Document of The World Bank FOR OFFICIAL USE ONLY INTERNATIONAL DEVELOPMENT ASSOCIATION PROGRAM DOCUMENT FOR A PROPOSED GRANT

Transcription

1 Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized Document of The World Bank FOR OFFICIAL USE ONLY INTERNATIONAL DEVELOPMENT ASSOCIATION PROGRAM DOCUMENT FOR A PROPOSED GRANT IN THE AMOUNT OF SDR 3.9 (US$6 MILLION EQUIVALENT) TO THE REPUBLIC OF RWANDA FOR A SECOND COMMUNITY LIVING STANDARDS GRANT FEBRUARY 16, 2010 Human Development Department Eastern Africa Country Cluster 2 Country Management Unit Africa Region Report No RW This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.

2 Rwanda - GOVERNMENT FISCAL YEAR July 1 June 30 CURRENCY EQUIVALENTS (Exchange Rate Effective as of December 31, 2009) Currency Unit = Rwandan Franc US$ 1.00 = RWF SDR = US$ Weights and Measures Metric System ABBREVIATION AND ACRONYMS AfDB Africa Development Bank CAS Country Assistance Strategy CFAA Country Financial Accountability Assessment CHW Community Health Workers CLSG Community Living Standards Grant CPAF Common Performance Assessment Framework DCDP Decentralization and Community Development Program DFID United Kingdom s Department for International Development DHS Demographic and Health Surveys DPO Development Policy Operation DRC Democratic Republic of Congo DS Direct Supports EC European Commission EDPRS Economic Development and Poverty Reduction Strategy EICV2 Household Living Standard Survey ( ) EU European Union FDLR Rwandan Democratic Liberation Forces GDP Gross Domestic Product GoR Government of Rwanda GTZ German Technical Cooperation HIDA Human Resources and Institutional Capacity Development Agency HIPC Heavily Indebted Poor Countries HIV/AIDS Human Immunodeficiency Virus/Acquired Immune Deficiency Syndrome HNP Health, Nutrition and Population HSSP Health Sector Strategy Plan IDA International Development Association IMF International Monetary Fund ITN Insecticide Treated Nets LDP Letter of Development Policy M&E Monitoring and Evaluation MAP Multi-Country Action Program MDGs Millennium Development Goals MINALOC Ministry of Local Government MINAGRI Ministry of Agriculture MININFRA Ministry of Infrastructure i

3 FOR OFFICIAL USE ONLY MINITERE MINECOFIN MOH MOU MTEF NBR OVC OPCS PBF PEFA PER PFM PRGF PRSG PRSP PW RPPA RWF SACCO SDR SIDA SP SWAp UN UNDP UNICEF US$ USAID VUP WFP Ministry of Lands, Environment, Forestry, Water, and Natural Resources Ministry of Finance and Economic Planning Ministry of Health Memorandum of Understanding Medium-Term Expenditure Framework National Bank of Rwanda Orphans and Vulnerable Children Operations Policy and Country Services Performance-based Financing Public Expenditure and Financial Accountability Public Expenditure Review Public Financial Management Poverty Reduction Growth Facility Poverty Reduction Support Grant Poverty Reduction Strategy Paper Public Works Rwanda Public Procurement Agency Rwandan Franc Savings and Credit Cooperative Organization Special Drawing Rights Swedish International Development Agency Social Protection Sector-Wide Approach United Nations United Nations Development Program United Nations Children s Fund United States Dollars United States Agency for International Development Vision 2020 Umurenge Program World Food Program Vice President: Country Director: Sector Manager: Task Team Leader: Obiageli K. Ezekwesili Johannes C. M. Zutt Lynne D. Sherburne-Benz Verdon S. Staines ii This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not be otherwise disclosed without World Bank authorization.

4

5 THE REPUBLIC OF RWANDA SECOND COMMUNITY LIVING STANDARDS GRANT TABLE OF CONTENTS GRANT AND PROGRAM SUMMARY... v INTRODUCTION... 1 COUNTRY CONTEXT... 3 RECENT ECONOMIC DEVELOPMENTS IN RWANDA... 3 MACROECONOMIC OUTLOOK AND DEBT SUSTAINABILITY... 5 THE GOVERNMENT S PROGRAM... 7 ECONOMIC DEVELOPMENT AND POVERTY REDUCTION STRATEGY (EDPRS) 7 VISION 2020 UMURENGE PROGRAM... 7 HEALTH SECTOR REFORM... 8 BANK SUPPORT TO THE GOVERNMENT S PROGRAM... 9 LINK TO CAS... 9 COLLABORATION WITH THE IMF AND OTHER DONORS... 9 RELATIONSHIP TO OTHER BANK OPERATIONS LESSONS LEARNED ANALYTICAL UNDERPINNINGS THE PROPOSED RWANDA SECOND COMMUNITY LIVING STANDARDS GRANT OPERATION DESCRIPTION POLICY AREAS OPERATION IMPLEMENTATION POVERTY AND SOCIAL IMPACTS ENVIRONMENTAL ASPECTS IMPLEMENTATION, MONITORING AND EVALUATION FIDUCIARY ASPECTS DISBURSEMENT AND AUDITING RISKS AND RISK MITIGATION ANNEXES ANNEX 1: LETTER OF DEVELOPMENT POLICY ANNEX 2: RWANDA COMMUNITY LIVING STANDARDS GRANT POLICY MATRIX ANNEX 3: FUND RELATIONS NOTE ANNEX 4: COUNTRY AT A GLANCE (includes country map) ANNEX 5: RECENT ECONOMIC DEVELOPMENTS & MEDIUM TERM PROSPECTS ANNEX 6: OVERVIEW OF UBUDEHE ANNEX 7: DECENTRALIZATION REFORMS IN RWANDA ANNEX 8: SOCIAL PROTECTION ANALYTICAL UNDERPINNINGS ANNEX 9: VUP IMPLEMENTATION PROCESS FOR RWANDA s NEW SP POLICIES ANNEX 10: INTERNATIONAL GOOD PRACTICE IN PUBLIC WORKS PROGRAMS ANNEX 11: INTERNATIONAL GOOD PRACTICE IN DIRECT SUPPORTS PROGRAMS ANNEX 12: SUMMARY OF VUP PUBLIC AWARENESS PROGRAM ANNEX 13: EXTRACT FROM VUP MONITORING & EVALUATION MANUAL ANNEX 14: UMURENGE SACCO STRATEGY EXECUTIVE SUMMARY ANNEX 15: RISKS OF USING SACCOS AS A CASH PAYMENT MECHANISM ANNEX 16: COMMUNITY PERFORMANCE-BASED FINANCING OF HEALTH ANNEX 17: ENVIRONMENTAL ASPECTS IN THE RWANDAN CONTEXT iii

6

7 TABLES TABLE 1: Rwanda: Definition of Poverty... 4 TABLE 2: Selected Economic Indicators, TABLE 3: Analytical Underpinnings for the Second Community Living Standards Grant TABLE 4: Prior Actions for the Second Proposed Operation and Triggers for the Third Operation.. 18 TABLE 5: Key Social Protection Performance and Monitoring Indicators in VUP Sectors TABLE 6: Community Health, Nutrition, and Population (HNP): Key Indicators TABLE 7: Intermediate Outcome Indicators to Link HNP Policy Actions with M&E Framework TABLE 8: Results-based Financing Impact Evaluation Design TABLE 5.1: Real Sector Growth and Contribution to GDP, TABLE 5.2: Selected Economic Indicators, TABLE 5.3: Consolidated Recurrent Expenditures, (% of total, incl. debt flows) TABLE 5.4: Donor Budget Support Commitments, (US$ million) TABLE 5.5: Financing Plan, (million US$) TABLE 6.1: Vision 2020 Umurenge-Targeting the most Disadvantaged Umurenge in each District TABLE 7.1: New Administrative Structures TABLE 8.1: Resource Gap to Reduce Moderate and Absolute Poverty in Rwanda TABLE 10.1: Program Wage Relative to Minimum and Market Wages in Public Works Schemes TABLE 10.2: Labor Intensity in Public Works Schemes BOXES BOX 1: Good Practice Principles on Conditionality BOX 5.1: Rwanda: Update of Selected DSA Indicators BOX 10.1: What are the Key Design Features of a Good Public Works Program? BOX 10.2: Targeting Effectiveness of Public Works Program: International Experience BOX 16.1: Development of the National PBF Model FIGURES FIGURE 5.1: Trend in Inflation FIGURE 5.2: Exports, Imports, and BOP Balance, (in US$ million) FIGURE 5.3: Central Government Expenditures (% of total) FIGURE 5.4: Donor Budget Support Commitments, (US$ million) FIGURE 5.5: Financing Plan, (million US$) FIGURE 16.1: The Purchaser-Provider Split in the National PBF Model FIGURE 16.2: Flow of Funds in the Community PBF FIGURE 16.4: Flow of Information in the Community PBF FIGURE 16.5: Summary of Incentives in the Community PBF and Links to VUP and Facility PBF The Second Community Living Standards Grant was prepared by a World Bank team consisting of Verdon Staines, Senior Economist, AFTSP (TTL); Gayle Martin, Senior Economist, AFTHE; Patrick Mullen, Senior Health Specialist, AFTHE; Josiane Niyonkuru, Team Assistant, AFMRW; Gert Van Der Linde, Lead Financial Management Specialist, AFTFM; Michael Karangwa, Consultant, AFTHE; Diego Garrido Martin, Monitoring & Evaluation Specialist, AFTRL; Anne Anglio, Senior Program Assistant, AFTSP Nanak Kakwani, Monitoring and Evaluation Consultant, AFTSP; Astania Kamau, Program Assistant, AFTSP; Alex Kamurase, Social Protection Specialist, AFTSP; Annika Kjellgren, Consultant, AFTSP; Kalanidhi Subbarao, Social Protection Consultant, AFTSP; Alan Thompson, Social Assistance Consultant, AFTSP; Aissatou Diallo, Finance Officer, CTRFC; and Sameena Dost, Senior Counsel, LEGAF. iv

8

9 GRANT AND PROGRAM SUMMARY THE REPUBLIC OF RWANDA SECOND COMMUNITY LIVING STANDARDS GRANT (CLSG-2) Borrower Implementing Agencies Financing Data Operation Type Main Policy Areas Key Outcome Indicators The Republic of Rwanda Ministry of Finance, Ministry of Local Government, and Ministry of Health Type: IDA grant. Terms: Single tranche to be released on effectiveness of the operation. Amount: SDR 3.9 million (US$6 million equivalent). CLSG-2 is the second of three programmatic development policy grant operations of similar size, approximately one year apart. Social Protection and Community Health, Nutrition, and Population. Social Protection: After three years, the Vision 2020 Umurenge Program (VUP) 1 supported by this grant will be assessed with reference to (i) of those households eligible for support, the percentage that are granted public works in a sample of VUP pilot Sectors (2008: 35%; 2009: 35%; 2011: 35%); (ii) of those households eligible for support, the percentage that are granted direct support in a sample of VUP pilot Sectors (2008: 0%; 2009: 60%; 2011: 80%); and (iii) the proportion of households in the bottom two categories of extreme poverty in a sample of VUP pilot Sectors (2008: 28.7%; 2009:< 28.7%; 2011: < 28.7%). Community Health, Nutrition, and Population (HNP): At the end of three years, HNP policy reforms will be assessed with reference to (iv) increase in the proportion of married women aged years using modern contraceptives (2008: 27%; 2011: 38%); (v) increase in the proportion of children under-five with diarrhea who receive oral rehydration therapy (2008: 30%; 2011: 50%); (vi) increase in the proportion of births that are cared for in an accredited health facility (2008: 45%; 2011: 50%); and (vii) increase in the annual per capita allocation to performance-based financing (PBF) for health facilities and community health cooperatives (2008: US$1.45; 2011: US$2.25). Program Development With one exception, these indicators are part of the Common Performance Assessment Framework (CPAF), which governs general budget support for the Government of Rwanda (GoR). 2 All indicators are agreed between GoR and development partners. The Program Development Objective is to support the GoR s social protection and health policy reforms designed to reduce extreme 1 A flagship program under the Government s Economic Development and Poverty Reduction Strategy. 2 The exception is Utilization of oral rehydration therapy, which is an intermediate outcome indicator relating to both child mortality and nutritional status, key outcomes of the Second Health Sector Strategy Plan (HSSP-II). v

10

11 Objective(s) and Contribution to Country Assistance Strategy (CAS) Risks and Risk Mitigation Operation ID poverty, initially in 30 pilot Sectors, and to expand access to highimpact health, nutrition and population interventions at the community level. The operation will support GoR s poverty reduction objectives, consistent with the country s fiscal framework and the second strategic theme of the CAS, which is to reduce social vulnerability and to ensure that the most vulnerable Rwandans also benefit from growth and to help Rwanda make further progress in building a more stable society. A smaller program will support the Flagship Vision 2020 Umurenge initiative; help Rwanda get on track on child and mortality goals 3 The main risks to the program arise from: (i) continued global economic slowdown; (ii) continued instability in the Great Lakes subregion and its effects on economic stability and donor support; and (iii) implementation challenges (including capacity challenges and the use of SACCOs 4 to make payments to beneficiaries). The first two risks are mitigated by increasing the predictability of aid flows specifically, by applying a higher proportion of international assistance on-budget and allocating more predictable financing towards the highest-priority activities. The GoR is also working closely with IMF to assess policy measures to help mitigate the impact of the global financial crisis. Related efforts include a Memorandum of Understanding (MoU) and Common Performance Assessment Framework (CPAF) agreed between GoR and donors to harmonize support and govern the provision of direct budget support for implementation of Rwanda s Economic Development and Poverty Reduction Strategy (EDPRS). In the health sector, the risk to implementation due to capacity constraints is mitigated by support to the GoR by the Bank, UNICEF, USAID and other donors. On the social protection side, the risk evolving from the use of SACCOs to make payments to beneficiaries under the VUP program is being raised with GoR. Implementation risks under the CLSG operation will be further addressed through the provision of support to assess the VUP pilots and their impact on targeted households, and to increase the sustainability of program activities within budget constraints. P Rwanda Country Assistance Strategy (CAS) Table 6 of the CAS refers to Health/Social Protection DPL (development policy lending), which is the CLSG operation. 4 Savings and Credit Cooperative Organizations are local cooperative financial institutions set up as legal entities to enable groups of individuals to save and invest money and to receive loans (see Annexes 14 and 15). vi

12

13 IDA PROGRAM DOCUMENT FOR A PROPOSED SECOND COMMUNITY LIVING STANDARDS GRANT TO RWANDA INTRODUCTION 1. Rwanda is one of Africa s most densely populated countries, with its modest geographical size accommodating an estimated population of 9.6 million. 5 It is also among the poorest, despite having made significant progress in the past decade. The proportion of the population that national household surveys identified as poor decreased slightly from 60 percent in to 57 percent in However, because of relatively rapid population growth estimated at 2.7 percent per year and an increase in income inequality, the estimated total number of people living in poverty rose from 4.8 million in to 5.4 million in More than 90 percent of the poor live in rural areas, and poverty remains deep and pervasive. 2. Despite strong economic growth, progress towards the Millennium Development Goals (MDGs) is uneven. Rwanda is on track to achieve the MDGs in access to primary education, gender parity, malaria incidence, and HIV prevalence. 6 However, with a poverty rate of 57 percent, considerable effort will be needed to meet the MDGs to reduce poverty and hunger, maternal mortality, and child mortality. 3. Rwanda s Economic Development and Poverty Reduction Strategy (EDPRS) for builds on strong achievements in human capital development and promotes three flagship programs to accelerate growth and reduce poverty. These flagship programs focus on (a) encouraging a pattern of growth that creates jobs and increases exports; (b) strengthening governance; and (c) accelerating poverty reduction by promoting pro-poor components of the national growth agenda and releasing the productive capacity of the poor in rural areas through a combination of public works, direct support and credit packages. 7 The Vision 2020 Umurenge 8 Program (VUP), announced in March 2007 and subsequently incorporated in EDPRS, constitutes the flagship poverty reduction program. A pilot that served as a precursor to VUP was also included in the Bank s Third Poverty Reduction Support Grant operation in FY2007. Additional complementary interventions to achieve EDPRS objectives include improving health status and slowing down population growth by increasing access to basic Health, Nutrition and Population (HNP) services. In support of 5 The population density was estimated in mid-2008 at 364 persons per square kilometer, the highest in the Africa Region (with the exception of one island in the Comoros). Demographic Growth and Development Prospects in Rwanda: Implications for the World Bank, June The Government of Rwanda s Ministry of Health and World Bank, Rwanda: A Country Status Report on Health and Poverty, June 2009 (Draft). 7 The Government of Rwanda s Economic Development and Poverty Reduction Strategy ( ). 8 The Government has progressively devolved responsibilities and transferred resources to decentralized levels of the administration Province, District, Sector (Umurenge), Cell (Akagari), and Village (Umudugudu). Of these, the last three are subordinate tiers of the district governments. The VUP and community health reforms are being implemented through this newly restructured, decentralized local government system. Also see Annex 7. 1

14 these objectives, the Government has initiated a set of policy reforms to expand access to basic health services at the community level. 4. The Bank s Country Assistance Strategy (CAS) for Rwanda has, as its second strategic theme, a reduction in social vulnerability by ensuring that the most vulnerable Rwandans also benefit from growth and by helping Rwanda make further progress in building a more stable society. 9 The proposed operation is the second phase of a three-year Community Living Standards Grant (CLSG) development policy operation that is consistent with the second CAS theme. It supports policy reforms in the public works and direct support components of the VUP and will also help Rwanda get on track with child and maternal mortality goals. It aims to achieve quick results in the areas of poverty reduction and improved health outcomes, even in the midst of the ongoing global economic and financial crisis. The operation s Program Development Objective is to support GoR s social protection and health policy reforms designed (i) to reduce extreme poverty, initially in 30 pilot Sectors, and (ii) to expand access to high-impact health, nutrition and population interventions at the community level. 5. Under the First Community Living Standards Grant (CLSG-1), implemented during the first year of the EDPRS ( ), policy actions were focused on ensuring that: Reforms followed international good practice in the domain of social protection policies, including the adoption of a non-distortionary wage policy for public works, direct payment of workers without intermediaries, 10 and determination of direct support benefits to poor families with no adult labor based on the household as a unit; and Health policies expanded poor households access at the community level to high-impact HNP interventions, 11 through the adoption of a community HNP service package, a community health policy implementation plan, a community performance-based financing strategy, and a medium-term expenditure framework (MTEF). 6. The proposed Second Community Living Standards Grant (CLSG-2) continues support in these areas over On the social protection side, policy actions will focus on ensuring that the Government s reforms include voluntary savings, and a process for determining eligibility that is clear and transparent to all potential recipients and non- 9 World Bank Rwanda s Country Assistant Strategy ( ), page Since December 2008, Rwandan policy has required that all wage payments and other payments made by the government to individuals and other private entities be made by direct deposit into an account for the individual or entity in a bank or financial institution. An individual entitled to receive such payments must have or create an account in a bank or financial institution into which the payment can be made. In the context of this operation, and within the Vision 2020 Umurenge Program (VUP), payment of wages under the public works program and of household benefits under the direct support program will be into such an account. Furthermore, the policy requires that private contractors are neither involved in making payments to VUP workers or beneficiaries nor in selecting the participants in the VUP public works or direct support programs, nor in determining their wage payments. However, private contractors can be used within the VUP for other purposes such as providing technical advice or supervising specific activities. 11 High-impact interventions are those that have had high impact on reducing disease burden in developing countries. Examples include: use of insecticide-treated bed-nets (ITN), integrated management of childhood illnesses, and the expanded program on immunization. 2

15 recipients because it makes program policy and procedures widely known and selection outcomes publicly available. On the HNP side, policy actions will focus on the creation of an incentive strategy to encourage demand for maternal health services, skills development for community health workers (CHWs), and design of a randomized evaluation to assess the impact of these reforms. 7. The second single-tranche development policy operation (DPO) is proposed to total SDR 3.9 million IDA financing. Co-financing of the program is expected from the Health Results Innovation Trust Fund. The first operation was SDR 4 million IDA financing, with co-financing from the Health Results Innovation Trust Fund. The proposed series of three operations is estimated to total SDR 12 million (of the order of US$18 million) IDA financing and potentially US$10 to 12 million from the Health Results Innovation Trust Fund. COUNTRY CONTEXT 12 RECENT ECONOMIC DEVELOPMENTS IN RWANDA 8. Rwanda s US$4.5 billion economy has remained resilient over the last five years, sustaining macroeconomic stability and real GDP growth despite the global economic slowdown and exogenous shocks that affected agricultural productivity. However, compared to other countries in the region, and Sub-Saharan Africa on average, Rwanda has experienced a relatively more severe reduction in growth trends in 2009, which poses a serious threat to poverty reduction. The 3 to 4 percentage point reduction in GDP growth, from 5-year historical trends to levels estimated for 2009, is due both to the global crisis and to domestic liquidity issues. Lower projected growth between 2008 and 2009 results from lower growth in agriculture and a substantial stagnation in the industry and services sector. 9. Overall, the global financial crisis, has led to lower export performance, GDP growth and revenues. The impact of the global economic and financial crisis is expected to be manifested through three main channels: (a) declining external demand for Rwandan goods and services (mainly for tourism and mineral exports); (b) slumping commodity prices, particularly for minerals; and (c) scarcer financial inflows, including foreign direct investment. 10. Prudent implementation of macro-economic policies and protection of priority spending have enabled the Government to manage the impact of the crisis. Inflationary pressures are expected to decrease in the rest of 2009 as international commodity and food prices fall and government manages the exchange rate, leading to its improved flexibility while increasing the coordination between monetary and fiscal policies. Import growth exceeded the increase in exports, leading to a deterioration of the balance of payments in 2008 and The current account deficit widened in 2009 compared to 2008 due, in part, to a worsening in the terms of trade. (For further information about recent economic developments in Rwanda, please refer to Annex 5). 12 This section draws heavily from the program document for Rwanda s Sixth Poverty Reduction Support Grant (PRSG-6) and is consistent with its macro framework. 3

16 11. The expansionary fiscal stance of the past year or two has moderated in A gradual reduction in resources was planned for the period 2009/10, based on anticipated lower levels of revenues and external resources (compared to disbursements in 2008). In the financial sector, aside from the liquidity ratio, financial indicators remained within their prudential benchmarks; and the stability of the banking sector has at present not been adversely affected by the global financial situation. Rwanda s financial sector moved from a situation of structural excess liquidity to a potential structural liquidity deficit, partly due to efficiency gains from ongoing reforms in the sector as well as National Bank of Rwanda s (NBR) low interest rate policy. The liquidity shortages exposed weaknesses in risk management and governance of commercial banks, i.e. a mismatch in credit maturities (for bank borrowing and lending). In response to the tightening liquidity situation, the NBR injected liquidity of approximately 1.5 percent of GDP into the economy. Reserve money remains the anchor of the monetary program. 12. Poor people have grown in number and live mostly in rural areas. The country s original target for MDG 1 was to reduce the proportion of the population living under the poverty line to 30 percent by The GoR s more recent EDPRS aims to lower the share of the population living under the upper poverty line to 46 percent, and those living in extreme poverty to 24 percent by 2012 (see Table 1). Based on the latest available household survey (EICV2, ), which preceded recent policy initiatives, only modest improvements in poverty levels were achieved in the period Rwanda extreme poverty line Defined as per capita consumption equivalent to less than RWF 175 (around US$0.29) per day 13. This represents the level of expenditure needed to provide minimum food requirements of 2,100 kilocalories per adult per day. In , one-third of the population, or about 3.5 million people, (93 percent of them living in rural areas), did not reach this level of average daily consumption. Table 1 Rwanda: Definition of Poverty Rwanda upper poverty line Defined at RWF 250 equivalent (around US$0.45) daily consumption per adult 14. This indicates that some 57 percent of the population, or 5.4 million people, were under this line in , of whom 92 percent lived in rural areas. Poverty has been modestly reduced from 60.4 percent of the population under the upper poverty line in to 56.9 percent in Of Rwanda s estimated population of 9.6 million in mid-2008, 44 percent are below the age of 15. With a current population growth rate estimated at 2.7 percent per year, leading to a doubling of the population in 26 years, 15 and with the second-highest population density in the Africa Region, Rwanda is facing increasing pressure to address challenges of poverty, nutrition, health and population in an increasingly youthful population that contributes to population momentum. 14. Child mortality and other health indicators have improved significantly since 2000, although malnutrition, maternal mortality, and socio-economic disparities in health outcomes remain concerns. Rwanda has made considerable progress in a number of health indicators in recent years, including under-five mortality, which declined from 196 per 13 Vision 2020 Umurenge Programme (VUP) Monitoring and Evaluation Framework, page 10, June (2006 base) 14 Ibid. 15 Population Reference Bureau,

17 1,000 live births in 2000 to 103 per 1,000 in Considerable progress is particularly evident in malaria control, with a reported two-thirds decrease in incidence since The use of modern contraceptives increased dramatically, from 4 percent to 27 percent, between 2000 and 2008, and the proportion of deliveries in health facilities rose from 27 percent to 45 percent in the same period. Maternal mortality for the period is estimated at 750 per 100,000 live births. Rates of both child and maternal mortality are still high. Consistent with poverty estimates, malnutrition rates are significant, with the prevalence of stunting among under-five children at 45 percent in Socio-economic disparities also remain an important concern, with 55 percent of under-five children in the lowest quintile showing stunting in , compared to 30 percent in the highest quintile. 16 MACROECONOMIC OUTLOOK AND DEBT SUSTAINABILITY 15. The GoR s medium term outlook for real GDP growth has been revised downward slightly to around 5 percent (see Table 2). A projected annual GDP growth of between 5 and 6 percent would be sustained through increased public and private sector investments to improve agricultural productivity and address infrastructure constraints to growth. The share of priority spending is expected to increase annually by about 2 percent of GDP, depending on the availability of sufficient resources. More resources will be dedicated to public infrastructure, agriculture, education, health and water. In the health sector, focus will be on financing programs to improve primary health care, financing of the mutuelle scheme 17 and the purchase of drugs. At the decentralized level, transfers to Districts are projected to increase as the Government implements its fiscal decentralization policy. Beyond the allocation of funds for poverty-related expenditures, improving the efficiency and targeting of such spending remains a significant challenge. The GoR is therefore focusing on strengthening the institutional framework for the public investment program on the prioritization process within sectors, and on increasing the accountability of line ministries. 16. Consistent with the medium-term policy objective of maintaining macroeconomic stability, the Central Bank is expected to contain inflationary pressures from any fiscal expansion and scaled-up external assistance. Whereas the consumer price index grew by 22.3 percent during 2008, it is expected to grow during 2009 and 2010 at 6 and 5 percent, respectively. The medium-term outlook assumes that Rwanda will continue to be able to cushion itself from the adverse impacts of economic shocks. The impact of the recent global slowdown over the medium term will depend on how strongly the drivers of growth are impacted. These include large-scale investments funded by external borrowing. All of this could potentially lead to stronger adverse growth effects in the medium to longer term, should the global economic slowdown persist and external resources become increasingly scarce. 16 GoR Ministry of Health and World Bank, Rwanda: A Country Status Report on Health and Poverty, June 2009 (Draft). 17 Mutuelles are community-based mutual health insurance schemes that provide access to a defined benefit package provided at public facilities. The mutuelle coverage experienced a dramatic increase in recent years, reaching 85 percent of the households in

18 Economic growth and inflation(percentage change) Table 2: Selected Economic Indicators, Actual 2006 Actual 2007 Actual 2008 Actual Real GDP (percentage change) Real GDP (per capita) Consumer price index (eop) Broad money(m2)* Central government budget (percent of GDP) Revenue Grants Government expenditure and net lending Est Proj. Current expenditure Capital expenditure Domestic fiscal balance (excl. demobilization spending) Overall balance (payment order) After grants Before grants National accounts (percent of GDP) Gross domestic investment Of which: Private Gross national savings Current account bal.(excl. grants) Balance of payments (percent of GDP) Exports of goods and services Imports of goods and services Current account balance (incl. grants) Overall balance External debt (percent of GDP) External debt service (percent of exports of G&S) Gross official reserves (months of imports of G&S) Nominal GDP (billions of Rwandan francs) Sources: Rwandan authorities; IMF and Bank staff estimates. Note: M2 is defined to be currency in circulation and deposits. The ratio of M2/reserves is defined as broad money (M2) as a ratio to foreign exchange reserves. M2 targets for 2009 allow private sector credit growth of about 20 percent. M2-multiplier is defined as broad money (M2) as a ratio to base money. 17. The most recent debt sustainability analysis, dated December 2008, concludes that Rwanda is now at a moderate risk of debt distress owing to recent debt relief and its narrow export base. The amount of domestic debt is small and expected to decline further as the BNR limits the use of domestic debt instruments and instead relies mainly on foreign exchange sales for sterilization so as to increase exchange rate flexibility. GoR has taken steps to prevent the reemergence of unsustainable debt by preparing a debt management strategy with limits for loans, government guarantees and contingent liabilities that will guide future borrowing. Annex 5 contains a detailed discussion of the macroeconomic outlook and debt sustainability Proj.

