Poverty and Social Impact Analysis in PRGF-Supported Programs

Contributor Notes

Author’s E-Mail Address: minchauste@imf.org

This paper aims to inform on the status of Poverty and Social Impact Analysis (PSIA) in IMF-supported programs, detailing the results presented in the recent review of PRGF-supported programs. The review showed that more needs to be done, both in undertaking PSIA when necessary, and in reporting the policy tradeoffs in program documents. Policy design should be continuously informed by the results of PSIA.

Abstract

This paper aims to inform on the status of Poverty and Social Impact Analysis (PSIA) in IMF-supported programs, detailing the results presented in the recent review of PRGF-supported programs. The review showed that more needs to be done, both in undertaking PSIA when necessary, and in reporting the policy tradeoffs in program documents. Policy design should be continuously informed by the results of PSIA.

I. Introduction

Poverty and social impact analysis (PSIA) is the assessment of intended and unintended consequences of policy interventions on the well-being of a country’s social groups, with a special focus on the vulnerable and the poor. Understanding the impact of policy changes on different groups is critical in designing effective poverty-reduction strategies. PSIA enables countries to evaluate different policy options before they are implemented, modify ongoing policies, or design countervailing measures to alleviate any adverse impact on poorer groups of the population.

An important feature of the IMF’s Poverty Reduction and Growth Facility (PRGF) is to include poverty and social impact analysis of measures that may have significant adverse effects on the poor. The recent PRGF review has documented the progress in implementing this key feature.2 The purpose of this paper is to inform on the status of PSIA in PRGF-supported programs and to provide details of the analysis presented in the recent PRGF review, with the aim of drawing lessons from the experience to date.

II. Background

The IMF has taken a progressively more active stance on social policies over the years, to ensure that they are well integrated into IMF-supported programs and IMF policy advice.3 In the early 1990s. IMF staff increasingly sought to incorporate social safety nets into adjustment programs, an approach that was endorsed by the IMF Executive Board in 1993. The purpose was to mitigate the adverse effects of program design on the poor. In 1997, IMF staff were issued guidelines for monitoring of social expenditures and social indicators, including the core set of international development goals. Programs supported by the Enhanced Structural Adjustment Facility (ESAF) sought general budgetary allocations for social safety nets. By the late 1990s, there had been some use of structural benchmarks and performance criteria to secure social protection objectives.4

The IMF has essentially relied on other institutions for advice on social issues. Collaboration with the World Bank has been critical for the implementation of the Highly Indebted Poor Countries (HIPC) Initiative, and more recently, in encouraging countries to develop country-owned Poverty Reduction Strategy Papers (PRSPs).5 More broadly, the IMF has intensified collaboration with other international agencies having social policy expertise and engaged representatives of civil society groups in dialogue on social concerns (see below).

Despite the IMF’s increased involvement in social issues, a 1997 review of ESAF-supported programs concluded that more was needed to be done to ensure that macroeconomic, structural, and social policies complement each other. In particular, both an internal review and external evaluation of the ESAF found that there was need for more comprehensive ex ante analysis of the likely social impact of key macroeconomic and structural reform measures.6 In addition, the internal review noted the need for appropriate social policy instruments before economic reforms are implemented, and for adequate follow-up of performance and monitoring of social safety nets during program implementation.

The Poverty Reduction and Growth Facility (PRGF) replaced the ESAF. A key feature of PRGF-supported programs is the inclusion of poverty and social impact analysis (PSIA) of major macroeconomic adjustment and structural reforms.7 Under the PSIA framework, analysis is undertaken with the objective of continually informing policy dialogue, choice, and implementation, within the broader objective of promoting sustainable poverty reduction and social inclusion.8 This represents a departure from previous practice, where the costs of social safety nets were incorporated into program targets as a way to mitigate the adverse impact of policies on vulnerable groups, but did not necessarily continually feed into the decision-making process, during and after program implementation.

PSIA would need to be integrated into poverty reduction strategy papers, since it enables countries to weigh the trade-offs of reforms, and to take ownership for the policies to be implemented.9 As a result, it is critical that country authorities take the lead in PSIA. However, given the data and technical capacity constraints in many developing countries, substantial technical assistance will be required.

