 3—The Fund should continue to invest in building a toolkit of models and monitors that can be applied as a basis for analysis of the adjustment-growth relationship and assessing growth-related developments in the program context.
▸ Functional departments could continue to take the lead in developing a suite of models suitable for analyzing the adjustment-growth relationship that are tractable and easily accessible for use by country desks to calibrate and apply in their country context. Particular attention should be paid to developing small-scale, easy-to-adapt macro/growth models for low-income countries where data are limited.
▸ Country teams should be encouraged to apply the models now being developed to achieve greater realism in program projections, to explore trade-offs between alternative policy mixes, and explain baseline projections and associated risks to authorities, which should help promote country ownership and mitigate the tendency towards growth optimism. Teams would determine case by case the models best suited to country circumstances and needs. Area departments could also contribute by undertaking in-depth case studies on program successes and failures.
▸ The Fund should increase efforts to keep track of whether structural reforms were sustained after the program concludes, for example by investing more in the Research Department’s new structural reform database.
▸ Further attention should be given to developing and deploying monitors to help support country desks’ capacity to track developments in key distributional indicators and to gauge program impact on key social distributional dimensions of growth, in close collaboration with the World Bank and other agencies.
Budgetary Implications
It should be recognized that full implementation of many of these recommendations would have significant resource costs. Most significantly, the recommendations to take a fuller and more rigorous approach to analyzing and supporting program growth strategies with greater attention in program documents could add considerably to the time needed for program work (including for effective collaboration with the World Bank and other partners). At the same time, much of this work is already well underway or at least anticipated in the Fund’s work program as part of the follow-up to the 2018 ROC. Taking on the additional commitments required would depend on a broader strategic decision to increase attention in the program context to ensure that IMF-supported programs not only deliver sufficient adjustment but also contribute in a more thorough way to sustained and inclusive growth.
Programs under the GRA provide non-concessional lending support, while programs under the PRGT provide lending on concessional terms to low-income countries.
