## A STRATEGY FOR IMF ENGAGEMENT ON SOCIAL SPENDING (ppea2019016)

## Source details

**Canonical URL:** [A STRATEGY FOR IMF ENGAGEMENT ON SOCIAL SPENDING (ppea2019016)](https://www.imf.org/-/media/files/publications/pp/2019/ppea2019016.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/pp/2019/ppea2019016.pdf.md)
- [Structured JSON version](/-/media/files/publications/pp/2019/ppea2019016.pdf.json)

---

### Executive summary: context, rationale, and headline findings
- Social spending comprises social protection, health and education and is a key policy lever for:
  - promoting inclusive growth;
  - addressing inequality;
  - protecting vulnerable groups during structural change and adjustment;
  - smoothing consumption over the life-cycle; and
  - stabilizing demand during economic shocks.
- Key external pressures increasing social spending needs: demographic change (ageing), labor market shifts (including rising female participation and large youth cohorts in many EMDCs), technological change (discontinuous careers, AI/robotics risks), climate change (extreme weather, poverty risks), and fiscal constraints.
- Recent Fund engagement and findings:
  - Fund engagement on social spending has increased, but unevenly; the 2017 IEO Report found scope for strengthening engagement.
  - The Fund’s operational framework introduced social-priority design features (PRGT in 2009) and increased technical assistance and research on inequality and SDG resource requirements.

### Strategy purpose, scope, and guiding principles
- Purpose:
  - Provide broad guidance on when and how the Fund should engage on social spending (social protection, health, education).
  - Serve as management’s response to Board-endorsed IEO recommendations.
- Scope:
  - Clarifies boundaries of IMF engagement; does not prescribe specific sectoral technical designs but identifies in-house resources and the role for IDIs.
- Guiding principle:
  - Engagement should be driven by an assessment of macro-criticality for specific social spending issues, evaluated through three channels: fiscal sustainability, spending adequacy, and spending efficiency.

### Macro-criticality: when to engage
- Definition: a structural issue is macro-critical if it affects, or could affect, domestic (growth, inflation) or external stability.
- Mission chiefs’ survey evidence:
  - Social spending was macro-critical in nearly 80 percent of countries.
  - 70 percent reported that policy advice was provided in this area.
  - The ratio of macro-criticality and advice provision was higher for EMDCs and program countries.
- Typical macro-critical channels:
  - Fiscal sustainability: ageing-related pension and health pressures in most AEs; age-related and service-delivery pressures in many EMDCs; shocks (conflict, refugees, disasters) can create acute pressures.
  - Spending adequacy: large gaps to meet SDGs in many EMDCs; LIDCs on average would require additional spending of 8 percentage points of GDP by 2030 to close health and education gaps.
  - Spending efficiency: high spending not always linked to outcomes; closing inefficiency gaps could lower SDG costs by about 4 percent of GDP in LIDCs and 2 percent of GDP in EMEs; in OECD economies addressing inefficiencies could reduce costs by 40–50 percent.

### Surveillance, research, and evidence on program outcomes
- Trends in Fund engagement:
  - Frequency of social spending terms rose in the 1980s–1990s, dipped around 2008, then recovered to early-2000s levels with increased IMF research linking social spending to inclusive growth.
  - An initiative launched in 2015 to operationalize inequality issues has produced studies for 42 countries, with 3 more ongoing.
- Program evidence on education and health spending:
  - On average, education and health spending in program countries increased by more than, or at the same rate as, in non-program countries.
  - Nevertheless, many instances exist where education and health spending decreased in program countries (as a share of GDP or in real per-capita terms).
  - Probit analysis findings:
    - High GDP growth increases the probability of a decline in spending as a share of GDP but lowers probability of a decline in real per-capita spending.
    - Greater short-term fiscal consolidation raises the probability of a spending decline; this effect is attenuated when consolidation relies more on revenue mobilization rather than expenditure cuts.
    - Declines are more likely where initial spending is high (potentially reflecting inefficiencies).

### IMF-supported programs, conditionality, and program design
- Use of social-priority conditionality:
  - Following the 2009 reform, social spending floors were used in over 90 percent of PRGT-supported programs.
  - Over sixty percent of programs since 2012 included quantitative targets on social and other priority spending, more than double the share during 2002–2011.
  - Across lending facilities, the vast majority of social spending floors were ITs; about 5 percent involved quantitative PCs.
  - Observance of PCs and ITs on social spending was about 70 percent during 2012–17.
- Mission chiefs’ survey:
  - 90 percent indicated social spending was protected in the program, mainly through ITs (70 percent).
  - Most mission chiefs reported social spending was maintained (53 percent) or increased (38 percent) despite nearly 80 percent indicating the program involved fiscal consolidation.
- Program objectives and tailoring:
  - Programs should combine uniformity of treatment with tailoring to country-specific macro-fiscal context, pace of adjustment, and structural challenges.
  - Program objectives may include: mitigating adverse effects on vulnerable groups, identifying fiscal space for adequacy, and improving spending efficiency.
- Strengthening quantitative conditionality:
  - Social spending floors should be carefully defined (share of public spending, share of GDP, or real per-capita terms) and tied to program objectives and projected financing.
  - Prioritize spending with greatest impact on the vulnerable; choose narrow or broad perimeters based on data availability and program goals.
  - Consider financing sources and contingency plans if donor or off-budget financing is relied upon.
  - Use PCs or ITs depending on data quality and criticality; avoid PCs where data quality is poor and instead include structural measures to improve data.

### Specific policy trade-offs: targeting, universalism, and financing
- Targeting methods and administrative feasibility:
  - Means-testing prevalent in many AEs; often infeasible in settings with large informal sectors and limited administrative capacity (particularly LIDCs).
  - Proxy-means testing can produce exclusion and inclusion errors; categorical targeting is simpler but can reduce poverty impact.
  - Expansion via progressive universalism requires fiscal space and improved tax capacity.
- Financing principles:
  - The combined distributional impact of transfers and taxes matters.
  - Key components of an efficient and progressive financing strategy include:
    1. effective strategies to improve tax compliance,
    2. progressive personal income taxes for higher-income groups,
    3. effective taxation of corporate income,
    4. a broad-based consumption tax,
    5. efficient taxation of goods with negative consumption externalities (fossil fuels, tobacco, alcohol),
    6. fighting tax evasion and avoidance.
  - As coverage expands, broadened consumption taxation should be accompanied by measures to protect the poor.
- Mission Chiefs’ Survey on targeting recommendations:
  - Introduction or expansion of means-tested schemes is recommended in 64 percent of cases.
  - Downsizing of schemes that do not require a means test is suggested in 18 percent of cases.
  - Expansion of social programs not based on a means test is recommended in about 18 percent of cases.
  - Under programs, the share of mission chiefs indicating they recommended introduction or expansion of means-tested safety nets is about 80 percent.

### Data, analytical tools, and technical assistance
- Data needs:
  - Good quality, timely, and disaggregated data are crucial for evaluation, advice, and monitoring.
  - Significant data gaps exist in EMDCs and LIDCs; reporting often limited to central government.
  - GFSM2014 and COFOG are the relevant international frameworks; about 80 countries report GFS data to the IMF.
- Technical assistance trends:
  - TA on tax policy and revenue administration has about doubled since 2010.
  - TA on PFM has increased significantly.
  - Over 2000–2018, all TA on social safety nets, and 94 percent on pensions, was provided to EMDCs; EMDCs accounted for 78 percent of TA on expenditure rationalization.
  - Recent SDG cost-estimation work underpins increased TA engagement in EMDCs.
- Diagnostic and analytical resources:
  - FAD to act as hub; available tools include templates for pension projections, SDG costing, social spending efficiency assessments, datasets (Gini indices, historical social spending, COFOG), and distributional analysis tools.