III. PSIA in PRGF-Supported Programs To Date

How has PSIA been implemented in PRGF-supported programs to date? As detailed below, although there has been an improvement over ESAF-supported programs, there is scope for significant improvement.

A. Methodology for the Review

The PRGF review analyzed programs approved or reviewed by the Executive Board between July 1, 2000 and September 30, 2001. The sample includes countries that started a new three-year PRGF-supported program (19 cases of “new” PRGFs). It also includes countries that had ESAF-supported arrangements but were then “transformed” into PRGF-supported arrangements through two or more reviews or a review supported by a full PRSP during this period (16 cases of “transformed” PRGFs). For comparison purposes, the review also included the last annual arrangement under ESAF-supported programs for the same sample of countries.

The review first took stock of reforms that could potentially have a negative impact on various groups in countries with PRGF-supported programs. Then, it sought to identify instances where PSIA had been undertaken. However, this task is inherently complex, since there could be cases where PSIA could have been undertaken, but not reported in program documents. As a result, two definitions were adopted. The first defines PSIA broadly to include instances where social consequences of policies were considered in program documents, whether these were based on a simple qualitative analysis or supported by analytical studies. Whenever IMF program documents made reference to potential social consequences of policy actions, it was assumed that PSIA was undertaken. Because such analyses were inferred from reports, the possibility of over- or underestimating PSIA cannot be ruled out. In the second definition, PSIA is accounted for in instances where analysis supported by rigorous analytical studies are mentioned in IMF program documents.10

B. How Prevalent Are Measures with Potentially Negative Social Impacts?

All PRGF-supported programs have at least one type of reform with potentially negative effects on the poor (Figure 1). In particular, 94 percent of PRGF-supported countries included at least one type of macroeconomic reform; 100 percent included at least one type of structural reform; and 29 percent had to face an exogenous shock.

Figure 1.
Figure 1.

Reforms in PRGF Countries and Poverty and Social Impact Analysis

(In percent of PRGF countries, number of countries undertaking reform in parentheses)

Citation: IMF Policy Discussion Papers 2002, 011; 10.5089/9781451969856.003.A001

Source: IMF staff reports.1/ Includes reforms in international trade policy and exchange rate adjustments.2/ Includes health, education, and pension reforms.3/ Includes agricultural, financial, and legal reforms; changes in user fees; and decentralization measures.4/ Includes terms of trade shocks, natural disasters, and security/refugee crises.

Macroeconomic reforms included measures in revenue, expenditure, exchange rate, and international trade policies. On the structural front, the most common measures include privatization, civil service reform, and changes to domestic prices.

Among macroeconomic reforms, common measures include reforms in tax policy (63 percent of PRGF-supported countries) and expenditure policy (74 percent of PRGF-supported countries). Interestingly, new PRGF-supported programs more often contain increases in public spending than transformed PRGF-supported programs do. This coincides with the finding that PRGF-supported programs are accommodating higher public spending than ESAF-supported programs (Table 1 and Figure 1).11

Table 1.

Policy Reforms in ESAF- and PRGF-Supported Programs

(Percent of countries)

article image
Source: IMF staff reports.

Includes reforms in international trade policy and exchange rate adjustments.

Includes health, education, and pension reforms.

Includes agricultural, financial, and legal reforms; changes in user fees; and decentralization measures.

Includes terms of trade shocks, natural disasters, and security/refugee crises.

C. Poverty and Social Impact Analysis

The review found that over half of all PRGF-supported programs include some form of PSIA for one or more measures. When using the first definition described above, about 63 percent of program documents include instances of PSIA. About 40 percent of the documents make reference to PSIA either on macroeconomic measures and/or structural measures.12, 13

Qualitative analysis has taken many forms. For example, in Guinea, formal PSIA was not conducted, but IMF staff held a series of limited consultations with the government and stakeholders (including trade unions) before its current PRGF-supported program, therefore providing a qualitative assessment of the potential impact of reforms. In The Gambia, IMF staff discussed the potential impact of some of the key policies contained in the PRGF-supported program as a first step toward a qualitative analysis. The next step is to work closely with the government, as part of the full PRSP, in developing some quantification of the impact of individual policies, using data from both the household surveys and the participatory poverty assessments.

