FOR OFFICIAL USE ONLY

Transcription

1 Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized South Asia Region Human Development Unit Document of The World Bank FOR OFFICIAL USE ONLY PROJECT APPRAISAL DOCUMENT ON A PROPOSED CREDIT IN THE AMOUNT OF SDR 40.2 MILLION (US$60 MILLION EQUIVALENT) TO THE ISLAMIC REPUBLIC OF PAKISTAN FOR A SOCIAL SAFETY NET TECHNICAL ASSISTANCE PROJECT May 8,2009 Report No: PK This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents mav not otherwise be disclosed without World Bank authorization.

2 CURRENCY EQUIVALENTS (Exchange Rate Effective March 3 1,2009) Currency Unit = Pakistan Rupee (Rs.) Rs = US$1 US = SDR 1 FISCAL YEAR July 1 - June 30 ABBREVIATIONS AND ACRONYMS AJK AJWNA APL BISP CAS CCT CNIC CPI CSP DG DGR DPC EOBI FSP FY GDP GoP ICB IBRD IDA M&E MOF MIS MSWSE NADRA NCB NDO NGO NSPS NWFP PBM PC-I PCO PHRD PIU PIHS PMT PPAF Azad Jammu and Kashmir Azad Jammu and KashmirlNorthern Areas Adaptable Program Loan Benazir Income Support Programme Country Assistance Strategy Conditional Cash Transfer Computerized National Identification Card Consumer Price Index Child Support Program Director General Directorate General of Registration Development Policy Credit Employee Old Age Benefits Food Support Program Fiscal Year Gross Domestic Product Government of Pakistan International Competitive Bidding International Bank for Reconstruction and Development International Development Association Monitoring and Evaluation Ministry of Finance Management Information System Ministry of Social Welfare and Special Education National Database Registration Authority National Competitive Bidding National Database Organization Non-Governmental Organization National Social Protection Strategy North West Frontier Province Pakistan Bait-&Mal Project Concept - I Population Census Organization Policy and Human Resource Development Project Implementation Unit Pakistan Integrated Household Survey Proxy Means Testing Pakistan Poverty Alleviation Fund

3 FOR OFFICIAL USE ONLY ABBREVIATIONS AND ACRONYMS (continued) PRES0 PRSC PRSP PSDP PSLM PSNS RD RSP RSPN SBD SSI Poverty Reduction and Economic Support Operation Poverty Reduction Support Credit Poverty Reduction Strategy Paper Public Sector Development Program Pakistan Social and Living Standards Measurement Survey Pakistan Safety Net Survey Regression Discontinuity Rural Support Program Rural Support Program (National) Standard Bidding Documents Social Security Institutions Vice President: Country Director: Sector Director Sector Manager: Co-Task Team Leaders: Isabel M. Guerrero Yusupha B. Crookes Michal Rutkowski Mansoora Rashid Andrea Vermehren / Iftikhar Malik / Cem Mete 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 PAKISTAN SOCIAL SAFETY NET TECHNICAL ASSISTANCE PROJECT CONTENTS Page STRATEGIC CONTEXT AND RATIONALE Country and Sector Issues... 1 Rationale for Bank Involvement... 3 Higher Level Objectives to which the Project Contributes... 4 PROJECT DESCRIPTION Lending Instrument Project Development Objective and Key Indicators Project Components Lessons Learned and Reflected in the Project Design Alternatives Considered and Reasons for Rejection... 9 IMPLEMENTATION Partnership Arrangements Loadcredit Conditions and Covenants Economic and Financial Analyses Technical Fiduciary Social Environment Safeguard Policies Policy Exceptions and Readiness... 22

6

7 Annex 1: Country and Sector Issues Annex 2: Major Related Projects Financed by the Bank and/or other Agencies Annex 3: Results Framework and Monitoring Annex 4: Detailed Project Description Annex 5: Project Costs Annex 6: Implementation Arrangements Annex 7: Financial Management and Disbursement Arrangements Annex 8: Procurement Arrangements Annex 10: Safeguard Policy Issues Annex 11: BISP Management Information System Annex 12: Development ofthe Poverty Scorecard for Pakistan Annex 13: Project Preparation and Supervision Annex 14: Documents in the Project File Annex 15: Statement of Loans and Credits Annex 16: Country at a Glance Annex 17: Map

8

9 PAKISTAN SOCIAL SAFET NET TECHNICAL ASSISTANCE PROJECT PROJECT APPRAISAL DOCUMENT SOUTH ASIA SASHD Sector Director: Michal Rutkowski Project ID: P Sector: Social Protection Themes: Social Services Project Financing Data: [I Loan [XI Credit [ 3 Grant [ ] Guarantee [ ] Other: For Loans/Credits/Others: IBRD/IDA Others Borrower: Government of Pakistan Estimated disbursements (Bank FY /US$60million) FY Annual Cumulative Project implementation period: Four years Expected effectiveness date: July 31, 2009 Expected closing date: July 31,201 3 Does the project depart from the CAS in content or other significant respects? 1 No Does the project require any exceptions from Bank policies? no Have these been approved by Bank management? I n/a Is approval for any policy exception sought from the Board? no Does the project include any critical risks rated substantial or high? Yes Does the project meet the Regional criteria for readiness for implementation? no Yes, except for: Advance procurement: All first year contracts to be finalized by the time the project goes to the Board. The Task Team received a waiver for this filter. Project Development Objective The objective of the proposed project is to enhance the operation and management of a nationwide, effective and transparent safety net system for the poor in Pakistan.

10 Project Description The proposed project aims at strengthening the administration and performance of the country s safety net with a particular focus on the Benazir Income Support Program (BISP) program as the national safety net platform. To this end, the project would support the following four components: ComDonent 1: Establishment of a National TargetinP System ($34.6 million): This component would support the roll out of the national targeting system based on the experience made and lessons learnt from the targeting test phase (start in March 2009) which is supported under a DFID trust fund managed by the Bank. The component would include the finalization and update of the targeting design, the data collection with the poverty scorecard at the national level, the data processing and cross-checking with existing data bases as needed, the administration of targeting appeals, as well as the maintenance of the data base. Component 2: StrenPtheninP Safety Net Operation ($12.9 million): This component would provide support for BISP to operate the program smoothly and transparently at the central and provincial/local levels. This would support the implementation of a communications strategy including training and communication with beneficiaries, materials, the implementation of grievance redressal and reconciliation mechanism, review of payment mechanism, as well as the implementation of monitoring and social accountability mechanisms. Component 3: Enhancing Safety Net Program Management, Accountability and Evaluation ($10.8 million): This component would provide funding for establishing and operating the safety net Management Information System (MIS), for strategic staffing and capacity building; the establishment of effective controls (including audits, spot checks, cross verification with other programs, etc.); process and impact evaluations, and analytical and pilot work on program graduation strategies and links to human development services. Component 4: Developing the Social Protection Policy and Strateply Monitoring ($1.7 million): This component would support the design o f the institutional and legal framework for execution of the National Social Protection Strategy (NSPS) and design and implementation of a monitoring and evaluation system of the NSPS. It would also provide training and capacity building for policy makers and implementing agencies in social protection policy and program implementation. The project would provide funds for technical assistance, equipment, materials, training, operational costs and minor office refurbishment costs over a four year period starting in July, BISP would be the implementing agency for all four components. However, the Planning Commission in its role to advance the social protection policy would provide the technical and strategic guidance for the implementation of the fourth component while BISP would be responsible for the administration of that component. Which safeguard policies are triggered, if any? No safeguard policies are triggered by the project

11 Significant, non-standard conditions, if any, for: Board presentation: None Loadcredit effectiveness: None Covenants applicable to project implementation: Conditions of Disbursement (i) Two FM specialists to be assigned to the PIU hired prior to Disbursement. (ii) Finance and Accounting guidelines for project issued prior to Disbursement. (iii) Two Procurement specialists to be assigned to the PIU hired prior to Disbursement. Dated Covenants By no later than six months after the Effective Date, BISP shall appoint an independent qualified third party, under terms of reference acceptable to the World Bank, to conduct regular spot checks/operational audits on the performance of the safety net program processes and reporting. By no later than sixty days after the Effective Date, a Memorandum of Understanding shall be signed between BISP and Planning Commission regarding implementation collaboration for component 4 of the project reached, including a work plan. By no later than one year after the Effective Date, BISP shall link BISP management information system with the beneficiary database of data processing agencies, to monitor the implementation of BISP program. By no later than three months after the Effective Date, BISP shall approve the Finance and Accounting Manual for the BISP in order to implement the safety net program. By no later than eight months after the Effective Date, BISP shall enter into contracts with data processing agencies for designing and managing a database and a management information system for carrying out targeting of eligible beneficiaries of BISP. By no later than four months after the Effective Date, the BISP shall develop and thereafter maintain, a publicly accessible website, in terms satisfactory to the Association, showing the Project procurement plans, the status of procurement of various contracts (the summary of proposalsibid evaluations and awards) and a list of procurement complaints and status thereof.

12

13 STRATEGIC CONTEXT AND RATIONALE 1. Country and Sector Issues 1. Despite recent periods of economic growth, poverty reduction in Pakistan is at risk of being reversed in the wake of financial crises and an uncertain macroeconomic context. Between 2001/02 and a sustained period of economic growth saw a fall in the national poverty rate by more than 11 percentage points. Yet recent estimates show that nearly a quarter of Pakistan s population remains poor, with a significant population clustered around the poverty line. Aside from low levels of income, vulnerability to adverse shocks is a concern. Both aggregate shocks, e.g., the recent global food, fuel and financial crises and natural disasters and idiosyncratic shocks, e.g., health and unemployment - have proven to impose significant costs on households, particularly the rural poor. Coping strategies typically have a negative impact on welfare and tend to perpetuate inter-generational poverty. For example, the poor often reduce food intake (both quality and quantity), increase labor supply, sometimes by withdrawing their children from school. Consequently shocks such as the recent food crisis of 2008 and the impacts of the global financial crisis are expected to reverse the declining trend unl ess policy measures are taken to protect poor and vulnerable households from adverse shocks. 2. Existing publicly financed social protection programs in Pakistan are limited in their coverage, administration, targeting efficiency, and ability to respond to shocks. Until the introduction of the Benazir Income Support Programme in 2008, Pakistan s safety net system (i.e., non-contributory transfers targeted to the poor) comprised basically of two cash transfer programs, namely, Zakat and the Food Support Program administered by the Bait-ul Maal. Both Bait-ul-Mal and Zakat are weakly targeted to the poor: only 46 percent of total expenditures of Bait-ul-Mal (and 43 percent of total Zakat expenditures) reach the poorest 40 percent of the population.2 More broadly, a disproportionate share of social protection expenditure appears to benefit the non-poor. Between 2004/05 to 2007/08 social protection expenditures increased from 3-4 percent of pro-poor PRSP spending to 13.4 percent. This increase comprised mainly in social security spending for formal sector workers (75 percent in 2007/2008) and food subsidies. Given that a large share of total social protection expenditures comprises social security spending and the sizeable leakage of safety net benefits to the non-poor, a significant share of social protection spending in Pakistan likely accrues to the non-poor. Government s Social Protection Strategy! 3. In light of the need to strengthen the country s social protection mechanisms, the Government of Pakistan, with support from the World Bank and other donors, started a collaborative process in 2006 to develop A Social Protection Strategy to Reach the Poor and the Vulnerable, which was approved by Cabinet in June The Social Protection Strategy is conceived within the overall framework of Just and Balanced Development envisaged in the Medium Term Development Framework ( ) and Vision The overarching vision of this strategy is to develop an integrated and comprehensive social protection system, covering all of the population, but especially the poorest and most vulnerable groups (for more detail, see Annex 1). 4. The cornerstone of this strategy is to use cash transfers targeted to poor households as a starting point to protect against the impacts of adverse shocks and to promote human development. Additional core instruments such as conditional cash transfers and skills development have also been foreseen to promote investment in human and physical assets of the poor. In this context, identified vulnerable groups include households with high dependency ratios, particularly those headed by single woman, as well as World Bank (2007) Social Protection in Pakistan: Managing Household Risks and Vulnerability. World Bank (2007), ibid. I

14 socially excluded groups defined along lines of religion, language gender, access to land, occupation (e.g., construction) and disability. In order to fulfill this mandate the need to enhance the operation and management of a nationwide, effective and transparent safety net program for the poor is paramount. Towards implementation of the strategy: Building a national social safety net platform 5. During the second half of 2008 the Government of Pakistan launched the Benazir Income Support Programme (BISP) as a flagship government initiative. The program provides cash transfers of Rs [$12] per month to eligible families and is expected to cover 3.4 million families in 2008/ The short term objective of BISP is to cushion the adverse impact of the food, fuel and financial crisis on the poor. Its broader objective - in line with the National Social Protection Strategy - is to establish a nationwide safety net program, marked by effective and transparent targeting and delivery mechanisms. The 2008/09 budget allocation for the BISP (Rs. 34 billion), doubles the federal Government s social safety nets spending from 0.3 percent in to 0.6 percent of GDP.4 The GoP intends to expand the budget for the BISP in the coming fiscal year 2009/10 to cover around 5 million families. The Government intends to increase the coverage of the safety net program, with safety net spending expected to reach approximately 1 percent of GDP in 2009/10. The Government s ultimate goal over the medium term is to expand coverage to 7 million families nationwide. 6. Given the imperatives of the evolving food and financial crises, combined with weak existing targeting capacity, the Government started the BISP program in mid 2008 through a legislator -based targeting system. BISP application forms were distributed to Members of the National Assembly (h4nas) and provincial parliaments in order to identify poor and needy families for the program. The limited information provided through these forms was processed by the National Database Registration Authority (NADRA) and cross-checked with other databases (e.g., with the Computerized National Identification Card (CNIC) and passport databases). 7. With the support of the World Bank and other donor agencies, the GoP is moving to consolidate its recent safety net reforms through a combination of activities that will strengthen program targeting, operations and management. At the centre of these efforts is the recognized need to enhance national capacity and expertise given the ambitious nature of the program, as well as the fragmented nature of existing programs. Consistent with international best practice, the Government has put in place separate agencies to conduct the enrollment process, determine program eligibility and pay benefits. The Government has also included monitoring, evaluation and other control and accountability mechanisms of the program as an important element of the reform agenda. The Government s commitment is particularly evident in the formal adoption of a poverty score card as an objective targeting instrument to identify safety net beneficiaries-a first in the South Asia region. Given the limitations of the legislators based targeting system, the Government of Pakistan (GoP) decided in December 2008 to improve BISP targeting through the adoption of a poverty scorecard that is based on a proxy means testing approach, thus creating a national targeting system that could serve multiple purposes. Roll-out of the poverty scorecard nationwide is an extensive task. A decision has therefore been made to start the scorecard roll out in March 2009 with a test phase in selected districts across all provinces of the country. This test phase will provide useful insights and the necessary learning for scaling up the program nationwide, including for improvements in the enrollment, payment, grievance/appeals and monitoring/evaluation processes. This level of coverage is approximately equal to about 25 percent of the population-around the national poverty headcount rate-according to the most recent estimates. In addition, the province of Punjab is allocating about 0.2 percent of GDP to its income suppodcash transfer program in A higher level of spending does not translate into more coverage for the poor if safety nets programs are poorly targeted. 2

15 8. Through the development of BISP and its supporting delivery mechanisms the GoP aims to achieve its long-run vision of an effective social safety net program that can promote human capital development and contribute to inclusive economic growth in Pakistan. This will be marked by a number of defining attributes. First, poor and vulnerable households will receive some relief to cope with adverse income effects of the r.ecent crises. The program will also provide basic income support to the chronic poor, consistent with the Government s re-distributive goals. Second, Pakistan will have a robust safety net system that can be scaled-up in response to future adverse shocks. Third, a nationwide, effective and transparent safety net system will provide the basis upon which it will be possible to leverage complementary graduation strategies such as micro-finance, skill development programs, or basic health and education services. Finally, the introduction of an effective safety net system may enable policy makers to take much needed structural reform measures (e.g., reduction of subsidies) while protecting the poor, thereby promoting dynamic efficiency. 2. Rationale for Bank Involvement 9. The Bank has supported the strengthening of safety net systems in Pakistan for several years, widening its engagement considerably to include: (i) analyses of the efficiency of social protection expenditures; (ii) technical assistance to formulate a National Social Protection Strategy (NSPS) that was approved by Federal Cabinet in June 2007; (iii) analytic work on the targeting performance of existing social assistance programs and the development of a poverty scorecard based on a Proxy Means Test (PMT) formula; (iv) policy dialogue with provincial governments on social protection policy and programs; (v) support for social care services for the disabled, and (vi) technical assistance to support the design and implementation of a pilot Child Support Program (CSP), a federal conditional cash transfer (CCT) scheme under the Ministry of Social Welfare. 10. The introduction of a national income support program presents an ambitious reform agenda, which the Bank is positioned to support through a programmatic mix of interventions: 1. The proposed Social Safe- Net Technical Assistance (SSN TA) project will be implemented over a four year period to provide the technical support to establish the national targeting system, strengthen the BISP operation, support BISP program management, accountability and evaluation, and advance the analytical and data work for the National Social Protection Strategy and its monitoring. 2. In parallel to this project, the World Bank has also been asked to provide budgetary support through a Social Safe- Net Development Policy Credit (SSN DPC) which would support policy actions to be taken by the Government on the design and implementation of the BISP and other social safety net measures. The proposed DPC focuses on establishing systems, strengthening institutions and governance through three pillars: (i) improve the targeting efficiency of the BISP; (ii) establish an institutional framework for program implementation; and (iii) enhance fiscal sustainability and strengthen the fiduciary environment for the program. 3. Preceding both operations the Bank has initiated a Pover- Reduction and Economic Support Operation PRESO) as a budgetary support operation, prepared for the WB Board approval in the third quarter of FY08/09. This included a prior action that the Prime Minister adopt the scorecard as the new targeting instrument for BISP for roll-out in the first half of Further support for the government s social protection agenda is envisioned to be detailed in the Bank s upcoming Pakistan Country Assistance Stratew, possibly in collaboration with other development partners some of which are interested in supporting the strengthening of the country s safety net system. 3

16 11. Informing these interventions the Bank draws from previous experiences of social protection involvement. This includes: (i) Cash transfers in earthquake-affected areas in Pakistan: The proposed project will build upon the experience and knowledge generated by this project regarding experiences with the development and implementation of cash transfers in earthquake affected areas, including for example grievance and redressal processes. (ii) Pakistan Poverty Alleviation Fund: The Bank has been supporting the Pakistan Poverty Alleviation Fund for many years and has just recently provided support for the implementation of the same poverty scorecard as a monitoring and evaluation mechanism. The field experience of this endeavor will be a valuable learning opportunity for the testing and roll-out of the poverty scorecard mechanism as a national targeting instrument, particularly since PPAF is one of the Partner Organizations of the BISP during the test phase of the scorecard. The organization also implements a World Bank financed project that includes a component to develop complementary interventions to safety nets. (iii) National Social Protection Policy in Pakistan: Strategic advice in the development policy framework, including Economic and Sector Work, and support of national dialogue. 12. In addition, the Bank has an established international expertise in the design and implementation of social safety nets including: (i) (ii) (iii) (iv) Provision of strategic advice in developing a unified targeting system building on objective and transparent mechanisms Substantial expertise in the design, implementation and evaluation of safety net programs and interventions, including cash transfer programs Advice on best practices on a system of controls and accountability mechanisms for safety net programs which in turn will enhance the program s impacts on the poor, and Access to international experience with exit strategies and complementary measures such as linking benefits to human development measures, providing innovative mechanisms for enhanced youth employability, etc. 3. Higher Level Objectives to which the Project Contributes 13. The proposed project is consistent with the Government s objectives of reducing poverty and improving income distribution, as well as human capital development as outlined in Pakistan s National Social Protection Strategy. Within these broad objectives, the Government seeks to support inclusive growth through the development and implementation of a financially sustainable, efficiently targeted, and well administered national safety net system in Pakistan. The safety net system will provide the chronic and transient poor with both basic income support and access to opportunities for graduating out of poverty. The SSN TA project would support these efforts by (i) building the foundations for a national, unified targeting system that could progressively be used for a defined set of targeted programs, (ii) strengthening the institutional capacity to roll out and manage an efficient and transparent national safety net system including the piloting of safety ladders through linkages with complementary measures, and (iii) supporting the evolution of the National Social Protection Strategy (NSPS) through the establishment of a monitoring and evaluation system for the strategy across programs and interventions. The project is therefore seen as the beginning of a longer term engagement to establish structures and systems for better social protection of the poor in the country - rather than a one-off project. Further support would be defined under the Bank s Country Assistant Strategy and in collaboration with interested donors. 4

17 PROJECT DESCFUPTION 1. Lending Instrument 14. The proposed lending instrument is a Technical Assistance Credit (TAC). A TAC was chosen because the Project mainly provides technical assistance to enhance the management and operation of a nationwide safety net program, which in its nature requires intense technical support to ensure appropriate targeting, operational and management mechanisms. The project design builds on the prior technical assistance to develop a National Social Protection Policy, as well as the poverty scorecard for targeting safety net benefits. A TAC is the most appropriate lending instrument to support the Government of Pakistan s efforts to enhance its institutional capacity for: (i) establishing the national targeting system; (ii) building the institutional capacity to operate a national safety net program in the mentioned dimension; (iii) strengthening the institutional framework to manage the program; and (iv) further developing the country s social protection strategy through analytical work, policy dialogue, and monitoring and evaluation. Strengthening government and non-government institutional capacity to work together, and providing essential tools for targeting, implementing and monitoring social safety nets is expected to lead to increased evidence-based decisions and budget allocation for social protection expenditures. 15. The timeliness and compatibility of the TA Credit with other interventions is an important consideration. The TA is being combined with a Social Safety Net Development Policy Credit (DPC) to support the immediate development of a safety net system. Pakistan s high vulnerability levels and unpredictable economic context provided a strong imperative for the Government and donor agencies to move quickly in establishing a national income support program. This has been enabled through Trust Fund financing provided by the Department for International Development (DFID) of the United Kingdom since January Through DFID support the Bank has been able to develop a series of institutional.and operational tools to help BISP establish itself as a federal safety net program. As the proposed project develops in the medium term the Bank expects continued financing interests by other development partners. 2. Project Development Objective and Key Indicators 16. The objective of the proposed project is to enhance the operation and management of a nationwide, effective and transparent safety net system for the poor in Pakistan. Progress toward this objective would be measured by key performance indicators, including: Key Performance Indicators Targeting: Percentage of disbursed BISP cash transfers received by beneficiaries in quintiles 1 and 2 (40 percent poorer households) Coverage: Number of households to which the poverty scorecard has been applied. Accountability: Percentage of participating districts in which beneficiary lists are publicly available. Target values Baseline5 End of Project 46%6 70% 0 At least 10 million households 0 80% Baseline data from March This figure is based on data from the PSNS, a nationally representative survey which shows that only 46 percent of total benefit expenditures of Bait-ul-Mal reach the poorest 40 percent of the population, while 43 percent of total Zakat expenditures reach the same population group. 5

18 Key Performance Indicators Beneficiary satisfaction: Percentage of beneficiaries satisfied with program implementation, measured by beneficiary assessment Sensitivity of selected safety net programs to vulnerable groups: NSPS reports that include information on vulnerable groups Target values Baselines End of Project nla 75% nla Safety net monitoring reports available with information on vulnerable 17. Annex 3 provides additional details on these indicators, the sources of information for their monitoring, as well as the agencies responsible for their collection. 18. Progress towards the ultimate objectives of poverty reduction and human capital accumulation will be difficult to assess and to attribute to the proposed project since it mainly provides technical assistance to enhance safety net administration. Nonetheless, in order to measure the impacts of the safety net reforms on this higher level, an impact evaluation will be carried out, partially financed by grant money (e.g., DFID Trust Funds) and partially by the proposed Pakistan SSN TA project. The impact evaluation would seek to measure improvements in family consumption, school attendance, in nutritiodhealth, employmentlchild labor and possibly other human development aspects. 3. Project Components 19. The proposed project aims at strengthening the administration and performance of the country s safety net with a particular focus on the new BISP program as the national safety net platform. Using cash transfers, the BISP seeks to reach up to 7 million families living below the poverty line, located in rural and urban areas in all districts of the country. In reaching 7 million families, it would roughly cover 35 million people or about 22% of the population. Payments will be made to the female head of the family every two months, providing families with Rs (about $150) income support per year which is about 26% of the poverty line. 20. To this end, the project would include four components: (i) establish a national targeting system; (ii) strengthen safety net operations; (iii) enhance safety net program management, accountability and evaluation; and (iv) develop the social protection policy and strategy monitoring. The project would provide funds for technical assistance, equipment, materials, training, operational costs and refurbishment of offices over a four year period. The Government counterpart funds would finance the cash grants in full, some equipment and the essential staff on central, provincial, and local levels. BISP would be the implementing agency for all four components. However, the Planning Commission in its function to advance the social protection policy would provide the technical and strategic guidance for the implementation of the fourth component while BISP would be responsible for the administration of that component. Component 1 : Establishment of a National Targeting System ($34.6 million) 21. This component would support the roll out of the national targeting system based on the experience made and lessons learnt from the targeting test phase (start in March 2009), which is supported under a DFID trust fund managed by the Bank. The component includes the finalization and update of the targeting design, the data collection with the poverty scorecard at the national level, the data processing and cross-checking with existing data bases as needed, the administration of targeting appeals, as well as the maintenance of the data base. 6