### Implementation, resource implications, and proposed support
- Staff Guidance Note:
  - To be completed by end–2020 to detail existing support, provide tools, and help country teams prioritize engagement; to be updated as needed.
- Current and incremental resource estimates (reported):
  - In the past 2 years, estimated that 8 in 10 country teams discussed social spending in staff reports and 5 in 10 produced a box, selected issues paper or working paper (at an estimated annual cost of US$3 million).
  - Random sample survey of 44 country teams: 18 of 44 anticipated needing more resources in FY20; 26 of 44 anticipated the same or less.
  - Person-weeks per country (time inputs):
    - FY18 Average 2.3; FY19 Average 3.3; FY20 Average 3.8.
    - Median: FY18 2; FY19 3; FY20 3.
    - Max: FY18 13; FY19 14; FY20 13.
    - Min: FY18 0; FY19 0; FY20 0.5.
    - Note: On average, teams spent 3.3 person-weeks in fiscal year 2019 and estimated spending 1 additional person-week per year in FY20 if the proposed strategy was implemented.
  - Estimated additional costs to implement the strategy:
    - Recurring additional costs: US$0.7 million (split equally across area and functional departments).
    - Area department estimate: US$0.36 million; would increase to about US$1 million under a higher-engagement upper-end scenario.
    - One-off (transitional) costs during FY20–FY21 to develop tools/databases, How-to Notes, and in-reach: US$1.3 million (to be financed through internal reallocation).
    - Functional departments baseline support: US$0.35 million (could rise to US$0.45 million under higher engagement).
  - Underlying survey assumption: engagement deepens in 60 countries per year at a marginal cost of 1 person-week; upper-end assumes more extensive engagement in 20 countries per year (marginal cost up to 2.8 person-weeks or US$1 million for area departments).

### Support, coordination, and external engagement
- FAD role:
  - Act as a hub to facilitate country-team engagement with IDIs and to provide tools, templates, and training.
  - Potential new tools: fiscal impact of age-related spending, SDG costing in EMDCs, safety net performance evaluation.
  - Proposed internal Knowledge Exchange website to aggregate social spending data, good practice, quantitative tools, and IDI contact information.
- Cooperation with IDIs and stakeholders:
  - Early dialogue with IDIs to align priorities; acknowledge differences in institutional focus and timelines.
  - Continue participation in SPIAC-B and the WHO-World Bank Technical Working Group on Health Financing.
  - Engage CSOs, academics, unions, and other stakeholders early and follow 2015 Staff Guidelines on IMF Staff Engagement with CSOs.
- Communication:
  - External messaging should clearly explain the Fund’s approach, mandate limits, and the macro-fiscal context of advice; involve COM early for contentious reforms.

### Analytical priorities and staff guidance
- Staff should follow a systematic approach when social spending is macro-critical:
  - Identify macro-critical issues across sustainability, adequacy, efficiency channels.
  - Understand country priorities and engage early.
  - Take stock of existing analysis and identify data gaps; involve IDIs and use FAD tools or in-house analyses.
  - Use TA and IDI collaboration to assess policy options and formulate recommendations reflecting macro constraints and sequencing.
  - Revisit the analysis as conditions evolve.
- Breadth and depth of engagement to be calibrated by urgency, staffing, authorities’ agenda, and Fund comparative advantage; surveillance can accommodate deeper analysis while programs focus on implementation and conditionality if critical to program success.

### Challenges, gaps, and areas for strengthening
- Identified needs:
  - A common definition of social spending to enhance clarity.
  - Better articulation of rationale for IMF engagement and operational steps for consistent engagement.
  - Strengthened in-house analytical capacity and greater attention to spending quality and efficiency.
  - Address perceived Fund bias toward means-tested targeting and insufficient engagement on universal approaches where appropriate.
  - Improve data quality and adherence to reporting frameworks (GFSM2014, COFOG); consider initiative to strengthen GFS reporting of social spending (noting potential costs).
  - Avoid duplication with IDIs and be mindful of Fund resource constraints.

### Conclusions and issues for Board discussion (as presented)
- Core strategy elements:
  - Engagement guided by macro-criticality (channels: fiscal sustainability, spending adequacy, spending efficiency).
  - Continue advice on sustainable financing and increase focus on quality and efficiency of social spending, leveraging IDI expertise.
  - Strengthen program design and conditionality to mitigate adverse effects on the vulnerable and address data gaps; adopt a medium-term sequencing focus.
  - Strengthen support for country teams (Staff Guidance Note by end–2020); FAD as hub for coordination; enhance collaboration with IDIs and stakeholders; improve external communication.
  - Discuss targeted vs universal transfers in context of safety net effectiveness, fiscal and administrative constraints, and social/political preferences.
- Questions for Directors included agreement on macro-criticality as guiding principle, continued policy advice on financing and quality, GRA program documentation of social impacts, collaboration with IDIs, and updating Fund engagement as issues evolve.

*Source: EXECUTIVE SUMMARY and selected chapters, A STRATEGY FOR IMF ENGAGEMENT ON SOCIAL SPENDING (ppea2019016).*

### EXECUTIVE SUMMARY

### EXECUTIVE SUMMARY

### Context and rationale
- Interest in social spending issues has intensified over the last decade, reflecting concerns about rising inequality and the need to support vulnerable groups, especially in the aftermath of the global financial crisis.
- There is a global commitment to continue support for inclusive growth as expressed in the 2030 Sustainable Development Goals (SDGs).
- Social spending is defined to comprise social protection, health and education spending and is viewed as a key policy lever for:
  - promoting inclusive growth;
  - addressing inequality;
  - protecting vulnerable groups during structural change and adjustment;
  - smoothing consumption over the life-cycle; and
  - stabilizing demand during economic shocks.
- Ongoing demographic, technological, and climate developments will pose new challenges and increase social spending pressures.

### Key trends and challenges highlighted
- Demographics:
  - Most Advanced economies (AEs) face rapid ageing; some Emerging Market and Developing Countries (EMDCs) will experience rapid ageing in coming decades.
  - Countries may need to strengthen health and pension systems to enhance spending efficiency and ensure adequate coverage.
- Labor markets:
  - Women’s labor force participation has significantly increased.
  - Many EMDCs face growing youth populations, requiring education and training reforms and social policies that support, not disincentivize, employment.
- Technological change:
  - Increases labor mobility and flexible work arrangements, potentially creating discontinuous careers and volatile income streams.
  - AI and robotics may render many current skills redundant; education and training must adapt to continuous skill upgrading.
  - Maintaining or expanding social protection coverage may require less reliance on employment-linked eligibility and greater reliance on financing from general government revenue.
- Climate change:
  - Expected to increase extreme weather risks, disproportionately affecting low-income and small island states and possibly pushing households into poverty, requiring enhanced social safety nets.
- Fiscal constraints:
  - Tight budgets mean spending pressures from these trends compete with other priorities; many countries would need to strengthen tax capacity to pursue universal access to key social services or expand safety nets.

### The Fund’s recent engagement and findings
- The Fund has increased work on inclusive growth and social spending, including analytical work on inequality and growth and resource requirements for achieving SDGs in education and health.
- Operational changes include the introduction in 2009 of the Poverty Reduction and Growth Trust (PRGT), which enhanced focus on social and other priority spending and incorporated explicit “targets” (typically social spending “floors”) into program design.
- The 2017 Independent Evaluation Office (IEO) Report on “The IMF and Social Protection” found Fund engagement has increased over time but been uneven with scope for strengthening; analysis for this paper confirms that finding and identifies additional challenges.
- The Fund’s engagement includes surveillance, IMF-supported programs, and technical assistance; there has been enhanced engagement on inequality issues and increased technical assistance to expand fiscal space for social spending.