PSIA based on rigorous analytical studies were referred to in 34 percent of PRGF-supported programs.14 In particular, 17 percent of the programs mention studies on the impact of macroeconomic measures, while 23 percent of programs mention studies on the impact of structural measures.15 Examples of these more rigorous studies include a joint World Bank-government analysis of the impact on employment of state-owned enterprise reform in Vietnam. In Chad, the government undertook PSIA on the consequences of reform in the cotton sector. In Mozambique, the Food and Agriculture Organization (FAO) undertook a cost-benefit analysis of the proposed elimination of import surcharges for sugar, while the World Bank sponsored an assessment of competitiveness and employment in the cashew-processing industry.

A focus on the number of countries undertaking PSIA for one or more measures could overstate the prevalence of PSIA. This is because some country programs could have many measures with potentially negative effects where PSIA would be required. In this light, the review found that under the first definition, PSIA was undertaken for less than half of each particular reform measure (Table 2 and Figure 2).

Table 2.

PSIA and Safety Nets in ESAF- and PRGF-Supported Programs

(Percent of reforms)

article image
Source: IMF staff reports.

Includes reforms in international trade policy and exchange rate adjustments.

Includes health, education, and pension reforms.

Includes agricultural, financial, and legal reforms, changes in user fees, and decentralization measures.

Includes terms of trade shocks, natural disasters, and security/refugee crises.

Figure 2.
Figure 2.

Poverty and Social Impact Analysis (PSIA) in PRGF-Supported Programs and Countervailing Measures

(In percent of countries undertaking reforms, number of countries undertaking reforms in parentheses)

Citation: IMF Policy Discussion Papers 2002, 011; 10.5089/9781451969856.003.A001

Source: IMF staff reports.1/ Includes reforms in international trade policy and exchange rate adjustments.2/ Includes health, education, and pension reforms.3/ Includes agricultural, financial, and legal reforms; changes in user fees; and decentralization measures.4/ Includes terms of trade shocks, natural disasters, and security/refugee crises.

Specifically:

  • As noted earlier, macroeconomic reforms were undertaken in 94 percent of all PRGF-supported programs, of which 18 percent were accompanied by PSIA, and 12 percent by countervailing measures. Depending on the composition of the expenditure, contractionary expenditure policy can have a negative impact on the poor. Although the potential negative effects are discussed in 38 percent of countries undertaking contractionary expenditure policy, none of the PRGF-supported program documents present a rigorous study assessing the poverty and social impact. Given the broad nature of these reforms, there are no specific safety nets incorporated in programs to protect the most vulnerable. However, 77 percent of the programs that include expenditure cuts mention increases or protection of social spending as a way to offset the negative impact of lower overall spending.

  • Trade and exchange rate measures were undertaken in 46 percent of all PRGF-supported countries, of which 50 percent had some form of PSIA but only 13 percent incorporated a safety net. On the structural front, despite the large number of countries undertaking some form of civil service reform (77 percent of programs) and privatization (83 percent of programs), program documents report some form of PSIA in only 7 and 14 percent of these cases, respectively.

These results show the substantial scope for improvement in undertaking and reporting on PSIA in PRGF-supported programs. As shown in Figure 2, PSIA is conducted for less than 50 percent of all reforms with potentially negative social effects.

Notwithstanding the above, PSIA is more common in PRGF-supported programs than it was under ESAF-supported programs. The latter very rarely reported social impact analyses or discussed the potential social implications of policy measures. Only 3 to 4 percent of the ESAF-supported programs reviewed discuss PSIA undertaken for revenue measures, trade and exchange rate, privatization, or other structural reforms (Table 2). More striking is the fact that no ESAF-supported program reports on any PSIA related to contractionary expenditure policies, although 53 percent included such policies. However, 61 percent of these cases also included measures to protect or increase social spending. This compares to 77 percent of the cases in PRGF-supported programs. Although this is not a targeted countervailing measure, the aim was to partly ameliorate the potential negative impact of adjustment.