19 Component 2: Strengthening Safety Net Operation ($12.9 million) 22. This component would provide support for BISP to operate the program smoothly and transparently at the central and provincial/local levels. This would support the implementation of a communications strategy, the improvement of beneficiary selection and program enrolment process, strengthening of mechanisms for payment provision and reconciliation mechanisms, the implementation of grievance redressal and monitoring, the establishment of field offices, as well as the establishment of social accountability mechanisms. Component 3: Enhancing Safety Net Program Management. Accountability and Evaluation ($1 0.8 million) 23. This component would provide funding for establishing and operating the safety net Management Information System (MIS), for strategic staffing and capacity building; the establishment of effective controls (including audits, spot checks, cross verification with other programs, etc.); process and impact evaluations, and analytical and pilot work on program graduation strategies and links to human development services. Component 4: Developing the Social Protection Policy and Strategy Monitoring ($1.7 million) 24. This component would support the design of the institutional and legal framework for execution of the National Social Protection Strategy (NSPS) and design and implementation of a monitoring and evaluation system of the NSPS. It would also provide training and capacity building for policy makers and implementing agencies in social protection policy and program implementation. 4. Lessons Learned and Reflected in the Project Design 25. Political leadership combined with clear organization structures and coordinating mechanisms are key to program acceptability and sustainability. International experience suggests that successful implementation and institutionalization of cash transfer programs requires both strong support from the highest level of Government (provided by the President, Ministry of Finance (MOF) as well as the Prime Minister s Office) and a sound operational design with clear institutional responsibilities for program coordination and implementation (as described in the Operations Manual, see also annex 4). A limitation of safety net interventions to date within Pakistan concerns the weak institutional settings and lack of coordination on the part of individual agencies and programs working in this area. 26. Cash transfers offer predictable and timely income support to smooth consumption and promote human development outcomes. The project design is based on the proven effectiveness of cash transfers to overcome shortcomings that traditional aid programs have not been able to solve. Impact evaluations of cash transfer programs in Latin America have shown that, whether conditioned or not, demand subsidies via cash or voucher transfer programs not only improve consumption of the family as a whole, but also improve school attendance, as well as health and nutrition outcomes. In Ethiopia, Lesotho, Mozambique and Zambia, children benefit from transfers even though they are not the programs primary targets. The better the cash transfers are targeted to the poor, the larger the impact on poverty and human development outcomes. Well-implemented cash transfer programs are among the most effective programs to distribute income and benefit the poorest p~pulation.~ Lindert, K., E. Skoufias and J. Shapiro (2006) in Redistributing Income to the Poor and the Rich: Public Transfirs in Latin America and the Caribbean. 7

20 27. Cash transfers also provide an entry point to address vulnerabilities and empower women. Proving the cash transfers to women helps ensure that money is spent to benefit the family and particularly children. International evidence on intra-household allocation and on the use of conditional cash transfers in other countries provided to women has repeatedly shown the increases in women s income translates in more expenditures for food, children, clothing, education supplies and other goods for children (shoes, medicine, etc.). In addition, female beneficiaries have reported an increase in their self-esteem and improved relationships in the family - with less incidence of domestic violence by their partners or other decision-makers (in-laws) as they are more able to contribute to the household s economy. 28. Available empirical evidence shows that the impact of cash transfer programs on poverty and inequality is significantly higher when effective and accurate targeting mechanisms are used. Designing and constructing such a mechanism requires that beneficiaries are identified and selected according to appropriate criteria. In addition, updating the mechanism so that it adequately reflects changes in the country s welfare is a complex and, generally, costly process as shown by the recent beneficiary recertification process in Colombia. A good targeting system may be complex to develop but can be used for many programs, not only for direct transfers in cash or in kind, but also for entry into programs that provide health carehsurance, schooling, training, etc. When effective targeting instruments and recertification procedures are in place, these can progressively be expanded to other programs. The shared overhead is not only efficient, but can lead to a more coherent overall social policy. 29. Detailed monitoring, evaluation and accountability mechanisms need to be designed and implemented from the outset of a safety net program. A strong Monitoring & Evaluation (M&E) system makes it possible to assess whether existing programs fulfill their objectives and how performance could be improved. Programs are monitored to evaluate their targeting, assess beneficiary take-up rates, and measure impact on selected program outcomes. For this purpose the M&E system collects information on selected process and result indicators. It tracks program caseloads and spending, estimates inclusion and exclusion targeting errors. The information generated by Management Information Systems (MIS) serves two critical functions: it allows program managers to assess program operations and monitor program targets, and it enables policy-makers, financiers and the general public to hold programs, and their management, accountable for their performance. Consequently, setting up an integrated MIS is considered crucial for any program-with continuous updates throughout the program life. In addition monitoring should be complemented with rigorous impact, process and beneficiary evaluations. The findings generated by such evaluations constitute key inputs for program fine-tuning and re-design to meet the program s main objectives. Furthermore, the dissemination of results from impact evaluations has proven effective in ensuring continued political and public support for safety net programs (e.g., Mexico, Colombia, Kenya). 30. A system of control and accountability mechanisms needs to be in place to provide transparent, timely and adequate information to all stakeholders. Bank reviews of the control systems in CCT programs in Latin America and South Asia highlight the following lessons which were taken into account in the project design: Institutional design. Define lines of authority and accountability, including administration and service delivery. Align incentives and financing. Identify risks and remedies up front. Targeting. Focus attention on possible exclusion errors (that is, eligible households not being included) and define rules for recertification of eligibility. World Bank (2007), Control and Accountability Mechanisms in Conditional Cash Transfer Programs: A Review of Programs in Latin America and the Caribbean. Operational Innovations in Latin America and the Caribbean, Vol. 1, No. 1, 8

21 e e e e e e e Conditionalities (for CCT programs only). Balance accuracy and the burden of verification. Examine reasons for non-compliance. Payment processes. Ensure cash-flow monitoring and associated control procedures. Management information systems. Strengthen focus on design and effective implementation of MIS systems. Where possible, encourage cross-checks of enrollment with other databases. Internal quality control. Include process evaluation, feedback loops, and correction mechanisms within programs, particularly through spot checks. Accountability. Ensure appropriate mechanisms to handle complaints and appeals in a timely manner, and publish program data. Financial management and auditing. Ensure payment accounting and clearly define reporting requirements. Focus audits on risk areas. Make use of supreme audit institutions if adequate. Procurement. Closely supervise the implementation of large service contracts. 31. Separation of operational functions in the implementation of a cash transfer program improves program efficiency. Cash transfers ideally build on existing capacities in the country. Keyfunctions such as targeting, operation, payments and controls should be carried out by separate actors to allow for optimal checks and balances. This, of course, requires a strong, effective central unit to coordinate the program (see Annex 6 for implementation arrangements). 32. Clear and transparent rules to provide support and apply sanctions ensure program credibility. Successful cash transfer programs have built a broad credibility and accountability among beneficiaries and other stakeholders. This requires clear and transparent rules and their strict application, with defined oversight institutions. It is necessary to define and communicate to all stakeholders the transparent criteria for selecting families and determined benefit amounts, and sanctions for fraud and mismanagement by beneficiaries, program staff and local officials (e.g., postal services). Finally, the credibility of the program also rests on its ability to keep an ongoing communication with and receive feedback from participating households. It is crucial to implement an efficient system to track complaints and ensure that they receive a timely and adequate answer. 33. Cash transfer programs can be devised to complement graduation strategies and bridge to human development services and programs. International experience suggests that efforts to link beneficiaries of transfer programs to other complementary services and to help them develop graduation strategies can be an effective way to lift them out of poverty. The proposed project envisions the BISP as the central axis of a broader social protection strategy that aims to aid the poor while helping them develop strategies for self-sufficiency (graduation) and gain access to complementary programs and services. This follows the success of similar income support programs, e.g., Jamaica s food stamps transformation into PATH (Program for Advancement Through Health and Education), Ecuador s Bono Solidario transformation into Bono de Desarrollo Humano, etc.) or workfare elements (as in Bulgaria and Romania s income generation programs). 34. Federal-provincial linkages need to be considered to minimize any overlaps and ensure full project synergies. Throughout the project s design this has been recognized through the involvement of local level administrations. This is particularly important to ensure strong operational underpinning of central and regional offices. Encouragingly, two regions have adopted the poverty scorecard for other programs. 5. Alternatives Considered and Reasons for Rejection 35. Lending instrument: Initially, an Adaptable Program Loan (APL) was considered for the project since APLs have proven useful in providing phased support for long-term development goals, while providing financial resources for required investments. This approach was eventually disregarded given 9

22 the Government s objective to revamping the safety net system in the short run (during the next 3-4 years). A Specific Investment Loan (SIL) was not considered because the project does not include any works or grants. The programmatic character of the support, however, is reflected in the combination of Bank instruments as described above (PESO, SSN DPC and TA operations) to address simultaneously the operational as well as the policy decision needs of the ambitious goals of Government to establish a unified safety net program at the federal level. The use of several Bank instruments in parallel also provides for flexibility for development partners to provide co-financing in accordance to their possibilities. 36. Supporting only the Pakistan Bait-ul-Mal to expand its Child Support Program (CSP): Since 2006, the Bank has been successfully supporting the PBM to pilot a conditional cash transfer program for children. While this initiative will be further supported by the Bank, the main focus of the proposed technical assistance project would be the establishment of the BISP as the country s main safety net program. Lessons from the PBM CSP pilot will be valuable as the Government develops graduation and exit strategies for the new safety net program. IMPLEMENTATION 1. Partnership Arrangements 37. The World Bank and DFID are actively working together in support of BISP under a common framework. In order to provide effective Technical Assistance to BISP in the short term, DFID approved a Trust Fund, managed by the Bank, to support and accompany the test phase of the poverty scorecard and to set up the initial organizational and operational systems for implementing the BISP. In addition, continued support is provided by DFID to the pilot of the PBM CSP. 38. Initial project preparation of a Social Protection Project was financially supported by a PHRD grant from the Government of Japan, which was executed by the MoSWSE and PBM. The PHRD grant was closed as planned on December 3 1, 2008, with little money disbursed as priorities of the Government had changed to establishing the BISP. The Japanese has shown interest in supporting the government s overall social safety net agenda. Direct support may evolve over time. Institutional and Implementation Arrangements 39. Towards the end of 2008, the Benazir Income Support Programme was transferred from the Finance Division to the Cabinet Division and a Special Secretary was appointed along with a National Program Coordinator. Moreover, the Special Secretary was declared as the Principal Accounting Officer with all administrative and financial powers as given in the Rules of Business of Pakistan for Federal Secretary, in charge of a Division. The Special Secretary was later replaced by the Managing Director enjoying the same powers as of a Federal Secretary. In April 2009, the President of Pakistan signed an Ordinance creating BISP as an autonomous safety net authority to, inter alia, reduce poverty and promote equitable distribution of wealth especially for the low income groups. 40. Governance Structure: Since the end of April 2009, the BISP is governed by a Management Board headed by the Chairperson and advised by a Council headed by the President and the Prime Minister of Pakistan which are the Chief Patron and Executive Patron of the Council, respectively. The Board is composed of members including the chairperson. Board members as well as the Secretary of the Program are appointed by the Chief Patron and the Executive Patron of the BISP Council. Board members include representatives of government, non-government organizations and technical experts who have relevance to the purpose of the program. The Council will comprise highly reputable, 10

23 distinguished and well accomplished national and international individuals. The Chairperson will be the ex-officio member of the Council, the Secretary will be the ex-officio member of the Board. 41. Management Structure: The head of BISP management is the Secretary as defined in the BISP Ordinance with all administrative and financial powers given in the Rules of Business of Pakistan for a Federal Secretary. As part of the BISP management structure, a Director General (DG) Operations and DG Administration have been appointed. In addition, Directors of Operations, Finance and Accounts, and Internal Audit along with other key staff have been appointed. At provincial level, a DG will head each province including AJWNA (a total of four DGs, 2 Directors AJK and Northern Areas). Provincial DGs shall be assisted by 28 Divisional Directors, who shall be further assisted by 125 Supervisors. So far, most of the DG and Director positions have been filled whereas some of the other staff still has to be appointed as the BISP rolls out its operation (see Annex 6 for more detail). 42. Project implementation will be supported by a Project Implementation Unit (PIU) under the Director General of Operations as Project Director. The PIU would include a Project Coordinator, as well as procurement and financial management specialists. Financial management specialists will report directly to the Director of Finance and Accounts. BISP Partners in Program Administration 43. BISP is presently requesting the services of the National Database and Registration Authority (NADRA), a parastatal organization, for the design and management of the targeting database; verification of eligible beneficiaries selected through the targeting process, and generation of beneficiary lists. This set of tasks may evolve over time depending on the availability of feasible alternatives in the private sector. 44. BISP has also signed an agreement with Pakistan Post, which stipulates that BISP beneficiaries shall be paid through Pakistan Post Money Order. This arrangement was favored keeping in view the vast coverage, easy accessibility and money handling and delivery capacity of Pakistan Post (for more information, see Annex 6). Over time, however, BISP management seeks to investigate alternative payment options using modern technology such as smartcard or others. The options will be analyzed and piloted under the proposed project. 45. In order to reach out to the entire population, BISP management has decided to contract selected Partner Organizations (to carry out the field work for the targeting process). For the targeting test phase, the PPAF and the RSPN (both NGOs) and the Pakistan Census Organization, a government agency, have been selected. Their performance will be evaluated during the test phase and their further participation in the program for the nationwide roll-out will be determined based on the predetermined performance evaluation criteria as established in the BISP Operational Manual. Control and Accountability Mechanisms, Monitoring and Evaluation of OutcomesIResults 46. Establishing a solid monitoring and evaluation system for BISP that includes effective checks and balances will be a key element of the implementation of the project. Such system of control and accountability mechanisms would include: (i) An integrated Management Information System (see annex 11 for more details) will be established and administered by BISP to monitor the program s physical and financial progress during implementation. The system would link the NADRA beneficiary database and the Postal Services database for payments with the central MIS managed by BISP. The MIS would produce timely reports and track achievement of selected project performance 11

24 (ii) (iii) (iv) (v) (vi) (vii) indicators included in the Results Framework (Annex 3). The web-based system will provide crucial data to the public on the selection of beneficiaries and overall program performance by region, district etc. Financial Management and Auditing Arrangements. In addition to financial audits carried out by the Auditor General s office, BISP will also introduce a spot-check system ( operational audits ), implemented by a third party, to verify the quality of information feeding the monitoring system and the program s operational systems. Spot-checks will occur regularly on a sample basis and will initially focus on the quality of data collection during the targeting process and the payment process. It is expected that BISP will also set up an internal audit unit within the institution as part of its long term institutional development. Procurement Arrangements: The procurement under the Project will be carried out in accordance with the World Bank s Guidelines: Procurement Under IBRD Loans and IDA Credits dated May 2004 and revised October 1, 2006; and Guidelines: Selection and Employment of Consultants by World Bank Borrowers dated May 2004, revised October 1, 2006; and the provisions stipulated in the Legal Agreement. The BISP will develop its own standard operating procedures in the form of a handbook. This handbook will be based on the Public Procurement Rules Exploring new information technologies (IT) for beneficiary payments. During the initial stages of project implementation BISP will review the options for using IT in the implementation of the program, particularly with respect to the payment cycle where government is keen to introduce a smartcard option for payments. The project makes provision for the potential implementation of appropriate IT solutions, particularly for payments to beneficiaries, as they become available. Appeals and complaint mechanisms. After the selection of beneficiaries through the targeting process, the list of eligible families will be published. Families that feel they have been mistakenly or unfairly excluded as potential beneficiaries and believe they meet the eligibility criteria can appeal the decision. The procedures for this appeals process are spelled out in detail in the BISP operational manual. In addition, beneficiaries are expected to make complaints about a variety of program features and procedures. Most complaints are usually made with respect to payments. All complaints will be processed systematically through the MIS, and follow up actions will be documented. Social accountability mechanisms will help make project implementation transparent and accountable through a variety of participatory monitoring techniques. These would include the publication of beneficiary lists at the appropriate sights, systematic oversight by local actors, a hot line for complaints and confidential allegations, and others. The social accountability mechanisms would provide valuable feedback and suggestions to program managers, and local and central government officials about the quality of BISP implementation and existing bottlenecks. The role and use of impact and other forms of evaluation. An evaluation strategy has been developed for BISP that includes qualitative and quantitative surveys to assess processes and impacts. The proposed project supports the integration of these elements within a comprehensive M&E system for the BISP. The new system will also extend to provincial and local entities as necessary. An evaluation, contracted to an external firm, will assess the impact of the BISP, particularly with regard to targeting outcomes, educational and nutritional impacts of benefiting families as well as impacts on family behaviors. The impact evaluation would consist of a baseline survey, collected prior to the scorecard roll out, and follow-up surveys of control and treatment groups (for details, see Annex 3). 12

25 47. Sustainability of the project investments can be expected at three levels: (0 (ii) (iii) At the household level, the increase of poor families income is expected to both increase their consumption, reducing poverty, as well as increase families investments in human capital (through school attendance of beneficiaries children, better nutrition, increased availability for health expenditures, etc.). The income support program would provide beneficiaries sustained benefits overtime including through higher earnings, better mental and physical health, and better management of crises. As shown by the Economic Analysis (see Annex 9), the poverty reduction impact and the potential return on investments in human capital are expected to be appreciable. At the institutional level, investments in establishing systems (national targeting system, management information system, evaluation system, etc.), training of staff at federal, provincial and local government levels as well as of community based organizations (e.g., partner organizations), and the development of public-private partnerships in the social protection sector would lead to more efficient and effective implementation of safety nets. At the fiscal level, the government is expected to considerably increase its social safety net expenditures over time (see fiscal analysis in Annex 9). In addition, many development partners are looking at the BISP with interest for possible support in the future. The project would develop systems with checks and balances to ensure that the incoming funds for social safety nets are spent in an effective way. 2. Critical Risks and Possible Controversial Aspects 48. The Project faces several critical risks, given the country context, the program s initial design, and its inherent features of being a safety net program. As seen above, the track record of safety net programs in Pakistan is not convincing. Many programs have been used or accused of being used for political purposes. The reference to Benazir Bhutto in the title of the BISP is for some an indication of its political nature. BISP s initial targeting strategy through legislators did not help in this respect. Furthermore, lessons from other countries suggest that proper implementation of cash transfer programs can present considerable challenges and inherent risks, particularly when they are scaled up quickly which increases the risk of errors. The weakness of BISP as a new institution underscores the need for strong technical assistance and capacity building. It may limit the ability of BISP management to fully control activities at the local level in addition to security concerns. Even though the proposed project will not finance transfers to families, its success in supporting government to improve the BISP will depend on the mitigation of these and other risks. 49. The main risks to project success involve the overall volatile country situation including security concerns that could, for example, impede the nationwide roll-out of the targeting process, as well as the risks pertaining to the newly created implementing agency and potential drawbacks to the institutional changes supported by the project. Therefore, the overall risk for the project was assessed as substantial despite mitigation measures as described below: Potential Risks Risk Risk Mitigation Measures Risk Level Level after Mitigation Fragmentation of delivery of social safety net programs across national institutions. S 0 An Ordinance creating BISP as an M autonomous safety net authority has been issued by the President. In 13

26 ~ Dhase Potential Risks teversal of decisions to use the joverty scorecard as the national argeting instrument due to lifficulties in implementation, :hange in government or other easons. The BISP has been established as an iutonomous agency under )residential Ordinance IX of The Pakistan Constitution provides or a sunset clause for all lrdinances. Ordinance IX of 2009 vi11 eventually require the 'arliament's adoption to remain in orce. The chance of the Ordinance ieing repealed are deem slim. rechnical Design Risk Level S M Risk Mitigation Measures addition, the GoP has decided on the principle mandates of BISP and PBM. The project will strengthen the institutional capacity of BISP to act as the country's main social safety net program, and support the development of a coherent policy framework for social protection. Linkages and graduation strategies established e.g. with related safety net interventions (e.g. CSP) The project would support the targeting roll out through selection of credible and locally anchored Partner Organizations based on established performance criteria, rapid roll out to gain momentum and economy of scale. There is bipartisan consensus at the moment about the need for an effective safety net program. The project will support the further institutional development, including legal aspects if needed. Risk Level after Mitigation M L 'oierty scorecard could be rolled )ut too quickly and/or not mplemented in accordance with argeting guidelines. rargeting process generates high nclusion and exclusion errors H H Roll out is being implemented in phases. The initial test phase which is now ongoing, will provide valuable insights about the pace of further roll out. All materials (training, public information, targeting guidelines) will be adjusted based on this experience. Supervision of Partner Organizations, process evaluation, as well as spot checks are an integral part of test and roll out. Sample survey to assess targeting efficiency in the field. Spot checks and targeting process evaluations will provide information on efficiency as well. 14

27 Potential Risks Risk Risk Mitigation Measures Risk Level Level after Mitigation Grievance Redress System will be set up to capture and resolve complaints. Transition to new poverty scorecard H 0 Clear deadline for acceptance of"old" S targeting system may be hampered forms established at April 30, by high expectations generated by Public information campaign would the legislators-based targeting focus on managing some of the mechanism. expectations. Payments under old system will only be replaced when new system is in place. BISP will adopt a transition policy to ease any social costs of the transition. Other agencies implementing social S 0 Some agencies, e.g., PBM and M safety net programs are reluctant to use the poverty scorecard targeting mechanism. Grievance redressal and other social accountability measures are not sufficiently implemented due to insufficient institutional capacity. Payments may be delayed due to logistical or budgetary bottlenecks. Payments through money orders may cause irregularities, particularly at the local level. government of Punjab have already shown interest in using the poverty scorecard. Credibility, quality and timeliness of the availability of the shared database will be critical. Operational Manual includes S provision for these mechanisms. The manual includes grievance mechanisms for former beneficiaries under the old system. a 1 Public information campaign will include information on social accountability features (e.g., hotline). 1 Discussions about the use of local level institutions are underway and likely - to produce - feasible arrangements to deal with grievances. M Transmission of funds to Pakistan Post L offices ensures regular payment. Budget provided surpasses the H payment needs so far. The payment agency has developed an S MIS and procedures for the entire payment process Agreement between BISP and pay agency includes provision for quick (25 days) processing of reconciliation data which are available as of end of April Beneficiary assessment and spot checks will both focus on this area. Toll free number / Complaints cell to be established. I i H 15

28 Potential Risks Risk Risk Mitigation Measures Risk Level Level after Mitigation ISP loses Government smmitment and budgetary llocations in the medium term, thus laking TA unnecessary..ack of capacity of BISP to rapidly cale up the program given that it is new institution.?inancia1 Management Government s medium-term commitment to program captured in current budget allocation. Government has declare that beneficiaries will continue to receive benefits for continued period of time, to be defined. TA could be continued for other safety net programs. Inclusion of third party capacities (e.g., Partner Organizations, NADRA, payment agency, etc.) help with rapid implementation Organizational guidelines have been developed and agreed. Technical assistance will serve to provide essential capacity building and strategic staffing Longer term institutional capacity defined under upcoming BISP law as an autonomous authority. M. s erceived and actual iisappropriation of funds,ack of reconciliation between fund:.ansferred to postal services and ctually delivered, leading to nonientification of leakages hability to properly record and.econcile cash transfers H o l Comprehensive finance and accounting guidelines being developed on a priority basis to include appropriate internal control framework Independent internal audit to enhance control environment Internal audit to be report directly to the Board. Sophisticated financial management module to form a seamless part of the MIS to ensure adequate reconciliation mechanisms Ring fenced financial management arrangements for the project, including hiring of professionally qualified staff. Deadline for implementation of Finance and Accounting guidelines including the reconciliation mechanism included therein. Dedicated staff to be assigned ensuring that all transactions executed prior to implementation of new H 16

29 Potential Risks lelay in procurement holds up other ctivities, including scorecard roll,ut. Jon transparent and inefficient use )f Bank funds in implementation lelays in implementation due to lac1 )f procurement capacity. HI Social and Environmental Safeguards Risk Risk Mitigation Measures Risk Level Level after Mitigation procedures are completely and accurately recorded and reconciled. Hiring of specialized procurement staff at BISP. Procurement training of existing and new staff on Bank s procedures and guidelines. Preparation of a Procurement Manual. Preparation of a Procurement Plan to be agreed at appraisal. Develop and maintain a website posting all procurement notices and awards. S Reputational and Other The BISP has become the current administration s principal social initiative, and the President himself has a keen interest in its rapid implementation. While this can help ensure timely program (BISP) and TA project implementation, it may also result in the identification of the program with the President and governing party, weakening its institutionalization, and undermining its sustainability. H The Prime Minister has endorsed the adoption of the poverty scorecard as the country s new mechanism for targeting social safety nets. The project includes support for further developing the national social protection strategy to provide for a broad consensus building mechanism for policy making across political lines. The TA project provides technical assistance for setting up a safety net system rather than supporting just one program. The TA project would provide the necessary technical underpinnings (such as MIS, evaluations, spot checks etc.) to enhance transparency of the BISP. A rapid evaluation of the targeting accuracy of the poverty scorecard will provide valuable information about the quality of BISP program implementation and also make results comparable to other, existing safety net programs. H 17