### A strategy for IMF engagement on social spending (core elements)
- Purpose: clarify scope, objectives, and boundaries of IMF engagement and serve as management’s response to Board-endorsed IEO recommendations.
- Engagement would be guided by:
  - an assessment of macro-criticality for a specific social spending issue; and
  - consideration of that issue in a program context and the existence of in-house expertise.
- Key channels through which social spending can become macro-critical:
  - fiscal sustainability;
  - spending adequacy; and
  - spending efficiency.
- The strategy encourages early engagement with authorities and continued staff development of policy advice on:
  - sustainable financing of social spending; and
  - increasing focus on the quality of social spending for improving social outcomes, drawing on expertise of International Development Institutions (IDIs).
- Collaboration and coordination:
  - Fiscal Affairs Department (FAD) would act as a hub to facilitate and strengthen engagement between country teams and IDIs.
  - Early engagement of country teams with relevant IDIs is encouraged.
  - Stronger engagement with other stakeholders, including Civil Society Organizations (CSOs), would be pursued.
- Program design and conditionality:
  - Strengthen consistent emphasis on mitigating adverse effects of adjustment on the vulnerable in IMF-supported programs where consistent with primary program goals (help member correct balance of payments problem and achieve external viability).
  - Conditionality can be instrumental in helping to strengthen tax capacity in support of social spending, improving quality of social spending, and addressing data and information gaps.
- Advice on transfers:
  - Advice on the use of targeted and universal transfers would be discussed in the context of social safety net effectiveness.
  - Appropriate design depends on countries’ social and political preferences and should be consistent with fiscal and administrative constraints.
- External communication:
  - In specific countries, focus on providing a clear explanation of Fund policy advice reflecting the overall macro-fiscal context and strategy, while clarifying what the Fund can and cannot do.

### Implementation, guidance, and resources
- A Staff Guidance Note, to be completed by end–2020, will:
  - detail existing support and initiatives for strengthening engagement;
  - provide further tools for country teams and improved knowledge management;
  - help country teams prioritize when and how extensively to engage on different social spending issues.
- The Note will be updated as needed to reflect evolving social spending issues and Fund experience.
- In the interim, country teams will continue extensive engagement on social spending drawing on existing resources.
- Because IMF engagement on social spending is already extensive, additional resources to implement the strategy are at this point not expected to be significant, beyond some set-up costs.

### Analytical and operational follow-up
- The strategy is informed by background analyses, internal and external consultations, and the 2017 IEO Report.
- It envisions strengthened collaboration across departments (FAD as hub) and improved engagement with IDIs and stakeholders.
- Implementation will require granular and gradually evolving guidance to staff and enhanced tools and knowledge management.

*Source: EXECUTIVE SUMMARY, A STRATEGY FOR IMF ENGAGEMENT ON SOCIAL SPENDING (June 14, 2019).*

### 4. This paper proposes a strategy to provide broad guidance on the Fund’s engagement

### A STRATEGY FOR IMF ENGAGEMENT ON SOCIAL SPENDING

### Purpose and scope of the strategy
- Proposes broad guidance on the Fund’s engagement on social spending (social protection, health and education).
- Central to Fund management’s response to the Board-endorsed recommendations of the 2017 IEO Report on “The IMF and Social Protection,” which found Fund engagement was uneven and identified areas for strengthening.9
- Recognizes strengthening social spending is often a complex, long-term process requiring sustained engagement on fiscal and implementation constraints, and depending critically on country capacity and support from International Development Institutions (IDIs).10
- Provides broad guidance to staff on when and how to engage with member countries on social spending issues across Fund activities, ensuring advice is transparent and evenhanded.
- Does not set out specific Fund policy advice on social spending, but identifies resources available to staff to develop such advice, including engagement with IDIs that have deeper expertise.
- Clarifies Fund advice on the appropriate use of targeted and universal transfers.
- Provides a basis for developing effective communication on the scope and boundaries of IMF engagement on social spending issues.

### Implementation: Guidance Note and resource implications
- A Guidance Note to be completed by end–2020 to help country teams prioritize when and how extensively to engage on different social spending issues, considering trade-offs, policy options, traction in countries, and internal expertise constraints.
- The Guidance Note will:
  - Detail existing resources and initiatives for strengthening support across country contexts and social spending areas.
  - Be updated as needed to reflect evolving social spending issues and Fund policy advice, and as the Fund gains more engagement experience.
- The paper includes an estimate of the resource costs of implementing the strategy, which may be updated as implementation experience is gained.
- In the interim, country teams continue existing engagement on social spending using current resources; cross-country departmental papers are noted as a potential vehicle to develop and operationalize guidance.

### Complementarity with other Fund initiatives
- Draws on recent and ongoing policy work including:
  - Social Safeguards and Program Design in PRGT and PSI-supported Programs (SSP2017).
  - Review of Conditionality (ROC2018).
  - Comprehensive Surveillance Review (CSR).
  - Initiative on How to Operationalize Inequality Issues in Country Work.

### Evidence base, consultations, and analytical inputs
- Extensive analysis and consultations informed the paper, summarized as:
  - Stakeholder consultations with IMF Executive Directors, IMF area departments, IDIs, CSOs, trade unions, and academics (including seminars during the 2018 Spring and Annual Meetings, an ITUC social protection conference, and a workshop at the LSE); a dedicated consultative group from CSO and academic communities; and an open online consultation.
  - Text-mining analysis of surveillance and IMF-supported program documents examined the evolution of Fund engagement on social spending issues.
  - A mission chiefs’ survey (MCS) provided information on extent, nature, and challenges of country-team engagement with authorities and IDIs.
  - A separate survey of randomly selected country teams estimated resource implications of the strategy.
  - Cross-country data and econometric analysis assessed whether social spending has been protected in IMF-supported programs, based on a newly constructed database on education and health spending.
  - A note on the appropriate use of universal and targeted transfers discussed trade-offs and emphasized evaluating both tax and transfer sides when designing fiscal redistribution systems.
  - Case studies on eleven countries provided granular information on engagement in surveillance and program contexts and staff challenges.

### Recent trends in IMF engagement on social spending
- IMF engagement on social spending issues has been increasing, especially over the 1990s, across surveillance and program activities, supported by capacity development work.
- The mission chiefs’ survey and country case studies confirm engagement on social spending issues is high.
- Text-mining analysis (Box 2) of Article IV Staff Reports and program documents over four decades shows:
  - Frequency of social spending terms increased significantly over the 1980s and 1990s, remained stable until the mid-2000s, dropped sharply around 2008 (coinciding with the global financial crisis and the 2007 Surveillance Decision (SD)), and then recovered to early-2000s levels.
  - The recovery after the late 2000s was accompanied by increased IMF research on social spending and its link with inclusive growth.

### Rationale: Social spending and inclusive growth
- Public investment in education and health is a significant driver of inclusive growth; expansion of access to quality basic education and health services increases human capital and decreases inequality in outcomes (Box 3).
- Social protection spending reduces income inequality via redistribution and can enhance resilience and investment by lower-income groups.
- Conditional cash transfer (CCT) programs can address current poverty and break intergenerational transmission of poverty.
- Social spending needs to be efficient and sustainably financed:
  - Studies find substantial inefficiencies across social spending components, indicating room to improve social outcomes within current budgets.
  - Spending efficiency requires effective public financial management (PFM) systems for budget formulation, execution, and monitoring.
  - Sustainable financing requires balancing financing, revenue mobilization, and spending across transfers and investments in education, health, and infrastructure—particularly important in EMDCs with large human and physical capital needs, low tax capacity, and rising debt, and also important in AEs that need to reduce high public debt.