19 18. Overall, the medium-term macro-economic framework is sound and adequate to support the Government s program. As noted in the completed sixth review of the IMF s Poverty Reduction Growth Facility (PRGF), 18 the medium-term outlook for inflation, economic growth, and financing remains favorable. GoR will continue with policies to mitigate the negative impacts of global and external shocks. It will exercise prudence in implementing macro-economic policy, with a focus on ensuring coordinated fiscal and monetary policies. The GoR will also continue with policies that support private sector-led growth. Investments are planned to improve agricultural productivity, reduce soil erosion and improve irrigation. Investments in the energy sector aimed at reducing the dependence on thermal and hydro-power are showing evidence of paying off. As well, a strategic plan for post-basic education is being developed to address the issue of skills development. These measures should help to expand productivity and establish the foundation for long-term sustainable growth. THE GOVERNMENT S PROGRAM ECONOMIC DEVELOPMENT AND POVERTY REDUCTION STRATEGY (EDPRS) 19. The Economic Development and Poverty Reduction Strategy emphasizes economic growth and poverty alleviation that continue to build on achievements in human development and basic social services (including progress toward the MDGs), while also focusing on decentralization, greater community involvement, and a greater role for the private sector. The EDPRS encompasses, inter alia, the hard-to-reach MDG 1 (reducing poverty and malnutrition), MDG 5 (reducing maternal mortality), MDG 4 (reducing child mortality), and three flagship programs: (a) sustainable growth for jobs and exports; (b) governance, including expanding decentralization and enhancing accountability; and (c) the Vision 2020 Umurenge Program (VUP). It also includes complementary interventions in a variety of sectors designed to alleviate rural poverty and improve productivity. These include labor-intensive public works, direct financial support for the most needy and vulnerable households, microcredit initiatives, water and sanitation, and HNP services. VISION 2020 UMURENGE PROGRAM 20. The Vision 2020 Umurenge Program (VUP) aims to ensure that economic growth is pro-poor and that the majority of the population could benefit from the improvements in living standards that the country as a whole is experiencing. The VUP is managed by MINALOC and is implemented through the newly restructured system of decentralized local government. The VUP is organized around the following: (i) labor-intensive public works that will provide support to extremely poor landless families and encourage saving and development of productive activities; (ii) support to the most vulnerable households that is 18 Rwanda: Sixth Review Under the Three-Year Arrangement Under the Poverty Reduction and Growth Facility and Request for Waivers of Nonobservance of Performance Criteria Staff Report; Press Release on the Executive Board Discussion; and Statement by the Executive Director for Rwanda, August 2009, IMF Country Report No. 09/264. 7

20 intended to alleviate poverty, and (iii) micro-credit and financial service initiatives intended to facilitate investment in income-generation and entrepreneurial activities. 19 The overarching aims of all VUP program elements are to raise living standards sustainably, encourage saving, and promote movement from dependency to self-sufficiency. 21. A key VUP policy reform has been the institutionalization of the Ubudehe 20 process of village-level community decision-making. Ubudehe incorporates what is essentially a poverty-mapping process, which has a systematic methodology and allocates each household to one of six ordinal income and poverty-related categories differentiated by well-defined qualitative criteria. The maps have become the basis for community decisionmaking under the VUP program about income support in cash or in kind to ultra-poor households from local and donor sources, as well as a guide to which households should receive subsidies for health insurance. Consistent with the GoR s significant push toward decentralization, implementation of the VUP is being planned and coordinated by the Sector administrations with participation at the local level and with the Districts responsible for technical support and supervision VUP complements the delivery of basic health and education services, which is now a responsibility of local government. Local governments are now playing an expanding role in coordinating the delivery of various services. Community-based participatory processes have played important roles in the identification of VUP beneficiaries and in the selection and implementation of local public works pilots that are currently being implemented to varying degrees in all 30 original pilot Sectors. In addition, since July 2009, 30 additional pilots have been added as part of scaling up the VUP, leading to pilots that are currently being implemented in 60 Sectors. Lessons learned (see Section IV below) will be applied to national scaling up of the program. HEALTH SECTOR REFORM 23. In the area of health, the EDPRS aims to maximize preventive health measures and build the capacity for pro-poor delivery of high-quality and accessible health services and population interventions. Since 2004, there has been significant progress in implementing a variety of programmatic and health financing reforms to improve the equity and efficiency of health financing and programmatic effectiveness. This has helped to translate increased funding into improved health outcomes. Under the first Health Sector Strategy Plan (HSSP-I), 22 there was substantial progress in facility-level performance-based financing (PBF) reforms. The GoR is now expanding the PBF to preventive and communitybased services, in the form of the Community PBF. This is a demand- and supply-side incentive-based program targeting women and community health workers (CHW), respectively, for improved health behaviors and health service utilization. The Community PBF is a key 19 GOR currently envisages VUP funding for public works, direct supports and credit packages in the ratio 50:20: See Annex 6 for details about the Ubudehe village-level community decision-making process. 21 See Annex 7 for details about recent decentralization reforms in Rwanda. 22 The HSSP-I was adopted in 2005 and covered the period

21 feature of HSSP-II. 23 A key remaining challenge is whether resource allocation under HSSP- II will reflect the increased prioritization of PBF. 24. In addition to health-related reforms, the GoR has prepared a new National Population Policy (superseding the 2003 policy), which is informed by the 2006 Family Planning Promotion Policy. The new policy has four priorities: to reduce the levels of infant, child, and maternal mortality; to promote education, especially for girls; to promote long-term family planning methods and make access universal; and to focus on the needs of youth. The new policy also addresses gender issues, centrally linked to population and reproductive health outcomes. In support of these policies, the CLSG has identified family planning as one of the priority policy areas. 25. Nutrition is a key part of the Ministry of Health s (MOH) community health activities, integrated with maternal and child health interventions. Chronic malnutrition among children under-five is one of five indicators to track progress toward the HSSP-II s objectives. Through the CLSG support to community health policy reforms, which includes strengthening delivery of nutrition-related interventions such as breastfeeding promotion and oral rehydration therapy, the program addresses in part the country s nutrition challenges. LINK TO CAS BANK SUPPORT TO THE GOVERNMENT S PROGRAM 26. The operation will support both the GoR s poverty reduction objectives and the CAS objective of reducing social vulnerability, consistent with the country s fiscal framework. A program in the CAS was specifically earmarked to support the GoR s Vision 2020 Umurenge initiative. COLLABORATION WITH THE IMF AND OTHER DONORS 27. The operation is closely aligned with the ongoing Poverty Reduction Support Grant (PRSG) series, which continues to be prepared in parallel with the IMF s Poverty Reduction and Growth Facility (PRGF) review. The IMF leads the macroeconomic dialogue in Rwanda and in August 2009 completed its sixth and final review of Rwanda s economic performance under its PRGF arrangements. The completion of the sixth review brings total disbursements under the arrangement to a total of SDR8.01 million (about US$12.4 million). The IMF s work complements the PRSG s focus on structural and institutional reforms. The proposed CLSG will work closely to align objectives, resources and dialogue with the IMF, primarily through its close collaboration with the Bank s PRSG 6 operation. 23 The HSSP-II covers the period and has 7 strategic areas: (i) institutional capacity; (ii) human resources; (iii) financial access to health services; (iv) geographical access to health services; (v) drugs, vaccines and consumables; (vi) quality assurance; and (vii) specialized services, national referral hospitals and research centers. 9

22 28. The Bank s portfolio of support has been coordinated with that of other donors. The first PRSC/G series contributed to strengthening the framework for harmonized budget support in Rwanda, which currently includes five development partners: Africa Development Bank (AfDB), Department for International Development (DFID), European Commission (EC), Germany, and the World Bank. In the health sector, three partners are providing sectorspecific budget support - Belgium, DFID, and Germany. Over the last few years, the overall dialogue among donors, and between donors and the GoR, has been strengthened and trust has been built. In September 2008 all budget support donors and GoR signed a Memorandum of Understanding governing the provision of budget support in the implementation of Rwanda s EDPRS (replacing the Partnership Framework of November 2003). The CLSG operation has been prepared to ensure that there is close collaboration and consultation with the donors involved in these critical sectors and to ensure country ownership of the program s objectives, policy actions and outcomes. Recently, DFID announced a firm commitment to provide support of approximately $US31.8 million 24 ( 19.7 million) to the VUP program over four years ( ) covering the remaining EDPRS period. This, coupled with additional funding from EC of approximately $US 14.5 million 25 ( 10 million) over a 2-year period (starting with fiscal year ), will provide additional financing to facilitate implementation of the VUP. In addition, UNICEF has provided computers for Sector staff and there is an agreement for UNICEF to fund a consultant to work on the beneficiary training and sensitization component of the VUP. UNICEF is also likely to undertake an impact assessment of the VUP program in ten Sectors, in addition to providing technical assistance to ensure that the interests of orphans, vulnerable children (OVCs) and women are considered under the VUP. Following a competitive process, Rwanda has received a grant from the multi-donor Health Results Innovation Trust Fund to finance the community health reforms supported by the CLSG, with particular focus on results-based financing and impact evaluation. 26 Several other partners finance on- and off-budget support to the community health policies supported by the CLSG, in particular the Global Fund, UNICEF, and USAID. RELATIONSHIP TO OTHER BANK OPERATIONS 29. The CLSG series strongly complements the PRSG series. The first PRSC/G series (PRSG 1-3) supported health sector reforms, including expansion of community health insurance, decentralization and autonomy of health service management, and performancebased financing. It also supported a pilot social protection initiative that was a forerunner of the VUP. Through broad policy measures, the new PRSG series, now in its second year, will further strengthen investment in human capital and skills for growth; improve water resources, supply and sanitation; complement investments under the recently approved Transport Sector Development Project; and raise agricultural productivity and value-added through exports. The proposed CLSG has complementary policy objectives, including support for policy development in the domain of targeted social assistance (direct support), public works, incentive-driven voluntary savings to move people out of poverty, clear and 24 Exchange as of January Ibid. 26 This is part of a global initiative in nine countries that is being supported by the multi-donor financed Health Results Innovation Trust Fund. Significant technical assistance for the impact evaluation of the Community PBF is financed by an additional US$1 million from the trust fund over the three-year period of the CLSG series. 10

23 transparent eligibility determination processes, and community-based HNP programs. All of these objectives complement PRSG goals, but are much more specifically focused on helping GoR to create the basis for a sound safety net and to attain its hard-to-reach MDGs. 30. In addition to complementing the Poverty Reduction Support Grant (PRSG) series, the proposed operation complements other IDA financial assistance for Rwanda. That includes the Decentralization and Community Development Program (DCDP), and the recently closed Multi-Country Action HIV/AIDS Program (MAP). 31. The Bank-supported Decentralization and Community Development Program (DCDP) has built an important foundation for the VUP, this proposed operation, and the first CLSG. In addition to various community-driven development initiatives that resulted in the creation of social infrastructure assets (mainly class-rooms and health facilities), and the country-wide capacity building at the District levels to support deepening of decentralization, from which VUP has drawn lessons, DCDP is also specifically supporting local technical assistance to manage and support VUP implementation on the ground. 32. IDA support in the health sector has been channeled through the PRSG series of budget support operations, grant financing for HIV/AIDS, and analytical work and technical assistance. Incentive-based mechanisms have been a common theme across these different instruments. Examples include performance-based contracts, retention strategies for human resources, a results-oriented MTEF 27 (supported by the PRSG series), piloting of performance-based financing strategies (under the MAP), and analytical work on the impact and achievements of the sector s performance-based mechanisms. PBF-related reforms supported through CLSG were directly informed by the latter two analytical products. LESSONS LEARNED 33. The Bank s experience with budget support for Rwanda has demonstrated that substantial outcomes can be achieved with strong Government commitment and with capacity building for implementation. Key lessons drawn from the first PRSC/G series highlighted the importance of: (a) cross-sectoral linkages and synergies; (b) strong stakeholder commitment; (c) sound joint analytical work to inform policy dialogue and reforms; and (d) a common harmonized accountability framework aligned around policy objectives to lessen transaction costs and enhance ownership. 34. Specific lessons for social protection reforms include the following. The change to direct support as an entitlement is very important but also presents multiple challenges. Six months is too short for extremely poor households to graduate, as longer support and exposure to sensitization are needed for households to be able to save, invest and become self-reliant after having met their basic needs. Some direct support households may not be able to graduate. Financial systems for social protection programs need to have timely, predictable and accurate payments to beneficiaries as their primary objective. As a result, policy and procedural adjustments have been, and will continue to be, made to streamline existing systems (such as the change to quarterly direct 27 IDA at Work. Rwanda: From Post-Conflict Reconstruction to Development, August

24 support disbursements and payments direct to sector accounts). Although Sectors are the primary focus of the VUP, Districts play a critical role in program delivery and their need for empowerment and support has been recognized and strengthened. There has previously been relatively little systematic policy development in the area of social protection. As a consequence, numerous programs have been launched by various line ministries, some with donor assistance. A diverse range of civil society activities also address social protection objectives. Data on these activities, if available, are often noncomparable. A myriad of programs by various ministries and other organizations renders coordination a difficult and time-consuming job, with specific implications for MINALOC. At the recipients end, it is hard to assess who is accessing which programs, the level and appropriateness of multiple-program participation, and inclusion and exclusion errors. Rwanda has successfully established a number of unique features within its governance structure that are effective, nationally operative, and well-owned by the population. Two key examples are the Ubudehe system of community decision-making and program management and the Imihigo system of holding national and local government accountable. Notwithstanding these unique governance features, the design, planning and management of VUP budgetary processes in an evolving decentralized budgetary framework (discussed further in Section V) remain a challenge. In particular, efforts need to be made to design budget allocation criteria, distributional rules between programs and Sectors (within each District), and processes that disburse funds accurately and in a timely way to those deemed eligible, while protecting the position of poor households and using evolving data to improve the overall distribution of program funding. The method of selection under the current Ubudehe process is helpful in identifying potential beneficiaries for public works and direct support programs within communities but may not treat similarly placed households across communities equally, which may result in horizontal inequity. Moreover, to improve targeting outcomes within and across communities, over the long term, objective criteria may need to be developed, pilot-tested, and supplemented with ongoing community selection processes. 35. Specific lessons for health policy reforms include the following. Monitoring and evaluation of the policy reforms is facilitated by linking policy actions and outcome indicators in the policy matrix and making explicit the intermediate outcomes that will be tracked. In development policy operations, there is a tendency to focus primarily on policy actions, with less emphasis on tracking achievement of the monitoring framework indicators (formally part of the policy matrix) for which the program is held accountable. Intermediate indicators, available with more frequency, may also need to be monitored to show progress towards the program s outcomes. This is critical to monitoring the development effectiveness of the operation, and was also a key conclusion of the 2009 review of development policy lending by the Operations Policy and Country Services Network of the World Bank OPCS Development Policy Lending Retrospective: Flexibility, Customization and Results. 12

25 In a country with low capacity, the MOH and donors need to budget complementary inputs (e.g., training, technical assistance, materials and supplies) to ensure successful policy reform and achievement of program outcomes. A program suffers serious programmatic risk if there is not full budgeting for the implementation of policy actions and realization of the monitoring framework indicators. The community health policy reforms supported by CLSG receive substantial support from other partners. However, because much of this support is off-budget, MOH needs to continue to ensure strong coordination and planning for effective implementation of the policy reforms. Budget support provides important advantages with regard to ownership, sustainability, and use of country systems for innovative policy reforms, but also poses challenges. The Health Results Innovation Trust Fund is intended to support results-based financing innovations and randomized impact evaluation of these innovations. Achieving these objectives through the instrument of budget support has entailed challenges but also significant advantages. The GoR s proposal to the trust fund, which was approved after a competitive process, provided considerable detail on the results-based financing interventions. The GoR s own tools for implementation - notably budgets, implementation plans, and manuals of procedures - allow for monitoring of the extent to which the GoR is translating reform into action and implementing what it proposed to the trust fund. However, this strategy greatly depends on GoR s commitment and capacity. At the same time, underlining the benefits of budget support, implementation of policy reforms has progressed with financing from the GoR budget, independent of disbursement from the trust fund. Similarly, policy implementation occurs fully through country systems, enhancing ownership and sustainability. The reforms and related programs put in place by the GoR will continue after the end of the CLSG series. ANALYTICAL UNDERPINNINGS 36. The Bank and the GoR have undertaken substantial analytical work in the last few years that has been used in preparing this operation. That work has also helped inform both the Bank s most recent CAS and GoR s EDPRS and VUP program, to which the proposed operation is closely aligned. Discrete but complementary strands within the body of analytical work underlying the proposed operation include public expenditure reviews (PERs) completed in social protection, agriculture, and water and sanitation, a series of analytical studies of the health sector, analyses of national household survey data, and syntheses of relevant international policy and programmatic experience. This analytical work is summarized in Table 3 below. 37. The poverty assessment has provided important insights on poverty and vulnerability, geographic distribution of poverty, households characteristics that contribute to a deepening poverty gap, and potential routes out of poverty. This operation makes full use of these findings, and builds on an understanding of past failures and promising approaches to poverty reduction in Rwanda. Of particular relevance are findings pertaining to income shortfall from the poverty line by deciles, which is very high for the poorest decile. These findings confirm the vast distance the poorest groups have to traverse to 13

26 escape poverty, which evidently is caused by a combination of lack of assets and lack of gainful employment (see Annex 8). This operation helps address these specific facets of poverty. 38. One question raised by some studies is whether more efficient or sustainable ways could be developed to deliver social protection assistance. The number of people reached by the programs examined was extremely limited, with expenditures fragmented and poorly aligned with aid policies. There was no systematic monitoring of the programs poverty impact and no coordination across interventions. Consequently, the poverty impact of the expenditures was unknown. Moreover, social protection interventions in the past were poorly coordinated. VUP is attempting to correct this by setting up architecture for effective coordination of all social protection interventions at the Sector (Umurenge) level. 29 In addition, under the VUP, previously missed vulnerable groups will be reached by public works and direct support. 39. Health sector analytical work guided the selection of policy areas included in the CLSG. A series of high-quality Demographic and Health Surveys (DHS; 2000, , and ) provided trend data on key service delivery and outcome indicators. The surveys showed substantial improvements in a variety of areas, including child mortality, but also highlighted large challenges in addressing key determinants of maternal and child health (such as facility-based deliveries, and community-level preventive and curative interventions). These challenges are the focus of policy actions in this operation. Through the 2009 Country Status Report on Health and Poverty - in particular, technical work on community health insurance and performance-based financing (PBF) - the Bank has supported GoR with analysis of the health system s performance and its financing. Together with the impact evaluation of the facilitybased PBF, the analyses validated the use of incentives to improve health worker performance, and led to the inclusion of the incentive-based approach to encourage desired behaviors by community health workers (CHW) and community members, particularly pregnant women and mothers. The literature on the design of incentive strategies is evolving. Under the CLSG series, the MOH will experiment with and rigorously evaluate various interventions, taking into account their programmatic effectiveness, efficiency, and distributional considerations. 30 In the area of population, the analysis points to the fact that, in recent years, annual increases in contraceptive use in Rwanda have far exceeded the Sub-Saharan African average. As Rwanda appears to be at the tipping point with respect to starting its fertility transition, the analysis pointed to the importance of continued support and emphasis in the Bank s policy dialogue on expanding modern contraceptive coverage. 29 See footnote A brief description of the impact evaluation is provided in Annex 16, and more detail can be gleaned from the research protocol. 14

27 Table 3. Analytical Underpinnings for the Second Community Living Standards Grant Area Study/Report Poverty Data EICV I/Household Living Standards Survey (2000/2001) Core Welfare Indicators Questionnaire (2001) Rwanda Poverty Assessment (2003) EICV II/Household Living Standards Survey (2006) Demographic Growth and Development Prospects in Rwanda: Implications for the World Bank [J. May, (2009)] Draft Extreme Poverty in Rwanda: Profile and Correlates Public Budgeting for Effectiveness in Rwanda : from Reconstruction to Reform (2008) Expenditure MINECOFIN Budget Framework Paper (2007) Social Protection Public Expenditure and Financial Accountability (2007) Systemic Shocks and Social Protection: Role and Effectiveness of Public Works Programs The World Bank: HDNSP Working Paper Series. [K. Subbarao (2003)] The World Bank: HDNSP Working Paper How to Make Public Works Work: A review of Country Experience [C. del Ninno, K. Subbarao, and A. Milazzo (2009)] Rwanda: Assessment of Risk, Vulnerability & Vulnerable Groups [Prywes & Verwimp (2005)] Social Protection Public Expenditure Review (Foster, 2006) Social Protection Strategy Paper (Draft, 2009) A Preliminary Analysis of Social Protection in Rwanda (draft Baseline Survey Results, 2009) Health Health, Nutrition, and Population Country Status Reports (2004 and 2009) Demographic and Health Surveys (2000, and ) Scaling Up to Achieve the MDGs in Rwanda: A Background Study for High-Level Forum Meeting in Tunis (2006) Capacity to Produce Qualified Health Personnel in Rwanda (2006) Marginal Budgeting for Bottlenecks Modeling ( ) Rwanda: Fiscal Space for Health and the MDGs Revisited (2008) Health Medium-Term Expenditure Framework 2009/ /12 (2009) Impact Evaluation of Facility PBF Scheme (2009) Demographic Growth and Development Prospects in Rwanda (June 2009) Transparency and Accountability Financial Accountability Review and Action Plan (2003/2004) Country Procurement Issues Paper (2004) Country Financial Accountability Assessment (CFAA) (2005) Financial Sector Assessment Program/Review of Observance of Standards and Codes (2005) Report of the Auditor General to Parliament (2005) THE PROPOSED RWANDA SECOND COMMUNITY LIVING STANDARDS GRANT OPERATION DESCRIPTION 40. The Program Development Objective is to support GoR s social protection and health policy reforms designed to reduce extreme poverty, initially in 30 pilot Sectors, and to expand access to high-impact health, nutrition and population interventions at the community level. The grant will help Rwanda to integrate international good practice within its poverty-reducing policies under the EDPRS flagship Vision 2020 Umurenge Program (VUP). The grant will help ensure the GoR s policy reforms follow international good practice in two areas: assisting extremely poor, landless households through publicworks employment or, where such households are also labor-poor, through cash transfers; and delivering high-impact HNP interventions at the community level using performance-based 15

28 financing (PBF) strategies. These policy actions are consistent with the Bank s CAS objectives and with the fiscal sustainability and effectiveness of public resource allocation as reflected in the EDPRS. GoR has scaled up its VUP program to an additional 30 pilot Sectors, in addition to the initial 30 pilot Sectors. 41. Expected benefits on the social protection policy side include a VUP policy framework and MTEF that are closely aligned with the PRSG, EDPRS, and CAS goals, while at the same time incorporating international good practice. GoR s social protection policies supported by this operation are expected to contribute to (a) reduction in the povertyincome gaps in pilot Sectors over a three-year period and (b) greater opportunities for beneficiaries to seek self-reliance and to graduate into income-generating activities from public works and in some cases from direct supports programs. The operation is thus closely linked with GoR s program, but enhances its quality, transparency, and outcomes. 42. Similarly, expected benefits on the health policy side include policy reform to improve community HNP services, in support of the EDPRS objective to improve health status and slow down population growth. EDPRS targets for Rwanda include expanding access to basic health services to improve hard-to-reach MDGs for reducing fertility, malnutrition, and maternal and child mortality, as well as combating infectious diseases, particularly HIV/AIDS, malaria, and tuberculosis. Building on substantial progress in improving facility-based services, this operation will support reforms in GoR s community HNP policies that complement the VUP, including a strategy to deliver high-impact interventions to households by a network of CHW, and thereby to contribute to improved access to services by poor rural households. 43. As poverty affects health and poor health contributes to poverty, the policy reforms in social protection and health supported by this operation are strongly complementary. Under the VUP, the social protection policies are aimed at providing income support to the poorest households in the poorest areas of the country, while the community health policies focus on improving access to services and health outcomes among poor households in rural areas. The results-based financing innovations supported by the operation are being piloted first in the 30 VUP Sectors before their nationwide scale-up. It is anticipated that support from the social protection and health interventions may provide multisectoral synergies that enhance cumulative benefits to poor households. Achieving progress on MDG 1 (poverty and nutrition) in particular is dependent on improvements in a broad range of socio-economic factors, while child and maternal mortality are similarly affected by numerous determinants outside the health sector. This is most immediately evident with regard to child nutrition, which is strongly influenced by both household poverty level and access to basic health services. 44. During CLSG-1, several key actions were completed by GoR by end-january, The following policies were the focus of the first year of operation. Social Protection Policy: Public-works wage rates were adopted that were no higher than the relevant market rates so as to promote self-selection of the poorest into the program and to avoid labor-market distortions; and payment of wages to workers was done directly without intermediaries. In this regard, MINALOC has collected public and market wage 16

29 rates in VUP sectors, showing that public work wage rates are not higher than the prevailing market rates for each of the pilot sectors. 31 Also, the ministry has monitored the practice of paying workers in its public works program directly, without intermediaries. During the CLSG-1, 6,655 and 9,666 households participated in the direct supports program during January through June, and July through December 2009, respectively. Similarly, 17,886 and 21,937 households participated in the public works program during the same two periods, respectively. A bank account was opened in each participant s name, and the wages or direct support payments were deposited directly into their accounts, thus completely eliminating intermediaries from the payment process in the public works program. In the one case that was found to have used an intermediary, the situation was corrected. In the area of direct support, the household was adopted as the unit for determining the level of benefit. The operational manual sets out a household s base entitlement, which is monitored by regularly collecting data from Sectors and Districts. All Sectors use the household as a unit when determining eligibility under the direct supports program. Health, Nutrition, and Population Policy: Through a joint planning process, the MOH developed a community health service package to address family-planning, health, and nutrition goals, along with an implementation plan and budget that integrate support financed by GoR with programs supported by partners. A community health worker training program started in early 2009 and, to date, 9,500 CHW have received training. The program targets 30,000 CHW, one male and one female elected by each community, and an additional 30,000 community-level health workers who are currently working under disease-specific programs. The community health package is part of the larger National Community Health Policy. Furthermore, a PBF strategy for community health services was developed and training and implementation have started. Out of the planned total of around 400, 160 CHW cooperatives and community PBF steering committees have been formed, and 2,000 steering committee members have been trained in community PBF. Two rounds of funds transfers have been made to the District administrations to finance the incentive system. MOH has decided to incentivize quality reporting, as well, to make changes in service provision indicators. MOH has committed to start implementation by end-2009 in the 30 VUP Sectors, expanding to at least 100 sectors during the first part of A health-sector MTEF was approved by the Ministry of Finance, including budget lines for the community health policy reforms. Budget execution of the MOH recurrent budget is close to 100 percent The following table summarizes the prior policy actions and triggers for CLSG-2 and CLSG-3. Details are provided in the Policy Matrix (Annex 2). As noted in Table 4, the GoR has progressed as agreed for the second year of the operation. It has completed all CLSG-2 policy prior actions. 31 In any cases where monitoring has found public work wages to be higher than the prevailing market wage, action has been taken to investigate and correct this practice. 32 MINECOFIN 2009, Budget Execution Report for

30 Table 4. Prior Actions for the Second Proposed Operation and Triggers for the Third Operation Current Status of Prior Actions Medium Term Objectives CLSG-2 Prior Action (by end 2009) Policy Area: Social Protection Establish a public Adopt policy of voluntary works policy savings by deleting framework that reference to mandatory incorporates savings in VUP public good works and direct supports international implementation manuals. practice Establish a direct support policy framework that incorporates good international practice Government to begin to implement and operationalize the VUP Policy Framework in the 30 pilot Sectors Publicize eligibility criteria at Sector and village level to promote transparency and to ensure a clear and transparent eligibility determination process by making program policy and procedures widely known. Disclose lists of eligible households and selected households at community level to promote transparency and to ensure a clear and transparent eligibility determination process by making selection outcomes publicly available. Completed. Policy has already been adopted and reflected in implementation manuals by removing the previous wording of mandatory savings ; is advanced in developing arrangements for implementation. Completed. Policies for both these prior actions have already been adopted and are advanced in developing arrangements for implementation. Ministry of Local Government is pursuing several media outlets in VUP pilot Sectors (at times at national level) to ensure policies are made known to potential beneficiaries, i.e., policies are discussed at community meetings; live television and radio are used to share information and prompt listeners to ask questions; local NGOs are trained to help share information; and pamphlets in local language are being distributed. See ANNEX 12 for summary. CLSG-3 Trigger (by end 2010) None (only benchmark) Pilot an appeals and complaints mechanism in five Sectors. None (only benchmark) None (only benchmark) Government of Rwanda to develop and adopt a VUP budget allocation policy for reform of the distribution of VUP funding so as to retain broadly the policy intention of a percent split between Direct Support- Financial Services-Public Works programs at the national level while allowing the budget to vary across Sectors depending on individual Sector needs. Policy Area: Community Health, Nutrition and Population Enhanced family planning; improved health and nutrition status Incentive strategy to increase antenatal care and deliveries in health facilities, including conditional in-kind transfers to mothers, is approved by MoH Completed. The government s fiscal year 2010 budget includes support for an incentive strategy to improve reproductive health care. MOH has approved the incentive strategy and implementation manual. As of February 2010, the Community PBF has started implementation in the 30 VUP 18 Based on a review of the effectiveness of the current community-based targeting system (Ubudehe) within the VUP, Government of Rwanda to develop and adopt a policy to improve targeting so that it remains consistent with the objectives of the VUP. Annual Community Health Strategy implementation plan and budget, including Community PBF, reproductive health incentive, and CHW management and training strategies, are adopted by MoH Incentive strategy to increase

31 Strengthened health financing and pro-poor approaches Evaluation of requirements for CHW competencies completed and used to inform the development of guidelines on CHW training Approval of the research protocol for randomized impact evaluation to test these policy modifications.. sectors while preparatory activities are underway for the demand-side incentives but implementation has not started. Completed. MOH is currently implementing a program of basic training for CHWs. An assessment of CHW competency requirements and training is being conducted and this has been used to inform the development of guidelines on CHW training. Completed. MOH has taken decisions on policy modifications to be measured by a randomized impact evaluation. A study protocol has been cleared by the National Ethics Committee. Baseline data collection is planned for the beginning of The evaluation will measure and compare the impact on health indicators of the following policies: (i) Community PBF whereby incentives are tied to changes in health service indicators in 50 Sectors; (ii) In-kind incentives for maternal health in 50 Sectors; and (iii) Combination of Community PBF and In-kind incentives for maternal health in 50 Sectors. utilization of reproductive health services is adopted by MoH in at least 100 Sectors/Umurenge Performance-based contracts are signed and transfers are made with at least 100 CHW Cooperatives, whereby payments vary according to changes in health service indicators 46. To ensure that the operation is successful and that there is strong ownership of supported policy actions, the operation is applying good practice principles, on conditionality, as outlined in the box below. 47. In addition, this operation has benefited from continuous interaction and collaboration with donor partners. The social protection policy framework has evolved over time with significant, regular interaction and feedback from donor partners and bilateral donors ((WB, UNDP, WFP, UNICEF, DFID, GTZ, EC, the Netherlands). The CLSG series has benefited a great deal from continuing consultations with donor partners. Bank missions to Rwanda routinely meet with other donors, and the Bank participates both in a VUP core donors group and a social protection sector working group co-chaired by DFID, which has also funded an international consultant team to support VUP management. 48. Similarly, coordination of partners in the health sector is in line with that in other sectors. It includes annual joint sector reviews, centered on assessing progress towards the EDPRS and sector strategy objectives, and health sector input into the Common Performance Assessment Framework (CPAF). In the policy areas covered by the CLSG, the Bank has worked closely with UNICEF and USAID in supporting the GoR s community health policy reforms, with particular focus on results-based financing innovations 19