30 Potential Risks The deteriorating security situation in Pakistan may affect the nationwide implementation of the BISP, including the roll out of the targeting survey. The general public might not be able to distinguish between the old targeting system and the new, thus sees Bank as involved in both. Overall Risk Risk Risk Mitigation Measures Risk Level Level after Mitigation H 0 BISP is contracting partner H organizations for the roll out process of the scorecard that have a strong presence at the local and district levels and therefore have access to households. H The TA project provides support for a S comprehensive public information campaign. The task team includes a strategic communications specialist. S 3. Loadcredit Conditions and Covenants Conditions of Disbursement (i) Two FM specialists to be assigned to the PIU hired prior to Disbursement. (ii) Finance and Accounting guidelines for project issued prior to Disbursement. (iii) Two Procurement specialists to be assigned to the PIU hired prior to Disbursement. Dated Covenants (i) By no later than six months after the Effective Date, BISP shall appoint an independent qualified third party, under terms of reference acceptable to the World Bank, to conduct regular spot checks/operational audits on the performance of the safety net program processes and reporting. (ii) By no later than sixty days after the Effective Date, a Memorandum of Understanding shall be signed between BISP and Planning Commission regarding implementation collaboration for component 4 of the project reached, including a work plan. (iii) By no later than one year after the Effective Date, BISP shall link BISP management information system with the beneficiary database of data processing agencies, to monitor the implementation of BISP program. (iv)by no later than three months after the Effective Date, BISP shall approve the Finance and Accounting Manual for the BISP in order to implement the safety net program. (v) By no later than eight months after the Effective Date, BISP shall enter into contracts with data processing agencies for designing and managing a database and a management information system for carrying out targeting of eligible beneficiaries of BISP. (vi)by no later than four months after the Effective Date, the BISP shall develop and thereafter maintain, a publicly accessible website, in terms satisfactory to the Association, showing the Project procurement plans, the status of procurement of various contracts (the summary of proposals/bid evaluations and awards) and a list of procurement complaints and status thereof. 18

31 APPRAISAL SUMMARY 4. Economic and Financial Analyses 50. Economic Analysis: The primary economic impact of the program will related to the direct effects on consumption and income levels of the beneficiaries. Given the emphasis transfers to the family units secondary impacts are also likely to be felt for children in the areas of education and health outcomes. The primary consumption effects of the program include: (9 (ii) (iii) Reduction in poverly: By receiving regular cash transfers households are expected to increase their food intake resulting in less food insecurity and the maintenance of productive assets. The transfer is expected to have a positive effect on consumption smoothing. The average transfer benefit from the program is substantial, reaching an equivalent of US$l50 per families per annum. It is estimated that up to 5 million families will receive US$150 worth of transfers during the first year, providing a total of US$750 million per year in cash to poor families. In addition to reducing income poverty, the cash transfer is also likely to reduce income inequality due to the redistributive effect of targeting the transfer to the poorest families within the households. Improvement in health and nutrition: Improvements in the health and nutrition outcomes of children and the wider family are expected due to the income effects of the transfer (e.g., increased food consumption, dietary diversity). In addition to the direct welfare benefits arising from improved health status, gains can also be expected through increased schooling levels, improved cognitive achievement as well as productivity, earnings and income for adults. Income generation: The program is also likely to increase income generation capacity among poor households, particularly by mitigating consumption expenses and negative coping strategies in response to shocks. Shock related expenses represent 54 percent of annual consumption for ultra-poor households, 27 percent for poor households and 18 percent for non-poor households. Households in rural areas spend 37 percent of their annual consumption coping with shocks compared to 30 percent for urban households in this group. 51. Fiscal Analysis: The fiscal impact of the BISP program is consistent with the fiscal costs of safety net programs in other countries with comparable levels of income and with other countries in the region. If the program is fully implemented, the Government is expected to spend Rs. 34 billion (US$425 million) for the current fiscal year 2008/09, doubling the federal Government s social safety net spending from 0.3 percent in 2003/04 to 0.6 percent of GDP. The budget for the FY 2009/10 is expected to double to reach around 5 million families increasing annual safety net expenditures to approximately 1 percent of GDP. This level of expenditure will align Pakistan s safety net spending with other countries at its level of income. 5. Technical 52. The Government of Pakistan decided in December 2008 to use a scorecard system to identify BISP beneficiaries. A scorecard is a targeting tool that includes a limited number of simple indicators that correlate well with poverty along with an intuitive scoring system that helps identify program eligibility. In addition, the province of Punjab is allocating about 0.2 percent of GDP to its income support/cash transfer program in 2008/09. A higher level of spending does not translate into more coverage for the poor if safety nets programs are poorly targeted. 19

32 The scorecard proposed for Pakistan s BISP has been developed with support by the Bank10 and is marked by two main characteristics - (i) Ease of implementation, and (ii) a Proxy Means Testing approach to improve targeting performance (for a more detailed description, see Annex 12). 6. Fiduciary Financial Management 53. The financial management arrangements for the project are being set up. The overall FM risk is rated High. Separate books of account will be maintained for the project on a cash basis, which will be subjected to the entity s overall system of internal controls including internal audit. The implementing entity has been established as a body corporate through an Ordinance. It is governed by a Management Board. Institutional arrangements are defined in the Ordinance and further elaborated in the Operations Manual. Disbursements will be made to a segregated designated account in US Dollars on the basis of cash forecasts provided in quarterly Interim Un-audited Financial Reports. Annual financial statements will be prepared in accordance with cash basis international public sector accounting standards and audited by the Supreme Audit Institution, Le., the Auditor General of Pakistan, acceptable to the Bank. 54. A Director of Finance and Accounts has recently been appointed. The position of Director General (DG) Finance & Accounts is vacant. There is also a vacancy in the Internal Audit unit. Two professionally qualified FM specialists will be hired for the Project. They need to be in place prior to Disbursements. Support staff will also need to be augmented once the entity is fully operational. Finance and accounting guidelines are under preparation for issuance within three months of the Project being declared effective. Once issued, these guidelines will provide details of processes and procedures for financial management of the implementing entity and include a separate chapter regarding financial management of the Project. Pending the finalization of the entity s guidelines, separate finance and accounting guidelines will be prepared for the financial management of the Project, which need to be issued prior to disbursement. Independence of the internal audit function has been ensured through a specific clause in the BISP Ordinance which provides for the internal auditor to report directly to the Management. The MIS for the entity is being developed and will include a module for accounting and financial reporting. Once these arrangements are in place, it is anticipated that the entity will be capable of implementing the project in accordance with the requirements of OP The implementing agency would ensure that Bank s guidelines on Preventing and Combating Fraud and Corruption in Bankfinanced Projects are followed in the project. Intensive supervision will be required to ensure that adequate fiduciary arrangements are implemented once the Project becomes effective. Procurement 55. The 2000 Pakistan Country Procurement Assessment Review (CPAR) highlighted a weak regulatory framework governing public procurement, and called for institutional and procedural reforms. CPAR revealed that Pakistan does not have a coherent law that establishes standards and an effective legal system that protects against collusion and corruption in the award of govement contracts. CPAR recommended the establishment of new and modern procurement rules conforming to international best practice. A Public Procurement Regulatory Authority (PPRA) was created in 2002 with powers to formulate public procurement legislation applicable to federal government line departments, state-owned enterprises, and semi-autonomous organizations. New public procurement rules were notified in 2004, and they depart from the old system in two important ways: they require posting of all bid notices on the lo See Vishwanath, Hou and Yoshida (2009) in Poverty Scorecard for Pakistan: A Recommended Approach for Targeting the Poor See Grosh and Baker (1995) for more details on the Proxy Means Testing approach. 20

33 PPRA s website, and prohibit post-bid negotiations. Moving forward, the development of monitoring and reporting mechanisms, and the creation of an autonomous second-tier appeals procedure to handle complaints in a timely and transparent manner will be critical for effective implementation of PPRA rules. PRA rules are widely recognized as the rules to follow for procurement, and the Auditor General s Office conducted the first set of audits using the new procurement rules in the previous fiscal year. As part of the process of rolling out the rules, the PPRA has trained more that procurement staff of Federal government and its autonomous bodies, and over 220 organizations now post their procurement notices on the PPAR website. As the next logical step, the government has now directed its ministries and autonomous bodies to commence posting of notices of awards for contracts over Rs.50 million on the PPRA website. 56. The BISP is a new organization, and although subject to the 2004 Public Procurement Rules, does not have a history of carrying out any procurement on its own. As a result of the assessment, additional actions have been agreed, which include the appointment of a dedicated procurement specialist, a set of training measures for the procurement staff, use of NCB documents agreed with the Bank and additional stipulations for procurement under NCB. These steps are discussed in more detail in Annex 8 to this PAD. The overall procurement risk is rated as high before mitigation and substantial after mitigation measures. BISP is also developing a procurement handbook based on the 2004 Public Procurement Rules. 7. Social 57. Socio-economic vulnerabilities at the household level strongly influenced the use of a cash transfers intervention as the basis for BISP. A survey of safety net recipients/applicant households (PSNS 11, 2005) finds that nearly two thirds of surveyed households-of which 80 percent are poor-suffered one or more severe shocks during the three years prior to the survey (for more detail, see annex 10). The same survey finds that the frequency, nature and impact of shocks vary with socio-economic characteristics and areas of residence, e.g., health-related shocks disproportionally affect poorer households, urban households are more likely to face health and economic shocks, rural households are prone to natural disasters. However, a common thread emerging in all cases is the potentially harmful welfare impacts of household risk coping strategies, e.g., depletion of savings, increase in household labor, food intake substitution. Shock-related expenses represent 54 percent of annual consumption for ultra-poor households, 27 percent for poor households and 18 percent for non-poor households. Households in rural areas spend 37 percent of their annual consumption coping with shocks compared to 30 percent for urban households in this group. In this context, there was a strong imperative to develop a cash transfer system to assist the ultra poor. 58. Social exclusion driven by gender, language or ethnic diversity has also been considered in the design and implementation modalities for the project. A country gender assessment (World Bank, 2005) highlighted the need for a series of policy interventions focused specifically at gender inequality and covering a range of areas including access to health, education and the labor market. The provision of the cash transfer to the adult women in the households marks a significant step towards the empowerment of women, and in turn an entry point to address the needs of vulnerable children who represent just over half of those vulnerable to chronic poverty. 59. Ethnic and linguistic diversity are additional concerns. Although Pakistan s official language is Urdu, a host of other languages - including Baluchi, English, Pahto, Punjabi, Saraiki and Sindhi - reflect the ethnic diversity of the population. In particular this reinforces the need to ensure outreach, targeting and grievance mechanisms are accessible, including for example the translation of poverty scorecard materials where appropriate. 21

34 60. The benefits of the Benazir Income Support Programme go exclusively to women. Since the program is implemented nationwide, it will reach out to benefit the poor segments of the various ethnic minority populations of the country. Ethnic and cultural considerations will be mainstreamed into the program design to ensure equal and culturally appropriate accessibility for ethnic minority populations. 0 Under Component One, ethnic sensitivities and cultural considerations will be taken into account in improving the targeting design, including updating of the targeting tools (e.g., manuals, materials, database), training of administering staff and consultants. 0 Under Component Two, the public communication strategy, grievance redress and social accountability mechanism will be improved with specific attention to build in special mechanisms and provisions for operating with ethnic minority population groups and in tribal areas. Under Component Three on the BISP Program, the BISP M&E system design will include ethnic and cultural considerations in its various components, including baseline and follow-up survey design, and beneficiary assessment. 0 Component Four will strengthen social protection policy and institutional framework. It will build in ethnic and cultural considerations in its activities, including in the design of a NSPS monitoring system and various capacity building efforts for policy makers and implementers. In order to analyze its effects along various social dimensions it has been agreed to include in the project a number of studies to effectively monitor the social effects of BISP, particularly on gender and ethnic minority groups. This work will be led by a social scientist and guided by the respective Bank colleagues. 8. Environment 61. The project has been classified as a category Cy project in terms of environmental safeguards since the BISP program is a pure cash transfer program and is not expected to have any direct environmental effects. The project only includes minor works of refurbishing and renovating existing offices and does not finance any new construction. 9. Safeguard Policies Safeguard Policies Triggered by the Project Yes N O Environmental Assessment (OP/BP 4.0 1) [XI Natural Habitats (OP/BP 4.04) Pest Management COP 4.09) Physical Cultural Resources (OP/BP 4.1 1) Involuntary Resettlement (OP/BP 4.12) Indigenous Peoples (OP/BP 4.10) Forests (OP/BP 4.36) Safety of Dams cop/bp 4.37) Projects in Disputed Areas (OP/BP 7.60) Projects on International Waterways (OP/BP 7.50) [XI [XI [XI [XI Piloting the Use of Borrower Systems to Address Environmental and Social Safeguard Issues in Bank-Supported Projects (OP/BP 4.00) [XI 10. Policy Exceptions and Readiness 62. The project complies with all applicable Bank policies. By supporting the proposedproject, the Bank does not intend to prejudice thejnal determination of the parties claims on the disputed areas 22

35 I. Introduction Annex 1: Country and Sector Issues PAKISTAN: Social Safety Net Technical Assistance Project 1. This annex presents information on the country and the social sector contexts in which the proposed project will be implemented. The discussion on the country context touches upon both macroeconomic and poverty issues, while the sector discussion focuses on the country s safety net. The picture that arises from this discussion is one where despite a decline in the poverty rate in this decade, inequality has increased over this period. Improvements among those at the bottom of the expenditure distribution have been modest while social assistance to this group has been weak. The introduction of the new Benazir Income Support Programme (BISP) by President Zardari s administration seeks to reverse this trend. 11. Country Context Macroeconomic Context 2. Pakistan s development record was strong during the first part of this decade. The economy grew at 7.3 percent on average per year during 2003/04 through , driven by solid performances in the services and industrial sectors. The manufacturing sector grew fast during the earlier part of the period, but slowed down considerably later in the period due to capacity and infrastructural constraints. At the same time, the services sector, catering primarily to the domestic demand, grew steadily throughout the period. The factors that contributed to growth included a benign external environment, availability of external financing, political and macroeconomic stability. Key measures include renegotiation of external debt with Paris Club creditors, pre-payment of extensive debt, greater reliance on concessional borrowing for new loans, and liberalization of the financial sector. They, together with an expansion in the economy, led to a decline in public debt as well as a reduction in debt servicing cost. Growth translated into rising household incomes, with per capita income growing to US$878 in 2006/07, an 18.3 percent increase fiom US$742 two years earlier. 3. In 2007/08, the sharp rise in international oil and food (specifically wheat) prices, in combination with policy inaction and internal political turmoil, led to rapidly expanding macroeconomic imbalances in Pakistan. Both fiscal and current account balances widened significantly. In the absence of adequate remedial policy measures to address the imbalances-in particular not passing on the international price increases to domestic consumers-the economy begun to adjust through a slowdown in growth and rising inflation. 23

36 Poverty headcount rate in Pakistan, 1987/ /06 (Percent of population) s I- ~~ ~ ~~ I Source: Pakistan Economic Surveys & Poverty and Inequality Trends 4. Recent years of relatively high levels of economic growth led to a decline in poverty in Pakistan. Although poverty increased over the to period, this trend was reversed in subsequent years (see figure above). Irrespective of the methodology deployed, poverty dropped significantly between 2001/02 and 2005/06, from 34.5 percent of the population to 22.3 percent of the population, equivalent to over 17 million people moving out of poverty. Even the more conservative estimate suggests that the national poverty rate fell by close to six percentage points. * The effect of growth on poverty reduction between 2001/02 and 2004/05 was dampened by increasing inequality over the same period. Though lower than in many regions of the world, the Gini coefficient increased from to over the period Aside from low levels of income, Pakistani households remain vulnerable to income shocks, both economic crisis, such as the recent global food, fuel and financial crisis; and natural disasters, such earthquakes and droughts, and from idiosyncratic shocks such as health and unemployment. These shocks have the potential to reverse recent gains in poverty reduction. A recent simulation of the welfare impact of the fuel price found that poverty rates may have increased between 2.3 and 3.6 percent, depending on the growth assumption used.13 Earlier studies have also documented the negative impact of shocks on household welfare. The clustering of a large share of the population around the poverty line means that small changes in income can translate into large changes in poverty. For example, between and , the drought in Sindh caused a sizable movement of the population in the middle part of the consumption distribution to its lower end. The Pakistan social protection report14 also provides some l2 Poverty estimates depend on how the poverty lines are updated. The official poverty rates are estimated based on poverty lines updated with CPI-based inflation rates. According to the official estimates, the overall poverty declined by more than 12 percentage points, from 34.5 to 22.3 percent between 2001/02 and Poverty in urban areas declined from 22.7 to 13.1 percent and in rural areas from 39.3 to 27 percent. If a survey based price index (SBPI) is used to update the poverty line since , the overall poverty declined by a little less than 6 percentage points, from 34.5 to 28.8 percent over the same period. Both CPI and SBPI have pros and cons. CPI is constructed with price data from major cities only; thereby it might not properly reflect inflation in rural areas where poverty is the greatest. By contrast, the survey-based price index collect price data in both urban and rural areas, but, it suffers from the limited coverage of itemsxovering only food and transport prices, which may be unrepresentative of the items households purchase. Vishwanath et. al (2009). l4 World Bank (2007), ibid. These results presented in the report are from a representative sample of safety net recipientsiapplicants collected in 2005, and are therefore not nationally representative. The sample is predominantly poor. Over 24

37 insights into the specific shocks that affect households and household risk coping strategies. The report found that nearly two-thirds of safety net applicants and recipients (about 80 percent of whom were poor) surveyed suffered from one or more major shocks in three years before the survey, comprising both idiosyncratic (health, unemployment) and aggregate shocks (economic and natural disasters). While income shocks imposed major costs on all affected households, the poor were the hardest hit (cost of the shock as a share of annual household consumption for the poor was twice that for non-poor households). Nearly 33 percent of these households lowered their food intake, 10 percent put a child to work, and 8 percent pulled a child out of school in response to the shock. While these households also received private assistance, its concentration in time and geographic space made it difficult to use for smoothing consumption against aggregate shocks. These results are consistent with findings on risk coping strategies from other countries in the region, and reveal that income shocks can both impoverish households at the time of the shock and, by adversely impacting human development, can also perpetuate poverty across generations Overview of Social Protection in Pakistan 6. Pakistan s social protection programs comprise safety nets, social security, employment promotion and protection. The country s safety net system includes two main federal cash transfer programs (Zakat and the Food Support Program (FSP)) and small, scattered programs that provide social welfare and care services to persons with disabilities, child laborers, and others. Although Pakistan has earlier implemented public workdworkfare programs, no large workfare program is currently in place. To address aggregate economic (price) shocks, till recently, Pakistan relied primarily on a wheat subsidy program. Although no permanent program is in place to help individuals cope with aggregate disasters, Pakistan has used a combination of cash transfers, housing and social care services programs to help those affected by the 2005 earthquake. The country also implements several employment promotion programs, e.g., micro credit programs and skills and training. Employment protection institutions (e.g., labor laws) cover workers in the formal sector. Pakistan s sociaz security system offers pension (old age, survivor and disability) benefits to formal sector workers. Public sector workers are provided civil service pensions, while private sector workers have access to pensions from the Employees Old Age Benefits (EOBI), but also provincially based pension and non-pension programs such as the Workers Welfare Fund (WWF) and the Employees Social Security Institutions (ESSI). 7. During the last four years, the overall social protection expenditures have fluctuated as a share of total PRSP/pro-poor expenditures. Specifically, expenditures pertaining to social protection (food subsidies, Bait-ul-Mal, Tawana school meals, social security, and low cost housing) remained between 3-4 percent of total PRSP spending between 2004/05 till 2006/07 but then sharply increased to 13.4 percent of total PRSP spending in 2007/08 (see table below). This increase in social protection expenditures is largely due to the increased spending on food subsidies and social security, both of which are likely not well targeted to the poor. half of the sampled households (54 percent) are ultra poor (those with consumption below the food poverty line), about a quarter (23 percent) are poor (between the food and national poverty lines), while the remaining were non-poor. 25

38 PRSP Budgetary Expenditures Source: PRSP-2, Finance Division, Government of Pakistan (2009) 8. The share of food subsidies in PRSP pro-poor social protection (SP) spending has increased from 51 percent in to 68.4 percent in 2007/08. In comparison, the expenditures on the Food Support Program steadily decreased from 25.8 percent of SP-PRSP expenditures in to 5.4 percent in 2007/08. Since 75 percent of Pakistan s social protection spending (excluding food subsidies) is devoted to social security spending (likely received mainly by the non-poor), these expenditures largely accrue to the non-poor. N. Overview of Safety Net Programs 9. The country s safety net sysrem includes three main federal cash transfer programs (Zakat, the Food Support Program (FSP)), and the recently established (2008) Benazir Income Support Programme (BISP), and small, scattered programs that provide social welfare and care services to persons with disabilities, child laborers, and others. Although Pakistan has earlier implemented public workdworkfare programs, no large workfare program is currently in place. To address aggregate economic (price) shocks, till recently, Pakistan implemented a wheat subsidy program. Although no permanent program is in place to help individuals cope with aggregate disasters, Pakistan has used a combination of cash transfers, housing and social care services programs to help those affected by the 2005 earthquake. The country also implements several employment promotion programs, e.g., micro credit programs and skills and training. Employment protection institutions (e.g., labor laws) cover workers in the formal sector. Pakistan s social security system offers pension (old age, survivor and disability) benefits to formal sector workers. Public sector workers are provided civil service pensions, while private sector workers have access to pensions from the Employees Old Age Benefits (EOBI), but also provincially based pension and non-pension programs such as the Workers Welfare Fund (WWF) and the Employees Social Security Institutions (ESSI). The following table provides an overview of Pakistan s expenditures on and beneficiaries of social protection programs in 2003/ * Social Protection in Pakistan: Managing Household Risks and Vulnerability World Bank Report No PK, October

39 Expenditures Share of Share of total Source of Number of GDP SP budget funding beneficiaries (Rs. billion) (YO) (Yo) support and basic (guzara: 0.8 mn) SOCIAL SECURITY Sources: Issues and Policies Consultants (2004); World Bank (2006a); and information from program managers. N.A. Not available A Number of beneficiaries of recurrent cash and non-cash benefits. Number of workers covered by insurance scheme. Budget of Bait-&Mal has since been increased to Rs. 4.5 billion (0.08% of GDP) and the number of beneficiaries of FSP, its lar est program, to 1.45 mn. Based on revised estimate of GDP (at current market prices) for 2003/04 of Rs. f 5,641 mn. Including military pensions would raise spending to Rs. 60 billion (1.07% of GDP), and beneficiaries to 1.9 mn retirees and 2.9 mn active workers. Total SP spending including military would then be around 1.4% of GDP. (Spending on Zakat and Bait-&Mal constitutes 0.14 percent of GDP in 2004/05) 10. Until the introduction of the BISP (see description in section VI1 below) in 2008, the two main cash transfer programs in Pakistan were Zakat and the Food Support Program (FSP) of Pakistan Bait-ul-Mal (PBM). Although similar in terms of their overall objective and target population, these programs have different histories and are operated and funded in substantially different ways. Zakat is managed by the Ministry of Religious Affairs and implemented by local committees with financing from individual contributions. In contrast, the FSP is administered by the Ministry of Social Welfare and Special Education (MoSWSE), through PBM, and is funded through public resources. 27