### Surveillance: evolving emphasis and work on inequality
- Since the 1990s IMF policy advice has increasingly recognized that growth and distributional objectives need not conflict and that social spending plays a key role in inclusive growth.
- An initiative launched in 2015 aims to ensure inequality issues are routinely addressed in IMF surveillance; since inception, studies have been undertaken for 42 countries, with analysis for another 3 countries ongoing.12
- These studies cover safety nets, price subsidies, pension systems, and evaluation of poverty and inequality impacts of proposed measures; more recent work addresses gender inequality and macroeconomic benefits from improving gender equity.13

### IMF-supported programs: protection of social spending and conditionality
- PRGT-supported programs are designed to safeguard social and other priority spending and, whenever appropriate, to increase it.
- Both PRGT and GRA-supported programs often include quantitative spending targets (“spending floors”) on social and other priority spending, and specific reform measures to protect vulnerable groups as prior actions or structural benchmarks (SBs).
- Findings from ROC2018 and related staff analysis:
  - The share of programs with SBs on social spending declined during 2012–17 compared to 2002–2011 (Figure 1).
  - Use of quantitative conditionality (performance criteria (PCs) and indicative targets (ITs)) related to social spending increased significantly.
  - Over sixty percent of programs since 2012 included quantitative targets on social and other priority spending, more than double the share during 2002–2011.16
  - Following the 2009 Reform of the Fund’s Facilities and Financing Framework for Low-income and Developing Countries, social spending floors were used in over 90 percent of PRGT-supported programs.
  - Across all lending facilities, the vast majority of social spending floors were ITs, with about 5 percent involving quantitative PCs.
  - Observance of PCs and ITs on social spending was about 70 percent during 2012–17, comparable to compliance levels for other quantitative targets; these findings are confirmed by the MCS.17
  - The decline in SBs on social spending likely reflects the 2009 review-based approach to structural conditionality; where SBs on social spending were included, they were met about 70 percent of the time.
  - Use of structural conditionality was somewhat lower in PRGT-supported programs, though PRGT SBs spanned a wider range of social sector reforms, including cash transfer programs, expenditure frameworks on health and education, data reporting enhancements, and quality-improvement measures (such as unified registries for targeting).
  - Staff analysis of ROC2018 noted limited focus on the quality of social spending and on inequality.

- Mission Chiefs’ Survey (MCS) highlights:
  - Almost all mission chiefs (90 percent) indicated that social spending was protected in the program, mainly through the use of ITs (70 percent), and that conditionality was met most of the time.
  - Most mission chiefs for program countries indicated social spending was either maintained (53 percent) or increased (38 percent) despite nearly 80 percent indicating the program involved fiscal consolidation. These results align with SSP2017 findings that 90 percent of PRGT programs included a social spending floor and that in two-thirds of these programs social spending ITs were met.

### Evidence on spending outcomes in program countries
- Cross-country analysis examined public education and health spending changes after program approval, measuring changes in spending in each program year after approval relative to the year before approval, with interquartile ranges shown for:
  - Change in Health Spending (Percent of GDP)
  - Change in Education Spending (Percent of GDP)
  - Change in Health Spending (Percent change in real per capita spending)
  - Change in Education Spending (Percent change in real per capita spending)
- Note: Only countries with observations in the year before approval and during program years are included.

*Source: IMF staff, ppea2019016 - 4.*

### 12. Although programs have protected education and health spending on average,

### 12. Although programs have protected education and health spending on average,

### Findings on education and health spending under programs
- Most studies found that, on average, education and health spending in program countries either increased by more than, or at the same rate as, spending in non-program countries.
- New analysis of spending trends since 2000 undertaken for this paper confirms these findings (see Background Paper II).
- Nonetheless, a significant number of instances were identified in which education and health spending decreased in program countries (whether measured as a share of GDP or in real per-capita terms; Figure 2).
- An analysis of factors explaining instances of large spending decline (i.e., those greater than the median negative changes) shows:
  - High GDP growth increases the probability of a decline in spending as a share of GDP but lowers the probability of a decline in real per-capita spending (Figure 3).
  - The probability of a decline in spending (both as a share of GDP and in real per-capita terms) is greater where short-term fiscal consolidation is high; this effect is attenuated the more fiscal consolidation is achieved through revenue mobilization rather than expenditure consolidation.
  - Declines in spending are more likely where initial spending is high, which in some cases may reflect inefficiencies (such as excessively high wages).

### Quantitative and empirical notes (as reported)
- The probit model reported average marginal effects by raising each explanatory variable from its 25th to 75th percentile value (Figure 3).
- The results are similar when focusing on all (and not just large) instances of social spending decreases.

### Technical assistance (TA) on social spending
- Technical assistance in support of social spending has increased significantly.
- Many EMDCs need to create fiscal space to finance increases in social spending.
- Recent research finds that countries seem to move to a higher growth path once tax revenue reaches around 15 percent of GDP (Gaspar and others, 2016).
  - About half of low-income and developing countries (LIDCs) and a third of EMEs have tax ratios below this threshold.
- TA on improving tax policy and strengthening revenue administration (areas where the IMF traditionally takes a lead role) has about doubled since 2010 (Figure 4).
- TA on PFM has also increased significantly, focusing on enhancing spending efficiency and governance through strengthening budget formulation, execution, and monitoring.
- The scale of expenditure policy TA is relatively low, reflecting the lead role played by other IDIs (particularly the World Bank), but the scope of spending issues covered is broad.
- TA directly linked to social spending is mainly provided to EMDCs.
  - Over 2000–2018, all TA on social safety nets, and 94 percent on pensions, was provided to EMDCs, which also accounted for 78 percent of TA on expenditure rationalization that often includes health and education.
- Recent work on estimating the cost of achieving the SDGs has laid the foundation for increasing TA (and policy) engagement on social spending in these economies, including in collaboration with other IDIs (Gaspar and others, 2019).

### Broad assessment and challenges for IMF engagement
- Engagement on social spending has increased over recent decades, but the level and effectiveness of engagement remain uneven.
- Identified needs and challenges include:
  - A common definition of social spending to provide more clarity.
  - Better explaining the rationale for IMF engagement (i.e., when to engage) and identifying steps for effective engagement (i.e., how to engage) to improve consistency.
  - Strengthening in-house analytical work and expertise to improve the quality of policy advice.
  - Greater attention to specific policy issues, including:
    - The (perceived) Fund bias in favor of narrowly targeted transfers and against more universal types of transfer systems.
    - Encouraging and assisting a member to pursue its own objectives related to protection of vulnerable groups and social spending during adjustment (if these objectives are consistent with the primary goal of helping the member correct its balance of payments problem and achieve external viability).
  - A focus on enhancing the quality and efficiency of social spending to maximize its impact and improve social outcomes.
  - Enhancing engagement with external stakeholders (IDIs, CSOs, external experts) in a more systematic and broader manner while preventing duplication.
  - Better communication on Fund engagement on social spending issues to help enhance understanding and develop stronger country ownership of reforms.
  - Being mindful of constrained Fund resources to avoid crowding out core activities and not duplicating the activities of other IDIs with greater social spending expertise.

### Strategy architecture and aims
- The strategy aims to make engagement more consistent by clarifying when and how to engage, including how to strengthen engagement in program contexts, identifying internal resources, and measures to strengthen external engagement and communications (Table 1).
- Key strategy elements highlighted include:
  - Providing a social spending definition.
  - Addressing selected policy design issues (e.g., discussion of targeted and universal approaches to social transfers; perceived narrow focus on means-testing).
  - Focus on quality and efficiency of spending to improve social outcomes.
  - Providing guidance on when/how to engage, guided by the macro-criticality principle and steps for effective engagement.
  - Discussing enhancing support for in-house analysis and leveraging external expertise/cooperation with IDIs.
  - Attention to program design/issues and conditionality, and resource implications of the strategy.