D. How Has PSIA Influenced Program Design?

PSIA has increasingly influenced programs in both the design of economic policies and the formulation of countervailing measures. In some cases, this has led to a modification of policies. For example, in Senegal, a 15 to 20 percent subsidy was maintained due to concerns regarding the impact of higher prices on the poor. In Mauritania, basic items (such as wheat, medications, and minimum levels of electricity and water consumption) were exempted from the VAT to protect the poor.

In other cases, PSIA has influenced the pace of implementation of reform measures. For example, in Cambodia, the large-scale retrenchment of civil servants was delayed until a safety net could be put in place. In Cameroon, the authorities advocated a gradual pass-through of higher world oil prices in order to cushion the impact on the poor. In Uganda, removal of the protection accorded to the sugar industry was delayed after PSIA outlined the potential negative social impact.

E. Countervailing Measures

About two-thirds of PRGF-supported programs include countervailing measures. In fact, one or more countervailing measures are present in 69 percent of PRGF-supported countries, up from 57 percent under ESAF-supported programs (Table 1). Countervailing measures are most common for structural reforms such as domestic prices or privatization, or in cases where there has been an exogenous shock (Table 2).16

The design of countervailing measures varies according to the type of reform. For example, most countervailing measures accompanying civil service or privatization reforms include some form of severance payments (Kenya, Vietnam, Mongolia). Tariff and other domestic price reforms are commonly accompanied by a safety net that could include cross-subsidies for poorer consumers (Albania, Ghana), or targeted poverty benefits in cases where individual information is available (Azerbaijan, Georgia). Finally, in cases where there is an exogenous shock, such as a drought or the end of a war, safety nets are put in place to give food aid or provide resources for demobilization.

When compared to ESAF-supported programs, PRGF-supported programs introduce countervailing measures more frequently for all types of reforms.17 About 56 percent of the ESAF-supported programs include some type of countervailing measure, with 18 percent aiming to ameliorate the impact of macroeconomic measures, 41 percent to reduce the negative impact of structural measures, and 15 percent to counteract negative exogenous shocks. In contrast, about 69 percent of PRGF-supported programs include countervailing measures, with macroeconomic (11 percent), structural measures (54 percent), and negative exogenous shocks (23 percent).18 It should be noted that in some cases, ESAF-supported programs have implemented general safety nets not in response to any specific reform, but intended nevertheless to protect the poor.19

Some PSIA in PRGF-supported programs is accompanied by specific countervailing measures, while in other cases no social impact analysis is discussed but countervailing measures are adopted. This may indicate that some qualitative assessment was undertaken, but not reported. It is possible that the proportion of programs with PSIA is underreported. In particular, countervailing measures were reported in some cases where there were increases in domestic prices, civil service reform, privatization, and changes in user fees, but there is no evidence of any PSIA (Table 2).

A broader definition of poverty and social impact analysis might address the possibility that there was some PSIA in countries that implemented countervailing measures. This broader definition counts the instances when there was either some kind of PSIA, or a countervailing measure, or both. The results reported in Figure 3 paint a more positive picture than presented under the second definition, but continue to show that less than 60 percent of measures were subjected to PSIA and/or a countervailing measure

Figure 3.
Figure 3.

PSIA or Countervailing Measures

(In percent of countries undertaking reforms, number of countries undertaking reforms in parentheses)

Citation: IMF Policy Discussion Papers 2002, 011; 10.5089/9781451969856.003.A001

Source: IMF staff reports.1/ Includes reforms in international trade policy and exchange rate adjustments.2/ Includes health, education, and pension reforms.3/ Includes agricultural, financial, and legal reforms; changes in user fees; and decentralization measures.4/ Includes terms of trade shocks, natural disasters, and security/refugee crises.