40 11. In addition to these two main programs, Pakistan has an array of smaller, scattered programs that provide social welfare and care services to persons with disabilities, child laborers, and others. Micro finance programs aim to give the poor access to credit. Although Pakistan has earlier implemented public works/workfare programs, no large workfare program is currently in place. To address aggregate economic (price) shocks, Pakistan implements a wheat subsidy program. Furthermore, Pakistan has used a combination of cash transfers, housing and social care services programs to help those families affected by the 2005 earthquake. Pakistan s social security system offers pension (old age, survivor and disability) benefits to formal sector workers. Public sector workers are provided civil service pensions, while private sector workers have access to pensions from the Employees Old Age Benefits (EOBI), but also provincially based pension and non-pension programs such as the Workers Welfare Fund (WWF) and the Employees Social Security Institutions (ESSI). Programs for Exiting Poverty 12. Pakistan is implementing several pilot safety net programs that aim to improve human development outcomes for poor children, thereby eliminating the inter-generational link to poverty. Globally, there is a trend for cash transfer programs to include incentives to help poor children acquire human capital and avoid inter-generational poverty. These conditional cash transfers have successfully increased enrollment, improved health outcomes and reduced child labor1 6. Pakistan is also experimenting with these programs: (i) provincial stipend programs, including the Punjab Secondary School Stipend Program targeted to girls; (ii) Bait-ul-Mal s Child Support Program, a pilot conditional cash transfer program targeted to the poorest households; (iii) a school meals program (Tawana) that aims to improve nutrition and education of poor girls in rural areas; and (iv) a cash transfer program that targets improvement of health status (TB) alone. These programs all have program evaluations in place that will allow the government to gauge how programs impact educational and health outcomes of target groups. 13. For the working poor, microfinance programs provide opportunities to the poor to exit poverty, but program coverage in Pakistan is amongst the lowest in the region. Microfinance is offered through the banking sector (Khushali Bank) and NGOs (such as Rural Support Programs -RSPs). Pakistan has the lowest penetration of microfinance in South Asia with less than 2 percent of poor households covered. In Bangladesh and Sri Lanka, over 60 percent of the poor are microfinance clients and this share is 9 and 14 percent in India and Nepal, respectively. However, the impact of microfinance on those who receive them is positive. The slow growth of microfinance in Pakistan is caused in part by the lack of a solid and profitable core business, relying on subsidized donor finds rather than profitability (Rasmussen, 2005). 14. In the past, Pakistan has implemented workfare programs to build infrastructure and offer temporary employment to the poor, but no major program currently exists. Past programs include workfare for Afghan refugees and the Khushal Pakistan Program, though their overall impact on poverty reduction is not known. Currently, no public works program exists to target the poor through low wages or conditions on public works contracts to employ local labor. Programs for Coping with Aggregate Shocks 15. The government regularly intervenes in the wheat market through a subsidy with the understanding that wheat is a major staple for Pakistani households, and therefore changes in wheat prices directly impact welfare, particularly of poor consumers and producers, but at significant economic and fiscal costs. Since resell prices are not high enough to cover payments to producers and storage and handling l6 See Chapter 3. Evidence on health outcomes is mainly from the evaluation of Mexico s Progress& Opportunidades program. 28

41 costs, the government incurs significant fiscal expenses. Over the last 10 years, annual wheat subsidy costs have averaged around Rs. 6 billion, with the subsidy fluctuating from 0.1 to 0.5 percent of GDP. About a third of the subsidy accrues to millers to cover the difference between the guaranteed purchase and resale prices, and the other two-thirds finance inefficient government grain procurement and handling procedures. Aside from creating distortions in wheat prices, and although more in-depth analysis is needed, preliminary evidence seems to suggest that the current system favors millers and the non-poor (Dorosh, 2004). The ability of the program to help the poor cope with wheat price shocks is therefore small. 16. The recent earthquake exposed the gap in the ability of the government to cope with natural disasters. Since no safety net structure existed that could be rolled out quickly, interventions had to be designed from scratch. The final relief package combined short-term income support (cash support) with long-term aid for reconstruction (housing). Community-based rehabilitation services anchored in four resource and information centers for persons with disabilities in the earthquake-affected areas are planned to complement this program. Following the earthquake, the government has developed the National Strategy and Plan of Action for Vulnerable Populations in Earthquake Affected Areas and the Earthquake Reconstruction & Rehabilitation Authority has approved its draft social protection strategy. V. Challenges of the Existing Social Safety Net 17. As Expenditures: Prior to the institution of the new safety net program, expenditures on safety net programs were low relative to neighboring countries and declining over time in Pakistan.17 Safety net spending on the two main safety net programs (Zakat and Bait-ul-Mal) declined sharply, from 0.4 percent of GDP in 1991/92 to less than half of this level, or 0.14 percent of GDP in 2004/05. Contributions to Zakat have declined significantly (from 0.3 percent of GDP in the 1980s to 0.08 percent in 2002/03, and 0.04 percent in 2006/07) as more people have opted out from paying contributions by either withdrawing funds before the day that Zakat is levied or by filing a declaration of exemption, leading to concomitant declines in expenditures on Zakat programs. Fiscal allocations for Bait-ul-Mal have increased in recent years, from 0.02 percent in 2002/03 to roughly of 0.04 GDP in 2006/07, but not enough to compensate for the decline in Zakat expenditures. In 2006/07, the PBM Food Support Program and regular Zakat programs together covered 3.2 million households corresponding to about 13 percent of population in a country where poverty estimates range from 22.3 percent to 28.8 percent depending on which price index is used to take into account inflation. 18. Benefit adequacy and timing: Zakat guzara transfers represent 14 percent of average recipient household income while PBM s Food Support benefits are equivalent to 11 percent of household income among the ultra-poor, and to 8 and 5 percent among the poor and the non-poor respectively. In comparison, conditional cash transfer programs in Mexico and Nicaragua provide benefits that amount to 2 1 percent of average household consumption. Furthermore, the stipulated amounts are often not received by households. For example, the average Zakat guzara cash transfer beneficiary during the 12 months before the survey was equivalent to 5-6 monthly installments, rather than the stipulated 12, and payments occurred in 6-8 months intervals. Given the small size of program benefits and non-poor targeting in Pakistan, poverty and inequality decline only slightly in response to transfers. 19. Benefit Incidence: Although both Zakat and Bait-ul-Mal display elements of pro-poor targeting, they could still perform better. The current fiscal allocation to cash transfer programs could be more efficient in covering the poor if program expenditures were better targeted. A sizeable share of benefits reaches the non-poor while many of the rejected applicants are poor. Nearly 46 percent of total benefit Safety net spending (most recent estimates; as a share of GDP) is approximately 1.6 percent in Bangladesh, over 2 percent in India and 1 percent in Bangladesh. The level of social protection expenditures is not an indicator of program efficiency in targeting the poor, however. 29

42 expenditures of Bait-&Mal reach the poorest 40 percent of the population, while 43 percent of total Zakat expenditures reach the same population group. This performance could be improved. For example, 80 percent of expenditures on conditional cash transfers reach the bottom 40 percent of the population in Honduras; and 62 percent of total expenditures reach this group in Mexico. 20. Governance: Weak governance and lack of transparent eligibility conditions reduce targeting effectiveness. The main reasons for poor targeting include (i) the lack of an objective targeting instrument: Both Zakat and Bait-ul-Mal target the deserving needy and poor, but no objective targeting tool is used; (ii) the lack of an operational definition of poverty or targeting mechanism, that leaves eligibility decisions in the hands of the chairmen and members of the local Zakat committees or of local authorities in the case of Bait-ul-Mal; and (iii) systematic differences in program access and eligibility which favor the non-poor-programs tend to be located in better off localities and benefits to increase with household income. 21. GraduatiodExit: Currently, safety net program recipients are not linked to income earning opportunities or access to education and health services. Globally, there is a trend for cash transfer programs to include incentives to help children in beneficiary households acquire human capital and avoid inter-generational poverty. These conditional cash transfers have successfully increased enrollment, improved health outcomes and reduced child labor. Pakistan is also piloting and evaluating similar programs: (i) provincial stipend programs, including the Punjab Secondary School Stipend Program targeted to girls; (ii) Bait-ul-Mal s Child Support Program, a pilot conditional cash transfer program targeted to the poorest households; (iii) a cash transfer program that targets improvement of health status (TB) alone. Microfinance programs also provide opportunities to the poor to exit poverty. Even though Pakistan had the lowest penetration of microfinance in South Asia with less than 2 percent of poor households covered (vs. over 60 percent of the poor in Bangladesh and Sri Lanka) in early 2000s, microcredit disbursements have increased consistently from Rs 1 billion in FY02 to Rs 9.9 billion in FY07. During the same period, number of beneficiaries increased from 100 thousand in FY02 to 906 thousand in FY Coping with Shocks: Safety net programs also are not able to expand to cope with shocks. The recent earthquake exposed the gap in the ability of the government to cope with natural disasters. Since no safety net structure existed that could be rolled out quickly, interventions had to be designed from scratch. The final relief package combined short-term income support (cash support) with long-term aid for reconstruction (housing). Community-based rehabilitation services housed in four resource and information centers for persons with disabilities in the earthquake-affected areas are planned to complement this program. Pakistan has also implemented workfare programs to build infrastructure and offer temporary employment to the poor, but no major program currently exists. Past programs include workfare for Afghan refugees and the Khushal Pakistan Program, though their overall impact on poverty reduction is not known. Currently, no public works program exists to target the poor through low wages or conditions on public works contracts to employ local labor. VI. The Government s Social Protection Strategy 23. In order to strengthen the social protection sector, the Government of Pakistan, with support from the World Bank and other donors, started a collaborative process in 2006 to develop A Social Protection Strategy to Reach the Poor and the Vulnerable, which was approved by the Cabinet in June The Social Protection Strategy is conceived within the overall framework of Just and Balanced Development envisaged in the Medium Term Development Framework ( ) and Vision The Vision of the National Social Protection Strategy is to develop an integrated and comprehensive social protection system, covering all the population, but especially the poorest and the most vulnerable. 30

43 The goals of the strategy are: (i) to support chronically poor households and protect them against destitution, food insecurity, exploitation, and social exclusion; (ii) to protect poor and vulnerable households from the impacts of adverse shocks to their consumption and wellbeing that, if not mitigated, would push non-poor households into poverty, and poor households into deeper poverty; and (iii) to promote investment in human and physical assets, including health, nutrition, and education, by poor households capable of ensuring their resilience in the medium run and of interrupting the intergenerational cycle of poverty. 25. Priority areas of intervention will ensure these goals are met. These were identified as: (i) reaching the poorest; (ii) addressing the needs of vulnerable children; (iii) addressing the needs of vulnerable and poor women; (iv) providing employment and income earning opportunities; (v) protection against environmental and natural disaster; (vi) improving access to social care services and shelter; (vii) protection against health shocks; and (viii) protection for workers in the formal and informal labor markets. 26. The core instruments proposed by the strategy include: (i) expanding the coverage of cash transfers using conditional cash transfers (CCTs) supplemented with unconditional transfers (through the Food Support Program (FSP) and Zakat); (ii) a new public works program based on low-wage employment; (iii) programs against child labor, and various new pilots such as (a) combination of cash transfers and basic skills development aimed at graduating the poor into micro finance clients, and (b) programs for bonded labor; and (iv) scaling up school feeding and social care services. VII. The Cornerstone of the strengthened Social Safety Net System: The Benazir Income Support Programme 27. Based on deficiencies highlighted in the National Social Protection Strategy, and with the objective to cushion the severe food, fuel and financial crisis emerging in 2007/08, the Government decided to create a new cash transfer program, the Benazir Income Support Programme. During the second half of 2008, the Government of Pakistan launched the Benazir Income Support Programme (BISP) as its new main social safety net program to cushion the negative effects of the food crisis and inflation on the poor. The Program provides for cash transfers of Rs.1000 [$12] per month to the eligible families. The budget allocated for BISP for the current fiscal year is Rs.34 billion (about $420 million), which is intended to cover about 3.4 million families or 13.6 million people. In order to quickly launch the program, parliamentarians (Members of the National Assembly and Senators) were asked to identify 8,000 beneficiary households each on a prescribed form which collected information on names, national ID card, and household income. Upon receipt of the completed forms, the National Database Registration Authority (NADRA) will apply the selectiodverification filters to finalize the beneficiary lists. Payments were then made via the post offices and delivered at the doorstep through money orders to the adult female in the household. A list of the first 127,000 beneficiaries has been provided to the Pakistan Postal Services in January 2009 and these payments are presently being processed. By April 2009, the program was paying approximately 1.6 million families based on more than 2.4 million applications received. 28. Setting up a national targeting system through BISP. In December 2008, GoP decided to improve the existing targeting mechanism of BISP. More specifically, the Government of Pakistan adopted the use of a poverty scorecard that will be used for the selection of beneficiaries. Roll-out of the scorecard nationwide is an extensive task. A decision has therefore been made to start the scorecard roll out in February 2009 with a test phase of up to 15 districts. This test phase will provide useful insights and the necessary learning for scaling up the program nationwide since GoP intends to expand the budget for the program in the coming fiscal year 2009/10 to cover around 5 million families with the ultimate goal over the medium term to expand coverage to 7 million families in BISP is a national program that covers all provinces of the country, AJK and Northern Areas. 31

44 Annex 2: Major Related Projects Financed by the Bank and/or other Agencies PAKISTAN: Social Safety Net Technical Assistance Project Bank Project Second Poverty Reduction Support Credit (Cr.4302-PAK) Public Sector Capacity Building Project (Cr PAK) PPAF-I1 Suppl. Social Mobilization (Cr PAK) PK Earthquake Disability (Grant No PAK) Additional financing for Rehabilitating Earthquake Affected Communities-11: Pakistan (Cr PAK) Pakistan Earthquake ERC (Cr PAK) Second Poverty Alleviation Fund Project (Cr PAK) NWFP Community Infrastructure Project -11 (Cr and 3906-PAK) Status Approved by the Board in May 2007: includes support for safety net and social protection program. Approved in May Bank financed TA, which includes components and policy making and implementation. Approved in October Follow-on of the Pakistan Poverty Alleviation Fund Project. The development objective remains: to reduce the incidence of poverty in the country through provision of resources and services to the poor and low income, particularly women. Approved in May To expand the coverage, use and quality of social care and rehabilitation services for people with disabilities in 34 Union Councils affected by the October 8, 2005 earthquake, by ensuring better mobility, improved physical and mental health, increased participation in social and economic life, and strengthened empowerment. Approved in February To complete the housing reconstruction in areas affected by earthquake Approval December The objective is to assist the Recipient in carrying out its emergency recovery and reconstruction program following the massive destruction in the earthquake area. Approved by the Bank December Follow on project to reduce the incidence of poverty in the country through provision of resources and services to the poor and low income, particularly women. Approved by the Bank May The main objective of the project is to promote the well being of low income communities of NWFP (a) supporting investments in community development activities, basic services and small scale. Amount 1 DO/ IP (US$m) Rating 350 Not available T Not available 5.O Mu S S S 32

45 World Bank Financed Analytical Work Social Protection in Pakistan: Managing Household Risks and Vulnerability, 2007, The World Bank. Poverty and Social Impact Assessment: 2009, The World Bank Poverty Assessment, 2002, The World Bank. For Protection and Promotion: The Design and Implementation of Effective Safety Nets, 2008, The World Bank. Child Support Program Rapid Assessment Report, 2009, The World Bank. Child Support Program Baseline Survey Report, 2009, The World Bank. Pakistan Country Gender Assessment: Bridging the Gender Gap: Opportunities and Challenges, 2005, The World Bank. Pakistan Public Expenditure Management Report, 2004, The World Bank. Pakistan CFAA PEPFM 2008 Draft. and Federal PEFA PFM Draft F Sector/ Project/ Amount PAK/ ASIE/2006/ : Community Based Livelihoods Recovery Programme for Earthquake Affected Areas of AJK and NWFP PAKl Sustainable livelihood initiatives in disaster prone areas of Pakistan PAK/ Institutional Support for Poverty Reduction Project (1DPR)- A ) ojects Finance1 Agency European Commission and United Nations Development Programme European Commission and Oxfam, Great Britain Canadian International Development Agency by other Developmeni Project/Activity/ Objective and Status To restore the socioeconomic fabric in the earthquake affected areas by reinstating basic socio-economic infrastructure and livelihoods. The project aims to reduce poverty, vulnerability & gender imbalance & improve the livelihoods of disasteraffected communities, particularly women, by creating the critical links between relief, rehablitation & sustainable development continuum at policy & practice level. The project aims to achieve equitable and sustainable improvements in the livelihoods of people in Pakistan s Northern Areas and Chitral by increasing their human and institutional capacities, expanding?artners Location North-West Frontier Province and Azad Jammu and Kashmir Balochistan, Sindh and Punjab North-West Frontier Province and Federally Administrated Northern Areas End Date 4/15/ /01 / /31/

46 Sector/ Project/ Amount PAKl Improved child health FATA PAK/OO Gender Support Progra m (GSP)-ATLAS ID PAKl Northern Pakistan Education Programme Agency United States Agency for International Development Government of Balochistan; CIDA; Commonwealth Judicial Education Institute; DFID; Norway; Swiss Agency for Development and Cooperation; UNDP European Commission ProjectIActivityi Objective and Status and diversifying economic opportunities, and creating an enabling environment in the program area and beyond. Improved child health in FATA is delivering a health package for FATA's under-five infants covering immunization and treatment for respiratory infection diarrhea newborn care and nutrition. Women's Political Participation; Economic Empowerment; Enabling Social Environment; Training and Capacity Building - Gender and Women Development; Gender Budgeting; Gender-based Initiatives To improve the access, quality and sustainability of education for inschool and out of school children with increased gender equity and the participation of communities in Northern Areas and Chitral Location Federally Administrated Tribal Areas Country Wide Federally Administrated Northern Areas End Date /20/ /08/

47 Annex 3: Results Framework and Monitoring PAKISTAN: Social Safety Net Technical Assistance Project Results Framework Project Development Objective Enhance the operation and management of a nationwide effective and transparent safety net system for the poor in Pakistan Project Outcome Indicators Percentage of disbursed BISP cash transfers received by beneficiaries in quintiles 1 and 2 (40 percent poorer households) Use of Project Outcome Information Assess quality of targeting Monitor coverage of targeting system Measurement for accountability of BISP at the local level To inform program management about program bottlenecks as perceived by beneficiaries. Use for policy dialogue about responsiveness of safety net for vulnerable groups. National Targeting System based on the Dovertv scorecard. Component 2. Strengthening Safety Net Operations. Component 3. Enhancing Safety Net Program Management, Accountability and Evaluation. Component 4: Developing the Social Protection policy and strategy monitoring targeting process Number of female beneficiaries that receive payments (millions) Percentage of grievance redressal claims settled within three months of application Percentage of BISP positions that are filled with qualified staff; Degree of dissemination of evaluation results; Number of pilots implemented for exit strategies; Regular operational audits and follow up actions on the findings; 0 Revised NSPS approved by Cabinet. Cabinet adopts consolidation of selected federal Social Safety Net Programs under BISP Use of Intermediate Outcome Monitoring Assess regional and local coverage Assess family coverage and female beneficiaries Assess operational functioning and responsiveness. Assess institutional capacity Assess implementation of accountability measures. Assess progress towards graduation Measure outcome of policy dialogue 35

48 2 M s 5 m " E M e a - m C 4 - m C 4 - m C 4 'E m? VI 0 0 \o 0 I- 0 \o VI 0 VI m W 0 0 m 0 m 0 VI N 0 N 9 d 0 0 m. d m..- C

49

50 Evaluation Strategy of BISP 1. The evaluation strategy of BISP aims to evaluate the impacts of the cash transfers on the socioeconomic conditions of the beneficiary families such as food and non-food consumption, and potentially on a number of human development indicators such as school enrollment and attendance, health outcomes, and child labor. Given that the cash transfer is provided to the female head of the beneficiary families, the evaluation also aims to assess the impact of BISP on women s perceived and actual social status. In addition, the evaluation strategy aims to assess the targeting process, including the poverty scorecard performance and the effectiveness of targeting centers and door-to-door approach. 2. There are two phases of evaluation- the first phase focuses on evaluation of the test phase, mainly on the implementation process, targeting effectiveness and the spot checks. The second phase focuses on the impact evaluation of BISP through a combination of quantitative and qualitative instruments. Phase I: Evaluation of the Test Phase (A) Process Evaluation of the targeting test phase 3. This process assessment emphasizes to evaluate the effectiveness and implementation quality of the operational processes with three main objectives: (i) to evaluate the consistency of activities of the poverty scorecard test phase with what has been specified in the targeting manual; (ii) to evaluate the efficiency and transparency of the payment process, and (iii) to evaluate whether the data entry, data quality verification, data formatting and report generation in accordance with the targeting guidelines. The main methods for this evaluation include direct observation, individual interviews with key stakeholders, and focus group interviews. The selection of the districts that are to be included in the assessment will depend on the roll-out process. (B) Assessment of the scorecard effectiveness 4. The development of the poverty scorecard is based on the simulation using the most recent household survey data (PSLM ), which unavoidably bears statistical errors. The main purpose of this evaluation component is to assess the effectiveness of the scorecard by comparing the relative score distribution versus the actual consumption distribution. That is, for a sample of households, both the scorecard data and actual consumption data are to be collected. The relative score distribution against the national representative scorecard distribution is to be compared with the relative consumption distribution against the national representative consumption distribution. If there is an insignificant difference (both in magnitude and statistics) in the relative distribution of the scores and the consumptions, the scorecard is proved to be an effective instrument. 5. There are a couple of data sources that will be used for this assessment. The first is the up-coming PSLM data, which contain both scorecard and consumption information in a national representative sample (the sample size is about 16,000 households); the second is a rapid assessment of the overlap between the existing BISP beneficiaries and the poverty scorecard-based beneficiaries (the sample size is about 2,500 households); and the third is a follow-up study planned to assess the impact of the food crisis and financial crisis on households (sample size is about 4,000), which also contains the scorecard and the full consumption information. (C) Spot Checks 6. The sp~t-checks~~ (or operational audits) are at the initial stages of program implementation mainly used to evaluate whether the data collected on the poverty scorecard is accurate. This objective 38

51 will be achieved by re-visiting a sample of randomly selected families and re-collecting the same information that has been collected in the poverty scorecard and comparing it with respect to results provided by Partner Organizations. If difference exists, causes are to be investigated and the targeting in the problematic villages/union Councils (UCs) is to be repeated. (D) Qualitative Survey to Identify the Potential Areas of Focus for the Cash Transfer Impact Evaluation 7. This activity is designed to collect qualitative information on how the families are planning to use the additional income provided by BISP and how the additional income can potentially alter households behaviors in order to better understand the outcomes of the BISP. These areas include (increased) consumption on some items, (increased) savings, (increased) investment in productive assets, preferences to send children to school or work, women s social status, labor participation, and others. The collected qualitative information would be used to develop the questionnaires for the program impact evaluation. Phase II: Impact Evaluation of BISP 8. For the roll out of the BISP program, a full-fledged impact evaluation would be carried out that would benefit from the gradual roll out at the national level. The impact evaluation study will assess how cash transfers affect food expenditure, education and nutrition outcomes, child labor supply as well as women s social status. The evaluation would include a baseline survey (between targeting and payment) and at least two follow up surveys to build a panel dataset on the same households to assess the impact of the program in the medium and long term. The follow-up survey should be carried out ideally twice a year in order to capture the seasonality effects on consumption but the frequency of the follow-up surveys is contingent on the availability of the financial resources and institutional capacity. 9. Decisions are yet to be made by the government whether the national rollout could be a randomized process. If the rollout is to be a randomized process, it should be at the U C level and this randomization will guide the national rollout - the UCs in the treatment group should be rolled out first, followed up with UCs in the control group. In both treatment and control groups, a random sample from beneficiary families and non-beneficiary families that are right above the cut-off score will be selected. If the rollout is constrained by various factors and can not be a randomization process, matching technique has to be used to identify a control group in the later-enrolled UCs/districts which are as similar as possible to the selected beneficiary families in the earlier-enrolled UC/districts. Alternatively, the impact evaluation has to be entirely relied on the regression discontinuity design - compare beneficiaries right below the cut-off and non-beneficiaries right above the cut-off. 10. The questionnaire should include a consumption module to capture the potential impact of BISP on consumption, a household decision making module to capture the potential impact of BISP on the dynamics of decision making within families given that cash transfer is given to the women of the families, a self perception of social status module for the same reason mentioned above, a child education and child labor module to capture the potential impact of BISP of keeping children from child labor given additional income received by families There are a number of instruments for the design of the impact evaluation study. The first and the best instrument is to use a randomized experiment with treatment group being enrolled first and control group being enrolled later. The assumption is that families in the treatment group and control group are similar in both observed characteristics and unobserved characteristics. Direct comparison of the outcome indicators between the treatment and control groups renders the average program treatment effect. The second instrument is to use a Regression Discontinuity (RD) technique. Targeting by proxy means test generates exogenous thresholds that discriminate between eligible and non-eligible households. 39

52 12. Identification in the RD settings stems from the assumption that the un-observables of those who were barely excluded from the program are similar to those of individuals who were barely accepted. Comparing a sample of individuals within a very small range around the cutoff score is analogous to conducting a randomized experiment at the cut-off score (quasi-experimental design). The regression discontinuity methodology identifies the treatment effect locally, i.e., identification stems from the discontinuity around the exogenous cut-off point, thus might have limited implications on the overall program effects. The third instrument is to use a matching technique to match beneficiary families enrolled later and beneficiaries enrolled earlier if randomization becomes unfeasible. Various matching techniques are available, such as non-parametric matching and propensity score matching. The assumption of matching technique is that if families are similar based on a vector of observed characteristics, families are likely to be similar in unobserved characteristics. 40