### When to engage: macro-criticality
- Macro-criticality is an important guiding principle for when to engage on social spending issues.
  - A structural issue is macro-critical if it affects, or has the potential to affect, domestic (e.g., growth and inflation) or external stability.
  - The Fund should avoid duplicating the efforts of IDIs that often have greater expertise in these areas.
- Mission chiefs indicated:
  - Social spending was macro-critical in nearly 80 percent of countries.
  - 70 percent reported that policy advice was provided in this area.
  - The ratio was higher for EMDCs and program countries.
- Reasons for macro-criticality varied by country income level:
  - Advanced economies (AEs): expected social spending pressures (especially from population ageing) and achieving authorities' distributional objectives.
  - Emerging market economies (EMEs): achieving authorities' distributional objectives, risks to social or political stability posed by insufficient spending levels, and large social protection gaps.
  - Low-income and developing countries (LIDCs): large coverage gaps in education and health as well as risks to social or political stability.

### Macro-criticality channels of social spending
- The channels through which social spending may be macro-critical are grouped into three, often interrelated, channels:
  - Is social spending sustainably financed?
  - Is social spending adequate?
  - Is social spending efficient?
- A social spending issue is considered macro-critical if one, or any combination, of these channels is a policy concern.
- Assessment of macro-criticality often requires joint consideration of multiple channels and possible trade-offs. For example, if social spending is inadequate and needs to be increased, it may need to be accompanied by measures to ensure spending efficiency and fiscal sustainability.
- Fiscal sustainability is a common channel of macro-criticality:
  - Most AEs face significant health and pension spending pressures arising from ageing populations.
  - Many EMDCs face age-related pressures linked to population growth putting pressure on education and health systems.
  - Conflicts, refugee crises, natural disasters, or exogenous shocks (e.g., collapse of commodity prices) can create social spending pressures or reduce fiscal resources, prompting the need to create fiscal space, increase tax revenue, reduce other spending, or increase social spending efficiency.
- The distributional impact of adjustment depends on the composition of social spending measures and the composition of tax measures.

*Source: ppea2019016 - 12. Although programs have protected education and health spending on average,*

### 20. Spending adequacy reflects the importance of social spending for achieving inclusive

### 20. Spending adequacy reflects the importance of social spending for achieving inclusive and sustainable growth

### Spending adequacy and SDG financing gaps
- In many EMDCs, increased social spending is required to fill gaps in basic education, health, and social protection coverage as recognized in the SDGs.
- In the case of LIDCs, closing health and education gaps to achieve SDGs would require additional spending of 8 percentage points of GDP on average by 2030.
- Social spending adequacy can also be a temporary concern during crises or large adjustments (example: Egypt case study), when social protection spending may need to be scaled up to protect the vulnerable and ensure a successful transition.
- Country-specific spending adequacy depends on historical, political, and social factors, and on the role of private sector or non-governmental institutions which may complement or be crowded out by public social spending.

### Spending efficiency and social outcomes
- Relatively high education and health spending is not always reflected in commensurately high social outcomes, indicating substantial room to improve spending efficiency (Figure 9; South Africa case study).
- Inefficiencies can lead to:
  - Poor social outcomes despite high spending.
  - Unnecessarily high spending that crowds out other priority spending or leads to higher taxation.
  - Work disincentives in AEs if poorly designed social protection systems increase benefit withdrawal rates and marginal tax rates, potentially raising unemployment and spending.
  - Higher administrative costs in EMDCs when small duplicative programs span numerous ministries.
- Estimated potential gains from closing inefficiency gaps:
  - Would lower additional spending needed to achieve SDG commitments by about 4 percent of GDP in LIDCs and 2 percent of GDP in EMEs.
  - In OECD economies, addressing inefficiencies could reduce costs by 40–50 percent.

### General engagement principle: systematic approach for staff
- Objective: help members resolve macro-critical policy challenges associated with social spending in surveillance and program contexts.
- Strategy components:
  - Set out a systematic approach to social spending, including guidance on breadth and depth of engagement.
  - Provide guidance on leveraging external expertise alongside in-house resources.
  - Balance uniformity of treatment with flexibility to tailor engagement to country-specific circumstances (macro-critical issues, development level, economic cycle, institutional capacity, social and political preferences).

### Steps for staff engagement on macro-critical social spending issues
- Identify social spending issues that are macro-critical using the macro-criticality assessment; large spending alone does not warrant engagement unless sustainability, adequacy, or efficiency concerns endanger macro stability.
- Understand the member’s needs and priorities and engage early to help prioritize issues and ensure adequate social protection during adjustment.
- Take stock of existing analysis (Fund reports, TA reports, working papers, external IDI work, academia, CSOs, private sector).
- Identify and close analytical and data gaps; involve IDIs early and use FAD tools or in-house analyses as staffing permits; area departments may conduct cross-country analysis if relevant.
- Use TA, in-house work, and IDI collaboration to assess options in-depth.
- Formulate policy recommendations with FAD support and IDI input; recommendations should reflect the broader macroeconomic context, financing constraints, and sequencing.
- Revisit and repeat the sequence to account for changing economic situations and evolving macro-critical issues.

### Breadth and depth of engagement
- Breadth: cover relevant social spending components; not all sub-components need consideration (e.g., if primary education is near-universal, focus may be on secondary education expansion plans).
- Depth should consider:
  - Urgency of the issue and available staffing/resources.
  - Authorities’ policy agenda and implementation capacity; TA may be important.
  - The Fund’s comparative advantage and in-house expertise; more granular sectoral reform design may be left to IDIs while ensuring macroeconomic implications are consistent with Fund advice.
- Surveillance vs program contexts:
  - Surveillance may permit in-depth analysis and formulation of recommendations.
  - Programs should focus on implementation of developed policy advice; conditionality may be appropriate if critical to program success.
  - Authorities’ TA requests can influence depth of advice.

### Medium-term focus and capacity building
- Addressing social spending issues typically requires medium-term reforms:
  - Strengthening revenue administration and tax policy for sustainable financing.
  - Improving spending efficiency via PFM strengthening, anti-corruption, and service delivery capacity.
- Country teams should consider how TA and IDI engagement can build capacity and influence sequencing.

### Specific policy issues and priorities
- Sustainable financing:
  - Where social spending is inadequate, creating fiscal space for increased social spending may be important, especially in LIDCs, including through efficient and progressive tax systems.
  - Where social spending needs to be contained, improving spending efficiency can help achieve social objectives.
- Quality and efficiency:
  - Increased focus needed on quality and efficiency to improve social outcomes (recommended in ROC2018).
  - Options include reallocating within social spending, streamlining duplicative programs, eliminating ghost beneficiaries or vacant public positions.
  - This may warrant more collaboration with IDIs (case studies: Cyprus, Jamaica, Ukraine).

### Data needs and statistical frameworks
- Effective implementation requires better data:
  - Good quality, timely, and disaggregated data are crucial for evaluating policies, formulating advice, and monitoring outcomes.
  - Significant data gaps exist in EMDCs and LIDCs, including high non-reporting and focus limited to central government.
- Existing standards and IMF role:
  - GFSM2014 (Government Finance Statistics Manual) describes internationally agreed standards for government finance statistics; about 80 countries report GFS data to the IMF.
  - COFOG classification (OECD/UN) is the international framework for reporting functional government spending, including social benefits and education (GFSM2014).
  - Universal price subsidies are not included in the definition of social protection consistent with IDI practice.
  - Consideration could be given to the IMF leading an initiative to strengthen adherence to GFS reporting of social spending, but this could involve considerable costs.