F. Regional Comparison of PSIA

There are some important differences across regions in terms of PSIA. First, PSIA (broadly defined) is more common in African countries than transition economies. In particular, 78 percent of the reforms in Africa contain some form of PSIA, compared to 27 percent for transition economies (Figure 4). Both qualitative and rigorous types of PSIA are more frequent in Africa.

Figure 4.
Figure 4.

PSIA and Countervailing Measures

(In percent of reforms)

Citation: IMF Policy Discussion Papers 2002, 011; 10.5089/9781451969856.003.A001

Source: IMF staff reports.1/ Includes reforms in international trade policy and exchange rate adjustments.2/ Includes health, education, and pension reforms.3/ Includes agricultural, financial, and legal reforms; changes in user fees; and decentralization measures.4/ Includes terms of trade shocks, natural disasters, and security/refugee crises.5/ Includes Albania, Armenia, Azerbaijan, Cambodia, Georgia, Lao PDR, Macedonia FYR, Moldova, Mongolia, Tajikistan, and Vietnam.

Countervailing measures, on the other hand, are more common in transition economies, with 73 percent of reforms accompanied by a safety net, versus 65 percent for African countries. This could reflect greater resource constraints in African countries, the availability of more social policy instruments (such as social insurance programs and social assistance), and the active involvement of the World Bank in transition economies. It could also indicate that some form of PSIA was carried out prior to the introduction of safety nets in transition countries but was not reported in IMF program documents. Finally, several PRGF documents for transition economies refer to general social safety nets that cannot be attributed to any single reform measure.

Countervailing measures vary across regions. Most of the safety nets in transition economies are related to reforms of the civil service, privatization, and reforms in the social sectors—consistent with the types of reforms they were undertaking. In contrast, the implementation of safety nets in African countries is more widespread across different reform measures. In all regions, most programs have provided safety nets for cases where there have been exogenous shocks, such as natural disasters, security or refugee crises, or terms-of-trade shocks.

IV. Summary

The review of PSIA in IMF program documents found that over half of the PRGF-supported countries had some form of PSIA, broadly defined to include simple qualitative analysis as well as more rigorous qualitative and quantitative approaches. However, when looking across reform measures, it is clear that much more remains to be done as countries typically undertook more than one measure with potential negative impacts. There is room for improvement, both in undertaking PSIA when necessary and in reporting the policy tradeoffs in program documents. Countervailing measures are common in PRGF-supported countries, even when there is no PSIA reported in IMF program documents, pointing to the possibility that some form of PSIA took place, but was not discussed. Finally, the review found that there are some regional differences, with PSIA being more common in Africa, and countervailing measures being more common in transition economies, particularly for privatization, civil service reform, and the social areas. These results point to the need for policy design and implementation to be more explicitly informed by the results of PSIA.

In future, PRGF-supported program reports are expected to include a description of PSIA being carried out in the country and a summary of IMF staff discussions with the country authorities, indicating how these have affected policy choices and the design of any accompanying countervailing measures.20 Several PSIA pilot studies have been undertaken by the World Bank and the Department for International Development (DFID) over the last few months.21 The results of these studies should be useful in guiding the program design and implementing countervailing measures in the context of a sustainable macroeconomic framework.

APPENDIX TABLES

Table 3.

Examples of Social Impact Analysis in Recent PRGF-Supported Programs

article image
article image
Table 4.

Examples of Countervailing Measures Implemented in Recent PRGF-Supported Programs

article image
Table 5.

Examples of PSIA and Countervailing Measures Implemented in ESAF-Supported Programs

article image
Table 6.

Examples of General Social Safety Nets Implemented in Countries with ESAF-Supported Programs

article image

References

  • Abed, George, and others, 1998, Fiscal Reforms in Low Income Countries: Experience Under IMF-Supported Programs, IMF Occasional Paper No. 160 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Gupta, Sanjeev, and others, 2000, Social Issues in IMF-Supported Programs, IMF Occasional Paper No. 191 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Gupta, Sanjeev, and others, 2003, Has the PRGF Lived Up to Expectations? IMF Occasional Paper No. 216 (Washington; forthcoming).

  • International Monetary Fund, 1998, External Evaluation of the ESAF: Report by a Group of Independent Experts (Washington).