53 Annex 4: Detailed Project Description PAKISTAN: Social Safety Net Technical Assistance Project I. The proposed project would include four components to support the strengthening of targeting, implementation and monitoring of social safety nets, with a particular emphasis on the federally administered cash grant program, BISP. It also includes support for further developing the country s social protection policy, of which safety nets is only one element. The four components are as follows: Component 1: Establishment of a National Targeting System 2. This component would support the roll out of the national targeting system based on the experience and lessons learnt from a test phase that was started in March To recap: The Government of Pakistan, through the Prime Minister s office, adopted a poverty scorecard developed in 2008 as the basis for a national targeting system that would, initially, select the hture beneficiaries of the Benazir Income Support Programme (BISP), but is envisioned to eventually be used for the selection and targeting of other programs (e.g., health benefits, provincial safety net programs, etc.). The test phase of the poverty scorecard started with the contracting of three Partner Organizations (POs, which are PPAF, RSPN, and the Pakistan Census Organization) with BISP funds. With support from the World Bank, the implementation guidelines (targeting manual) were developed, as well as the training modules for field workers, the logistic plans, etc. The field work started in early April 2009 in 15 districts across all provinces. The districts were selected based on poverty criteria, accessibility, and presence of POs. The test phase is accompanied by a variety of monitoring activities, including a process evaluation of the field work of the POs, spot checks or operational audits which would assess the quality of the scorecard application by the POs, and regular monitoring of the scorecard implementation by BISP staff and regionalaoca1 offices. 3. The results of the test phase, properly evaluated through the above mentioned instruments, would then be used to define the implementation process of the roll out of the scorecard targeting process, financed under this project and expected to start in July To date, Government has decided to use door to door visits for the test phase and roll out of the scorecard survey. However, this will be revisited based on the experience of the test phase, the quality and availability of capable Partner Organizations, and the unit costs and speed of the implementation experienced during the test phase. Targeting centers could potentially be used selectively to speed up the roll out process and bring down unit costs (costs of survey per family), but a decision on this modality will still need to be made for the full roll out. 4. The national roll-out process will proceed in phases, gradually rolling out district by district, depending on the quality, capacity and availability of the POs. BISP is considering contracting additional Partner Organizations for the national roll out, but this will need to be determined based on the test phase for which performance evaluation criteria have been established. Based on the implementation guidelines, materials, training modules and logistic support developed during the test phase (supported by a DFID TF managed by the World Bank), the targeting roll out will continue with massive training of field staff of partner organizations in a cascade process. 5. Once the field work is underway, the scorecard forms will be collected by the POs and sent to NADRA which has developed a Management Information System (MIS), with support and oversight from BISP and the Bank, that will process the information contained in the forms, and select eligible beneficiaries based on the proxy means formula described in Annex 12 and cross checked with existing data bases that NADRA is managing (e.g., CNIC, public employees). The list of eligible beneficiaries will then be sent to the Pakistan Post for payment. In case the information contained in the form is not complete, or contains internal or external inconsistencies (internal inconsistencies are those occurring from differing information within the form, external inconsistencies may occur when information is cross 41

54 checked with external sources (e.g., CNIC names). These cases will be handled as pending and will be addressed through an appeals mechanism until they are solved - either as rejected or as eligible cases. To this end, the project would support the establishment of the institutional capacity to handle such claims and resolve them in a timely and client oriented manner as possible. 6. Maintenance of the targeting database will be crucial for making the most of the targeting system. It would include keeping the family information accurate, namely registering changes in the family composition, location, etc. Based on international experience, it is not envisioned to carry out a recertification process of beneficiaries during the duration of the project. Typically, these are implemented after longer periods of time. 7. In short, this component would finance the following activities: Finalization of targeting design: This would include updating of manuals, materials, and other targeting tools, as well as training to stakeholders in order to keep the targeting data base updated and arrange for any necessary adjustments. Nationwide rollout of targeting survey: The project would provide support for the following activities: (i) training for targeting activities, (ii) Service contracts and operational costs for targeting data collection, and (iii) service contracts for data processing and cross-checking with existing data bases. Administration of targeting appeals: During the test phase, the processes, systems and manuals for targeting related appeals will be developed. The proposed project would therefore assist in (i) providing services for necessary changedadjustments, and (ii) administration of appeals for timely responses to any inquiries. Maintenance of the targeting database. This would include the processing of updated information from beneficiaries as well as any incoming requests and applications. Component 2: Strengthening Safety Net Operation 8. This component would support the implementation of the safety net program beyond the targeting process. It would provide assistance for the entire program cycle of the cash transfer program, starting with the strategic design and implementation of a public information campaign that would be directed at (a) the beneficiaries to inform them about program implementation, their responsibilities, and social accountability mechanism, (b) the general public about program objectives, progress and participating players, (c) selected stakeholders that would be targeting specifically in accordance with the needs of the program (e.g. teacherdhealth staff, local authorities, MNAs, etc.). The project would support all aspects of the public information campaign including materials, training, technical assistance, etc. 9. After the targeting process (see component l), the project would support the verification of the collected data, the application of the proxy means test, as well as the selection and enrolment of beneficiary families. This process is presently carried out by NADRA with oversight by BISP (see annex 6 for more details on the operations/program cycle). 10. To date, payments to beneficiaries are made through the Post Offices. An agreement has been signed between BISP and the Postal Services to make payments to the eligible families through the money orders delivered by the postmen. The female adult of the family needs to sign a receipt when given the money at her door step, which is then being processed by the Postal Services for the reconciliation of payments. This system is not yet fully developed and will receive support by the project through technical assistance and some software development.. 42

55 1 1. The component would consist of the following activities: 0 Implementation of a public communication strategy. The project will finance the following activities: (i) services and materials to implement a communication strategy and to design and produce accompanying communication materials; (ii) training for government and partner organizations; and (iii) printing of materials, etc. Improvement of beneficiary selection and program enrolment processes: The project would finance (i) consultancies to develop the necessary protocols to access beneficiary eligibility information within the national targeting database and to feedback updated beneficiary information collected during the selection and enrolment processes (ii) hiring of temporary support field staff; and (iii) rental of goods, including hardware. 0 Strengthening of mechanisms for payment provision and reconciliation mechanisms: The project would finance the following activities: (i) assessment of options for the optimal distribution of payments at national level; (ii) design of revised payment system;and (iii) pilot experimentation with and evaluation of new technologies for provision of payment services. Establishment of BISP field offices: The project would support the establishment of field offices to ensure access and client responsiveness of the program at the local level. BISP is still in the process of defining its local structure. The project would support this process through the provision of technical assistance, equipment, refurbishing of offices, and training. 0 Implementation of grievance redressal (for payments) and monitoring. The project will finance the following activities: (i) operation of grievance redressal process; and (ii) monitoring and supervision mechanism to oversee timely feedback and corrective actions for BISP operations. 0 Establishment of social accountability mechanisms: The project would finance the following activities: (i) establishment of hotline, training of staff, and hot line operation; and (ii) support for social accountability mechanisms at the local level (e.g., publication of beneficiary lists, village meetings, etc.). Component 3: Enhancing Safetv Net Propram Management, Accountability and Evaluation 12. This component seeks to build the institutional capacity for the national social safety net platform. It would support the safety net administration by providing an integrated Management Information System that would link the targeting data base with the selection of beneficiaries, payments, and grievance redressal mechanism, and provide timely information and reports on the physical as well as financial progress of the program. Furthermore, the project would provide financing for critical staffing at the central and provincialaoca1 levels, particularly during the initial stage of safety net program implementation when more staff is needed than later on due to the nature of the program. It would also provide capacity building and training opportunities for participating institutions at all levels as needed to ensure that the program can continue to operate beyond the life of the project. In order to further develop the safety net program conceptually and operationally, the project would also provide technical assistance, training and materials to design and pilot key interventions that would help beneficiaries to graduate from the program - from safety net to safety ladder. This could take the form of piloting coresponsibilities for beneficiaries, training for youth from families receiving the income support, or others. The project would only finance the initial stages of these pilots with a view to engage other development partners and interested parties in their expansion and further implementation. 13. In order to strengthen the management of the safety net program, a set of tools will be developed and financed under the project, including operational audits/spot checks (to review the diligent application.of the implementation guidelines provided to participating institutions in the Operational Manual of the 43

56 program), process and impact evaluation (see annex 3 for more detail) and financial and possibly performance audits for which training will be provided to the competent authorities. 14. Concretely, the component would consist of the following activities: Design and roll out of MIS at national level: The project would assist in: (i) completing the full design of the MIS (started under the previous DFID TF), (ii) providing hardware at the necessary levels, servers, and connectivity; (iii) software and licensing including training, manuals, etc., and (iv) maintenance of the system. Strategic staffing and capacity building: The project would finance (i) technical assistance in the areas of project coordination, monitoring & evaluation, training, MIS, financial management, procurement and communications;(ii) capacity building and training of stakeholders; (iii) implementation of periodical spot checks, internal audits, and other control mechanisms, (iv) technical staff at the provincial and local level; and (v) transportationhehicles, office equipment and furniture, etc.. Analytical and pilot work to develop program graduation strategies and links to human development services: (i) design and supervision of at least two pilots; (ii) pilot implementation through service contracts; (iii) evaluation of pilots, and (iv) development and publication of dissemination materials and organization of workshops and seminars. Process and Impact Evaluations: The project would finance the following activities: (i) develop the methodological design, including the survey(s) sampling strategy for process and impact evaluations (both qualitative and quantitative); (ii) collect and analyze baseline and follow up survey data; (iii) implementation of beneficiary and other assessments to learn about the quality of program implementation, and (iv) development and publication of dissemination materials and organization of workshops and seminars. Operational auditshpot checks: The project would finance biannual spot check surveys that would audit selected elements of the program cycle. This would be done based on a sample of beneficiaries and other stakeholders. Safety net program specific training for external auditors: The project would finance specific training for the Office of the Auditor General in Pakistan related to the auditing for safety net programs. This would be based on international experience and best practices of auditing practices of safety net programs. Component 4: Developing the Social Protection Policy and Strateq Monitoring 15. This component would provide the government with support for advancing the country s Social Protection agenda. Its implementation would be led by the Planning Commission which would be responsible for the content of the component. It would include analyses of the institutional and legal framework of the overall social protection system of the country. Based on this analysis, an M&E system for monitoring the social protection system would be established, that would include the evaluation of selected programs in order to inform the Social Protection policy making process, including budget allocations. Finally, this component would enable the key Social Protection stakeholders to access international knowledge and experience on selected social protection topics, and provide much needed capacity building for social protection policy making. 44

57 16. Concretely, the component would seek to: 0 0 Strengthen the institutional and legal framework for social protection: The project would finance the following activities: (i) consultancies to define institutional roles and prepare and/or revise the necessary Operations Manuals and other documentation; and (ii) technical assistance for the design of a service delivery system for the NSPS. Build a Monitoring and Evaluation System for the National Social Protection Strategy: Technical assistance to streamline monitoring and evaluation arrangements across selected safety net programs. Train and build capacity of policy makers and implementers in social protection policy: This would include training of government representatives as well as key stakeholders on social protection issues, and would include training courses, study tours, workshops, etc. 45

58 Annex 5: Project Costs PAKISTAN: Social Safety Net Technical Assistance Project Pakistan Social Safety Net Technical Assistant Project Com ponents Project Cost Summary 'YO 'YO Total (Rupees Million) (US% Million) Foreign Base Local Foreign Total ~-- Local Foreign Total Exchange Costs 1. Establishment of a National Targeting System 2, , Strengthening Safety Net Operation Enhancing Safety Net Program Management, Accountability and Evaluation Developing the Social Protection Policy and SP Strategy Monitoring Total BASELINE COSTS 3, , Physical Contingencies Price Contingencies Total PROJECT COSTS 3, , Expenditure Accounts Project Cost Summary I. Investment Costs A. Equipmenflehicle B. Printing Materials C. Technical Assistance D. Training E. Admin. CostsiOperation Total Investment Costs 11. Recurrent Costs Incremental Operating Costs Total Recurrent Costs Total BASELINE COSTS Physical Contingencies Price Contingencies Total PROJECT COSTS 'YO YO Total (Rupees Million) (US% Million) Foreign Base Local Foreign Total --- Local Foreign Total Exchange Costs , , , , , , , , , ,

59 Pakistan Social Safety Net Technical Assistant Project Expenditure Accounts by Components - Totals Including Contingencies (US$ Million) I. Investment COS& A. EquipmentNehicle B. Printing Materials C. Technical Assistance D. Training E. Admin. Costs/Operation Total Investment Costs II. Recurrent Cos& Incremental Operating Costs Total Recurrent Costs Total PROJECT COSTS Taxes Foreign Exchange Developing Enhancing the Social Establishment Safety Net Protection of a Program Policy and National Strengthening Management, SP Targeting Safety Net Accountability Strategy System Operation and Evaluation Monitoring Total o Note: Due to rounding, the total figures may not add up. 47

60 Annex 6: Implementation Arrangements PAKISTAN: Social Safety Net Technical Assistance Project BISP Institutional and Legal Framework 1. The Benazir Income Support Programme (BISP) was initially established under the Ministry of Finance. In September 2008, the Program was transferred to the Cabinet Division under the Prime Minister s office through a federal government s notification. In order guide the program, the BISP Management Board was created, comprising of a Chairperson (with the status of a Federal Minister); Vice Chair (Member of the National Assembly); 3 MNAs; the Deputy Chairman of the Planning Commission; the Advisor Finance; and the Managing Director who also officiates as Secretary to the Board. A Managing Director was appointed and assigned all administrative and financial powers as the Principal Accounting Officer as provided to all Federal Secretaries under the Rules of Business of Pakistan. 2. In order to establish BISP as the permanent national safety net platform, the Government passed an Ordinance in April 2009 to establish BISP as an autonomous safety net authority in Pakistan. The purpose of creating BISP as an autonomous program is to provide the organization requisite and financial and administrative autonomy for the design and implementation of the government s main safety net program with appropriate provisions for program operations. 3. BISP Governance Structure: BISP is governed by a Management Board headed by the Chairperson and advised by a Council headed by the President and the Prime Minister of Pakistan which are the Chief Patron and Executive Patron of the Council, respectively. The Board is composed of 9-11 members including the chairperson. Board members as well as the Secretary of the Programme are appointed by the Chief Patron and the Executive Patron of the BISP Council. Board members include representatives of government, non government organizations and technical experts who have relevance to the purpose of the program. The Council will comprise highly reputable, distinguished and well accomplished national and international individuals. The Chairperson will be the ex-officio member of the Council, the Secretary will be the ex-officio member of the Board. 4. Powers and Functions of the Council. The new BISP Ordinance establishes a Council of the Programme which shall comprise highly reputable, distinguished and well accomplished national and international individuals. The President and the Prime Minister of Pakistan shall be the Chief Patron and Executive Patron of the Council, respectively. The Chairperson shall be the ex-officio member of the Council. The Council would meet at least once in a year. The council shall mobilize financial resources for the Programme and advise the Board on the matters of policies on poverty reduction, enhancement of the Programme, reach out to the donors through the Chairperson of the Programme, and conducting Board affairs and performing its functions. 5. Powers and functions of the Board. The Board s powers and functions include: (i) (ii) (iii) (iv) (v) (vi) approval of the budget of the Programme prepared by the Management; take decisions on the financial aspects of the program submitted by the management for Board consideration monitor the Programme in a transparent manner; make regulations and approve policies and manuals in order to carry out the purposed of the Ordinance; approve criteria of eligible families for financial assistance under the Programme; and present annual progress reports to the Council and consider its recommendations. 48

61 ~ 6. BISP Management Structure: The head of BISP management is the Secretary as defined in the BISP Ordinance with all administrative and financial powers given in the Rules of Business of Pakistan for a Federal Secretary. At the federal level, presently the Secretary along with a Director General (DG) Administration and DG Operations have been appointed. Moreover, Director Administration, Director IT, Director Finance & Accounts, Director Internal Audit and a Deputy Director Media, along with some support staff have also been appointed. Some key positions at the federal level, namely DG Finance & Accounts, M&E Officer, Director Media and certain support staff are yet to be filled. The Organogram of the BISP at the federal level is placed below: Organizational Structure of Benazir lnwme Support Program (BISP) I I I I + - DG 7 DG 7 Operations P.A Admin PA Cmplaiol Cell UDCRDC UDCRDC 10 X ASSEIS NQ N.Q Dri. Dri. T 1 v t 1 I 1 T Diredor Diredor Diredor Dlredor Direclor DIreclor Diredor Director Punjab' DJ% %% AJWNA ISL!T!F c o ~ ~ ~ i o, IT Admi-i Admi-li Media 1 1 PA PA PA PA PA PA P.A PA = P.A PA ASSlSl ASSbl ASIS1 Ass!51 ASSiSl NQ N.Q N.Q N.Q N.Q d NQ= Chowk= 4 swep= 4 DR= 2 PMO= I Rep= 2 Dri= 17 - RBI Asst= 3 7. At the provincialhegional level, four Regional Director Generals for provinces have been appointed. Moreover, two Directors for AJK & NA have also been appointed. Provincial DGs shall be assisted by 28 Divisional Directors, who shall be further assisted by 127 Supervisors. The Organogram of the BISP at the regional level is placed below: 49

62 I Organizational Structure of BISP Balochietan Supervisor A.C=ASSistanl Complaint *A M 0- Asn Media otflur *NC!=Nub 1 'PA-PerSOnaIAssJstan! 1 Qasa 8. BISP Operational Structure. As per the BISP Operational Manuel, the program is composed of three main departments: (i) the Operational Department, (ii) the Administration Department, and (iii) the Department for Finance and Accounting. (il The Operations Wing will be mainly responsible for the program design and its implementation, which, among others, include coordinating payments to the payment agency through the Finance & Accounts Wing; providing input in procurement related matters essential for program implementation; providing input for policy decision; handling matters relating to survey organizations, payment agency and data processing firm, and concluding agreements with the stakeholders. Five Directors of the Operations Wing will coordinate the responses from the provincial and regional offices as well. While the regional/provincial offices will handle the complaints, perform M&E function and provide for grassroots interaction with the beneficiaries; however, a complaint cell shall also be established at the BISP headquarter to augment the efforts for complaints redressal, which shall primarily remain the responsibility of provincial/regional offices. (ii) (iii) The Administration Department will handle all issues related to human resources, procurement and general program administration and should have units accordingly. The Finance and Accounting Department is responsible for all financial budgeting, financial transactions and reporting for the program. The Internal Auditor will conduct annual audits and will prepare all necessary documents for third party audits whenever required. 9. Project Implementation Unit: Implementation of the project would be supported by a small unit headed by a Project Coordinator who would help coordinate consultancies, facilitate communication with BISP Management and departments for the smooth implementation of the project. The project would 50

63 support the hiring of Financial Management Specialists that would work directly under the Director Finance, and Procurement Specialists that would work under the DG Administration. BISP Partners in Program Administration 10. As part of setting up the national targeting system that separates the enrollment of beneficiaries from the determination of their eligibility for the BISP and to improve administrative efficiency, three organizations, namely the Pakistan Poverty Alleviation Fund (PPAF), the Rural Support Program Network (RSPN) and the Population Census Organization (PCO) have been selected to for data collection during the test phase of poverty scorecard. The rationale for selecting these Partner Organizations (POs) is to capitalize on their outreach and ability to galvanize the community to apply for the program by completing the scorecard applications during the targeting process. The Government will also establish clear criteria for evaluating their performance of each organization regarding efficiency, maintaining confidentiality of household information, quality of trainings and of information campaigns, and the methodology of approaching culturally and geographically diverse populations. After the test phase, BISP will engage in similar partnerships with these and if needed other organizations that have the capacity and capability of managing the scorecard data collection process The BISP has also entered into an agreement with National Database and Registration Authority (NADRA) for database design and management; MIS; verification of eligible beneficiaries selected through the Targeting Process, and generation of payment lists. The arrangement will remain effective till completion of the poverty scorecard test phase in 16 districts, starting from May On completion of the test phase, the partnership arrangement will be re-visited for the nationwide rollout. 12. BISP has also signed an agreement with Pakistan Post, which stipulates that presently, the BISP beneficiaries shall be paid through Pakistan Post Money Orders (a money order is an order issued by the post office for the payment of a sum of money through the agency of the post office). This arrangement was favored keeping in view the vast coverage, easy accessibility and money handling and delivery capacity of Pakistan Post. Efforts are underway to design technology based payment systems such as smart cards, which will be piloted and evaluated for delivery of the program benefit at an appropriate stage. Implementation Arrangements 13. The overall implementation of BISP will be guided by an Operational Manual which has been approved by the Management Board in early April, 2009, but will be amended and refined from time to time, in accordance with program progress and experience made during implementation. 14. The Operational Manual specifies the overall process of the program cycle which is composed of three main phases: Targeting, Enrolment and Payments. During this process the beneficiaries might interact with the program by updating their data, filing complaints in case of non-selection, and appealing in case of grievances - all those processes are part of the Case Management. The Program is overseen during the entire implementation through close Monitoring activities and the quality of service delivery, as well as program outcomes and even the larger impact of the BISP are investigated through Evaluations. The interaction between those different processes is summarized in the Project Cycle Chart (Figure 1). 15. The Targeting includes the filling of the Scorecard-based application forms, processing and verification of scorecard data in the program MIS, and calculation of the PMT score. The Enrolment includes the selection of the beneficiary families through application of the cut-off score approved by the BISP Management Board and sending out the enrolment letters to selected beneficiaries. The Payment includes delivery of program benefit amount to eligible families over a determined period. 51

64 16. It is expected that some of the families that perceive their own circumstances as severe poverty, are not chosen as beneficiaries and like to appeal. For such cases the program provides the opportunity to file for Grievances. For all other grievances such as those related to payments, the BISP divisional offices will be entrusted the task of their resolution through involvement of local structures. Detailed procedures for Grievances are provided in the Operational Manual. 17. During the program implementation, families enrolled in the program as beneficiaries might have changes in the family composition on account of marriage of one member or the moving out of one family member from the household, etc. The program design provides for reporting such changes, which might affect even the eligibility for the program. The Operational Manual provides the procedures for Updating of beneficiary data besides setting up the rules for the beneficiaries to comply with the requirement. Finally, to maintain the quality of service delivery to the beneficiaries, the Operational manual also provides the Monitoring and Evaluation system for effective and efficient program administration. 52

65 Country Issues Annex 7: Financial Management and Disbursement Arrangements PAKISTAN: Social Safety Net Technical Assistance Project 1. The Bank has carried out extensive analytical work on public financial management (PFM) systems in the country, both at national and sub-national levels. A Country Financial Accountability Assessment (CFAA) was completed in December 2003 which highlighted weaknesses in budget execution and inadequate accounting, financial reporting and internal controls mainly due to low reliability of data below the federal level. In May 2007, Public Financial Management and Accountability Assessments (PFMAA), using the PEFA PFM Performance Measurement Framework, were completed for Baluchistan, Punjab, and NWFP and a federal level PFMAA using the same framework is currently underway for delivery during FY The implementing agency, BISP is a statutory corporation established through an Ordinance. It has independent financial management arrangements lying outside the domain of PFM systems. Project Components Component 1 : Establishment of a National Targeting System Component 2: Strengthening Safety Net Operations Component 3: Enhancing Safety Net Program Management, Accountability and Evaluation Component 4: Developing the Social Protection Policy and Strategy Monitoring TOTAL Amount in Nature of USD Expenditures millions 34.6 Consultancies, training and incremental operating costs 12.9 Goods and works, Consultancies, training, and incremental operating costs 10.8 Goods, Consultancies, training and incremental operating costs 1.7 Consultancies, training and incremental operating costs 60 Risk Analysis a) Inherent risk: The inherent risk is High. Country level risk is Substantial to Moderate but the implementing agency lies outside the domain of PFM systems. Since it is a new entity in the process of being established, it is subject to the inherent risks in such situations which can be effectively mitigated once clearly defined institutional arrangements are in place. b) Control risk: The control risk is High. Professional financial management staff dedicated for the project will be hired from the market prior to Disbursement. Director General Finance and Accounts for BISP is yet to be appointed and the finance and accounting guidelines are also in the process of being developed. Once staffing is complete and the guidelines are issued it is anticipated that there will be adequate assurance that project funds will be used for intended purposes, economically and efficiently. Public Expenditure and Financial Accountability 53

66 c) Residual risk: The residual risk rating of the project is considered High. This will be revisited once risk mitigating measures are put in place. Risk Risk Rating Risk Mitigating Measures Incorporated into Project Design Condition of Negotiations, Board or Effectiveness Country Level Substantial - Autonomous implementing agency N with independent financial management arrangements A Entity Level High - Well defined institutional N arrangements. Project Level High - Ring fenced arrangements for the project; - FM staff with professional qualifications and experience to be Y Hiring of FM hired and trained under the project; specialists prior to - Project activities subject to the Disbursement entity s internal controls including internal audit. Control Risk Budgeting Accounting Internal Control Funds Flow Financial Reporting Modest High High Modest High - Normal GoP rules will amlv. - Accounting and financial reporting to be automated as part of entity s MIS; - Induction of professionally qualified staff. - Strong internal control framework, including independent internal audit built into operational manual; - Issuance of finance and accounting guidelines % financing of all categories; - IUFR based disbursement in segregated designated account. - Induction and training of professional FM staff with appropriate experience and qualifications; N Y Hiring of FM specialists prior to Disbursement Issuance of Finance and Accounts guidelines for the Project prior to Disbursement N N Auditing Overall Risk Modest - Automation of accounting and reporting. - External audit by Supreme Audit Institution. N 54