32 Box 1: Good Practice Principles on Conditionality Principle 1. Reinforce Ownership Ensure that the GoR has a strong track record in committing to and implementing reforms. The second CLSG grant continues support for EDPRS, which was developed in a participatory manner involving extensive consultations with a wide range of stakeholders. The process was led by a National Steering Committee, comprised of Ministers and Governors, providing high level guidance. The Bank has actively supported GoR in this process. The GoR has completed a formal Aid Policy that articulates its vision for improved aid management and policies. In September 2008, GoR and all budget support donors signed a new Memorandum of Understanding (MoU) governing the provision of direct budget support in the implementation of the EDPRS. These actions underscore the strong commitment by GoR to implement prior policy actions and accomplish the triggers of this operation. Principle 2. Agree up front with GoR and other financial partners on a coordinated accountability framework GoR and budget support donors have agreed on a common assessment framework for public expenditure and financial accountability (PEFA). In a further step, GoR and donors have elaborated a common performance assessment framework (CPAF) that is aligned with the EDPRS. The CPAF framework is also identified as one of the three complementary monitoring and performance management frameworks under the Memorandum of Understanding between budget support donors and GoR. Sector-specific mechanisms, including a sector-wide approach (SWAp) MoU and a sector strategy governing budget support have been agreed between government and partners in the health sector. The GoR also passed an Accountants Bill in December 2007 and adopted and appointed the Rwanda Public Procurement Authority (RPPA) in March, Rwanda is widely recognized in the region for its low levels of corruption. Principle 3. Customize the accountability framework and modalities of Bank support to country circumstances The CLSG series is fully aligned with the EDPRS overall objectives to accelerate progress towards the MDGs, including poverty alleviation and improvements in basic social services. The CLSG should be seen as the overall umbrella for Bank support and policy dialogue in the area of social protection and community HNP policy areas, complemented both by the second PRSG series and by specific investment projects. Principle 4. Choose only actions critical for achieving results as conditions for disbursement The CLSG focuses on a few selected triggers and prior actions, which represent critical steps for implementing GoR s program. These triggers and prior actions have evolved from extensive discussion with GoR and other relevant stakeholders and are aligned with EDPRS objectives, the CPAF, and sector-specific frameworks related to social protection and community HNP. Principle 5. Carry out transparent progress reviews conducive to predictable and performance-based financial support A full review of the program will take place in line with the review agreed in the MOU governing the provision of budget support in the implementation of Rwanda s EDPRS. Joint budget reviews and joint sector reviews take place twice a year, aligned with the budget cycle. 49. For performance and monitoring indicators and pertinent quantitative benchmarks for the proposed operation, please refer to Section VI under implementation, monitoring and evaluation. POLICY AREAS 50. Although the prior actions under the CLSG-1 development policy grant were successfully completed, further policy reforms need to be undertaken to address implementation challenges in both the social protection and health sectors. Details of 20

33 policy challenges, policy reform measures being undertaken, and their expected results are outlined below. Policy Area: Social Protection 51. Description: GoR s innovative VUP flagship program is aimed at addressing the challenges of deep and pervasive poverty. The CLSG series will support and facilitate the GoR s efforts, incorporating global good practices. 52. Challenges: Building on the CLSG-1 reforms, the following additional key policy challenges will be addressed in CLSG-2 and CLSG-3. Public Works and Direct Support The need to ensure that saving is voluntary rather than mandatory. Originally, the GoR advocated mandatory savings for beneficiaries under the VUP to inculcate a saving habit and, at least for some households, create the possibility of launching small-scale incomegenerating activities. However, this policy intention in some areas led to wage rates for public works being set above market wages, in contradiction to the other policy intention of ensuring that the public works wages would not distort the private market (which is also a good practice policy objective supported under CLSG-1). Furthermore, mandatory saving limits the ability of extremely poor households to meet their immediate and basic needs, a key objective of the VUP. The need to establish an eligibility process that is both clear and transparent to all potential recipients and non-recipients and is open to public question and review, in accordance with good international practice. Such a policy would establish program policy and procedures that are widely known and would make selection outcomes publicly available. This policy challenge is being addressed under CLSG-2. The need for a framework to handle appeals and complaints by beneficiaries and nonbeneficiaries. Maintaining public confidence in the Ubudehe selection process requires that the public fully understands and accepts the system, and that there is a process by which the public can complain. This challenge will be addressed under CLSG-2 and CLSG-3. Operationalizing the VUP The need to improve targeting and reduce potential inclusion and exclusion errors under the current system. When the VUP under the EDPRS was originally developed by the GoR, the GoR decided that Ubudehe, the traditional Rwandan process of village-level community decision-making, would be used as basis for the program s targeting mechanism (see Section III; paragraph 21 and Annex 6) together with additional land and labor availability criteria. While this is considered a solid targeting mechanism with strong community-based ownership, and as such highly desirable as VUP is scaled up, household baseline survey data from December 2008 suggest that there are divergences between the intended targeting of poor households under the current Ubudehe structure and the actual patterns of inclusion and exclusion of households found in the analysis of relevant data. Although this is not surprising in a new program for which implementation 21

34 systems are still under development, it does suggest that serious efforts to improve the targeting efficiency of VUP are warranted. Given plans to scale up the program quickly with a view to achieving national coverage within a few years, it is particularly important to address these issues while the program is still in its pilot phase. The need to strengthen the design, planning and management of VUP budgetary processes in an evolving decentralized budgetary framework. Implementation challenges have emerged as MINALOC seeks to implement the coordinated but disparate elements of public works, direct supports and financial services. The challenges involve (a) managing budgets across and within Districts and Sectors, reflecting population size and local poverty rates in the geographic distribution of expenditures; and (b) managing local area outlays so that the entitlement character of direct support expenditures is respected and the intention of allocating 50 percent of total VUP funding to public works is realized. Underlying these implementation challenges are multiple and conflicting government strategies. One is to devolve budgets to the District and Sector level and to pay all benefits through Districts and Sectors, rather than paying beneficiaries directly from the national level. Another is to implement a notional percent split of the overall VUP budget between public works, direct support and financial services, respectively. A third is to respect these sub-programs different budget implications. Only direct support is an entitlement, but it could potentially crowd out public works by reducing its share to less than 50 percent of the overall VUP budget. Financial services are open to many, but the program promises nothing specific. Moreover, poverty rates and population size affect local demand but not budgets, inducing horizontal inequities across Sectors. Scaling up VUP will intensify these problems, as more Sectors per District must share the District s budget. So efforts are needed (a) to design budget allocation criteria, interprogram and inter-sector distributional rules, and processes that accurately disburse funds to those deemed eligible, while protecting the position of poor households; and (b) to improve the consistency between the distributions of poverty and program funding, using evolving data. 53. Actions to Date/Supported Measures: Given the policy challenges highlighted above, the CLSG continues to ensure good international practices by addressing these challenges in the following way. Public Works and Direct Support The adoption of a clear national policy for voluntary (as opposed to mandatory) saving under the VUP. With support under the CLSG-2 operation, this policy has been incorporated into GoR s revised public works and direct support manuals. In addition, discussions of the content of these manuals at workshops, training programs, and community forums and through publicized public documents are all helping to ensure that private sector market wages are not distorted; that poor households have options for consumption to meet their basic needs; and that the GoR s overall social protection policy intentions are in line with international good practice and are widely understood. Adoption of a clear national policy for making the eligibility determination process clear and transparent. With support from the CLSG-2 operation, the eligibility policy has been incorporated into the GoR s revised public works and direct supports manuals making program policy, procedures and selection outcomes publicly available to all potential 22

35 recipients and non-recipients. In addition, clear and simple descriptions of eligibility processes and selections outcomes are being disclosed publicly at local government centers, adapted for publication in newspapers and discussed at local communitymeetings. Adoption of an appeals and complaints mechanism, to be piloted in 5 Sectors, that relies on a number of steps that are open to public meetings and therefore public debate and question. With the ongoing support of CLSG-2, the individual s right to question decisions will be extended by providing a more formalized complaint mechanism under CLSG-3. That would allow a household to seek a review of the decision not to include them in direct support or public works. Under CLSG-2, a mechanism will be developed to allow for a review conducted independently of the direct Ubudehe process and at a higher level. In line with international good practice, it is envisaged that, under CLSG-3, this mechanism will be further developed into a more formal appeals procedure, at least for the entitlement-based direct support. Operationalizing the VUP Improve targeting under the two VUP components (direct support and public works) to ensure that VUP interventions reach the intended beneficiaries. Emerging statistical analysis suggests that further fine-tuning of the Ubudehe process is necessary to ensure that targeting is considerably enhanced and inclusion and exclusion errors are reduced as the program is scaled up to more Sectors. Measures have already been taken during CLSG-1 to improve targeting and minimize errors, including sensitization on the policy and procedures, development of detailed targeting guidelines and formats, targeting training (for Sector and village-level staff and facilitators), all of which underpinned the most recent targeting exercise in July CLSG-2 is supporting the GoR to ensure that statistical and administrative processes and tools for identifying, listing, validating and enrolling eligible beneficiaries are reviewed, streamlined and made familiar to implementers, so that the VUP scale-up meets its VUP objectives. The proposed improved targeting mechanism under CLSG-3 will be based on lessons learned from the use of the existing Ubudehe process and on analysis of data collected over the course of pilot implementation, supported by the two previous CLSG operations, and from the findings of the planned household surveys when they become available. Strengthening financial management of VUP, including budgetary processes, at decentralized levels of GoR. As discussed above, the underlying budgetary implementation challenges that MINALOC faces, combined with multiple and conflicting government strategies, necessitate careful consideration of how budget allocations are managed, disbursed, and monitored to ensure that Sectors receive the funds they require to successfully implement the VUP program. A key indicator of success would be Rwanda s adoption of a budget allocation mechanism for fiscal year 2010/11 that describes how funds will be distributed among administrative Districts or Sectors, and among the three main program components. 54. Expected Results: Implementation of social protection good practice policies in Rwanda should ensure that programs are better administered and that assistance is delivered on time to intended beneficiaries, thus contributing to the overall policy goal of reducing the poverty gap over a three-year horizon. The emphasis in this second operation on 23

36 incorporating transparent processes in the VUP will be complemented by analysis to ascertain whether the people included in, excluded from, or assisted by the VUP program conform with the intention of the policy and thus whether further refinement of targeting criteria and processes is required. Good practice in dissemination would involve transparent disclosure of eligibility criteria and the list of selected beneficiaries at the community level, and an upward feedback of the results of the adoption of the policy. Such a transparent dissemination will also make it possible to assess whether the targeting criteria are being implemented on the ground, the nature of any errors in targeting, and what kinds of improvement in targeting policy are needed. Policy Area: Community Health, Nutrition and Population 55. Description: Community HNP policy reforms address key challenges to making progress on the hard-to-reach MDGs. The operation supports a series of policy reforms that complement the VUP, including strategies to deliver high-impact interventions to households through a network of CHW and to incentivize demand for maternal health services. There is progression in the policy actions between the CLSG-1, 2 and 3. The CLSG-1 policy actions focused on the Community PBF strategy development, the CLSG-2 triggers focus on development of the reproductive and maternal health strategy and the community health worker training strategy; and the CLSG-3 triggers focus on consolidation and implementation of these policy reforms. 56. Challenges: Implementation of policy reforms in order to maintain progress towards the hard-to-reach MDGs remains a challenge. The challenges to be addressed under the second CLSG operation involve the following. The need to improve demand for reproductive and maternal health services. Rwanda has made substantial progress in improving reproductive health services and outcomes in recent years. Modern contraceptive prevalence has increased from 4 percent in 2000 to 27 percent in 2008, while the total fertility rate has declined slightly from 5.8 in 2000 to 5.5 in In 2008, more than 95 percent of pregnant women had at least one antenatal care visit and 66 percent had at least two visits. Although the proportion of deliveries done in health facilities rose from 27 percent in 2000 to 45 percent in 2008, this level of utilization remains low. As part of its strategy to reduce fertility and maternal mortality, the government is continuing to pursue improvements on the supply side in the quality of services, as well as in extending coverage through the community health program. A remaining challenge is low demand for reproductive and maternal services. To this end, an incentive strategy is being devised that targets women of reproductive age (to stimulate demand for family planning) and pregnant women (to stimulate demand for antenatal and delivery care). The development of the incentive strategy is advanced but implementation poses important challenges to MOH. The need to expand coverage of basic health services to poor and rural populations. To continue its progress towards the health-related MDGs, the government is committed to addressing urban-rural and socio-economic inequalities in access to health services. The community health program, designed to improve health services in poor and rural areas, involves basic training of CHW, the establishment of technical supervision and support by health centers, and performance-based incentives for community health worker cooperatives. MOH has started implementation through training of trainers, focusing on 24

37 familiarizing CHW with the package of services they are to deliver. Further improvement in community health worker training, as well as the elaboration of a skills-development strategy to improve quality of service provision, are among the challenges faced by MOH 33. The need for community-based action to complement facility-based efforts to achieve the hard-to-reach MDGs. The outreach model of the community PBF is particularly well suited to address behaviors by pregnant women and mothers that critically determine child health and nutrition outcomes. For example, precisely because the first well-baby visit at a health facility is too late to promote behaviors such as exclusive breastfeeding and use of colostrums, this outreach model is uniquely suited to bring messages to these women at the community-level. The need to support and motivate CHW through performance-based payments tied to improved health indicators. Rwanda has successfully implemented nationwide a PBF system to improve the supply and quality of facility-based health services. In recognition of the fact that CHW require both technical support and adequate incentives to deliver services effectively, MOH is extending this approach to community-based services. Implementing the community PBF system, and effectively tying incentive payments to changes in key health indicators, present considerable technical and organizational challenges. As discussed in the CPAF review (April 2009), the expected level of financing for the facility and community PBF schemes (US$1.45 per capita) was not achieved in 2008 and substantial additional resources are needed from MINECOFIN to ensure that the target for 2011 (US$2.25) will be met. The need to rigorously evaluate policy innovations to ensure that they are cost effective in achieving improvements in health services and outcomes. Rwanda successfully undertook a randomized impact evaluation of the facility-based PBF system which definitively showed that the strategy leads to improved quality and utilization of basic health services. The evaluation also made a significant contribution to the international literature and evidence base in this area. A similar rigorous evaluation will be done to assess the effectiveness and impact of extending the PBF strategy to community-based services, and of using in-kind incentives on a large scale to encourage demand for reproductive and maternal health services. Given the scale of the evaluation, whereby four groups of 50 Sectors will be randomized into a control and several intervention groups, the implementation challenges are significant. 57. Actions to Date/Supported Measures: With the support of the CLSG-1, the GoR achieved significant policy reforms to improve the provision of basic health services at the community level. Progress to date on policy reforms, as well as further actions to be supported under CLSG-2, are described below. 33 For example, in 2008 it was found that in rural areas, 42 percent of deliveries are in a health facility, while in urban areas, the proportion is 62 percent. Among children in the lowest socio-economic quintile who suffer from diarrhea, 32 percent receive oral rehydration therapy, while the proportion among the highest quintile is 49 percent. 25

38 Policy on demand-side incentives for reproductive and maternal health. In line with its proposal to the Health Results Innovation Trust Fund, MOH has decided the broad parameters of a policy to incentivize utilization of maternal family planning and antenatal and delivery care. However, in order to achieve the desired outcomes, including the relevant program indicators relating to delivery care in health facilities and contraceptive prevalence, these policy decisions need to be operationalized. Training policy to improve the delivery of basic services at the community level in poor rural areas. MOH has developed a community health service package and is currently implementing a basic training program for CHW. The requirements for community health worker competencies and the current training strategy have been evaluated to inform further development of the CHW management and training strategy. On the basis of this evaluation, MOH will take decisions on its community health worker training policy and operationalize the policy through an implementation plan and budget. Performance-based financing (PBF) strategy to incentivize the delivery of basic health services at the community level. In the context of its community health program, MOH has taken the necessary policy decisions to begin implementation of a strategy to incentivize CHW based on their performance. These building blocks include the legal establishment of community health worker cooperatives and Sector-level steering committees, design of the payment system, selection of indicators, and transfers to District administrations for funding of the scheme. Community health worker incentives will be linked to quality reporting on a set of indicators, as well as to changes in specific service indicators. MOH will implement this initially in the first 30 VUP Sectors and then in a further 100 randomly selected Sectors. Evidence base for policy decisions on the most cost-effective strategies to expand basic health services to achieve progress towards the MDGs. In line with its proposal to the Health Results Innovation Trust Fund, and consistent with its commitment to evidencebased policy-making, MOH is undertaking policy reforms in such a way that their effectiveness and impact can be rigorously evaluated through a randomized design. The reforms to be assessed (community PBF and in-kind maternal health incentives) have been determined by MOH, as have the design and methodology of the impact evaluation, for which baseline data collection is planned in the beginning of The triggers for CLSG-3 include: (a) a trigger that ensures that the MOH budget and implementation plan sufficiently reflects the Community PBF; (b) a trigger that captures the implementation of the maternal health (demand-side incentive) strategy and (c) a trigger that captures implementation of the performance-based (supply-side incentive) contracts with the CHW cooperatives. 59. Expected Results: Improved coverage of high-impact HNP interventions at the community level is expected to respond to the health needs of poor households. The transfer of knowledge and health services to communities will help households better manage their own health care, which will contribute to improved productivity and higher incomeearning potential. The program is expected to contribute to: (a) improved outcomes such as use of modern contraceptives and ultimately reductions in fertility and population growth; (b) improved child health and nutrition outcomes (for example, reduced dehydration from 26

39 diarrhea, and improved referral for malnourished children); and (c) improved maternal health outcomes through expanded use of facility-based deliveries. POVERTY AND SOCIAL IMPACTS OPERATION IMPLEMENTATION 60. In line with GoR s overall objective under the EDPRS of alleviating poverty, the VUP social protection program and the policy reforms to be supported under this operation aim to reduce extreme poverty among Rwandan households and to expand access to high-impact health, nutrition and population (HNP) interventions at the community level. 61. The social protection policies supported by this operation are expected to contribute to (a) a reduction in the income-poverty gaps in pilot Sectors over a three-year period; and (b) greater opportunities for beneficiaries to seek self-reliance and to graduate into income-generating activities from public works and in some cases from direct support programs. 62. Community HNP policies are intended to improve access to high-impact services by poor households, particularly in rural areas, contributing to addressing persistent socioeconomic disparities in utilization of health services and in health outcomes. The expected outcomes will be closely monitored by specific indicators and rigorous impact evaluation of the VUP will be undertaken with support from the Health Results Innovation Trust Fund. ENVIRONMENTAL ASPECTS 63. Under Operational Policy (OP) 8.60, on Development Policy Lending, the Bank is required to determine whether specific country policies supported under the operation are likely to cause significant effects on the country s environment, and, in the event of such likely effects, to assess the country systems for reducing such adverse effects and enhancing such positive effects. In the event of significant gaps in the analysis, or shortcomings in the country systems, the Bank is required to identify how such gaps or shortcomings would be addressed before or during program implementation, as appropriate. Following these policy requirements, specific country policies supported by the CLSG in the form of budget support are not likely to cause significant positive or negative effects on Rwanda's environment, forests and other natural resources. 64. While GoR s social protection policies relating to the CLSG-2 operation are, in themselves, not likely to cause significant effects on the country s environment, forests, or natural resources, MINALOC has sought to facilitate cooperation in this area by incorporating prevailing environmental and technical assistance guidelines in the operational manual for its public works program (see Annex 17). Environmental challenges are being addressed through the environmental framework of the GoR s Strategic Plan for Agricultural Transformation for marshland development (see Annex 17). 27

40 65. In the area of health policy, MOH adopted an Environmental Health Policy in 2008, which provides the overall context for a more specific "Policy on Injection Safety, Prevention of Transmission of Nosocomial Infections and Healthcare Waste Management, approved in The policy aims to put in place mechanisms, systems, and practices to prevent transmission of infection through injections and other medical procedures, and ensure that medical waste is disposed in a safe manner that does not pose any risk to health personnel, patients, or the community. The implementation details are spelled out in the National Standards and Guidelines for Injection Safety, Prevention of Transmission of Hospital Infections and Medical Waste Management, developed by MOH s Epidemiology and Public Hygiene Unit. 34 Both policies are intended to address the risk of HIV transmission, but also to comprehensively respond to other infectious blood-borne diseases such as hepatitis and Ebola. Based on analysis of HIV transmission routes, there is no evidence in Rwanda that infectious waste is a major source of HIV transmission. 66. In sum, the country systems have adequate capacity to mitigate the potentially small negative effects, if any. The framework for environmental management has evolved in the last decade, resulting in noticeable improvements in institutional capacity. Arrangements have been made to ensure that appropriate safeguards are in place to address any minor effects that might arise under the GoRs s VUP and health sector programs. The operation therefore complies fully with the requirements of OP IMPLEMENTATION, MONITORING AND EVALUATION 67. In September 2008, GoR and all budget support donors signed a Memorandum of Understanding (MoU) governing the provision of direct budget support for implementation of the EDPRS. The MoU, which replaces the 2003 Partnership Framework, outlines monitoring and performance management responsibilities and guidelines. 68. In line with the MoU, the CLSG-1 was aligned with the CPAF, with the exception of one health indicator that, by mutual agreement, is an intermediate indicator relating to child mortality and malnutrition. This alignment was achieved through close collaboration between the GoR and budget support donors. The CLSG-2 operation will continue this dialogue and will support the ongoing evolution of the CPAF. 69. The Policy Matrix (Annex 2) for the proposed operation provides a results framework that describes expected immediate and short-term outcomes as well as indicators to measure progress during the annual operations, and identifies agencies accountable for reporting. The matrix and indicators are consistent with the CPAF and sector-specific frameworks. Benchmarks will be regularly evaluated and interim steps taken by GoR will be monitored In addition to specifying guidelines and standards, it also specifies the roles and responsibilities with regard to medical waste management at the central and district levels (as well as for the private sector), and their respective roles in: waste management at the facility level and in waste production, storage, collection, transportation, and disposal, with special emphasis on guidelines for infected waste. 35 GoR s detailed monitoring framework for the VUP program is described in Annex

41 Social Protection M&E framework 70. Monitoring of the VUP program and policy actions under CLSG-2 will draw on data and findings from a range of sources. The VUP monitoring system is now welldeveloped. From each Sector (and District), program-relevant data are transmitted on a regular basis to the VUP program management unit, housed within MINALOC s Community Development Fund (CDF). Data include number of workers participating in public works programs, the number of days of employment generated, the number of households under the direct support program, and the amounts disbursed. 71. The established baseline values and determined targets are presented in Table 5. These indicators have been discussed with MINALOC and MINECOFIN and are consistent with the CPAF. Indicator Table 5: Key Social Protection Performance and Monitoring Indicators in VUP Sectors Actual End 2008 Baseline year (2008) Target Intermediate (2009) Target (2011) (1) Of those households eligible for support, % 35% 35% 35% 35% granted public works 36 (2) Of those households eligible for support, % 0% 0% 60% 80% granted direct support (3) The VUP program at end of three years will be assessed with reference to the proportion of households in the bottom two categories of extreme poverty (as identified by the Ubudehe process). This exercise is expected to be done in a sample of pilot Sectors % 22.7% 38 < 28.7% < 28.7% 72. To learn from experience and to modify the design of the program accordingly, an impact evaluation will be undertaken three years after the inception of the operation. This evaluation will examine the impact of the VUP program and the modified policy framework underlying this program on households socioeconomic status, and health and education outcomes. Community Health, Nutrition and Population M&E framework 73. The main indicators and targets for the community HNP policy relate to increased financing and coverage of high-impact interventions on maternal and child health, and family planning. The indicators in the policy matrix are among those that have been agreed between GoR and partners within the Common Performance Assessment Framework (CPAF) or by separate agreement in the case of indicator 2 (see Table 6). The MOH annually tracks the progress toward these (and other) CPAF and EDPRS indicators. The 2008 Demographic and Health Survey (DHS) provides baseline estimates; follow-up measures will be provided by routine reporting systems, including the reporting system being 36 To ensure that the proportion of eligible households will not be reduced due to any intervening factor, such as budget constraints, this key monitoring indicator has been fixed at a minimum of 35% throughout the period of this operation. 37 Baseline figure was calculated for 18 VUP pilot sectors for which Ubudehe data are well analyzed. 38 Figures for July

42 set up as part of the performance-based payment mechanism. Another DHS is planned for In the years between the two DHS surveys, the project will track intermediate outcome indicators, which will be used to ensure that MOH is on track to achieve the outcome indicators that have been set for the project. While not formally part of the Monitoring Framework, these intermediate outcome indicators will assist with supervision and mitigate the programmatic risk posed by the long interval (2-3 years) between formal surveys (see Table 7) 39. Because the implementation is to be scaled up in selected Sectors, the indicators will be disaggregated by geographic area. The indicators from DHS will also be disaggregated by income quintile to track progress among the poorest quintiles. 74. Performance-based financing (PBF) has special data requirements, and an information system has been developed in support of Community PBF that builds on the existing health information system capacity in health centers (see Annex 16). The system is currently being streamlined to ensure data quality and address some practical issues that arose during its initial implementation. Innovative data collection strategies are also being explored, such as the use of cell phones by CHW to upload data from the field. 75. Key indicators for monitoring progress are shown in Table 6. The responsible agency is MoH. Table 6. Community Health, Nutrition, and Population: Key indicators Indicator Baseline Actual Target (2008) End 2009 (2011) Policy Area: Community Health, Nutrition, and Population (1) Annual per capita allocation to PBF for health US$1.45 US$ US$2.25 facilities and community health cooperatives (2) % of under-five children with diarrhea who receive 30% See footnote 41 50% oral rehydration therapy (3) % married women aged using modern 27% See footnote 42 38% contraceptives (4) % of births in an accredited facility 45% See footnote 43 50% 39 The choice of intermediate indicators was informed by the HNP Results Chain developed through an extensive consultation process under the leadership of the Bank s Africa Human Development (AFTHD) and Results Learning (AFTRL) departments. 40 This is the only health CPAF indicator that was not met as reported in the 2009 Budget Support Review. This raises doubts as to whether the target for 2011 is realistic. However, the target is not formally changed here in order to be consistent with the targets in CPAF. 41 Indicators (2) to (4) require household survey data and no new data will be available before the 2010 DHS. 42 Ibid. 43 Ibid. 30

43 Table 7. Intermediate Outcome Indicators to Link HNP Policy Actions with Monitoring Framework Indicators Output Indicators Intermediate Outcome Indicators Monitoring Framework Indicators - Family planning promoters trained (number). - Contraceptives distributed to HCs (number). - CHW training in 5 key areas: family planning; nutrition; diarrhea (dehydration); fever (malaria); etc.. - Materials developed for CHW training. - CHW cooperatives established (number). - PBF contracts signed with CHW cooperatives. - MTEF supports Community PBF and maternal health incentives. - Facilities with no stock-out in past quarter (number, percent). - Contraceptives distributed by CHWs (number). - Women (15-49 years) with knowledge of contraceptive methods (percent). - IEC in ORT implemented. - Training of CHWs for implementation of the new Incentive Strategy. - Women (15-49 years) with knowledge of ORT methods (percent). - People reached by the CHW outreach (number, percent of eligible) women. - Mothers receiving in-kind transfer (number, percent of eligible women). - ANC utilization - Women delivering at accredited health facilities (number). - MINECOFIN annual allocation to PBF (US$ million, RWF million). - Donors annual allocation to PBF (US$ million, RWF million). - Number of payments made to CHW Cooperatives. (4) % of married women aged using modern contraceptives increases from 27% (2008 DHS) to 38% ( ) [CPAF] Source: Household surveys. (5) % of under-five children with diarrhea who receive oral rehydration therapy increases from 30% (2008) to 50% ( ) [Health Sector Matrix] Source: Household surveys. (6) % of births in an accredited facility increases from 45% (2008 DHS) to 50% ( ) [CPAF] Source: HMIS and Household surveys. (7) annual per capita allocation to PBF for health facilities and community health cooperatives increases from US$ 1.45 (2008) to US$ 2.25 ( ) [CPAF] Source: Budget and expenditure data. 76. A rigorous impact evaluation is a key part of the evaluation framework to assess the impact of the community HNP policy reforms. The randomized design entails data collection through baseline and follow-up household surveys of target and control Sectors. The follow-up surveys will collect data on service quality and utilization by socio-economic status in order to assess the extent to which the community health reforms are helping the poor. The factorial design of the impact evaluation is illustrated by Table 8. Baseline survey is expected to begin in January of Table 8. Results-based Financing Impact Evaluation Design T1: In-kind incentives Sectors: 50 Households: 600 T3: In-kind incentives + CHW incentives Sectors: 50 Households: 600 T2: CHW incentives Sectors: 50 Households: 600 C: Control group Sectors: 50 Households:

44 FIDUCIARY ASPECTS 77. The public financial management system is reasonably adequate to support the CLSG. In addition, a key focus of the complementary PRSG operation is providing continuing support to strengthening the GoR s 5-year Public Financial Management (PFM) Reform Strategy in its second year of implementation. The strategy was prepared on the basis of the 2007 Public Expenditure and Financial Accountability (PEFA) outcomes which provided a strong foundation for a better understanding of the PFM environment. Assessments of the fiduciary aspects are largely based on results of several diagnostics. These include the Country Financial Accountability Assessment of 2005 and the PEFA of They indicate that public financial management and procurement systems have been improving steadily, in line with government commitment to reform (as outlined in its Financial Accountability Review and Action Plan), and that the dialogue between the Government and donors has been constructive. GoR has also recently completed a Public Financial Management (PFM) Action Plan to ensure comprehensiveness and consistency of reforms. This plan has been endorsed by key donors engaged in supporting the GoR s reform agenda. Fiduciary assessments are also based on results from recently completed Public Expenditure Reviews (PER) in the education, health, agriculture, and water sectors, as well as from the 2004 Heavily Indebted Poor Countries (HIPC) Assessment and the 2005 Country Financial Accountability Assessment (CFAA). In addition, the most recent safeguards assessment of the National Bank of Rwanda (NBR) was completed by the IMF in the context of the PRGF arrangement approved on June 12, The updated assessment proposed recommendations to address continuing vulnerabilities in the external audit and financial reporting areas. The implementation of these measures is being monitored under the Fund program. Finally, the Bank is undertaking a policy-driven technical assistance PER to provide capacity building support to Government in expenditure analysis and monitoring. GoR has also committed to a second PEFA in DISBURSEMENT AND AUDITING 78. The IDA grant will be disbursed as a single tranche of SDR 3.9 million (US$6 million equivalent) upon effectiveness. The proposed grant will follow the Bank s disbursement procedures for development policy operations. The grant will be disbursed against satisfactory implementation of the development policy and not tied to any specific purchases. Subsequent to Bank Board approval and grant effectiveness, the Bank will disburse the proceeds of the grant into an account designated by the Recipient at the central bank of Rwanda (National Bank of Rwanda) and forming part of Rwanda s foreign exchange reserves. The Recipient will ensure that upon deposit of the grant proceeds into said account, an equivalent amount is credited in the Recipient s budget management system, in a manner acceptable to IDA, with an indication of the exchange rate applied, where applicable. IDA reserves the right to request an audit of such account. If the proceeds of the grant are used for ineligible purposes as defined in the Financing Agreement, the Recipient will be requested to refund to IDA an amount equal to the amount of said payment, promptly upon notice from IDA. Amounts refunded to the Bank upon such request shall be cancelled. The administration of this grant will be the responsibility of the Ministry of Finance and Economic Planning. 32