### Targeting, universalism, and financing of social assistance
- IMF advice often centers on targeting mechanisms based on means testing rather than alternative approaches (e.g., presence of children or elderly).
- Targeting methods:
  - Sophisticated means testing is common in many AEs but often infeasible in countries with large informal sectors and limited administrative capacity (particularly LIDCs).
  - Proxy-means testing estimates income from household characteristics correlated with low income but can produce exclusion and inclusion errors.
  - Categorical targeting (e.g., children, elderly) is simpler but can reduce poverty impact due to targeting errors.
  - Expanding coverage via progressive universalism requires fiscal space.
- Appropriate mix of universal and targeted transfers depends on administrative, financing, social, and political constraints and may change over time (e.g., biometrics improving administrative capacity).
- Financing transfers:
  - What matters ultimately is the combined distributional impact of transfers and the taxes that finance them.
  - Key components of an efficient and progressive financing strategy include:
    1. effective strategies to improve tax compliance,
    2. progressive personal income taxes for higher-income groups,
    3. effective taxation of corporate income,
    4. a broad-based consumption tax,
    5. efficient taxation of goods with negative consumption externalities (fossil fuels, tobacco, alcohol),
    6. fighting tax evasion and avoidance.
  - Strengthening tax system dimensions is key to promoting progressive universalism.
  - Expanded safety net coverage should accompany broader consumption taxation to protect the poor.
- MCS findings on targeting recommendations:
  - Introduction or expansion of means-tested schemes is recommended in 64 percent of cases.
  - Downsizing of schemes that do not require a means test is suggested in 18 percent of cases.
  - Expansion of social programs not based on a means test is recommended in only about 18 percent of cases.
  - Under programs, the share of mission chiefs indicating they recommended introduction or expansion of means-tested safety nets is about 80 percent.

_Italic: Source — IMF staff (chapter content provided in the supplied PDF excerpt)._

### 32. Program objectives and design related to social spending should be tailored to

### 32. Program objectives and design related to social spending should be tailored to country-specific circumstances.

### Program objectives and tailoring to macroeconomic context
- Programs should follow the principle of uniformity of treatment while being consistent with country-specific macroeconomic context: intensity of short-term macro-fiscal pressures, pace and size of adjustment needed, and longer-term structural challenges.
- Case studies illustrating tailored approaches include Jamaica (tax reform and scaling up the safety net), Ukraine (subsidy reform and scaling up an energy compensation mechanism), and Egypt (strengthening social spending while addressing long-standing imbalances).
- Programs can aim at one or a combination of the following objectives:
  - Mitigate the adverse effects of adjustment measures on the vulnerable, and ensure conditionality supports social objectives where critical for program success.
    - In line with staff recommendations of ROC2018, increase focus on the quality of social spending, in addition to the level, underpinned by more analysis of the program’s impact on the poor and vulnerable groups.
    - PRGT-supported programs have an established principle of reducing poverty and protecting the vulnerable, including through appropriate conditionality (social safeguards).
    - Under Fund policies for GRA financing there is no similar general focus, but there is extensive engagement on social spending issues in GRA programs. Based on existing good practice, in GRA-supported programs staff should analyze and, as appropriate, document the social impact of adjustment and measures to protect the vulnerable to build public support and political ownership.
    - PRGT-supported programs are expected to explicitly incorporate measures, possibly as conditionality, to protect the vulnerable and generate public support for adjustment; GRA programs would do so where critical for achieving program success and may require the introduction or expansion of social safety nets.
  - Identify fiscal space to address the adequacy of macro-critical social spending.
    - Especially relevant to PRGT-supported programs whose core objective is to foster sustainable and inclusive growth.
    - Fiscal space can be created through increasing domestic revenue or donor grants, expenditure rationalization, and better PFM systems.
  - Improve the efficiency of social spending to cost-effectively achieve social objectives.
    - Examples: streamlining administratively costly social programs and enhancing PFM systems to strengthen safety nets.
    - IDIs are expected to lead on design and implementation issues aimed at promoting spending efficiency.
    - Case studies (e.g., Cyprus and South Africa) illustrate promoting policies to improve spending efficiency as an important component of IMF engagement.

### Strengthening quantitative conditionality on social spending
- There is room to strengthen quantitative conditionality on social spending.
- Mitigating adverse effects on vulnerable groups and improving spending adequacy can usually be addressed by including quantitative conditionality—social spending floors—typically relying on existing social spending schemes.
- Factors to consider when designing program targets to strengthen the use of quantitative conditionality:
  - Quantitative targets for social spending (“floors”) should be carefully defined.
    - Program objectives may be expressed as a share of public spending, as a share of GDP, or in real per-capita terms; objectives are translated into program targets set in nominal terms.
    - Where relevant, floors should be specified at sectoral or social program level and tied to a specific program objective and projected financing.
    - Coverage of social spending floors will vary across countries depending on program objectives and data availability and should be explained in program documents and linked to program objectives and predictable financing.
  - Prioritize social spending with the greatest impact on the vulnerable.
    - Staff should exercise judgment in defining the perimeter of targets.
    - Narrowly defined targets (e.g., spending on specific programs that cost-effectively protect vulnerable groups or enhance their human capital) may be desirable when resources are limited or many components are inefficient.
    - Broader targets (e.g., covering all education and health spending) may be needed when disaggregated data are unavailable and could be supported by SBs to improve data quality supported by capacity development.
  - Carefully consider financing sources.
    - Authorities should ideally have adequate control over financing sources to avoid unexpected shortfalls.
    - If targets rely in part on financing outside authorities’ control (e.g., donor support), contingency plans should be considered.
    - When social spending is directly financed by donors, the availability and size of off-budget spending should be discussed and documented in program documents.
  - Review availability and quality of social spending data.
    - Country teams can choose to set quantitative targets as PCs or ITs, depending on criticality to program objectives.
    - Using PCs should be avoided if there are concerns about data quality; programs can incorporate structural measures to improve data quality and timeliness.

### Structural measures to strengthen safety nets and improve medium-term outcomes
- Where relevant, programs should consider structural measures to strengthen social safety nets and improve quality and efficiency of social spending and outcomes in the medium-term.
  - Structural measures can strengthen existing schemes (e.g., removing inefficiencies) or introduce new schemes.
  - Structural measures can be incorporated as conditionality if critical to program success.
  - Such reforms may contribute to creating fiscal space over time.
- Factors to consider when designing structural measures:
  - Criticality: whether a measure is set as structural conditionality depends on its criticality for program success and would typically apply to reform milestones; program design needs to follow the principle of parsimony, particularly in fragile states. Other measures can be commitments in the MEFP with implementation timelines.
  - Type: measures may include introducing new social assistance schemes or redesigning existing ones; programs should include measures to enhance spending efficiency (e.g., road map for increasing value for money in education). Use of such measures typically requires leveraging IDI advice.
  - Appropriate sequencing and timeline: consider phasing and sequencing, economic cycle, and fiscal implications; take a realistic approach to time needed for effective implementation.
  - Capacity constraints: successful implementation may require training, redeployment, or hiring of staff by the authorities.

### Documentation, discussion with authorities, and conditionality consistency
- Social spending-related issues need thorough discussion with authorities and consistent documentation in staff reports.
  - Program documents need to explain the criticality of specific measures for program objectives—particularly important in PRGT and PSI programs given the critical importance of poverty reduction.
  - If a country cannot effectively monitor social spending or faces obstacles in designing and implementing measures, program documents should discuss such challenges and ideally present plans or measures to build capacity.
- Quantitative conditionality includes performance criteria (PCs) and indicative targets (ITs) per the 2002 Conditionality Guidelines; PCs are formal conditions for purchases or disbursements, while ITs are used where substantial uncertainty exists.