67 Risk rating summary is in the working papers file. Weaknesses and Action Plan 3. The implementing entity is newly established and institutional arrangements are being finalized. Weakness Lack of key staff in Finance and Accounts Internal audit not independent of operations MIS for accounting and finance not in place Lack of documented finance and accounts policies and procedures Implementing Entity Action Hiring of two FM specialists Appointment of Director General Finance and Accounts Revision of organogram to change reporting of Internal audit function to Management Board Implementation of accounting and finance module in MIS Issuance of finance and accounting guidelines for BISP including a separate chapter for the Project Issuance of Finance and Accounting guidelines for the Project Responsibility MD, BISP MD, BISP Completion date Prior to Disbursement MD, BISP July 15,2009 Completed. Incorporated in BISP Ordinance MD, BISP Within first year of effectiveness MD, BISP MD, BISP June 1,2009 Not later than 3 months after effectiveness Prior to Disbursement 4. All aspects of the Project s implementation will be managed by BISP. It has been established as a statutory corporation with an independent Management Board. Accordingly, BISP will notify its own accounting procedures within the framework of the GoP s General Financial Rules. 5. The current staffing structure is inadequate for the anticipated nature and scale of operations and will be unable to satisfactorily manage accounting and financial reporting functions once the entity fully commences operations and sub-national formations are functional. The Finance and Accounts guidelines under preparation will also provide the minimum staff and capacity required. Currently an Accounts Officer and two Assistants are in place. The former is a retired officer from the Military Accounts service. The assistants are SAS qualified. The position of Director General for Finance and Accounts is vacant. A Director Finance has recently been appointed. He is from the Income Tax service. The support staff will need to be significantly augmented, preferably with staff deputed from the Office of the Auditor General. Training and capacity building activities will be done under the Project for the Finance and Accounts staff. 6. Two Financial Management Specialists will be hired prior to Disbursements to ensure adequate financial management of the project. They will advise and assist the Director Finance in building the financial management capacity of the entity. 55

68 7. An internal audit department has been set-up. The Director and an Audit Officer are in place and an Assistant is required. The incumbent Director Internal Audit has more than 15 years of experience and is a Certified Internal Auditor. As with Finance and Accounts, once the entity is fully functional, staff in Internal Audit will also need to be enhanced to ensure that the internal audit work program can be effectively implemented. Budgeting 8. Normal Government of Pakistan rules and procedures will apply. One hundred percent of Project activities will be financed through the technical assistance and there shall be no counterpart funds. Annual budget for the project will be prepared and submitted on the basis of planned activities. Accounting 9. Separate books of account, on cash basis, will be maintained for the Project activities using the Chart of Accounts under the New Accounting Model. Sufficient subsidiary records will be kept to facilitate preparation of quarterly reports and annual financial statements providing details of receipts and expenditures by Project components and disbursement categories. An accounting and financial reporting module will form part of the entity s MIS. 10. BISP is currently maintaining the following books: - CashBook - Cheque Book Register - Cheque Register - Stock Register - Contracts Register - Budget Control Register - Disbursement Register - Register of Advances 11. Records to be kept for the funds transferred to Pakistan Post for distribution include the following: - Statement of Releases - Monthly Statement of Expenses - Monthly Funds Reconciliation 12. However, complete reconciliations for funds transferred to Pakistan Post are yet to be made. Internal Control and Internal Auditing 13. Comprehensive accounting and finance guidelines are being developed. These will incorporate- a robust internal control framework and ensure adequate segregation of duties. Until the issuance of the entity s guidelines, finance and accounting guidelines for the financial management of the Project will be adopted. This will be done prior to disbursement. The guidelines for the Project will be incorporated in the entity s guidelines in the form of a separate chapter describing the procedures and controls required for the financial management of the Project. The entity s guidelines will be issued within three months of the Project being declared effective. Invoice register will track payment of invoices. Separate record of assets procured from project funds will be maintained. These will be tagged for identification and regularly subjected to a physical verification. Bank reconciliation will be prepared at the end of each 56

69 ~ ~~ month and approved by the Director Finance. Draft accounting procedures have been submitted to the Office of the Auditor General of Pakistan for review. 14. The initial organizational structure provided for an internal audit function but currently its reporting line is to the DG Finance and Accounts. To ensure independence and enhance the control environment, this has now been revised to be placed directly under the Management Board as per the BISP Ordinance. An internal audit policy has been drafted and is under internal review. Funds Flow and Disbursement Arrangements 15. Project activities will be fully financed by the IDA credit, including taxes. A segregated designated account in US Dollars will be opened for the Project. This will be operated by joint signatories ensuring adequate segregation of duties. Funds will be front-loaded using a report-based system. The Project will submit quarterly Interim Un-audited Financial Reports (IUFRs) providing six monthly cash forecasts. Initial advance into the designated account will be provided by the Bank on the basis of projections for the first two quarters, which will be replenished quarterly on the basis of actual expenditures incurred and forecast for the following two quarters. 16. The proposed IDA credit would be disbursed over a period of four years. Allocation of Credit proceeds by disbursement category and related percentages to be financed under the IDA Credit are shown in the following Table. Allocation of Credit Proceeds Cat ego ry Goods, Works, Consultants Services, Training (including workshops and study tours) and Incremental Operating costs TOTAL AMOUNT Incremental Operating Costs Amount of Credit (expressed in USD 60,O 00,O Percentage of Expenditures to be Financed 100% 17. The incremental operating costs will cover expenditures for salaries of contractual staff (other than consultants), per diem and allowances, office rental, office supplies, utilities, conveyance, travel and boardingaodging allowances, operating and maintenance of office equipment and vehicles, advertising expenses and bank charges, insurance, media projections, newspaper subscriptions, periodicals, printing and stationary costs in connection with the management and coordination of Project activities, which expenditures would not have been incurred absent the Project, but excluding salaries of the Recipient s civil servants. Retroactive Financing 18. Retroactive financing will be provided for the expenditures incurred one year prior to the Credit signing date to the tune of US$2,000. Financial Reporting 19. The project will submit IUFRs within 45 days of the end of each calendar quarter. Format and content of these reports will be agreed during negotiations. 57

70 20. Annual financial statements for the Project will be prepared in accordance with cash basis international public sector accounting standards. These will include details of expenditure incurred both by disbursement category and Project components. Auditing 21. External audit will be conducted by the Supreme Audit Institution, Le., the Office of the Auditor General of Pakistan which is acceptable to the Bank. Acceptable audited financial statements for the project will be submitted within 6 months of the close of each financial year. Audit Report Type Project Financial Statements for Financial Year ending June 30 each year. Due Date December 3 1 each year. Conditionality (i) Two FM specialists to be hired prior to Disbursement. (ii) Finance and Accounting guidelines for the Project to be issued prior to disbursement. (iii) Finance and Accounting guidelines for BISP including a separate chapter for the Project to be issued not later than three months after effectiveness. Supervision Plan 22. Intensive FM supervision will be required. 58

71 Annex 8: Procurement Arrangements PAKISTAN: Social Safety Net Technical Assistance Project A. General 1. Procurement for the proposed project will be carried out in accordance with the World Bank's "Guidelines: Procurement under IBRD Loans and IDA Credits" dated May 2004 revised October 1,2006; and "Guidelines: Selection and Employment of Consultants by World Bank Borrowers" dated May 2004, revised October 1, 2006, and the provisions stipulated in the Legal Agreement. The various items under different expenditure categories are described in general below. For each contract to be financed by the Loan, the different procurement methods or consultant selection methods, the need for pre-qualification, estimated costs, prior review requirements, and time frame are agreed between the Borrower and the Bank in the Procurement Plan. The Procurement Plan will be updated at least annually or as required to reflect the actual project implementation needs and improvements in institutional capacity. The GPN describing all major procurements under this project shall be published by Apri124, 2009 followed by SPNs where required. 2. Procurement of Consultants: Major consultancy services under this project would include: IT Consultancy firm(s) for Software, Hardware, and Data Network - for system design, analysis, development, testing and execution of overall system; Quality check - Process Evaluation, Spot Checks, and Social Accountability; Monitoring and Supervision, Management; Communications Strategy; Exit Strategy - Human Development Strategies; Financial Consultants; Grievance redress system; Legalhnstitutional technical assistance, Audit for payments disbursed to beneficiaries, Capacity Building firm for PMU. For contracts with consulting firms estimated to cost less than $200,000 equivalent per contract, the shortlist of consultants may comprise entirely national consultants in accordance with the provisions of paragraphs 2.7 of the Consultant Guidelines. The first Consultancy services contract, regardless of value and thereafter, Consultancy contracts estimated to cost $200,000 or more per contract and all single source selection of consultants (firms) will be subject to prior review by the Bank. 3. Selection of Individual Consultants: Services for assignments that meet the requirements set forth in the first sentence of paragraph 5.1 of the Consultant Guidelines may be procured under contracts awarded to individual consultants in accordance with the provisions of paragraphs 5.2 through 5.3 of the Consultant Guidelines. Individual Consultants will be hired for PMU Capacity Building. These include Individual Consultants/Specialists for Procurement, Financial Management, Social, IT etc. 4. Procurement of Works: Works procured under this project, will include some minor works of refurbishment and renovation of offices,. Accordingly, all civil works under the project will be procured through National Competitive Bidding (NCB) procedures acceptable to Bank's Standard Bidding Documents (SBD) for all National SBD agreed with (or satisfactory to) the Bank. However if foreign firms wish to participate in these contracts, they will be permitted. 5. Procurement of non-consulting services: Trainings, Workshops 6. Procurement of Goods: Goods procured under this project would include: Vehicles, Office Furniture, Computer Hardware/Software, Office accessories (Misc. Items). ICB procedures shall be followed for each Goods contract estimated to cost US$200,000 equivalent or more. Goods estimated to cost less than US$200,000 per contract may be procured through NCB procedures acceptable to the Bank. Small value off-the-shelf goods estimated to cost up to US$50,000 equivalent may be procured following shopping procedures in accordance with the Bank's procurement guidelines 59

72 7. Improvement of Bidding Procedures under National Competitive Bidding: The following improvements in bidding procedures will apply to all procurement of Goods and Works under National Competitive Bidding, in order to ensure economy, efficiency, transparency and broad consistency with the provisions of Section 1 of the Guidelines: a. b. C. d. e. f. g. h. 1. j. k. 1. m. n. 0. P. Invitation to bid shall be advertised in at least one national newspaper with a wide circulation, at least 30 days prior to the deadline for the submission of the bid. Bid documents shall be made available, by mail or in person, to all who are willing to pay the required fee; Foreign bidders shall not be precluded from bidding and no preference of any kind shall be given to national bidders in the bidding process; Bidding shall not be restricted to pre-registered firms; Qualification criteria shall be stated in the bidding documents; Bids shall be opened in public, immediately after the deadline for submission of bids; Estimates shall be based on market rates and bids shall not be rejected merely on the basis of a comparison with an official estimate without the prior concurrence of the Association; Before rejecting all bids and soliciting new bids, the Association's prior concurrence shall be obtained; Bids shall be solicited and contracts shall be awarded on the basis of unit prices and not on the basis of a composite schedule of rates; Contracts shall not be awarded on the basis of nationally negotiated rates; Single bids shall also be considered for evaluation. Contracts shall be awarded to the lowest evaluated and qualified bidder; Post-bid negotiations shall not be allowed with the lowest evaluated or any other bidders. Draft NCB contract would be reviewed by the Association in accordance with the prior review procedures; A firm declared ineligible by the Association, based on a determination by the Association that the firm has engaged in corrupt, fraudulent, collusive, coercive or obstructive practices in competing for or in executing an Association-financed contract, shall be ineligible to be awarded an Association-financed contract during the period of time determined by the Association; and Each contract financed from the proceeds of a Credit shall provide that the suppliers, contractors and subcontractors shall permit the Association, at its request to inspect their account and records audited by auditors appointed by the Association. The deliberate and material violation by the supplier, contractor or subcontractor of such provision may amount to obstructive practice B. Assessment of the agency's capacity to implement procurement 8. Procurement activities will be carried out by Benazir Income Support Programme (BISP). Assessment of the capacity of BISP for the project has been carried out by Procurement staff of the task team. 9. The assessment reviewed the organizational structure, current procurement procedures and practices; staff responsible for procurement, types of procurement taking place etc. Most of the issues/ risks concerning the procurement component for implementation of the project have been identified. These risks along with mitigation measures are: 60

73 Assessment of the Benazir Income Support Programme 10. Procurement activities will be carried out by BISP which is a Federal Government entity. An assessment of the capacity of the Implementing Agency to implement procurement actions for the project has been carried out. The assessment reviewed the organizational structure for implementing the project and the existing capacity within the organization to carry out the implementation of the procurement under this TA. Presently, an officer of the level of Deputy Secretary, in addition to other duties, is designated to carry out all procurement along with an associated committee of two members on some minor procurement for basic office requirements like computers, printers, fax machines, LAN, Laptop, Servers, Operating software and Furniture. Procurement is handled by a single person through a flat structure with no supporting staff. No previous experience of procurement using Bank s guidelines exists. As the entity is newly established, there is no procurement unitkpecialized staff present. Procurement capacity is very low, as there is nothing on ground and the entity is at embryonic stage The 2000 Pakistan Country Procurement Assessment Review (CPAR) highlighted a weak regulatory framework governing public procurement, and called for institutional and procedural reforms. CPAR revealed that Pakistan does not have a coherent law that establishes standards and an effective legal system that protects against collusion and corruption in the award of government contracts. CPAR recommended the establishment of a new and modern procurement rules conforming to international best practice. 12. A Public Procurement Regulatory Authority (PPRA) was created in 2002 with powers to formulate public procurement legislation applicable to federal government line departments, state-owned enterprises, and semi-autonomous organizations. New public procurement rules were notified in 2004, and they depart from the old system in two important ways: they require posting of all bid notices on the PPRA s website, and prohibit post-bid negotiations. Moving forward, the development of monitoring and reporting mechanisms, and the creation of an autonomous second-tier appeals procedure to handle complaints in a timely and transparent manner will be critical for effective implementation of PPRA rules. PPRA rules are widely recognized as the rules to follow for procurement, and the Auditor General s Office, conducted the first set of audits using the new procurement rules in the previous fiscal year. As part of the process of rolling out the rules, the PPRA has trained more that procurement staff of Federal government and its autonomous bodies, and over 220 organizations now post their procurement notices on the PPAR website. As the next logical step, the government has now directed its ministries and autonomous bodies to commence posting of notices of award for contracts over Rs.50 million on the PPRA website. Mitigation Measures Agreed 13. Based on the assessment and ensuing discussions with the government, the following action plan has been agreed: Hiring of specialized procurement staff at BISP prior to disbursement; Procurement training of existing and new staff on Bank s procedures and Preparation of a Procurement Manual Procurement Plan agreed at Appraisal Develop and maintain a website posting all procurement notices and awards The overall project risk for procurement is high, after mitigation substantial. 61

74 C. Procurement Plan 14. The Project Management Unit in coordination with BISP has1 developed a procurement plan for the project implementation which provides the basis for the procurement methods. This plan shall be made available on the BISP websites. The procurement plan shall be updated in agreement with the Project Team quarterly or as required to reflect the actual project implementation needs and improvements in institutional capacity. Review of Procurement by the Bank 15. Prior Review threshold as will be proposed while laying a procurement plan after negotiations with implementing agency. (a) (b) (c) (d) (e) (f) (8) all goods contracts procured by implementing agency on the basis of ICB; all works contracts procured by implementing agency on the basis of ICB; the first goods contract procured by implementing agency on the basis of NCB as well as Shopping regardless of its value; all contracts for goods and works estimated to cost more than US$ 200,000 equivalent per contract; the first consultancy services contract with a firm procured by implementing agency regardless of its value: thereafter all consultants services contracts with firms procured by implementing agency estimated to cost USD 200,000 equivalent per contract or more; the first individual consultant contract procured by implementing agency regardless of its value thereafter all individual consultants contracts procured by implementing agency estimated to cost more than USD 50,000 equivalent per contract; and the first non-consultancy services contract procured by implementing agency regardless of its value. 16. All other contracts will be subject to post review by the Bank and emphasis would be placed on ex-post procurement and end-use audits which will be carried out on a regular basis to verify that the activities financed under the project are procured in accordance with agreed procedures and are used for the intended purposes. D. Frequency of Procurement Supervision 17. In addition to the prior review supervision to be carried out fiom Bank offices, the capacity assessment of the Implementing Agency has recommended six monthly supervision missions to visit the field to carry out post review of procurement actions. 62

75 E. Details of the Procurement Arrangements Involving International Competition Goods, Works, and Non Consulting Services (a) List of contract packages to be procured following ICB and direct contracting: Consulting Services (a) List of consulting assignments with short-list of international firms Description of Assignment + out of targeting Updating manuals, materials & other targeting tools 3 Data Processing 4 Implementation of PIC 5 Communication Strategy 6 Design of new payment mechanism 7 Design MIS Comprehensive 8 I Implementation of grievance redressal & monitoring 9 Process & Impact Estimated Selection Review Expected Comments cost Method by Bank Proposals (Prior / Submission Post) Date Rs.1040M QCBS Prior

76 Ref. No Description of Assignment Evaluation Operational Audit- Spot Checks Communication to Beneficiaries (letters) M&E for National Social Protection Strategy Networking & Software Development Rs.5.9M 1 CQS Method Review by Bank (Prior I Post) Prior Prior Post Post Expected Proposals Submission Date Comments 64

77 Annex 9: Economic and Financial Analysis Pakistan: Social Safety Net Technical Assistance Project Introduction 1. Despite recent periods of economic growth, poverty reduction in Pakistan is at risk of being reversed in the wake of financial crisis and an uncertain macroeconomic context. Irrespective of the methodology deployed, economic growth reduced poverty between 2001/02 and 2005/06. The recently completed Risk and Vulnerability Asse~sment'~ (World Bank 2007) reports that about 56 percent of the population are poor and remain vulnerable with the highest levels of risk at the rural household, particularly those employed in the agriculture sector. The majority of poor households are susceptible to both aggregate (economic and natural disaster) and idiosyncratic (health, unemployment) shocks. One survey indicated that about two-thirds of safety net recipients experienced at least one major shock during the three years before the survey2'. An area of particular concern relates to adverse coping strategies and the tendency towards labor substitution by withdrawing and withholding children from school. In response to shocks nearly 33 percent of households lowered their food consumption, 10 percent put their child at work and 8 percent put a child out of school in response to the shock. These patterns were more severe among ultra-poor households (36 percent) and households in urban areas (35 percent) due to higher levels of food insecurity associated with lack of access to land. To this end the economic analysis considers in close detail the possible protective, consumption smoothing and human development benefits of the cash transfer program, also seen through the educational benefits of the transfer. 2. The analysis considers the possible economic and financial benefits of the Benazir Income Support Programme (BISP). The potential economic impacts of the program are outlined based on available data and assessment information. Supporting evidence is also provided based on the global experience of cash transfer programs. The fiscal impact and sustainability aspects of the program are also considered. The main conclusion is that BISP is sustainable both economically and financially. Economic Analysis and Expected Benefits of BISP 3. The Benazir Income Support Programme (BISP) is a new social safety net program introduced by the Government in 2008 to cushion the negative effects of the food, fuel and financial crises of the poor and in the longer term establish a nationwide, effective and transparent cash transfer program for the poor. The program provides monthly cash transfers of Rs (US$12) to the eligible families. The primary beneficiaries are expected to be women and children with the transfer being provided to female head of eligible families. The multi-dimensionality and start-up nature of the BISP makes it difficult to quantify the different social and economic benefits. This section therefore considers the primary consumption and income effects of the program, as well as the secondary potential effects on an education outcomes, 4. The BISP program will have direct effects on consumption and income levels of beneficiaries: the (i) Reductions in poverty: By receiving regular cash transfers households are likely to increase their food intake resulting in less food insecurity and the maintenance of productive assets. This is expected to have a positive effect on consumption smoothing. The average transfer benefits from the program are substantial, reaching an equivalent of US$l50 per families per annum. It is estimated that up to 5 million families will receive US$l50 worth of transfers during the first year, providing a total of US$750 million. 19 Social Protection in Pakistan: Managing Household Risks and Vulnerability. *' Pakistan Safety Net Survey (PSNS 11). 65

78 This will reduce both income poverty and income inequality due to the redistributive effect of targeting the transfer to the poorest families within the households. (ii) Improvement in health and nutrition: Improvements in the health and nutrition outcomes of children and the wider family are expected due to the income effects of the transfer (e.g., increased food consumption, dietary diversity). In addition to the direct welfare benefits arising from improved health status, gains can also be expected through increased schooling levels, improved cognitive achievement as well as productivity, earnings and income for adults. (iii) Income generation: The program is also likely to increase income generation capacity among poor households, particularly by mitigating consumption expenses and negative coping strategies in response to shocks. Shock related expenses represent 54 percent of annual consumption for ultra-poor households, 27 percent for poor households and 18 percent for non-poor households. Households in rural areas spend 37 percent of their annual consumption coping with shocks compared to 30 percent for urban households in this group. 5. The BISP transfer is also expected to impact on human development, particularly amongst children given that transfers are focused on the family unit. For example, based on comparable programs elsewhere it is reasonable to assume that families may devote anything up to 15% of this transfer on education which could have the potential to increase school enrollment. This will be an issue returned to during impact evaluation, together with the potential health impacts of the program. Global Experience of Cash Transfer Programs 6. We also consider international experiences in the implementation of cash transfer programs to gauge the likely impact of the BISP. In recent years non-contributory cash transfedsafety net programs have proven effective instruments to address poverty and strengthen human development outcomes related to education, health and nutrition. There are two major forms of cash transfers: conditional and unconditional. Many of the large and well-established cash transfer programs are conditional with payments dependent on certain behavioral changes within the household e.g. regular school attendance or use of preventive health services or other specified conditions. Unconditional cash transfers (e.g., income support, social pensions, child support grant) are not tied to service use. This may occur for a variety of reasons including limited service supply, ease of administration, timeliness of delivery (particularly during a crisis) or as a starting point towards more elaborate safety net interventions in the future. Under both types of cash transfers, it is expected that there may be direct impact on poverty reduction (through consumption) and human development outcomes may be improved through increased use of education and health services or indirectly through increasing food consumption. 7. In recent years there has been growing attention concerning the use of cash and conditional cash transfer programs as part of broader poverty reduction strategies. Since the late 1990 s Conditional Cash Transfer (CCT) programs have spread rapidly2. While first generation CCT programs originated in Brazil and Mexico, it is now estimated that CCT programs can be found in at least 33 developing countries (Basset, 2008). Because they require complementary supply-side inputs (usually schools and health centers), CCTs have been mainly implemented in middle-income countries, predominantly Latin America, as well as South and East Asia. More recent pilot adaptations are testing CCT s to international emergency settings and urban development programs, including the US context. In parallel cash transfer programs have become an increasingly important element particularly in low income settings (Ethiopia, South Africa), as well as for emergency recovery and reconstruction projects. Robust impact evaluations 21 Conditional Cash Transfer programs (CCTs) provide cash payments to poor households that meet certain behavioral requirements, generally related to children s health care and education. 66

79 I of well designed CCT programs shows that cash transfers conditioned on households participation in education, health and nutrition services have demonstrated significant impacts on poverty and human development outcomes mainly for children. The most significant impact can be seen in the areas of school attendance rates and enrollment. The evidence on health outcome improvements are not as clear but some studies demonstrate that CCT programs have resulted in increased total food consumption and improved child 8. A growing number of low-income countries in Africa like Zambia, Malawi, South Africa and Ethiopia have begun to design social cash transfer scheme without any conditions. All of these programs focus on households that are extremely poor and are often tailored to address both chronic and acute aspects of poverty. For example transfers have been tailored to promote consumption smoothing during lean seasons; to provide timely and predictable cash assistance to protect against asset depletion and harmful coping strategies (e.g., livestock sales, increasing child labor supplies); and to enhance productive benefits for smallholder farmerdrural communities in the areas of agricultural productivity, market access and microfinance. Empirical evidence highlights that these interventions are also improving human development outcomes in the areas of education and health There has been a major debate occurring over whether cash transfers should be conditional or unconditional and there is little research to date that directly compares conditional cash transfers (CCTs) and unconditional cash transfers (UCTs) in a given setting.24 UCTs give poor families most flexibility and are administratively less challenging to implement. However, conditions reassure governments and donors that money will be spent on desired goals. In practice, compliance with conditions is not always enforced rigorously (Chapman 2006).25 The Kalomo Cash Transfer Program in Zambia found that recipients made rational spending decisions, including on health and education for families. UCTs are likely to impact on human development where vulnerable groupshouseholds have sufficient access to information to make informed choices about health and education. This is also confirmed by findings in the broader literature. Significantly, cash transfers have been found to have an indirect benefit in terms of empowering women and socially excluded groups through increased control of household finances Supporting evidence on the impact of Cash Transfers tends to come from CCT programs which have been well designed, developed at scale and more generously funded to allow robust monitoring and evaluation. In contrast Unconditional Cash Transfers (UCT) Programs tend to be smaller in scale, developed without control groups and implemented in lower income contexts where analysis can be constrained by weaknesses in data, administration and contextual fact01-s.~. Despite these limitations a body of supporting evidence is emerging particularly in the African contexts which show that unconditional cash transfers have had an impact on consumption and human capital, even if the magnitude of those impacts and isolation of program impacts are not definitive Fiszbein and Schady, Conditional Cash Transfers for Attacking Present and Future Poverty, See next section. 24 Adato and Bassett, What Is The Potential Of Cash Transfers To Strengthen Families Affected by HIV/AIDs? A Review of the Evidence on Impacts and Key Policy Debates, Monetary incentives in primary health care and effects on use and coverage of preventive health care interventions in rural Honduras: cluster randomized trial Lancet 2004; 364: Zambia and Malawi programme. 27 Adato and Bassett, What Is The Potential Of Cash Transfers To Strengthen Families Affected by HIV/AIDs? A Review of the Evidence on Impacts and Key Policy Debates, Adato and Bassett, What Is The Potential Of Cash Transfers To Strengthen Families Affected by HIV/AIDs? A Review of the Evidence on Impacts and Key Policy Debates,