45 79. Within two business days, the National Bank of Rwanda (NBR) will credit the Rwanda Franc equivalent of the grant proceeds to the consolidated account maintained on behalf of GoR which finances budgeted expenditures. The Director of Treasury will be notified accordingly. The NBR will not impose any charges or commissions on GoR for these transactions. The conversion from US$ to Rwanda Franc will be based on the prevailing exchange rate on the date that the funds are credited to the consolidated account. GoR, through the MINECOFIN, will (i) provide written confirmation within 30 days to the Bank that an amount equivalent to the grant proceeds from the Bank has been credited to the consolidated account, with an indication of the exchange rate applied; (ii) provide evidence that the Rwanda Franc equivalent of the grant proceeds was recorded as financing for GoR budget; and (iii) ensure that the Rwanda Franc equivalent of the grant proceeds are subject to controls to ensure its use for eligible budgeted public expenditures only. RISKS AND RISK MITIGATION 80. There are four main potential risks associated with the program: (a) macroeconomic and financial sector risk; (b) regional/country risk; (c) program risk; and (d) fiduciary risk. 81. Macroeconomic and Financial Risk. Given GoR s strong record in macroeconomic management and its overall satisfactory rating, as evidenced by reviews under the PRGF, the recent economic reforms, and experience with budget support under the first PRSC/G series, the macroeconomic and financial risk is low. Nonetheless, with high aid dependency and a low tax base, managing injections of liquidity into the economy may present dilemmas. To mitigate these risks, the Bank and the IMF will continue to provide support for macroeconomic reforms and a commitment to use market-based instruments to contain inflation and exchange rate appreciations. 82. Country and Regional Risk. Currently, Rwanda s internal security and political situation remains stable but continued effort is needed to address the remaining political effects of the 1994 genocide and the recent flare-ups of violence in Eastern DRC in late International and regional efforts have been made to mitigate instability in the Great Lakes sub-region, and the Bank supports efforts aimed at fostering peace and security between Rwanda and its neighbors. In January 2009, the governments of GoR and DRC signed a joint military agreement to disarm the Rwandan Democratic Liberation Forces 44 (FDLR). The joint efforts could mark an important breakthrough in achieving greater peace and stability in the region. 83. Program Risk. Given Rwanda s high dependency on international assistance, particularly in the areas of social protection and health, there is a significant risk that unpredictable donor aid flows could hinder the implementation of GoR policies. While Rwanda managed to cushion delays in donor funding in partially related to withdrawal of budget support by SIDA and the Netherlands following flare-ups in Eastern Congo (see above) - by strong revenues, accumulated deposits, and increased budget spending, the GoR s programs remain vulnerable to the unpredictability of donor aid. However, there are strong 44 Forces Démocratiques de Libération du Rwanda 33

46 signs of efforts by GoR to prioritize on-budget resources towards the most effective services, and of improved coordination between GoR and donors to mitigate this risk. Budget support donors and GoR demonstrated their commitment by signing a Memorandum of Understanding in September 2008 governing provision of direct budget support for the implementation of the EDPRS. Rwanda also explicitly states in its EDPRS that it seeks to decrease its aid dependency. 84. More specific risks to implementation of the VUP and health sector policies involve challenges in local governments. The DCDP project (see Section IV and related paragraphs), which plays a key role in addressing capacity-constraints at decentralized levels, is helping to strengthen overall implementation mechanisms at local levels. Ensuring the sustainability of technical support is critical to maintaining current implementation levels and to the ambitious scale-up of VUP. There is a risk that DCDP, now in its final phase of implementation, will not be continued, leading to challenges of program management at central levels, implementation, and regular monitoring and evaluation. Options to mitigate this risk include: (a) a follow-up program to the DCDP program scheduled to formally close in December 2010, and (b) encouraging other donor support, particular from those already involved in the VUP, such as DFID. 85. On the social protection side, particular risks revolve around the use of SACCOs 45 to make payments to beneficiaries under the VUP. While the GoR is developing a new SACCO structure to cover all Sectors, it is essential that the interests of poor households are fully considered, excessive charges and transaction costs are reduced if not removed, and the full inclusion and equal rights of such households are ensured within the SACCO structure. Under the CLSG-2 operation, dialogue with the GoR is ongoing to address this key aspect of payments under the VUP. Further support to address implementation risks under the CLSG operation includes providing ongoing support to the GoR to assess the impact of pilots on targeted households, as well as to address challenges on the ground to making the implementation of social protection policies sustainable within budget constraints. Similar capacity constraints pose a risk to implementation of the health sector policy reforms, and these are being mitigated by support to the government by the Bank and other partners, particularly UNICEF and USAID, on technical and implementation issues. 86. Fiduciary Risks. The rapid decentralization undertaken by GoR has heightened potential procurement and public financial management (PFM) risks. GoR has committed to accelerating its efforts to address these risks and progress has been made in recent years. This includes the recent passage of the procurement law and the Accountants Bill and completion of Rwanda s first Public Expenditure and Financial Accountability Assessment (PEFA) in Fiduciary Risk Assessments have been carried out by DFID at the Country Assessment Plan level in 2008 and specifically for the VUP in January and December While PFM reforms are advancing, including the new Public Procurement Agency to regulate and oversee public procurement, and finalization of a monitoring framework for the EDPRS, some PFM 45 A Savings and Credit Cooperative Organization is a local cooperative financial institution set up as a legal entity to let a group of individuals save and invest money and to provide loans (see Annex 14). For issues bearing on the use of SACCOs as a delivery mechanism for transfers of cash under direct support and for wage payments under public works, please see Annex

47 areas still need strengthening. GoR is addressing these issues with the help of development donors, including the Bank. To further address fiduciary risks, particularly those related to rapid decentralization, both the Bank and other donors are providing ongoing support and guidance. (See also paragraph 77 in the section above on fiduciary aspects and the annex for Recent Decentralization Reform). 35

48

49 ANNEX 1: LETTER OF DEVELOPMENT POLICY I{~li. o ~ fey, 16~~..., ~hl ( lio{ lip! MIN M INlSn V Of.,"",..,. "'~D <'<;DNO"IC...." St."".0... '" ~''''' T..... ~ ~,_""'" "'-*"'--'- III" J... _ l;l!it C_.,'1' DI... JI_ _ TlltW_Bo...,..._ ~ IMri.. It,p.. L h m'. """... C!IfIk_ R,,_. _~ <I00I... ;,..... livin,..-. f«ito PI. " ihio.,. """'... """- ""..., " iii...,.u.!be ""'"""..... io ""'10 -.,... i.. _ -..I,.-.""'"... ;.....,o4iy _ ill iofioo_ loci.. Ibe..., I... ~..!be 0(.... illhlokd witli c_~ ad"'"... """... i.. "",,-; mol m~ liatmooy. Ioocludioo.c """"'"...,..,.... ~ loci tiviol _, -..dacinl... pol;";" "",.. I ~ empioym<ol boa< fot _y... ".; ""'" _~ ,ff<d;..,..."....., mol ioc... l""---- I. Cjmwdtopi"""",,,- II... ' us lhi bil",_y toas,.."";,,.,t.. oi(;n< it)'... GOP.",..,.. oinpdo... ""'" Ibe... r.w,... _...,...,.. """ '"" ;;.~... "'_ """"'.... Ike"';..-,.-...,mty. Tho _..,. _ io... 11,1 _ io looi. no.. /Up... _ ;. 100II_ from -'"... of "'" I~...,.. io... """'00ial _.!be... ioc fi..,... Tho...,"""'" _ "eo """"'_P'O"'" 0(""'" "'" ""'''''D... ill <100...,..... ~... y """.. "'" ft,... of... '''''' ,...,.... powiio... pioyo6 va)' -.",It.. oem......, ()[)I'...,... 0(1,6_ f«!beperiod looo-iii >««1<... EDPRS... 0( 7 _, Deop;I< _""""I... ~ _ofl.7..._ boo ;"'I"dod """""" ioo pa capito GOP;... ioof\ouon... """"'>'.. i...,.. """""'"... " pa"';.. 40<... by... y _I ''''"'''''''"1'''' food p!'i«s foil _ moj...,!lo.iblt o«mtof... ~ _ r.ooor"oo.;... Tho. Iobo.I pot;c;n.. d... 36

50 .,. «do.. pti<h _ -,... ""... l.o inflalioo. ~... loa b<c. 0;0;_ by......; ,. tool 'Y. by ~ in,,,,,_. 11M ill!po<l of... ""'"" r...r.:ioi crisis.. _ w" _tin '. '... '" be..,..,;p;ti<_ ".,... ""... _ po"; ,. ~ <rill r..... ""'...,_. of liquioiory; _... ;po.... alto.xpo<... ftii iow.o.,!. Tko< _ ~' ;II hlel, "",. "" f_io,... ""... of... ~DPRS.....,.... _Iopmem,... "'... R_'.>biI... _a.... _ioo w <le.. "-... 1'11<...,., ~...,ha'",." ~l<hoo... _ _... pete... in 2IJOO.01..!1 _.. io 2001.(16, tho of pool' iodi_... _ U n'iooo '" of ~ P"I""'ti ill... _Iy """'"... Tho H ouil... ;... 1"... _ p"'... y _ doli... 1o>IcI ~ '"""""..,.,..,"" """Y ot._ Iiae....,_ no", 1,...' '" I. -'... poriod, "... "" '... poor b... oon ;... ooi)'byl_.. _ R... '!(I.day.... _,...;...,...J7 1'"""''' 'bon 'i... ;....,.,...".... «>Odd... fjord... t-l roqio_ofl,iooke.l.doy. ""-ty... be... n-tyhip(91 _I ill f..,;ti<> _...,. dq><_ "" op'oc\iijufai ""'" Iobot ("to;c.o typio:oiiy ,_ ""... iiioqiioiit)' _ "hiclo... ait<-.!y bi... by AIricoo _. Tho< "" , _... of ooonom;" po... 0/"'... _, b_'. ""--"" 0..."- _, bw... _ ""_ , ~ ;~~~;t:;.. II.""'" roo jobo..._ "',u "... vioioo 2010 _il. U_ I"rop'am (VU!'). VUP',"*" objoan.. i,.. 0«<_"......,. prnmot" pro-poor... ot... "'"... by... ptt>du«i" of... "., J _bonouoo ofpoble _ '"""'..t... _," n.. """"' WO<b.....,.... booo&... _... of a-.... yl"""... '_"""ha,..._"r...-"""0)'.",_ io..."..."'.~, Tk [);reo. S_ ""lie)'... i....,.,; WI _, _ I...tlen 0Di...,. po<><..._... that au...,.... u..;, vi'", """ _... Ip...,..,.. ~... ~... poii<y... "... 1«1.. _....,... ' ''''"... -w _r. INnI i""""ved.. ~ P'ikl< po>jocio" p..bil:...u""""it= _ hoi"" loop......,;,,;,j 10 ""'" """""" ""db JO... ';10<_ abo _.....,... VU' P«;'. 0Di impio _..,. _._... "'"."...-,;,to witll... vur..,me.....,.;...;u, objoo:ti... nd ""',,_ d,i d rot... VU.,...,... R_.. """ _n.. ""wtii.... F.DPII.'i..._... ; _ of,. Tbc "9',..... IiDfRS ODd 1M Mjl "";... Do.. ' r",. (MOC,!). Tk U>.RS io<1_ -.;eo 000II..,..... _. ~ lp'irl<'"... MIlGo. io<1odioi io... '" kohb. -.d... _ Willi... oojo<io.. of ioipru... IInIti Kr......,... ''''~of_oood,,",",_on: (i) WI ~ all ~ lib _ add...-,.., _10; (iij WI ""_... _.,. add qoalio)' of... _... _..._. ""~...- to impov< labildy of... 1I.,~.. IiI;.<I... ~.... ol... -,.... """,;;;;;;;; (W 37

51 .,. pricululy.. ronoi _ (i.,.,-... "'"' b<aid, ""'" ii...obio.. all "" """"... ~ o\>ilily 10 po", ~.).,...-,*,aj_lit)' to qiooiity hooltt. ""'"..mo..; (v) '" ""'<GX ""';lobolily ud.f oniabili", of ~... am "",,,,,,,,bl<o, (';).. """""" III< quoidy of. ro<... i<... '.. """""" of di-...; ODd (".;;) '" _......,...i... """""I.._. t>o"... ",.. ot<!>... ilm... I. o..m.... tid pg,p- Cplt""'01\;,,,.. ~ zooo. "" "".a=.", ODd oil budft... M olU_ =:t... ~ of... 0<1_~..... rt... ofr_'owns.'... _..._ b)'... Go... m; t 'poh<y.._ '"... _... l _...,... _Iioo -.01 h<aith "'" Will ~ _...duos...,..po (SWG) "bio:ii.. _ al ""';." ~ ODd.~ -' ~ aod. jooo<... _...,"";..". _.,... poi;e,.... -V i~..., n.-...,... lor> bo.. P""_ _ow..._.. r.....",... "" '... oiip<nm io... "dh... ""'... OIl 001 df..._ The _ Cl.S(J _ <00<... _ <be """"'Y" 1IDn.'I..n;.b wo. _~ ill poni<;po.,.,. ;.""VVOC... j.. of.... _ 0( ;,.,1... f_... 80M.. do.. _... _ policy... ~ by" CUG,'" BaM...-,_Iy ",ith UNICEF _ USAJIl io... io """""""... ~'. ---,. hea"" poiii:)'.._.nib """'" 0'0<lIo OIl..."'... S"""-ty.,.. _ "'" _,-.d..."... h, 0....", _;., "' SWG (UNICEF. Dl'J!), W. WFP _ lindp}. 6.!'rom-" nil' ja Sosio! b'''&... T <hal... of P""Ui.,... doop.-.-_..,"",,"",ipolly" nni... Go,"_..... ;..... vioioo /020 u (VUI')... of.. _ Ropoip ~ _, III< WPRS _..,..,... of..."...,... ~ pol",>,..._... 01>0..._... dot~.. ~.,.ii... >d _ Oy d<cidoo _... jm_ """" io... ioily _y... pp;... poooooo,..."".... io -.JoIowy ono!.. 100:... to """ of oil< _, _...,. _y..._ "*"'1O"i«.t;fI"a... by _n... '1""_" <rit<ti.. or _. _ in... Iowso """'~..., ~ to be,-"" """,. _... Iao.d a6d Ioboor -. OR I>elo&...,.~.....,;ty d~bio iih <Ii<o<O """"'" or I"'i>Ii<...,... on! _... ''1>I0I0... at< "'I;,;bIo. no...,.... """- "'" b,,,is r.. _WII. y... """.. obo&... ia-kaoi i"'-m>oro<o... _..... _r tII... C-_'Y"1r.1 1""""'" _... """".. 4<lermino vii"" priori... ~ 10<.. """"""'"' _ 10 be,_ """"P... pub..:... ~_ fi>i~... "00,.. IO<ti.. ly """ _...,.... WFti... io.-u.,... piio:o """'" 7.!'toW".. pw j. HQ Itb HugjIj<p 1M! "mo"lie' The ~ bu ~,;p;/kan poiky..- io... ioo... poi"y.-. In _ of... "'_ I!oe p«i<jnno_ -U (I...,..,)!>tow... "'_ of... RepubIj< onoi 38

52 llfplllwhllj H!!!l 1E PIll Jl iii I ''-!l;l"il'lll}:1 ~l -. ; H,,J " I! n!'._ - _I. 1 1'1>... jl '!I!}ll!!],,-'! ".,.,... 11il~il:ilililfl' II!!; '{Hili ll}h illt liljf~i!!iriihh!1!lin ~ii!l{il" i Ull! 'ih j 39 " "OIl ' "',"'.", Jr'jl s' l 1 ' ~! J lli I'i! d"i' II'l! Pl';~ 1"1' I l~!! ~l t 1jl,11 I '-'j l.'!!jh af -Ii!] ~. 1 Uh h q, ~ l'll t: ~jjsd r~l'!;. i ;;r~ j '''''ililill!l~ ;'jlil Ihl,!l:llHj "i.!l;.!h 1H1IHd. :,I!! 1t;!.,!1 ~1!'1111'ili!ll!i ~h [I ' " I" 1!1'1' h'i 11 'I' ~IIJ!llllfll~liil!'1"I!".! ill fill I,PI 1 -, -' J!J is. e.i!~ l.!j ~.J~.. t t 2 ~ g ~ '" -li

53 ,,. I'oodiop '"' P'- 01 VI,IP... " )(I c-.-.., _... VUP_... 1'01.,..".. U'I,,"' '_01..,... 01" ~"" (. J ibod.--...,_.... ',VU, _. C ry.. =... _.. _. 4 _,, " n vup_ - r _.. _... VUI'._..... O n... _... -' -_ -j' '.._"U...' _-.. ~oi' ",...,....,01 ~, _... U.t'S'"..._"'......_' 4 ~;,...--.io. ~_,... _... _0(.. t,,_vu, _ ""11&_ ' I.,..om.... "ill... _.,...,... ain.ody... _... CLSo-,.. """""", _... VU._ ""......_... _ ".."'.. _... _10... _... ot..._-"'id... _ no._...-.._ --_.---' ~.;;;.;i lot..._, t Iaa VUf' ;.,... '_ DoGe,,, ;, ' _ r... o...io,,."",.,_oivvi'... oo,~"",_ "",,,,"'-oi_ t ~,'II lim_iii.. _... _... _ ' r '7..-.,..._01 ~""" _, 00<1...,.,... I. ruli<iic... JOOlo....rot It "ioo... -.od: (0) e '''p",... """'"'...0 _.. ~)rdio«ioa;; I I '......, r... (c) w.. )'1... _ """",,_' ot.~i0 " _"'... 50_.,... """'".....,...,_. Q -. 0_"..., "' _ Fl _"""'p,.. -.' _.., ~- -""....._-._ TOio... _...'... '... c... "'0/1 pon;ripoooo -'>'... '-''',, r oo "".. _""' lo ';1000. Oi_... ",lit ~ pioo......,., of... """"" So... er..to C...-;.. Orpoo;"'_ (IACCOo, '"" _ "PO'"..., _ vur.1;pb!< _110:_,,,;11 fi-.i.... $ACCOo... bl< fi.-;01 1_..._.,_... ir_1\io.,"",00..._.._' '... ct.!i --.,.._ "' SACCo--. 'I'loo 1ooo.1;lo..,.,$AC()()--..-., '..._... SJoCOOo..._,...,..._,..._..... _ot_" ' " _... "...,I "' '" 40

54 10 St."., "Nrw I'pIjq h"""' _ ~......;. ~pinc... poli<y <hol~ to ""pletneotiq til< vup!he ~'. fbpiup I'fOF"'",;1I'IOd ol...."or- ~ ,.-.. _ ".._ i_i rood pn<tict ~ ill _..... th< po,.". _...,.,... CI.S()."...,... "'~ """"""'''_ '''Go.n liodn (;'-S() l... Go..._',,..&,..._ DR«s.,.,.oru pol;"", ",-",... ~... )O' io!be domaia ocl"ubli< WOW aod.. Adop<"'po'"'r~"""'''''''''''''''dc....-.,",.. '.'''~ to. i. VUP... " """" Mod otitt<, ~ io.,..._,..-_it: Pubti<iDao; dip.;b!y ori... s.:... 'il... level....- r aod <1<. Mol ~bility... '_"" """"". by... _N"'"*l:r~ C ""'",..,. Ii,,, oc <lijil>io... _"" aod oeieaod ~..., to _., ""'> -.:I... c tlip.;liiy <kin. ' rico JI<OC"< by... _"" poobioo:ly.. ';10Il10; d. ""'0" ~ vup ood... rood pn<ik>< pol""" ~ '" _ pul>ii< """,... '""'".-"'''''''''-'~",y, eo po''''>' ' won f«_iq«wplo"~ f. Adopt... pol;'. of r-iol '*"'.. vup', _;,,-.. _-'_... ;.,p,... orv\jp _... _~... OC ~,..... IldopOini' pol.". of ~.. r",ou-._ """'"""'" of _ public _ &lid _ -' iiici<odina poiioy of _an r...-.ocfiadio,p... J _ U"",,<:U(;-J. Ih<Goou ", poi;.y.. _ io... domaif.of N>IK Woru... DR<t ~poi"ylo<_: Pi""i... _ ODd ~ _,,,;,,,, iri n... ~ b ' """"'"" VlJI' ocMion,...;.y "'" _ of 110. cbo<tiboo"", of W P Iimi... to -'" ~... pol"" ~ JO.SO P'<""<. opj;t... 0;,.", s...- -F~iaI ~"'_ Wooko... _ at"... while,1iow;... b.odc<'.....,.... _ ~.. oolaolirioiool Soc"" -.II: on.! c...., 00...;.~. of... dio«;_ "'..._b.. "'FIio&... tile VUP." o... 1op and ~. p<>6oy...,...,..... ;, 10 _.. _... ~"'... VUP. :... 41

55 I, I', ' "\ 1 1,!' Ii. I '1!'1~11.,II! I I 1 ". 'j hfl 1t?::"~.I", "i JIll'! ~, >~ ~~. ~ :I.,.j, l 1'1. 1 j ~ h.qip " I -... ~ n. I!!.-, 'Jl ; "'.Ii ~1 ~J.~ II I111 ~ ll' I. 11 (ll'llll Jill Ill I! I. I,!! '111-11;. ",!. i"! 'j!. li'!!'!1 to-t. ~. d. i:-8 I1 'l! f Hi, 1'1'!, Ii. Pl~ Llllit. ~~~ sis 1~ t!il ~ljli;~; :. loll Jjl'llll. 1! ' fi ~llh i!o Ililllllll Ii I. : 'II""!!'!l1!IJ'1 lj tl"!! tl.j lihi,lh.llli tl]' rli,..~, i. I!-l~ fl~ luh!i\1. ml!~lth H II iti'ld I!'pl!!,! 11~1,1'" ~. ~ ij ~ ~ dll:,fjj 42

56 _IV«!... "" ~.. i, 0001.,.... end.. pt<>... <OOnI... of.. in;04.r_ty... 1<10 _ < ~ pol;"" rrl... io "'i. domoio.- C\..S() l _Iud<.--..., IS for CllW -'<'«I... okvclopm<no 0( p;ddiooo.. CHW IrOiald&.... _ I,...,., wi.....-k!on... _ioie lilt "'...,....._ CJ.S(].J "" Go_", polie, r<f... i.d..., ~~d """""_ or< Iipod... ".."i'<n _..., \oat 100 CIlW Coop<m;... 'Ok,...,. 1"')11><0"..., ~ '" """- i. hooldo """"""'". '1. errfqpo_elfd fcw;.. (PBfl no. Go.. ".".,',.. "" _ of PBF pdiey,.rom..-. Cl.$(l-l...: T10e _01 of...,., ~ lot twnd.wn_ ""'*"."...,.... """,,,.." ". MoH, Ailditjm.!ly, "" (;0,."..-.. _iu«i "'...!hot ""food,...,._ <OOtrow... _po<! on<i... _ al l ~ CHilo' C~_ ot>d.. --.l..- bo<it«.... _"""" _ MTEf, \0 roily -"'"... ~ '. """"'..., "" M<I pop.laboc. poii<y"- Tho Go ", p>ii<y.._... to ~ I-I<atl. a-nonity K.. hi> ond_"t I s..-io&_cl$g.j""'_: /u,... C_.. ',,1'- Pol;',.opk pioo... bo.o<itd. ""'h>oiioa c-oairy PDP,... v.<... ;,.,..q_... CHilo' I""... _... "OOC;"......, ~ by MoH. I. I...,.., "- p<>iie;'" "" c.o...t.. _ lui"... j-. Mol rvaluot... """"'''''""""-_ ofp>li<ie...,... '" ~ PDF, "J" ;,." Ii... ond CHilo' ~ ~.. "",I...;_ 'YO"";'" furo-<.... """Ydosip. IJ. 110< Go><mne.. i.... _.. rip<juil,..w... po,""..... tiuo...., I'IOc<w... """'"""'" """"' RWMdo _fully ~...-:",.t... '......", of III< taom.,-booal... F., Ii<II "'" d<ftnili..., III< to impro>'«i _ lily on<i util_ or...., too.hb...-; u... MIod oipi&..m...-",,;, ,-. ""'. in..., A... i... ri..., ~I 110_ ">_... _~ of... too PBF ,._boood..me n..... ;"_0;,,0... loop 0<aI< to..,.,."... _ for ~i.. Ioeol<!> ~. Il "'va- 0.. <lo,u,",.,.' lo.-...;......,11'li< Md... U bo... of "" k>.,.-i>iltpd...,._ or too R....,. The VUP 1""1""" i <

57 _01.."". " ,- "..,..... _",..,,,", _... ' jooolioovt)' nqui poii<y..._. rot...- "'" 00.."... _',.....'...", w.,.poiiq... n..... filwdi_... U EDPltSI_... R I'.. -., '... SIrootcY... (H.'lC)_ ;I ' '''~'''_ '11...., r.. _,. d~""_"'..."'""_"",,: at<i.. MDG... II ~!7-~... _~_1 ' _10_. W.... ~ '-01 Do. " "",;.y.. _I tao _ ' Do..',..._of... _... Cl.S<l...,ot... S6_... _of.._ol... a..tg_... T_...-.'O. w. _mo ,..,,,. "$ n ' t '.-01_..... I< r... _..... _... prioriiuoqoo or u. ""'"... _ _... 0( tho,... V_.--y,... 44

58

59 ANNEX 2: RWANDA COMMUNITY LIVING STANDARDS GRANT POLICY MATRIX Medium Term Objectives (1) CLSG-1 COMPLETED (prior actions are bolded) (2) RWANDA SECOND COMMUNITY LIVING STANDARDS GRANT POLICY MATRIX Policy Actions Expected Outcomes & Responsible Entity CLSG-2 By end-2009 (prior actions are bolded) (3) CLSG-3 By end-2010 (triggers are bolded) (4) EDPRS Strategic Objective: 2. Managing population growth rates and enhancing population development [CPAF] Policy Area: Social Protection (SP) Establish a public works policy framework that incorporates international good practice [CPAF 2.8 Improved social safety nets] Adopt international good practice policy of fixing public works wage no higher than the prevailing local market wage in different pilot Sectors. Adopt policy of voluntary savings by deleting reference to mandatory savings in VUP public works and direct support implementation manuals. (5) Outcome: Public-works employment implemented using international goodpractice approach. Outcome: Avoidance of distortions and inefficiency in private labor markets that are associated with policy designs for public-works employment programs that pay above market wages. Adopt a wage payment policy that ensures direct payment of wages to workers without intermediaries. Adopt a dissemination strategy regarding VUP program components and international good practice policies pertaining to both public works and direct support, to ensure total transparency. Conduct a study of lessons learned in public works as a basis for evaluating the need to make adjustments to program policies and procedures. Responsible: District governments; MINALOC Outcome: Participants have bank accounts, which could be used for income-generating activities Outcome: Public works income transfer is offered to identified, eligible poor households in a transparent and timely manner. Responsible: District governments; MINALOC 45 Monitoring Framework Indicators (6) (1) Of those households eligible for support, % granted public works in VUP Sectors (2008: 35%; 2009: 35%; 2011: 35%) [CPAF] Source: MINALOC and local government.

60 Medium Term Objectives (1) Establish a direct support policy framework that incorporates international good practice [CPAF 2.8 Improved social safety nets] CLSG-1 COMPLETED (prior actions are bolded) (2) Adopt a policy of community input into selection of VUP beneficiaries and projects. Adopt international good-practice policy of using the household (rather than family or individual) as the unit for determining the level of benefit. RWANDA SECOND COMMUNITY LIVING STANDARDS GRANT POLICY MATRIX Policy Actions Expected Outcomes & Responsible Entity CLSG-2 By end-2009 (prior actions are bolded) (3) CLSG-3 By end-2010 (triggers are bolded) (4) (5) Outcome: Creation of community assets that are of use to communities and are identified in Sector Development Plans, and incremental wage income in the hands of extremely poor households. Publicize eligibility criteria at Sector and village level to promote transparency and to ensure a clear and transparent eligibility determination process by making program policy and procedures widely known. Disclose lists of eligible households and selected households at community level to promote transparency and to ensure a clear and transparent eligibility determination process by making selection outcomes publicly available. Develop an implementation framework for handling complaints. Conduct a study of lessons learned in direct supports as a basis for evaluating the need to make adjustments to program policies and procedures. Pilot an appeals and complaints mechanism in five Sectors. Responsible: District governments; MINALOC Outcome: Direct Support policy is implemented using international good practice approach. Outcome: Social assistance income transfer is offered to identified, eligible poor households in a transparent and timely manner. Responsible: District governments; MINALOC 46 Monitoring Framework Indicators (6) (2) Of those households eligible for support, % granted direct support in VUP Sectors (2008: 0%; 2009: 60%; 2011: 80%) [CPAF] Source: MINALOC and local government.

61 Medium Term Objectives (1) Government to begin to implement and operationalize the VUP Policy Framework in the 30 pilot Sectors [CPAF 2.8 Improved social safety nets] CLSG-1 COMPLETED (prior actions are bolded) (2) Adopt a policy of institutional strengthening for VUP management through establishment of a core national management team, including national coordinator, financial management specialist, and monitoring and evaluation specialist to implement the VUP policy framework, consistent with GoR s decision on VUP policy. RWANDA SECOND COMMUNITY LIVING STANDARDS GRANT POLICY MATRIX Policy Actions Expected Outcomes & Responsible Entity CLSG-2 By end-2009 (prior actions are bolded) (3) CLSG-3 By end-2010 (triggers are bolded) (4) (5) Adopt a policy of financial management that strengthens VUP s capacity to monitor and manage the implementation of VUP through decentralized levels of government. Government of Rwanda to develop and adopt a VUP budget allocation policy for reform of the distribution of VUP funding so as to retain broadly the policy intention of a percent split between Direct Support-Financial Services-Public Works programs at the national level while allowing the budget to vary across Sectors depending on individual Sector needs. Based on a review of the effectiveness of the current community-based targeting system (Ubudehe) within the VUP, Government of Rwanda to develop and adopt a policy to improve targeting so that it remains consistent with the objectives of the VUP. Outcome: Inclusion of extremely poor households from the bottom two Ubudehe categories in one of the safety net programs under the VUP in pilot Sectors, with adequate budgetary provision Outcome: An institutional policy framework capable of implementing social protection programs fully integrating VUP program, at central and Sector levels. Responsible: District governments; MINALOC Integrate VUP program into the broader social protection policy framework. 47 Monitoring Framework Indicators (6) (3) The VUP program at end of three years will be assessed with reference to the proportion of households in the bottom two categories of extreme poverty (as identified by the Ubudehe process) in a sample of pilot Sectors. (2008: 28.7%; 2009 & 2011: < 28.7%) [CPAF] Source: MINALOC and local government.