  • International Monetary Fund, 2000, “Key Features of IMF Poverty Reduction and Growth Facility (PRGF) Supported Programs,” August 16. Available via the Internet: http://www.imf.org/external/np/prgf/2000/eng/key.htm

    • Search Google Scholar
    • Export Citation
  • International Monetary Fund, 2002a, “IMF Executive Board Reviews the Poverty Reduction and Growth Facility (PRGF),” Public Information Notice (PIN) No. 02/30, March 8. Available via the Internet: http://www.imf.org/external/np/sec/pn/2003/pn0230.htm

    • Search Google Scholar
    • Export Citation
  • International Monetary Fund, and World Bank, 2001a, “Social Impact Analysis of Macroeconomic and Structural Policies,” Joint Implementation Committee Factsheet, April 26. Available via the Internet: http://www.imf.org/external/np/exr/facts/sia.htm

    • Search Google Scholar
    • Export Citation
  • International Monetary Fund, and World Bank, 2001b, “Guidelines for Joint Staff Assessment of a Poverty Reduction Strategy Paper,” Annex 2 of Poverty Reduction Strategy Papers—Progress in Implementation. Available via the Internet: http://www.imf.org/external/np/prsp/2001/042001.htm.

    • Search Google Scholar
    • Export Citation
  • International Monetary Fund, and World Bank, 2003, Poverty Reduction Strategy Papers—Progress in Implementation (Washington; forthcoming).

    • Search Google Scholar
    • Export Citation
  • World Bank, 2002, A User’s Guide to Poverty and Social Impact Analysis, draft work in progress. Available via the Internet: http://www.worldbank.org/poverty/psia/draftguide.pdf

    • Search Google Scholar
    • Export Citation
3

For a review of the evolution of the IMF’ social policy advice see Gupta and others, 2000.

4

Ibid.

5

For a description of the PRSP approach see http://www.imf.org/extcrnal/np/exr/facts/prsp.htm

10

This approach did not assess the extent of public debate on PSIA and the associated risks, or the quality of the PSIA, but rather focused on quantifying the number of instances where PSIA was present. For an in-depth review of the broader issues involved in PSIA, see World Bank, 2003.

11

See Gupta and others, 2003 (forthcoming).

12

Examples of PSIA in PRGF documents are given in Appendix Table 3.

13

Note that in some instances countries undertook PSIA for more than one measure, therefore numbers do not necessarily add up.

14

The analysis could be extended beyond economic groups, to different social groups, and there are different tools and techniques for undertaking this analysis. However, this paper does not review these tools and techniques, but rather takes stock of whether any type of analysis was undertaken at all.

15

Note that in some instances countries undertook PSIA for more than one measure, therefore numbers do not necessarily add up.

16

Examples of countervailing measures in PRGF documents are given in Appendix Table 4.

17

For examples of PSIA and countervailing measures under ESAF-supported programs, see Appendix Table 3.

18

Countries could have countervailing measures for more than one measure. Therefore, the numbers do not necessarily add up to the totals.

19

See examples in Appendix Table 4

21

The World Bank is conducting PSIA in Chad, Guyana, the Kyrgyz Republic, Malawi, Mongolia, Pakistan, and Madagascar. DFID is supporting studies in Rwanda, Mozambique, Uganda, Honduras, Armenia, Indonesia, and in the State of Orissa, India.

Poverty and Social Impact Analysis in PRGF-Supported Programs
Author: Ms. Gabriela Inchauste
  • View in gallery

    Reforms in PRGF Countries and Poverty and Social Impact Analysis

    (In percent of PRGF countries, number of countries undertaking reform in parentheses)

  • View in gallery

    Poverty and Social Impact Analysis (PSIA) in PRGF-Supported Programs and Countervailing Measures

    (In percent of countries undertaking reforms, number of countries undertaking reforms in parentheses)

  • View in gallery

    PSIA or Countervailing Measures

    (In percent of countries undertaking reforms, number of countries undertaking reforms in parentheses)

  • View in gallery

    PSIA and Countervailing Measures

    (In percent of reforms)