80 Impact on Consumption 11. Evidence from various programs suggests that UCT s have made substantial impacts on consumption mostly related to expenditures on food, the quantity and quality of food consumed and changes in child s nutrition. In most programs beneficiaries spend substantial amounts of the transfer on food and this led to increased food consumption in the beneficiary households. Impact eval~ation~~of Ethiopia s safety net program3 found that 80 percent of the beneficiaries used cash transfers on staple foods and 11 percent used some of the transfer for other food items. In addition, participating households consumed more and better quality food after receiving the transfer and 60 percent of the beneficiaries were able to keep more of the food they produced for their own consumption. The Malawi Cash Transfer Program3 also showed improvement in food consumption where 93 percent of the intervention households reported improved food consumption. In addition, in Malawi32 and Zambia33 the transfer also increased dietary diversity and number of meals consumption per day. In Malawi, average meals per day among beneficiaries significantly increased from 1.5 meals per day for children and adults at baseline to an average of 2.4 meals per day meals per day during e~aluation.~~ In addition, impact evaluations from these programs also suggest that UCT s can be appropriate responses to acute food crisis; long term cash transfers as part of a social protection strategy are feasible and help to alleviate chronic poverty and increase the resilience of households to deal with periodic shocks. Impact on Education 12. UCT programs in South Africa, Malawi and Zambia show a positive impact on education. In Zambia and South Africa the largest impacts indentified on enrollment rates concern very young children suggesting that improved nutrition and health might have increased school attendance. 13. Evidence from the Zambia35 scheme shows that between the baseline data and program evaluation school enrollment for children between the ages of 5-18 increased significantly with the largest increase between years (see table below). Case, Hosegood and Lund (2005) hypothesize that the availability of school feeding and removal of school fees were two key factors promoting increased primary school enrollment. In terms of gender disparities girls were disadvantaged compare to the boys showing with a low transfer level parents often send only one child and it is usually a boy. Samson et al. (2004) found that the household receipt of the South African grant36 is associated with a 20 to 25 percent reduction in the school non attendance gap, possibly reflecting the important role of care-giving provided to orphans by grandparents. Samson hypothesizes that grant receipts have different spending patterns, prioritizing school attendance more than non-grant recipients. It also found that recipient households in the poorest quartile (measured by expenditure) will increase the likelihood of school attendance for boys and girls. The Malawi3 program provided a protective function with respect to schooling: after one year, 29 Devereux, S., Sebates-Wheeler, R, Tefera, M., & Taye, H. (2006). Ethiopia s Productive Safety Net Programme (PSNP): Trends in PSNP transfers within targeted households. 30 Ethiopia Productive Safety Net Program (PSNP) 3 Malawi Mchinji Cash Transfer Programme. 32 Malawi Dowa Emergency Cash Transfer Programme. 33 Zambia Social Cash Transfer Programme. 34 Devereux, S., Mthinda, C., Power, F., Sakala, P., & Suka, A. (2007). An evaluation Concern Worldwide s Dow Emergency Cash Transfer Project (DECT) in Malawi. 35 MCDSSiGTZ (Ministry of Community Development and Social Services, Government of ZambidGerman Technical Cooperation) Monitoring Report 2 d Edition: Pilot Social Cash Transfer Schene, Kalomo District, Zambia. Lusaka. 36 South Africa Old Age Pension (OAP) 37 Adato and Bassett, What Is The Potential Of Cash Transfers To Strengthen Families Affected by HIV/AIDs? A Review of the Evidence on Impacts and Key Policy Debates,

81 children in beneficiary households had doubled their enrollment rates compared to comparison households3*, with drop rates half as large as children in comparison households. The beneficiary group also had fewer school absences per month. Education Impact of UCT in African countries Program School Enrollment Attendance South Africa Child Support Grant I +8.1% (age 6) +25% (CSG) +1.8% (age 7) South Africa Old Age Pension % (OW +3% (boys); +7% (girls) Zambia SCTS Malawi Mchinji Cash Transfer +5% newly enrolled month. -3% dropout rate Sources: Case, Hosegood, & Lund, 2005; MCDSS/GTZ, 2006; Samson et al, 2004; Miller et al., Impact on Health and Nutrition +10.4% (age 5-6) +3% (age 7-18) +8% (age 14-15) -2% (age 16-18) +12% enrollment rate -1.3 days absent in previous 14. Positive effects of UCT s on health were also found in Malawi, South Africa and Zambia. In Malawi, the cash transfer led to improvements in health care access for adults and children among the beneficiary households. The percent of children under 18 who were sick in the previous month before the survey was 13 percent lower among intervention households compared to comparison households (42 vs 55 percent). In addition, 8lpercent reported that their children s health had improved compared to 15 percent among comparison households. 15. The beneficiaries of the South Africa program3 also reported a significant impact on their child s anthropometric growth. Children who first received the transfer after age two or who benefitted less than 20 percent of nutrition window of opportunity experienced no statistically significant impact. However, the children who received the transfer before age 2 and continued to receive benefits for at least two thirds of their first three years of life experienced significant improvement in height attainment. A male child who received the CSG before age 1 and received benefits for the next two years experienced an increase in height for age of 0.4 or approximately 3.5 cm leading to a 2.1 percent gain in adult height. In addition, the evidence also suggests that cash transfer programs targeted to the elderly particularly if they are female have positive impacts on children s nutritional status. 38 The group is referred comparison as oppose to control group because the ntervention and comparison households were not demographically identical at baseline as children appear to have been prioritized in the intervention areas, whereas elderly households appear to have been prioritized in the comparison areas. Households were the same in terms of monthly expenditures (Miller et al., 2008). 39 South Africa Child Support Grant (CSG). 69

82 Program Malawi Mchinji Cash Transfer South Africa OAP South Africa CSG Use of transfer for health 75% HH spend some of the transfer on health 3% of pension +1.5% in all food consumption; + 1.2% in basic food items Health Outcomes Reduced illness in past month: -2 1 % (adults) -13% (children) Improvements in child health: +67% improved health -56% no change YO worse health Improved health of all household when income is pooled cm increase in height on average if received CSG the first 3 years. 16. For all of the above cases, the expenditure patterns show that in addition to spending on basic necessities (90 percent) households also spent on health and education. The proportion spent on education showed that the grant helped parents to afford education. The large expenditures on food in all countries may have an indirect effect on schooling via nutrition and health improvements. Cash grants usually lead to increased food consumption of beneficiary households and in most of the programs evaluated, grants were also associated with an increase in average meals per daythus reducing hunger. Beneficiary households generally seem to have understood the purpose of the transfers and made rational choices. Fiscal Impact of BISP 17. The Government has recently increased its attention and commitment to social protection in the wake of recent global financial crisis. The fiscal impact of the BISP program is consistent with the fiscal costs of safety net programs in other countries with comparable levels of income and with other countries in the region. If the program is fully implemented then the Government is expected to spend Rs 34 billion (US$ 425 million) for the current fiscal year , doubling the federal Government s social safety nets spending from 0.3 percent in 2003/04 to 0.6 percent of GDP.40 Government s intention is to cover five million families in FY2009/10, and 7 million in 2010/11. This implies budgetary needs of approximately one percent of GDP in 2009/10. This level of expenditure will make Pakistan s safety net spending consistent with other countries at its level of income. BISP budget is adequately funded and fits within the Government s macroeconomic framework to implement the program. Therefore, fiscal impacts of a nationwide safety net program like BISP seem manageable. 40 In addition, the province of Punjab is allocating about 0.2 percent of GDP to its income suppodcash transfer program in A higher level of spending does not translate into more coverage for the poor if safety nets programs are poorly targeted. 70

83 Annex 10: Safeguard Policy Issues PAKISTAN: Social Safety Net Technical Assistance Project 1. Despite large reductions in poverty, many households in Pakistan are poor and remain vulnerable to poverty. This report discusses vulnerability at some length to help scope out interventions that could help the country deal with the remaining burden of poverty and vulnerability. Poverty is a dynamic phenomenon, also at the level of the household. Although some households are both poor (non-poor) and likely to remain poor (non-poor) in the future because of their endowments and other characteristics, most others experience shocks that affect their income, consumption and assets and can trigger entry and exit from poverty. In addition, certain characteristics of the poor such as limited savings capacity and limited access to financial markets and safety nets make them more vulnerable to shocks and, consequently, diminish their capacity to escape poverty in a risky or uncertain en~ironment.~~ 2. This section explores these issues in detail. It examines the question of risk by analyzing data on the incidence and impact of various types of shocks faced by safety net recipients/applicants, (PSNS 11, 2005). In doing this, special attention is paid to potential differences between ultra poor (those with consumption less than the food poverty line), poor (those with consumption above the food poverty line but below the national poverty line) and non-poor households (those with consumption above the poverty line) among this group, both in terms of exposure to shocks and their impact. As noted above, while the results from this survey are not nationally representative, it provides some insight that shocks are important for the poor and that their potential impact on consumption levels and other welfare measures can be substantial. 3. The main conclusion from this exercise is that the introduction of risk significantly expands the set of households in potential need of assistance at a certain point in time. And although in many ways vulnerable households are similar to poor households, there also exist important differences with potential policy implications. Incidence of shocks 4. The PSNS I1 showed that safety net recipient/applicant households suffered one or more major shocks during the three years prior to the survey. Health and economic shocks appeared to be the most prevalent types of shocks, accounting for 54 and 28 percent of all such occurrences respectively, followed by shocks related to natural disasters (7 percent), agricultural and demographic (family-related) shocks (4 percent each), and shocks related to law and order (3 percent) The concentration of individuals around the poverty line in Pakistan means that small changes in income can result in large changes in poverty. 42 Health shocks include illness, birth and death in the family, disability, surgical procedures and accidents. Economic shocks include change in employment status (e.g., job loss) and failure of own business. Shocks related to natural disasters include weather related events.(e.g., drought, storms, flooding, landslides), pest attacks, fire, and earthquakes. Demographic shocks include marriages and other family events that may be associated with high financial costs. Finally, shocks related to law and order (or lack thereof), include theft and crimehiolence, riots, land disputes, legal costs and police trouble. In interpreting these numbers it is important to note that the data was collected prior to the earthquake of October 8, 2005 (see Box 2.1 for more information on the earthquake) and that the PSNS focuses on households that benefited from or applied to social assistance programs, most of whom are poor, Moreover, rural Pakistan is often hit by drought, which can register as both an agricultural shock (for farmers) and employment shock (for laborers). 71

84 5. Some of these shocks affected only the reporting household (Le., idiosyncratic shock), while others affected a large demographic group or even a whole community (i.e., covariate or aggregate shocks). According to the survey, the incidence of idiosyncratic shocks (health, family matters) was 58 percent, significantly higher than the incidence of aggregate shocks (economic and natural/agricuitural). Although idiosyncratic shocks appear more prevalent by these estimates, natural disasters and aggregate shocks generally clearly remain important in rural Pakistan, as evidenced by the earthquake and by the recurring droughts and floods. World Bank (2002a), for example, found that 63 percent of villages had experienced drought within a five-year period. While aggregate shocks can exhaust the coping capacity of community-based informal safety nets, individual shocks (such as non-epidemic disease) can potentially be managed through community-risk pooling as different households are affected at different points in time. Incidence and Nature of Shocks Incidence of shocks Health I Natural and 1 Family matters 1 Law agricultural and order Source: Staff estimates based on data from Phase I1 of Pakistan Safety Net Survey (PSNS 11), which is representative of safety net recipients/applicants and not nationally representative. Note: Ultra-poor households are here defined as those with consumption below the food poverty line, poor households as those with consumption below the poverty line but above the food poverty line, and non-poor households as those who consume above the poverty line. 6. The frequency and nature of shocks for this group varies with household characteristics and across areas of residence. Approximately 62 percent of the ultra-poor and 65 percent of the poor sampled in PSNS I1 report to have suffered at least one shock in the three years prior to the survey, compared to 68 percent of the non-poor. Conditional on suffering a shock, however, ultra-poor and poor respondent^^^ were more likely to suffer health shocks than non-poor respondents, while the latter were more likely to suffer economic shocks than the former. In addition urban respondents are more prone to shocks than rural ones and, conditional on the occurrence of a shock, they are more likely to suffer health and economic shocks and less likely to suffer shocks related to natural disasters. 7. These differences can be attributed to a number of factors. First, poor households exhibit lower levels of access to health care and formal labor markets than non-poor households. Second, lack of adequate infrastructure in rural areas, combined with higher dependence on the agricultural sector for survival, makes them more prone to weather-related shocks. 43 We define ultra-poor households as those with consumption below the food poverty line, and the poor households as those with consumption below the poverty line but above the food poverty line. Data is from Phase I1 of Pakistan Safety Net Survey which focused on households that benefited from or applied to safety net programs; therefore not representative of the entire population of Pakistan. 72

85 Box.1: The Earthquake of October 8,2005 The earthquake caused great devastation. The worst affected parts of four districts including Mansehra, Basham, Muzzafarabad and Bagh were completely wiped out. According to the joint ADB-World Bank Needs Assessment, approximately 73,000 people died and more than 70,000 have been severely injured or disabled. Many of the injured had amputated limbs or severe spinal cord injuries that left them paralyzed or disabled. A very large and disproportional share of deaths and injuries were sustained by schoolchildren and women. An estimated 15,000 people sustained spinal cord injuries or amputation, and many more suffered psychological trauma. Over 2.8 million persons lost their shelter. The earthquake reminds us that the poor in rural mountain areas in particular face a considerable risk of loss of life or assets to natural disasters, small or large. There are few if any public programs offering prevention, mitigation, or coping with natural disasters - after the earthquake, an assistance package and delivery mechanism had to be designed largely from scratch. Impact of shocks 8. Shocks experienced by this subset of households are generally associated with income and asset losses and, potentially, a decline in household consumption levels. These losses arise from the direct impact of the shock (e.g., foregone wages associated with the loss of employment) and the cost of coping with the shock after it happens (e.g., medical costs associated with an illness). The average total cost of a shock as reported by households in the PSNS I1 survey was Rs. 14,000 (or $233).44 This amount is equivalent to more than one third of average annual household expenditures, or five months of income from daily labor, of this group of househ~lds.~~ 9. Total costs vary with the nature of the shock. These range from Rs. 42,000 for demographic or family-related shocks to Rs. 7,700 for economic shocks46 (corresponding to approximately 100 and 19 percent of annual household expenditures, respectively, for the low-income PSNS sample). The actual burden associated with each type of shocks also depends on how frequent they are. A better measure of the actual burden can then be calculated by multiplying the average cost times the frequency of each type of shock. Based on this measure health shocks account for most of the total shock costs to the PSNS respondents (5 1 percent), followed by economic shocks (20 percent), natural disasters (14 percent), family events (9 percent), agricultural shocks (4 percent), and law and order (2 percent). Within economic shocks, seasonal unemployment is widespread and very costly, with male workers on average idle for about two months in a year. This corresponds to an income loss of 17 percent, or nearly Rs. 6,000 per worker per year among the low income PSNS respondents. More stable employment would significantly reduce poverty. 10. Information on average costs or average total burden masks important differences between poor and non-poor and between urban and rural safety net recipients/applicants in the impact of shocks. Specifically lower income and consumption levels among the poor and in rural areas generally translated into higher relative costs of shocks for these groups, even in the absence of significant differences in the 44 Based on Phase I1 of Pakistan Safety Net Survey and therefore not nationally representative. 45 This amount is much higher than that reported in the PIHS or PSLS and can potentially be explained by several factors (i) PSNS I1 data pertain to the worst shock experienced by households over three years; (ii) self-reported total losses and coping costs of shocks include loss of employment and foregone income as well as out-of-pocket costs (the PIHSPSLS only report out-of-pocket health expenditures. Finally, (iii) it is also important to note that the sample of both surveys is different; the PIHS is nationally representative, while the PSNS I1 is representative of safety net recipients/applicants the majority of whom are poor. 46 Average costs of other shocks are: natural disasters, Rs. 20,700; law and order, Rs. 17,000; health, Rs. 13,000; agricultural, Rs. 12,300. As mentioned in the main text, these self-reported coping costs are defined differently and differ in magnitude from the out-of-pocket expenditures reported in the expenditure modules of PIHS or PSLS which, for example in the case of health expenditures, are no larger than 5 percent of total expenditures. 73

86 absolute costs of shocks between them and others. Shock-related expenses represent 54 percent of annual consumption for ultra-poor households in the PSNS (defined as those with consumption below the food poverty line), compared to 27 percent for poor households (defined as those with consumption below the poverty line but above the food poverty line) and 18 percent for non-poor households. Similarly, conditional on experiencing a shock, households in rural areas incur costs/losses equivalent to 37 percent of their annual consumption, compared to 30 percent for urban households (see table below). I Coping with shocks can be a high share of household expenditures Direct and indirect cost of coping with shock (YO of annual household expenditure)* Percent of households whose coping costs exceeded annual household expenditure Income level Ultra-poor Poor Non-Poor 1 Location Rural Total Sample size I 1,501 Source: Staff estimates based on data from Phase I1 of Pakistan Safety Net Survey (PSNS II), which is representative of safety net recipients/applicants and not nationally representative. Note: Calculations include households that reported zero coping costs. * Data pertain to the worst shock experienced by households over three years and self-reported total losses and coping costs of shocks include loss of employment and forgone income as well as out-of-pocket costs Finally shocks can also cause non-monetary losses. The PSNS I1 contains information on events triggered by shocks, such as increases in child labor or reduction in household food intakes. These changes are not conceived as coping strategies per se but rather result from other existing constraints (e.g., unavailability of spare adult labor or low pre-shock consumption levels). Eight percent of all shocks among the low-income households sampled for the PSNS resulted in a child being taken out of school, and 10 percent led to a child being put to work. These percentages were higher for more costly shocks (see table below). Similarly, one third of all PSNS respondents subject to a shock saw their food intake decrease as a consequence. This effect was more severe among ultra-poor households (36 percent) and households in urban areas (35 percent) due to higher levels of food insecurity associated with lack of access to land. For regional and international evidence see Box 2. 74

87 Children withdrawn from school or put to work in response to shocks Reactions to economic shocks (households with children that experienced an economic shock) Children I Children put 1 Average cost Source: Staff estimates based on data from Phase I1 of Pakistan Safety Net Survey (PSNS 11) representative of safety net recipients/applicants and not nationally representative 12. While this survey is not nationally representative, these figures suggest that the potential impact of shocks can be devastating both in the short- and the medium-term. This is compounded by the fact that often coping mechanisms used by these households, although effective in the short term, severely jeopardize future livelihoods, as discussed below. 13. The discussion on shocks and their short- and long-term consequences makes it clear that a static notion of welfare such as that based on the traditional concept of monetary poverty is inadequate when it comes to informing social protection policy. Box 2: Regional and international evidence on the incidence and impact of shocks Growing evidence suggests that most' households across South Asia and other developing regions face similar risks. Surveys have shown that health, employment, and natural disaster risks are the top three risk factors in other South Asian countries. The relative importance varies-in Sri Lanka, natural disaster tops the list, while disease is the leading risk in Maldives and, by some evidence, in Andhra Pradesh, India, where drought is another major concern. In Afghanistan, employment risk is the single most prevalent risk, closely followed by disease and death. Indeed, there is mounting evidence from many developing countries that health shocks are among the most sizeable and least predictable shocks, imposing very large coping costs in terms of medical expenses and loss of earnings. While for other shocks, increasing the labor supply is a common response, this is often not possible when breadwinners face ill health, and medical costs are therefore compounded by income loss and an absence of coping options (see Kochar, 1995 for India). A detailed study in China also fmds a substantial and significant reduction in income and labor supply due to health shocks. The loss in income is a consequence of a reduction in labor supply for the head of household, which is not compensated by increase in labor supply of other members. Health insurance in China appears to offer very limited protection-surprisingly out-of-pocket health care expenditures increase more for the insured than for the uninsured in response to health shocks (Wagstaff and Lindelow, 2005). In Indonesia, research by Gertler and Gruber (1997) using panel data found significant economic costs associated with illness, albeit more from income loss than from medical expenditures. The study found that households were unable to self-insure against illness. In fact, for illnesses that severely limit physical function, families were able to smooth less than 30 percent of the income loss from these illnesses. However, there is evidence that microfinance savings and lending institutions in Indonesia help families to self-insure consumption against health shocks (Gertler, Levine, and Moretti, 2003). 75

88 Vulnerability 14. We define vulnerability as expected poverty, or the likelihood that households and individuals will remain or become poor in the (near) future. Our method assumes that the variation in households consumption over time is similar to the variation in household consumption across regions of Pakistan. The results are therefore only indicative, but can potentially help develop a better sense of who the vulnerable are (see Box 3). Although most ultra-poor and poor households will be considered vulnerable under this definition, the concept of vulnerability is broader than that of poverty and will therefore include some currently non-poor households with a high probability of falling into poverty. Similarly, it will exclude some currently poor households with a high probability of exiting poverty in the future given their endowments and other characteristics. In the absence of recent panel data representative of the whole of Pakistan, we employ a method for econometric estimation of vulnerability based on a single cross-section survey, following the methodology pioneered by Chaudhuri and others. The vulnerability estimation was carried out with the household data of the 2001/02 PIHS. The method consists in first estimating a consumption function based on cross-sectional survey data, and using the estimated model to generate household-level predictions of the probability of becoming poor in the future, defined in this report as a two to three year horizon. Households are classified as vulnerable if their probability of being poor is more than 50 percent. The crucial assumption behind this method is that the household inter-temporal variance in consumption can be proxied by the cross-sectional variance. This is a strong assumption-it amounts to saying that the variation in households consumption over time is similar to the variation in household consumption across regions of Pakistan. The results should therefore be taken with a grain of salt: the results are potentially useful for establishing a profile of the vulnerable, but not as a point estimate of the level of vulnerability. Moreover, the method does not capture the potential impact of large aggregate shocks. We use the cross-section estimates at regional level to estimate the structure of variability, which we then use to derive an estimate of the vulnerability at bter-temporal level. After estimating the level of vulnerability for each household, we partition the sample kccording to predicted vulnerability and current poverty status. Sources: Del Ninno, Vecchi and Hussain (2006), based on Christiansen and Boisvert (2000), Christiansen and Subbarao (2005), Chaudhuri (2000, 2001 and 2003), and Chaudhuri, Jalan and Suryahadi (2003). Further details of the estimation methodology are discussed in Appendix By broadening our definition of welfare to account for the risk of future deprivation, we also gain some insight into the role and relative importance of structural and transitory poverty. Specifically, we distinguish between the following three types of households according to their vulnerability status: Chronically poor or vulnerable to chronic poverty. Vulnerability among these households is the result of low levels of consumption. Their poverty is structural in that their basic characteristics make them likely to remain poor. They correspond broadly to the group of ultra-poor defined above. Transitorily poor or at risk of transitory poverty. Vulnerability among these households stems from income/expenditure volatility (most likely associated with positivehegative shocks). They are likely to transition in and out of poverty over time as a result of shocks and other factors. They overlap broadly with the group of poor (but not ultra-poor) defined above, but include also some who are not currently poor. Non-poor or infrequently poor. Levels of vulnerability to poverty are extremely low among these households since they enjoy both high levels of current consumption or low levels of future consumption volatility. They are the most secure households and, therefore, unlikely to fall into poverty. 16. Using a nationally representative survey (PIHS 2001/02), we find that approximately 56 percent of households can be classified as vulnerable when a 2-year horizon is considered, according to the above 76