62 Medium Term Objectives (1) CLSG-1 COMPLETED (prior actions are bolded) (2) RWANDA SECOND COMMUNITY LIVING STANDARDS GRANT POLICY MATRIX Policy Actions Expected Outcomes & Responsible Entity CLSG-2 By end-2009 (prior actions are bolded) (3) Adopt a policy of institutional strengthening to facilitate regular monitoring of both public works and direct support programs, including a policy of upward feedback of findings relating to monitoring outcomes. CLSG-3 By end-2010 (triggers are bolded) (4) Conduct a study of the implementation of the monitoring and evaluation framework. Using the standardized, monthly, administrative data of the VUP program, produce the first in a series of Annual VUP Reports summarizing the first years of implementation of the VUP program (5) Policy Area: Community-Health, Nutrition and Population Enhanced family planning [CPAF 2.1] Population and family planning promotion policy is developed including IEC strategy [CPAF 2008] Promulgation of new tariffs establishing free family planning services [CPAF Jan-Jun 2009] Training in family planning of all associations of promoters conducted [CPAF ] Outcome: Improved utilization of family planning, contributing to reduced fertility and population growth Responsible: MoH Improved health and nutrition status [CPAF 2.2] Community health, nutrition and population service package (including high-impact interventions, such as ITNs), and Community Health Policy implementation plan and budget, including partners programs, are approved by MoH. Annual Community Health Policy implementation plan and budget, including partners' programs, are adopted by MoH Annual Community Health Strategy implementation plan and budget, including Community PBF, reproductive health incentives, and CHW preand in-service training strategies, are adopted by MoH. Outcome: Increased access to community health, nutrition and population services Responsible: District and Sector administrations and MoH 48 Monitoring Framework Indicators (6) (4) % of married women aged using modern contraceptives increases from 27% (2008 DHS) to 38% ( ) [CPAF] Sources: HMIS & household surveys (5) % of under-five children with diarrhea who receive oral rehydration therapy increases from 30% (2008) to 50% ( ) [Health Sector Matrix] Source: household surveys Source: HMIS

63 Medium Term Objectives (1) Strengthened health financing and propoor approaches [CPAF 2.3] CLSG-1 COMPLETED (prior actions are bolded) (2) Facility PBF policy is modified by MoH to increase payment for deliveries CHW Integrated Training curriculum, training plan and budget are developed by MoH CHW Cooperatives are established in at least 30 Umurenge Community PBF strategy is approved by MoH Health sector MTEF that supports community health, nutrition and population policy reforms, including community PBF, CHW training and incentive strategy for maternal health, is drafted by MoH for submission to Ministry of Finance RWANDA SECOND COMMUNITY LIVING STANDARDS GRANT POLICY MATRIX Policy Actions Expected Outcomes & Responsible Entity CLSG-2 By end-2009 (prior actions are bolded) (3) CLSG-3 By end-2010 (triggers are bolded) (4) (5) Incentive strategy to increase antenatal care and deliveries in health services, including conditional in-kind transfers to mothers, is approved by MoH. Evaluation of requirements for CHW competencies completed and used to inform the development of guidelines on CHW training Performance-based contracts are designed and signed with all targeted CHW Cooperatives Annual government budget operationalizes health sector MTEF that supports community health, nutrition and population policy reforms, including community PBF, CHW training and incentive strategy for maternal health Incentive strategy to increase utilization of reproductive health services is adopted by MoH in at least 100 Sectors/Umurenge Performance-based contracts are signed and transfers made to at least 100 CHW Cooperatives, whereby payments vary according to changes in health indicators Annual government budget operationalizes health sector MTEF that supports community health, nutrition and population policy reforms. Outcome: Reduced risk of maternal mortality due to improved delivery and referral care Responsible: District and Sector administrations, Ministry of Finance and MoH Outcome: Improved availability and retention of skilled health workers in rural areas Responsible: MoH Outcome: Improved community health, nutrition and population services due to performance-based financing strategy Responsible: District and Sector administrations, Ministry of Finance and MoH Outcome: Sufficient financing of improved community health, nutrition and population services Responsible: District and Sector administrations, Ministry of Finance and MoH 49 Monitoring Framework Indicators (6) (6) % of births in an accredited facility increases from 45% (2008 DHS) to 50% ( ) [CPAF] Sources: HMIS and household surveys (7) annual per capita allocation to PBF for health facilities and community health cooperatives increases from US$ 1.45 (2008) to US$ 2.25 ( ) [CPAF] Source: Budget and expenditure data

64 Medium Term Objectives (1) CLSG-1 COMPLETED ( prior actions are bolded) (2) M&E plan and impact evaluation design for Community PBF and maternal health incentive policies are determined by MoH RWANDA SECOND COMMUNITY LIVING STANDARDS GRANT POLICY MATRIX Policy Actions Expected Outcomes & Responsible Entity CLSG-2 By end-2009 (prior actions are bolded) (3) CLSG-3 By end-2010 (triggers are bolded) (4) (5) Approval of the research protocol for randomized impact evaluation to test these policy modifications. Monitoring and evaluation reports on implementation of policies relating to Community PBF, reproductive health incentives and CHW training and skills-development, as well as the impact evaluation follow-up study design, are distributed by MoH Outcome: Improved data and analysis to inform community health, nutrition and population policy Responsible: MoH 50 Monitoring Framework Indicators (6)

65 ANNEX 3: FUND RELATIONS NOTE R wanda-.. h, ~,' ''n~nt L~tI~r for,h ~ " 'orld Bank January 14, 2010 TIlls ]<11<'1 pro\ci&s an "ss ~s=1 o f macr~onomic &n]~ms in Rwanda in 2009 and prosp<'cts for FlU'" in,'o]v"""'nt in 20]0 and kyond. P ffiortl1ll.nc ~ aga in5llb ~ aulhoril;~ ' 2009f1 0 macro.-conomic ~work prtpar.-d ;n cohaoor-ai;on with FlU'" staff bas k... bmadly satisfaclory. Strucmu] r.-forms w<'l~ imp]e",",n",d as ~ x p<'cl~d_ although so",", hy fiscal ind;cal;ve larg<1s w ~ r~ mis... d. R~la, i o n ' with Ih~ DIF The, ixlh and fiua] r~, i~w uuder a,hl"p~, y ~at ' PRGF "rt'aug~tll~ui for R wanda was c OlII]l]e,~d iu Ju])' 20(19 and the a't"u \i:~meot npir~d in.-\.u \i:u, t For monitoring pu1pos~. The authoril;~ pr."ar.-d. in collaboration with F und staff. a macr~onomic framework foc 2009f1 O. with quarterly indicativ ~ targels. A FlU'" mission V;,.;I.-d Kigali duling October 27- November to r"",jew r.-cent &,'dopmcnts and perfortll"nce agaimt The 2009/ ] 0 ma clo~ conomic framework. The autborities indicated Th ~ ir ;mention to requesl F luld suppon foc Their economic program. STarting on July]. 20]0. to align The progr"'" ",jth Ihe n ~w b''''g<1 cycl ~ (Ju]y,JtlUe} A Fund mission is p]ann.-d foc MaITb 20]0 10 discuss Th ~ auth""ities program. Rec~nl DH~] O]lmeot ' Ecooomic growth, low ~d, hal'ply iu 2009, f"oiii Ihe \' ~t "Y hi\i:h I"3le in 200S. Rca] GDP gro",h in 2009 is ~s Timat ~ d al 3 percent. compared perc.-nt proj ~ c",d in mid and down from ] ].2 pcit... 1 in 200S. Th ~ slowdown re flects mainly a downrum in comtruclion. flllancia! 5<'I\cic~s. comm<'lc~ and louri5lll.... clors of Th ~.-conomy panicular]y '111n..-ab]~ 10 Th ~ g]oba] slowdown and The tighten.-d domestic liquidity conditions. Ba]ance of paymems pressufes M,-e so far bc= avoi<kd d,,,, to high<'l aid flows. Inflation in 2009 bas d.-clin~d 10 single digits. " 'hi]e the fi,ea] Pt'Oll t''''" r~main ed on 'l"3ck through Jun~ 2009, ]lhfo nll anc~ was mind in,h~,hi"d 'l"art~r of 2009 r~fleclinll delay. in dollor d i, blll"s~mellts and challenge, in budge, pnculion and tllana g e lll ~ut. TOla] re,,"llu ~S w<'ie in ]ine ",jlh budgeted amounts Ihrough md, Scplember as a significant shortfall in non,tax re'-cllucs (mainly fcrti]i:z ~ r 1.-c~ipTS "'''' dri,' ~rs lice."""s) was off... t by highel,than projcct.-d coll.-ction of taxes on goods and...,jc ~s and international trade tax ~s. TOla] ~ X pendilufes w<'le a]50 1..,,1 wilhin budgel ~ d amoums. The authoriti ~s ma& a capita] comribution in a new convmlion c ~ nlcl_ which was nol included in The b''''get. and was off... 1 by ]OW<'l domesticahy,fltllluc.-d capila] =pcndinll ~s Howe,'<'I. three out of se,-en indical;ve fiscal targets W<'l ~ miss.-d al md, Stplembcr The ceiling on n",t cr ~ ditlo Th ~ go,-c-rnmmt was mi...d owing 10 shortfalls in donor b''''g<1ary suppon. while accumubtion of dotllcst ic arrears was also high<'l, Than' '''''P'''Cled owing to delays in paymenls to supp]i<'l" pending an ov<'lbaul of procuf""",m practic~ ; 51

66 2 th ~ paymnj.1s arr now being r~glllariz..d_ Priority s~ding was also slightly hdow larg~1 31 ~"d - S'1'lrmkr arising in part from low...- dolll<'stically-fmancm capilal s~nding_ The monelar" proi:taln n mained o n I,"ark throu\:h end-septembe,". Th.. authorities /If.. dteclivdy using [cpo o~ a tiolls 10 Il1<'r l,.. ""rye money IMgets on a monthly a,-rng" rather than =d-quart...- basis a. recommmckd by Fund staff. Bank liquidity has improh'(i. as &pos;t inl",rsl rales have httomr positi... in [.,,,.11=<, but banks,.. main cautious in =tr nding crr dit amid [eer nt higher l... ds of noil-p<rlonning loans (th.. ratio ofnpls 10 10la1100n ponfolio increa'iro from 12.6 ~fcrnl in Decnnber 200S perce1ll in S'1'lemkr 2009) Credit to the private s«:i01" has comidu..! 10 <kclinr in falling by 4 p<'fcr nl (y-o-y) in OClohn this year. The c<1ltral bank iow...-ed its hy policy ralr from 9 percent p<'fcr nt OIl D..crmbn hdp spur c[..dilto the private s",lor. At th... """" tim.\ banks ar ~ d ~an.ing up th~ ir balanc~.h~.s and impro"ing rid IlJ.'lnag ~ mnl1. with a 5si.tanc ~ from th~ Fwm. Th ~ cnllrnl bank has also.t"'pp"d up its banking su~f\' ision. whil~ conlinuing to mak ~ progr~ss in ils d fort 10 widen acc ~ss 10 financial sn-vi.c<'s. The Rwandan fl'anc ba.' b... n.-elatihly slabl.. iu 2009 I'efiectinl: in pal't tbe low demand by tbe pl'ivate, ectm' fol' ro... igu euhan::e. Th ~ CMlu-a1 bank is taking ""'3S\If<'S to cr~at.. a mor~ dtecliv.. intn-bank for ~ ign ~xchang.. mark... but progr<'ss has ~= slow. Th~ r.. al df~cti,-~ =chang ~ ral ~ depreciat.-d by 7 p.-rc=t in th.. first nin ~ months of r ~ fi.. ning th~ dedi... in th ~ inflalion '"t r ~v..._sing in pm1 th ~ larg.. r~al appreciation that ocrurr.-d in 2008 Stl'nctul'al I'efol'm, WHe implemented a, expected. On th.. fiscal front. CU5tOIllS administralion conlinu'" to irnprov.. wilh th.. rec=1 inrrodoctioll of cargo scann...-s. With significant a 5Si5lanc~ from donors. th.. gov..-nmnj.i is r.-doubling its..!torts 10 improv.. ils public financial rnrulllg=m systnn, including by.-nhancing th.. cov..-ag~ of exl<"r1lillly fulllnc.. d proj.-.:ts and audits of gov=1 3CCOWlIS. On th ~ financial. ide. 3 c.. mus of micro flllancial ins,inllions was luld.-rtak= wilh a vi.. w 10 pro,,;de pr"",,", O\-~rsighl and ~ nhanc~ public " cc~.s 10 cr..rur. Th ~ curr~nt social S<"C uri!y.ysl.-m is also ~ing r... ci ~w.. d "cith a,,; ~w 10...rnanc~ Ih ~ dfici.. ncy and deplh of th ~ financial,,",,ioc. Th ~ authooli... also impl.-m"nt ~ d,h~ l ~g al and imtructional fra"",work n ec~ ssary for " pri\"al ~ cr..rul bur~au 10 ~ fully op.-taliollll1 in ~arly 20 I 0 Outloo k Looking forwal'd,,be economy i, expected 10 l:t'3duan ~.-e co\ '"", ta, tinl: in 2010, albeit a somewbat mo,'e sl u::::i,b parp thau lll e, ' iou." ~ expected, due ' 0 uel:ati,... 'pillo.-e,", fl'om ~al GDP growth of.. ound 5 ~rc=1 is expect.-d foc "I'M recov.-ry "cill co",," from anothn- good agriculm.-al haf\ ~S1." gradual pick up in con,truction and srf\cic~ conling~nt on a turn.around in priva...,,",clor cr.-di, gro",b------and compl ~ tion of ou-going infrastrucmr ~ projects in wa'n- and dectrici!y.ectors. Inflalion is expect.-d 10 r.-main in singl~ digi... Th.. o,'..-all balanc~ Ofpaytll<"lllS is expect.-d 10 de... iora... Iightly in 2010 from lower tramf..-s and" pickup in impons r<'suiting from,h~ ~stic reco\"m)'. a!though inl...-national re...,,-<'s.hould r.-main al comfortabl ~ I~vds. D ~s pil~ a r.-duction in.~nding (a., p.-rcnll ofgdp) and UllChang.. d r~\" ~ nu<'s. th ~ o\" ~ rall fi,..,al balatlc ~ (including grants) is project.-d 10 widen in Staff supports th~ fisc"l.. a",, ~> which was ~ arlin- mdor-srd by th ~ Fund Board in th ~ con... xt ofth ~. ixth PRGF r~'';<"w. The fiscal deficil is ~ x pecl ~d 10 narrow o\" ~ r th ~ medium I..-m.. 52

67 3 Illlh~ phi",1 ahead, the authorities DHd to step up effort, 10 lnobiliu dolnestie res,,","r e, ami eu.>u1"p sufficient I, oli~' f1nibili~ ' lo maiuuill marroerollomic, tabili!;>-. Spttifically_ dtorts arr nrr<k<i 10 r :q>alld the tax base 10 the informal...,10.- and accdnak the r~o,-~ of budgt1ffl non-tax [ <vellu<'s. tl.iimdy f..mli:z~ [,""ripts and dri\-~s lic..us~_ AI the sa"", t;m.\ the authorities should connnue to strength..., fo<" ;&'1 rxchang.. op<ntions 10 allow ill.. =changr rate to adjust as n~<kd 10 changes in N:OllOmiC fun~ntals_ Likrw;s.-. proactive monetary policy and positih' rral int.-rrsi rates,,~ll hdp 10 control inflation and ~ n the IllliIuc;al sn:iof. 53

68 '''''' '~''''''''''''''''_(O 4.,......,,,0>0,,"'- NO <>O. {po<,"' ) 0"'... _ 1 Con... _{ond"... 1,-"...,,,,..,_....,...,'....,,_,'~.... U.. _ I.. _... tob i"' U~ _... ).. ""'_...,...' -...-,..... (100..., 1.. -~,-, "..... "'... ~ _ {OO,..,;... I H.."'......,...._ --"'...,. _. ~ ". "'on, "'-.'... 1 '''''''''''''{'--') C.p"..._ C..-t...._... _..,,...,_...,...,..,... -""".-." _o=.<...,,_... ~ {...", ' ' '''.._ tpo "",, I """,...,... {10..""... "'.....",.._ r., ~"'... ~ ""P _.."... I..., II"'"_..._. ~_;.. ",.,,... ", ". ",.. ",,-, H.' " ".1 ", '" ". _LO.... _102 ' 1.0 '",.... "..... '" " H.'..,.... ".. 1>.'..... _1O ~ ~, -, '-' ", "..,.. ". ". '" ".1... >1.1.,.. '".. _11.1 '",.. n,.. n, 13.1 _1M " ,., """ "_... 1>.'..... "'." ".., "... >.4.. "..... "', 13.1 ".... ",.,.. _1.'..."., _' fa "'...,.,.,.' "",... _-,- ~ ---.."'''""'...,.._ ,..- '" "'!-.. ".0 ".0.>S.'."., ".,.>.0 _>.0.. " _1.7.. "....., ". 11.',<.1 1, 11'~,~.. ~ un. "'2.',..'"" on ~"'''' "'_ '." 11' '''''''''''~ 2< ',,~... ""', 11' """"'" = ',""""'...,.".,'_...-"""---",, n ,..,-...,...,.. '.._.ou...- "' "..-.-" "'..._..._"'_"" _..--1"' '..."... <*> _.{.....,.,...""""... '_"' n.... "' ""''''_ 'A '''"''''''''-'''''--'''''-...,.,... " ".0...,, u " 20".".. "_od.... '", >l.' " '" " '" '" ".. ", ".0 ".0 " '" 11.' ~, -,'-' ". ".,,~.. ".... ",...,.. ", 11.' ''-I "., 10.' ".0 " ~, -"., ".. _,7.7,

69 ANNEX 4: COUNTRY AT A GLANCE (includes country map) Rwanda at a glance 11/ 28/ 09 Sub- Key Development Indicators Saharan Low Rwanda Africa income (2008) Population, mid-year (millions) ,296 Surface area (thousand sq. km) 26 24,242 21,846 Population growth (%) Urban population (% of total population) GNI (Atlas method, US$ billions) GNI per capita (Atlas method, US$ ) GNI per capita (PPP, international $) 860 1,869 1,489 Age distribution, 2007 Male Female GDP growth (%) GDP per capita growth (%) percent of total population (most recent estimate, ) Poverty headcount ratio at $1.25 a day (PPP, %) Poverty headcount ratio at $2.00 a day (PPP, %) Life expectancy at birth (years) Infant mortality (per 1,000 live births) Child malnutrition (% of children under 5) Adult literacy, male (% of ages 15 and older) Adult literacy, female (% of ages 15 and older) Gross primary enrollment, male (% of age group) Gross primary enrollment, female (% of age group) Under-5 mortality rate (per 1,000) Access to an improved water source (% of population) Access to improved sanitation facilities (% of population) Rwanda Sub-Saharan Africa Net Aid Flows a (US$ millions) Net ODA and official aid Top 3 donors (in 2007): United Kingdom United States European Commission Aid (% of GNI) Aid per capita (US$) Long-Term Economic Trends Consumer prices (annual % change) GDP implicit deflator (annual % change) Exchange rate (annual average, local per US$) Terms of trade index (2000 = 100) Growth of GDP and GDP per capita (%) GDP GDP per capita (average annual growth %) Population, mid-year (millions) GDP (US$ millions) 1,163 2,584 1,735 4, (% of GDP) Agriculture Industry M anufacturing Services Household final consumption expenditure General gov't final consumption expenditure Gross capital formation Exports of goods and services Imports of goods and services Gross savings Note: Figures in italics are for years other than those specified data are preliminary... indicates data are not available. a. Aid data are for Development Economics, Development Data Group (DECDG). 55

70 Rwanda Balance of Payments and Trade (US$ millions) Total merchandise exports (fob) Total merchandise imports (cif) Net trade in goods and services Current account balance as a % of GDP Workers' remittances and compensation of employees (receipts).... Reserves, including gold Governance indicators, 2000 and 2007 Voice and accountability Political stability Regulatory quality Rule of law Control of corruption Central Government Finance 2007 Country's percentile rank (0-100) 2000 higher values imply better ratings (% of GDP) Current revenue (including grants) Source: Kaufmann-Kraay-Mastruzzi, World Bank Tax revenue Current expenditure Technology and Infrastructure Overall surplus/deficit Paved roads (% of total) Highest marginal tax rate (%) Fixed line and mobile phone Individual.... subscribers (per 100 people) 1 7 Corporate.... High technology exports (% of manufactured exports) External Debt and Resource Flows Environment (US$ millions) Total debt outstanding and disbursed 1, Agricultural land (% of land area) Total debt service Forest area (% of land area) Debt relief (HIPC, M DRI).... Nationally protected areas (% of land area) Total debt (% of GDP) Freshwater resources per capita (cu. meters) 1, Total debt service (% of exports) Freshwater withdrawal (billion cubic meters) Foreign direct investment (net inflows) CO2 emissions per capita (mt) Portfolio equity (net inflows) 0 0 GDP per unit of energy use (2005 PPP $ per kg of oil equivalent).... Composition of total external debt, 2007 Short-term, 18 Energy use per capita (kg of oil equivalent).... Private, 0 IBRD, 0 Bilateral, 91 IDA, 204 World Bank Group portfolio (US$ millions) Other multilateral, 262 IMF, 8 IB RD Total debt outstanding and disbursed Disbursements Principal repayments Interest payments US$ millions IDA Total debt outstanding and disbursed Disbursements Private Sector Development Total debt service 11 2 Time required to start a business (days) 14 IFC (fiscal year) Cost to start a business (% of GNI per capita) Total disbursed and outstanding portfolio 2.. Time required to register property (days) 315 of which IFC own account 2.. Disbursements for IFC own account 0.. Ranked as a major constraint to business Portfolio sales, prepayments and (% of managers surveyed who agreed) repayments for IFC own account 0 0 Electricity Tax rates MIGA Gross exposure Stock market capitalization (% of GDP).... New guarantees Bank capital to asset ratio (%) Note: Figures in italics are for years other than those specified data are preliminary. 11/28/09.. indicates data are not available. indicates observation is not applicable. Development Economics, Development Data Group (DECDG). 56

71 RWANDA SELECTED CITIES AND TOWNS AKARERE (DISTRICT) CAPITALS INTARA (PROVINCE) CAPITALS NATIONAL CAPITAL ~ RIVERS -- MAJNROADS AKARERE (DISTRICT) BOJNDARIE Kilometers!!! i 10 i 20 i 30 Miles -- INTARA (PROVINCE) BOUNDARIES INTERNATIONAL BOUNDARIES 1 ~30'S DEM. REP. OF CONGO TANZANIA Th;. map ""'. prodvced by IheMopDe,jgnUnilofThe World flank. Thebotmdari..., co!or>,d"nominalion.ancl onyolmrinformalionshown ~~:::tno~:t:&:k Group,anrio.dgmenlonlhe 1egaI.IoIu.of"nylerrilory, or any enci"," " menr or i:,.,c:j,::.~c e of.vch 29"OO'E 29"30'E BURUNDI 57

72

73 ANNEX 5: RECENT ECONOMIC DEVELOPMENTS & MEDIUM TERM PROSPECTS 46 A. RECENT ECONOMIC DEVELOPMENTS Recent Growth Performance 1. Rwanda s US $ billion economy has remained resilient over the last five years, sustaining macroeconomic stability and real GDP growth despite the global economic slowdown and exogenous shocks that affected agricultural productivity. The economy grew by 11.2 percent in The high growth rate in 2008 was mainly due to the strong recovery in agriculture, which grew by nearly 15 percent, as a result of the expansion of the crop intensification program (CIP). This has contributed to an estimated average annual GDP growth for the period of 7.8 percent, which exceeds the EDPRS target of 7 percent. Table 5.1: Real Sector Growth and Contribution to GDP, Sector (Estimate) % GDP Growth Rate % GDP Growth Rate % GDP Growth Rate % GDP Growth Rate % GDP Growth Rate Agriculture Industry Services Adjustments Total GDP (Billion RWF) Source: National Institute of Statistics, Ministry of Finance, and staff estimates. Note: Base year is However, compared to other countries in the region, and Sub-Saharan Africa on average, Rwanda has experienced a relatively more severe reduction in growth trends in 2009, which pose a serious threat to poverty reduction. The World Economic Outlook estimates a 4 percentage point decrease in growth for Sub-Saharan Africa (SSA) on average to an estimated level of around 1 percent growth in In the Great Lakes region, growth is expected to slow from 5.8 to 4.3 percent, or a reduction of 1.5 percentage points. In Kenya, Tanzania and Uganda, growth is expected to slow down by roughly 2 percentage points for each country, leading to estimated growth levels in 2009 of 2, 5 and 7 percent, respectively. These trends contrast with the severe reduction in growth for Rwanda from 11.2 percent in 2008 to an estimated 3.3 percent growth rate in This represents a 7.9 percentage point reduction in growth between 2008 and 2009, or is close to a 4 percentage point lower level of growth in 2009 compared to the average growth of 7.5 percent between 2005 and These trends raise substantial risks to meeting the target of halving poverty by 2020 since the EDPRS estimates that GDP will need to growth by at least 8 percent per annum to achieve this target. 3. The 3 to 4 percentage point reduction in GDP growth, from 5-year historical trends, to levels estimated for 2009 is due both to the global crisis and domestic liquidity 46 PRSG-6 Program Document, February 2, The estimate shown is consistent with the latest Bank/IMF analysis. The GoR s LDP cites a lower figure. 58

74 issues. Lower estimated growth between 2008 and 2009 results principally from stagnation in the industry and services sector (as shown in Table 5.1). Estimated agricultural growth during 2009 reflected both lower commodity export growth and sustained high growth in the food crop sector. Continued investment under the Government s Crop Intensification program helped to raise productivity and mitigate the effects of the global crisis. 4. Overall, the global financial crisis, has led to lower export performance, GDP growth and revenues. The impact of the global economic and financial crisis is expected to manifest through three main channels: (a) declining external demand for Rwandan goods and services (mainly for tourism and mineral exports); (b) slumping commodity prices, particularly for minerals; and (c) scarcer financial inflows, including Foreign Direct Investment (FDI). To date, the global crisis has had a limited effect on FDI, tourism and remittance receipts. The situation for these areas will continue to be closely monitored, as there are likely to be larger lagged effects. However, there has been a substantial impact on export receipts which dropped by 32 percent in the first six months of 2009, while exports of goods and services are projected to decrease by 25 percent (f.o.b., US Dollar terms) in 2009 compared to Lower export growth as well as the stagnation in the industry and services sectors stems both from the liquidity constraints that arose in the latter part of 2008 due to structural issues in the banking sector coupled with the effect of the global financial crisis on domestic expectations. Credit to the private sector dropped by as much as 24 percent or from RwF 94.4 billion in the first half of 2008 to RwF 71.7 billion over the same period in The details of the issues that contributed to tight domestic liquidity are outlined below (under the section that provides an overview of financial and monetary performance). 5. Prudent implementation of macro-economic policies and protection of priority spending has enabled the Government to manage the impacts of the crisis. Government adopted various measures in order to protect priority spending and limit the extent of the growth slowdown: (i) it reduced spending in non-priority areas (while maintaining priority spending); (ii) it drew down foreign reserves; and (iii) it allowed the fiscal deficit to widen. Policies aimed to reduce inflation and improve the liquidity situation were also adopted and are discussed in further detail below. 6. Inflationary pressures decreased in 2009 Figure 5.1: Trend in Inflation as international commodity and food prices fell and government managed the exchange rate leading to its improved flexibility while increasing the coordination between monetary and fiscal policies. Slow growth in housing and food prices led to a decline in inflation by July 2009 to 5.8 percent, down from the high level of 22.3 percent in December Finally, both the global slowdown as well as government policies to contain prices contributed to the reduction in inflation. Government policies included: (i) controlling pump prices by offering taxation subsidies to importers; (ii) tightening monetary policy to reduce broad money growth from 30 percent in 2008 to around 11 percent in 2009; and (iii) tightening fiscal policy by reducing expenditures. As a result, the annual average and end of period (December 2009) inflation rates are projected at 12.0 percent and 6.0 percent respectively. 59

75 External Balance 7. Import growth exceeded the increase in exports leading to a deterioration of the balance of payments in 2008 and The growth in imports is primarily led by public investment which drove imports of machinery and equipment in 2007 and The fall in exports in 2009 is due to slowing global demand and will be partially offset by a 6 percent reduction in imports. Due to the slight decline in coffee and tea prices, it is projected that the value of their exports (coffee and tea) will increase moderately from US$ 46.9 million and US$ 40 million respectively in 2008, to US$ 50.6 million and US$ 47.9 million respectively in Foreign Direct Investment (FDI) and other capital inflows are also estimated to be lower in 2009, resulting in the deterioration of the overall balance of payments, which is estimated to be in a deficit of US$ 27.4 million in 2009, compared to a surplus of $64.7 million in Figure 5.2: Exports, Imports, and BOP Balance, (in US$ million) Source: Rwanda authorities and IMF, Estimates. 8. There will be a widening current account deficit in 2009 compared to 2008, resulting in part from a worsening in the terms of trade. The current account deficit (including official transfers) is estimated to have worsened from 5.5 percent of GDP in 2008 to about 6.6 percent of GDP in 2009, driven by the continued slowdown in global demand. Accrual of the special SDR allocation from the IMF helped to reconstitute foreign reserves from a level of 4.7 months of imports at end-2008 to 5.5 months of imports by end Current projections are that over the fiscal year 2010/11, (beginning in July 2010), the ongoing global financial crisis will lead to lower donor commitments, lower export receipts and a projected financing gap. Therefore, a widening of the current account deficit could be compounded by the projected decrease in donor grants should there be a delayed recovery from the global financial crisis. This would pose significant risks to the Government s program, making it difficult to continue to protect priority spending. 60

76 Table 5.2: Selected Economic Indicators, Economic growth and inflation(percentage change) 2005 Actual Actual 2007 Actual 2008 Actual Real GDP (percentage change) Real GDP (per capita) Consumer price index (eop) Broad money(m2)* Central government budget (percent of GDP) Revenue Grants Government expenditure and net lending Est Proj. Current expenditure Capital expenditure Domestic fiscal balance (excl. demobilization spending) Overall balance (payment order) After grants Before grants National accounts (percent of GDP) Gross domestic investment Of which: Private Gross national savings Current account bal.(excl. grants) Balance of payments (percent of GDP) Exports of goods and services Imports of goods and services Current account balance (incl. grants) Overall balance External debt (percent of GDP) External debt service (percent of exports of G&S) Gross official reserves (months of imports of G&S) Nominal GDP (billions of Rwandan francs) Sources: Rwandan authorities; IMF and Bank staff estimates. Note: M2 is defined to be currency in circulation and deposits. The ratio of M2/reserves is defined as broad money (M2) as a ratio to foreign exchange reserves. M2 targets for 2009 allow private sector credit growth of about 20 percent. M2-multiplier is defined as broad money (M2) as a ratio to base money. 9. To ensure Government s ability to protect priority spending, additional funds from the Crisis Response Window of $29.8 million will be disbursed under PRSG-6, bringing the total amount of the operation to $115.8 million. These funds will be used to maintain poverty reducing spending in education, agriculture, and infrastructure. In education, funds will go toward accelerating the 9-year basic education program for the payment of teachers' salaries, capitation grants, books and other materials and feeding costs. In agriculture, funds will help maintain the Government s planned activities under the Crop Intensification program to improve agriculture productivity with an emphasis on capacity support to farmers in the use of improved inputs and access to rural finance. As shown above, 2011 Proj.