### Support for country teams and IMF resources
- Functional departments provide significant support to country teams on social spending; FAD leads many efforts with involvement from ICD, RES, SPR, STA.
- Types of support:
  - Policies: Board papers on pensions, health, social protection, food and energy subsidies, governance, and fiscal space via tax design and administration.
  - Technical assistance (TA): develops country capacity to evaluate and formulate fiscal policies, enhance revenue systems, strengthen institutions and policies for management and efficiency of social spending; TA informs conditionality and training supports nationally-owned strategies; considerable TA on COFOG in GFS.
  - Diagnostic tools and databases: information on spending pressures (e.g., aging) published in Fiscal Monitor; templates for assessing social spending efficiency have enabled increased analytical work.
  - Analytical work: Staff Discussion Notes, Selected Issues Papers, How-to-Notes, Technical Notes and Manuals contribute to internal thinking and external dialogue.
  - Review departments: provide feedback on policy discussions and evenhandedness when social spending is in IMF documents; support operationalizing inequality work through Article IV consultations.
  - Synergies: engagement on social spending benefits from mainstreaming inequality in surveillance and integrating SDG cost estimates into country work and TA for Medium-Term Revenue Strategies.
  - Seminars and training: cover energy subsidy reform, food and other subsidies, safety nets, wage bill spending, and PFM.
  - Access to external resources: facilitated through IMF-internal websites and interaction with IDIs, including participation in SPIAC-B, conferences, and seminars.
- Evidence from internal surveys and mission chiefs:
  - The Mission Chiefs’ Survey (Background Paper III) found that in 92 percent of GRA countries staff assessed social spending to be macro-critical and in 96 percent staff recommended social spending reforms.
  - Authorities’ responses to the ROC2018 survey suggest scope for significant improvement in assessing the social and welfare impact of program policies.

### IMF resources (selected categories summarized)
- General papers and templates for operationalizing inequality, SDG costing, and expenditure efficiency assessments.
- Datasets (internal): Gini Income Inequality Indices; historical country social spending data; COFOG in Annual GFS Database.
- Pension: papers, pension template, country-level long-term spending projections.
- Social assistance: papers on Universal Basic Income, energy subsidy reform, distributional analysis tools.
- Health and education: papers and country-level historical and projection datasets; EAT templates for education.
- Creating fiscal space: papers on corporate taxation, fiscal capacity in fragile states, revenue mobilization.

*Source: ppea2019016 - 32. Program objectives and design related to social spending should be tailored to country-specific circumstances.*

### 37. FAD could further strengthen support for country teams, resources permitting,

### 37. FAD could further strengthen support for country teams, resources permitting

### Strengthening support for country teams
- New dedicated tools recommended for FAD (building on existing initiatives and in collaboration with IDIs) to:
  - assess the fiscal impact of age-related spending (pensions and health) and evaluate alternative policy options;
  - cost the spending required to achieve the SDGs in EMDCs;
  - evaluate the performance of social safety net systems in EMDCs, including identifying coverage gaps.
- Seminars and courses for staff could be expanded in collaboration with ICD and external academic and IDI expertise.
- New internal Knowledge Exchange website proposed to:
  - facilitate access to social spending data and information (examples: good practice on social safeguards in surveillance; links to relevant analytical work, blogs, and other relevant websites; quantitative tools; and IDI country and expert contacts);
  - provide guidance on how to engage on social spending issues, including a list of key IDI country contact persons on social spending and a series of How-to-Notes on different components of social spending.
- These measures aim to enhance the capacity of country teams to leverage available and new resources, expertise, and analytical products.

### Country-level engagement with IDIs and other stakeholders
- Close engagement with IDIs is valuable; IMF country teams mainly rely on their own resources but often leverage external expertise, most frequently from the World Bank.
- Barriers to effective cooperation at the country level include:
  - differences in institutional focus (Fund’s short-term stabilization vs IDIs’ longer-term development focus);
  - different institutional priorities for individual countries;
  - lack of information on IDIs’ organizational setup, misalignment of work plans, and uneven country engagement.
- Strategy responses:
  - emphasize early dialogue with IDIs when a country strategy is being developed to align institutional priorities and foster effective collaboration;
  - FAD would act as a hub for identifying relevant counterparts and provide guidance on where IDIs take the lead or provide complementary support (drawing on IMF Resident Representatives’ knowledge and contacts);
  - where IDI support is not available and the social spending issue is macro-critical or critical to program success, IMF staff may need to develop internal expertise.
- Institutional-level engagement:
  - Since early 2018 the Fund participates in SPIAC-B and in the WHO-World Bank-led Technical Working Group on Health Financing;
  - continue interaction with academic experts and coordination with multilateral development banks on policy-based lending;
  - explore enhanced engagement of IDI experts in developing technical analysis.
- Engagement with other stakeholders (CSOs, academics, think tanks, labor unions, business, local experts) should occur early to enrich country understanding, improve policy design, increase ownership, and strengthen traction; staff interactions should follow the 2015 Staff Guidelines on IMF Staff Engagement with CSOs.

### Communicating IMF advice externally
- External communication should build awareness of the Fund’s engagement and make policy advice more impactful by:
  - explaining the Fund’s approach clearly in specific country contexts and setting realistic expectations about the Fund’s mandate, resources, and expertise;
  - articulating IMF policy advice where issues are likely contentious (example: sequencing reforms to strengthen social safety nets in the context of energy subsidy reform);
  - underscoring the Fund’s long-standing engagement on social spending issues and commitment to evenhandedness.
- Collaboration with the Communications Department (COM) should begin early in the surveillance cycle and program design, based on a communications plan that:
  - explains why social spending is macro-critical and its role for economic development;
  - clarifies the Fund’s area of responsibility when social spending policies are designed in consultation with external stakeholders;
  - spells out specific policy recommendations in a broader policy context, including synergies, complementarities, constraints, and trade-offs.
- Use the 2015 Staff Guidelines on IMF Staff Engagement with CSOs as a basis for interactions with external stakeholders.

### Resource implications and risks
- Recent activity and estimated baseline costs:
  - In the past 2 years, an estimated 8 in 10 country teams discussed social spending in their staff reports and 5 in 10 produced a box, selected issues paper or working paper (at an estimated annual cost of US$3 million).
- Survey of a random sample of 44 country teams:
  - 18 of 44 country teams surveyed anticipated needing more resources in FY20;
  - 26 of 44 country teams anticipated the same or less resources would be needed.
- Time inputs (person-weeks, per country):
  - FY18 Average 2.3; FY19 Average 3.3; FY20 Average 3.8.
  - Median: FY18 2; FY19 3; FY20 3.
  - Max: FY18 13; FY19 14; FY20 13.
  - Min: FY18 0; FY19 0; FY20 0.5.
  - Note: On average, teams spent 3.3 person-weeks in fiscal year 2019. Teams estimated spending 1 additional person-week per year in FY20 if the proposed strategy was implemented. In a few cases teams spent much more (up to 14 person-weeks; one country with 31 person-weeks was excluded from summary statistics).
- Estimated additional costs of implementing the strategy:
  - Recurring additional costs from increased country team engagement and functional department support: US$0.7 million (based on a random sample survey and split equally across area and functional departments).
  - Area department engagement estimate: US$0.36 million; would increase to about US$1 million if resource needs increase more rapidly than in the baseline (upper-end estimate calibrated on more extensive engagement).
  - One-off (transitional) costs during FY20–FY21 to develop tools and databases (largely in FAD), prepare How-to Notes, and conduct in-reach: US$1.3 million (to be financed through internal reallocation).
  - Functional departments baseline support: US$0.35 million (could rise to US$0.45 million under higher engagement scenario).
- Underlying assumptions and sensitivity:
  - Cost estimate based on the survey assumes engagement will deepen in 60 countries per year at a marginal cost of 1 person-week.
  - Upper-end estimate based on more extensive engagement in 20 countries per year points to marginal cost for area departments of up to 2.8 person-weeks or US$1 million and would increase demand for functional departments support from US$0.35 million to US$0.45 million.
- Expected risk mitigation:
  - If implemented, the strategy should reduce risks to surveillance and lending operations by improving consistency, strengthening program design and conditionality, mitigating adverse effects on vulnerable households, and improving quality of spending.
  - Proposed steps to support country teams and deepen engagement with IDIs would contain budget and human capital risks; strong communication would mitigate reputational risks.