89 definiti~n.~~ That is, in 2001/02 around 56 percent of households were either poor or expected to be poor at some point in the next two years. And vulnerability levels increase to 67 percent when a 3-year horizon is considered instead (although, as mentioned, not too much should be read into these point estimates due to the strong assumptions of the estimation methodology). The observed growth in the incidence of vulnerability as we move further into the future is a direct reflection of increased consumption risk-i.e., the probability of suffering (at least) one negative shock rises over time (see table below). 17. Approximately half of Pakistan s burden of vulnerability stems from vulnerability to chronic poverty, and the other half stems from vulnerability to transient poverty, according to these estimates. One third of the population is estimated to be chronically poor or vulnerable to chronic poverty. Because this group is poor and likely to remain poor over time despite other developments, their relative numbers do not change as the time horizon expands from 2 to 3 years. An additional 24 percent of households are transitorily poor or vulnerable to transitory poverty. Because these households are exposed to high volatility in consumption, their relative numbers increase to 35 percent of the total as the time horizon expands from 2 to 3 years. Category Total Total chronically poor or vulnerable to chronic poverty [Low consumption] Vulnerable to chronic poverty Chronic poor Total transitory poor or vulnerable to transitory poverty [High volatility] Vulnerable to frequent poverty Frequent poor Total non-vulnerable 1 [High consumption, low volatility] Time horizon Iyear! 2-year I 13.2 I 10.2 I I Note: The estimated level of vulnerability is larger than the poverty rate since the vulnerable include households that are currently poor and expected to remain poor in the future as well as those who are not currently poor but face a larger than 50 percent risk of becoming poor in the next 2-3 years. 18. Following from the above, an effective social protection system would then have to achieve the following: (i) provide (minimum) income and other support for those with a small probability of exiting poverty (Le., the chronic poor); (ii) support the implementation of mechanisms and strategies to exit poverty among those with a higher likelihood of succeeding, including increased access to economic opportunities and the promotion of human capital accumulation-particularly of children; and (iii) support (directly or indirectly) the implementation of effective risk mitigation strategies to help households smooth income over time. In addition, high exposure to risk and high vulnerability incidence for particular segments of the population (e.g., informal sector workers) requires that, to the extent administratively and fiscally feasible, the design and nature of social protection programs must be such as to allow the system to quickly react to individual and aggregate shocks. This implies that programs must 47 Analysis is based on the PIHS Since the PIHS was collected, poverty and vulnerability have reduced, but remain at high levels, and the characteristics and composition of the poor and vulnerable are unlikely to have shifted in very major ways. 77

90 be capable of deploying assistance at short notice and that they must be able to expand and contract their beneficiary base over time. 19. The proposed operation is designed to support Pakistan in setting up an essential part of such system: A basic income support program for the poorest families. The project s development objective is to establish a nationwide effective and transparent cash transfer program for the poor. This is motivated by the recommendations of the Government s Social Protection Strategy (2007), a national framework developed through by policy dialogue with a series of stakeholders and the World Bank (2007) and further informed by the findings of a national participatory poverty assessment supported by DFID (2003). To this end, the primary beneficiaries of the program are expected to include women and children, with transfers being provided to the female head of eligible families. 20. The BISP includes a number of features that have been designed to cushion shocks and promote human development outcomes amongst most vulnerable groups. First, the establishment of a national targeting system based on objective and transparent criteria, supported by strong and transparent institutional structures for collecting enrollment data, determining eligibility, benefit payment, and for addressing grievances, with a strong emphasis on monitoring and evaluation, has the objective of ensuring that benefits reach the poor, with minimum inclusion and exclusion errors. The proposed operation would also allow improve the pro-poor targeting of public expenditures on other poverty alleviation programs. Second, information campaigns will be initiated to ensure credibility and accountability amongst all stakeholders. Communication with participating households is central to this and will be achieved via simple and transparent case management. Third, the operation will link income support with productive opportunities, such as skill training activities, micro-finance and other programs, in addition to supporting the development of conditional cash transfer schemes that provide incentives to poor households to send their children to school (and in the future nutritional and health services). Evidence from Latin America indicates that well designed and administered conditional cash transfer programs both reduce poverty but also have positive impact on school enrollment, a critical ingredient for improving welfare of the future generation 21. During project implementation, a series of beneficiary assessments will be carried out to monitor the main social impacts and results of the project. The following themes are particularly important for monitoring, process and impact evaluation: The impact of income support on human development outcomes, e.g., food consumptions, education, nutrition, child labor supply and women s social status. Impacts on women and children are of particular interest. Concerning child labor, evidence suggests that many poor families do not send their children to school owing to resource constraints and lost income opportunities. Concerning women s status, more supporting evidence is required that female headed households are more likely to increase resources for girls nutrition and education in Pakistan. The targeting of vulnerable population groups in rural and urban areas. Vulnerability is high among socially excluded groups, defined along lines such as religion, language, gender, access to land, occupation, physical disability and/or social status. Allied to this is the differences in poverty between rural and urban households are of interest. Approximately sixty-three per cent of rural households are considered vulnerable. The ability of the income support program to protect against chronic and transient poverty. In recent years Pakistan has experienced a series of adverse aggregate income shocks, such as a drought and earthquake and recently, the global economic crisis. The unfolding global financial crisis will reduce aggregate demand and will inevitably affect - albeit with a lag - employment and wages. The BISP, supported by this operation, has the potential to mitigate some of the negative economic impacts of the financial crisis by providing well-targeted basic income support 78

91 to the poorest households. A Poverty and Social Impact Analysis for Pakistan indicated that while the recent food and fuel crisis households could have increased poverty between 2 and 3 percent, depending on whether growth did not occur or the economy grew at 6 percent, a safety net system could be much more effective than subsidies for protecting the poor, assuming it was well targeted. 22. The benefits of the Benazir Income Support Programme go exclusively to women. Since the program is implemented nationwide, it will reach out to benefit the poor segments of the various ethnic minority populations of the country. Ethnic and cultural considerations will be mainstreamed into the program design to ensure equal and culturally appropriate accessibility for ethnic minority populations Under Component One, ethnic sensitivities and cultural considerations will be taken into account in improving the targeting design, including updating of the targeting tools (e.g. manuals, materials, database), training of administering staff and consultants. Under Component Two, the public communication strategy, grievance redress and social accountability mechanism will be improved with specific attention to build in special mechanisms and provisions for operating with ethnic minority population groups and in tribal areas. Under Component Three on the BISP Program, the BISP M&E system design will include ethnic and cultural considerations in its various components, including baseline and follow-up survey design, and beneficiary assessment. Component Four will strengthen social protection policy and institutional framework. It will build in ethnic and cultural considerations in its activities, including in the design of a NSPS monitoring system and various capacity building efforts for policy makers and implementers. 23 In order to analyze its effects along various social dimensions it has been agreed to include in the project a number of studies to effectively monitor the social effects of BISP, particularly on gender and ethnic minority groups. This work will be led by a social scientist and guided by the respective Bank colleagues. 79

92 ~ A. The BISP MIS Annex 11: BISP Management Information System PAKISTAN: Social Safety Net Technical Assistance Project 1. The BISP Management Information System (MIS) is designed as a comprehensive information management system designed to support the BISP in all its program functions at the national level. The diagram below presents a functional view depicting the system's main processes and data flows that will support the BISP operations. Box 1: BISP M IS Functional View 1 BENAZIR INCOME SUPPORT PROGRAM - MANAGEMENT INFORMATION - ~ SYSTEM I I - ; ELIGIBILITY +,,, COMPLAINTS i I 1 j PAYMENT '-+ BISPMIS COMPLAINTS I f f \ 80

93 2. A complementary physical distribution of this functionality is shown below in the system architecture view, which shows access to system functions from the central as well as regional operations segment. Box 2: BISP MIS Architecture View Province A - I - District! District 2 P- NAD+ titsr bervers Province B Operations Database This architecture has the following features: MIS and Database Servers - Hosted at NADRA headquarters in Islamabad during the initial year of operation. Hosting will be transferred to BISP headquarters ofice on the second year. Form-1 Data scanning and data entry - Centralized operation at NADRA in Islamabad. This operation will continue to be performed at NADRA to utilize the installed data scanning infrastructure. BISP Management and Operations Departments - Online secure Web access to BISP MIS. Provincial and Regional BISP Offices - Online secure Web access to BISP MIS B. MIS Processes and User Groups 2. With respect to the services provided by the MIS there are four main processes and user groups - subject to change over time - namely: (i) Targeting and Eligibility - Modules 1.O through 6.0 support the data entry process, validation, verification and the application of the Proxi Means Test (PMT) to determine BISP eligibility. This is a process performed at present at the central (headquarters) NADRA office, managed 81

94 by the MIS Data Entry Team. Inherent to this activity is the handling of a very large volume (millions) of applicant forms, a continuous attention to quality, organization and performance to ensure an efficient operation. (ii) Grievance Redressal - Module 7.0 provides a system function to enable BISP Officers, located in provincial and/or regional/district offices, to obtain online applicant or beneficiary information and have the necessary elements to evaluate and resolve complaints. Complaints may be of different types, including those related to an applicant not being eligible (i.e., missing information, invalid CNIC, invalid name, etc.) or related to payments (Le., not receiving it, wrong address, etc.). The MIS will provide a function for the Officer to interact with the database and update it as needed in solving the applicant s or beneficiary s issues. (iii) Payment Process - Modules 8.0 and 9.0 provide system functions to manage the payment process. The MIS may be prompted to generate a bi-monthly list of eligible beneficiaries who will receive a BISP payment. This list will be produced by BISP Central Operations Department, and once it is verified to be complete and correct, will be delivered to the Pakistan Post BISP Program Manager for payment production. After payments have been distributed to recipients and the respective reports and payment reconciliation process is completed, these will be provided to BISP Operations for updating and final reconciliation on the BISP database. (iv) Monitoring and Reuorting - Module 10.0 consists of a business intelligence capability to provide BISP Management the facility to generate activity information on any program activity, from targeting to payment and monitoring. Reports can be produced by geographical levels or by any attributes with respect to demographics or other data available data descriptors. Additionally, statistics and even more sophisticated tools may be used, such as data mining and graphic tools to analyze program operations and performance. C. MIS Modules 4. The BISP M IS is integrated by the following software modules. 1.O Form-] Scanning - The system uses high speed scanners to scan Form-1 (cover page and scorecard), which are provided to NADRA by the Participating Organizations (POs). Forms and pre-numbered and this number may be referenced by the applicant through the stub provided to himher, which will have the same number that is pre-printed on the scanned form. Scanned forms are stored in a database for subsequent processing. 2.0 Double Data Entry - The MIS performs a routing of the stored scanned forms to two distinct data entry operators for double (concurrent) data entry. When the system detects differences on the entered data, a third person (supervisor) will review and make a decision on the final value to be recorded in each case. 3.0 Form-1 Data Validation - MIS applies data validation rules based on the technical specifications provided for the BISP MIS design. 4.0 Verifv Household and Family Members -Verify the data provided for each member of the family unit using NADRA s biological database and their CNICs, to ensure that data on the scorecard is consistent. 82

95 5.0 Verify Other Social Protection Programs - A verification is performed to determine if the family is enrolled on the Pakistan Bait ul Mal program, in which case the family will be enticed to enroll in BISP and exit PBM. 6.0 Apply PMT to Determine Eligibilitv - The Proxy Means Test (PMT) is applied to BISP applicant records that are already validated and complete, including valid CNICs, and also, that are not currently beneficiaries of other social protection programs. Those applicants who obtain a total score below or equal to the program cutoff mark, will become beneficiaries in BISP. 7.0 Grievance Redressal - This function allows BISP Officers to retrieve an applicant or beneficiary record based on the application form number, or the CNIC of the head of the household or of any member of the family. Based on the application status (incomplete, ineligible, incorrect CNIC, etc.) the Officer will request additional information and, if provided, will update the database. The BISP Officer may also enquire on the status of a payment (no answer, incorrect address, person not available, etc.) and will make the necessary annotations to correct the problem. Any unresolved issues may be directed to the central BISP Operations Office for further evaluation and decision on the case. 8.0 Generate List of Payments - BISP Operations personnel authorized to perform this function will generate a list of payments for the current period. This list will be reviewed for completeness and comparison with the current budget numbers and approved for distribution. The list, along with an electronic file of payment recipients will be provided to the Pakistan Post Program Manager for generating the payment money orders and to arrange for payment distribution by geographical location. Payment distribution will be duly recorded by each Postman and these lists will be summarized and a master payment report will be produced and delivered to BISP Management. 9.0 Update BISP Database - BISP Management receives a payment report and reconciliation report from Pakistan Post, in hardcopy as well as an electronic payment file with detailed information on the disposition of each household payment. The MIS will process this file and update each beneficiary s record as to the distributed payment, receiver, date and the postman s identification. A separate update file of payments will be kept by the MIS, as an alternate control mechanism of payments Monitoring and Reporting - the MIS will provide a reporting facility to meet the needs of BISP management as well as the Operations Units. Information should be available for all program activities and phases, from Targeting to Payments and case management. Information will be used to manage program operations, to handle specific grievances and to help in decision making by management Spot Checks Process - The MIS will provide a function to manage spot checks processing which performs all targeting functionality as described above (A.1 - A.5) as a separate process for a specified set of BISP applicants usually pertaining to a designated village or all villages within a Thesil. The purpose is to be able to compare the data quality and eligibility outcomes for the same area, as data is collected by a special team of enumerators to accomplish a spot check process. A determination for repeating the data collection process will be made if there is a significant difference on the results of the two data collection exercises. 83

96 Annex 12: Development of the Poverty Scorecard for Pakistan PAKISTAN: Social Safety Net Technical Assistance Project 1. The poverty scorecard that the Government of Pakistan has now officially adopted was developed based on the proxy means testing method. This development required the following steps: (i) selection of a proper dataset, (ii) selection of an indicator for household welfare, (iii) selection of a group of poverty correlates, (iv) selection of a regression model based on some selection criteria, and (v) conversion of regression coefficients to a simple scoring system48. (i) Data Selected for the Analysis 2. The data used for this exercise is the Pakistan Living Standard Measurement Survey (PSLM), conducted by the Federal Bureau of Statistics in This is a multi-topic household survey, with modules on consumption, income, employment, health, education, and living conditions. The PSLM covered around 15,000 households drawn from four major provinces of Pakistan and was designed to be representative at the national and provincial levels. For several reasons, the PSLM appears to be well suited to the purpose of developing a scorecard following the Proxy Means Testing approach. First, it has rich and detailed information on household and individual characteristics that are correlates of poverty. Second, the PSLM data are representative at the national level, which enables us to estimate key program statistics like program coverage, program leakage, and budgetary needs. Third, this dataset is used for estimating the official poverty headcount rates of making it easy to check what percentage of the poor defined in the official estimation will be identified by the proposed scorecard. (ii) Selecting an Indicator for Household Welfare 3. Household consumption expenditure (monthly) per adult equivalent were chosen as the welfare measure. The same measure is used for producing the official poverty estimates in Pakistan; therefore, it is easy to identify the poor. Household expenditure includes all expenditures on non-durables, the imputed value of non-durables received as gifts or produced in the household, while it excludes expenditures on durable goods and assets. 4. In the economic development literature, consumption expenditure is generally considered a more accurate measure of welfare than income for several reasons (see Deaton 1997). First, because consumption expenditures tend to be less variable than income over seasons, it is more likely to indicate the household s true economic status (as a result of households with sporadic incomes smoothing their consumption patterns over time). Second, in practice, consumption is generally measured with far greater accuracy than income in a household survey. For example, households sources of income may include home-based production, own farms and businesses. Calculating the flow of net incomes from these sources turn out to be a big problem since the flow of costs and returns from these activities are often inaccurately reported by households. (iii) Choice of Poverty Predictors 5. Selection of variables to predict welfare as measured by per capita consumption should take into account two separate criteria: correlation between the welfare measure (consumption) and the predictor, which will determine accuracy of the prediction, and verifiability of the predictor, which will determine 48 This section is taken from the draft paper Poverty Scorecard for Pakistan: A Recommended Approach for Targeting the Poor, by Vishwanath, Hou and Yoshida (2009). 84

97 the accuracy of information used to impute welfare. The types of predictors used for this exercise, discussed below, were arrived at after judging all possible predictors on the basis of these two riter ria.^' 6. Household and individual characteristics such as the number of members and dependents, and the age, education and occupation of the household head, as well as the education of children not only strongly correlate with consumption levels they are relatively easy to verify and are less likely to be mi sreported. 7. Ownership of durable goods and housing characteristics: These variables are verifiable by inspection. However they can be misrepresented if a household removes the goods from its home during an expected visit by the social worker, which is easier to do with small or mobile items than with larger items such as stoves or refrigerators. The general presumption in the literature is that people are more willing to lie about ownership of such items as opposed to household characteristics. Nevertheless, these variables tend to have high predictive power for welfare, and therefore including them should reduce mis-targeting substantially. 8. Ownership of productive assets, especially land holding, livestock and farm equipment, is critical to proxy the welfare status of rural households. Verifiability of these variables is reasonably high. For example, even for a variable like land ownership, which may not be measured perfectly, one can reasonably expect that program officers belonging to the community would be able to deter misreporting using local knowledge. (iv) Selection of a Regression Model Based on some Criteria 9. While each variable listed above is already a good poverty predictor, a combination of them can be an even more powerful predictor of poverty. However, combining them is not a simple task and combining variables in an arbitrary way can produce a worse predictor of poverty. For example, some welfare programs simply count the number of poverty predictors satisfied by a household and use that as a score to identify whether the household is poor or not. There is no guarantee that the resulting score is a better predictor of poverty than a single variable is. 10. In the Proxy Means Testing approach, the OLS regression method is used to find a good combination of variables in predicting poverty. The OLS method regresses the consumption variable (household expenditure per adult equivalent in the case of Pakistan) on various sets of variables listed above. Predicted values from a regression model are by construction better proxies of poverty than any single variable included in the regression model. The main task here is to find a regression model from many possible combinations of variables based on the following selection criteria: 11. A good model should be able to identify target groups and exclude unintended groups as accurately as possible. However, any targeting mechanism is not perfect and entails certain targeting errors. Following the literature, this paper assesses the targeting performance by examining undercoverage, leakage, and program incidence. 85

98 12. Targeting performance is measured by the extent to which there is an overlap between the true target population (intended beneficiaries) and program beneficiaries as selected by the targeting mechanism. Since the target population of the BISP was not known when this analysis was conducted, targeting performance was examined for several different target groups, i.e., the poorest 10, 20, 25, and 30 percent of population. Assuming that misreporting in household expenditure data of PSLM is limited, the target population is identified in the sample of PSLM data. 13. Two issues are central in determining targeting performance at any given cut-off (a) The extent to which intended beneficiaries are excluded. This gives a measure of program undercoverage ; (b) The extent to which the targeting mechanism mistakenly includes non-target groups. The most important issue here is that a fraction of the program budget is lost due to leakage to an unintended population. The goal of a good targeting mechanism is to minimize both types of errors, but in practice, there is often a trade-off such that improving one worsens the other. 14. More formally, the concepts of both undercoverage and leakage can be defined using Table 1. Table 1 s columns represent whether a person actually belongs to the target group or not. This is determined by the person s actual household expenditure per adult equivalent. Table 1 s rows represent whether a person is determined to be eligible by the scorecard that is developed using the PMT. This is determined by whether a person s estimated score is below a cutoff. N1 to N4 correspond to population of each cell. The undercoverage rate is represented by N2/(Nl+N2) while the leakage rate is by N3/(Nl+N3). 15. Program incidence is another key statistic of targeting performance. This is measured by computing program coverage for each decile of household expenditure. Undercoverage treats all the poor equally although there is a substantial variation in household welfare even among the poor. In reality, exclusion of the extremely poor, say the poorest 10 percent of the population, is likely to have more serious welfare consequences than exclusion of poor households near the poverty line. Conversely, leakage treats all the non-poor equally, but leakage to the non-poor who are very close to the poverty line is not as detrimental as leakage to the richest 10 percent of population. Program incidence enables us to see a finer picture of targeting performance across poverty groups. 16. Regression statistics like R-squared and t-statistics are also important to find a good regression model. R-squared is a measure of model prediction by comparing the actual expenditure and the expenditure predicted by a model. It is preferable that the R-squared of a regression model is at least 40 percent. Also, t-statistics for each variable in a regression model is important. This statistic is a measurement of whether the coefficient of a particular variable is statistically relevant for predicting household expenditure. If a variable takes a too low a value of the t-statistic, the variable is typically dropped. These two sets of statistics usually help in identifying good models along with the targeting errors discussed above. (v) Conversion of the Predicted Values to Scores 17. Once the preferred regression model is determined, the predicted value is a good proxy for household expenditure per adult equivalent. However, the predicted value is not necessarily intuitive and easily processed for implementation. To simplify the calculation of scores, the predicted values were normalized so that all scores are integers ranging between 0 and 100. To minimize the risk of reversing the ranking, this normalization was made by first revising regression coefficients of all variables proportionally and then rounding them. 86

99 (vi) Reaching a Final Poverty Means Testing Formula (PMTF) and Proposing a Poverty Scorecard 18. After examining 99 different models, a final PMTF model was selected. The regression results of the final model include a total of 23 variables. Consistent with the standard literature on poverty diagnostics, the number of dependents, the educational attainment of household head, children s enrollment status, and housing conditions and asset ownership remain in the final model. 19. The coefficients of all 23 variables are statistically significant at the 5 percent level and the R- squared of the model is more than 55 percent. This indicates that the predicted values of the regression model fit the actual values of household expenditure well and all the selected variables are relevant for identifying the poor. Table 2: Targeting performance: Coverage, 20. The targeting performance of this model in terms of undercoverage and leakage improves as the target group shifts from the poorest 10 percent to 30 percent. If the poorest 25 percent of population is set as the target group, the under-coverage rate is 52 percent and the leakage rate is 37 percent (see Table 2). This means 52 percent of the poor (the poorest Target group (poorest 10% 20% 25% 30% Coverage Undercoverage iource: The Wodd Bank staff estimation 25 percent of population) will be excluded while 37 percent of beneficiaries are non-poor (or do not belong to the poorest 25 percent of population). However, as noted below, most of the reported leakage is to the population that is vulnerable to poverty. 21. Note that the targeting performance of the final model is comparable with others in the region. For example, in Sri Lanka s PMTF, if the poorest 25 percent of the population is selected as the target group, the corresponding undercoverage rate is 52 percent and the leakage rate is 39 percent. 22. Figure 1 provides a more disaggregated and encouraging picture of program incidence. Note that at a coverage rate of 25%, the excluded among the poor are far more likely to belong to the upper deciles of expenditure while over 60 percent of the poorest 10 percent of population are selected. It also shows that most leakage (close to 85 percent) goes to a population that is vulnerable to falling into poverty with a small income shock (the bulk of unintended beneficiaries come from the third, fourth, fifth and sixth deciles, whose monthly household expenditure per adult equivalent is less than Rs. 1436, around 150% of the official poverty line). In other words, the model does a fairly good job of excluding the non-poor from program beneffis. I 62.3 Figure 1: Program Coverage (YO) by expenditure deciles 70, % n I Source: The World Bank staff estimation using PSLM data 87

100 (vii) Comparison to other models 23. Compared to other models, this final model has several advantages particularly in relation to preparation and implementation of a scorecard. An example would illustrate this. Table 3 shows the results of an alternative model, which is one of 99 regression models examined in this note. This model is one of the best models in terms of R-squared and targeting performance, which are better than in the final model (see Table 3). However, that regression model requires 41 variables to produce predicted values of household expenditure while the final model needs only 23 variables. It is not impossible to prepare a scorecard with 41 variables, but a model with only 23 variables is certainly an advantage at the data collection and processing stages. Furthermore, note that, despite the limited number of variables, the targeting performance of the final model is almost as good as that of the alternative model. Unlike the alternative model and others, the final model also excludes household size, since this variable, though it is highly predictive of consumption levels, gives a large weight to small households which can lead to the systematic exclusion of small but vulnerable households. During the pilot phase of Sri Lanka s Samurdhi program reform a similar issue arose. While household size was a good poverty predictor in general, its inclusion led to a very high rate of exclusion of small but poor households with a few elderly members. Since such households were known to be highly vulnerable in Sri Lanka, the Government of Sri Lanka prepared a special selection formula for small households to ensure the inclusion of such small but highly vulnerable households. In the case of Pakistan also, this issue was noted, and to Source: The World Bank staff estimation using PSLM data avoid errors, the household size variable was,dropped from the regression model, but the effect of household size on consumption is retained indirectly through the inclusion of dependents and the number of rooms per person-both of which capture essential effects of large households on poverty. (viii) Multiple scorecards vs. one scorecard 24. Although only one scorecard is proposed for the BISP here, in theory multiple scorecards can be prepared. For example, each province can have a different scorecard, as can urban and rural areas. In Pakistan, it is well known that consumption patterns differ largely across provinces and between urban and rural areas. Preparation of separate scorecards for each province or urbadrural areas might be a good idea to reflect the differences in consumption patterns since it tends to improve the predictive power of the regression model. 25. On the other hand, the use of different scorecards can increase administratiodimplementation costs. Training for enumerators and for data collection/processing systems is also far more complex. Also, the use of different scorecards can lead to concerns about equity across locations. For example, it is possible that a household in Punjab that is rejected based on a Punjab-based scorecard could be selected if a scorecard of Sindh (with the cutoff) is applied and over time this may also cause beneficiaries to misreport residence. The urban-rural breakdown also faces similar difficulties, perhaps even more so since the definition of urban areas can change from survey to survey. It is possible that some urban households who are rejected based on an urban scorecard would be selected if the rural scorecard is 88