77 Government interventions to improve agriculture productivity and access to markets have been instrumental in mitigating the effects of the crisis and improving Rwanda s resilience to external shocks. In the infrastructure sector, funds will support investments to improve the road network with an emphasis on road maintenance, particularly in rural areas. The Crisis Response Window funds will therefore ensure that ongoing efforts in these areas are not compromised, particularly if the global slowdown persists. Fiscal Performance 10. The expansionary fiscal stance of the past year or two has slowed down in Total expenditure and net lending remained more or less at the same level in 2009 as in Spending increases in priority sectors (i.e., education, health, agriculture and transport) that occurred in 2008 have been maintained in 2009 (see Table 5.3) with the exception of education. (However, Crisis Response Window funds will be used to fill the financing gap for the 9-year basic education program and help protect priority spending in agriculture and roads) 48. Government has also continued to support increasing transfer of funds to the Districts and to maintain investments (as shown by the trend in transfers in Figure 5.3). However, slowing growth and lower inflation will lead to a decline in revenues to historical levels (from 14.9 percent of GDP in 2008 to an estimated 12.5 percent in 2009). This will result in a widening of the fiscal deficit before grants from 11 to 13.8 percent. 48 Specifically, receipt of these funds will enable GoR to accelerate the 9-year basic education program, including the one child one computer project. Funds will be used for the payment of teachers salaries, capitation grants, provisions of books and other materials and feeding costs. The implementation of these programs has resulted in scaling up of education sector expenditures considerably and the Fast Track Initiative funds are not enough to carter for these outlays. 62

78 Table 5.3: Consolidated Recurrent Expenditures, (% of Total, incl. Debt Flows) Jan-Jun 2009 Primature Internal affairs Agriculture Education Health Infrastructure o/w road maintenance Land, Natural Resources, Environment Local administration o/w CDF Minicom All other Ministries Total Source: MINECOFIN, Budget Department and World Bank Staff estimates. Note: CDF stands for the Common Development Fund, an investment fund that channels resource to Districts. 11. A gradual reduction in resources was planned for the period 2009/10, based on anticipated lower levels of revenues and external resources (compared to disbursements in 2008). The 2009/2010 budget submitted to Parliament envisaged a reduction in spending of around 0.4 percent of GDP (or 1 percent of GDP during the period January to June 2009). In addition, to maintain priority spending, Government drew down foreign reserves. Looking beyond 2010/11, projected lower levels of donor grants currently imply a financing gap of around 2 percent of GDP. Additional funds will therefore be needed to ensure continued protection of priority spending in line with the EDPRS priorities and to provide a buffer against potential future shocks. In response to this additional need, the Bank has applied to the Crisis Response Window (CRW), to receive an additional US $29.8 million, to be disbursed through the PRSG-6 operation. Monetary and Financial Sector Policies and Performance 12. In the financial sector, aside from the liquidity ratio, financial indicators remained within their prudential benchmarks; and the stability of the banking sector has at present not been adversely affected by the global financial situation. All 8 commercial banks have been in compliance with the minimum capital requirements of approximately US $9.2 million. Reported foreign exchange exposure ratios have been within the prudential BNR benchmarks; and the quality of the loan portfolio has improved. However, the current liquidity shortage in the banking sector could be exacerbated by the impact of the global financial crisis leading to a reduction in the availability of credit. This effect coupled with the real effective appreciation of the exchange rate in 2008 by 22 percent (due to high domestic inflation), would increase the demand for foreign exchange and imports, and undermine Rwanda s competitiveness. 13. Rwanda s financial sector moved from a situation of a structural excess liquidity to a potential structural liquidity deficit partly due to efficiency gains from ongoing 63

79 reforms in the sector as well as NBR s low interest rate policy. The past policy of low interest rates led to strong credit growth in while discouraging savings. As a result, credit to the private sector increased above 30 percent in late The global financial situation also contributed to these trends. Low nominal interest rates combined with high inflation encouraged depositors to look for better yielding investments, which resulted in partial withdrawal of deposits from the banking system. 14. The liquidity shortages exposed weaknesses in risk management and governance of commercial banks, i.e. a mismatch in credit maturities (for bank borrowing and lending). This prompted the banks to tighten credit standards leading to a decline in credit growth. The liquidity problems in the banking sector became apparent in the fourth quarter of 2008, as the loan-to-deposit ratios increased and banks reduced their purchases of NBR and government s financial instruments. 15. In response to the tightening liquidity situation, the BNR injected liquidity of approximately 1.5 percent of GDP in the economy. This was aimed at providing immediate liquidity, sustaining credit expansion, and boosting economic growth in Government repaid RwF 18 billion worth of T-bills by end-2008 and reduced the reserve requirements from 8 to 5 percent. Also, during the first part of 2009, government eased its low interest rate policy. This caused the average real interest rate for deposits to increase, although it remained negative. The stock of private sector credit increased to RwF 7.1 billion in June 2009, and remained above its mid-2008 level (of RwF 5.1 billion). 16. Despite the injection, broad money supply (M2) growth slowed and is projected to be negative in Growth of broad money (M2) is projected to have slowed from a level of 24.2 percent in 2008 to 2.6 percent in This contrasts with growth of 30.8 percent in 2007, and reflects the tighter monetary policy and reduction in the monetary overhang. 17. Reserve money remains the anchor of the monetary program. GoR s reserve money objective consists of an active use of interest rate policy supported by a proper communication strategy to help anchor interest rate expectations, reduce volatility, and strengthen the monetary policy transmission mechanism. The quantitative targets for reserve money were rebased to reflect the reduction of the reserve requirement and to accommodate a buffer of excess reserves (RwF 4 to 5 billion) that would support the efficient operations of the payments system and the inter-bank market for short-term liquidity. Debt Sustainability 18. The most recent DSA dated December 2008, concludes that Rwanda is now at a moderate risk of debt distress owing to recent debt relief and its narrow export base. With the exception of the net present value of exports to debt which breaches the threshold of 150 under extreme stress tests, all debt indicators remain below their respective thresholds. Exogenous shocks to exports or imprudent borrowing on non-concessional terms could cause a rapid deterioration in the medium-term outlook. A more recent and preliminary assessment presented in the sixth review of the PRGF indicated that the medium-term external debt sustainability has improved since the December 2008 assessment, and that the nonconcessional borrowing, for upgrading and rehabilitation of roads in and around Kigali, would not significantly worsen the debt outlook (see Box 5.1). 64

80 19. The amount of domestic debt is small and expected to decline further as the BNR limits the use of domestic debt instruments and instead relies mainly on foreign exchange sales for sterilization so as to increase exchange rate flexibility. Greater flexibility of the nominal exchange rate will be essential to ensure the adherence to quantitative monetary targets while containing excessive real appreciation of the Rwandan franc. In addition, the GoR remains committed to limiting external debt and keeping to a minimum 50 percent grant element. 20. GoR has taken steps to prevent the reemergence of unsustainable debt by preparing a debt management strategy with limits for loans, government guarantees and contingent liabilities that will guide future borrowing. The Government continues to take strong steps to improve debt management practices. In 2008, the GoR developed and published a Debt Policy and Medium Term Debt Strategy. This ensures conformity to the best practice in public debt and the associated management of risk(s) to reduce the possibility of falling into an unsustainable debt position. The Bank is providing technical support in the implementation of the strategy. In addition, Government has undertaken complementary reforms to ensure that public spending is effective and sustains growth. Specifically, a Public Financial Management (PFM) action plan was recently adopted which aims to consolidate and strengthen reforms to improve the implementation of the MTEF and the budget monitoring system. A framework for implementing a Public Investment Program (PIP) was also recently elaborated and adopted in February The aim is to ensure good coordination of investment across and within sectors, establish an analytical and information base for prioritizing investments and support the strengthening of capacity in this area. Box 5.1: Rwanda: Update of Selected DSA Indicators The ratio of the PV of external public and publicly guaranteed debt to exports of goods and services (traditionally the weakest DSA indicator for Rwanda) has improved considerably. Part of this improvement is attributable to the change in estimation methodology for tourism receipts. The old methodology yields lower receipts compared to the recently adopted methodology. There has been some concern expressed by the authorities and staffs of both the Bank and the Fund on the potential over-estimation of tourism receipts. The Bank and Fund staffs support the authorities efforts to refine these estimates. Nevertheless, whether the new or old methodology for estimating tourist receipts is employed, Rwanda remains at a moderate risk of debt distress. Rwanda. External PPG Debt-to-Exports Ratio Under Alternative Scenarios Most extreme stress test Baseline using old tourism receipts data Baseline 100 Baseline Without one-time relaxation of concessionality policy Note: 1 Export value growth at historical average minus one standard deviation in Source: Sixth Review Under the Three-Year Arrangement Under the Poverty Reduction and Growth Facility 65

81 B. MEDIUM TERM PROSPECTS AND FINANCING REQUIREMENTS 21. The GoR s medium term outlook for real GDP growth has been revised downward slightly to roughly 5 percent. Growth will continue to be based on increased productivity in the agricultural and mining sectors, and strong public and private sector investment activity. This outlook is contingent on favorable weather conditions in the next few years, continued implementation of policies to promote commodity export and successful implementation of the medium-term policy agenda as outlined in the EDPRS on five fronts: (i) carrying out Government s planned investments in infrastructure; (ii) improving agricultural yields, supported by actions in the new fertilizer strategy; (iii) sustaining progress on the private sector development program; (iv) accelerating overall export growth and diversifying the export base; and (v) taking actions aimed at enhancing the efficiency of investment as outlined in the new Public Investment Policy. 22. A projected annual GDP growth of between 5 and 6 percent would be sustained through increased public and private sector investments to improve agriculture productivity and address infrastructure constraints to growth. Reflecting the strong emphasis that GoR is placing on public infrastructure investment, overall investment rates are projected to rise fairly steadily from about 21 percent of GDP in 2007, to about 24 percent of GDP by Public investment will increase from 8.6 percent in 2007 to close to 10 percent of GDP by With the private sector s engagement in the social sectors and also in infrastructure expected to be strengthened, private investment would also trend upward gradually to about 12 percent of GDP per year, in the medium-term. 23. Strong import growth stemming from higher capital imports associated with the expanding public investment program, is expected to outpace moderate expansion in exports, resulting in a higher current account deficit. Export volumes led by receipts from traditional exports and tourism, are forecasted to increase at a relatively slower rate as the GoR implements its export promotion strategy. This is expected to widen the external current account deficit considerably from about 5 percent of GDP (including grants) in 2007, to about 9 percent of GDP (including grants) by Government capital spending would increase strongly over the medium term, in line with the need for increased investment to accelerate and sustain growth, causing the overall fiscal balance to remain in deficit. The deficit (including grants) would widen to about 3.2 percent of GDP by 2011, up significantly from under 1.5 percent in This increase would be due to higher investments. 25. The share of priority spending will increase annually by about 2 percent of GDP depending on the availability of sufficient resources. More resources will be dedicated to public infrastructure, agriculture, education, health and water. Increased spending in the education sector will go towards improving delivery of primary and post-basic education and vocational training and skills development. In the heath sector, focus will be on financing programs to improve primary health care, financing of the mutuelle scheme and purchase of drugs. Infrastructure spending will be geared towards addressing the issue of limited and insufficient energy through construction of the hydro power dams, the generation of power from the Lake Kivu Methane gas power plant, and rehabilitation of infrastructure, including in the water and transport sectors. At the decentralized level, transfers to Districts are projected to increase as the Government implements its fiscal decentralization policy, with increases in 66

82 capitation grants in education and performance based financing in the health sector, feeder roads financing in the transport sector and funds to cover District overhead costs. 26. Beyond the allocation of funds for poverty related expenditure, improving the efficiency of such spending and its targeting to the poor remains a significant challenge. The GoR is therefore focusing on improving the prioritization and efficiency of expenditures. Efforts are being made to strengthen the institutional framework for the Public Investment Program. To date, a public investment policy has been adopted and guidelines for identifying investment selection criteria and project prioritization that maximize the returns on investment, and safeguards public debt sustainability are being developed. The Government has also embarked on the process to institutionalize public expenditure reviews with the aim of strengthening public expenditure management and monitoring of spending. The intent is to strengthen the expenditure prioritization process within sectors and improve accountability of line ministries. 27. Over the medium term, budgetary grants will continue to be the main source of financing for the EDPRS. The projected increase in investment will continue to rely on grant financing. To date, the global financial crisis has not had an adverse effect on donor commitments. Table 5.4 summarizes commitments over the medium term. Precise comparisons are difficult to make since projections are now based on the newly adopted fiscal year (July to June) as opposed to the previously used calendar year. Although, current commitment levels have not yet been affected by the global slowdown, projections beyond 2010/11 are based on firm commitments and assume lower levels of disbursements. Government is prepared to limit Table 5.4: Donor Budget Support Commitments, (US$ million) Jan.- June 2009 Actual July June 2010 July June 2011 July June 2012 General Budget Support* World Bank AfDB DfID EC Germany COMESA (CET) 15.2 Sector Budget Support Other Budget Support Total Source: MINECOFIN and Donors. Note: Figures for July 2009-June 2010 is net of potential CRW funds of $29.8 million. spending and draw down on reserves should there be a reduction in commitments due to a sudden deterioration of global economic conditions; or perceived violation by donors of the underlying principles of the Partnership Agreement as occurred with the recent DRC- Rwanda joint security operation. 28. Consistent with the medium-term policy objective of maintaining macroeconomic stability, the Central Bank is expected to contain inflationary pressures from any fiscal expansion and scaled-up external assistance. Average inflation is expected to decline from around 15 percent in 2008, due to rising food, transport and energy prices, to single digits by The reduction in inflation reflects the price pass through from declining international food prices and domestic policy. Projections are that with strong implementation of monetary policy, inflation could decline to about 5.2 percent per annum during

83 Table 5.5: Financing Plan, (millions US$) / / / Actual Estimated Projected of FYs* Financing requirements (incl. IMF) Current account deficit (excl. off. transfers) Long-term amortization (excl. IMF) Change in official reserves (increase +) Change in other foreign liabilities (increase +) IMF repayments Change in arrears (decrease -) Financing source Private flows FDI Other private financing flows (net) Public sector Official transfers o/w PRSG 4-8a/ Official capital grants Debt forgiveness Long-term disbursements excl. IMF of which IDA IMF purchases Financing gap Sources: World Bank and International Monetary Fund staff projections. Note: 1. PRSG 6 disbursements (for 2009/10) include projected amount for Rwanda of $29.8 million, from the Crisis Response Window, on top of the originally committed amount of $86 million. *Sum of three fiscal years (FY) 2009/10, 2010/11, and 2011/12. Adequacy of the Macro-Economic Policy Framework 29. Overall, the medium term macro-economic framework is assessed to be sound and adequate to support the Government s program. As noted in the completed sixth review of the IMF s PRGF, the medium-term outlook, including inflation, economic growth and financing remains favorable. GoR will continue with policies to mitigate the negative impacts of global and external shocks. It will exercise prudence in implementation of macroeconomic policy with a focus on ensuring coordinated fiscal and monetary policies. Government will also continue to pursue policies that support private sector led growth through the development of a Public Investment Program, implementation of the Financial Sector Development Plan and the Rural Development Program. Investments are planned to improve agriculture productivity, reduce soil erosion and improve irrigation. There is also evidence that the investments in the energy sector aimed at reducing the dependence on thermal power is paying off with the progress made in power generation in the Lake Kivu electricity project. This along with investments in the Nyaborongo hydro-power project will help reduce the cost of energy to firms and households. GoR is also now in the process of developing a strategic plan for post-basic education to address the issue of skills development. These measures should help to expand productivity and establish the foundation for long-term sustainable growth. 68

84 ANNEX 6: OVERVIEW OF UBUDEHE 1. An important, relatively early aspect of SP development within Rwanda was the reestablishment of Ubudehe, a traditional Rwandan process of village-level community decision-making based on discussion, gathering of systematic information, and decisionmaking through community meetings in which all villagers participate. 2. Ubudehe is a traditional practice and culture of collective action to solve community problems. It has been adopted by GoR as an approach to fight poverty. It is present at 9,000 cells all over the country. The 2006 Ubudehe survey results suggest it should be a core method of GoR s approach to fighting poverty because it is embedded in Rwandan mind sets, is seen as an approach that benefits the poor, primarily through increasing access to livestock, promotes social capital and unification, increases participation in planning and decision making, and empowers the poor and women Under Ubudehe, two village committees are elected: one to oversee implementation of communal decisions, and the other to function as a sentinel or oversight committee that activates itself only if concerns arise about the functioning of the first committee. These community-driven processes have been used in all villages for multiple purposes, including identifying and overseeing projects to address community problems, identifying village priorities as part of local government planning processes, and identifying the poorest households in the village that would receive community-managed income support. The last of these functions entails a systematic process through which each village has created its own geographic and social map of the villages which shows physical features and structures, each household s home, and codes for relevant socioeconomic characteristics of each family. This poverty-mapping process, which has a systematic methodology, allocates each household to one of six ordinal income and poverty-related categories differentiated by well-defined qualitative criteria. The maps have become the basis for community decision-making about income support in cash or in kind to ultra-poor households from local and donor sources, and about which households should receive subsidies for health insurance. More recently, similar processes have been used to identify community priorities for consideration within District government plans for public works programs. Similar approaches will be used, at least in the initial phases of the pilot VUP programs, to select households to receive assistance and, separately, to select community assets to be created through the VUP public works component. 4. Development partners played a significant role in the redevelopment of Ubudehe by providing technical support for development of related skills within local communities (e.g., DFID) and by offering operational funding (e.g., European Union) for each village community to select, design, and oversee the contractual implementation of a substantial locally selected investment project to address what each village perceived to be its most challenging solvable problem. Resulting projects included modern market premises, roadbuilding, new primary health care or education facilities or their foundations, and even a small, hydroelectric system powered by a local stream. 49 EDPRS, p.3, footnote 2 69

85 5. The pilot Sectors for the implementation of the VUP program have been selected based on a set of criteria such as food security, water access, distance to education and health center, and level of village settlement (see Table below): Table 6.1: Vision 2020 Umurenge - Targeting the Most Disadvantaged Umurenge in each District Selection criteria: Food security, water access, distance to education, distance to health center, and level of village settlement (Imidugudu) Columns 1 to 3 in each panel show the province name, district name, and selected pilot sectors. Kigali 1 Gasabo Gikomero 2 Nyarugenge Mageragere 3 Kicukiro Gahanga Nothern 4 Gicumbi Rubaya 5 Rulindo Cyinzuzi 6 Burera Gitovu 7 Musanze Gashaki 8 Gakenke Minazi Eastern 9 Bugesera Kamabuye 10 Gatsibo Kiziguro 11 Kayonza Nyamirama 12 Kirehe Mahama 13 Ngoma Gashanda 14 Nyagatare Rwempasha 15 Rwamagana Fumbwe Western 16 Karongi Ruganda 17 Ngororero Muhororo 18 Nyabihu Rurembo 19 Nyamasheke Mahembe 20 Rubavu Rubavu 21 Rusizi Nyakarenzo 22 Rutsiro Manihira Southern 23 Gisagara Gishubi 24 Huye Maraba 25 Kamonyi Nyarubaka 26 Muhanga Rugendabari 27 Nyamagabe Kibumbwe 28 Nyanza Kibirizi 29 Nyaruguru Rusenge 30 Ruhango Mwendo 70

86 ANNEX 7: DECENTRALIZATION REFORMS IN RWANDA 1. GoR s economic reforms have focused on improving public financial management, deepening the financial sector and removing constraints in key economic sectors such as agriculture, power, transport and water. Attention has also been given to improving the legal and regulatory environment for business as well as to strengthening the public service and building its capacity. Several measures have been taken, including introduction and increased application of the MTEF. 2. GoR adopted a National Decentralization policy in May 2000 to achieve three main goals: (i) good governance, (ii) pro-poor service delivery, and (iii) sustainable development. More generally, the reforms sought to boost community participation, accountability and transparency, sensitivity and responsiveness of public administration, sustainable capacities at local levels, and effective, efficient planning, monitoring and service delivery. This resulted in the restructuring of local governments into 5 provinces, 30 Districts, 416 Sectors and 2,148 cells, covering an estimated total of 15,000 villages across the country. The provinces serve as the coordinating arm of the central GoR while the Districts have their own selected councils and mayors and prepare their own budgets, implying that Districts have their own taxing powers and are responsible for service delivery, economic development coordination, and planning. Sectors and cells, through elected representation, serve as constituencies within Districts and hence provide an important vehicle for the citizens voices. Although Sectors are assigned more fiscal and service delivery responsibilities, including administration of basic services, management of community development, and collection of statistical data and tax revenue collection (of which 50 percent is handed over to the Districts), the cells main responsibility remains community mobilization 50. To support decentralization reforms, GoR is strengthening monitoring and appropriate oversight of transfers to the Districts to ensure accountability and performance and to align these transfers with monetary and fiscal policies. Table 7.1: New Administrative Structures Entity # of Units Responsibilities Administrative Structure Province (incl. Kigali City Council (KCC)) 5 - Coordination of local development plans - Strategic planning for urban development; monitoring & evaluation; provision of complex services and those cutting across district governments. - Governor for Provinces - Mayor for CC - Administrative units District 30 - Economic development coordination - Planning & coordination of services Sector Administration of basic services (plan, implement & monitor service delivery) - Management of community development - Collection of statistical data - Collection of taxes (50% of taxes collected by sectors is handed over to the districts) - Elected council (mayor) - Executive committee - Technical staff - Elected council - Technical staff Cell 2,148 - Community mobilization - Council (all adult population) - Executive committee Source: "Making Decentralization Service Delivery Work in Rwanda", Republic of Rwanda, September 2006, pg Source: Making Decentralization Service Delivery Work in Rwanda, The Republic of Rwanda, September 2006, pg

87 3. Implementation has occurred in three phases. Phase one ( ) established democratic and community development structures and enhanced their capacities through legal, institutional, and policy reforms. These covered roles and responsibilities, financing of services, and mechanisms to ensure accountability. Phase two ( ) consolidated and expanded decentralization by emphasizing local service delivery through well-integrated, accountable networks that empower communities to participate more fully in planning and managing local affairs. Phase three (post-2008) will continue to improve, support and sustain the efforts of phases one and two. 4. A key component of Rwandan decentralization is the introduction of IMIHIGO, a performance management tool for reform introduced in Borrowed from Rwanda s traditional culture and values, the term now designates a written one-year performance management contract between the President and each District mayor on behalf of local constituents and extending to all levels of local government from State, to Provinces, to Districts, to Sectors, and culminating in a series of pledges made by households to their communities or villages (imidugudu). The contract is recorded publicly and presents a set of development targets backed by specific, measurable performance indicators, creating visible accountability mechanisms and incentives for local officials and communities to ensure implementation of decentralization policies so as to meet regional and national development targets such as those set out in the EDPRS and VUP. IMIHIGO contracts are now an integral part of GoR s governance and decentralization reforms and are important in the context of the VUP program as it provides a basis for mainstreaming the development of behavioral changes needed to make the VUP program a success, including commitment to a savings culture and commitment to undergo needed training. 5. Empowerment of local government and citizens rights down to village level is a key objective of both decentralization and VUP implementation. The Government has progressively devolved responsibilities and transferred resources to decentralized levels of the administration Province, District, Sector (Umurenge), Cell (Akagari), and Village (Umudugudu). Of these, the last three are subordinate tiers of the District governments. Rwanda uses the term local government to encompass the District through village levels. 6. The second phase of decentralization has highlighted the urgent need to put in place reporting mechanisms and elaborate a framework for oversight. To this end, GoR developed the Rwanda Decentralization Strategic Framework (RSDF) in 2007 that provides a framework for the coordination of technical and financial resources in support of decentralization. This builds on key national strategies, the MDGs, EDPRS, VUP, and the Aid Development Strategy. Its Decentralization Implementation Program (DIP) serves as a road map toward phasing out the provincial level of Government. 7. The current system of intergovernmental transfers is based largely on a recurrent block grant to finance administrative costs (including salaries), and a development grant through the Common Development Fund (CDF) to fund capital projects. In addition to these flows, a significant level of international funding contributes to local development priorities on either an on- or off-budget basis. The Fiscal and Financial Decentralization Policy, adopted by GoR in 2006, sets the objectives and guiding principles for the future reform work. This transferred to Districts 30 percent of budget resources and associated implementation and reporting responsibilities. 72

88 8. When the MTEF was introduced in the central administration, GoR decided to extend the approach simultaneously to local governments. Progress has been in the development of local government budgets and the incorporation into the MTEF process. Ministry of Finance is drafting new planning and budgeting guidelines that will cover the various phases in the local planning and budgeting process to streamline and strengthen District financial management. Local governments are required to submit monthly reporting of financial transactions that meet transaction requirements to ensure receipt of intergovernmental transfers, which remain the main source of revenues for Districts. 9. Decentralization has devolved decision-making and budget implementation to District-level authorities, and responsibility for service delivery to the Sectors. As weak administrative capacity has been a major challenge to decentralization and implementation of GoR s reform agenda, efforts in the area of capacity building and civil service reform have been, and remain, critical. Donors are providing support for capacity building, including the Decentralization and Community Development Project (DCDP) and the PRSC/G companion Public Sector Capacity Building Project which supports Rwanda s Human Resources and Institutional Capacity Development Agency (HIDA). HIDA is GoR s institution to coordinate all capacity building interventions, though its role is one of facilitator and not implementer. 10. The implementation of the VUP and of the delivery of basic health services through results-based financing will play a key role in operationalizing the roles, functions, and services at Sector levels. These administrative structures require additional support in capacity building and technical assistance beyond that already provided by the Districts. The VUP implementation will be coordinated by Sector administrations with participation at the Umudugudu level with technical support and supervision from the Districts; results-based financing of community health, nutrition and population services will rely on administrative capacity in health centers at the Sector level, which has been strengthened by the reforms and improved financing of recent years. The community-based participatory processes, Ubudehe, will play an important role within the VUP program, while also strengthening governance and accountability at Sector level. 11. Key initiatives that complement the decentralization reform include the public service reform strategy (2004), the territorial reform (2006), and ongoing civil service reform program. The last of these reforms incorporated the autonomization of service providers, the block-granting of schools and health facilities, the introduction of performancebased incentives, the delinking of all health workers from the civil service and the possibility given to service providers to hire and fire. 73

89 ANNEX 8: SOCIAL PROTECTION ANALYTICAL UNDERPINNINGS 1. The poverty assessment has provided important insights on poverty and vulnerability, geographic poverty, households characteristics that contribute to deepening poverty gap, and potential escape routes out of poverty. This operation makes full use of these findings, and builds on an understanding of past failures and future promising approaches to poverty reduction in Rwanda. Of particular relevance are the findings pertaining to the absolute level of the poverty gap by decile group (Table 8.1) confirming the vast distance the poorest groups have to traverse, evidently caused by a combination of a lack of assets and of gainful employment. This operation helps address these very facets of poverty. Table 8.1: Resource Gap to Reduce Moderate and Absolute Poverty in Rwanda Mean per capita consumption expenditure (RWF) Per capita income shortfall in 2007 prices from moderate poverty line (RWF) 51 Per capita income shortfall in 2007 prices from extreme poverty line (RWF) Consumption decile Household size Poorest Richest 10% All households Source: Extreme Poverty In Rwanda: Profile & Correlates, World Bank, Poor have income shortfall from the poverty line (negative values) and non-poor have income surplus over the poverty line (positive values). 74

90 ANNEX 9: VUP IMPLEMENTATION PROCESS FOR RWANDA s NEW SP POLICIES 1. The social protection (SP) policy changes that the CLSG series supports incorporate good international practice within the Public Works (PW) and Direct Support (DS) elements of Rwanda s new Vision 2020 Umurenge Program (VUP). VUP is one of three flagship programs under Rwanda s new five-year EDPRS. 2. Identifying Those Eligible. While other aspects of EDPRS assist poor agricultural households that own land, the PW and DS elements of VUP are directed at households that are both landless and extremely poor. In addition, for PW eligibility, all adults in the household who are capable of working must lack employment opportunities. Landless is defined as owning less than 0.25 hectares. Extremely poor households are those identified as falling within the two poorest of six household categories that are ranked ordinally by living standards through Rwanda s community-based decision-making process known as Ubudehe. (See Ubudehe overview in Annex 6). Potential VUP recipients will be identified through the poverty-map process, which covers all extremely poor households in the village. The VUP targeting at village level is a two step process (carried out in one day at village level): Step 1: The information is taken from the existing Ubudehe social/poverty map on the morning of the targeting day and filled in the targeting format. Step 2: A community meeting in the afternoon to which the Umudugudu (village) population is invited. At the meeting, the social map information is validated each household s Ubudehe category is updated, and information required to apply the VUP eligibility criteria is collected and added to the targeting format (household land and labor availability). The completed targeting formats are then sent up to Sector level. The DS and PW eligible households are then extracted from the village targeting lists and discussed and validated at the Sector Joint Action Development Forum (a forum including different partners, including civil society). The lists of mutually exclusive households eligible for direct support and public works are then sent back down to cells and villages for validation. These lists serve as the gate-keeping mechanism to identify and rank eligible households in each community. Each community specifies the criteria that distinguish the characteristics of the Ubudehe household categories based on living standards that are identified in the community mapping process. These are similar, but not identical, across the country. Some local observers consider that lower-income communities tend to set more stringent standards than higherincome communities when defining poverty levels. 3. Assistance Offered Public Works. For eligible households, the PW element of VUP will offer paid, intensive work under the pilot program to one member of the household for an initial (potentially renewable) period of six months. Where possible, to facilitate female participation, work-sites will be kept within an hour s walk of the community s center. Employment will be subject to project and funding availability and may be concentrated during slack agricultural seasons when private employment is less available. If the number of positions available is less than the number of eligible household members, priority will be 75