### Conclusions (summary of intent and policy directions)
- Rationale:
  - Social spending is increasingly important for IMF members to address inequalities, protect vulnerable groups, smooth consumption over the life cycle, stabilize demand after shocks, and achieve 2030 SDG commitments.
- Observations:
  - Fund engagement on social spending increased significantly over the last decade but can be further strengthened.
  - Country teams regularly draw on IDI expertise but face barriers (knowledge of counterparts, IDIs’ longer-term focus).
  - Analysis shows spending has on average been protected but there are many instances where it declined; there has often been an overly narrow focus on means-tested targeting and insufficient focus on spending quality.
- Core elements of the proposed strategy:
  - Engagement guided by assessment of the macro-criticality of specific social spending issues (channels: fiscal sustainability, spending adequacy, spending efficiency).
  - Continue policy advice on sustainable financing of social spending and increase focus on the quality of spending, while relying more on IDI expertise.
  - Strengthen program design and conditionality to mitigate adverse effects on the vulnerable and address data and information gaps; inform conditionality by country’s macro-fiscal context and social objectives with a medium-term sequencing focus.
  - Strengthen support for country teams (Staff Guidance Note by end–2020 to present existing support and initiatives; Note to be periodically updated).
  - Improve collaboration with IDIs with FAD acting as a hub and encourage early engagement to align institutional priorities; seek closer engagement with other stakeholders.
  - Discuss advice on universal and targeted social benefits in the context of overall effectiveness of social safety nets, fiscal and administrative constraints, and social and political preferences; ensure progressive and efficient financing as coverage expands.
  - Focus external communication on explaining how IMF engagement fits into the overall medium-term macro-fiscal context and policy strategy, clarifying what the Fund can and cannot do.

### Issues for discussion (as presented)
- Do Directors agree that the proposed strategy provides appropriate guidance to staff on the Fund’s engagement on social spending issues?
- Do Directors concur that macro-criticality is an important guiding principle for when to engage on social spending issues?
- Do Directors support continued provision of policy advice by staff on sustainable financing of social spending and an increased focus on the quality of such spending?
- Based on existing good practice, do Directors agree that GRA-supported programs should analyze and, as appropriate, document the social impact of adjustment and measures to protect the vulnerable?
- Do Directors agree that advice on the use of targeted and universal transfers would be discussed in the context of the overall effectiveness of the social safety net, administrative constraints, the broader macro-fiscal environment, and countries’ social and political preferences?
- Do Directors agree to strengthen collaboration with IDIs with greater expertise in social spending issues?
- Do Directors support improving the clarity of communicating social spending issues, including by embedding them into a broader policy context?
- Do Directors agree that the nature of social spending issues is evolving and that the Fund should update its engagement as it acquires more experience?

*International Monetary Fund — “37. FAD could further strengthen support for country teams, resources permitting,” excerpt.*

### Annex I.   Examples of Social Spending Policy Issues and Support

### Annex I.   Examples of Social Spending Policy Issues and Support

### Pensions
- IMF Policy Analysis (Areas where Fund teams can provide specific policy advice)
  - Is pension spending projected to increase significantly under current policies (sustainability)?
  - Does the government have plans to expand pension coverage over the medium term (adequacy, efficiency)?
  - Is pension spending high in comparison to peers?
  - If so does this reflect demographics, low retirement ages, or generous pension benefits, in-payment and accrual of rights?
  - Is the pension system running deficits? Is it sustainable?
  - What are the efficiency and equity implications of reform options?
  - Are there currently large gaps in the pension system? Should gaps be filled with contributory (earnings related) or non-contributory (social) pension systems?
  - What are the spending implications of policy options over the short and medium terms?
  - What are the distributional implications of the current pension system?
  - What are the distributional implications of alternative reform options for containing pension spending?
- External Expertise (Areas where Fund teams do not typically provide specific policy advice)
  - How can new contributory schemes be designed and implemented, or existing ones be scaled up?
  - How can pension systems be expanded to cover difficult-to-reach groups?

### Social Safety Net (SSN)
- IMF Policy Analysis (Areas where Fund teams can provide specific policy advice)
  - Is SSN spending high/low compared to peers (efficiency; adequacy)?
  - Is SSN spending effective at achieving social objectives (efficiency)?
  - Does the SSN create strong disincentives to work (efficiency)?
  - Is the government planning to expand the coverage or generosity of the SSN (sustainability; adequacy)?
  - Does the government have the fiscal space and administrative capacity to strengthen the SSN?
- External Expertise (Areas where Fund teams do not typically provide specific policy advice)
  - Is high/low spending due to high/low coverage or high/low generosity?
  - Is spending progressive and coverage of vulnerable groups high?
  - Are means-tested benefits quickly withdrawn as income increases? Does an adequate tax registry exist, and is it effectively integrated with complementary databases?
  - What options is the government considering for strengthening the SSN, and what are the fiscal implications?
  - Does the country have the appropriate mix of programs?
  - Can the administrative system be improved to consolidate and streamline programs and administrative costs?
  - How can a beneficiary registry be developed and integrated into the implementation of the SSN?
  - How can the targeting mechanisms be designed to ensure adequate coverage of the vulnerable and minimal leakage to higher-income group?

### Health
- IMF Policy Analysis (Areas where Fund teams can provide specific policy advice)
  - Is health spending projected to increase significantly under current policies (sustainability, efficiency)?
  - Does the government have plans to expand health coverage over the medium term (adequacy)?
  - Is health spending high in comparison to peers?
  - If so does this reflect demographics, generous benefits, or spending inefficiency?
  - What are the efficiency and equity implications of reform options?
  - Are there currently large gaps in health coverage? Do plans exist to scale up access, consistent with achievement of the SDGs?
  - Should gaps be filled with expansion of health insurance coverage or general revenue financing?
  - What would be the spending and distributional implications of policy options over the short and medium terms?
- External Expertise (Areas where Fund teams do not typically provide specific policy advice)
  - How can the primary health care system and hospital system be better integrated to increase efficiency?
  - Can the system for paying providers be reformed to better incentivize efficiency?
  - Can the system for purchasing drugs be reformed to cut costs?
  - How can the insurance system be reformed to enhance coverage among lower-income and middle-income groups?

### Education
- IMF Policy Analysis (Areas where Fund teams can provide specific policy advice)
  - Has spending declined in line with population ageing? (efficiency; sustainability)?
  - Does the government have plans to expand coverage and increase spending (adequacy)?
  - Is spending high compared to peers?
  - Is spending efficient? Are teachers’ wages too high or too low?
  - Are there currently large gaps in education access? Do plans exist to scale up access, consistent with achievement of the SDGs? What are the fiscal implications of expansion plans?
  - Has the government the administrative capacity to deliver high quality education?
- External Expertise (Areas where Fund teams do not typically provide specific policy advice)
  - Is the education system capable of delivering quality education?
  - How can schools be consolidated to increase spending efficiency? What complementary policies are needed to reduce excessive inconvenience?
  - Is the system capable of ensuring quality education while expanding access?

*Annex I. Examples of Social Spending Policy Issues and Support*

---


_Source: https://www.imf.org/-/media/files/publications/pp/2019/ppea2019016.pdf_