91 given to employing workers from the poorest categories of households. The PW element will be managed by staff from the Sector and cell tiers of local government. Payment mechanisms available for public works include direct deposit to workers bank accounts every two weeks, where feasible without any third-party involvement. Plans are in place for the provision of banking services at least on a part-time basis at locations convenient to most villages. Financial controls covering transparent payment mechanisms and accounting have been formulated. The majority of projects will be selected by the local communities through Ubudehe processes and incorporated within their Sector or District plans. Technical support will be arranged by program managers under established arrangements with local governments and national ministries. Environmental and technical matters will be handled in conformity with national policies and procedures. These have been incorporated within the VUP public works and financial management program manuals. 4. Assistance Offered Direct Support. The level of each eligible household s direct support payment will be based on the number of adults and children in the family, using a national scale, and will reflect economies of scale in larger families. Support available from other sources will be offset against this amount. Eligibility lists will be updated every six months. Monthly direct support will build on the payment process for public works Payments will be made by direct bank transfer. A list of clients and amounts received will be disclosed publicly at village, cell and Umurenge levels. Direct support recipients will have the option of saving a small part of their monthly allowance. Policy incentives for such personal saving are being considered. 76

92 ANNEX 10: INTERNATIONAL GOOD PRACTICE IN PUBLIC WORKS PROGRAMS Public Works Programs: Some Good Practice Design Features Drawn from International Experience A number of countries placed in diverse socio-economic circumstances countries emerging from long periods of conflict, or serious macro economic collapse, or serious natural disasters such as earthquakes or floods have introduced public works programs. There is much international experience that enables one to delineate some good practices in the management of such programs. In particular, five design features are critical for the success of any public works program. These are: the level of the wage rate, labor intensity, seasonality of operation, project selection procedures, and monitoring and evaluation. Box 10.1 provides a summary of these five and other good-practice design features based on the experience of various countries that launched such a program. Box What are the Key Design Features of a Good Public Works Program? To realize the full potential of a public works program as a poverty-reducing and risk-coping instrument, it is recommended that: The wage rate should be set at a level which is no higher than the prevailing market wage for unskilled manual labor in the setting in which the scheme is introduced. Restrictions on eligibility should be avoided; the fact that one wants work at this wage rate should ideally be the only requirement for eligibility. If rationing is required (because demand for work exceeds the budget available at the wage set) then the program should be targeted to poor areas, as indicated by a credible "poverty map". However, flexibility should be allowed in future budget allocations across areas, to reflect differences in demand for the scheme. If further rationing is required within a community, it is best to identify some critical, easily observable, household characteristics that best predict an individual or household s vulnerability, and select such individuals for participation. The labor intensity (share of wage bill in total cost) should be as high as possible, but not too high at 100 percent. No public works activity of some quality can be done with labor alone. The level of labor intensity will depend on the relative importance attached to immediate income gains versus (income and other) gains to the poor from the assets created. This will vary from setting to setting. The projects should be targeted to poor areas, and try to assure that the assets created are of maximum value to poor people in those areas. Any exceptions in which the assets largely benefit the nonpoor should require co-financing from the beneficiaries, and this money should go back into the budget of the scheme. Communities must be consulted in project selection. Public works should be synchronized to the timing of agricultural slack seasons. In order to encourage female participation, the appropriate form of wages is important for example, women can benefit from piece rates or task-based wages; sometimes wage in the form of food has attracted more women to work sites. Also, provision of childcare or preschool services can improve participation by women. Transaction costs to the poor are kept low one important means to accomplish this is through locating project sites close to villages. It is also necessary to ensure appropriate mediation of NGOs for protecting the rights of the poor vis-à-vis program managers. The program should include an asset maintenance component. Source: Subbarao (1997, 2003), Ravallion (1999, 2000). 2. Wage rate. To promote self-selection of participants into the program, it is best to fix the wage rate at a level equal to, or lower than, the prevailing market wage for unskilled labor. 52 This short note is drawn from Subbarao (2003), and Carlo del Ninno, K. Subbarao and A. Milazzo (2008). 77

93 International experience suggests that in countries where public works programs enabled propoor targeting (such as Bangladesh, Pakistan, India before 1988, and Republic of Korea), the public works program wage was maintained at a level lower than the prevailing market wage for unskilled labor (Table 10.1). Table 10.1: Program Wage Relative to Minimum and Market Wages in Public Works Schemes Country Program PW in relation to minimum wage (MW) and market wage (MKW) Bangladesh Cash for work, PW<MKW Pakistan IGPRA III, 1992 PW<MKW Philippines Cash for work, 1990 Food for work, 1987 PW>MKW PW>MKW Botswana Cash for work PW<MW but >MKW Kenya Cash for work, PW=MW>MKW Thailand 1998 crisis program PW = MW Republic of Korea 1998 crisis program PW = MW < MKW Maharashtra/India MEGS till 1988 MEGS after 1988 PW = MW <MKW PW = MW >MKW Chile Cash for work 1987 PW = c. 70 percent of MW/MKW Argentina Trabajar, date PW=MW<MKW PW<MW<MKW Source: Subbarao (2003). 3. Labor intensity. Depending upon the nature of the asset created and various other factors, there is much variation in the proportion of non-labor inputs needed to run a public works program. In general, a non-labor share of 40 percent appears to be the norm reached in most low income countries (Table 10.2). Table 10.2: Labor Intensity in Public Works Schemes Country Program Average labor cost as share of total cost Bangladesh FFW percent Argentina Trabajar percent range, and percent average Korea 1998 crisis program Around 70 percent Maharashtra/India EGS percent South Africa FFW (Western Cape) 39 percent, with range of 6-86 percent Source: Subbarao (2003). 4. Seasonality of operation. Ideally public works programs should operate at times of the year when the opportunity cost of labor is low. Experience with public works programs around the world suggests that indeed in most countries the program is operating intensively in agricultural slack seasons. However, in most countries the program is available throughout the year so it acts as a proxy program of unemployment insurance in countries where formal unemployment insurance systems have not developed. In other words, whenever an individual is unemployed for a short period of time, the program is available for him/her to seek work at the going market wage for unskilled labor. The maximum number of days a single individual could be employed is either left open-ended, or restricted (to, say 50 days or 75 days). 78

94 5. Targeting effectiveness. In general, in comparison with other safety net programs, public works programs around the world seem to reach ultra poor households better than other programs, provided the above mentioned design features are incorporated. Of these design features, the most critical one is of course the level of the wage rate. A relatively low wage rate seems to ensure that the poorest benefit from the program more than the rest, though it is possible that near-poor households can also be expected to participate in the program. Box 10.2 provides a summary of available international experience with respect to targeting effectiveness of public works programs. Box 10.2: Targeting Effectiveness of Public Works Programs: International Experience While there is a need to keep in mind the ultimate objectives of static and dynamic poverty reduction, targeting effectiveness clearly remains an important factor in assessing cost effectiveness of public works programs. In general, targeting performance has been good relative to other targeted public interventions. In Chile, almost 100 percent of participating households were poor. In Argentina s Trabajar program, around 80 percent of households were from the poorest quintile, and 60 percent from the poorest decile. In addition to the low program wage rate, this outcome was achieved by including the local level poverty rate in the scoring formula for determining program allocations. Various researchers have found positive targeting outcomes on the India s Maharashtra Employment Guarantee Scheme (MEGS), at least prior to the 1988 increase in minimum (and hence program) wages. A dedicated survey of MEGS participants from found 90 percent of workers living below poverty line, even including EGS earnings in the calculation, at a time when the poverty headcount similarly calculated was 49 percent. A later survey in found mean incomes of participating MEGS households to be around 20 percent below the poverty line. The positive targeting outcomes are confirmed by other analysis for There is also evidence that the scheme s targeting performance deteriorated when the wage rate rose substantially in In Bangladesh, targeting of the Food-for-Work Program was also found to be good, with program participants coming from poorer households than the rural population as a whole, and the total number of person days of work created even more pro-poor. While around 25 percent of all rural households had incomes below 1500 taka in , around 60 percent of FFWP participants had incomes below this level. Equally, around 70 percent of program employment went to the 25 percent of rural households with income below this level. Under the Bolivia Social Investment Fund, 77 percent of participants came from the poorest 40 percent of the population. Assessments of other SIFs in Honduras, Nicaragua and Peru have also found pro-poor incidence of SIF benefits. In Philippines, most public works participants were from marginally poor and non-poor households rather than the very poor. This outcome was largely driven by program compensation which provided for official MW in cash plus in-kind benefits which took the total compensation package above the market wage. In Indonesia, assessment of safety net programs introduced after the Asian crisis found that public works schemes which relied on self-targeting were much more likely to reach households which had suffered large shocks than programs which relied on administrative targeting methods, including subsidized rice, scholarships and health subsidies. Analysis from all-india NSS data on public works schemes for indicates pro-poor targeting performance, but that inclusion errors remained significant. Targeting performance was also better in comparison to both PDS (then untargeted) and micro finance program (IRDP). Sources: K. Subbarao (2003) and Carlo del Ninno, K. Subbarao and A. Milazzo (2007). 6. The reviewed experience suggests that it is important to bear in mind the good practice design features prior to introduction of a public works program. Failure to pay attention to the design features may render public work-fare programs both inefficient and expensive. 79

95 7. Project selection procedures. There is much cross-country variation in the actual selection of projects. Some good practices are worth noting. In Korea, for example, a combination of community selection, and the line Ministry selection, was followed. 50 percent of projects were selected by communities, and 50 percent by line Ministries. The Government was keen to ensure a igh labor intensity. To accomplish this goal, the Government estimated labor coefficients for various types of activities such as soil conservation, afforestation, garbage disposal, etc. and gave a list of potential candidates (projects) to communities. Communities selected projects from this list of labor-intensive activities that are most needed by communities. International experience suggests that local governments (and communities) are capable of selecting and implementing small projects (with a penchant for maintaining quality), but lack capacity to implement larger and more complicated projects. (Subbarao 2003). Even when projects are selected by communities, it is useful to ensure that these projects are well integrated with the overall rural development plans of local communities. While community selection of projects is generally the good practice, care needs to be taken that project benefits are not captured by local elites. 8. Monitoring and evaluation. By far the most critical missing link in the experience of public works programs is the presence of a credible monitoring and evaluation framework. Though over 40 countries, both developed and developing, have implemented public works projects at various stages in their development, only three or four countries have documented their performance with rigorous monitoring and impact evaluations. Two countries with extensive evaluation experience of public works programs are India and Argentina. Several ongoing programs in countries such as Ethiopia and Kenya now have incorporated an extensive monitoring and evaluation framework. 9. Important indicators for monitoring of public works programs include, on the inputs side, budget expenditures and/or available food (to be used as wage), program staff composition and number etc. and, on the outcome side, number of projects by type and labor intensity, wages paid and periodicity and regularity of payments and/or amount of food distributed as wages, number of workers participating in programs and their key characteristics (by gender, age, previous economic activity) and beneficiaries of any potential training or other incentives offered to escape from poverty. Impact evaluation would require a good baseline on a random sample of households gathering information on all household characteristics of both participating and nonparticipating households. 80

96 ANNEX 11: INTERNATIONAL GOOD PRACTICE IN DIRECT SUPPORTS PROGRAMS Some Good Practice Design Features for Social Assistance (Direct Support) Drawn from International experience This note provides a brief summary of good practices in the design of social assistance (direct support) programs, drawn from a few select countries. The note addresses four design features: (a) selection of households for direct support; (b) determination of the level of benefits; (c) the mode of payments and selection of the delivery mechanism; and (d) monitoring and feedback and mid-course correction. A. Selection of households for direct support. 2. It is common knowledge that the more generous the definition of eligibility, the larger will be the pool of applicants for social assistance and the cost of the program. The cost of the program also depends on the level and duration of benefits (discussed in the section below). For example, for the old age social assistance pension program, Nepal defined the cut-off age limit for eligibility as 75. This undoubtedly restricted the pool of beneficiaries and kept the program within the limits set by the available budget, but the program could not reach some of the critical vulnerable groups even slightly below the age 75. On the other hand, some countries set the limit for eligibility low at 60 (or even lower), leading to the opposite effect of too many beneficiaries and a very large budget. To overcome difficulties of this kind, many countries now resort to limiting the cash transfer social assistance to, say, the poorest 10 percent of the population. One such example is the Kalomo District Pilot Social Cash Transfer Program in Zambia, which limited the outreach to the bottom 10 percent of the population. 3. Determining the eligibility and selection of eligible beneficiaries has varied a great deal from one country to another, depending upon (a) administrative feasibility, and (b) the available information. Where both these sets of conditions are weak, countries (such as Rwanda) have resorted to community targeting approaches. However, good practice dictates validation of such selections via a transparent communication system. For example, in Mexico s Oportunidades program (a conditional cash transfer program), beneficiary lists are presented at community meetings which has given communities a chance to pick both exclusion and inclusion errors. In all cases, a complaints mechanism is critical for ensuring community satisfaction with the targeting approach. 4. Where the information constraint is less severe, countries have adopted a proxy means test (PMT). This is a targeting method by which a score for each applicant is generated based on household characteristics that are fairly easy-to-observe usually non-income characteristics) such as the location and quality of housing unit, ownership of durable assets, number of children, level of education, etc. A threshold score level is set below which a 53 This short note draws from M. Grosh et. al. The Design and Implementation of Effective Safety Nets: For Protection and Promotion. World Bank,

97 household becomes eligible for the benefit. When a community targeting approach such as the one described in paragraph 3 above is adopted, it is still possible to move gradually to a PMT method, short-list households for the benefit, and then use community meetings to ensure transparency and avoid exclusion and inclusion errors. Many low income countries are resorting to this combination of a PMT and community validation as a means to select beneficiaries for safety net programs generally. A recent example is Bangladesh, which is now sponsoring a safety net program using a combination of PMT and community validation. B. Determining Benefit Levels 5. Determining the size of social assistance direct support is a tricky issue in all countries. It is hard to provide clear-cut policy advice based on international experience, but some guiding principles can be offered. Typically, in last resort programs such as the one proposed for Rwanda, which aims to reduce extreme poverty, the benefit levels are set as a fraction of the income gap of target beneficiaries. How high or low that fraction should be depends on the available budget and the number of people in extreme poverty. Using a proxy means test, Armenia and Georgia have used this principle. 6. In this regard, some number crunching might be helpful with the available household level information. For example, information on the number of extremely poor households can be combined with information on their income (poverty) gap and, from both these sets of information, one can derive the financial requirements for a given level of benefit. One can then see the feasibility (affordability) of alternative benefit levels and decide on the level that can be defended within the available budget envelope. 7. Benefit levels need not be fixed at a flat level for all types of households. Instead, the levels can be varied. Variable benefit formulas are often the norm in many countries, mainly because such formulas allow for variation in household circumstances (such as number of children, presence of a disabled child or person, long term sickness of a household head, etc). A very good (successful) example of such a variable benefit formula is that of Brazil s Bolsa Familia Program. This program provides two types of benefits: a base benefit to all families in extreme poverty, and a variable benefit that depends on family composition and income. 8. Whatever method is adopted to determine the benefit level, it is useful to assess the level of benefit as a percentage of the consumption expenditure of extremely poor households. How generous the program is can be assessed from this proportion. The higher the benefit level as a percentage of the household s consumption expenditure, the more generous the program. Maintaining a generous benefit level is likely to impact on labor supply through adverse disincentive effects: i.e., households, even when provided with opportunities to work in the labor market, may opt to stay in the generous program. While this concern for the impact of benefit levels on work disincentives is theoretically valid, it does not apply to programs that target extremely poor households with no adult labor to participate in the labor market, or the disabled, or the elderly. C. Delivery Mechanisms and Payment modalities 9. Four principles generally guide the delivery mechanism: (a) ensuring reliability and regularity of payments; (b) maintaining accountability (governance issues) and prevention of 82

98 fraud; (c) reducing transaction costs to the beneficiaries; and (d) minimizing the administrative cost of delivery. While a number of delivery agencies or routes are available bank branches, mobile banks, post offices, decentralized government agencies, NGOs the selection of the delivery mode eventually must satisfy the above four principles, and be available and suitable for a given country situation. Not surprisingly, countries have varied a great deal in this regard. If contractors or a specific agency is selected, performance-based incentive contracts can be developed as was done in the Brazil s Bolsa Familia program. Kenya has tried a pilot program for a cash transfer to orphans and vulnerable children through a bidding process to select a lowest-cost service agency. The advantage of these contractbased service agencies is that the contracts can be revised based on performance. In countries with somewhat developed IT infrastructure, debit cards and smart cards are being used to transfer cash assistance. In the state of Gujarat in India, a pilot program is being tried to use smart cards to transfer in-kind social assistance. It is hard to recommend one specific option or delivery mechanism: the main challenge is to adapt any one reliable mechanism to country circumstances, and avoid any unintended adverse effects. D. Monitoring and Evaluation of Cash Transfer programs 10. Program monitoring is extremely important for any safety net program and especially for cash transfer programs. Systematic monitoring helps one to assess how well the program is being implemented at all levels, and helps mid-course correction in the event of poor implementation. Evaluation complements the monitoring system, inasmuch as it allows an assessment of the distributive effects of cash transfer programs. Despite the critical importance of monitoring and evaluation, unfortunately most safety net programs lack a credible M&E system in place. 11. Monitoring is a continuous activity, and is typically done at all levels village, district and at the national levels. Its main role is to assess whether or not the program is being implemented in accordance with its design with outcomes as expected. Its annual cost must be factored into the program costs and it must become an integral part of the programmatic framework. A good monitoring system must collect information on the program s key outcomes. A good practice procedure is that monitoring should be done by an independent agency, outside of the agency or institution implementing the program. A well-documented international experience of good monitoring is from Zambia. The Kalmo District Pilot Social Cash Transfer program (which operated with technical assistance from Germany) implemented third-party monitoring that focused on the quality of program management, the effectiveness of targeting, regularity of transfer payments, and even beneficiaries use of the transfers. 12. It is not enough to know the program s outcomes; it is important also to know the impact of the program on household welfare, which is the ultimate goal of a cash transfer program. Several techniques are available to do an impact evaluation. Two approaches can be distinguished. One is a quantitative approach that collects information on a random sample of households belonging to both the treatment group and the control group, both at the launch of the program and after a given period of time (say, one year). Econometric techniques are 83

99 then used to assess the impact of the program. 54 A complementary approach is qualitative evaluation, which is based on focus group interviews, key informant interviews, and direct observation. Though qualitative evaluations are not representative, they do offer rich information on the program s functioning, its merits, and shortcomings. 13. A variant of descriptive evaluation is process evaluation, which is probably the most common evaluation technique followed in many countries. Its approach is to assess and document how each of the processes underlying a cash transfer program is being implemented. It helps address the question: what is happening through the program. Process evaluation strongly complements, but does not substitute for, an internal monitoring system and other evaluations mentioned above. For example, the Zambia example is worth repeating: it included a process evaluation by external evaluators in addition to other evaluations which helped improve the monitoring capacity as it revealed specific flaws in specific processes underlying the program s implementation. 14. The short review above is not meant to be a substitute for the vast literature that exists on cross-country experience of cash transfer programs. Readers requiring further information are advised to refer to the World Bank s most recent publication, viz., M. Grosh et. al. For Protection and Promotion: The Design and Implementation of Safety Net Programs (World Bank, 2008). 54 For a good understanding of the techniques and applications, see World Bank website on Impact Evaluation. 84

100 ANNEX 12: SUMMARY OF VUP PUBLIC AWARENESS PROGRAM Summary of Ministry of Local Government s Efforts to Ensure Transparency of the VUP Policy and Procedures, Eligibility Determination Process and Criteria for Ensuring Transparency of the Selection Outcomes of Eligible and Non-Eligible 1. The Ministry of Local Government is continuing to pursue international best practice of both its Public Works and Direct Supports program under the VUP to ensure total transparency of all program policy and procedures. 2. The main outlet for ensuring that the VUP eligibility determination process is both clear and transparent to all potential beneficiaries and non-beneficiaries by making program policy and procedures widely known and to ensure that selection outcomes are publicly available is the use of traditional community meetings. 3. The targeting process uses the home-grown and traditional Ubudehe process (see Annex 6). This helps ensure community input into beneficiary selection as well as transparency. The process is conducted at Umudugudu (village) level. It is a two-step operation which involves taking information from the existing Ubudehe social map followed by a community meeting to which the Umudugudu population is invited. At the meeting the social map information is validated, each household s Ubudehe category is updated, and information required to apply the VUP eligibility criteria is added (the process is detailed in the VUP Targeting List Training Manual, May 2006 ). At the community meetings the facilitators present the meeting objectives, the VUP program, and explain each step in the process as they go along. 4. The targeting data is used to determine eligibility lists. Once Direct Support lists are approved by the Sector Joint Action Development Forum (JADF) they are posted publically at Sector offices. A meeting is then held with eligible households to explain the policy, procedures and next steps (why they are eligible, what they will be paid, when, the need to opening bank accounts etc.). For Public Works, once recipients are selected from the eligibility list to work on a specific project, before work starts, a meeting is held with them to explain the policy, procedures and next steps (how they were selected, how work will be organized (the need to make groups and nominate a supervisor), how long they will work, wage rate etc.). 5. In future, there is a plan for eligibility lists to be publically posted at Cell level and to be read by Imidugudu Coordinators at community work days (Umuganda). 6. Local government officials are also responsible for sharing and explaining the VUP program broadly with all the communities under their mandate. This includes non-recipient households in VUP operational sectors. Furthermore, the VUP team, as part of its monitoring activities, takes the opportunity to ask questions and share information with direct support and public works recipients and when visiting work-sites during field visits. 7. The Ministry of Local Government is also pursuing other ways of sharing the VUP program s objectives, policy and procedures and the need for transparency in sharing 85

101 selection outcomes. The list below provides some examples currently being pursued by the Ministry for this purpose: Live radio and television interactive shows where representatives of local government, central government and beneficiaries explain the VUP program and answer questions. Live, interactive events took place on national television and on radio in April 2009 and on radio in June Radio and television programs offer toll-free numbers for listeners to call in. Workshops with mayors, vice-mayors, and District and Sector executive secretaries are being organized to present the VUP program and provide guidance on how to communicate the VUP program effectively and accurately to community members. Change agents are being trained through an organization called Itorero to go out in communities and provide information of the VUP program. These change agents are covering the entire country, not only the VUP sectors. Leaflets have been prepared in Kinyarwanda, the local language, and 16,000 copies are being printed and will be distributed to every village (approximately 15,000 across the country). VUP program information posted on the MINALOC website. Articles are published in national newspapers on the VUP program and progress. An electronic newsletter was distributed in November 2009 and it is planned to be produced every 2 months. A documentary film is in the process of being made to further share the message of the VUP program, and will be shown on the national TV channel (Rwanda Television). Districts are encouraged to prepare a proposal and receive financing from the Ministry to help promote the VUP program on local radio and television. As of October 2009, the Districts of Gasabo and Huye (Maraba Sector) have submitted proposals for this purpose. 8. Additional plans for the coming year to further share the program s policy and procedures and enhance transparency are: Development of a communication strategy. Terms of Reference have already been drafted to hire a consultant to assist with this. User-friendly poster on targeting and eligibility to the different program sub-components to be posted at community level. A more technical booklet that further explains the VUP program and selection process, in the local level, to be disseminated to District and Sector levels Sensitisation and training activities to include opinion leaders within sectors (such as teachers, change agents, health post managers, Cell coordinators, Cell executive coordinators, local bank/mfi staff, members of sector CDCs, local cooperative representatives ) for even broader dissemination of messages 86

102 Live interactive radio shows on regional community radio (Rubavu, Ngoma, Huye and Rusizi) 9. In addition to promoting transparency of the VUP policy, procedures and eligibility determination process and criteria for ensuring transparency of the selection outcomes of eligible and non-eligible beneficiaries, the Ministry of Local Government is using the above media outlets to ensure that the policy of encouraging voluntary savings is well understood and being implemented across sectors. 10. Also, as part of the promotion of the VUP program, the Ministry is working to ensure that the policy under VUP of ensuring community input into the selection of VUP projects under its Public Works program is maintained. The main process for this is through community participation in the development of Sector Development Plans, and this is being monitored by local governments. VUP will continue to raise awareness on the importance of community participation in sector plans (and improvements to existing plans, and development of new plans) through forum like the JADFs. 87

103 ANNEX 13: EXTRACT FROM VUP MONITORING & EVALUATION MANUAL Complete VUP M&E Framework Manual (June 2009) including Annexes, is available upon request. 88

104 2.4 Beyond tile income poverty rabonal, the rno<e global ooootation 01 tile EDPRS and its flagship VU P progr:lmme are clearly supported by a human poverty concern ecoodm>c gro'niil and income must materialize into meeting basic human needs in different domains: health, education, wate<i5anitation, nutrition and food security, housing cond itions, social integr:ltion (participation), etc.? This concern Os taken into account in two monitoring tools: again the Community Based Human Poverty Household Ques ~onmjire, section 1 (demography), sections 2 and 4 (education), section 3 (health), sectioo 5, (labou r force), section 6 (income) section 7 (rnltrition), section 8 (main expend~ures & investments), sections 9, to and tl (agricultural production), section t2 (habitat and accommodation ), section 13 (durables and productive assets ), section 14 (access 10 infr:lstructure and markets), section 15 (participation); second, the Cell DevekJpment Infra structure Ouestionnaire. Indicators gene.-ated by these tools are ' sted in Part II, section 4, Indicators Methodologies. Their precise location and role in the M&E architecture are given in section 2.t, Dynamic Archilecture. The Ce" Development In frastruc ture Ques60nnaire is tound in Anne.. E. Among its specific teatu res, it can already be observed that ~ incllkles a roste< of the lmidugudu incllkled in the Cell, analogous 10 the household members roster usually tound in household level questionnaires. It allows to locate the different infrastructures at the umudugudu level and to measure accessibi lity at this lev," also. 2.5 The different monilofing tools introduced aoove, all interlinked to measure multidimensional poverty at different levels, are the key elements suworling the main components of the M&E a rchitecture 2.1 Dyn3mic 3rchitecture: Mo nitoring Cycle 2.6 The word accelerate used in the foflllulabon of the VUP purpose requires that tile dimensions of lime. and movement be emphasized across the M&E design. The dymjrr.-c side of too VUP thus appears as a centra l architectural component To mon~or thos acceleration means mon iloring closely tile path of extreme poverty reduction, in the context ot a two-level targebng.ltategy aiming precisely 10 K1en~fy and to tackle this poverty wherever ~ exists more intensively: a first tarye~ng at the sector level, expressed in a poverty ordering of too sectors with in each district, determming the priority a ccess to VUP services. and tile annual budget allocation bety,'oon VU P sectors; a second targeting at the OOusehold level, specific 10 each component and relying on the UBUDEHE social mob4lizalion system 2.7 The structure of this dynamic mon~oring refers primarily 10 the scaling up strategy. The assumption oore is that every year, starting in the financial year 2OIJ , in each ot the 30 districts, a new group of sectors joins the VUP program. Such a group of seclors is named here a Cohort. The thirty pilot sectors foflll Coho<t If f, the new coming sectors Cohort /f2, and so on. The size of these cohorts is determined by the level of financial resources available. To each cohort Iii of sectors eo<responds the sea li ng up date T', defined as the 1" day of the month woore VUP operations and spending begins. Too dynamic structure of the M&E is tilen expressed tilroug h the Monitoring Cycle summarized in Figu re t below. Some boxes in Figure 1 are ma rked with 3 X: it means that they are specifically involyed and critical in the impact e valuation scheme. 89 "

105 Figure 1 VU P sector Monilofing Cycle -_. --~ " 90

106 2.8 Figure 1 shows the.,cation of the different mon~oring toos in a ti me frame refening to five sectors cohorts defined by the annu3j scaling up strategy. Comments are needed to cl;iiify the rofes of the different statistical operations associated to the moo~oring toos. First, from bot!om to top, there are two parts separated by the time axis. Too upper part, ines 1 to 5, refers to the household level mon ~ oring toofs: the Targeting imland the Community Based HUrmJ" Poverty Household Oue3tioonaire The "'... er part, lines , refer to the cell and umud"9udu level monitoring toof the Cell Devek>pmenl Infrastruclure Questionnaire. Dates"" the time axis are defined by a symtx> T in reference to months, too superscript idootitying the cohorl of sectors, the subscript,eferring to the month before (minus symtx» Of afte, (plus symbof) relative to the starting date specific to the cohort. If we want to refer to the calendar, a realistic assumption is that T 2 0 corresponds to July The moniloring cycle is n<ni described by refening to whal should be the oormaf process for cohorl N2, the one comin9 in the f,rst scaling up operation foflowing the 30 p<lot sectors identified with cohorl Nl. It appears in Figure I, date T J _ 1 = T'.,,, that before scali r.g up with a new cohort of sectors, a one-yea, (13 months) preparation period is needed. First, accofding to column A, li nes -1 to -5, a coosus of all ce lls in the country, using the UBUDEHE system and the Ceil Development Infrastruduro Questionnaire, allows to rank all sectors, within each district, according to the popu lation size (/I hlds) and socioecooomic infrastructure ava ilabfe (see more in.j.13 et seq. Pari II). Then a scaling up prog ram can be defined according to the targetir.g cri!erilm of the poorest sectors comir.g in first In addition, every year, th is eel questionnaire can pro,,;oo an important input for the alocation of VUP budget according to seclor poverty level. This ce l census is completed, in each sector of the new cohorl cell "'2, with a Needs Asses:smenf Household Survey, using the Community Based Human Poverty Household Que3fionnaire. This survey is done on a sample of households representative of the whofe popolation in the secto<, 001 only some Ubudehe bottom categories. Sampling structure is detailed in section 4, as. well as the rational beh ind. This Needs AsseMmenfsurvey pro,,;oo s information for a more equitat>le annual budget al.,cation between VUP sectors. It al.,,,,,,, to bridge the two poverty concepts (principle 117, Bridge befween subjedi,.., and objective poverty measuremenf.). Oepending 00 sampfe sire, this survey can p<'ovide a SeclOf poverty profile which, combined w~h the ce ll developmoot infrastructure coosus, should inform the sector developmenf plan, a key component o f the decootralization p<'ocess (see priflaple 111 ). Finaly, this Needs Asse3Smenl Survey constitutes one of the pilars for sourod impact e valuation, by providing a cootrof group at the same time, T' _ " the previous cohorl is conducting a baseline survey just before VUP operations. start (cell AI ). As indicated in Figure I, the baseline survey conducted in previous. cohort sectors uses tile same Community Based Human Poverty Household Questionnaire. Th is ooe--year p<'eparation phase is used for VUP staff recruitment, pruject idootification for pobfic works. training, and budget arod operations planning 2.10 Two months before implemootation, T'.2, too targeting operation is runned, using the UBUDEHE p<'ocess arod the Targeting List toof (cell C2). This operation covers all households in each umljdugljdu, and thus provides an update of the UBUDEHE social map and household distribution across too six ubudehe categories list of eligible benef,ciaries for OS and PW are tmn ava ilab" 00\ late< than one mooth before im ple mentation 91

107 2.11 During!he month before implementation, T'." as seen in cell 02, a Baseline Household Survey is run in each secto.- of the cohort The questionnaire used is!he same as lor!he Needs As:;es:;ment Survey run twelve rnomns before,!he Community Based Human Poverty Household Questionnaire. The household sample is also!he same, ",oooing a pan," of OOuseholds, essootialfy in 'fiew of impact evaluation. It should be... bced that at the same t me, T'.", ce ll 01, a Short Term HouooOO/d Impact Survey, using the same questionnaire, is run!he previous cohort, and at T'_13, cell 03, a Needs Assessment Survey is run in the upcoming cohort 00><1 year 2.12 The targeting list is updated in each UmudlJgoou of the cohort at ti me T'.., cell E2, as stated in!he VUP poicy that lists of beneficiaries are to be revised every six moo The annual operational cycle then ends f,rst with!he Short Term Howsehold Impact Survey at time T'.", cell G2, with!he household pane l sample. The short term impact ana lysis is made possible with the standard BeforeJAller Unequivafent Control Group seheme because, at the same time T'." cell G3, a Baseline Houoohold Survey is run in cohort. 3. Second,!he annual end of!he cycle ineludes, at the same time T'.", cell G-2, with the update of!he cell development infrastructure, ineo<por:lted \0 the UBUDEHE "",ces:; as an annual update of each UmudlJgoou social map. Thus this update has a national """'...-age, in ",ina pie The annuaj cycle of monitoring data coflection operations, lor an y VUP sec\of, is schematized in Figure 3 befow. Figure 2 highlights the short te<m impact evaluation seheme as we i as upcoming statistcal surveys required for sealing up Figure 2 Short Term Impact Evaluation Scheme and upcoming surveys 92

108 COHORT 111 COHORT I' COHORT

109 Figure 3 Sector Mon itoring Cycle '.... Hid '. II... T...,.n '.... U. t,. ' U. t 2.2 Static architecture: RBM and LogFrame 2.15 The sta~c a rch itectu re refe", \0 the intervention strategy shaping the VUP. ThOs strategy. also frequently ca lled the logic. of intervention. is drive n by too Re~ult Based Management (RBM) a pproach. relying on tile Logical Frameworl< (logframe ) as its main technical Standa rd definitions relative \0 these well-known methodologies are provided ill Annex A. The VUP M& E contributes to structure the VUP program according 10 the formal norms regulating lilese appmadles. IttMn """rationalizes lilis c.onceptuaj a nd formal structure as a system of concrete measureme nt operations a rticulated in a coherent Managemenllnformalion Sy3tem (MIS ). This 94 "

110 MIS is lile object 01 Part II of lilis Manual. This section locuses only Of! defining!he more cooceprual part 01 too static arcmecru re: objectives, main expec1ed resuks at different objective leve ls, and types of indicators used to measure the progress 3Qainstlllese expected re suks. Too technical definitions of tmse indicators are given in section 4, Indicalors Melhodo/ogie~, as well as the ir precise location within!he moo~oring toos introduced previously or beklw PU!ROSe and expected effects The pu'pooe of VUP, already referred to in section 2.1, is: To accelerate the reduction of em-erne poverty This e"l'ression re ' es {)f1 the rational justifying VUP as expressed in two official policy documents: Economic Development and Poverty Reduction strategy, ()12 {2007}, and Vision 202() Umurenge, EDPRS FlafF'hip Program Document {August 2007}. Iv1 outcome is expected from this purpose: Ovtl;ome. 1: Extreme ly poor OOusehokl are K1entified by VUP. taryeted and ra pidly taken oul of this poverty status. Three ind icators are used to measure the degree of achievement of th is outcome n",y are: : Share of population living in e xtreme poverty in VUP sec10rs : Poverty depth ifl VUP sectors. 01 OOuseholds in too bottom tno UBUD EHE The flrsl tno indicators are primarily defined relatively to ';"come poverty, in relerence to iflcome poverty lines give n in 2.2. The first ind icator is the standard extreme pove rty incide nce indicator,!he percentage of population in households below the extreme poverty line. The second ind icator measures, ifl terms of percentage,!he average gap betneen the pove rty line and the income of OOuseOOlds below!he pove rty line. Through a compw~e indica/of of human poverty, a nalogous ind icators will be computed for huma n pove rty. The lilird ind icator refers 10 the social assessment 01 the relative poverty status with in lile umudugudu. Since these tno bottom categories are at the core of the targeting process, the percentage is expected to decrease significantly To reduce income poverty and inequal~y..,.~,.",."."!."~, ~:",.':.~through which VU P is.";'.;;., _.";;;;... ~ The contri>utes. goal is lilen 11 is Without any more specification 01 impacts exp<essing this goal in terms of expected resuks, IIlree indicators measure!he global impac1 associated to lilis goal. They are: Silare 01 population below poverty line. Gin i coefficienl H 95

111 Indicator. Gl Share of population hving in extreme ""verty All three indicators are moneymelric ind icators_ All three require a methodology for measuring houseoold income_ Two of them, Gland G3, use the income ""verty li oos given in 2_2_ The second one, G2, is the most we l~k oown iooqua.ty index oomputed from the lorenz CU<Ve li nking income shares to the pop<llation distribution across income leve l. This iooqua. ty index thus requ ires that households samples be representative of the whole population, not only of VU P targeted categories for OS and PW_ The measurement of these ind icators, at the national level, is not the VUP responsibility, OOtthe NISR responsibi lity_ Nevertheless, our PI indicalof above is in lact the same than G3, in VUP sectors on~. Th"", a dialog"" is oocessary on income measurement methodology bet",een NISR and VUP M& E unit 2_18 Qu\oots The VUP programme is structured in fll'e components. 1 Publi<: Works. 2 Direct Suppor1. 3 Financial Services JI4 Tra ining. 5 PM U (Program Management Unit) AccOfd i ng~, programme outputs are first grouped accofd ing to this five component structure_ But all outputs are not oocessarily specific to one of these five components Some oot;>uts, Cfoss-cu1ting all components, are resuks expected from the s)'ll"'lly bet",een activities run in al components_ It should be remembered that VUP, articulated within the de centralization process, intends precisely to coord inate all developmen t ope.-ations happening in a given sector so that the overall efficiency is en hanced_ 2_19 Two outputs are directly inked to this program Extremely poor households have access to granted publ i<: worts OvIput n 2: Re levant and operational community assets, are produced according to intemationaj good practi<:es The fi rst output is expected from an efficient targeting system Fou r ind icators are identified in view of measuring this efficie""y_ households granted pubfic % of UOOd ehe elig ible households granted public worts ",';;;;;;;:ijja Distribution of PW beneficiaries (persons) according to sex The forst indicator refers to an objective concept of ""verty, usually the moneymetr.c definition of poverty, with relerence to the income e><1reme poverty line_ It is the most diff,cuk to measure, OOt alows to bridge between t>oth ""verty concepts (pri""iple " 96

112 117). It requires!he Human Poverty Howsehold QUfMtionnaire and ParmI Lists. More precision is given in Chapter 4. Indicators 112 and 113 refer to the Ubudehe poverty concept and dire ct~ to the targeting methodology Indicators 112 to 114 are extracted from the PW Payroll List, gene.-ated from the Targeli.-.g ill/. This PW Payroillist is one 01 the primary componoots of the VUP Financial Information System These first IIlree indicators are disaggregated according to lile sex of the household head. This gender analysis is com;>leted by lile fou rth indicator which refers to the sex of the workers perfofming public woos The second output catches the relevance and good management 01 lile projects identified Iocalty for pub lic woos. It expresses too expectation 01 getting high quality assets, especially infrastructures, from these projects. Si>: indicators describe the ach ievement 01 this output Number 01 operational com mun ity assets by cate!jo'y f roads, schools, health posts, etc), In each of population! number 01 elioots served by these : % 01 Public Woos projects paying wage rates oot exceeding lor unskilled labo.- PW beneficiaries paid directly in!heir bank acrount from!;~~~~.~:'eek afler a two week work Indicator % 01 PW projects ~,,,',';., specific to their sector 01 intervention in their bank account, nol environmental protection norms These six ind icators will be captured mainly by tno important subsystems 01 the whole M&E system, the VUP PW Projects Database and the VUP Financial Information System. The first su bsystem is develoj>ed in fun consistency with the CD F Project Database currently being implemooted. Too VUP PW Projects Database is described in Part II, secbon 4. Too VUP Financial Information System is described in!he Vision 2O2O Umurenge Financial Management Fmmewori< and Opemb-',nal Manual. Too Cell Deve/opmentlnfraslructure Questionnaire is also required '>:<:1 Suooort QyIDuls. Two oul;>uts are directly linked to this program componenl OvtpvJ 111 /1: Extre mely p<>of households are assisted through direct support paclages Ovtput ' Direct cash transfers implemented according to international good practices The first OS output is expected from an efficient targeting system. Three ;oofcalors are identified in view of measuring this efficiency " 97

113 ~;~~~~f.,:"~~"m~~"j~':o:, :":""':,;"",,::eligible OOuseholds assisted through ~ households assisted through direct As for PW, the first indicator refers to an objective concept of poverty, usualty the moneymelric definition of poverty, w ith reference to the ir>come extreme poverty li oo 11 requires the Human Poverty Household Questionnaire and Payroll /..N;I~_ More precision is give n in Chapter 4. Other tno indicators refer \0 the Ubudehe poverty COr>Cept and direclly to the targeting methodology. These tno indicators are obtained through the Targeting /..N;/ and the DS Payrol/usl These first three indicators are disaggregated according to the sex of the household head_ The secood OS output expresses the preoccupabon of taking aceoon! of "'SSOOS and norms arising from otilef COUIl!ries experience in such a difficult and rontrovers;al intervention area_ Three indicators describe the achievement of this olllpot Indicator % of VUP ImKlugudu where the form TargeHng List U$ng Ubudehe has been publicly f,11ed Ind icator : % of VUP ImiduglJdu where the list of OS beneficiaries has been ""t>lidy pub~shed, allowing fof a 2-y,eek delay lor individual revision requests Ind icator' % of as beneficiaries timely paid in their bank accollllt, the lit day of the reierooce monlll 2.21 The Financial Services component is stil under ;;: ~~:~~:~~;~~~~~,;';"~~.:~... ~n terms fu lly designed of outputs for a nd implementation, ind icators. At thts ~ is III : OS and PW beooficiary households access cred~ and sa"'"-gs savings laciities im~mented according to The Output e xpresses an important VUP policy: the Financial Services compone nt is a lso designed as a targeted program, reserved to households lilat already benefited from any of the f..-st tno components, PW or as. In the devek>ping countries, there is now a long trad~ion in microcredit Output '111.2 requires that this experience be taken into account Three ind icators can be associated 10 Output NIII.1 n'"'""",,' OS and PW beneficiary households ge tting a 98

114 Indicator A' Tolal amollllt of voluntary savings adlieved by OS aoo PW beneficiary OOusehokls Ind icator A' III 1 3 Repayment rate achieved by microcredit beooficiaries. 222 Tfilioioo Qutoots Thi,; component of VUP i,; currentty being designed. ~ will be integrated in the LogFrame 0000 a first draft has been circu lated aoo discussed 2.23 PM U (Proofil m Management Unit) OvIoots "'Ltc.",", VUP ma nagement at MINALOC aoo UMURENGE Five iooicators measure the achievement of this output 1 ~~~~~!~;2;'i~~;'~.;';'i",i";'i'~'~':i":;';;'ioom~~ance acrordar>ee with the financial with the prescribed in the M&E manual Develo!>ment Fora (JAOF) in pla ce at Imire nge 2.24 Cross cullioo Outputs. Some ve ry important expected results, at the output level, canoot be imputed to a specific component of VUP. They should be adlieved through the syne<gy expected ffom a harmooious integra tion of a ll components constantty aiming at the same aoo unique purpose, 10 accelerate /he reduction of extreme poverty. That's why they are qua lified as "<:ross culling" ouit>uts. Two of these are iden~fied at th is bme 225 A f,rst cross cutting ootput is: OvIput _VI 1 Benef'cia... s from Publi<: WOIks aoo ~irect support graduate. The probo.m with this output is we l~ koowfl of graduation? which ooofi)tiooal definition can be give n 226 To define graduation, tho approaches a re proposed, relative in fact 10 the two concepts of extreme pove rty ( 1.6, principle 117, a oo 2 16, purpose iooicatorn). 227 From the Ubudehe relative poverty con<:ept, as <>perationalized in the VUP largeting process, a beneficiary hoosehold can be cooside<ed of "having graduated ffom OS" d, afler some time. ~ is no more eligible to OS. In a similar way, a beooficiary OOusehoid can be considered of "having graduated from PW" rt, afte r some time, it is no more eligible to PW and is 001 elig ible to OS. Thi,; approach is ve ry ";mple 10 im plement since ~ relies on the very clealty defined elig i b~ity criteria 1m eadl of the two components. AccOfd ing 10 the VUP Sector Monitorir.g Cycle (F igures 1 aoo 2, Dynamic Ardl~e cture), the eligib< lily list is revised e ve ry six months, aoo " 99

115 lilus the regular remion pmcess would give an immediate measurement of graduation. But ~ must be ob-noos lilat this approach w~1 generally mean that!he umoougoou population COflfinns that the household has moved from catego<y 1 Of 2 at least 10 category 3 (u mukene ) A second approach has been recentty proposed, in line with the ab""uie poverty concept, income based and objective. Thts proposibon ts reproduced here in Annex G. Since monetary income and cash transfer is re a l ~ the core of the VUP as a Social Protectioo program, it is suggested, in view of developing an objective criterium fo.- graduation, to focus on income poverty and Of! potential cash Of in kind Income. Thus, ~ is P"'l"'sed to define graduation as f<>~: A household receiving 01) a monthly basis an amoulll of X FRW, froli) PWor DS, is considered as having graduated if it effectively generates, in a sustainable way, an equivalent monthly net income of X FRW, additiolwl to its income level previous to its participalioll to PW or to DS Ob-noosiy, this approach requ ires 10 mon~of closely the new income generating activities and the net income that they generate. Thts type of mon~oring ts precisely lile core of the M&E as presented above in the d)'flamk: arcmecture Whatever be the definition of "g raduation", five ind icalors are proposed to captu re the level of graduation achieved IIlrough the programme. 01 PW beooficiaries of PW beneficiaries These ind icators are disaggregated accofding to gender A second cross cutting output is: Output VI 2 Decentralised Local Governments are strengthened in planning local develo!>ment Accord ing 10 principle ' l (Decentralization. a community b,.,;ed M&E), the VUP M&E must contooute to enhance lile capacity 10 better manage the local developmenl Data collection operations are cooceived in such a way that there ts an important involltement of lile commun~ies and kxal authorities at all levels, particularly through lile Ubudeoo process. One indicatof is specified fof this output, otoors will eventualy be added 100

116 Indicator NI2 1 Number of VUP sectors produ<:ing regularly a sector poverty profile based on objectivety verifiat>le indicators measured through an impmved Utmdehe system as part 01 a locally manageable information system There is oore a focus on the sector poverty profile, which in fact play different rofes in the decentralization process: - it helps to elaborate an evidence based Sec/Of development plan, - ~ provides an objective information basis for an equitat>le allocation of public funds across sectors, - as an important social asset, the iillormation generated at the sectof level enhance the advocacy pov,er 01 the.. cal communily_ 2_31 The VUP M&E stali<: arch itecture, presented according to the logical framework technique, is summarized in Anoox B, Logframe Table. This logframe Table stops at the output level, since the input line deta ils (activities and fi nancial structure ) a re to be found in the other VUP operational manuals_ It must be emphasized that in the context of the M&E manual, locussing on a conceptual framewori<, monilo.-ing tools and methodologies, ~ is inappropriate and deli berately avoided to introduce in the logframe numerical baseline values and targets_ These va lues, regularly revised according to the availabfe financial resources (budget) and to ind icators measurement, distract from the substance of the M&E framework. Consequentfy, the logframe Table concentrates on Objectives, Indicators, Means of Verification. These Means of Verifi<:ation are themselves the ma in subject of Part If of this ma nual, Mon~oring Information System 2_33 It can be seen, from Annex B, that at this stage of development of VUP, the logframe includes 1 goal, 1 purj>ose, 9 outputs and 38 indicators. The development of the financial Services and Training components will generate new outputs and indicators 101

117 ANNEX 14: UMURENGE SACCO STRATEGY EXECUTIVE SUMMARY Executive Summary 1. Credit Unions and SACCOs are nothing new to Rwanda where diverse small SACCOs and Banque Populaire du Rwanda (BPR), the most important credit union, have operated for many years. However despite a long history behind, SACCOs coverage remain very limited, with only 3 percent of the population saving with all MFIs in general, SACCOs included. 2. To address this issue analyzed as a major factor hindering economic development, a recent National Dialogue Meeting held in December 2008 recommended the creation of at least one SACCO at the level of each Administrative Sector (Umurenge). 3. In line with this a task force composed of representatives from diverse Government institutions, especially from MINECOFIN, MINICOM, MINALOC, and BNR was formed and was assigned to propose a strategy to implement this policy directive. The strategy proposed in this document is in line with the National Micro Finance Policy, the National Microfinance Policy Implementation Strategy, and the recently validated National Savings Mobilization Strategy. Towards this aim, the structure of this document is designed as follows: the first step is to explain what is a SACCO, what are its major characteristics and roles, related advantages and specific challenges. 4. Then after having analyzed the current level of development of SACCOs which cover 18 percent of adult Rwandan population, the document highlights the specific major challenges hampering their development which are a little education on them; poor governance and little human resource capacity, little information on bad clients of; inappropriate legal framework and taxes; lack of adequate infrastructure (energy, roads, offices, equipments ; real return on deposits little or negative. 5. To address the issues analyzed, the strategy recommends to develop a successful SACCO Model in line with national policies and international best practices; the strategy proposes a number of actions to be implemented in three phases: phase one is related to preliminary actions which will pave the way for next phases; phase two is related to creation of SACCOs and making them operational, and phase three is the development phase once SACCOs will be created and effectively operational. Detailed action plan for all those steps, as well as related time frame, budget and responsibilities, are proposed. The major item in the proposed budget is related to a subsidy to cover in a digressive way the operating costs of starting SACCOs until they reach self-sustainability. This would cost around four billion RwF for the first three years of the plan. 6. To be effective and successful, this strategy requires a conducive macroeconomic environment. The strategy also needs total involvement of all players: Central Government, Local Governments, Central Bank, community based organizations, and the concerned population. This general mobilization for a radical revolution requires total political involvement at all levels. 102

118 ANNEX 15: RISKS OF USING SACCOS AS A CASH PAYMENT MECHANISM Background 1. The GoR, in an attempt to reduce fraud and corruption, has mandated that all payments of public funds to private-sector entities (including individuals) should be made directly into accounts these entities have established in financial institutions and that no government payments to be made in cash. Payments under the direct support (DS) and public works (PW) programs fall into this category and therefore have to be delivered to a secure personal bank account and cannot be delivered in cash. As far as the VUP is concerned, this is beneficial for it provides an opportunity for the poor to access Financial Services (FS). However, this presents the VUP administration with two challenges. First, most recipients of DS and PW have never had a personal bank account. Second, because most DS and PW recipients live in the more remote villages, there are no convenient bank offices where they could access their accounts and withdraw cash. 2. SACCO s grew up as cooperatives formed by small numbers of people at village level who act together to find a secure mechanism to pool their savings and earn interest, or to provide small credits. As such, they appear to offer an ideal mechanism to channel DS and PW payments to the local recipients. 3. The GoR has encouraged financial institutions (such as Bank Populaire and microfinancing institutions (MFIs) including the private SACCO s) to take on the role of providing the mechanism to pay benefits by allowing DS and PW recipients to open accounts and save or withdraw their cash at the local village level. Risks 4. SACCO s as currently structured, and in the future as structured under the new SACCO strategy (see Annex 14), do offer a good option to deliver DS and PW benefits through a secure banking personal account system at local level. 5. However the use of SACCO s for this function is not without risk. For the DS and PW recipients, there are two risks. First, as for all members of SACCO s, there are the general risks of a small scale financial cooperative. Second, there are additional risks specific to DS and PW recipients because of their status within the SACCO. General risks 6. The general risks, common to DS and PW recipients and non-recipients alike, are associated with cooperatives of this type and are spelt out in the new SACCO strategy. Generally weak governance inadequate regulation limited range of products, and above all the fact that the SACCO is likely to be set up and run by a combination of the local elite (the non-poor) and people who are likely to be net borrowers who have specific personal interests in how the SACCO is run. These groups have a great incentive to control the operation and rules of the SACCO and their priorities would not necessarily be pro-poor. The strategy identifies these risks and attempts to address them with a combination of better regulation and governance. However, the great advantage of SACCOs is that they are set up and managed 103

119 by a group of local people who (a) have a common bond, possibly as a group of workers from the same industry or just a geographical bond from a small village, and (b) they are monitored and governed by the members who have a direct interest in how the rules and management affect their personal financial interest in the SACCO. 7. Increased regulation and a move to greater external governance, would reduce the risk of SACCOs being run by a small group of self-interested people, but could undermine one of the main advantages of a SACCO, which is self-governance by the people with a vested interest. Specific Risks to DS and PW Recipients 8. The structure of current private SACCOs allows a DS or PW recipient, as a matter of choice, to become a full member. However the cost of becoming a member is too high for most DS and PW recipients to pay and therefore SACCOs are dealing with these people by treating them as clients rather than members. Although some private SACCOs reportedly achieve considerably higher membership rates, particularly among PW beneficiaries, in some of the existing SACCO MFIs operating in the VUP Sectors we have observed, only a negligible proportion of DS and PW recipients have become members. As clients, they avoid the start up cost of purchasing a share in the SACCO. However, PW beneficiaries still have to pay the setup cost of opening an account, and both PW and DS beneficiaries have to pay monthly charges for operating the SACCO. Also, as clients, who are non-members as opposed to members, they have no voting rights and therefore no control over how the MFI (private SACCO) is run. In consequence, the rules surrounding the payment of interest are set by members in a way that makes it almost impossible for a client ever to be able to save in an interest bearing account. In practice, therefore, DS and PW recipients who are dealt with as clients rather than members of SACCOs are likely to face risks related to the following: No voting rights and no influence on how the SACCO is run. Pay setup costs that can be of the order of 4-5 days benefit for a household. Pay monthly service charges that that can be in the order of 6% of benefit costs. Receive no interest on any of their savings, unless they can lock in a significant amount of savings for a fixed period. 9. To address these risks, and to deal with the problems specific to the DS and PW recipients, the implementation of the Umurenge SACCO strategy needs to be accelerated so as to provide a mechanism whereby these individuals can become full members, have some control over how individual SACCO s are run, and receive the full value of their benefits without monthly administration charges. 10. Umurenge SACCOs seek to address a number of the drawbacks and risks associated with existing private SACCOs and other MFIs through (i) greater membership of PW and DS beneficiaries; (ii) sensitization and training on the management of Umurenge SACCOs; and (iii) members roles and responsibilities in decisions that affect their transactions, with the result that they seek to provide greater rights and benefits. 104

120 ANNEX 16: COMMUNITY PERFORMANCE-BASED FINANCING OF HEALTH The impetus for performance-based financing (PBF) in Rwanda came from the experience of NGOs who paid health workers salary bonuses, yet the outputs of the health services were stagnating and even deteriorating. The NGOs efforts at improving service delivery yielded various models, and from these the National PBF was developed.57 PBF became a major pillar within the Ministry s first Health Sector Strategic Plan and it was implemented in order to provide incentives to health workers to improve efficiency and outcomes, and provide additional income to health workers. Nationwide roll-out of the National PBF took place in April The National PBF is divided into two main models: (i) the PBF model for health centers and hospitals, called the Facility-based PBF (or the Facility PBF), and (ii) the PBF model for community health, called the Community-based PBF (or the Community PBF). Facility PBF 2. In the Facility PBF, payments for performance are based on the quantity of outputs (through case-based remuneration) conditional on the quality of services delivered. The quantity of outputs are measured monthly while the quality is measured quarterly through the use of an elaborate supervisory checklist. Health facility staff can increase their performance, and hence their earnings, by increasing the quantity of outputs, increasing the quality of services delivered, or both. The health center PBF model is based on contracts between different levels: (i) contract between the CAAC and the Steering Committee (Comité de Pilotage); (ii) contract between the Steering Committee and health facilities; and (iii) contract between health facilities and individual health workers. In-depth verification activities are done by one focal point per administrative District for quantitative evaluation and by hospital for qualitative evaluation. 58 Community PBF 3. The Community PBF was embedded in the National Community Health Policy, and through this policy the Ministry of Health provides the policy direction, while implementation is by the Ministry of Local Administration. Initial implementation was in January 2006 in 23 experimental Districts following a broad consultative process and based on 6 communitylevel PBF indicators: (i) mutuelle enrollment; (ii) deliveries at health center; (iii) use of insecticide treated nets; (iv) management of dehydration due to diarrhea among under-fives; (v) personal hygiene; and (vi) reporting. The model was based on experience gained during the implementation of the health center and hospital PBF models, and benefits from a close fit with these models. However, although the Ministry of Local Administration transferred a 56 This Annex draws extensively from the Rwanda Health Country Status Report, Examples of the NGO models are: Ville de Kigali model; the Cyangugu model and the Butare model 58 The hospital PBF model for use in district hospitals consists of 52 indicators, 27 of which are singular output measures or binary variables and 25 are composite indicators with 99 composite variables in all. Fifteen indicators assess administration and represent 20 percent of the total performance score; 19 indicators assess quality assurance and represent 30 percent of the total performance score, and 18 indicators assess clinical activities and represent 50 percent of the total performance score. 105

121 budget related to community PBF, there have been delays in implementation. This is due to the fact that, unlike the health center and hospital models, the community PBF model did not succeed in linking payment to changes in outcomes (as opposed to submission of reports) and suffered from a weak incentives to report reliable and timely data from the community-level. Following these initial experiences the Community PBF has been redesigned to address the design weaknesses while learning from the initial experiences. Box 16.1: Development of the National PBF Model The national PBF model for health centers was designed during the first quarter of 2006 through an extensive consultative process involving various key stakeholders. The model is predominantly based on the Butare model it has steering committees with representation of health authorities, health facilities, donors and civil society and contracts between each of these entities. It also has key quality elements of the Cyangugu model, such monthly control at the health center level and incorporation of the quality supervisory tool as a performance measure. A complicating factor during the introductory phase of the National PBF was the parallel implementation of the decentralization reforms at the time. These reforms included far-reaching administrative reforms, decentralized decision-making power and budgetary control, making the Mayor the decentralized authority on sector budgets. In the health sector there were also significant reforms for example, the former District Health Teams (headed by the District Director of Health who was also the hospital Director) were disbanded, and merged with the hospital teams. A new district health department called Unité Santé et Famille et Protection des Droits de l Enfant (USF) was created, with a new staffing, new District Director of Health and under direct authority of the Mayors office (Ministry of Local Administration). Figure presents the purchaser-provider split as defined in the national model between the quality controllers, the purchasers and the service providers. Figure 1. National PBF model During the second and third quarters of 2006, the national model was rolled-out through the 23 PBF districts. The implementation of the national model is guided through a rigorous intervention study, and a the roll-out plan. There were three phases of the PBF roll-out plan. Phase 0 districts corresponds to the three PBF pilot projects ( ): (i) Nyamasheke and Rusizi, the Cyangugu Pilot ; (ii) Huye and Gisagara, the Butare Pilot ; and (iii) the BTC Ville de Kigali Pilot. Phase 1 districts are districts in which PBF was started since January Finally, Phase 2 districts are the seven control districts in which PBF was implemented in April Figure 16.1 presents the purchaser-provider split that is proposed for the Community PBF: Purchasers: the Health Center Management on behalf of all Donors; Provider: the CHW Cooperative; Controllers: the Sector PBF Steering Committee comprising the Sector administrative officials responsible for health, the health center management, and other public institutions or members like teachers, representatives of civil society organizations (local NGOs, faith-based organizations, etc.). 106

122 Figure The Purchaser-Provider Split in the National Community PBF Model 5. The Community Health Desk of the MoH provides overall policy guidance and support to the national Community PBF. The implementation of the Community PBF reflects a decentralized control and decision making model (see Figure 16.2 and 6. Figure 16.3). The Local Sector Administration signs a contract with the Community Health Worker Cooperative detailing the Community PBF performance indicators and the content of the reports that need to be submitted before a certain date. The Community Health Worker Cooperative reports to the Sector PBF Steering Committee as per the contract. The District health department and District hospital provide support to the PBF Sector Steering Committee: (i) general support and capacity building in the institutional mechanisms governing the Community PBF system; and (ii) support organizing data input using the Web application. The PBF Sector Steering Committee decides on the level of achievement: (i) discusses the quarterly sector community performance report and validates it; (ii) discusses any observations brought to the table by the Health Center; and (iii) approves the payment after determining successful performance based on the key performance indicators. CHW Cooperative receives payment from the health center after performance has been certified by the PBF Sector Steering Committee. The funds received by the CHW Cooperatives are pooled with other resources from income generating activities. Figure Flow of Funds in the Community PBF 107